Method and system for establishing power system database

Abstract

The invention relates to a method and system for establishing a database of an electric power system, and relates to the technical field of electric power, and the method comprises the following steps: obtaining various electric power related data of the electric power system, the power related data comprises power grid operation data, equipment state data, load data, electric quantity data, fault and accident data, market transaction data, meteorological data and user power consumption data; the particle swarm is initialized, and each particle in the particle swarm represents a power data cleaning strategy; determining a fitness function for each particle; according to the fitness function, parameters of the particle swarm are continuously adjusted and optimized through iterative optimization, so that a final cleaning strategy is obtained; and according to the final cleaning strategy, cleaning the preliminarily acquired power related data to obtain cleaned power data. The method improves the quality and utilization efficiency of power data.

Description

Technical Field

[0001] The present invention relates to the field of electric power technology, and in particular to a method and system for establishing an electric power system database. Background Art

[0002] The power system can not only balance the load of the power grid and improve the stability and regulation ability of the power system, but also effectively store and utilize renewable energy and reduce energy waste. However, with the widespread application of the power system, higher requirements are also put forward for the processing and management of power data.

[0003] In traditional technical solutions, the processing and management of power data may have certain limitations in some aspects, especially in the process of data cleaning and integration.

[0004] First, traditional data cleaning methods mainly rely on simple rules or thresholds to screen data. This approach may be slightly insufficient when dealing with complex and changeable power data, which may affect the cleaning effect and data quality.

[0005] Secondly, in the data integration process, traditional solutions may not be flexible enough when integrating power data from different sources and in different formats, which may, to a certain extent, restrict the in-depth analysis and application of the data. Summary of the invention

[0006] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a method for establishing a power system database, thereby improving the quality and utilization efficiency of power data.

[0007] In order to solve the above technical problems, the basic concept of the technical solution adopted by the present invention is:

[0008] In a first aspect, a method for establishing a power system database comprises:

[0009] Obtain various power-related data of the power system;

[0010] Initialize the particle swarm, where each particle in the particle swarm represents a power data cleaning strategy;

[0011] Determine a fitness function for each particle;

[0012] According to the fitness function, through iterative optimization, the parameters of the particle swarm are continuously adjusted and optimized to obtain the final cleaning strategy;

[0013] According to the final cleaning strategy, the initially acquired power-related data is cleaned to obtain cleaned power data;

[0014] Using time series analysis technology, the cleaned power data is supplemented to obtain supplemented data, and the supplemented data is corrected to obtain sorted power data;

[0015] The sorted power data is stored and classified using the k-clustering analysis algorithm to build a power system database.

[0016] In a preferred embodiment of any of the above schemes, the power-related data includes power grid operation data, equipment status data, load data, power data, fault and accident data, market transaction data, meteorological data and user power consumption data.

[0017] In a preferred embodiment of any of the above schemes, the calculation formula of the fitness function is:

[0018]

[0019] Among them, F i represents the value of the fitness function, α, β, γ, ∈ are the weight coefficients of each indicator, A is the amount of data that has been cleaned correctly, T is the total amount of data, M is the amount of missing data after cleaning, O is the amount of original data, N is the amount of noisy data that still exists after cleaning, C is the total amount of data after cleaning, L is the number of logical errors in the data after cleaning, P is the amount of data processed by cleaning, and S is the time required for cleaning.

[0020] In a preferred embodiment of any of the above schemes, according to the final cleaning strategy, the initially acquired power-related data is cleaned to obtain cleaned power data, including:

[0021] According to the cleaning strategy, the preliminary data is traversed to obtain preliminary data;

[0022] Deleting or replacing invalid, erroneous, and missing data points from the preliminary data to obtain preliminary cleansed data;

[0023] The data of the preliminary cleaning data are standardized and normalized and the data format is corrected to obtain the power data after cleaning.

