CN116307773A - Reliability estimation method for secondary equipment of transformer substation - Google Patents
Reliability estimation method for secondary equipment of transformer substation Download PDFInfo
- Publication number
- CN116307773A CN116307773A CN202310306309.2A CN202310306309A CN116307773A CN 116307773 A CN116307773 A CN 116307773A CN 202310306309 A CN202310306309 A CN 202310306309A CN 116307773 A CN116307773 A CN 116307773A
- Authority
- CN
- China
- Prior art keywords
- secondary equipment
- transformer substation
- reliability
- substation
- fault rate
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000012360 testing method Methods 0.000 claims description 33
- 230000007787 long-term memory Effects 0.000 claims description 32
- 238000013528 artificial neural network Methods 0.000 claims description 26
- 238000005315 distribution function Methods 0.000 claims description 23
- 230000006403 short-term memory Effects 0.000 claims description 20
- 238000012549 training Methods 0.000 claims description 20
- 239000011159 matrix material Substances 0.000 claims description 19
- 230000015654 memory Effects 0.000 claims description 17
- 125000004122 cyclic group Chemical group 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 10
- 230000006872 improvement Effects 0.000 claims description 7
- 230000002452 interceptive effect Effects 0.000 claims description 5
- 230000004913 activation Effects 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000000528 statistical test Methods 0.000 claims description 2
- 238000011161 development Methods 0.000 description 5
- 230000018109 developmental process Effects 0.000 description 5
- 230000000875 corresponding effect Effects 0.000 description 4
- 238000012423 maintenance Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 208000037408 Device failure Diseases 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000001276 controlling effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- CIWBSHSKHKDKBQ-JLAZNSOCSA-N Ascorbic acid Chemical compound OC[C@H](O)[C@H]1OC(=O)C(O)=C1O CIWBSHSKHKDKBQ-JLAZNSOCSA-N 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000003631 expected effect Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
Images
Classifications
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
- G06Q10/06375—Prediction of business process outcome or impact based on a proposed change
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
-
- 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)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- Computing Systems (AREA)
- Educational Administration (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Operations Research (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
Abstract
The invention discloses a reliability estimation method for secondary equipment of a transformer substation, and belongs to the field of product reliability analysis. According to the method, an expanded fault rate data sample in an operation time period is obtained according to original fault rate statistical data and a substation secondary equipment fault rate prediction model; and estimating fault rate model parameters by using a least square method according to the extended samples, finally converting to obtain a time expression of the reliability of the secondary equipment of the transformer substation, and determining the reliable operation life of the device by a preset failure threshold. The method can analyze and predict the operation failure rate and the failure life of the secondary equipment of the transformer substation, and has the characteristics of high precision, high reliability, high speed and low power consumption.
Description
Technical Field
The invention relates to the technical field of reliability analysis of secondary equipment, in particular to a reliability estimation method for secondary equipment of a transformer substation.
Background
At present, the scale of the power system is continuously enlarged, the structural complexity is increasingly improved, and the power system is developed towards the ultra-high voltage direction. In order to meet the development needs and meet the high-quality requirements of people on electric power, the relay protection device is fastened to serve as a first defense line for safe and stable operation of the transformer substation, and the reliability of the relay protection device is particularly important to be accurately and rapidly estimated.
Under actual working conditions, environmental factors such as temperature and humidity, vibration and dust can influence the reliability of secondary equipment, and along with the increase of the operation years, the failure rate of a large number of electronic components in the device can be gradually increased, so that the residual service life of the device is lost. For a transformer substation system with a highly complex structure, accurate and effective reliability evaluation brings great convenience to maintenance and overhaul of secondary equipment, has great significance to safe and stable operation of the transformer substation, and is beneficial to improving the utilization rate of a device and reducing cost loss.
Mature reliability models exist for assessing life, but for high reliability relay protection systems, very little operational failure data presents a significant challenge for model parameter estimation. In recent years, students apply the prediction method in the machine learning field to the parameter prediction of the secondary equipment, and the evaluation effect and the accuracy are effectively improved. However, when the average failure rate data of the secondary equipment is processed, the time-related data is time-series data, and the BP neural network and the artificial neural network have some defects, such as slow learning speed and easy sinking into local minima, which greatly affect the prediction accuracy of the BP neural network and the artificial neural network. Also in the conventional neural network model, from the input layer to the hidden layer to the output layer, wherein the hidden layers are not connected, but the average failure rate time series data are not independent of each other.
