CN116108760B - Main heating characteristic monitoring method and system - Google Patents
Main heating characteristic monitoring method and system Download PDFInfo
- Publication number
- CN116108760B CN116108760B CN202310383188.1A CN202310383188A CN116108760B CN 116108760 B CN116108760 B CN 116108760B CN 202310383188 A CN202310383188 A CN 202310383188A CN 116108760 B CN116108760 B CN 116108760B
- Authority
- CN
- China
- Prior art keywords
- main transformer
- oil temperature
- sequence
- value
- error
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 93
- 238000010438 heat treatment Methods 0.000 title claims abstract description 74
- 238000012544 monitoring process Methods 0.000 title claims abstract description 63
- 238000004364 calculation method Methods 0.000 claims abstract description 333
- 230000002159 abnormal effect Effects 0.000 claims abstract description 79
- 238000001514 detection method Methods 0.000 claims abstract description 56
- 230000008859 change Effects 0.000 claims abstract description 55
- 238000013528 artificial neural network Methods 0.000 claims abstract description 26
- 230000005856 abnormality Effects 0.000 claims description 63
- 239000013598 vector Substances 0.000 claims description 51
- 238000012360 testing method Methods 0.000 claims description 38
- 238000012549 training Methods 0.000 claims description 30
- 230000007935 neutral effect Effects 0.000 claims description 26
- 239000011159 matrix material Substances 0.000 claims description 16
- 238000010276 construction Methods 0.000 claims description 12
- 238000012216 screening Methods 0.000 claims description 6
- 238000010792 warming Methods 0.000 claims 1
- 238000013178 mathematical model Methods 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 36
- 230000008569 process Effects 0.000 description 19
- 238000012423 maintenance Methods 0.000 description 11
- 230000007613 environmental effect Effects 0.000 description 4
- 238000005065 mining Methods 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000009191 jumping Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000000737 periodic effect Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Data Mining & Analysis (AREA)
- Operations Research (AREA)
- Pure & Applied Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Computational Mathematics (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Evolutionary Biology (AREA)
- Quality & Reliability (AREA)
- Probability & Statistics with Applications (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Entrepreneurship & Innovation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Algebra (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
Abstract
The invention discloses a main heating characteristic monitoring method and a main heating characteristic monitoring system. Based on the main transformer operation data, a main transformer oil temperature numerical calculation input quantity is determined. And calculating an input quantity, a least square method and a radial basis function neural network based on the oil temperature value, and constructing an oil temperature value calculation model corresponding to the main transformer. And performing longitudinal and transverse anomaly detection by using the oil temperature numerical calculation model and the main transformer historical operation data to generate a main transformer set with abnormal thermal characteristics. When the main transformer does not belong to the main transformer set with abnormal thermal characteristics, the oil temperature numerical calculation model and the main transformer historical operation data are adopted to conduct space-time dimension difference comparison, and the main transformer set is required to be paid attention to for generating the thermal characteristic difference change. And finally integrating the thermal characteristic abnormal main transformer set and the thermal characteristic difference change to pay attention to the main transformer set and generate a main heating characteristic monitoring report. By establishing a mathematical model and an early warning strategy, the monitoring data are toolized, so that the function expansion and popularization and application are facilitated.
Description
Technical Field
The invention relates to the technical field of main heating characteristic monitoring, in particular to a main heating characteristic monitoring method and system.
Background
The main transformer, which is abbreviated as a main transformer, is a main step-down transformer mainly used for power transmission and transformation in a unit or a transformer substation, and is also a core part of the transformer substation. The main transformer is a key device of an electric power system, and the main heating characteristic is an important factor for determining the load capacity and insulation aging speed of the device. As a core index characterizing the main heating characteristics, the oil temperature may reflect the thermal state inside the main transformer. In order to monitor the internal temperature change condition of the main transformer, an oil temperature gauge is arranged on the top layer of the main transformer, and oil temperature data are remotely transmitted to a dispatching center through a collecting device.
At present, the on-site operation and maintenance of the thermal characteristics of the transformer is to grasp the change condition of the main temperature and the operation condition of the oil temperature gauge by performing periodic inspection by operators. And carrying out periodical pre-testing on the oil temperature gauge by a tester, and determining the health state of the oil temperature gauge. And carrying out on-site maintenance and overhaul work by an overhaul worker according to the related defect report of the main heating state and the state evaluation result.
However, the on-site operation and maintenance modes of the thermal characteristics of the transformer are large in investment, and the problems that the input and output are low, the safety risk is high, the thermal state of the transformer cannot be mastered in real time, the abnormal condition of the main heating characteristics cannot be found in time and the like exist in on-site operation such as periodic inspection of the main temperature change and periodical pre-test of the temperature measuring device. Moreover, the situation awareness and the abnormality identification capability of workers on the main heating characteristic are poor, and the related business development lacks effective support and powerful grip. Therefore, the existing on-site operation and maintenance mode of the main heating characteristics cannot find abnormal conditions of the main heating characteristics in time, and the temperature data of the transformer is excavated to a low degree, so that the main heating characteristics are not monitored in time, not in place and not accurate.
Disclosure of Invention
The invention provides a main heating characteristic monitoring method and a main heating characteristic monitoring system, which solve the technical problems that the existing main heating characteristic on-site operation maintenance mode cannot find out the abnormal condition of the main heating characteristic in time and the temperature data of a transformer is low in mining degree, so that the main heating characteristic monitoring is not in time, out of place and inaccurate.
The invention provides a main heating characteristic monitoring method, which comprises the following steps:
determining a main transformer oil temperature numerical value to calculate an input quantity according to main transformer operation data;
calculating an input quantity, a least square method and a radial basis function neural network by adopting the main transformer oil temperature numerical value, and constructing an oil temperature numerical value calculation model corresponding to the main transformer;
performing longitudinal and transverse anomaly detection by adopting the oil temperature numerical calculation model and the main transformer historical operation data to generate a main transformer set with abnormal thermal characteristics;
when the main transformer does not belong to the main transformer set with abnormal thermal characteristics, carrying out space-time dimension difference comparison by adopting the oil temperature numerical calculation model and the main transformer historical operation data, and generating the main transformer set with attention to the thermal characteristic difference change;
and integrating the thermal characteristic abnormal main transformer set and the thermal characteristic difference change to pay attention to the main transformer set and generate a main heating characteristic monitoring report.
Optionally, the main transformer operation data comprises load data, voltage data, neutral point direct current component data, oil temperature data and environmental temperature data; the step of determining the main transformer oil temperature numerical value calculation input quantity according to the main transformer operation data comprises the following steps:
the load data, the voltage data, the neutral point direct current component data, the oil temperature data and the environment temperature data are adopted according to a time sequence, and a load sequence, a voltage sequence, a neutral point direct current component sequence, an oil temperature sequence and an environment temperature sequence corresponding to the main transformer are constructed;
respectively calculating the correlations among the oil temperature sequence, the load sequence, the voltage sequence, the neutral point direct current component sequence and the environment temperature sequence to generate a main-transformer correlation index set;
screening sequences corresponding to the correlation index threshold values with indexes larger than the preset correlation index threshold values in the main transformer correlation index set, and calculating input parameter sequences as oil temperature values;
calculating an autocorrelation index corresponding to the oil temperature sequence, and calculating a partial autocorrelation index corresponding to the oil temperature sequence by adopting the autocorrelation index;
when the partial autocorrelation index does not meet a preset confidence interval, taking a time lag parameter corresponding to the partial autocorrelation index as an input variable order;
And calculating an input parameter sequence by adopting the input variable order, the oil temperature sequence and the oil temperature numerical value, and constructing a main transformer oil temperature numerical value calculation input quantity.
Optionally, the step of constructing the oil temperature numerical calculation model corresponding to the main transformer by adopting the main transformer oil temperature numerical calculation input quantity, the least square method and the radial basis function neural network includes:
performing main transformer top layer oil temperature numerical calculation modeling by adopting a nonlinear least square method and the main transformer oil temperature numerical calculation input quantity, and generating a first target oil temperature numerical calculation model corresponding to the main transformer;
generating a first oil temperature value corresponding to the main transformer through the first target oil temperature value calculation model;
performing main transformer top layer oil temperature numerical calculation modeling by adopting a radial basis function neural network and combining the main transformer oil temperature numerical calculation input quantity, and generating a second target oil temperature numerical calculation model corresponding to the main transformer;
generating a second oil temperature value corresponding to the main transformer through the second target oil temperature value calculation model;
acquiring the true oil temperature value corresponding to the main transformer, and comparing the first oil temperature value and the second oil temperature value with the true oil temperature value at the corresponding moment respectively to generate corresponding calculation errors;
Constructing a calculation error information matrix corresponding to the calculation error by taking the minimum sum of squares of errors corresponding to the calculation error as a construction target;
respectively determining a first optimal weighting coefficient corresponding to the first target oil temperature numerical calculation model and a second optimal weighting coefficient corresponding to the second target oil temperature numerical calculation model through the error information matrix;
and constructing an oil temperature numerical value calculation model corresponding to the main transformer by adopting the first target oil temperature numerical value calculation model, the second target oil temperature numerical value calculation model, the first optimal weighting coefficient and the second optimal weighting coefficient.
Optionally, the step of performing main transformer top layer oil temperature numerical calculation modeling by using a nonlinear least square method and the main transformer oil temperature numerical calculation input quantity to generate a first target oil temperature numerical calculation model corresponding to the main transformer includes:
calculating an input quantity by adopting the main transformer oil temperature value, and constructing a first initial oil temperature value calculation model corresponding to the main transformer;
calculating input quantity by adopting a nonlinear least square method and the main transformer oil temperature numerical value, and determining a calculation error function corresponding to the first initial oil temperature numerical value calculation model;
Performing partial derivative calculation by taking the minimum value of the obtained calculation error function as a target, and generating a model coefficient corresponding to the main transformer;
and updating the first initial oil temperature numerical value calculation model by adopting the model coefficient to generate a first target oil temperature numerical value calculation model corresponding to the main transformer.
Optionally, the step of performing main transformer top layer oil temperature numerical calculation modeling by adopting a radial basis function neural network and combining the main transformer oil temperature numerical calculation input quantity to generate a second target oil temperature numerical calculation model corresponding to the main transformer comprises the following steps:
initializing a radial basis function neural network to generate a second initial oil temperature numerical calculation model corresponding to the main transformer;
calculating the input quantity of the main transformer oil temperature value to be used as a training sample set, and selecting a plurality of training samples from the training sample set to be respectively used as initial sample clustering centers according to a preset selection interval;
determining a middle sample clustering center corresponding to the training sample set by adopting the initial sample clustering center and the distance from each training sample to the initial sample clustering center;
when the middle sample clustering center is consistent with the corresponding initial sample clustering center, taking the initial sample clustering center as a target sample clustering center;
Determining radial basis function variance corresponding to the second initial oil temperature numerical calculation model by adopting a maximum distance value between the target sample cluster centers and the hidden layer node number corresponding to the initial sample cluster center;
determining a weight corresponding to the second initial oil temperature numerical calculation model by adopting a least square method, the target sample clustering center, the hidden layer node number and the maximum distance value;
and updating the second initial oil temperature numerical value calculation model by adopting the weight and the radial basis function variance to generate a second target oil temperature numerical value calculation model corresponding to the main transformer.