[0024] In any of the above schemes, a preferred embodiment uses time series analysis technology to complete the cleaned power data to obtain completed data, including:

[0025] Missing data points were identified by examining the timestamps in the dataset to determine missing data.

[0026] Using the random forest method, the missing data points are predicted to obtain the predicted missing data points;

[0027] Fill the predicted missing data points into the corresponding missing data point positions to obtain the completed data;

[0028] The supplemented data is integrated with the original data to obtain the sorted power data.

[0029] In a preferred embodiment of any of the above schemes, correcting the padded data to obtain sorted power data includes:

[0030] Identify outliers through standard deviation statistics to obtain data outliers;

[0031] Based on the analysis of data outliers, determine the strategy for replacing outliers and correcting them;

[0032] Applying a correction strategy to the detected abnormal data points to obtain corrected data;

[0033] Verify and evaluate the corrected data to obtain the sorted power data.

[0034] In a preferred embodiment of any of the above schemes, the sorted power data is classified by a k-cluster analysis algorithm, including:

[0035] According to the sorted power data, the cluster center is initialized through To get the distance from each data point to the cluster center, where x i is each point in the data set, x ij is the data point x i The value of the jth feature, j is the index, n is the dimension of the data, m is the kj is the value of the k-th cluster center on the j-th feature in the iteration, t is the number of iterations, and k is the number of data points;

[0036] According to the distance from each data point to the cluster center, Update the cluster center and get the clustering result, where x i is each point in the data set, k is the number of data points, t is the number of iterations, represents the set of data points of the kth cluster in the tth iteration, represents the center of the kth cluster after the tth iteration;

[0037] According to the clustering results, the characteristics of each cluster are analyzed and the sorted power data are classified.

[0038] In a second aspect, a system for establishing a power system database includes:

[0039] The acquisition module is used to acquire various power-related data of the power system, wherein the power-related data includes power grid operation data, equipment status data, load data, power data, fault and accident data, market transaction data, meteorological data and user power consumption data; initialize the particle swarm, wherein each particle in the particle swarm represents a power data cleaning strategy; determine a fitness function for each particle; according to the fitness function, continuously adjust and optimize the parameters of the particle swarm through iterative optimization to obtain the final cleaning strategy;

[0040] The processing module is used to clean the initially acquired power-related data according to the final cleaning strategy to obtain cleaned power data; use time series analysis technology to complete and correct the cleaned power data to obtain sorted power data; store the sorted power data, and classify the sorted power data through the k clustering analysis algorithm to build a power system database.

[0041] In a second aspect, a system for establishing a power system database includes:

[0042] The acquisition module is used to acquire various power-related data of the power system; initialize the particle swarm, where each particle in the particle swarm represents a power data cleaning strategy; determine a fitness function for each particle; according to the fitness function, continuously adjust and optimize the parameters of the particle swarm through iterative optimization to obtain the final cleaning strategy;

[0043] The processing module is used to clean the initially acquired power-related data according to the final cleaning strategy to obtain cleaned power data; use time series analysis technology to complete and correct the cleaned power data to obtain sorted power data; store the sorted power data, and classify the sorted power data through the k clustering analysis algorithm to build a power system database.

[0044] According to a third aspect, a computing device includes:

[0045] one or more processors;

[0046] The storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method described.

[0047] In a fourth aspect, a computer-readable storage medium stores a program, and when the program is executed by a processor, the method described is implemented.

[0048] After adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art: the cleaning strategy determined by the particle swarm optimization algorithm can more accurately identify and clean the outliers and noise in the power data, thereby significantly improving the quality and accuracy of the data. This method not only cleans the data, but also uses time series analysis technology to complete and correct the cleaned data, effectively integrating data from different sources and different time points, and ensuring the integrity and consistency of the data. Classifying the sorted power data through the k clustering analysis algorithm can more accurately reveal the internal structure and laws of the data. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. Some specific embodiments of the present application will be described in detail in an illustrative and non-restrictive manner with reference to the drawings. The same reference numerals in the drawings indicate the same or similar components or parts. It should be understood by those skilled in the art that these drawings are not necessarily drawn to scale. In the drawings:

[0050] Figure 1 The present invention is a flowchart of a method for establishing a power system database.

