CN110728395A - Main transformer short-term power load calculation method and device, computer and storage medium - Google Patents
Main transformer short-term power load calculation method and device, computer and storage medium Download PDFInfo
- Publication number
- CN110728395A CN110728395A CN201910835477.4A CN201910835477A CN110728395A CN 110728395 A CN110728395 A CN 110728395A CN 201910835477 A CN201910835477 A CN 201910835477A CN 110728395 A CN110728395 A CN 110728395A
- Authority
- CN
- China
- Prior art keywords
- load data
- data
- cluster center
- sample load
- clustering
- 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
- 238000004364 calculation method Methods 0.000 title claims abstract description 38
- 238000000034 method Methods 0.000 claims abstract description 62
- 238000004422 calculation algorithm Methods 0.000 claims description 32
- 238000004590 computer program Methods 0.000 claims description 23
- 230000008569 process Effects 0.000 claims description 23
- 238000012545 processing Methods 0.000 claims description 22
- 238000004140 cleaning Methods 0.000 claims description 20
- 238000013507 mapping Methods 0.000 claims description 17
- 230000004931 aggregating effect Effects 0.000 claims description 11
- 230000006870 function Effects 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 238000005259 measurement Methods 0.000 description 8
- 238000003064 k means clustering Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000007405 data analysis Methods 0.000 description 3
- 230000002776 aggregation Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 238000013075 data extraction Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000008676 import Effects 0.000 description 2
- 230000001932 seasonal effect Effects 0.000 description 2
- 238000011524 similarity measure Methods 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000013506 data mapping Methods 0.000 description 1
- 238000013213 extrapolation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 230000003068 static effect Effects 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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Development Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Game Theory and Decision Science (AREA)
- Evolutionary Biology (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Educational Administration (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Primary Health Care (AREA)
- Probability & Statistics with Applications (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application relates to a main transformer short-term power load calculation method, a main transformer short-term power load calculation device, computer equipment and a storage medium. The method comprises the following steps: acquiring current load data within a preset time before the current moment; acquiring a plurality of pre-stored cluster center curves formed by clustering historical data; comparing the current load data with the central curves of all clusters, and calculating the Euclidean distance between the current load data and the central curves of all clusters; determining a cluster center curve which is most matched with the current load data according to the calculated Euclidean distances; and obtaining short-term power load data of the main transformer according to the determined most matched cluster center curve. The short-term data of the load of the main transformer can be calculated, external data such as weather and the like are not depended on, the accuracy of the calculated data is high, and the short-term data can be efficiently calculated.
Description
Technical Field
The application relates to the technical field of main transformer short-term power load calculation, in particular to a main transformer short-term power load calculation method, a main transformer short-term power load calculation device, computer equipment and a storage medium.
Background
In power plants and substations, transformers for delivering power to an electrical power system or to a customer are called main transformers, or simply main substations. The load of the main transformer is affected by periodic weather, holidays and the like, the number of terminal users is small, the power load changes are complex, all data cannot be guaranteed to be accurate in the whole process of collecting, transmitting and exchanging the power load measurement data, and all factors cannot be considered completely when load prediction is carried out.
The conventional prediction methods include trend extrapolation, time series method, regression analysis method, grey mathematical theory, expert system method, fuzzy load prediction and the like. The method has very high accuracy on the load prediction of the bus, but the load of the main transformer is difficult to build an accurate model by adopting a common prediction method because the change rule is complex, and the load of the main transformer cannot be accurately calculated.
Disclosure of Invention
In view of the above, it is necessary to provide a method, an apparatus, a computer device, and a storage medium for calculating a short-term power load capable of changing a main transformer.
A method of primary transformer short term electrical load calculation, the method comprising:
acquiring current load data within a preset time before the current moment;
acquiring a plurality of pre-stored cluster center curves formed by clustering historical data;
comparing the current load data with each cluster center curve, and calculating the Euclidean distance between the current load data and each cluster center curve;
determining the cluster center curve which is most matched with the current load data according to the calculated Euclidean distances;
and obtaining short-term power load data of the main transformer according to the determined most matched cluster center curve.
In one embodiment, the step of obtaining a plurality of pre-stored cluster center curves clustered by historical data comprises:
acquiring a plurality of sample load data;
clustering the sample load data to obtain a plurality of cluster center curves;
and storing each cluster center curve.
In one embodiment, the step of clustering each sample load data includes:
calculating to obtain an initial cluster center of each sample load data by adopting an extreme distance vertex algorithm;
and performing iterative clustering on the data according to a K-means algorithm determined by the initial cluster center to obtain a plurality of cluster center curves.
In one embodiment, the step of clustering each sample load data to obtain a plurality of cluster center curves further includes:
and cleaning each sample load data, and removing burr data in each sample load data to obtain the cleaned sample load data.
In one embodiment, after the step of performing the cleaning process on each sample load data, removing the burr data in each sample load data, and obtaining the cleaned sample load data, the method further includes:
and aggregating the cleaned sample load data according to a preset rule to obtain the aggregated sample load data.