Optionally, the main transformer historical operation data comprises historical annual same-period month operation data, historical month operation data and adjacent main transformer historical month operation data; the step of detecting longitudinal and transverse anomalies by adopting the oil temperature numerical calculation model and the main transformer historical operation data to generate a main transformer set with abnormal thermal characteristics comprises the following steps:
updating the oil temperature numerical calculation model by adopting historical annual and lunar operation data corresponding to the main transformer to generate a first abnormality detection model;
inputting the historical month operation data corresponding to the main transformer into the first abnormality detection model to generate a first error sequence corresponding to the main transformer;
Calculating entropy values by adopting the first error sequence, and determining entropy values of the first error sequence corresponding to the main transformer;
when the first error sequence entropy value is larger than a preset sequence entropy threshold value, judging that the main transformer has thermal characteristic abnormality from a longitudinal dimension, and incorporating the main transformer into a thermal characteristic abnormality main transformer set;
when the first error sequence entropy value is smaller than or equal to a preset sequence entropy threshold value, updating the oil temperature numerical calculation model by adopting the adjacent main transformer historical month operation data to generate a second abnormality detection model;
inputting the historical month operation data corresponding to the main transformer into the second abnormality detection model to generate a second error sequence corresponding to the main transformer;
performing entropy calculation by adopting the second error sequence, and determining a second error sequence entropy corresponding to the main transformer;
and when the second error sequence entropy value is larger than the preset sequence entropy threshold value, judging that the main transformer has thermal characteristic abnormality from a transverse dimension, and incorporating the main transformer into the thermal characteristic abnormality main transformer set.
Optionally, the step of calculating the entropy value by using the first error sequence and determining the entropy value of the first error sequence corresponding to the main transformer includes:
Converting the first error sequence into an error vector sequence according to a preset vector dimension;
respectively calculating vector distances corresponding to vectors in the error vector sequence;
obtaining the number that the vector distance is smaller than a preset similarity tolerance, and generating a number statistics value corresponding to the error vector sequence;
calculating the ratio of the number statistics to the total distance corresponding to the error vector sequence, and generating a sequence ratio corresponding to the error vector sequence;
calculating the average value of the sequence ratio, and generating a sequence average value corresponding to the error vector sequence;
and when the number of data points corresponding to the sequence average value meets a preset value, determining a first error sequence entropy value corresponding to the main transformer by adopting the sequence average value and the corresponding historical sequence average value.
Optionally, when the main transformer does not belong to the main transformer set with abnormal thermal characteristics, performing space-time dimension difference comparison by adopting the oil temperature numerical calculation model and the main transformer historical operation data, and generating the main transformer set with attention to the thermal characteristic difference change includes:
when the main transformer does not belong to the thermal characteristic abnormal main transformer set, updating the oil temperature numerical calculation model by adopting historical annual and contemporaneous month operation data corresponding to the main transformer to generate a third abnormal detection model, inputting the historical annual and contemporaneous month operation data corresponding to the main transformer, and generating a third error sequence corresponding to the main transformer;
Updating the oil temperature numerical value calculation model by adopting the historical month operation data corresponding to the main transformer to generate a fourth abnormality detection model, and inputting the historical month operation data corresponding to the main transformer to generate a fourth error sequence corresponding to the main transformer;
determining a first error sequence mean value and a third error sequence mean value corresponding to the main transformer by adopting the first error sequence and the third error sequence;
determining a first error sequence standard deviation and a third error sequence standard deviation corresponding to the main transformer by adopting the first error sequence mean and the third error sequence mean;
determining a first test statistic corresponding to the main transformer by adopting the first error sequence mean value, the third error sequence mean value, the first error sequence standard deviation and the third error sequence standard deviation;
when the absolute value of the first test statistic is larger than a preset difference threshold, performing entropy calculation by adopting the third error sequence, and determining a third error sequence entropy corresponding to the main transformer;
when the first error sequence mean value is larger than the third error sequence mean value and the first error sequence entropy value is larger than the third error sequence entropy value, judging that the main transformer has thermal characteristic difference from a longitudinal dimension, and taking the main transformer into a main transformer set which needs to be concerned by thermal characteristic difference change;
And when the absolute value of the first test statistic is smaller than or equal to a preset difference threshold, the second error sequence and the fourth error sequence are adopted for transverse comparison, and the main transformer set is required to be concerned for generating the thermal characteristic difference change.
Optionally, when the absolute value of the first test statistic is less than or equal to a preset difference threshold, the step of using the second error sequence and the fourth error sequence to perform lateral comparison to generate the thermal characteristic difference change needs to pay attention to a main transformer set includes:
when the absolute value of the first test statistic is smaller than or equal to a preset difference threshold value, determining a second error sequence mean value and a fourth error sequence mean value corresponding to the main transformer by adopting the second error sequence and the fourth error sequence;
determining a second error sequence standard deviation and a fourth error sequence standard deviation corresponding to the main transformer by adopting the second error sequence mean and the fourth error sequence mean;
determining a second test statistic corresponding to the main transformer by adopting the second error sequence mean value, the fourth error sequence mean value, the second error sequence standard deviation and the fourth error sequence standard deviation;
When the absolute value of the second test statistic is larger than the preset difference threshold, determining a fourth error sequence entropy value corresponding to the main transformer by adopting the fourth error sequence;
and when the second error sequence average value is larger than the fourth error sequence average value and the second error sequence entropy value is larger than the fourth error sequence entropy value, judging that the main transformer has thermal characteristic difference from a transverse dimension, and taking the main transformer into a main transformer set for thermal characteristic difference change.
The invention also provides a main heating characteristic monitoring system, comprising:
the main transformer oil temperature numerical calculation input quantity determining module is used for determining main transformer oil temperature numerical calculation input quantity according to main transformer operation data;
the oil temperature numerical value calculation model construction module is used for constructing an oil temperature numerical value calculation model corresponding to the main transformer by adopting the main transformer oil temperature numerical value calculation input quantity, the least square method and the radial basis function neural network;
the thermal characteristic abnormal main transformer set generation module is used for carrying out longitudinal and transverse abnormality detection by adopting the oil temperature numerical calculation model and the main transformer historical operation data to generate a thermal characteristic abnormal main transformer set;
the main transformer set generation module is used for carrying out space-time dimension difference comparison by adopting the oil temperature numerical calculation model and the main transformer historical operation data when the main transformer does not belong to the main transformer set with abnormal thermal characteristics, and generating the main transformer set with the main transformer set required to be concerned for the thermal characteristic difference change;
And the thermal characteristic monitoring report generation module is used for integrating the thermal characteristic abnormal main transformer set and the thermal characteristic difference change main transformer set to generate a main heating characteristic monitoring report.
From the above technical scheme, the invention has the following advantages:
the invention determines the numerical calculation input quantity of the main transformer oil temperature based on the main transformer operation data. And calculating an input quantity, a least square method and a radial basis function neural network based on the main transformer oil temperature value, and constructing an oil temperature value calculation model corresponding to the main transformer. And performing longitudinal and transverse anomaly detection by using the oil temperature numerical calculation model and the main transformer historical operation data to generate a main transformer set with abnormal thermal characteristics. When the main transformer does not belong to the main transformer set with abnormal thermal characteristics, the oil temperature numerical calculation model and the main transformer historical operation data are adopted to conduct space-time dimension difference comparison, and the main transformer set is required to be paid attention to for generating the thermal characteristic difference change. And finally integrating the thermal characteristic abnormal main transformer set and the thermal characteristic difference change to pay attention to the main transformer set and generate a main heating characteristic monitoring report. The method solves the technical problems that the existing main heating characteristic field operation maintenance mode can not find the abnormal condition of the main heating characteristic in time and has low mining degree on the temperature data of the transformer, so that the main heating characteristic is not timely, not in place and not accurate in monitoring. On the basis of not increasing hardware investment, the abnormal and differential changes of the main heating characteristics are revealed through data depth analysis, remote, centralized and real-time monitoring of the main heating characteristics is assisted, the differential monitoring and abnormal diagnosis analysis of the main heating characteristics are realized, and the pertinence and the efficiency of field checking and maintenance work are improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained from these drawings without inventive faculty for a person skilled in the art.
FIG. 1 is a flow chart of steps of a method for monitoring a main heating characteristic according to a first embodiment of the present invention;
FIG. 2 is a flowchart illustrating a method for monitoring a main heating characteristic according to a second embodiment of the present invention;
FIG. 3 is a block flow diagram of a method for monitoring main heating characteristics according to a second embodiment of the present invention;
fig. 4 is a block diagram of a main heating characteristic monitoring system according to a third embodiment of the present invention.
Detailed Description
The embodiment of the invention provides a main heating characteristic monitoring method and a main heating characteristic monitoring system, which are used for solving the technical problems that the existing main heating characteristic on-site operation maintenance mode cannot find out the abnormal condition of the main heating characteristic in time and the temperature data of a transformer is low in mining degree, so that the main heating characteristic monitoring is not in time, not in place and not accurate.
In order to make the objects, features and advantages of the present invention more comprehensible, the technical solutions in the embodiments of the present invention are described in detail below with reference to the accompanying drawings, and it is apparent that the embodiments described below are only some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, fig. 1 is a flowchart illustrating a method for monitoring a main heating characteristic according to an embodiment of the invention.
The first embodiment of the invention provides a main heating characteristic monitoring method, which comprises the following steps:
and 101, determining a main transformer oil temperature numerical value to calculate an input quantity according to main transformer operation data.
The main transformer operation data refer to all data related to the main transformer operation process, including load data, voltage data, neutral point direct current component data, oil temperature data, environment temperature data and the like.
In the embodiment of the invention, the load sequence, the voltage sequence, the neutral point direct current component sequence, the oil temperature sequence and the environment temperature sequence corresponding to the main transformer are constructed by adopting the load data, the voltage data, the neutral point direct current component data, the oil temperature data and the environment temperature data according to the time sequence. And respectively calculating the correlations among the oil temperature sequence, the load sequence, the voltage sequence, the neutral point direct current component sequence and the environment temperature sequence, and generating a main transformer correlation index set. And screening a sequence corresponding to a correlation index threshold value with indexes larger than a preset correlation index in the main transformer correlation index set, and calculating an input parameter sequence as an oil temperature value. And calculating an autocorrelation index corresponding to the oil temperature sequence by calculating the autocorrelation index corresponding to the oil temperature sequence and adopting the autocorrelation index to calculate a partial autocorrelation index corresponding to the oil temperature sequence. When the partial autocorrelation index does not meet the preset confidence interval, taking the time lag parameter corresponding to the partial autocorrelation index as the input variable order. And finally, calculating an input parameter sequence by adopting the input variable order, the oil temperature sequence and the oil temperature value, and constructing a main transformer oil temperature value calculation input quantity.
And 102, calculating an input quantity, a least square method and a radial basis function neural network by adopting the main transformer oil temperature value, and constructing an oil temperature value calculation model corresponding to the main transformer.
In the embodiment of the invention, nonlinear least square method and main transformer oil temperature numerical calculation input quantity are adopted to carry out main transformer top layer oil temperature numerical calculation modeling, a first target oil temperature numerical calculation model corresponding to the main transformer is generated, and a first oil temperature numerical corresponding to the main transformer is generated through the first target oil temperature numerical calculation model. And carrying out main transformer top layer oil temperature numerical value calculation modeling by adopting a radial basis function neural network and combining main transformer oil temperature numerical value calculation input quantity, generating a second target oil temperature numerical value calculation model corresponding to the main transformer, and generating a second oil temperature numerical value corresponding to the main transformer through the second target oil temperature numerical value calculation model. And acquiring the true oil temperature value corresponding to the main transformer, and comparing the first oil temperature value and the second oil temperature value with the true oil temperature value at the corresponding moment respectively to generate corresponding calculation errors. And constructing a calculation error information matrix corresponding to the calculation error by taking the minimum sum of squares of errors corresponding to the calculation error as a construction target. And respectively determining a first optimal weighting coefficient corresponding to the first target oil temperature numerical calculation model and a second optimal weighting coefficient corresponding to the second target oil temperature numerical calculation model through the error information matrix. And constructing an oil temperature numerical value calculation model corresponding to the main transformer by adopting the first target oil temperature numerical value calculation model, the second target oil temperature numerical value calculation model, the first optimal weighting coefficient and the second optimal weighting coefficient.