[0051] Figure 2 The present invention is a system schematic diagram of a method for establishing a power system database. DETAILED DESCRIPTION

[0052] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present application.

[0053] The following embodiment of the present application takes a method for establishing a power system database as an example to explain the solution of the present application in detail, but this embodiment cannot limit the protection scope of the present application.

[0054] like Figure 1 As shown, the present invention provides a method for establishing a power system database, comprising:

[0055] Step 11, obtaining various power-related data of the power system;

[0056] Step 12, initializing the particle swarm, wherein each particle in the particle swarm represents a power data cleaning strategy;

[0057] Step 13, determine a fitness function for each particle;

[0058] Step 14, according to the fitness function, through iterative optimization, continuously adjust and optimize the parameters of the particle swarm to obtain the final cleaning strategy;

[0059] Step 15, according to the final cleaning strategy, cleaning the initially acquired power-related data to obtain cleaned power data;

[0060] Step 16, using time series analysis technology, supplementing the cleaned power data to obtain supplemented data, and correcting the supplemented data to obtain sorted power data;

[0061] Step 17, storing the sorted power data, and classifying the sorted power data through a k-cluster analysis algorithm to construct a power system database.

[0062] In an embodiment of the present invention, the cleaning strategy of power data is iteratively optimized through a particle swarm optimization algorithm, which can more effectively identify and remove noise, outliers and missing values ​​in the data, thereby improving the accuracy and reliability of the data; the quality of the cleaned power data is significantly improved, providing a more solid foundation for subsequent data analysis and decision-making; the cleaned data is completed and corrected using time series analysis technology, which can fill in the gaps in the data and correct possible errors, making the data more complete and accurate; the sorted power data is classified through a k-cluster analysis algorithm, and a structured power system database is constructed to facilitate data storage, retrieval and analysis.

[0063] In a preferred embodiment of the present invention, the power-related data includes power grid operation data, equipment status data, load data, power data, fault and accident data, market transaction data, meteorological data and user power consumption data.

[0064] In an embodiment of the present invention, the grid operation data includes real-time operation parameters such as voltage, current, power factor, frequency, etc. of the grid. The equipment status data involves parameters such as the working status, temperature, pressure, etc. of power equipment such as transformers, generators, circuit breakers, and disconnectors. The load data is to record the power load conditions of each time period, including total load, peak load, load curve, etc. The power data includes power generation, power consumption, power loss, etc. The fault and accident data includes recording fault events, accident causes, and processing processes in the power system. The market transaction data involves the sale and purchase of electric energy, including electricity prices, transaction volumes, transaction times, etc. Meteorological data includes wind speed, wind direction, temperature, humidity, etc., which have an important impact on the stable operation of the power system and the utilization of new energy (such as wind energy and solar energy). User power consumption data includes each user's power consumption, power consumption period, power consumption habits, etc., which help to understand user needs and optimize power supply. When constructing a power system database, these "various power-related data" need to go through steps such as cleaning, filling, and correction to ensure the accuracy and integrity of the data. Through data mining techniques such as cluster analysis, we can further explore the potential value of these data and provide strong support for the planning, operation and management of power systems.

[0065] In a preferred embodiment of the present invention, the above step 12 includes:

[0066] Get the basic parameters of the particle swarm,

[0067] According to the basic parameters of the particle swarm, the particle position and velocity are initialized to complete the particle initialization.