The current output of a sequence in the recurrent neural network is correlated with the previous output, which provides the possibility to solve the predictions of the average failure rate data with time sequence. However, the calculation of the traditional cyclic neural network is related to all the calculation of the previous n times, which causes the calculation amount to be exponentially accumulated, so that the training time is long, which is obviously inapplicable to the relay protection device with extremely long rated life, and the improved long-term memory network (LSTM) based on the cyclic neural network solves the problem of long-term memory, and avoids the problem that all the calculation results of the previous n times participate in the calculation.
Disclosure of Invention
Aiming at the defects in the prior art, the technical problem to be solved by the invention is to provide a reliability estimation method for secondary equipment of a transformer substation with high accuracy and good robustness.
The technical scheme adopted by the invention for achieving the purpose is as follows: a reliability estimation method for secondary equipment of a transformer substation comprises the following steps:
acquiring original time sequence statistical data; the original time sequence statistical data is an operation index of a plurality of time points in the actual operation of the secondary equipment, which is counted according to the time sequence, wherein the operation index is the average failure rate of the secondary equipment under the operation time limit; the average failure rate is an average value of the number of failures of the secondary equipment of the transformer substation every year;
inputting the data of a plurality of continuous operation time points in the original time sequence statistical data into a substation secondary equipment failure rate prediction model to obtain average failure rates corresponding to the plurality of continuous operation time points;
obtaining an average fault rate data sample which is expanded by the secondary equipment of the transformer substation and accords with a fault distribution model rule according to actual statistical data of average fault rates corresponding to a plurality of continuous operation time points of the secondary equipment, carrying out parameter estimation on a fault rate distribution function by the average fault rate data sample to obtain reliability model parameters of the secondary equipment of the transformer substation, and obtaining the service life of the secondary equipment of the transformer substation by the reliability model of the secondary equipment of the transformer substation.
The construction method of the transformer substation secondary equipment failure rate prediction model comprises the following steps:
acquiring original time sequence statistical data, wherein the original time sequence statistical data is an operation index of a plurality of time points in actual operation of secondary equipment counted according to a time sequence, the operation index is an average failure rate of the secondary equipment under the operation time limit, and the average failure rate is an average value of the number of times of failure of the secondary equipment of a transformer substation every year;
training a long-period memory network based on the cyclic neural network improvement according to the original time sequence statistical data to obtain a fault rate prediction model of the secondary equipment of the transformer substation;
the input of the long-short-period memory network based on the cyclic neural network improvement is a plurality of continuous operation time points in the original time sequence statistical data, and the output of the long-short-period memory network based on the cyclic neural network improvement is an operation index corresponding to the plurality of continuous operation time points in the original time sequence statistical data.
The improved long-term and short-term memory network based on the cyclic neural network comprises the following components:
an input layer for acquiring the current time point test index x t ；
An implicit layer for testing the index x according to the current time point t Short-term memory h at previous time point t-1 And long-term memory c at the previous time point t-1 Obtaining the short-term memory h of the current time point t And long-term memory c at the current time point t ；
An output layer for memorizing the current time point in short term h t Output y as the current point in time t Outputting the current time point output y t Namely the later time point test index x t+1 。
The hidden layer comprises a plurality of hidden modules, each hidden module comprises four mutually interactive full-connection layers, and the four mutually interactive full-connection layers are respectively:
g t a layer for testing the index x for the current time point t And said previous point in time short term memory h t-1 Analyzing;
forgetting door, from f t A control, wherein the forgetting gate is used for controlling whether the long-term memory c of the previous time point is discarded or not t-1 Part of the content of (a);
input gate, consisting of i t Control, the input gate is used for screening the g t The analysis result of the layer is combined with the screening result and the forgetting result of the forgetting door to form the long-term memory c of the current time point t ；
Output gate, from o t A control for controlling whether to read and output the current time point long-term memory c t Is a part of the content of the file.