And 103, performing longitudinal and transverse anomaly detection by adopting an oil temperature numerical calculation model and main transformer historical operation data to generate a main transformer set with abnormal thermal characteristics.
The main transformer historical operation data comprises historical annual same-period month operation data, historical month operation data and adjacent main transformer historical month operation data. The historical annual month operation data refer to the data of the past annual month of the main transformer to be analyzed. The historical month operation data refers to data of the last month of the main transformer to be analyzed. The historical month operation data of the adjacent main transformer refer to the data of the same transformer substation with the same type and the adjacent main transformer with the same type for the last month.
In the embodiment of the invention, the oil temperature numerical calculation model is updated by adopting the historical annual same-period month operation data corresponding to the main transformer, and a first abnormality detection model is generated. And inputting the historical month operation data corresponding to the main transformer into a first abnormality detection model to generate a first error sequence corresponding to the main transformer. And calculating an entropy value by adopting the first error sequence, and determining the entropy value of the first error sequence corresponding to the main transformer. When the first error sequence entropy value is larger than a preset sequence entropy threshold value, judging that the main transformer has thermal characteristic abnormality from the longitudinal dimension, and incorporating the main transformer into a thermal characteristic abnormality main transformer set. And when the first error sequence entropy value is smaller than or equal to a preset sequence entropy threshold value, updating the oil temperature numerical calculation model by adopting the adjacent main transformer historical month operation data, and generating a second abnormality detection model. And inputting the historical month operation data corresponding to the main transformer into a second abnormality detection model to generate a second error sequence corresponding to the main transformer. And calculating an entropy value by adopting the second error sequence, and determining a second error sequence entropy value corresponding to the main transformer. And when the second error sequence entropy value is larger than a preset sequence entropy threshold value, judging that the main transformer has thermal characteristic abnormality from the transverse dimension, and incorporating the main transformer into a thermal characteristic abnormality main transformer set.
And 104, when the main transformer does not belong to the main transformer set with abnormal thermal characteristics, performing space-time dimension difference comparison by adopting an oil temperature numerical calculation model and main transformer historical operation data, wherein the main transformer set is required to be concerned for generating the thermal characteristic difference change.
In the embodiment of the invention, when the main transformer does not belong to the thermal characteristic abnormal main transformer set, the oil temperature numerical calculation model is updated by adopting the historical annual same-period month operation data corresponding to the main transformer, and a third abnormal detection model is generated. And inputting the historical annual and lunar operation data corresponding to the main transformer into a third anomaly detection model to generate a third error sequence corresponding to the main transformer. And updating the oil temperature numerical calculation model by adopting the historical month operation data corresponding to the main transformer to generate a fourth abnormality detection model. And inputting the historical month operation data into a fourth abnormality detection model to generate a fourth error sequence corresponding to the main transformer. And determining a first error sequence mean value and a third error sequence mean value corresponding to the main transformer by adopting the first error sequence and the third error sequence. And determining the first error sequence standard deviation and the third error sequence standard deviation corresponding to the main transformer by adopting the first error sequence average value and the third error sequence average value. And determining a first test statistic corresponding to the main transformer by adopting the first error sequence mean value, the third error sequence mean value, the first error sequence standard deviation and the third error sequence standard deviation. And when the absolute value of the first test statistic is larger than a preset difference threshold, performing entropy calculation by adopting a third error sequence, and determining a third error sequence entropy corresponding to the main transformer. When the first error sequence average value is larger than the third error sequence average value and the first error sequence entropy value is larger than the third error sequence entropy value, judging that the main transformer has thermal characteristic difference from the longitudinal dimension, and taking the main transformer into the thermal characteristic difference change needs to pay attention to the main transformer set. And when the absolute value of the first test statistic is smaller than or equal to a preset difference threshold, the second error sequence and the fourth error sequence are adopted for transverse comparison, and the main transformer set is required to be concerned for generating the thermal characteristic difference change.
And 105, integrating the thermal characteristic abnormal main transformer set and the thermal characteristic difference change to pay attention to the main transformer set and generating a main heating characteristic monitoring report.
In the embodiment of the invention, the main transformer set with abnormal thermal characteristics and the main transformer set with thermal characteristic difference change needing attention are subjected to set analysis, and the main heating characteristic monitoring report corresponding to the main transformer is constructed, so that whether the main transformer is the main transformer with thermal characteristic difference change needing attention and the thermal characteristic abnormality degree corresponding to the main transformer are determined.
In the embodiment of the invention, based on main transformer operation data, the main transformer oil temperature numerical value calculation input quantity is determined. And adopting a main transformer oil temperature numerical value to calculate an input quantity, a least square method and a radial basis function neural network, and constructing an oil temperature numerical value calculation model corresponding to the main transformer. And performing longitudinal and transverse anomaly detection by using the oil temperature numerical calculation model and the main transformer historical operation data to generate a main transformer set with abnormal thermal characteristics. When the main transformer does not belong to the main transformer set with abnormal thermal characteristics, the oil temperature numerical calculation model and the main transformer historical operation data are adopted to conduct space-time dimension difference comparison, and the main transformer set is required to be paid attention to for generating the thermal characteristic difference change. And finally integrating the thermal characteristic abnormal main transformer set and the thermal characteristic difference change to pay attention to the main transformer set and generate a main heating characteristic monitoring report. The method solves the technical problems that the existing main heating characteristic field operation maintenance mode can not find the abnormal condition of the main heating characteristic in time and has low mining degree on the temperature data of the transformer, so that the main heating characteristic is not timely, not in place and not accurate in monitoring. On the basis of not increasing hardware investment, the abnormal and differential changes of the main heating characteristics are revealed through data depth analysis, remote, centralized and real-time monitoring of the main heating characteristics is assisted, the differential monitoring and abnormal diagnosis analysis of the main heating characteristics are realized, and the pertinence and the efficiency of field checking and maintenance work are improved.
Referring to fig. 2, fig. 2 is a flowchart illustrating a method for monitoring a main heating characteristic according to an embodiment of the invention.
Another method for monitoring main heating characteristics provided in the second embodiment of the present invention includes:
Further, the main transformer operation data includes load data, voltage data, neutral point direct current component data, oil temperature data, and ambient temperature data. Step 201 may include the following sub-steps S11-S16:
and S11, adopting load data, voltage data, neutral point direct current component data, oil temperature data and environment temperature data according to a time sequence to construct a load sequence, a voltage sequence, a neutral point direct current component sequence, an oil temperature sequence and an environment temperature sequence corresponding to the main transformer. And S12, respectively calculating correlations among the oil temperature sequence, the load sequence, the voltage sequence, the neutral point direct current component sequence and the environment temperature sequence, and generating a main transformer correlation index set. S13, screening sequences corresponding to the correlation index threshold values with indexes larger than the preset correlation index threshold values in the main transformer correlation index set, and calculating an input parameter sequence as an oil temperature value. S14, calculating an autocorrelation index corresponding to the oil temperature sequence, and calculating a partial autocorrelation index corresponding to the oil temperature sequence by adopting the autocorrelation index. And S15, when the partial autocorrelation index does not meet the preset confidence interval, taking the time lag parameter corresponding to the partial autocorrelation index as an input variable order. S16, calculating an input parameter sequence by adopting the input variable order, the oil temperature sequence and the oil temperature value, and constructing a main transformer oil temperature value calculation input quantity.
In the embodiment of the invention, load data, voltage data, neutral point direct current component data, oil temperature data and environmental temperature data are extracted from an information system and are assembled into a load sequence curve, a voltage sequence curve, a neutral point direct current component sequence curve, an oil temperature sequence curve and an environmental temperature sequence curve according to time sequence respectively, namely, a corresponding load sequence, a voltage sequence, a neutral point direct current component sequence, an oil temperature sequence and an environmental temperature sequence are constructed. Calculating correlation indexes between the oil temperature sequence and the load sequence, between the oil temperature sequence and the voltage sequence, between the oil temperature sequence and the neutral point direct current component sequence and between the oil temperature sequence and the environment temperature sequence respectively:
in the method, in the process of the invention,
is a sequencexAnd (3) withyIs a correlation index of (2); sequence(s)xRepresenting the sequence of oil temperatures, the sequenceyRepresenting any one sequence of a load sequence, a voltage sequence, a neutral point direct current component sequence and an ambient temperature sequence; />
The symbols are calculated for the expected values.
If it is
Wherein b is a correlation index threshold, indicating a sequencexSum sequenceyHigh correlation of (C) then sequenceyCalculating an input parameter sequence by selecting the main transformer oil temperature value; otherwise, indicate the sequencexSum sequenceyLow correlation of (C) then sequenceyThe input parameter sequence is calculated without being selected as the oil temperature value. In the preferred embodiment, the correlation index threshold b is set to be 0.5, and the load sequence, the environment temperature sequence and the oil temperature sequence are selected as main transformer oil temperature values to calculate the input parameter sequence.
Computing a sequencexI.e. the time lag of the oil temperature sequence is
Autocorrelation index->
：
In the method, in the process of the invention,
is a sequencexIs the first of (2)tA number of values; />
Is a sequencexIs>
A number of values; />
Is the sequence error sequence average value.
Using autocorrelation index
Calculating the time lag of the oil temperature sequence as +.>
Partial autocorrelation index->
：
The preset confidence interval is an interval range which needs to be met by the partial autocorrelation index set based on actual needs. Is generally arranged as
，nFor the total number of training samples.
Is provided with
Is output when the hysteresis order is +.>
In the time of partial autocorrelation index->
At 95% confidence interval, i.e. preset confidence interval +.>
In addition, the time lag parameter corresponding to the partial autocorrelation index +.>
Selected as the input variable order.
And finally, calculating an input parameter sequence by adopting the input variable order, the oil temperature sequence and the oil temperature value, and constructing a main transformer oil temperature value calculation input quantity.
And 202, performing main transformer top layer oil temperature numerical calculation modeling by adopting a nonlinear least square method and main transformer oil temperature numerical calculation input quantity, and generating a first target oil temperature numerical calculation model corresponding to the main transformer.
Further, step 202 may comprise the following sub-steps S21-S24:
s21, calculating input quantity by adopting the main transformer oil temperature value, and constructing a first initial oil temperature value calculation model corresponding to the main transformer. S22, calculating the input quantity by adopting a nonlinear least square method and the main transformer oil temperature numerical value, and determining a calculation error function corresponding to the first initial oil temperature numerical value calculation model. S23, taking the minimum value of the obtained calculation error function as a target to perform partial derivative calculation, and generating a model coefficient corresponding to the main transformer. S24, updating the first initial oil temperature numerical calculation model by using model coefficients to generate a first target oil temperature numerical calculation model corresponding to the main transformer.
In the embodiment of the invention, the load sequence in the oil temperature numerical calculation input quantity comprises a plurality of main transformer loads, namely a main transformer current load value and a main transformer historical load value. The oil temperature sequence includes a plurality of main transformer oil temperatures, i.e., includes a main transformer historical oil temperature value. The ambient temperature sequence includes a plurality of main transformer oil temperatures, i.e., includes a current ambient temperature and a historical ambient temperature.
Taking a current load value of a main transformer, a historical load value of the main transformer, a historical oil temperature value of the main transformer, a current environment temperature and a historical environment temperature as input quantities, and taking the current oil temperature value of the main transformer as output quantities, and constructing a first initial oil temperature numerical calculation model corresponding to the main transformer:
in the method, in the process of the invention,
and->
Respectively istTime of day and time of dayt-iA main transformer oil temperature value at moment; />
Is thatt-iAmbient temperature at time; />
Is thatt-iMain transformer load at moment; />
、/>
、/>
And->
The coefficients corresponding to the main transformer oil temperature, the environment temperature, the main transformer load and the constant term are respectively adopted.