In one embodiment, the step of aggregating the sample load data after the cleaning process according to a preset rule to obtain the aggregated sample load data further includes:
performing interval mapping on the aggregated sample load data,
the step of clustering each sample load data to obtain a plurality of cluster center curves comprises:
and clustering the sample load data subjected to interval mapping to obtain a plurality of cluster center curves.
A main transformer short-term electrical load computing device, the device comprising:
the current load data acquisition module is used for acquiring current load data within preset time before the current moment;
the cluster center curve acquisition module is used for acquiring a plurality of cluster center curves which are pre-stored and formed by clustering historical data;
the Euclidean distance calculation module is used for comparing the current load data with each cluster center curve and calculating the Euclidean distance between the current load data and each cluster center curve;
a cluster center determining module, configured to determine, according to the calculated euclidean distances, the cluster center curve that is most matched with the current load data;
and the short-term power load number acquisition module is used for acquiring the short-term power load data of the main transformer according to the determined most matched cluster center curve.
In one embodiment, the method further comprises the following steps:
the system comprises a sample load data acquisition module, a data processing module and a data processing module, wherein the sample load data acquisition module is used for acquiring a plurality of sample load data;
the cluster processing module is used for carrying out cluster processing on the sample load data to obtain a plurality of cluster center curves;
and the cluster center curve storage module is used for storing each cluster center curve.
A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the following steps when executing the computer program:
acquiring current load data within a preset time before the current moment;
acquiring a plurality of pre-stored cluster center curves formed by clustering historical data;
comparing the current load data with each cluster center curve, and calculating the Euclidean distance between the current load data and each cluster center curve;
determining the cluster center curve which is most matched with the current load data according to the calculated Euclidean distances;
and obtaining short-term power load data of the main transformer according to the determined most matched cluster center curve.
A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
acquiring current load data within a preset time before the current moment;
acquiring a plurality of pre-stored cluster center curves formed by clustering historical data;
comparing the current load data with each cluster center curve, and calculating the Euclidean distance between the current load data and each cluster center curve;
determining the cluster center curve which is most matched with the current load data according to the calculated Euclidean distances;
and obtaining short-term power load data of the main transformer according to the determined most matched cluster center curve.
According to the main transformer short-term power load calculation method, the main transformer short-term power load calculation device, the computer equipment and the storage medium, the Euclidean distance between the current load data and the cluster center curve is calculated by comparing the current load data with the cluster center curve, the cluster center curve which is most matched with the current load data is selected, and therefore the main transformer short-term power load data is obtained according to the selected cluster center curve, calculation of the short-term data of the load of the main transformer is achieved, external data such as weather are not relied on, the calculated data are high in accuracy, and the short-term data can be efficiently calculated.
Drawings
FIG. 1 is a diagram of an exemplary embodiment of a method for calculating a short-term electrical load of a primary substation;
FIG. 2 is a schematic flow chart illustrating a method for calculating the short-term electrical load of the main transformer according to an embodiment;
FIG. 3 is a block diagram of a main transformer short-term electrical load computing device according to an embodiment;
FIG. 4 is a diagram illustrating an internal structure of a computer device according to an embodiment;
fig. 5A is a schematic diagram illustrating sample load data after interval mapping in an implementation process of the main transformer short-term power load calculation method according to an embodiment;
fig. 5B and 5C are schematic diagrams of sample load data subjected to clustering processing in the implementation process of the main transformer short-term power load calculation method in one embodiment;
fig. 5D and 5E are schematic diagrams illustrating the adjustment of the cluster number by using an elbow criterion algorithm in the implementation process of the main transformer short-term power load calculation method according to an embodiment;
fig. 5F is a schematic diagram of a calculation result in an implementation process of the main transformer short-term power load calculation method in one embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The main transformer short-term power load calculation method can be applied to the application environment shown in fig. 1. The terminal 102 is connected to the server 104 through a network, and is connected to communicate with the server 104. The terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices, and the server 104 may be implemented by an independent server or a server cluster formed by a plurality of servers. The server 104 stores a plurality of cluster center curves formed by clustering historical data, the terminal 102 acquires current load data in a preset time before the current time, the current load data in the preset time before the current time is sent to the server 104, and the server 104 acquires and acquires a plurality of pre-stored cluster center curves formed by clustering historical data; comparing the current load data with each cluster center curve, and calculating the Euclidean distance between the current load data and each cluster center curve; determining the cluster center curve which is most matched with the current load data according to the calculated Euclidean distances; and obtaining short-term power load data of the main transformer according to the determined most matched cluster center curve.