The fault rate prediction model of the secondary equipment of the transformer substation is realized by the following formula:
wherein sigma and tanh correspond to sigmod and tanh nonlinear activation functions in the neural network respectively; w corresponds to a weight coefficient matrix, and the four full connection layers test the index x relative to the current time point t The weight matrix of (2) is { W ] xi ,W xf ,W xo ,W xg -short-term memory h of said four fully connected layers with respect to said previous point in time t-1 The weight matrix of (2) is { W ] hi ,W hf ,W ho ,W hg }，c t-1 Long-term memory for the previous time point, h t-1 Is a short-term memory of the previous time point.
And performing error loss calculation on the output value and the theoretical value of the fault rate prediction model of the secondary equipment of the transformer substation, and adjusting the weight coefficient matrix W according to the error loss calculation result.
The original time sequence test data comprises training data and test data, wherein the training data is used for training the improved long-term and short-term memory network based on the cyclic neural network, and the test data is used for testing the accuracy of the fault rate prediction model of the secondary equipment of the transformer substation;
the testing data testing the accuracy of the fault rate prediction model of the secondary equipment of the transformer substation comprises the following steps:
inputting the test data into a fault rate prediction model of the secondary equipment of the transformer substation;
and comparing the output of the obtained substation secondary equipment failure rate prediction model with the original time sequence statistical test data to obtain the accuracy of the substation secondary equipment failure rate prediction model.
After obtaining the accuracy of the substation secondary equipment failure rate prediction model,
comparing the accuracy of the obtained substation secondary equipment failure rate prediction model with a preset accuracy;
and when the accuracy of the obtained substation secondary equipment failure rate prediction model is lower than the preset accuracy, training the improved long-term and short-term memory network based on the cyclic neural network again.
The failure rate distribution function is represented by the following formula:
wherein lambda (t) is a fault rate distribution function, t is time, alpha is a scale parameter, and beta is a shape parameter.
Carrying out logarithmic preprocessing on the average fault rate data sample before carrying out parameter estimation on the fault rate distribution function by the average fault rate data sample, and converting the parameter estimation of the fault rate distribution function into a linear form:
ln[λ(t)]＝m+nlnt (5)
wherein t is time, lambda (t) is a fault rate distribution function, and m and n are introduced simplification parameters.
The parameter estimation of the fault rate distribution function is realized by the following formula:
wherein alpha is the scale parameter of the fault rate distribution function, beta is the shape parameter of the fault rate distribution function, and m and n are the introduced simplified parameters.
The reliability model of the secondary equipment of the transformer substation is represented by the following formula:
the reliability model parameters alpha and beta are the fault rate distribution function parameters, t is time, and R (t) is the reliability of secondary equipment of the transformer substation.
The service life calculation method of the secondary equipment of the transformer substation comprises the following steps:
taking reliability R (t) of secondary equipment of transformer substation 0 ) =0.9, and the reliable life prediction result of the secondary device is t 0 And exits the run time as a recommendation for the substation secondary device.
The invention has the following advantages and beneficial effects:
1. the invention adopts a long-short-time memory network (LSTM) improved by a cyclic neural network to process and predict data, adopts the running time of a group of secondary equipment to construct a multidimensional input variable, predicts a multidimensional average failure rate matrix, carries out parameter estimation of a failure rate model by expanded failure rate data, and finally predicts the reliable service life of the secondary equipment. Compared with the traditional reliability evaluation method, the method effectively expands small data samples, achieves higher parameter estimation precision, has good robustness, and keeps the prediction error in a smaller range along with the change of the training data amount.
2. The method can effectively estimate the service life of the secondary equipment, and save the time and cost of operation, maintenance and overhaul; meanwhile, the prediction accuracy is high, the applicability is good, the quick and reliable service life estimation can be provided for secondary equipment products, the development period is shortened, the development iteration of the industry is accelerated, and the guarantee is provided for quick and wide application of secondary equipment.
3. Due to the structural characteristics of the long-short-period memory network, the reliability of the algorithm can be ensured in long-term real-time life prediction of the secondary equipment aging test, and the calculation training time consumption can not be greatly increased due to the accumulation of data quantity, so that the system prediction timeliness and accuracy are not affected.
Drawings
FIG. 1 is a schematic diagram of a long-term memory network architecture according to the present invention.
FIG. 2 is a schematic diagram of hidden layer cell modules in a long-short term memory network according to the present invention;
FIG. 3 is a schematic view of fault rate model parameter estimation for a preferred embodiment of the present invention;
fig. 4 is a reliability model and life prediction diagram of a preferred embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples.
