Power big data-based main transformer peak load prediction method

Abstract

The invention discloses a main transformer peak load prediction method based on electric power big data, and relates to the technical field of electric power systems. According to the method, the external influence data is subjected to combination and dimension reduction through rapid classification of the data by using a data classification result, so that the operand is greatly reduced, the prediction speed is improved, and the association relationship between the load data and the external influence data can be reserved to the greatest extent by the classification and dimension reduction method; therefore, the prediction accuracy is not influenced by using the external influence data after dimension reduction. According to the method, a large amount of historical data generated by operation and management of a power system is counted, analyzed, predicted and evaluated, information for scientific decision-making of a power enterprise is rapidly and accurately extracted from the historical data, a basis is provided for rapidly and accurately predicting the peak load change of a main transformer in the future, and the data mining processing method is established by establishing a data warehouse system. Accurate and reliable peak load prediction of the main transformer is realized, and the system is reasonably dispatched to operate safely and economically.

Description

Technical Field

[0001] The present invention belongs to the technical field of power systems, and particularly relates to a main transformer peak load forecasting method based on power big data. Background Art

[0002] With the commercialization and marketization of power systems, the accuracy of power load forecasting is of great significance to the safe and economic operation of power systems and the development of the national economy. The level of power load forecasting work has become one of the significant signs of whether the management of a power enterprise is moving towards modernization. As an important basis in electricity trading, the power system load forecast value provides a necessary guidance for power companies to formulate electricity price quotations, operation plans, and grid planning. Its forecasting accuracy will closely affect the economic benefits of power enterprises. Especially today when China's power industry is developing unprecedentedly, with electricity management moving towards the market, the power load forecasting problem has become an important and arduous task faced by power systems.

[0003] The load of the power system is a dynamic process that changes in real time. To ensure the stable operation of the power system, it is necessary to forecast the load to achieve the purpose of setting the power supply plan in advance. Existing load forecasting methods all forecast based on historical data. However, due to the large number of power consumption terminals in the system and the irregular changes of influencing factors, directly using the original data for forecasting consumes a long operation time and has poor real-time performance, and it is impossible to make a new forecasting result in time when sudden situations occur in the power system. Therefore, a main transformer peak load forecasting method based on power big data is proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide a main transformer peak load forecasting method based on power big data to solve the existing problems.

[0005] To solve the above technical problems, the present invention is realized through the following technical solutions:

[0006] The present invention is a main transformer peak load forecasting method based on power big data, including the following steps:

[0007] S1. Collect the load data of each transformer in the power system and the corresponding external influence data, and at the same time, statistically analyze the historical data of the power system to establish a data warehouse system;

[0008] S2. Adopt sequence pattern analysis and clustering analysis data mining methods to analyze the extracted knowledge according to the decision-making purpose of the end user, distinguish the most valuable information, and submit it to the user;

[0009] S3. Merge and reduce the dimension of the external influence data categories corresponding to each type of load data;

[0010] S4. Preprocess the main transformer peak load data based on the sorted historical data, and input the external influence data into the prediction model to predict the power system load.

[0011] Further, the data mining process in step S2 includes data preparation, data mining, and interpretation and evaluation. Among them, the data preparation includes the following steps:

[0012] S201. Data integration: Merge the data in a multi-file or multi-database operating environment, solve semantic ambiguity, handle omissions in the data, and clean up useless data.

[0013] S202. Data selection: According to user requirements, use some database operations to process the data and extract the data set to be mined from the data.

[0014] S203. Data preprocessing: Re-process the data in step S202, check the integrity and consistency of the data, process the noise data therein, fill in the missing data using statistical methods, prepare for further analysis, and determine the type of mining operation to be performed.

[0015] S204. Data transformation: According to the needs of data mining, perform operations such as mutual conversion between discrete value data and continuous value data, grouping and classification of data values, and calculation and combination between data items.

[0016] Further, in step S3, the merging and dimensionality reduction of the external influence data categories corresponding to each type of load data includes the following steps:

[0017] S301. Establish a relationship graph between each type of load data and the corresponding external influence data, with each type of external influence data as a node.

[0018] S302. Set the node priority according to the number of load data associated with the node. The more the number of associated load data, the higher the node priority.

[0019] S303. Perform PCA transformation on the load data associated with the node, and generate a dimensionality reduction matrix using the processed load data.

[0020] S304. Map the external influence data categories to a low-dimensional space using the dimensionality reduction matrix to achieve merging and dimensionality reduction.

[0021] Further, in step S4, the prediction model is a neural network model or a decision tree model.

[0022] Further, in step S3, generating a dimensionality reduction matrix using the processed load data includes the following steps: calculating the differential features of the processed load data, centralizing the differential feature matrix, calculating the eigenvectors of the differential feature matrix, and using the eigenvectors to form the dimensionality reduction matrix.

[0023] The present invention has the following beneficial effects:

[0024] Through the rapid classification of data, the present invention uses the results of data classification to merge and reduce the dimensionality of external influence data, which not only greatly reduces the amount of computation and improves the prediction speed, but also, since the classification and dimensionality reduction method of the present invention can retain the correlation between the load data and the external influence data to the greatest extent, using the dimensionality-reduced external influence data will not affect the accuracy of the prediction.

[0025] By analyzing a large amount of historical data generated by the operation and management of the power system, effectively statistically analyzing, predicting, and evaluating these large amounts of data, extracting information for the scientific decision-making of power enterprises quickly and accurately from them, providing a basis for quickly and accurately predicting the future peak load changes of main transformers, establishing a data warehouse system, establishing a data mining processing method, and using the support vector machine method to predict the load changes of main transformers, the present invention realizes accurate and reliable prediction of the peak load of main transformers and reasonably schedules the system to operate safely and economically.

[0026] Of course, it is not necessary for any product implementing the present invention to simultaneously achieve all the above-mentioned advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

[0028] Figure 1 It is a flowchart of a method for predicting the peak load of a main transformer based on power big data. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0029] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0030] In the description of the present invention, it should be understood that the terms "upper", "middle", "outer", "inner", etc. indicating orientation or positional relationships are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the components or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation on the present invention.

[0031] In the description of the present invention, it should be noted that unless otherwise clearly specified and defined, the terms "installed", "provided with", "connected", etc. should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0032] Please refer to Figure 1 As shown, the present invention is a method for predicting the peak load of main transformers based on power big data, including the following steps:

[0033] S1. Collect the load data of each transformer in the power system and the corresponding external influence data, and at the same time, statistically analyze the historical data of the power system to establish a data warehouse system;

[0034] S2. Adopt sequence pattern analysis and clustering analysis data mining methods, analyze the extracted knowledge according to the decision-making purpose of the end user, distinguish the most valuable information, and submit it to the user;

[0035] S3. Merge and reduce the dimension of the external influence data categories corresponding to each type of load data;

[0036] S4. Through the organized historical data, preprocess the main transformer peak load data, and input the external influence data into the prediction model to predict the power system load.

[0037] In one embodiment, the data mining process in step S2 includes data preparation, data mining, and interpretation and evaluation. Among them, data preparation includes the following steps:

[0038] S201. Data integration, merge and process the data in a multi-file or multi-database operating environment, solve semantic ambiguity, handle omissions in the data, and clean up useless data;

[0039] S202. Data selection, according to the user's requirements, use some database operations to process the data, and extract the data set to be mined from the data;

[0040] S203. Data preprocessing, reprocess the data in step S202, check the integrity and consistency of the data, process the noise data therein, fill in the missing data using statistical methods, prepare for further analysis, and determine the type of mining operation to be performed;

[0041] S204. Data transformation, perform operations such as the mutual conversion between discrete-value data and continuous-value data, the grouping and classification of data values, and the calculation and combination of data items according to the needs of data mining.

[0042] In one embodiment, in step S3, the merging and dimensionality reduction of the external influence data categories corresponding to each type of load data includes the following steps:

[0043] S301. Establish a relationship graph between each type of load data and the corresponding external influence data, with each type of external influence data as a node;

[0044] S302. Set the node priority according to the number of load data associated with the node. The more load data associated with the node, the higher the node priority;

[0045] S303. Perform PCA transformation processing on the load data associated with the node, and generate a dimensionality reduction matrix using the processed load data;

[0046] S304. Use the dimensionality reduction matrix to map the external influence data categories to a low-dimensional space to achieve merging and dimensionality reduction.

[0047] In one embodiment, in step S4, the prediction model is a neural network model or a decision tree model.

[0048] In one embodiment, in step S3, generating a dimensionality reduction matrix using the processed load data includes the following steps: calculate the differential features of the processed load data, centralize the differential feature matrix, calculate the eigenvectors of the differential feature matrix, and use the eigenvectors to form the dimensionality reduction matrix.

[0049] In the description of this specification, the descriptions referring to terms such as "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0050] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the present invention, so that those skilled in the art can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A main transformer peak load prediction method based on power big data, characterized in that: It includes the following steps: S1. Collect the load data of each transformer in the power system and the corresponding external influence data. At the same time, statistically analyze the historical data of the power system and establish a data warehouse system; S2. Adopt sequence pattern analysis and clustering analysis data mining methods, analyze the extracted knowledge according to the decision-making purpose of the end user, distinguish the most valuable information, and submit it to the user; S3. Merge and reduce the dimension of the external influence data categories corresponding to each type of load data; S4. Through the sorted historical data, preprocess the main transformer peak load data, and input the external influence data into the prediction model to predict the power system load.

2. The method for predicting the peak load of main transformers based on power big data according to claim 1, wherein, In the data mining process of step S2, it includes data preparation, data mining, and interpretation and evaluation. Among them, the data preparation includes the following steps: S201. Data integration, merge and process the data in a multi-file or multi-database operating environment, solve semantic ambiguity, handle omissions in the data, and clean up useless data; S202. Data selection, according to the user's requirements, use some database operations to process the data, and extract the data set to be mined from the data; S203. Data preprocessing, reprocess the data in step S202, check the integrity and consistency of the data, process the noise data therein, fill in the missing data using statistical methods, prepare for further analysis, and determine the type of mining operation to be performed; S204. Data transformation, according to the needs of data mining, perform operations such as mutual conversion between discrete value data and continuous value data, grouping and classification of data values, and calculation and combination between data items.

3. A method for predicting the peak load of a main transformer based on power big data according to claim 1, characterized in that, In step S3, the merging and dimension reduction of the external influence data categories corresponding to each type of load data includes the following steps: S301. Establish a relationship graph between each type of load data and the corresponding external influence data, and each type of external influence data is used as a node; S302. Set the node priority according to the number of load data associated with the node. The more the number of associated load data, the higher the node priority; S303. Perform PCA transformation on the load data associated with the node, and generate a dimension reduction matrix using the processed load data; S304. Use the dimension reduction matrix to map the external influence data categories to a low-dimensional space to achieve merging and dimension reduction.

4. A method for predicting the peak load of main transformers based on power big data according to claim 1, characterized in that, In step S4, the prediction model is a neural network model or a decision tree model.

5. A main transformer peak load prediction method based on power big data according to claim 3, characterized in that In step S3, generating a dimension reduction matrix using the processed load data includes the following steps Calculate the differential features of the processed load data, centralize the differential feature matrix, and calculate the eigenvectors of the differential feature matrix, Use the eigenvectors to form a dimension reduction matrix.

Images