In an embodiment, as shown in fig. 2, a method for calculating a short-term power load of a main transformer is provided, which is described by taking the application scenario in fig. 1 as an example, and includes the following steps:
Specifically, the preset time before the current time refers to a period of time before the current time, and the preset time is a preset time length. In this embodiment, load data several hours before the current time is collected, for example, power load data four hours before the current time is collected. In addition, the current load data is the current power load data of the main transformer, and in this embodiment, the current power load data of the main transformer is acquired. In the present embodiment, the first and second electrodes are,
Specifically, a plurality of cluster center curves are obtained by clustering historical data and are pre-stored in a memory. The historical data is historical power load data, and may also be referred to as historical load data. The cluster center curve is used for representing the load change and the change trend of the main transformer in the historical period.
Specifically, in this step, the current load data is compared with the data of the time period corresponding to each cluster center curve, and the euclidean distance between the current load data and each cluster center curve is calculated. In this embodiment, because the current load data is acquired within the preset time of the current time, the current load data has a corresponding time period, and the euclidean distance between the current load data and each cluster center curve can be calculated by comparing the cluster center curve of the corresponding time period with the current load data. According to the current time and the preset time, the time period corresponding to the cluster center curve is obtained, the current load data is compared with the data of the time period corresponding to each cluster center curve, and the Euclidean distance between the current load data and each cluster center curve is calculated.
For example, the current time is PM 18: 00, the current load data acquisition time period is PM 14: 00-18: 00, then the corresponding time period of the cluster center curve is also PM 14: 00-18: 00, PM 14: 00-18: and comparing the data of the cluster center curve in the 00 time period with the current load data, and calculating to obtain the Euclidean distance.
Specifically, each cluster center curve is calculated to obtain one euclidean distance, and therefore, a plurality of cluster center curves are calculated to obtain a plurality of euclidean distances, and each cluster center curve corresponds to one euclidean distance. The Euclidean distance can be used to determine the cluster center curve with the highest matching degree with the current load data, and it is worth mentioning that the best matching means the highest matching degree or the best matching. In one embodiment, the cluster center curve corresponding to the minimum euclidean distance is the cluster center curve that most matches the current load data.
And step 250, obtaining short-term power load data of the main transformer according to the determined most matched cluster center curve.
In this embodiment, the most matched cluster center curve is determined, and short-term power load data of the main transformer after the current time can be obtained according to the most matched cluster center curve. Specifically, the short-term power load data of the main transformer can be obtained according to a cluster center curve, and specifically, the short-term power load data of the main transformer can be obtained according to data on the cluster center curve corresponding to a short-term time node, wherein the data on the cluster center curve corresponding to the short-term time node is the short-term power load data of the main transformer. The cluster center curve is formed by clustering historical data, so that the cluster center curve comprises a plurality of historical data, the power load data of the main transformer in a period of time after the current moment can correspond to the historical data of the cluster center curve in the period of time, and the historical data in the period of time is the short-term power load data of the main transformer.
According to the current time and second preset time, a second time period corresponding to the cluster center curve is obtained, and data in the second time period of the determined best-matched cluster center curve is obtained to serve as short-term power load data of the main transformer. Therefore, short-term power load data of the main transformer can be accurately obtained.
In the embodiment, the Euclidean distance between the current load data and the cluster center curve is obtained through comparison between the current load data and the cluster center curve, the cluster center curve which is most matched with the current load data is selected, so that the short-term power load data of the main transformer is obtained according to the selected cluster center curve, the calculation of the short-term data of the load of the main transformer is realized, the calculation is independent of external data such as weather, the accuracy of the calculated data is high, and the short-term data can be obtained through efficient calculation.
In one embodiment, the step of obtaining a plurality of pre-stored cluster center curves clustered by historical data comprises: acquiring a plurality of sample load data; clustering the sample load data to obtain a plurality of cluster center curves; and storing each cluster center curve.
Specifically, the sample load data is sample power load data, which may also be referred to as historical load data or historical power load data, the sample load data is collected and stored in advance before the current time, and a large amount of sample load data is obtained, and the greater the data amount of the sample load data is, the higher the accuracy of comparison is. In this embodiment, each sample load data is clustered, so as to obtain a plurality of cluster center curves, and each cluster center curve is stored.
In one embodiment, the step of clustering each sample load data includes: calculating to obtain an initial cluster center of each sample load data by adopting an extreme distance vertex algorithm; and performing iterative clustering on the data according to a K-means algorithm determined by the initial cluster center to obtain a plurality of cluster center curves.
Specifically, during clustering, an extremely-far vertex algorithm is firstly adopted to determine an initial cluster center, namely the initial cluster center. Taking clustering into K clusters as an example, K pieces of data can be found in all data by the extreme vertex, the sum of the distances between each point is larger than the sum of the distances of any other K pieces of data, and the maximum difference of the centers of each initial cluster can be ensured. And then carrying out iterative clustering on the data by adopting a K-means algorithm (K-means clustering algorithm) specified by the initial cluster center until convergence. The K-means algorithm is a hard clustering algorithm and is a typical representation of an objective function clustering method based on a prototype, a certain distance from a data point to the prototype is used as an optimized objective function, an adjustment rule of iterative operation is obtained by using a function extremum solving method, the Euclidean distance is used as a similarity measure, the optimal classification of a vector V corresponding to a certain initial clustering center is solved, the evaluation index J is enabled to be minimum, and a square error sum criterion function is used as a clustering criterion function.