The invention relates to a reliability estimation method of secondary equipment of a transformer substation, which can obtain a reliability function by a fault rate distribution function of determined parameters so as to predict the reliable service life of the secondary equipment, and specifically comprises the following steps:
in step S1, taking a 110kV substation as an example, according to the operation maintenance record of the protection device, average fault rate data of secondary equipment (including a relay protection device, a measurement and control device, an intelligent integrated device, etc.) of the substation at each operation time point is arranged as shown in table 1.
Table 1 secondary device failure rate statistics table
S2, carrying out standardized preprocessing on the statistical data of the original time sequence test data; and dividing the processed average fault rate data into training data and test data. The lifetime distribution of the secondary device is a continuous random variable, the reliability model adopts a two-parameter Weibull distribution, and the reliability function can be expressed as follows:
wherein t is time, alpha is a scale parameter, and beta is a shape parameter. Accordingly, the failure rate distribution function may be expressed as:
and (3) making:
λ(t)＝e m ·t n (8)
the invention carries out logarithmic processing on the average fault rate data and converts the formula (8) into a linear form:
ln[λ(t)]＝m+nlnt (5)
the training data and the test data both comprise an input matrix and an output matrix, wherein the input matrix X t Composed of lnt, output matrix Y t From ln [ lambda (t)]Construction, preferred embodiment of the invention input matrix X t Output matrix Y t The dimensions of (2) are all selected to be 4, namely four run-time matrices and corresponding average failure rate matrices respectively. Specific training data and test data are shown in table 2.
TABLE 2 Long-short term memory network training set and test set
And S3, designing an improved long-term and short-term memory network framework based on the recurrent neural network. As shown in fig. 1 and 2.
As in fig. 1, the hidden layer module of the long and short term memory network can be regarded as a cell (C t ) Its state is split into two vectors: h is a t And c t 。h t Is in a short-term state (output Y from this time t t Equal), c t Then long term memory is represented. c t The core of the operation is that: what information is read, what information is held, and what information is discarded in the long-term state of network learning. c t Running horizontally throughout above the hidden layer cells, less information is interacted and better is kept. Current input X t And short-term memory h at the previous time t-1 Input cell C t Then enter four interactive full connection layers g t The layer performs basic analysis of these two inputs, which in the standard RNN will then produce an output, the module terminates, and the cells of the long and short term memory network are designed with a further three-layer "gate" structure to choose to pass and process the information. Forgetting door (f) t Control) decides which long-term memories should be forgotten, input gate (i) t Control) decides g t Which contents of (a) should be added to the long-term memory, output gate (o t Control) determines which long term memories should be read and output. f (f) t 、i t 、o t The three terms are three fully connected layers with sigmod activation function, expressed as a function, and the input is the current time point test index x t And short-term memory h at the previous time point t-1 For calculating the values of the input gate, the forget gate and the output gate.
As shown in fig. 2, when long term memory c t-1 When passing through the neural network from left to right, the neural network firstly passes through a forgetting gate, and is formed by f t Control loses some memory and then screens memory with the input gate (i t And g t Control) is combined with the present information to produce a long-term memory result c t Is directly output. In addition, long-term memory c t Is copied and a tanh function is applied, while the current input X t And a short-term state h at the previous moment t-1 Filtering with an output gate, and mixing with tanh (c t ) Together generate the current short-term state h t I.e. the output Y at the current time t . The output process in fig. 2 can be represented by the following formula:
wherein sigma and tanh correspond to sigmod and tanh nonlinear activation functions in the neural network respectively; w corresponds to a weight coefficient matrix, specifically, { W xi ,W xf ,W xo ,W xg Four fully connected layers are related to input vector X t Weight matrix of { W } hi ,W hf ,W ho ,W hg Four full connection layers are related to short term memory h t-1 Is a weight matrix of (a). And calculating error loss of an output value (average fault rate at a certain time point) and a theoretical value of the fault rate prediction model of the secondary equipment of the transformer substation, and adjusting the weight coefficient matrix W according to the error loss calculation result.
And S4, determining the number of input layers, hidden layers and output layers, and training the long-term memory network by inputting a training set. In the preferred embodiment, the number of input layers of the long-short-period memory network is 7, the number of output layers is 3, and the number of hidden layers is 14 after repeated trial and error.
And S5, after training, applying the long-short-period memory network to the test set to verify the prediction accuracy. The predicted output of the test set was [0.0541,0.0633,0.1097,0.1726], the average error was 4.56% compared to the actual value of table 2, and the accuracy was good.
S6, inputting a plurality of groups of random operation time limit matrixes which are arranged in ascending order within a limited time range into the long-period memory network determined by the step, and generating an average failure rate prediction result of secondary equipment at the operation time; and (3) combining the predicted average failure rate data of the secondary equipment with the actual statistical data in the step (S1) to obtain an extended average failure rate data sample of the secondary equipment conforming to the failure distribution model rule, wherein the average failure rate data sample is shown in a table 3.
Table 3 secondary device failure rate expansion statistics table
S7, parameter estimation is carried out on the fault rate distribution function by the extended average fault rate data; and combining the reliability function according to the reliability experience threshold value to obtain a secondary equipment life prediction value. As shown in fig. 3, the secondary equipment expansion failure rate data in table 3 is subjected to logarithmic processing, and is represented in the figure, and parameters m and n in the formula (5) are estimated by adopting a least square method, so as to obtain m= -8.3319 and n= 1.8119; next, α= 27.9600, β= 2.8119 can be obtained according to formula (6).
Thus, a specific expression of the reliability function is obtained:
as shown in FIG. 4, the reliability of the relay protection device is 0.9, R (t 0 ) =0.9, the reliable lifetime prediction result of the secondary device is t 0 = 12.559, whereby the lifetime can determine that the proposed exit run time of the relay protection device is 12.559 years. The method can effectively estimate the service life of the secondary equipment and save the time and cost of operation, maintenance and overhaul; simultaneous pre-emptionThe method has the advantages of high measurement accuracy and good applicability, can provide quick and reliable life estimation for secondary equipment products, shortens development period, accelerates development iteration of industry, and provides guarantee for quick and wide application of secondary equipment. Particularly, on the premise that only the early-stage average fault rate data exists, the method can still expand the later-stage fault rate data to realize effective prediction, and the service life of the relay protection device is acquired in advance, so that the expected effect of the prediction behavior is achieved.
The foregoing is merely a preferred embodiment of the present invention, and it should be noted that modifications and variations could be made by those skilled in the art without departing from the technical principles of the present invention, and such modifications and variations should also be regarded as being within the scope of the invention.
Claims (12)
1. The reliability estimation method for the secondary equipment of the transformer substation is characterized by comprising the following steps of:
acquiring original time sequence statistical data; the original time sequence statistical data is an operation index of a plurality of time points in the actual operation of the secondary equipment, which is counted according to the time sequence, wherein the operation index is the average failure rate of the secondary equipment under the operation time limit; the average failure rate is an average value of the number of failures of the secondary equipment of the transformer substation every year;
inputting the data of a plurality of continuous operation time points in the original time sequence statistical data into a substation secondary equipment failure rate prediction model to obtain average failure rates corresponding to the plurality of continuous operation time points;
obtaining an average fault rate data sample which is expanded by the secondary equipment of the transformer substation and accords with a failure distribution model rule according to actual statistical data of average fault rates corresponding to a plurality of continuous operation time points of the secondary equipment, carrying out parameter estimation on a fault rate distribution function by the average fault rate data sample to obtain reliability model parameters of the secondary equipment of the transformer substation, and obtaining the service life of the secondary equipment of the transformer substation by a reliability model of the secondary equipment of the transformer substation.
2. The method for estimating reliability of secondary equipment of a transformer substation according to claim 1, wherein the method for constructing the failure rate prediction model of the secondary equipment of the transformer substation is as follows:
acquiring original time sequence statistical data, wherein the original time sequence statistical data is an operation index of a plurality of time points in actual operation of secondary equipment counted according to a time sequence, the operation index is an average failure rate of the secondary equipment under the operation time limit, and the average failure rate is an average value of the number of times of failure of the secondary equipment of a transformer substation every year;
training a long-period memory network based on the cyclic neural network improvement according to the original time sequence statistical data to obtain a fault rate prediction model of the secondary equipment of the transformer substation;
the input of the long-short-period memory network based on the cyclic neural network improvement is a plurality of continuous operation time points in the original time sequence statistical data, and the output of the long-short-period memory network based on the cyclic neural network improvement is an operation index corresponding to the plurality of continuous operation time points in the original time sequence statistical data.
3. The method for estimating reliability of secondary equipment of a transformer substation according to claim 2, wherein the long-term and short-term memory network based on the cyclic neural network improvement comprises:
an input layer for acquiring the current time point test index x t ；
An implicit layer for testing the index x according to the current time point t Short-term memory h at previous time point t-1 And long-term memory c at the previous time point t-1 Obtaining the short-term memory h of the current time point t And long-term memory c at the current time point t ；
An output layer for memorizing the current time point in short term h t Output y as the current point in time t Outputting the current time point output y t Namely the later time point test index x t+1 。
4. A method for estimating reliability of secondary equipment of a substation according to claim 3, wherein the hidden layer comprises a plurality of hidden modules, each hidden module comprises four mutually interactive fully connected layers, and the four mutually interactive fully connected layers are respectively:
g t a layer for testing the index x for the current time point t And said previous point in time short term memory h t-1 Analyzing;
forgetting door, from f t A control, wherein the forgetting gate is used for controlling whether the long-term memory c of the previous time point is discarded or not t-1 Part of the content of (a);
input gate, consisting of i t Control, the input gate is used for screening the g t The analysis result of the layer is combined with the screening result and the forgetting result of the forgetting door to form the long-term memory c of the current time point t ；
Output gate, from o t A control for controlling whether to read and output the current time point long-term memory c t Is a part of the content of the file.
5. The method for estimating reliability of secondary equipment of a transformer substation according to claim 4, wherein the failure rate prediction model of the secondary equipment of the transformer substation is implemented by the following formula:
wherein sigma and tanh correspond to sigmod and tanh nonlinear activation functions in the neural network respectively; w corresponds to the weightA coefficient matrix, the four full connection layers test the index x relative to the current time point t The weight matrix of (2) is { W ] xi ,W xf ,W xo ,W xg -short-term memory h of said four fully connected layers with respect to said previous point in time t-1 The weight matrix of (2) is { W ] hi ,W hf ,W ho ,W hg }，c t-1 Long-term memory for the previous time point, h t-1 Is a short-term memory of the previous time point.
6. The method for estimating reliability of secondary equipment of a transformer substation according to claim 5, further comprising performing error loss calculation on an output value and a theoretical value of a failure rate prediction model of the secondary equipment of the transformer substation, and adjusting the weight coefficient matrix W according to the error loss calculation result.
7. The substation secondary equipment reliability estimation method according to claim 2, wherein the original time series test data comprises training data and test data, the training data is used for training the improved long-term and short-term memory network based on the cyclic neural network, and the test data is used for testing the accuracy of the substation secondary equipment failure rate prediction model;
the testing data testing the accuracy of the fault rate prediction model of the secondary equipment of the transformer substation comprises the following steps:
inputting the test data into a fault rate prediction model of the secondary equipment of the transformer substation;
and comparing the output of the obtained substation secondary equipment failure rate prediction model with the original time sequence statistical test data to obtain the accuracy of the substation secondary equipment failure rate prediction model.
8. The method for estimating reliability of secondary equipment of a transformer substation according to claim 7, further comprising, after obtaining accuracy of the failure rate prediction model of the secondary equipment of the transformer substation,
comparing the accuracy of the obtained substation secondary equipment failure rate prediction model with a preset accuracy;
and when the accuracy of the obtained substation secondary equipment failure rate prediction model is lower than the preset accuracy, training the improved long-term and short-term memory network based on the cyclic neural network again.
9. The substation secondary device reliability estimation method according to claim 1, wherein the failure rate distribution function is represented by the following formula:
wherein lambda (t) is a fault rate distribution function, t is time, alpha is a scale parameter, and beta is a shape parameter.
10. The method for estimating reliability of secondary equipment of a transformer substation according to claim 9, wherein, before the step of performing parameter estimation on the fault rate distribution function by the average fault rate data sample, performing logarithmic preprocessing on the average fault rate data sample, and converting the parameter estimation of the fault rate distribution function into a linear form:
ln[λ(t)]＝m+nlnt (5)
wherein t is time, lambda (t) is a fault rate distribution function, and m and n are introduced simplification parameters.
The parameter estimation of the fault rate distribution function is realized by the following formula:
wherein alpha is the scale parameter of the fault rate distribution function, beta is the shape parameter of the fault rate distribution function, and m and n are the introduced simplified parameters.
11. The method of reliability estimation of secondary equipment of a substation according to claim 1, wherein the reliability model of the secondary equipment of the substation is represented by the following formula:
the reliability model parameters alpha and beta are the fault rate distribution function parameters, t is time, and R (t) is the reliability of secondary equipment of the transformer substation.
12. The method for estimating reliability of secondary equipment of a transformer substation according to claim 11, wherein the method for calculating life of the secondary equipment of the transformer substation is as follows:
taking reliability R (t) of secondary equipment of transformer substation 0 ) =0.9, and the reliable life prediction result of the secondary device is t 0 And exits the run time as a recommendation for the substation secondary device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310306309.2A CN116307773A (en) | 2023-03-27 | 2023-03-27 | Reliability estimation method for secondary equipment of transformer substation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310306309.2A CN116307773A (en) | 2023-03-27 | 2023-03-27 | Reliability estimation method for secondary equipment of transformer substation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116307773A true CN116307773A (en) | 2023-06-23 |
Family
ID=86799480
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310306309.2A Pending CN116307773A (en) | 2023-03-27 | 2023-03-27 | Reliability estimation method for secondary equipment of transformer substation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116307773A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117540618A (en) * | 2023-09-28 | 2024-02-09 | 中国长江电力股份有限公司 | Relay protection device fault rate prediction method based on HS-LSTM algorithm |
-
2023
- 2023-03-27 CN CN202310306309.2A patent/CN116307773A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117540618A (en) * | 2023-09-28 | 2024-02-09 | 中国长江电力股份有限公司 | Relay protection device fault rate prediction method based on HS-LSTM algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112838946B (en) | Method for constructing intelligent sensing and early warning model based on communication network faults | |
| CN109214708B (en) | Electric power system risk assessment method based on cross entropy theory optimization support vector machine | |
| CN110598854A (en) | GRU model-based transformer area line loss rate prediction method | |
| CN113516310B (en) | Transformer fault early warning method and system | |
| CN111798095A (en) | Power cable state evaluation method based on neural network | |
| CN114707712A (en) | Method for predicting requirement of generator set spare parts | |
| CN116307773A (en) | Reliability estimation method for secondary equipment of transformer substation | |
| CN110674984A (en) | Tri-Training-Lasso-BP network-based static voltage stability margin prediction method | |
| CN112734209A (en) | Risk early warning method and system for safety and stability control of load-intensive urban power grid | |
| CN111210049A (en) | LSTM-based method for predicting degradation trend of lead-acid valve-controlled storage battery of transformer substation | |
| CN113791351B (en) | Lithium battery life prediction method based on transfer learning and difference probability distribution | |
| CN115438897A (en) | Industrial process product quality prediction method based on BLSTM neural network | |
| CN109412152B (en) | Power grid loss calculation method based on deep learning and elastic network regularization | |
| CN111061708A (en) | Electric energy prediction and restoration method based on LSTM neural network | |
| CN111241629A (en) | Intelligent prediction method for performance change trend of airplane hydraulic pump based on data driving | |
| CN114638421A (en) | Method for predicting requirement of generator set spare parts | |
| CN111935762B (en) | Distribution network fault diagnosis method and system based on EWT and CNN under 5G load-bearing network | |
| CN115564310A (en) | Reliability evaluation method for new energy power system based on convolutional neural network | |
| CN114861969A (en) | LSTM-based electric power system oscillation risk prediction method | |
| CN115062533A (en) | Construction and prediction method and medium of multiple regression prediction model | |
| CN113780356A (en) | Water quality prediction method and system based on ensemble learning model | |
| CN114780609A (en) | Regional carbon emission calculation method based on electric power big data and time series | |
| CN114198266A (en) | Parallel monitoring method and device for running state of wind turbine generator | |
| CN114510469A (en) | Method, device, equipment and medium for identifying bad data of power system | |
| CN113570129A (en) | Method for predicting strip steel pickling concentration and computer readable storage medium |
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 |