Calculating input quantity by adopting a nonlinear least square method and a main transformer oil temperature numerical value, and determining a calculation error function corresponding to a first initial oil temperature numerical value calculation model
：
In the method, in the process of the invention,
is thattReal oil temperature of main transformer at moment; />
Calculated for a first initial oil temperature numerical calculation modeltA main transformer oil temperature value at moment;Dis the total time of data.
The model coefficients refer to coefficients corresponding to main transformer oil temperature, ambient temperature, main transformer load and constant terms. And taking the minimum error function value as a target, solving a partial derivative, enabling a first derivative to be 0, solving a model coefficient corresponding to the main transformer, and updating the first initial oil temperature numerical calculation model by adopting the model coefficient to generate a first target oil temperature numerical calculation model corresponding to the main transformer.
And 203, generating a first oil temperature value corresponding to the main transformer through a first target oil temperature value calculation model.
In the embodiment of the invention, a first target oil temperature numerical calculation model constructed based on a nonlinear least square method is used for calculating a first oil temperature numerical value corresponding to a main transformer, which is a numerical calculation value of the top oil temperature of the main transformer
。
And 204, carrying out main transformer top layer oil temperature numerical calculation modeling by adopting a radial basis function neural network and combining the main transformer oil temperature numerical calculation input quantity, and generating a second target oil temperature numerical calculation model corresponding to the main transformer.
Further, step 204 may include the following substeps S31-S37:
s31, initializing the radial basis function neural network to generate a second initial oil temperature numerical calculation model corresponding to the main transformer. S32, calculating the input quantity of the main transformer oil temperature value to be used as a training sample set, and selecting a plurality of training samples from the training sample set to be respectively used as initial sample clustering centers according to a preset selection interval. S33, determining a middle sample clustering center corresponding to the training sample set by adopting the initial sample clustering center and the distance from each training sample to the initial sample clustering center. And S34, when the middle sample clustering center is consistent with the corresponding initial sample clustering center, taking the initial sample clustering center as a target sample clustering center. S35, determining radial basis function variances corresponding to the second initial oil temperature numerical calculation model by adopting maximum distance values among target sample clustering centers and hidden layer node numbers corresponding to the initial sample clustering centers. S36, determining a weight corresponding to the second initial oil temperature numerical calculation model by adopting a least square method, a target sample clustering center, hidden layer node numbers and a maximum distance value. And S37, updating the second initial oil temperature numerical value calculation model by adopting the weight and the radial basis function variance, and generating a second target oil temperature numerical value calculation model corresponding to the main transformer.
In the embodiment of the invention, a radial basis function neural network is initialized to obtain a second initial oil temperature numerical calculation model corresponding to the main transformer. The main transformer oil temperature numerical value is adopted to calculate the input quantity as a training sample set, and the training sample set is randomly selected according to a preset selection intervalhThe training samples are used as initial clustering centers
（i=1,2,…,h），hIs the number of hidden layer nodes. Grouping the input training samples by nearest neighbor, and calculating the +.>
（p=1,2,…,n，nTotal number of training samples) and an initial cluster center +.>
Distance between training samples is assigned to the nearest cluster set +.>
. Re-calculating the cluster centers of all cluster types to obtain the middle corresponding to the training sample setInter-sample cluster center->
（j=1,2,…,h），hIs the number of hidden layer nodes.
In the method, in the process of the invention,
is a cluster set; />
Is a training sample.
If the clustering center is unchanged, namely the middle sample clustering center is consistent with the corresponding initial sample clustering center, obtaining
And (3) the target sample cluster center of the radial basis function neural network is obtained, otherwise, the S33 is returned.
Using maximum distance values between cluster centers of target samples
Hidden layer node number corresponding to initial sample clustering centerhDetermining radial basis function variance corresponding to the second initial oil temperature numerical calculation model >
。
Training samples using least squares
Target sample cluster center->
Number of hidden layer nodeshAnd maximum distance value->
Calculate the weight between the hidden layer and the output layer, namely the firstWeight corresponding to two initial oil temperature numerical calculation models
. And finally, updating the second initial oil temperature numerical calculation model by adopting the weight and the radial basis function variance, thereby generating a second target oil temperature numerical calculation model corresponding to the main transformer.
And 205, generating a second oil temperature value corresponding to the main transformer through a second target oil temperature value calculation model.
In the embodiment of the invention, the numerical calculation value of the top oil temperature of the main transformer, namely the numerical value of the second oil temperature corresponding to the main transformer, is calculated through a second target oil temperature numerical calculation model constructed based on the radial basis function neural network
。
And 206, acquiring the true oil temperature value corresponding to the main transformer, and comparing the first oil temperature value and the second oil temperature value with the true oil temperature value at the corresponding moment respectively to generate corresponding calculation errors.
In the embodiment of the invention, the true value of the oil temperature is combined
And oil temperature numerical value calculation value +.>
、/>
、…、
Can obtain the firstiThe numerical calculation model of the oil temperature is as followstCalculation error of time ∈>
The method comprises the following steps:
in the method, in the process of the invention,
is thattThe true value of the oil temperature at the moment; / >
Is thattTime of day (time)iThe oil temperature numerical value calculation values of the respective target oil temperature numerical value calculation models,i=1,2,…,m，mcalculating the number of models for single oil temperature values; />
Is thattTime of day (time)iCalculating errors of the numerical calculation model of the target oil temperature.
In the embodiment of the invention, it is provided thatFCalculating the error square sum for each single numerical calculation model, the oil temperature numerical calculation problem with the minimum of the error square sum as the target can be expressed as the following optimization problem:
in the method, in the process of the invention,
and->
Respectively the firsti、jCalculating the weighting coefficient of the model by using the numerical value of each target oil temperature; />
And->
First, thei、jCalculating errors of the numerical calculation model of the target oil temperature. Will beThe formula in the optimization problem is specially changed into a matrix form to be obtained:
in the method, in the process of the invention,
；/>
the coefficient matrix is constructed by adopting all weighting coefficients;Jis thatmDimension full 1 column vector; />
A column vector matrix constructed by column vectors;Eis thatm×mA dimension calculation error information matrix, wherein,
is the firsti、jCovariance of calculation errors of numerical calculation models of the respective target oil temperatures, in particular, < >>
Is the firstiThe sum of squares of the errors is calculated by the numerical calculation model of the individual target oil temperatures.
In the embodiment of the invention, a first optimal weighting coefficient corresponding to a first target oil temperature numerical value calculation model and a second optimal weighting coefficient corresponding to a second target oil temperature numerical value calculation model are calculated by combining an error information matrix with the first target oil temperature numerical value calculation model and the second target oil temperature numerical value calculation model.
And 209, constructing an oil temperature numerical calculation model corresponding to the main transformer by adopting the first target oil temperature numerical calculation model, the second target oil temperature numerical calculation model, the first optimal weighting coefficient and the second optimal weighting coefficient.
In the embodiment of the invention, a first target oil temperature numerical calculation model, a second target oil temperature numerical calculation model, a first optimal weighting coefficient and a second optimal weighting coefficient are adopted to obtain a final numerical calculation result of the top-layer oil temperature of the main transformer, and then the oil temperature numerical calculation model corresponding to the main transformer is obtained
. Wherein (1)>
The oil temperature numerical value calculation model corresponding to the main transformer is adopted; />
And->
Respectively the first A first partA weighting coefficient of the target oil temperature numerical calculation model and the second target oil temperature numerical calculation model; />
Is thattTime of day (time)A first partAn oil temperature numerical value calculation value of the target oil temperature numerical value calculation model; />
Is thattAnd calculating the oil temperature numerical value of the model according to the second target oil temperature numerical value at the moment.
And 210, performing longitudinal and transverse anomaly detection by adopting an oil temperature numerical calculation model and main transformer historical operation data to generate a main transformer set with abnormal thermal characteristics.
Further, the main transformer historical operation data comprises historical annual month operation data, historical month operation data and adjacent main transformer historical month operation data. Step 210 may include the following substeps S41-S48:
s41, updating the oil temperature numerical calculation model by adopting historical annual and lunar operation data corresponding to the main transformer to generate a first anomaly detection model. S42, inputting the historical month operation data corresponding to the main transformer into a first abnormality detection model, and generating a first error sequence corresponding to the main transformer. S43, performing entropy calculation by using the first error sequence, and determining the entropy of the first error sequence corresponding to the main transformer. And S44, judging that the main transformer has thermal characteristic abnormality from the longitudinal dimension when the first error sequence entropy value is larger than a preset sequence entropy threshold value, and incorporating the main transformer into a thermal characteristic abnormality main transformer set. And S45, when the first error sequence entropy value is smaller than or equal to a preset sequence entropy threshold value, updating the oil temperature numerical calculation model by adopting the adjacent main transformer historical month operation data, and generating a second abnormality detection model. S46, inputting the historical month operation data corresponding to the main transformer into a second abnormality detection model, and generating a second error sequence corresponding to the main transformer. S47, performing entropy calculation by using the second error sequence, and determining a second error sequence entropy corresponding to the main transformer. S48, when the second error sequence entropy value is larger than a preset sequence entropy threshold, judging that the main transformer has thermal characteristic abnormality from the transverse dimension, and incorporating the main transformer into a thermal characteristic abnormality main transformer set.
In the embodiment of the invention, the data of the past calendar history of the main transformer to be analyzed, namely the historical annual month operation data, are combined, the oil temperature numerical calculation model is updated by adopting the historical annual month operation data based on the steps of constructing the oil temperature numerical calculation model, the main transformer oil temperature numerical calculation model A, namely the first anomaly detection model, is established, the data of the last month of the main transformer to be analyzed, namely the historical month operation data, is utilized to carry out the main transformer oil temperature numerical calculation to be analyzed, and the first error of each numerical calculation result is obtained
。
In the middle of
For the moment of timetA first error of a main transformer oil temperature calculation result to be analyzed, which is obtained by the main transformer oil temperature numerical calculation model A; />
For the moment of timetA main transformer oil temperature calculation result to be analyzed is obtained by the main transformer oil temperature numerical calculation model A;
for the moment of timetAnd analyzing the actual value of the main transformer oil temperature.
Determining entropy value of first error sequence corresponding to main transformer through first error sequence
. If it is
And judging that the main heat characteristic to be analyzed is suspected to be abnormal from the longitudinal dimension, namely judging that the main transformer has the abnormal heat characteristic from the longitudinal dimension, and taking the main transformer into a main transformer set with the abnormal heat characteristic and jumping to the step 212. Wherein,,
for the first error sequence->
Calculating a first error sequence entropy value, < > >
The entropy threshold value of the sequence is preset and can be set according to historical values and expert experience.
If it is
The oil temperature numerical calculation model is updated by adopting the adjacent main transformer historical month operation data based on the step of constructing the oil temperature numerical calculation model, a main transformer oil temperature numerical calculation model B, namely a second abnormality detection model is established, the main transformer oil temperature numerical calculation model B to be analyzed is utilized to carry out the main transformer oil temperature numerical calculation to be analyzed according to the data of the main transformer of the same transformer substation of the same model of the same type, namely the adjacent main transformer historical month operation data, and the second error of each numerical calculation result is obtained>
。
In the method, in the process of the invention,
obtaining a main transformer oil temperature to be analyzed at moment for the main transformer oil temperature numerical calculation model BtCalculating a second error of the result; />
For the moment of timetCalculating the result of the main transformer oil temperature to be analyzed by a main transformer oil temperature numerical calculation model B>
For the moment of timetAnd analyzing the actual value of the main transformer oil temperature.