[0068] In an embodiment of the present invention, by randomly initializing the positions and velocities of particles, it can be ensured that the algorithm can cover a wide search space at the beginning. The particle swarm algorithm is a parallel optimization algorithm. By simultaneously updating the positions and velocities of multiple particles, the efficiency of the optimization process can be significantly improved. Since the particle swarm algorithm does not depend on the specific form of the problem, it has good adaptability and robustness to different types of power data cleaning strategies.

[0069] In a preferred embodiment of the present invention, in the above step 13, the calculation formula of the fitness function is:

[0070]

[0071] Among them, F i represents the value of the fitness function, α, β, γ, ∈ are the weight coefficients of each indicator, A is the amount of data that has been cleaned correctly, T is the total amount of data, M is the amount of missing data after cleaning, O is the amount of original data, N is the amount of noisy data that still exists after cleaning, C is the total amount of data after cleaning, L is the number of logical errors in the data after cleaning, P is the amount of data processed by cleaning, and S is the time required for cleaning.

[0072] In an embodiment of the present invention, the fitness function can comprehensively and objectively evaluate the effects of different cleaning strategies by comprehensively considering multiple indicators such as the amount of data correctly cleaned, the total amount of data, the amount of data missing after cleaning, the amount of original data, the amount of noise data still existing after cleaning, the total amount of data after cleaning, the number of logical errors in the data after cleaning, the amount of data processed by cleaning, and the time required for cleaning; by adjusting the weight coefficient, the quality of data cleaning can be optimized for different application scenarios and data characteristics; the inclusion of an indicator of the time required for cleaning in the fitness function helps to consider the cleaning efficiency while optimizing the quality of data cleaning. Through weight adjustment, the time required for cleaning can be reduced as much as possible while ensuring data quality; the fitness function takes into account the amount of data missing after cleaning and the number of logical errors in the data, which helps to reduce data missing and logical errors during the cleaning process and improve the integrity and consistency of the data; by adjusting the weight coefficients of various indicators, the optimization target can be flexibly defined.

[0073] In a preferred embodiment of the present invention, the above step 14 includes:

[0074] According to the fitness function, the corresponding fitness value is obtained;

[0075] Compare the fitness value with the historical optimal fitness value to update to the global optimal position;

[0076] According to the global optimal position, boundary processing is performed and iteration is repeated to obtain the final cleaning strategy.

[0077] In the embodiment of the present invention, iterative optimization can continuously approach the optimal solution, improve the quality and effect of the cleaning strategy, and the continuous adjustment and optimization of the particle swarm algorithm can ensure that the algorithm maintains diversity and convergence during the search process; through comparison, it can be judged whether the current strategy is better than the historical optimal strategy, and the global optimal position represents the best cleaning strategy currently found; through boundary processing, it can be avoided that particles exceed the effective range during the search process, thereby maintaining the effectiveness and efficiency of the search; through continuous iteration, the algorithm can fully search in multi-dimensional space and gradually approach the global optimal solution. The cleaning strategy finally obtained will be a high-quality strategy that has been optimized and verified many times.

[0078] In an embodiment of the present invention, the amount of cleaning data of each particle is substituted into the fitness function to obtain the corresponding fitness value. If the fitness value of the current particle is better than its historical optimal fitness value, its individual optimal position is updated, all particles are traversed, the particle with the optimal fitness value is found, and its position is updated to the global optimal position. It is checked whether the global optimal position exceeds a preset range. If it exceeds, boundary processing is performed to limit it within the boundary; it is determined whether the preset number of iterations is reached or other stopping conditions are met. If the stopping conditions are not met, the above steps are repeated; otherwise, the global optimal position is output as the final cleaning strategy.

[0079] In a preferred embodiment of the present invention, the above step 15 includes:

[0080] Step 151, traversing the preliminary data according to the cleaning strategy to obtain preliminary data;

[0081] Step 152, deleting or replacing invalid, erroneous and missing data points from the preliminary data to obtain preliminary cleansed data;

[0082] Step 153 , standardize and normalize the preliminary cleaned data and correct the data format to obtain cleaned power data.