In the embodiment, an elbow criterion is adopted, the sum of the square roots of errors from each data to the center of each cluster is calculated according to different cluster numbers, a cluster number K exists, and the condition that the difference values of the sum of the square roots of the errors are within a set threshold range when the clusters are clustered into K + 1-K +5 clusters is met. The cluster number K is the number of clusters to be clustered finally, and the corresponding cluster center curve is the final result generated by the data analysis model.
In one embodiment, the step of clustering each sample load data to obtain a plurality of cluster center curves further includes: and cleaning each sample load data, and removing burr data in each sample load data to obtain the cleaned sample load data.
Specifically, the sample load data is cleaned, and the sample load data with incomplete data, which is the glitch data, is removed, for example, data with zero value, null value, and significantly too high numerical value, which is the glitch data.
In one embodiment, after the step of performing the cleaning process on each sample load data, and removing the burr data in each sample load data to obtain the cleaned sample load data, the method further includes: and extracting the cleaned sample load data at intervals, and extracting a piece of sample load data at intervals of a third preset time to obtain the sampled sample load data.
In this embodiment, the data cleaning process needs to process the garbage data in the data, such as the glitch data with zero value, null value, and obviously too high numerical value. Data extraction is done, if one measurement point every 15 minutes is required, while actual data is one measurement point every 15 minutes, from which one piece of data needs to be extracted every 15 minutes.
In one embodiment, after the step of performing the cleaning process on each sample load data, removing the burr data in each sample load data, and obtaining the cleaned sample load data, the method further includes: and aggregating the cleaned sample load data according to a preset rule to obtain the aggregated sample load data.
Specifically, the aggregating is grouping, in this embodiment, the sample load data after the cleaning processing is aggregated according to a preset rule, that is, the sample load data after the cleaning processing is grouped according to the preset rule, so as to obtain multiple groups of sample load data. The preset rule is a grouping rule, and the preset rule may be time or date, for example, the sample load data after the cleaning process is grouped by day to obtain the grouped sample load data.
In this embodiment, the cleaned data are grouped according to day, the start time and the end time of the grouping are 24 hours, such as 8: 00-07: 45 days, the number of samples of the sample load data in each day is guaranteed to be the same after the grouping, and if there is missing data, the sample load data in the day is discarded.
In one embodiment, the step of aggregating the sample load data after the cleaning process according to a preset rule to obtain the aggregated sample load data further includes: carrying out interval mapping on the aggregated sample load data; the step of clustering each sample load data to obtain a plurality of cluster center curves comprises: and clustering the sample load data subjected to interval mapping to obtain a plurality of cluster center curves.
In particular, interval mapping refers to scaling up sample load data over a period to a larger period. It is worth mentioning that the natural increase of the load and the periodicity of seasonal variation are considered, the total load of the data in the latest chronology year is taken as a reference standard, and other historical data are correspondingly amplified in a comparation mode according to chronology year grouping.
The following is one example:
in this embodiment, the implementation process includes offline learning and online decision, where the offline learning is to obtain a training model including a plurality of cluster center curves through pre-training, and the online decision is to determine a cluster center curve through comparison between current load data and the cluster center curve, so as to obtain short-term circuit load data of the main transformer.
First, offline learning
The off-line learning is to construct a main transformer short-term power load prediction model and complete the processes of data import, processing, analysis and the like.
1. Data import
The historical data can be imported from database tables or variable files (such as txt, cvs and the like). When a variable file is imported, the file needs to be standardized, for example, a header, an empty line, and a repeat line are removed, so that the data is structured and usable.
2. Data cleansing
The data cleaning link needs to process the garbage data in the data, such as the burr data with zero value, null value and obviously too high numerical value. Data extraction is done, if one measurement point every 15 minutes is required, while actual data is one measurement point every 15 minutes, from which one piece of data needs to be extracted every 15 minutes.
3. Aggregating data by day
Grouping the cleaned data according to days, wherein the starting time and the ending time of the grouping are 24 hours, such as 8: 00-07: 45 days, ensuring that the number of data points in each day is the same after grouping, and discarding the data of the day if missing data exists.
4. Interval mapping
And (4) correspondingly amplifying other historical data in a chronologic year grouping mode in a comparably way by taking the total load of the data of the latest chronologic year as a reference standard in consideration of the natural increase of the load and the periodicity of seasonal variation.