Determining entropy value of second error sequence corresponding to main transformer by second error sequence composed of all second errors corresponding to main transformer
If the entropy of the second error sequence is +.>
Greater than a preset sequence entropy threshold
Then it is determined from the lateral dimension that the main heating characteristic is suspected to be abnormal, the main transformer is included in the set of thermal characteristic abnormal main transformers and the process proceeds to step 212. Otherwise, go to step 211.
Further, S43 may include the following substeps S431-S436:
s431, converting the first error sequence into an error vector sequence according to a preset vector dimension. S432, respectively calculating vector distances corresponding to the vectors in the error vector sequence. S433, the number of vector distances smaller than the preset similarity tolerance is obtained, and the number statistics corresponding to the error vector sequence is generated. S434, calculating the ratio of the number statistics to the total distance corresponding to the error vector sequence, and generating the sequence ratio corresponding to the vector sequence. S435, calculating the average value of the sequence ratios, and generating a sequence average value corresponding to the error vector sequence. And S436, when the number of data points corresponding to the sequence average value meets a preset value, determining a first error sequence entropy value corresponding to the main transformer by adopting the sequence average value and the corresponding historical sequence average value.
In the embodiment of the invention, the preset vector dimension is the dimension corresponding to the pointing quantity sequence. And calculating entropy values by adopting the first error sequence, and constructing the first error sequence. Sequence the first error
Form a group ofmSequence of error vectors for dimensions:
wherein,,
，1≤i≤N-m+1，Nthe number of data points for the error sequence.
Defining vectors
And->
The vector distance between them is:
In the formula, 1 to less than or equal toj≤N-m+1 andj≠i；0≤a≤m-1。
the preset similarity margin refers to a critical value for measuring the vector distance, which is set based on actual needs. Setting similar toleranceslStatistical vector
And->
The distance between them is smaller thanlCount value of (2) and total distanceN-mTo obtain the sequence ratio corresponding to the vector sequence +.>
：
Wherein: num represents the statistics of the number; 1-1j≤N-m+1 andj≠i。
calculating the average value of the sequence ratio to generate a sequence average value corresponding to the error vector sequence
：
Dimension of sequence frommChange tom+1 dimension, repeating steps S431-S436, and calculating to obtain corresponding sequence average value
. When the number of data points corresponding to the sequence average value meets a preset value, namelyNWhen the value is limited, determining a first error sequence entropy value corresponding to the main transformer by adopting a sequence average value and a corresponding historical sequence average value>
：
In the method, in the process of the invention,mandlthe constant values 2 and 0.2std, std are the standard deviation of the sequence of timing errors. Similarly, the step of obtaining the entropy value of the second error sequence corresponding to the main transformer is the same as that of steps S431 to S436.
Further, step 211 may include the following substeps S51-S58:
and S51, when the main transformer does not belong to the thermal characteristic abnormal main transformer set, updating the oil temperature numerical calculation model by adopting the historical annual same-period month operation data corresponding to the main transformer to generate a third abnormal detection model, inputting the historical annual same-period month operation data corresponding to the main transformer, and generating a third error sequence corresponding to the main transformer. S52, updating the oil temperature numerical calculation model by adopting the historical month operation data corresponding to the main transformer to generate a fourth abnormality detection model, inputting the historical month operation data corresponding to the main transformer, and generating a fourth error sequence corresponding to the main transformer. S53, determining a first error sequence mean value and a third error sequence mean value corresponding to the main transformer by adopting the first error sequence and the third error sequence. S54, determining a first error sequence standard deviation and a third error sequence standard deviation corresponding to the main transformer by adopting the first error sequence mean and the third error sequence mean. S55, determining first test statistics corresponding to the main transformer by adopting the first error sequence mean value, the third error sequence mean value, the first error sequence standard deviation and the third error sequence standard deviation. And S56, when the absolute value of the first test statistic is larger than a preset difference threshold, performing entropy calculation by using a third error sequence, and determining a third error sequence entropy corresponding to the main transformer. S57, when the first error sequence average value is larger than the third error sequence average value and the first error sequence entropy value is larger than the third error sequence entropy value, judging that the main transformer has thermal characteristic difference from the longitudinal dimension, and taking the main transformer into the thermal characteristic difference change needs to pay attention to the main transformer set. And S58, when the absolute value of the first test statistic is smaller than or equal to a preset difference threshold, performing transverse comparison by adopting the second error sequence and the fourth error sequence, wherein the main transformer set is required to be concerned for generating the thermal characteristic difference change.
In the embodiment of the invention, when the main transformer does not belong to the thermal characteristic abnormal main transformer set, namely the main transformer is not included in the thermal characteristic abnormal main transformer set, the baseThe step of constructing the oil temperature numerical calculation model adopts the data of the same calendar history month of the past calendar of the main transformer to be analyzed, namely the data of the same calendar history month of the past calendar to update the oil temperature numerical calculation model, and the main transformer oil temperature numerical calculation model C, namely the third anomaly detection model, uses the data of the same calendar history month of the past calendar of the main transformer to be analyzed to carry out the calculation of the oil temperature numerical of the main transformer to be analyzed, and obtains the third error corresponding to each numerical calculation result
：
In the method, in the process of the invention,
obtaining the main transformer oil temperature to be analyzed at moment for the main transformer oil temperature numerical calculation model CtCalculating a third error of the result; />
For the moment of timetCalculating a main transformer oil temperature result to be analyzed by a main transformer oil temperature numerical calculation model C; />
For the moment of timetAnd analyzing the actual value of the main transformer oil temperature.
Based on the step of constructing the oil temperature numerical calculation model, updating the oil temperature numerical calculation model by adopting the data of the main transformer to be analyzed in the last month, namely analysis historical month operation data, wherein the main transformer oil temperature numerical calculation model D is a fourth abnormality detection model, and carrying out the calculation of the oil temperature of the main transformer to be analyzed by utilizing the data of the main transformer to be analyzed in the last month, namely the historical month operation data, so as to obtain a fourth error of each numerical calculation result 
：
In the method, in the process of the invention,
is a modelDObtaining the temperature of the main transformer oil to be analyzed at the momenttCalculating a fourth error of the result; />
For the moment of timetModelDCalculating a main transformer oil temperature result to be analyzed; />
For the moment of timetAnd analyzing the actual value of the main transformer oil temperature.
Respectively constructing a first error sequence by adopting all the first error, the second error, the third error and the fourth error
Second error sequence->
Third error sequence->
And fourth error sequence->
。/>
Respectively calculating a first error sequence mean value and a first error sequence standard deviation corresponding to the first error sequence by adopting an error sequence mean value formula and an error sequence standard deviation formula, wherein the error sequence mean value formula is as follows:
the error sequence standard deviation formula is:
；
in the method, in the process of the invention,
is a first error sequence; />
Is the first error sequence mean; />
As a first error-sequence standard deviation,Nthe number of data points for the error sequence. Similarly, the second error sequence +.>
Third error sequence->
And fourth error sequence->
Corresponding second error sequence mean +.>
Average value of third error sequence
Fourth error sequence mean->
Second error sequence standard deviation +.>
Third error sequence standard deviation->
Standard deviation of fourth error sequence->
。
Calculating a first test statistic of the first error sequence and the third error sequence by adopting the first error sequence mean value, the third error sequence mean value, the first error sequence standard deviation and the third error sequence standard deviation 
：
In the method, in the process of the invention,
is a first test statistic; />
Is the first error sequence mean; />
Is the third error sequence mean; />
Is the standard deviation of the first error sequence; />
Is the third error sequence standard deviation;Nthe number of data points for the error sequence.
Setting the significance level α=0.05 if
>1.96, determining that the thermal characteristics of the main transformer to be analyzed change in the time dimension, finding a difference in longitudinal comparison of the main thermal characteristics, and calculating according to step S43 to obtain a third error sequence
Entropy value of the third error sequence of +.>
. If->
>/>
And->
Marking the main transformer to be analyzed as thermal characteristicsThe difference change is to pay attention to the main transformer, that is, the main transformer is determined to have thermal characteristic difference from the longitudinal dimension, the main transformer is included in the thermal characteristic difference change, and the main transformer set is to be paid attention to, and the process goes to step 212. If->
And (5) jumping to the step S58, wherein the number of the jumping points is less than or equal to 1.96.
Further, step S58 may include the following substeps S581-S585:
s581, when the absolute value of the first test statistic is smaller than or equal to a preset difference threshold, determining a second error sequence mean value and a fourth error sequence mean value corresponding to the main transformer by adopting the second error sequence and the fourth error sequence. S582, determining a second error sequence standard deviation and a fourth error sequence standard deviation corresponding to the main transformer by adopting the second error sequence mean and the fourth error sequence mean. S583, determining a second test statistic corresponding to the main transformer by adopting the second error sequence mean value, the fourth error sequence mean value, the second error sequence standard deviation and the fourth error sequence standard deviation. S584, when the absolute value of the second test statistic is larger than a preset difference threshold, determining a fourth error sequence entropy value corresponding to the main transformer by adopting a fourth error sequence. S585, when the second error sequence average value is larger than the fourth error sequence average value and the second error sequence entropy value is larger than the fourth error sequence entropy value, judging that the main transformer has thermal characteristic difference from the transverse dimension, and taking the main transformer into the thermal characteristic difference change to pay attention to the main transformer set.
In the embodiment of the invention, a second error sequence is obtained by adopting an error sequence mean value formula and an error sequence standard deviation formula
Fourth error sequence->
Corresponding second error sequence mean +.>
Fourth error sequence mean->
Second error sequence standard deviation +.>
Standard deviation of fourth error sequence->
. The preset variance threshold is set to 1.96.
Calculating a second test statistic of the second error sequence and the fourth error sequence by adopting the second error sequence mean value, the fourth error sequence mean value, the second error sequence standard deviation and the fourth error sequence standard deviation
：
In the method, in the process of the invention,
is a second test statistic; />
Is the second error sequence mean; />
Is the fourth error sequence mean; />
Is the second error sequence standard deviation; />
Is the fourth error sequence standard deviation;Nthe number of data points for the error sequence.
Setting the significance level α=0.05 if
>1.96, determining that the thermal characteristics of the main transformer to be analyzed change in the spatial dimension, and finding a difference in the transverse comparison of the main thermal characteristics, and calculating according to step S43To fourth error sequence->
Entropy value of the fourth error sequence of +.>
. If it meets->
>/>
And->
And marking the main transformer to be analyzed as the thermal characteristic difference change, namely judging that the main transformer has the thermal characteristic difference from the transverse dimension, and focusing the main transformer into the thermal characteristic difference change. Otherwise, jump to step 212.
And 212, integrating the thermal characteristic abnormal main transformer set and the thermal characteristic difference change to pay attention to the main transformer set and generating a main heating characteristic monitoring report.
In the embodiment of the present invention, the implementation process of step 212 is similar to that of step 105, and will not be repeated here.
In the embodiment of the invention, as shown in fig. 3, the main transformer oil temperature value calculation input quantity is determined by calculating the correlation index. And (3) establishing an oil temperature numerical calculation model based on a least square method and a radial basis function neural network by taking the least square sum of errors as a target. And establishing a main transformer oil temperature numerical calculation model A by combining the data of the past calendar history and the same month of the main transformer to be analyzed. Carrying out oil temperature numerical calculation by utilizing data of the last month of main transformer to be analyzed, and analyzing numerical calculation first error
Constructing a first error sequence and calculating a first error sequence entropy value corresponding to the first error sequence>
. If it is
Namely AA>And if limit, judging that the main heating characteristic to be analyzed is suspected to be abnormal from the longitudinal dimension, and finishing the detection of the thermal characteristic abnormality, namely, the main transformer has the thermal characteristic abnormality, and finishing the monitoring corresponding to the main transformer. If it is
And further combining data of the nearest month of adjacent main transformers of the same transformer substation model to establish a main transformer oil temperature numerical calculation model B.