[0083] In an embodiment of the present invention, by traversing preliminary data and deleting or replacing invalid, erroneous and missing data points, the quality of the data can be significantly improved, which is helpful for subsequent data analysis and decision-making processes. The data standardization and normalization steps ensure the consistency and comparability of the data, and the data format correction step ensures the uniformity and correctness of the data format. Through the automated and systematic data cleaning process, the time and labor costs required for manual data cleaning can be reduced, and the efficiency of data cleaning can be improved. The cleaned power data is easier to use and understand.

[0084] In the specific implementation process of the present invention, a cleaning strategy and methods and rules for data standardization and normalization are defined, preliminary data are traversed, and erroneous data points are replaced with appropriate values ​​according to the rules defined in the cleaning strategy. For missing data points, these data points can be deleted; data standardization or normalization methods are applied to convert the data to a unified scale; the format of the data is checked to ensure that all data conform to the expected format standards; the cleaned power data is saved in a file or database for subsequent analysis and use.

[0085] In a preferred embodiment of the present invention, the above step 16 includes:

[0086] Step 161 , identifying missing data points by checking the timestamps in the data set to determine missing data.

[0087] Step 162, using the random forest method, predicting the missing data points to obtain predicted missing data points;

[0088] Step 163, filling the predicted missing data points into the corresponding missing data point positions to obtain completed data;

[0089] Step 164, integrating the supplemented data with the original data to obtain sorted power data.

[0090] In the embodiment of the present invention, by identifying and predicting missing data points and filling them in the corresponding positions, the integrity of the data can be significantly improved. By using the random forest method to predict missing data, a relatively accurate prediction value can be obtained; missing data may cause deviation or uncertainty in the data analysis results. By filling in the missing data, the results of data analysis can be optimized to make them more accurate and reliable; the filled data is easier to use and understand, and the automatic identification and prediction of missing data can save a lot of time and resources, improve the efficiency of data processing, and allow more resources to be used for other important tasks.

[0091] In the specific implementation process of the present invention, an appropriate data processing library is used to read the power data set, and the timestamp is checked. According to the missing timestamp, the random forest method is used to predict the missing data points to obtain the predicted missing data points, and the characteristic variables and the target variables are extracted from the original data. The data set is divided into a training set and a test set, and the random forest algorithm is used to predict the missing data points to obtain the predicted values ​​of the missing data points; the predicted values ​​of the missing data points are filled into the corresponding missing positions in the original data set. After the filling operation, the padded power data set is obtained, and the padded data is integrated with the original data to obtain the sorted power data.

[0092] In a preferred embodiment of the present invention, the above step 16 further includes:

[0093] Step 165, identifying outliers by using a standard deviation statistical method to obtain data outliers;

[0094] Step 166, determining a correction strategy for replacing outlier values ​​based on the analysis of data outliers;

[0095] Step 167, applying a correction strategy to the detected abnormal data points to obtain corrected data;

[0096] Step 168, verifying and evaluating the correction data to obtain sorted power data.

[0097] In an embodiment of the present invention, outliers are identified and processed by standard deviation statistical methods, which can significantly improve the quality of data. In the process of data modeling and machine learning, outliers will cause the performance of the model to decline, affect the effect of data visualization, and cause misleading information in charts or graphs. By correcting outliers, the effect of data visualization can be improved, and the charts can more clearly and accurately reflect the characteristics and trends of the data. By processing outliers, it can be ensured that the decision-making process is based on high-quality data, thereby improving the reliability and effectiveness of the decision. Automatic identification and correction of outliers can save a lot of time and resources.

[0098] In the specific implementation process of the present invention, for each data point, the deviation of the average value is calculated, and the deviation is compared with the standard deviation. If the deviation exceeds a preset multiple of the standard deviation, the data point is regarded as an outlier, and all data points identified as outliers and their location information are recorded; the outliers are further analyzed to determine the cause of their generation, and according to the analysis results of the outliers, a correction strategy of replacing with the average value, replacing with the median, replacing with the mode, or replacing with the previous or next valid data is selected, and a detailed correction plan is formulated. According to the correction plan, the detected abnormal data points are corrected by applying the selected correction strategy; the corrected data points are replaced back into the original data set to obtain a corrected power data set, and the corrected data are verified and evaluated to obtain a final sorted power data set.

[0099] In a preferred embodiment of the present invention, the above step 17 includes:

[0100] Step 171, based on the sorted power data, initialize the cluster center, through To get the distance from each data point to the cluster center, where x i is each point in the data set, x ij is the data point x i The value of the jth feature, j is the index, n is the dimension of the data, m is the kj is the value of the k-th cluster center on the j-th feature in the iteration, t is the number of iterations, and k is the number of data points;

[0101] Step 172, based on the distance from each data point to the cluster center, Update the cluster center and get the clustering result, where x i is each point in the data set, k is the number of data points, t is the number of iterations, represents the set of data points of the kth cluster in the tth iteration, represents the center of the kth cluster after the tth iteration;

[0102] Step 173, based on the clustering results, analyze the characteristics of each cluster and classify the sorted power data.

[0103] In an embodiment of the present invention, cluster analysis can be used to more deeply understand the internal structure of power data. The clustering results can classify the sorted power data into different groups, providing strong support for data labeling and classification. Cluster analysis helps to identify outliers or outliers in the data. Through clustering, the core features of the data can be extracted, data dimensionality reduction can be achieved, and subsequent data analysis and model building processes can be simplified. The results of cluster analysis can provide strong support for decision-making. Analyzing the time series data in the clustering results can reveal the trends and patterns of power data, configure and optimize power resources, improve resource utilization efficiency, and reduce costs.

[0104] In the specific implementation process of the present invention, K data points are selected as initial cluster centers, and the initial centers are selected by selecting the K-means algorithm. For each point in the data set, its distance to each cluster center is calculated, and each data point is assigned to the nearest cluster center to form K clusters; for each cluster, the mean of all its data points is calculated, and the mean is used as the new cluster center, and the above process is repeated until the cluster center no longer changes or reaches a preset number of iterations. After iterative updates, the final cluster center and clustering results are obtained; for each cluster, the mean, variance, maximum value and minimum value of its data points are analyzed, and according to the clustering results, the sorted power data are classified into different clusters, and each cluster can be regarded as a category or group for subsequent data analysis and decision-making; the clustering results are output as reports or visualization graphics.

[0105] like Figure 2 As shown, a system 20 for establishing a power system database includes the following steps:

[0106] The acquisition module 21 is used to acquire various power-related data of the power system, wherein the power-related data includes power grid operation data, equipment status data, load data, power data, fault and accident data, market transaction data, meteorological data and user power consumption data; initialize the particle swarm, wherein each particle in the particle swarm represents a power data cleaning strategy; determine a fitness function for each particle; according to the fitness function, continuously adjust and optimize the parameters of the particle swarm through iterative optimization to obtain the final cleaning strategy;

[0107] The processing module 22 is used to clean the initially acquired power-related data according to the final cleaning strategy to obtain cleaned power data; use time series analysis technology to complete and correct the cleaned power data to obtain sorted power data; store the sorted power data, and classify the sorted power data through the k clustering analysis algorithm to construct a power system database.

[0108] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for establishing a power system database, characterized in that: The following steps are involved: Obtain various power-related data of the power system; Initialize the particle swarm, where each particle in the particle swarm represents a power data cleaning strategy; Determine a fitness function for each particle; According to the fitness function, through iterative optimization, the parameters of the particle swarm are continuously adjusted and optimized to obtain the final cleaning strategy; According to the final cleaning strategy, the initially acquired power-related data is cleaned to obtain cleaned power data; Using time series analysis technology, the cleaned power data is supplemented to obtain supplemented data, and the supplemented data is corrected to obtain sorted power data; The sorted power data is stored and classified using the k-clustering analysis algorithm to build a power system database.

2. The method for establishing a power system database according to claim 1, characterized in that: Electricity-related data include grid operation data, equipment status data, load data, electricity data, fault and accident data, market transaction data, meteorological data and user electricity consumption data.

3. The method for establishing a power system database according to claim 2, characterized in that: The calculation formula of the fitness function is: Among them, F i represents the value of the fitness function, α, β, γ, ∈ are the weight coefficients of each indicator, A is the amount of data that has been cleaned correctly, T is the total amount of data, M is the amount of missing data after cleaning, O is the amount of original data, N is the amount of noisy data that still exists after cleaning, C is the total amount of data after cleaning, L is the number of logical errors in the data after cleaning, P is the amount of data processed by cleaning, and S is the time required for cleaning.

4. The method for establishing a power system database according to claim 3, characterized in that: According to the final cleaning strategy, the initially acquired power-related data is cleaned to obtain cleaned power data, including: According to the cleaning strategy, the preliminary data is traversed to obtain preliminary data; Deleting or replacing invalid, erroneous, and missing data points from the preliminary data to obtain preliminary cleansed data; The data of the preliminary cleaning data are standardized and normalized and the data format is corrected to obtain the cleaned power data.

5. The method for establishing a power system database according to claim 4, characterized in that: The cleaned power data is supplemented by time series analysis technology to obtain supplemented data, including: Missing data points were identified by examining the timestamps in the dataset to determine missing data. Using random forest method, the missing data points are predicted to obtain predicted missing data points; Fill the predicted missing data points into the corresponding missing data point positions to obtain the completed data; The supplemented data is integrated with the original data to obtain the sorted power data.

6. A method for establishing a power system database according to claim 5, characterized in that: Correct the padded data to obtain the sorted power data, including: Identify outliers through standard deviation statistics to obtain data outliers; Based on the analysis of data outliers, determine the strategy for replacing outliers and correcting them; Applying a correction strategy to the detected abnormal data points to obtain corrected data; Verify and evaluate the corrected data to obtain the sorted power data.

7. A method for establishing a power system database according to claim 6, characterized in that: The sorted power data is classified by the k-cluster analysis algorithm, including: According to the sorted power data, the cluster center is initialized through To get the distance from each data point to the cluster center, where x i is each point in the data set, x ij is the data point x i The value of the jth feature, j is the index, n is the dimension of the data, m is the kj is the value of the k-th cluster center on the j-th feature in the iteration, t is the number of iterations, and k is the number of data points; According to the distance from each data point to the cluster center, Update the cluster center and get the clustering result, where x i is each point in the data set, k is the number of data points, t is the number of iterations, represents the set of data points of the kth cluster in the tth iteration, represents the center of the kth cluster after the tth iteration; According to the clustering results, the characteristics of each cluster are analyzed and the sorted power data are classified.

8. A system for establishing a power system database, characterized in that: include: The acquisition module is used to acquire various power-related data of the power system, wherein the power-related data includes power grid operation data, equipment status data, load data, power data, fault and accident data, market transaction data, meteorological data and user power consumption data; initialize the particle swarm, wherein each particle in the particle swarm represents a power data cleaning strategy; determine a fitness function for each particle; according to the fitness function, continuously adjust and optimize the parameters of the particle swarm through iterative optimization to obtain the final cleaning strategy; The processing module is used to clean the initially acquired power-related data according to the final cleaning strategy to obtain cleaned power data; use time series analysis technology to complete and correct the cleaned power data to obtain sorted power data; store the sorted power data, and classify the sorted power data through the k clustering analysis algorithm to build a power system database.

9. A computing device, characterized in that include: one or more processors; A storage device, used for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images