5. Clustering
During clustering, an extremely-far vertex algorithm is firstly adopted to determine the initial cluster center. Taking clustering into K clusters as an example, K pieces of data can be found in all data by the extreme vertex, the sum of the distances between all points is larger than the sum of the distances of any other K pieces of data, and the maximization of the initial center difference of all clusters can be ensured. And then, carrying out iterative clustering on the data by adopting a K-means algorithm specified by the initial cluster center until convergence. The K-means algorithm is a hard clustering algorithm and is a typical representation of an objective function clustering method based on a prototype, a certain distance from a data point to the prototype is used as an optimized objective function, an adjustment rule of iterative operation is obtained by using a function extremum solving method, the Euclidean distance is used as a similarity measure, the optimal classification of a vector V corresponding to a certain initial clustering center is solved, the evaluation index J is enabled to be minimum, and a square error sum criterion function is used as a clustering criterion function.
6. Cluster center
The cluster center can be obtained by clustering, but the elbow criterion is introduced because the clustering into a plurality of clusters is difficult to judge. And calculating the sum of the square roots of errors from each data to the center of each cluster according to different cluster numbers, wherein one cluster number K exists, and the difference values of the sum of the square roots of errors are within a set threshold range when the data are clustered into K + 1-K +5 clusters. The cluster number K is the number of clusters to be clustered finally, and the corresponding cluster center is the final result generated by the data analysis model.
7. Storing
And (4) persisting the processing result and storing the processing result into a cache server or a memory so as to be used in online decision making. When the data is stored, a result is required to be stored in a variable file so as to be used in association with the power dispatching robot data.
Subsequently, making an online decision
And comparing the real-time data of the previous four hours of the current moment with the corresponding time periods of the cluster centers stored in the cache server, and selecting the best matched cluster center curve from the real-time data by calculating the Euclidean distances of the corresponding points, so that the cluster center can be used as a main transformer short-term power load prediction result.
1. Predicting the main transformer short-term power load based on an improved K-means clustering algorithm: the patent discloses a main transformer short-term power load prediction method applying a clustering machine algorithm. The clustering algorithm is an improved K-means clustering algorithm, and the improvement is that the initial cluster center is not randomly assigned and is calculated according to the extreme vertex. The cluster family number of the main transformer short-term power load prediction is calculated through an elbow criterion algorithm and is not manually specified. The main transformer short-term power load prediction model has the capabilities of off-line learning, on-line decision and model updating.
2. The method for determining the power load abnormal data initial cluster center based on the extreme distance vertex algorithm comprises the following steps: clustering analysis is needed after historical data are grouped, for example, clustering is carried out to form K clusters, K pieces of data can be found in all data by the extreme vertex, the sum of distances between all points is larger than that of any other K pieces of data, and the maximization of the initial center difference of each cluster can be ensured. Therefore, the K pieces of data with the largest difference are selected initially, and each cluster of subsequent clustering results can have better convergence.
3. The method for determining the predicted cluster number of the main transformer short-term power load based on the elbow criterion algorithm comprises the following steps: the cluster center can be obtained through clustering, and is regarded as a characteristic curve of historical power load measurement data, and due to the influences of factors such as weather, holidays, double holidays, seasons and the like, the cluster center is difficult to artificially predict to have several types. An elbow criterion algorithm may solve the problem. And calculating the sum of the square roots of errors from each piece of data to the center of each cluster according to different cluster numbers, wherein one cluster number K exists, and the difference values of the sum of the square roots of errors are within a set threshold range when the data are clustered into K + 1-K +5 clusters. The cluster number K is the number of clusters to be clustered finally, and the corresponding cluster center is the final result generated by the data analysis model.
4. The self-learning capability of the main transformer short-term power load prediction: the method comprises the following steps of having offline learning capability, online decision-making capability and model updating capability: the off-line learning capability refers to a process of obtaining a cluster center, namely power load characteristics, by carrying out data standardization, data cleaning, data aggregation, interval mapping, extreme vertex calculation, K-means clustering and elbow criterion calculation on a large amount of power load historical data. The online decision is a process for predicting the short-term power load of the main transformer by using real-time data and power load characteristics. And the model updating means that the data accumulated in the online decision process is added into historical data periodically and automatically, and the power load characteristics are acquired again.
The following is a specific example:
the target is as follows: predicting short-term power load of a main transformer 2019.02.2112: 00-15: 45.
Constraint conditions are as follows:
the main transformer power load historical data has enough time span, and the optimal time span is 3 years and is at least 1 year.
The method comprises the following steps:
1. 2019.02.21, making preliminary judgment to find out the point of missing measurement data, which can be determined as collection error or missing point;
2. filling up missing points, such as adopting median values of adjacent points;
3. for data prior to 2019.02.21, such as data of two years, the inter-region mapping adjustment is performed and grouped after considering the load increase. As shown in fig. 5A.
4. The waveforms for all data were clustered by day (using a modified Kmeans algorithm, i.e. using the very far vertices to determine the initial cluster center). As shown in fig. 5B and 5C.
5. And adjusting the number of the clustered clusters by using an elbow criterion algorithm to obtain a central curve of each cluster. As shown in fig. 5D and 5E.