Carrying out oil temperature numerical calculation by utilizing data of the last month of the main transformer to be analyzed, and analyzing the numerical calculation to calculate a second error
And calculating a second error sequence entropy value corresponding to the second error sequence>
. If the entropy of the second error sequence is +.>
Greater than a preset sequence entropy threshold +.>
I.e. AB>And if limit, judging that the main heating characteristic is suspected to be abnormal from the transverse dimension, and incorporating the main transformer into a main transformer set with abnormal thermal characteristics to finish the detection of the abnormal thermal characteristics, namely, the main transformer has abnormal thermal characteristics, and finishing the monitoring corresponding to the main transformer.
If the main transformer does not have thermal characteristic abnormality, namely the main transformer does not contain a thermal characteristic abnormal main transformer set, establishing a main transformer oil temperature numerical value calculation model C by combining data of the main transformer to be analyzed in the same period of one calendar history, carrying out oil temperature numerical value calculation by using the data of the main transformer to be analyzed in the same period of one calendar history, and analyzing a third error of the numerical value
. Combining the data of the main transformer to be analyzed in the last month, establishing a main transformer oil temperature numerical calculation model D, and developing the oil temperature by utilizing the data of the main transformer to be analyzed in the last monthNumerical calculation, analysis of the numerical calculation fourth error +.>
。
Calculating a first test statistic for the first error sequence and the third error sequence
. If->
>1.96, the difference in the longitudinal alignment of the main heating characteristics is found, and a third error sequence +. >
Entropy value of the third error sequence of (2)
。
Judging whether or not to meet
>/>
And->
If yes, marking that the main transformer to be analyzed is the thermal characteristic difference, namely judging that the main transformer has the thermal characteristic difference from the time dimension when the main transformer is concerned, and taking the main transformer into the thermal characteristic difference change, namely paying attention to the main transformer set to complete thermal characteristic difference comparison, namely that the main transformer has the thermal characteristic difference, and completing the monitoring corresponding to the main transformer. If not, calculating a second test statistic of the second error sequence and the fourth error sequence by adopting the second error sequence mean value, the fourth error sequence mean value, the second error sequence standard deviation and the fourth error sequence standard deviation ≡>
. If->
>1.96, determining that the thermal characteristics of the main transformer to be analyzed change in the spatial dimension, finding a difference in transverse comparison of the main thermal characteristics, and calculating a fourth error sequence +.>
Entropy value of fourth error sequence of (2)
。
Judging whether or not to meet
>/>
And->
And if so, marking that the main transformer to be analyzed needs to be focused on the thermal characteristic difference change, taking the main transformer into a main transformer set which needs to be focused on the thermal characteristic difference change, and completing thermal characteristic difference comparison, namely, the main transformer has the thermal characteristic difference, thereby completing the monitoring corresponding to the main transformer. If not, the thermal characteristic difference comparison is completed, namely, the main transformer has no thermal characteristic difference, and the monitoring corresponding to the main transformer is completed. And finally integrating the thermal characteristic abnormal main transformer set and the thermal characteristic difference change to pay attention to the main transformer set and generate a main heating characteristic monitoring report. On the basis of not increasing hardware investment, a mathematical model and an early warning strategy are established, and monitoring data are tooled, so that the function expansion and popularization and application are facilitated. The data depth analysis reveals the abnormal and differential changes of the main heating characteristics, the remote, centralized and real-time monitoring of the main heating characteristics is assisted, the differential monitoring and the abnormal diagnosis analysis of the main heating characteristics are realized, the pertinence and the efficiency of the on-site checking and maintenance work are improved, and the goals of 'load reduction and efficiency increase and risk pre-control' are realized.
Referring to fig. 4, fig. 4 is a block diagram illustrating a main heating characteristic monitoring system according to a third embodiment of the present invention.
The third embodiment of the invention also provides a main heating characteristic monitoring system, which comprises:
the main transformer oil temperature value calculation input amount determining module 401 is configured to determine a main transformer oil temperature value calculation input amount according to main transformer operation data.
The oil temperature numerical calculation model construction module 402 is configured to construct an oil temperature numerical calculation model corresponding to the main transformer by adopting the main transformer oil temperature numerical calculation input quantity, the least square method and the radial basis function neural network.
And the thermal characteristic abnormal main transformer set generating module 403 is used for performing longitudinal and transverse anomaly detection by adopting the oil temperature numerical calculation model and the main transformer historical operation data to generate a thermal characteristic abnormal main transformer set.
The main transformer set generating module 404 is configured to perform space-time dimension difference comparison with the historical operating data of the main transformer by using the oil temperature numerical calculation model when the main transformer does not belong to the abnormal main transformer set, and generate the main transformer set.
The thermal characteristic monitoring report generating module 405 is configured to integrate the thermal characteristic abnormal main transformer set and the thermal characteristic difference change main transformer set to generate a main heating characteristic monitoring report.
Alternatively, the main transformer operation data includes load data, voltage data, neutral point direct current component data, oil temperature data, and ambient temperature data, and the main transformer oil temperature numerical calculation input quantity determination module 401 includes:
the load sequence, the voltage sequence, the neutral point direct current component sequence, the oil temperature sequence and the environment temperature sequence construction module is used for constructing a load sequence, a voltage sequence, the neutral point direct current component sequence, the oil temperature sequence and the environment temperature sequence corresponding to the main transformer by adopting load data, voltage data, neutral point direct current component data, oil temperature data and environment temperature data according to time sequence. And the main transformer correlation index set generation module is used for respectively calculating correlations among the oil temperature sequence, the load sequence, the voltage sequence, the neutral point direct current component sequence and the environment temperature sequence to generate a main transformer correlation index set. And the oil temperature numerical calculation input parameter sequence screening module is used for screening a sequence corresponding to a correlation index threshold value with an index larger than a preset correlation index in the main transformer correlation index set, and taking the sequence as an oil temperature numerical calculation input parameter sequence. And the partial autocorrelation index calculation module is used for calculating an autocorrelation index corresponding to the oil temperature sequence and calculating a partial autocorrelation index corresponding to the oil temperature sequence by adopting the autocorrelation index. And the input variable order determining module is used for taking the time lag parameter corresponding to the partial autocorrelation index as the input variable order when the partial autocorrelation index does not meet the preset confidence interval. And the oil temperature numerical calculation input quantity determination submodule is used for calculating an input parameter sequence by adopting the input variable order, the oil temperature sequence and the oil temperature numerical value to construct a main transformer oil temperature numerical calculation input quantity.
Optionally, the oil temperature numerical calculation model construction module 402 includes:
the first target oil temperature numerical calculation model generation module is used for carrying out main transformer top layer oil temperature numerical calculation modeling by adopting a nonlinear least square method and main transformer oil temperature numerical calculation input quantity to generate a first target oil temperature numerical calculation model corresponding to the main transformer. And the first oil temperature value generation module is used for generating a first oil temperature value corresponding to the main transformer through a first target oil temperature value calculation model. The second target oil temperature numerical calculation model generation module is used for carrying out main transformer top layer oil temperature numerical calculation modeling by adopting a radial basis function neural network and combining main transformer oil temperature numerical calculation input quantity to generate a second target oil temperature numerical calculation model corresponding to the main transformer. And the second oil temperature value generation module is used for generating a second oil temperature value corresponding to the main transformer through a second target oil temperature value calculation model. The calculation error sequence generation module is used for acquiring the oil temperature true value corresponding to the main transformer, and comparing the first oil temperature value and the second oil temperature value with the oil temperature true value at the corresponding moment respectively to generate a corresponding calculation error. The calculation error information matrix construction module is used for constructing a calculation error information matrix corresponding to the calculation error by taking the minimum sum of squares of errors corresponding to the calculation error as a construction target. The first optimal weighting coefficient and the second optimal weighting coefficient determining module are used for respectively determining a first optimal weighting coefficient corresponding to the first target oil temperature numerical value calculation model and a second optimal weighting coefficient corresponding to the second target oil temperature numerical value calculation model through the error information matrix. The oil temperature numerical calculation model construction submodule is used for constructing an oil temperature numerical calculation model corresponding to the main transformer by adopting a first target oil temperature numerical calculation model, a second target oil temperature numerical calculation model, a first optimal weighting coefficient and a second optimal weighting coefficient.
Alternatively, the first target oil temperature numerical calculation model generation module may perform the steps of:
calculating input quantity by adopting the main transformer oil temperature value, and constructing a first initial oil temperature value calculation model corresponding to the main transformer; calculating input quantity by adopting a nonlinear least square method and a main transformer oil temperature numerical value, and determining a calculation error function corresponding to a first initial oil temperature numerical value calculation model; performing partial derivative calculation by taking the minimum value of the obtained calculation error function as a target, and generating a model coefficient corresponding to the main transformer; and updating the first initial oil temperature numerical calculation model by adopting model coefficients to generate a first target oil temperature numerical calculation model corresponding to the main transformer.
Alternatively, the second target oil temperature numerical calculation model generation module may perform the steps of:
initializing a radial basis function neural network to generate a second initial oil temperature numerical calculation model corresponding to the main transformer; taking the main transformer oil temperature numerical value calculation input quantity as a training sample set, and selecting a plurality of training samples from the training sample set as initial sample clustering centers according to a preset selection interval; determining a middle sample clustering center corresponding to the training sample set by adopting the initial sample clustering center and the distance from each training sample to the initial sample clustering center; when the middle sample clustering center is consistent with the corresponding initial sample clustering center, taking the initial sample clustering center as a target sample clustering center; determining radial basis function variances corresponding to the second initial oil temperature numerical calculation model by adopting maximum distance values among target sample cluster centers and hidden layer node numbers corresponding to the initial sample cluster centers; determining a weight corresponding to the second initial oil temperature numerical calculation model by adopting a least square method, a target sample clustering center, an implicit layer node number and a maximum distance value; and updating the second initial oil temperature numerical calculation model by adopting the weight and the radial basis function variance to generate a second target oil temperature numerical calculation model corresponding to the main transformer.
Optionally, the main transformer historical operation data comprises historical annual month operation data, historical month operation data and adjacent main transformer historical month operation data. The thermal characteristics abnormal main transformer set generation module 403 includes:
the first abnormality detection model generation module is used for updating the oil temperature numerical calculation model by adopting historical annual and lunar operation data to generate a first abnormality detection model. The first error sequence generation module is used for inputting the historical month operation data corresponding to the main transformer into the first abnormality detection model to generate a first error sequence corresponding to the main transformer. The first error sequence entropy value determining module is used for calculating entropy values by adopting the first error sequence and determining a first error sequence entropy value corresponding to the main transformer. And the first submodule is used for judging that the main transformer has thermal characteristic abnormality from the longitudinal dimension when the first error sequence entropy value is larger than a preset sequence entropy threshold value, and taking the main transformer into the thermal characteristic abnormal main transformer set. And the second abnormality detection model generation module is used for updating the oil temperature numerical calculation model by adopting the adjacent main transformer historical month operation data when the first error sequence entropy value is smaller than or equal to a preset sequence entropy threshold value, so as to generate a second abnormality detection model. And the second error sequence generation module is used for inputting the historical month operation data corresponding to the main transformer into the second abnormality detection model to generate a second error sequence corresponding to the main transformer. And the second error sequence entropy value determining module is used for calculating entropy values by adopting the second error sequence and determining a second error sequence entropy value corresponding to the main transformer. And the second submodule is used for judging that the main transformer has thermal characteristic abnormality from the transverse dimension when the second error sequence entropy value is larger than the preset sequence entropy threshold value, and taking the main transformer into the thermal characteristic abnormal main transformer set.
Alternatively, the first error sequence entropy value determination module may perform the steps of:
converting the first error sequence into an error vector sequence according to a preset vector dimension; respectively calculating vector distances corresponding to vectors in the error vector sequence; obtaining the number of vector distances smaller than a preset similarity tolerance, and generating a number statistics value corresponding to an error vector sequence; calculating the ratio of the number statistics to the total distance corresponding to the error vector sequence to generate a sequence ratio corresponding to the error vector sequence; calculating the average value of the sequence ratio, and generating a sequence average value corresponding to the error vector sequence; when the number of data points corresponding to the sequence average value meets a preset value, determining a first error sequence entropy value corresponding to the main transformer by adopting the sequence average value and a corresponding historical sequence average value.
Optionally, the thermal property difference change attention main transformer set generating module 404 includes:
and the third error sequence generation module is used for updating the oil temperature numerical calculation model by adopting the historical annual and period month operation data corresponding to the main transformer when the main transformer does not belong to the thermal characteristic abnormal main transformer set, generating a third abnormal detection model, inputting the historical annual and period month operation data corresponding to the main transformer, and generating a third error sequence corresponding to the main transformer. And the fourth error sequence generation module is used for updating the oil temperature numerical value calculation model by adopting the historical month operation data corresponding to the main transformer to generate a fourth abnormality detection model, inputting the historical month operation data corresponding to the main transformer and generating a fourth error sequence corresponding to the main transformer. The first error sequence average value and the third error sequence average value determining module is used for determining a first error sequence average value and a third error sequence average value corresponding to the main transformer by adopting the first error sequence and the third error sequence. The first error sequence standard deviation and third error sequence standard deviation determining module is used for determining the first error sequence standard deviation and the third error sequence standard deviation corresponding to the main transformer by adopting the first error sequence mean value and the third error sequence mean value. The first test statistic determining module is used for determining a first test statistic corresponding to the main transformer by adopting the first error sequence mean value, the third error sequence mean value, the first error sequence standard deviation and the third error sequence standard deviation. And the third error sequence entropy value determining module is used for calculating an entropy value by adopting the third error sequence when the absolute value of the first test statistic is larger than a preset difference threshold value, and determining a third error sequence entropy value corresponding to the main transformer. The first submodule is used for judging that the main transformer has the thermal characteristic difference from the longitudinal dimension when the first error sequence average value is larger than the third error sequence average value and the first error sequence entropy value is larger than the third error sequence entropy value, and taking the main transformer into the thermal characteristic difference change to pay attention to the main transformer set. The thermal characteristic difference change needs to pay attention to the main transformer set to generate a second submodule, and the second submodule is used for carrying out transverse comparison by adopting a second error sequence and a fourth error sequence when the absolute value of the first test statistic is smaller than or equal to a preset difference threshold value, so that the thermal characteristic difference change needs to pay attention to the main transformer set.
Alternatively, the generating the second sub-module by focusing on the main transformer set may perform the following steps:
when the absolute value of the first test statistic is smaller than or equal to a preset difference threshold value, a second error sequence and a fourth error sequence are adopted to determine a second error sequence mean value and a fourth error sequence mean value corresponding to the main transformer; determining a second error sequence standard deviation and a fourth error sequence standard deviation corresponding to the main transformer by adopting a second error sequence mean value and a fourth error sequence mean value; determining a second test statistic corresponding to the main transformer by adopting a second error sequence mean value, a fourth error sequence mean value, a second error sequence standard deviation and a fourth error sequence standard deviation; when the absolute value of the second test statistic is larger than a preset difference threshold, determining a fourth error sequence entropy value corresponding to the main transformer by adopting a fourth error sequence; when the second error sequence mean value is larger than the fourth error sequence mean value and the second error sequence entropy value is larger than the fourth error sequence entropy value, judging that the main transformer has thermal characteristic difference from the transverse dimension, and taking the main transformer into the thermal characteristic difference change needs to pay attention to the main transformer set.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein.
In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods may be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of elements is merely a logical functional division, and there may be additional divisions of actual implementation, e.g., multiple elements or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (9)
1. A method of monitoring a main heating characteristic, comprising:
determining a main transformer oil temperature numerical value to calculate an input quantity according to main transformer operation data;
calculating an input quantity, a least square method and a radial basis function neural network by adopting the main transformer oil temperature numerical value, and constructing an oil temperature numerical value calculation model corresponding to the main transformer;
performing longitudinal and transverse anomaly detection by adopting the oil temperature numerical calculation model and the main transformer historical operation data to generate a main transformer set with abnormal thermal characteristics;
when the main transformer does not belong to the main transformer set with abnormal thermal characteristics, carrying out space-time dimension difference comparison by adopting the oil temperature numerical calculation model and the main transformer historical operation data, and generating the main transformer set with attention to the thermal characteristic difference change;
integrating the thermal characteristic abnormal main transformer set and the thermal characteristic difference change to pay attention to the main transformer set and generating a main heating characteristic monitoring report;
the main transformer historical operation data comprise historical annual same-period month operation data, historical month operation data and adjacent main transformer historical month operation data; the step of detecting longitudinal and transverse anomalies by adopting the oil temperature numerical calculation model and the main transformer historical operation data to generate a main transformer set with abnormal thermal characteristics comprises the following steps:
Updating the oil temperature numerical calculation model by adopting historical annual and lunar operation data corresponding to the main transformer to generate a first abnormality detection model;
inputting the historical month operation data corresponding to the main transformer into the first abnormality detection model to generate a first error sequence corresponding to the main transformer;
calculating entropy values by adopting the first error sequence, and determining entropy values of the first error sequence corresponding to the main transformer;
when the first error sequence entropy value is larger than a preset sequence entropy threshold value, judging that the main transformer has thermal characteristic abnormality from a longitudinal dimension, and incorporating the main transformer into a thermal characteristic abnormality main transformer set;
when the first error sequence entropy value is smaller than or equal to the preset sequence entropy threshold value, updating the oil temperature numerical calculation model by adopting the adjacent main transformer historical month operation data to generate a second abnormality detection model;
inputting the historical month operation data corresponding to the main transformer into the second abnormality detection model to generate a second error sequence corresponding to the main transformer;
performing entropy calculation by adopting the second error sequence, and determining a second error sequence entropy corresponding to the main transformer;
and when the second error sequence entropy value is larger than the preset sequence entropy threshold value, judging that the main transformer has thermal characteristic abnormality from a transverse dimension, and incorporating the main transformer into the thermal characteristic abnormality main transformer set.
2. The main warming characteristic monitoring method according to claim 1, wherein the main transformer operation data includes load data, voltage data, neutral point direct current component data, oil temperature data, and ambient temperature data; the step of determining the main transformer oil temperature numerical value calculation input quantity according to the main transformer operation data comprises the following steps:
the load data, the voltage data, the neutral point direct current component data, the oil temperature data and the environment temperature data are adopted according to a time sequence, and a load sequence, a voltage sequence, a neutral point direct current component sequence, an oil temperature sequence and an environment temperature sequence corresponding to the main transformer are constructed;
respectively calculating the correlations among the oil temperature sequence, the load sequence, the voltage sequence, the neutral point direct current component sequence and the environment temperature sequence to generate a main-transformer correlation index set;
screening sequences corresponding to the correlation index threshold values with indexes larger than the preset correlation index threshold values in the main transformer correlation index set, and calculating input parameter sequences as oil temperature values;
calculating an autocorrelation index corresponding to the oil temperature sequence, and calculating a partial autocorrelation index corresponding to the oil temperature sequence by adopting the autocorrelation index;
When the partial autocorrelation index does not meet a preset confidence interval, taking a time lag parameter corresponding to the partial autocorrelation index as an input variable order;
and calculating an input parameter sequence by adopting the input variable order, the oil temperature sequence and the oil temperature numerical value, and constructing a main transformer oil temperature numerical value calculation input quantity.
3. The method for monitoring main heating characteristics according to claim 1, wherein the step of constructing a main transformer-corresponding oil temperature numerical calculation model by calculating an input amount, a least square method, and a radial basis function neural network using the main transformer oil temperature numerical value comprises:
performing main transformer top layer oil temperature numerical calculation modeling by adopting a nonlinear least square method and the main transformer oil temperature numerical calculation input quantity, and generating a first target oil temperature numerical calculation model corresponding to the main transformer;
generating a first oil temperature value corresponding to the main transformer through the first target oil temperature value calculation model;
performing main transformer top layer oil temperature numerical calculation modeling by adopting a radial basis function neural network and combining the main transformer oil temperature numerical calculation input quantity, and generating a second target oil temperature numerical calculation model corresponding to the main transformer;
generating a second oil temperature value corresponding to the main transformer through the second target oil temperature value calculation model;
Acquiring the true oil temperature value corresponding to the main transformer, and comparing the first oil temperature value and the second oil temperature value with the true oil temperature value at the corresponding moment respectively to generate corresponding calculation errors;
constructing a calculation error information matrix corresponding to the calculation error by taking the minimum sum of squares of errors corresponding to the calculation error as a construction target;
respectively determining a first optimal weighting coefficient corresponding to the first target oil temperature numerical value calculation model and a second optimal weighting coefficient corresponding to the second target oil temperature numerical value calculation model through the calculation error information matrix;
and constructing an oil temperature numerical value calculation model corresponding to the main transformer by adopting the first target oil temperature numerical value calculation model, the second target oil temperature numerical value calculation model, the first optimal weighting coefficient and the second optimal weighting coefficient.
4. The method for monitoring main heating characteristics according to claim 3, wherein the step of performing main transformer top-layer oil temperature numerical calculation modeling using a nonlinear least square method and the main transformer oil temperature numerical calculation input amount to generate the first target oil temperature numerical calculation model corresponding to the main transformer comprises:
Calculating an input quantity by adopting the main transformer oil temperature value, and constructing a first initial oil temperature value calculation model corresponding to the main transformer;
calculating input quantity by adopting a nonlinear least square method and the main transformer oil temperature numerical value, and determining a calculation error function corresponding to the first initial oil temperature numerical value calculation model;
performing partial derivative calculation by taking the minimum value of the obtained calculation error function as a target, and generating a model coefficient corresponding to the main transformer;
and updating the first initial oil temperature numerical value calculation model by adopting the model coefficient to generate a first target oil temperature numerical value calculation model corresponding to the main transformer.
5. The method for monitoring main heating characteristics according to claim 3, wherein the step of performing main transformer top-layer oil temperature numerical calculation modeling by using a radial basis function neural network in combination with the main transformer oil temperature numerical calculation input quantity to generate the second target oil temperature numerical calculation model corresponding to the main transformer comprises the steps of:
initializing a radial basis function neural network to generate a second initial oil temperature numerical calculation model corresponding to the main transformer;
calculating the input quantity of the main transformer oil temperature value to be used as a training sample set, and selecting a plurality of training samples from the training sample set to be respectively used as initial sample clustering centers according to a preset selection interval;
Determining a middle sample clustering center corresponding to the training sample set by adopting the initial sample clustering center and the distance from each training sample to the initial sample clustering center;
when the middle sample clustering center is consistent with the corresponding initial sample clustering center, taking the initial sample clustering center as a target sample clustering center;
determining radial basis function variance corresponding to the second initial oil temperature numerical calculation model by adopting a maximum distance value between the target sample cluster centers and the hidden layer node number corresponding to the initial sample cluster center;
determining a weight corresponding to the second initial oil temperature numerical calculation model by adopting a least square method, the target sample clustering center, the hidden layer node number and the maximum distance value;
and updating the second initial oil temperature numerical value calculation model by adopting the weight and the radial basis function variance to generate a second target oil temperature numerical value calculation model corresponding to the main transformer.