6. The method comprises the steps of taking load history measurement data of the main transformer in the first 4 hours of a load starting point to be predicted, namely data in 2019.02.2108: 00-11: 45 time periods, comparing the data in the time periods with curves of corresponding time periods (namely 08: 00-11: 45) of cluster centers, selecting the most matched cluster center by calculating Euclidean distances of corresponding points, and obtaining the curve of the time periods of 12: 00-15: 45 of the cluster center as a calculation result of short-term power load of the main transformer 2019.02.2112: 00-15: 45. The calculation results are shown in fig. 5F.
In one embodiment, as shown in fig. 3, there is provided a main transformer short-term electric load calculation apparatus including, wherein:
a current load data obtaining module 310, configured to obtain current load data within a preset time before a current time;
a cluster center curve obtaining module 320, configured to obtain a plurality of cluster center curves clustered by historical data;
the euclidean distance calculating module 330 is configured to compare the current load data with each cluster center curve, and calculate a euclidean distance between the current load data and each cluster center curve;
a cluster center determining module 340, configured to determine the cluster center curve that is most matched with the current load data according to the calculated euclidean distances;
and a short-term power load number obtaining module 350, configured to obtain short-term power load data of the main transformer according to the determined most matched cluster center curve.
In one embodiment, the main transformer short-term electrical load calculation device further comprises:
the system comprises a sample load data acquisition module, a data processing module and a data processing module, wherein the sample load data acquisition module is used for acquiring a plurality of sample load data;
the cluster processing module is used for carrying out cluster processing on the sample load data to obtain a plurality of cluster center curves;
and the cluster center curve storage module is used for storing each cluster center curve.
In one embodiment, the cluster processing module comprises:
an initial cluster center calculation obtaining unit, configured to calculate and obtain an initial cluster center of each sample load data by using an extreme distance vertex algorithm;
and the cluster center curve calculation obtaining unit is used for carrying out iterative clustering on the data according to the K-means algorithm determined by the initial cluster center to obtain a plurality of cluster center curves.
In one embodiment, the main transformer short-term electrical load calculation device further comprises:
and the cleaning processing module is used for cleaning each sample load data, removing burr data in each sample load data and obtaining the cleaned sample load data.
In one embodiment, the main transformer short-term electrical load calculation device further comprises:
and the data aggregation module is used for aggregating the cleaned sample load data according to a preset rule to obtain the aggregated sample load data.
In one embodiment, the main transformer short-term electrical load calculation device further comprises:
a data mapping module for performing interval mapping on the aggregated sample load data,
the clustering processing module is further used for clustering processing the sample load data subjected to interval mapping to obtain a plurality of cluster center curves.
For specific limitations of the main transformer short-term power load calculation device, reference may be made to the above limitations of the main transformer short-term power load calculation method, which are not described herein again. All or part of the modules in the main transformer short-term electric load calculation device can be realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 4. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is used for storing a plurality of cluster center curves. The network interface of the computer device is used for connecting and communicating with an external node such as a terminal through a network. The computer program is executed by a processor to implement a method of calculating a main transformer short-term electrical load.
Those skilled in the art will appreciate that the architecture shown in fig. 4 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the following steps when executing the computer program:
acquiring current load data within a preset time before the current moment;
acquiring a plurality of pre-stored cluster center curves formed by clustering historical data;
comparing the current load data with each cluster center curve, and calculating the Euclidean distance between the current load data and each cluster center curve;
determining the cluster center curve which is most matched with the current load data according to the calculated Euclidean distances;
and obtaining short-term power load data of the main transformer according to the determined most matched cluster center curve.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
acquiring a plurality of sample load data;
clustering the sample load data to obtain a plurality of cluster center curves;
and storing each cluster center curve.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
calculating to obtain an initial cluster center of each sample load data by adopting an extreme distance vertex algorithm;
and performing iterative clustering on the data according to a K-means algorithm determined by the initial cluster center to obtain a plurality of cluster center curves.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
and cleaning each sample load data, and removing burr data in each sample load data to obtain the cleaned sample load data.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
and aggregating the cleaned sample load data according to a preset rule to obtain the aggregated sample load data.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
performing interval mapping on the aggregated sample load data,
the step of clustering each sample load data to obtain a plurality of cluster center curves comprises:
and clustering the sample load data subjected to interval mapping to obtain a plurality of cluster center curves.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of:
acquiring current load data within a preset time before the current moment;
acquiring a plurality of pre-stored cluster center curves formed by clustering historical data;
comparing the current load data with each cluster center curve, and calculating the Euclidean distance between the current load data and each cluster center curve;
determining the cluster center curve which is most matched with the current load data according to the calculated Euclidean distances;
and obtaining short-term power load data of the main transformer according to the determined most matched cluster center curve.
In one embodiment, the computer program when executed by the processor further performs the steps of:
acquiring a plurality of sample load data;
clustering the sample load data to obtain a plurality of cluster center curves;
and storing each cluster center curve.
In one embodiment, the computer program when executed by the processor further performs the steps of:
calculating to obtain an initial cluster center of each sample load data by adopting an extreme distance vertex algorithm;
and performing iterative clustering on the data according to a K-means algorithm determined by the initial cluster center to obtain a plurality of cluster center curves.
In one embodiment, the computer program when executed by the processor further performs the steps of:
and cleaning each sample load data, and removing burr data in each sample load data to obtain the cleaned sample load data.
In one embodiment, the computer program when executed by the processor further performs the steps of:
and aggregating the cleaned sample load data according to a preset rule to obtain the aggregated sample load data.
In one embodiment, the computer program when executed by the processor further performs the steps of:
performing interval mapping on the aggregated sample load data,
the step of clustering each sample load data to obtain a plurality of cluster center curves comprises:
and clustering the sample load data subjected to interval mapping to obtain a plurality of cluster center curves.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A method of primary transformer short term electrical load calculation, the method comprising:
acquiring current load data within a preset time before the current moment;
acquiring a plurality of pre-stored cluster center curves formed by clustering historical data;
comparing the current load data with each cluster center curve, and calculating the Euclidean distance between the current load data and each cluster center curve;
determining the cluster center curve which is most matched with the current load data according to the calculated Euclidean distances;
and obtaining short-term power load data of the main transformer according to the determined most matched cluster center curve.
2. The method of claim 1, wherein the step of obtaining a plurality of pre-stored cluster-center curves clustered by historical data is preceded by:
acquiring a plurality of sample load data;
clustering the sample load data to obtain a plurality of cluster center curves;
and storing each cluster center curve.
3. The method of claim 1, wherein clustering each of the sample load data, the plurality of cluster-center curves comprises:
calculating to obtain an initial cluster center of each sample load data by adopting an extreme distance vertex algorithm;
and performing iterative clustering on the data according to a K-means algorithm determined by the initial cluster center to obtain a plurality of cluster center curves.
4. The method of claim 1, wherein the step of clustering each of the sample load data to obtain a plurality of cluster center curves further comprises:
and cleaning each sample load data, and removing burr data in each sample load data to obtain the cleaned sample load data.
5. The method according to claim 1, wherein the step of performing the cleaning process on each sample load data to remove the glitch data in each sample load data to obtain the cleaned sample load data further comprises:
and aggregating the cleaned sample load data according to a preset rule to obtain the aggregated sample load data.
6. The method according to claim 1, wherein the step of aggregating the sample load data after the cleaning process according to a preset rule to obtain the aggregated sample load data further comprises:
carrying out interval mapping on the aggregated sample load data;
the step of clustering each sample load data to obtain a plurality of cluster center curves comprises:
and clustering the sample load data subjected to interval mapping to obtain a plurality of cluster center curves.
7. A main transformer short-term electrical load calculation apparatus, the apparatus comprising:
the current load data acquisition module is used for acquiring current load data within preset time before the current moment;
the cluster center curve acquisition module is used for acquiring a plurality of cluster center curves which are pre-stored and formed by clustering historical data;
the Euclidean distance calculation module is used for comparing the current load data with each cluster center curve and calculating the Euclidean distance between the current load data and each cluster center curve;
a cluster center determining module, configured to determine, according to the calculated euclidean distances, the cluster center curve that is most matched with the current load data;
and the short-term power load number acquisition module is used for acquiring the short-term power load data of the main transformer according to the determined most matched cluster center curve.
8. The apparatus of claim 1, further comprising:
the system comprises a sample load data acquisition module, a data processing module and a data processing module, wherein the sample load data acquisition module is used for acquiring a plurality of sample load data;
the cluster processing module is used for carrying out cluster processing on the sample load data to obtain a plurality of cluster center curves;
and the cluster center curve storage module is used for storing each cluster center curve.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the steps of the method of any of claims 1 to 6 are implemented when the computer program is executed by the processor.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910835477.4A CN110728395A (en) | 2019-09-05 | 2019-09-05 | Main transformer short-term power load calculation method and device, computer and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910835477.4A CN110728395A (en) | 2019-09-05 | 2019-09-05 | Main transformer short-term power load calculation method and device, computer and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110728395A true CN110728395A (en) | 2020-01-24 |
Family
ID=69217837
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910835477.4A Pending CN110728395A (en) | 2019-09-05 | 2019-09-05 | Main transformer short-term power load calculation method and device, computer and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110728395A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111401796A (en) * | 2020-04-27 | 2020-07-10 | 新智数字科技有限公司 | Method and device for establishing equipment energy efficiency model |
| CN113762373A (en) * | 2021-08-30 | 2021-12-07 | 广东电网有限责任公司 | Load characteristic abnormity identification method and device, electronic equipment and medium |
| CN113972664A (en) * | 2021-10-29 | 2022-01-25 | 国网上海市电力公司 | Electric power data complement method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012244897A (en) * | 2011-05-13 | 2012-12-10 | Fujitsu Ltd | Apparatus and method for predicting short-term power load |
| CN107918639A (en) * | 2017-10-19 | 2018-04-17 | 广东电网有限责任公司云浮供电局 | Based on electric power big data main transformer peak load forecasting method and data warehouse |
| CN109309382A (en) * | 2018-09-13 | 2019-02-05 | 广东工业大学 | A kind of Short-Term Load Forecasting Method |
| CN109740648A (en) * | 2018-12-21 | 2019-05-10 | 广州供电局有限公司 | Electric load disorder data recognition method, apparatus and computer equipment |
-
2019
- 2019-09-05 CN CN201910835477.4A patent/CN110728395A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012244897A (en) * | 2011-05-13 | 2012-12-10 | Fujitsu Ltd | Apparatus and method for predicting short-term power load |
| CN107918639A (en) * | 2017-10-19 | 2018-04-17 | 广东电网有限责任公司云浮供电局 | Based on electric power big data main transformer peak load forecasting method and data warehouse |
| CN109309382A (en) * | 2018-09-13 | 2019-02-05 | 广东工业大学 | A kind of Short-Term Load Forecasting Method |
| CN109740648A (en) * | 2018-12-21 | 2019-05-10 | 广州供电局有限公司 | Electric load disorder data recognition method, apparatus and computer equipment |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111401796A (en) * | 2020-04-27 | 2020-07-10 | 新智数字科技有限公司 | Method and device for establishing equipment energy efficiency model |
| CN111401796B (en) * | 2020-04-27 | 2024-03-29 | 新智数字科技有限公司 | Method and device for establishing equipment energy efficiency model |
| CN113762373A (en) * | 2021-08-30 | 2021-12-07 | 广东电网有限责任公司 | Load characteristic abnormity identification method and device, electronic equipment and medium |
| CN113972664A (en) * | 2021-10-29 | 2022-01-25 | 国网上海市电力公司 | Electric power data complement method and system |
| CN113972664B (en) * | 2021-10-29 | 2024-02-20 | 国网上海市电力公司 | Electric power data complement method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109740648B (en) | Method and device for identifying abnormal data of power load and computer equipment | |
| CN111368890A (en) | Fault detection method and device and information physical fusion system | |
| CN111210024A (en) | Model training method and device, computer equipment and storage medium | |
| CN106600463B (en) | Local shape similarity ultra-short term load prediction method and device | |
| WO2017076154A1 (en) | Method and apparatus for predicting network event and establishing network event prediction model | |
| CN109861206B (en) | Transient power angle stability fault screening system and method based on support vector machine | |
| CN111160626B (en) | Power load time sequence control method based on decomposition fusion | |
| US9928467B2 (en) | Apparatus for forecasting water demand | |
| CN110391840B (en) | Method and system for judging abnormality of telemetry parameters of sun synchronous orbit satellite | |
| CN110728395A (en) | Main transformer short-term power load calculation method and device, computer and storage medium | |
| CN116227637A (en) | Active power distribution network oriented refined load prediction method and system | |
| CN112085285A (en) | Bus load prediction method and device, computer equipment and storage medium | |
| CN115656673A (en) | Transformer data processing device and equipment storage medium | |
| JP7422272B2 (en) | Method and apparatus for facilitating storage of data from industrial automation control systems or power systems | |
| CN106874511A (en) | A kind of database for corroding quantity of electric charge forecasting system based on insulator metal accessory | |
| CN112421770A (en) | Automatic auxiliary inspection method and system for power distribution automation system | |
| CN112614010A (en) | Load prediction method and device, storage medium and electronic device | |
| CN113110961B (en) | Equipment abnormality detection method and device, computer equipment and readable storage medium | |
| CN117787915A (en) | Digital twin intelligent brain construction method for power distribution network | |
| CN111679953A (en) | Fault node identification method, device, equipment and medium based on artificial intelligence | |
| CN108123436B (en) | Voltage out-of-limit prediction model based on principal component analysis and multiple regression algorithm | |
| CN114756722A (en) | Multi-data centralized management control system and method | |
| CN110766286B (en) | Power distribution network topology checking method and device, computer equipment and readable storage medium | |
| CN113591266A (en) | Method and system for analyzing fault probability of electric energy meter | |
| CN113361750A (en) | Electricity sales amount prediction method based on business expansion large data |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200924 Address after: 510620 Tianhe District, Guangzhou, Tianhe South Road, No. two, No. 2, No. Applicant after: Guangzhou Power Supply Bureau of Guangdong Power Grid Co.,Ltd. Address before: 510620 Tianhe District, Guangzhou, Tianhe South Road, No. two, No. 2, No. Applicant before: GUANGZHOU POWER SUPPLY Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200124 |
|
| RJ01 | Rejection of invention patent application after publication |