6. The method for monitoring main heating characteristics according to claim 1, wherein the step of calculating an entropy value by using the first error sequence and determining the entropy value of the first error sequence corresponding to the main transformer comprises:
Converting the first error sequence into an error vector sequence according to a preset vector dimension;
respectively calculating vector distances corresponding to vectors in the error vector sequence;
obtaining the number that the vector distance is smaller than a preset similarity tolerance, and generating a number statistics value corresponding to the error vector sequence;
calculating the ratio of the number statistics to the total distance corresponding to the error vector sequence, and generating a sequence ratio corresponding to the error vector sequence;
calculating the average value of the sequence ratio, and generating a sequence average value corresponding to the error vector sequence;
and when the number of data points corresponding to the sequence average value meets a preset value, determining a first error sequence entropy value corresponding to the main transformer by adopting the sequence average value and the corresponding historical sequence average value.
7. The method for monitoring main heating characteristics according to claim 1, wherein when the main transformer does not belong to the set of abnormal main transformers of thermal characteristics, performing space-time dimension difference comparison by using the oil temperature numerical calculation model and the historical operation data of the main transformer, and generating the set of main transformers to be concerned with the change of the thermal characteristics, comprises the steps of:
When the main transformer does not belong to the thermal characteristic abnormal main transformer set, updating the oil temperature numerical calculation model by adopting historical annual and contemporaneous month operation data corresponding to the main transformer to generate a third abnormal detection model, inputting the historical annual and contemporaneous month operation data corresponding to the main transformer, and generating a third error sequence corresponding to the main transformer;
updating the oil temperature numerical value calculation model by adopting the historical month operation data corresponding to the main transformer to generate a fourth abnormality detection model, and inputting the historical month operation data corresponding to the main transformer to generate a fourth error sequence corresponding to the main transformer;
determining a first error sequence mean value and a third error sequence mean value corresponding to the main transformer by adopting the first error sequence and the third error sequence;
determining a first error sequence standard deviation and a third error sequence standard deviation corresponding to the main transformer by adopting the first error sequence mean and the third error sequence mean;
determining a first test statistic corresponding to the main transformer by adopting the first error sequence mean value, the third error sequence mean value, the first error sequence standard deviation and the third error sequence standard deviation;
When the absolute value of the first test statistic is larger than a preset difference threshold, performing entropy calculation by adopting the third error sequence, and determining a third error sequence entropy corresponding to the main transformer;
when the first error sequence mean value is larger than the third error sequence mean value and the first error sequence entropy value is larger than the third error sequence entropy value, judging that the main transformer has thermal characteristic difference from a longitudinal dimension, and taking the main transformer into a main transformer set which needs to be concerned by thermal characteristic difference change;
and when the absolute value of the first test statistic is smaller than or equal to a preset difference threshold, the second error sequence and the fourth error sequence are adopted for transverse comparison, and the main transformer set is required to be concerned for generating the thermal characteristic difference change.
8. The method for monitoring main thermal characteristics according to claim 7, wherein when the absolute value of the first test statistic is less than or equal to a preset difference threshold, the step of using the second error sequence and the fourth error sequence to perform lateral comparison, and generating the thermal characteristic difference change requires attention to a main transformer set, includes:
when the absolute value of the first test statistic is smaller than or equal to a preset difference threshold value, determining a second error sequence mean value and a fourth error sequence mean value corresponding to the main transformer by adopting the second error sequence and the fourth error sequence;
Determining a second error sequence standard deviation and a fourth error sequence standard deviation corresponding to the main transformer by adopting the second error sequence mean and the fourth error sequence mean;
determining a second test statistic corresponding to the main transformer by adopting the second error sequence mean value, the fourth error sequence mean value, the second error sequence standard deviation and the fourth error sequence standard deviation;
when the absolute value of the second test statistic is larger than the preset difference threshold, determining a fourth error sequence entropy value corresponding to the main transformer by adopting the fourth error sequence;
and when the second error sequence average value is larger than the fourth error sequence average value and the second error sequence entropy value is larger than the fourth error sequence entropy value, judging that the main transformer has thermal characteristic difference from a transverse dimension, and taking the main transformer into a main transformer set for thermal characteristic difference change.
9. A main heating characteristic monitoring system, comprising:
the main transformer oil temperature numerical calculation input quantity determining module is used for determining main transformer oil temperature numerical calculation input quantity according to main transformer operation data;
the oil temperature numerical value calculation model construction module is used for constructing an oil temperature numerical value calculation model corresponding to the main transformer by adopting the main transformer oil temperature numerical value calculation input quantity, the least square method and the radial basis function neural network;
The thermal characteristic abnormal main transformer set generation module is used for carrying out longitudinal and transverse abnormality detection by adopting the oil temperature numerical calculation model and the main transformer historical operation data to generate a thermal characteristic abnormal main transformer set;
the main transformer set generation module is used for carrying out space-time dimension difference comparison by adopting the oil temperature numerical calculation model and the main transformer historical operation data when the main transformer does not belong to the main transformer set with abnormal thermal characteristics, and generating the main transformer set with the main transformer set required to be concerned for the thermal characteristic difference change;
the thermal characteristic monitoring report generation module is used for integrating the thermal characteristic abnormal main transformer set and the thermal characteristic difference change main transformer set to be concerned with, and generating a main heating characteristic monitoring report;
the main transformer historical operation data comprise historical annual same-period month operation data, historical month operation data and adjacent main transformer historical month operation data; the thermal characteristic abnormal main transformer set generation module comprises:
the first abnormality detection model generation module is used for updating the oil temperature numerical calculation model by adopting historical annual contemporaneous month operation data corresponding to the main transformer to generate a first abnormality detection model;
the first error sequence generation module is used for inputting the historical month operation data corresponding to the main transformer into the first abnormality detection model to generate a first error sequence corresponding to the main transformer;
The first error sequence entropy value determining module is used for calculating entropy values by adopting the first error sequence and determining a first error sequence entropy value corresponding to the main transformer;
the first submodule is used for judging that the main transformer has thermal characteristic abnormality from the longitudinal dimension when the first error sequence entropy value is larger than a preset sequence entropy threshold value, and taking the main transformer into the thermal characteristic abnormal main transformer set;
the second abnormality detection model generation module is used for updating the oil temperature numerical calculation model by adopting the adjacent main transformer historical month operation data when the first error sequence entropy value is smaller than or equal to the preset sequence entropy threshold value, so as to generate a second abnormality detection model;
the second error sequence generation module is used for inputting the historical month operation data corresponding to the main transformer into the second abnormality detection model to generate a second error sequence corresponding to the main transformer;
the second error sequence entropy value determining module is used for calculating entropy values by adopting the second error sequence and determining a second error sequence entropy value corresponding to the main transformer;
and the second sub-module is used for judging that the main transformer has thermal characteristic abnormality from a transverse dimension when the second error sequence entropy value is larger than the preset sequence entropy threshold value, and incorporating the main transformer into the thermal characteristic abnormality main transformer set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310383188.1A CN116108760B (en) | 2023-04-12 | 2023-04-12 | Main heating characteristic monitoring method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310383188.1A CN116108760B (en) | 2023-04-12 | 2023-04-12 | Main heating characteristic monitoring method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116108760A CN116108760A (en) | 2023-05-12 |
| CN116108760B true CN116108760B (en) | 2023-06-27 |
Family
ID=86258274
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310383188.1A Active CN116108760B (en) | 2023-04-12 | 2023-04-12 | Main heating characteristic monitoring method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116108760B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109856299A (en) * | 2018-11-26 | 2019-06-07 | 国家电网有限公司 | A kind of transformer online monitoring differentiation threshold value dynamic setting method, system |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105512474B (en) * | 2015-12-02 | 2017-12-12 | 国网山东省电力公司电力科学研究院 | A kind of method for detecting abnormality of Transformer's Condition Monitoring data |
| CN109871660B (en) * | 2019-03-26 | 2020-09-18 | 国网江苏省电力有限公司扬州供电分公司 | Early warning method and fault positioning method for heating fault of main transformer |
| CN111461922B (en) * | 2020-04-02 | 2023-04-21 | 国网冀北电力有限公司唐山供电公司 | Real-time prediction method for hot spot temperature of transformer based on extreme learning machine |
| CN111797365B (en) * | 2020-06-09 | 2022-11-25 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | Converter transformer temperature abnormity judgment method and system |
| CN112464566B (en) * | 2020-12-07 | 2023-01-24 | 广东电网有限责任公司佛山供电局 | Transformer oil temperature prediction method based on genetic algorithm and BP neural network |
| CN113343564A (en) * | 2021-05-28 | 2021-09-03 | 国网江苏省电力有限公司南通供电分公司 | Transformer top layer oil temperature prediction method based on multi-element empirical mode decomposition |
| CN115358355B (en) * | 2022-10-24 | 2023-01-24 | 广东电网有限责任公司佛山供电局 | Method and device for judging main transformer oil temperature gauge and top layer oil temperature abnormity |
| CN115878958B (en) * | 2022-11-18 | 2023-09-22 | 贵州电网有限责任公司 | Transformer oil temperature prediction method, device, equipment and storage medium |
| CN115655516B (en) * | 2022-12-12 | 2023-03-21 | 广东电网有限责任公司佛山供电局 | Transformer top layer oil temperature abnormity checking method and related device |
-
2023
- 2023-04-12 CN CN202310383188.1A patent/CN116108760B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109856299A (en) * | 2018-11-26 | 2019-06-07 | 国家电网有限公司 | A kind of transformer online monitoring differentiation threshold value dynamic setting method, system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116108760A (en) | 2023-05-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gorjian et al. | A review on degradation models in reliability analysis | |
| Tabesh et al. | Assessing pipe failure rate and mechanical reliability of water distribution networks using data-driven modeling | |
| CN108763654A (en) | A kind of electrical equipment fault prediction technique based on Weibull distribution and hidden Semi-Markov Process | |
| CN112101480A (en) | Multivariate clustering and fused time sequence combined prediction method | |
| Yang et al. | The multiplicative consistency threshold of intuitionistic fuzzy preference relation | |
| CN116150897A (en) | Machine tool spindle performance evaluation method and system based on digital twin | |
| CN110530650B (en) | Method for monitoring performance state of heavy-duty gas turbine based on generalized regression neural network and box diagram analysis | |
| Jurado et al. | Fuzzy inductive reasoning forecasting strategies able to cope with missing data: A smart grid application | |
| CN112417627A (en) | Power distribution network operation reliability analysis method based on four-dimensional index system | |
| CN115713320A (en) | Operation and maintenance strategy optimization method and device for equipment, equipment and storage medium | |
| CN114091600B (en) | Data-driven satellite association fault propagation path identification method and system | |
| Nastro et al. | A machine learning approach based on neural networks for energy diagnosis of telecommunication sites | |
| Hoskins et al. | Modelling the degradation of condition indices | |
| CN117553840A (en) | Instrument based on intelligent management and system thereof | |
| CN116108760B (en) | Main heating characteristic monitoring method and system | |
| CN112800674A (en) | Power system element reliability parameter calculation method based on extreme learning machine | |
| CN117191147A (en) | Flood discharge dam water level monitoring and early warning method and system | |
| CN116384732A (en) | Intelligent assessment method, system, storage medium and computing device for station pipeline risk | |
| Sim et al. | Missing data imputation algorithm for transmission systems based on multivariate imputation with principal component analysis | |
| CN116451911A (en) | Equipment technical system evaluation method and system based on capability gap | |
| CN116957534A (en) | Method for predicting replacement number of intelligent electric meter | |
| Kestelyn et al. | Towards a hybrid twin for infrastructure asset management: Investigation on power transformer asset maintenance management | |
| CN113919204B (en) | Comprehensive importance analysis method for availability of multi-state manufacturing system | |
| CN109376957A (en) | A kind of prediction technique of thermal power plant's load | |
| Uddin et al. | Comparison of some statistical forecasting techniques with GMDH predictor: A case study |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |