A remote control method and control system for electric power equipment

Abstract

The present invention provides a remote control method for electric equipment and a control system thereof, belonging to the technical field of electric equipment. Multivariate data of electric equipment obtained by preprocessing is obtained to obtain preprocessed multivariate data; then, feature selection is performed on the preprocessed multivariate data to obtain an optimal feature set; then, the optimal feature set is input into a pre-constructed abnormality detection model, and the abnormal result is determined; finally, based on a control strategy mapping library self-constructed by using a genetic algorithm, the determined abnormal result is matched with the control strategy in the control strategy mapping library, the corresponding preset operation of the electric equipment when the abnormal result occurs is determined, and the electric equipment is regulated according to the determined preset operation. The present invention can not only monitor the multivariate data of the electric equipment in real time, realize fast and accurate identification of abnormalities by using a feature set optimized by a genetic algorithm and a neural network model, but also intelligently match and execute preset operation instructions, and effectively ensure the safe operation of the electric equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of electric power equipment, and in particular to a remote control method for electric power equipment and a control system thereof. Background Art

[0002] With the development of smart grids and the continuous improvement of the automation level of power systems, remote monitoring and intelligent control of power equipment have become one of the key technologies to ensure the stable operation of power systems and improve energy efficiency. Traditional power equipment monitoring mainly relies on manual regular inspections and on-site maintenance. This method is not only inefficient, but also difficult to detect and handle potential faults in the power system in a timely manner, increasing the operating risk of the system. Therefore, it is particularly important to develop a remote control method that can monitor the status of power equipment in real time, efficiently identify anomalies, and automatically take corresponding control measures.

[0003] In recent years, with the rapid development of technologies such as big data and artificial intelligence, more and more researchers and companies have begun to explore data-driven remote control methods for power equipment. This method obtains multivariate data of power equipment in real time, processes and analyzes the data using machine learning, deep learning and other technologies, thereby achieving precise control of power equipment.

[0004] However, there are still some problems with the existing data-driven remote control methods for power equipment. First, there is often a lack of effective preprocessing methods for the acquired multivariate data, resulting in low data quality and affecting subsequent analysis and control effects. Second, in terms of feature selection, existing methods are often based on experience or simple statistical methods, and it is difficult to mine truly useful features from large amounts of data. In addition, the construction and training of anomaly detection models is also a difficult problem, and an effective method needs to be found to improve the accuracy and generalization ability of the model.

[0005] Therefore, it is necessary to provide a remote control method for electric power equipment and a control system thereof to solve the above technical problems. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides a remote control method for power equipment and a control system thereof, which monitors the multivariate data of the power equipment in real time, uses a feature set optimized by a genetic algorithm and a neural network model to realize fast and accurate identification of anomalies, and can also intelligently match and execute preset operating instructions, effectively ensuring the safe operation of the power equipment and improving the stability and efficiency of the entire power system.

[0007] The present invention provides a remote control method for electric power equipment, the control method comprising the following steps:

[0008] Acquire multivariate data of the power equipment in real time, and preprocess the multivariate data to obtain preprocessed multivariate data, wherein the multivariate data includes sensor data and environmental data corresponding to the power equipment;

[0009] Based on the preprocessed multivariate data, feature selection is performed according to a predefined feature screening rule to obtain an optimal feature set, wherein the feature screening rule is set in the processing of historical multivariate data of the power equipment according to a genetic algorithm;

[0010] Inputting the optimal feature set into a pre-built anomaly detection model and determining an abnormal result, wherein the anomaly detection model is a neural network model trained by extracting the optimal feature set of samples from historical multivariate data of the power equipment based on a genetic algorithm;

[0011] Based on a control strategy mapping library self-constructed using a genetic algorithm, the determined abnormal result is matched with the control strategy in the control strategy mapping library, the corresponding preset operation of the power equipment when the abnormal result occurs is determined, and the power equipment is regulated according to the determined preset operation to ensure the safety of the power equipment.

[0012] Preferably, the real-time acquisition of multivariate data of the power equipment and preprocessing of the multivariate data to obtain the preprocessed multivariate data include:

[0013] Using data acquisition equipment to collect sensor data and environmental data of the power equipment in real time, wherein the sensor data includes current, voltage, temperature and humidity, and the environmental data includes external climate;

[0014] Performing data cleaning on sensor data and environmental data collected from power equipment in real time, wherein the data cleaning includes missing value processing, outlier processing, duplicate data processing and consistency checking;

[0015] The sensor data and the environmental data after cleaning are combined to obtain preprocessed multivariate data. Preferably, the feature selection is performed based on the preprocessed multivariate data according to a predefined feature screening rule to obtain an optimal feature set, wherein the feature screening rule is set according to a genetic algorithm in the processing of historical multivariate data of the power equipment, including:

[0016] Extract sample features from historical multivariate data and use genetic algorithms to encode sample features;

[0017] Initializing the sample features after feature encoding to obtain a sample initial solution set, wherein each solution in the sample initial solution set represents a sample feature set;

[0018] Using different sample feature sets in the sample initial solution set to train an anomaly detection model, and obtaining a fitness value according to the anomaly detection model training result, wherein the historical multivariate data includes normal data and abnormal data;

[0019] Based on the fitness value, a sample optimal feature set is screened out from different sample feature sets, and the features in the preprocessed multivariate data are matched with the sample optimal feature set to obtain an optimal feature set consisting of multiple features contained in the preprocessed multivariate data.

[0020] Preferably, the optimal feature set is input into a pre-built anomaly detection model and the abnormal result is determined, wherein the anomaly detection model is a neural network model trained by extracting the optimal feature set of samples from the historical multivariate data of the power equipment based on a genetic algorithm, and comprises:

[0021] Design the architecture and parameters of the neural network model based on the genetic algorithm, and use the optimal feature set of the sample to train the initial neural network model to obtain the anomaly detection model;

[0022] The optimal feature set of the preprocessed multivariate data obtained in real time is input into the anomaly detection model for forward propagation, and the detection result corresponding to the optimal feature set is output;

[0023] The detection result is judged as abnormal based on a preset abnormal threshold, and an abnormal result is obtained.

[0024] Preferably, the control strategy mapping library self-constructed based on the genetic algorithm matches the determined abnormal result with the control strategy in the control strategy mapping library, determines the preset operation corresponding to the power equipment when the abnormal result occurs, and regulates the power equipment according to the determined preset operation, including:

[0025] A control strategy mapping library is constructed using a genetic algorithm, wherein the control strategy mapping library includes sample abnormal results and control strategies of a mapping relationship;

[0026] Based on the established mapping relationship, the determined abnormal result is matched with the sample abnormal result of the control strategy mapping library, and the matched sample abnormal result is output;

[0027] Extracting the control strategy corresponding to the matched sample abnormal result, and determining the preset operation for the power equipment according to the extracted control strategy;

[0028] Control electrical equipment according to preset operations.

[0029] Preferably, the control strategy mapping library is constructed by using a genetic algorithm, wherein the control strategy mapping library includes sample abnormal results and control strategies of a mapping relationship, including:

[0030] Predefine control strategies and perform initialization operations;

[0031] Based on the control strategy, the genetic algorithm is iteratively optimized in combination with the historical multivariate data of the power equipment to obtain the sample abnormality results;

[0032] The iteratively optimized control strategy is mapped to the sample abnormality results to form a control strategy mapping library. Each sample abnormality corresponds to a control strategy or a combination of multiple control strategies.

[0033] Preferably, the preset operation includes one or a combination of load adjustment, voltage and frequency regulation, line switching, power factor correction, harmonic suppression, emergency shutdown and reclosing.

[0034] The present invention also provides a remote control system for electric equipment, which is applied to the remote control method for electric equipment, and the control system comprises:

[0035] A preprocessing module, used for acquiring multivariate data of the power equipment in real time, and preprocessing the multivariate data to obtain preprocessed multivariate data, wherein the multivariate data includes sensor data and environmental data corresponding to the power equipment;

[0036] An optimal feature set determination module, used for selecting features based on the preprocessed multivariate data according to predefined feature screening rules to obtain an optimal feature set, wherein the feature screening rules are set according to a genetic algorithm in processing historical multivariate data of the power equipment;

[0037] An anomaly detection module, used for inputting the optimal feature set into a pre-built anomaly detection model and determining an abnormal result, wherein the anomaly detection model is a neural network model trained by extracting the optimal feature set of samples from the historical multivariate data of the power equipment based on a genetic algorithm;

[0038] The control strategy matching module is used to match the determined abnormal result with the control strategy in the control strategy mapping library based on the control strategy mapping library self-constructed by using the genetic algorithm, determine the corresponding preset operation of the power equipment when the abnormal result occurs, and regulate the power equipment according to the determined preset operation to ensure the safety of the power equipment.

[0039] Compared with the related art, the remote control method and control system of electric power equipment provided by the present invention have the following beneficial effects:

[0040] The present invention obtains multivariate data of power equipment in real time and preprocesses the multivariate data to obtain preprocessed multivariate data; then, based on the preprocessed multivariate data, feature selection is performed according to predefined feature screening rules to obtain an optimal feature set; then the optimal feature set is input into a pre-constructed anomaly detection model, and the abnormal result is determined; finally, based on a control strategy mapping library self-constructed by using a genetic algorithm, the determined abnormal result is matched with the control strategy in the control strategy mapping library, the corresponding preset operation of the power equipment when the abnormal result occurs is determined, and the power equipment is regulated according to the determined preset operation to ensure the safety of the power equipment. The present invention forms an efficient and intelligent remote monitoring and control system by integrating real-time data acquisition, genetic algorithm-assisted feature selection, neural network anomaly detection model and a control strategy mapping library constructed based on a genetic algorithm. It can not only monitor the multivariate data of power equipment in real time, use the feature set optimized by the genetic algorithm and the neural network model to realize rapid and accurate identification of anomalies, but also intelligently match and execute preset operation instructions, effectively ensure the safe operation of power equipment, and improve the stability and efficiency of the entire power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A flowchart of a remote control method for electric power equipment provided by the present invention;

[0042] Figure 2 A flowchart of step S1 of a remote control method for electric power equipment provided by the present invention;

[0043] Figure 3 A flowchart of step S2 of a method for remotely controlling electric power equipment provided by the present invention;

[0044] Figure 4 A flowchart of step S3 of a method for remotely controlling electric power equipment provided by the present invention;

[0045] Figure 5 A flowchart of step S4 of a method for remotely controlling electric power equipment provided by the present invention;

[0046] Figure 6 A module structure diagram of a remote control system for electric power equipment provided by the present invention. DETAILED DESCRIPTION

[0047] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only the parts related to the present invention, rather than all structures, are shown in the accompanying drawings. In addition, the embodiments of the present invention and the features in the embodiments may be combined with each other without conflict.

[0048] It should also be noted that, for ease of description, only the part relevant to the present invention but not all the content is shown in the accompanying drawings. It should be mentioned before discussing the exemplary embodiments in more detail that some exemplary embodiments are described as the processing or method described as a flow chart. Although the flow chart describes each operation (or step) as being processed sequentially, many operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of each operation can be rearranged. When its operation is completed, the processing can be terminated, but it can also have additional steps not included in the accompanying drawings. The processing can correspond to a method, a function, a procedure, a subroutine, a subprogram, etc.

[0049] Embodiment 1

[0050] The present invention provides a remote control method for electric power equipment, referring to Figure 1 As shown, the control method comprises the following steps:

[0051] S1: Acquire multivariate data of the power equipment in real time, and preprocess the multivariate data to obtain preprocessed multivariate data, wherein the multivariate data includes sensor data and environmental data corresponding to the power equipment.

[0052] In this embodiment, the sensor network collects sensor data of power equipment in real time, and the sensor data includes but is not limited to humidity, temperature, voltage, power and current, and reads weather data provided by the network through the weather API, including wind speed, precipitation and other information.

[0053] In specific implementation, the original multivariate data usually contains noise, missing values, and inconsistency, which directly affect the accuracy of data analysis. Therefore, preprocessing multivariate data is the basis for improving data quality and the reliability of subsequent analysis. Preprocessing includes but is not limited to missing value processing, outlier processing, duplicate data processing, and consistency checking. Finally, the required features are selected from the original multivariate data to better reflect the equipment status, improve data quality, make subsequent feature selection and model training more effective, and improve the accuracy and reliability of the control strategy.

[0054] S2: Based on the preprocessed multivariate data, feature selection is performed according to predefined feature screening rules to obtain an optimal feature set, wherein the feature screening rules are set according to a genetic algorithm in processing historical multivariate data of the power equipment.

[0055] In this embodiment, the multivariate data of the power equipment contains a large number of features, but not all features contribute to anomaly detection and control strategy formulation. The selection of effective features can reduce the curse of dimensionality and improve the efficiency and interpretability of model training.

[0056] In the specific implementation, a genetic algorithm is used to evaluate the feature combination through an iterative process based on the historical multivariate data of power equipment. Each generation of genetic algorithms will perform "natural selection" according to the predictive performance of the feature set (evaluated through cross-validation), retain the feature combinations with good performance, and explore better feature sets through "crossover" and "mutation" operations to finally obtain the optimal feature set.

[0057] S3: Inputting the optimal feature set into a pre-constructed anomaly detection model and determining an abnormal result, wherein the anomaly detection model is a neural network model trained by extracting the optimal feature set of samples from historical multivariate data of the power equipment based on a genetic algorithm.

[0058] In this embodiment, timely and accurate anomaly detection is the key to preventing power equipment failures. The neural network model is trained based on the data after preprocessing and feature selection, which can automatically learn abnormal patterns and achieve efficient monitoring.

[0059] In the specific implementation, the optimal feature set screened by the genetic algorithm is used as input to train a neural network model. The model learns the boundary between normal and abnormal data during the training process. The training data comes from the historical multivariate data of power equipment. The weights are adjusted through back propagation to minimize the loss function and achieve the best fit. The anomaly detection model can quickly and accurately determine whether an anomaly occurs when new data flows in, providing a basis for timely control measures.

[0060] S4: Based on the control strategy mapping library self-constructed by using genetic algorithm, the determined abnormal result is matched with the control strategy in the control strategy mapping library, the corresponding preset operation of the power equipment when the abnormal result occurs is determined, and the power equipment is regulated according to the determined preset operation to ensure the safety of the power equipment.

[0061] In this embodiment, taking action directly against abnormal results is the key to ensuring safe and stable operation of the power system. At the same time, control strategy matching based on the mapping library can achieve rapid response and precise regulation.

[0062] In the specific implementation, according to the abnormal result determined in step S3, the corresponding control strategy pre-set in the control strategy mapping library is searched. The control strategy mapping library is optimized and constructed through a genetic algorithm to ensure a high degree of matching between the control strategy and the abnormal type. Once the match is successful, the corresponding preset operation is immediately executed, realizing rapid mapping of abnormalities to control actions, enhancing the autonomous maintenance capability and emergency response capability of the power system, and reducing the risk of failure.

[0063] In this embodiment, reference Figure 2 As shown, step S1 includes the following steps:

[0064] S101: Using data acquisition equipment to collect sensor data and environmental data of power equipment in real time, wherein the sensor data includes current, voltage, temperature and humidity, and the environmental data includes external climate.

[0065] In this embodiment, the sensor network deployed on the power equipment continuously monitors the key parameters of the equipment, and calls the third-party weather API through the network interface to capture the meteorological information of the current location in real time, such as wind speed, precipitation, etc., to provide a comprehensive data source for subsequent data analysis.

[0066] S102: Performing data cleaning on sensor data and environmental data collected from power equipment in real time, wherein the data cleaning includes missing value processing, abnormal value processing, duplicate data processing and consistency checking.

[0067] In this embodiment, data cleaning mainly includes four steps: missing value processing, such as using the mean, median or forward and backward filling method to supplement missing data; outlier processing, identifying and removing or correcting outliers through statistical methods (such as box plot analysis); duplicate data processing, identifying and deleting completely identical records in the data set; consistency check, ensuring the consistency and rationality of the data in the time series. The quality of the cleaned data is significantly improved, reducing the possibility of erroneous analysis and improving the accuracy of feature extraction and model training.

[0068] S103: Combining the cleaned sensor data and environmental data to obtain pre-processed multivariate data.

[0069] In this embodiment, the cleaned sensor data and environmental data are merged in a unified format and time series to form preprocessed multivariate data. The merging method includes data standardization or normalization processing to ensure that data from different sources and dimensions can be directly compared and analyzed.

[0070] In this embodiment, reference Figure 3 As shown, step S2 includes the following steps:

[0071] S201: Extract sample features from historical multivariate data, and use a genetic algorithm to perform feature encoding on the sample features.

[0072] In this embodiment, extracting features from historical multivariate data is the first step of feature selection, and the purpose is to convert it into a format that can be understood by the model. Using a genetic algorithm for feature encoding converts the features into a chromosome form that can be processed by the genetic algorithm for genetic operations.

[0073] In specific implementation, various features closely related to the equipment status are extracted from the historical multivariate data of the power equipment. Subsequently, each feature is converted into an individual (chromosome) in the genetic algorithm by using binary coding or real number coding, where "1" or a specific value may indicate that the feature is selected, and "0" or another value indicates that it is not selected. The feature extraction and feature coding provide a basis for the subsequent operation of the genetic algorithm, ensuring that the algorithm can efficiently explore the feature space and find the optimal feature combination.

[0074] S202: Initializing the sample features after feature encoding to obtain a sample initial solution set, wherein each solution in the sample initial solution set represents a sample feature set.

[0075] In this embodiment, an initial solution set, ie, a candidate set of feature combinations, is generated through initialization so that the algorithm can be optimized on this basis.

[0076] In the specific implementation, a series of feature codes are generated randomly to form an initial population. Each solution (chromosome) represents a feature set, that is, a possible solution. The initialization provides a starting point for the genetic algorithm search space. Through the diversified initial solutions, the possibility of finding the global optimal solution is increased.

[0077] S203: Train an anomaly detection model using different sample feature sets in the sample initial solution set, and obtain a fitness value according to the anomaly detection model training result, wherein the historical multivariate data includes normal data and abnormal data.

[0078] In this embodiment, training the anomaly detection model and evaluating its performance are key steps in optimizing the feature set. Meanwhile, the fitness value reflects the prediction ability of the model under a specific feature combination and is the goal of genetic algorithm optimization.

[0079] In the specific implementation, different feature sets obtained by initialization are used to train the anomaly detection models built based on neural networks, and cross-validation is performed through normal data and abnormal data in historical multivariate data. The fitness value is calculated based on indicators such as the prediction accuracy and recall rate of the model. Through fitness evaluation, the contribution of different feature combinations to the anomaly detection effect can be quantified, providing a basis for the evolutionary selection of the genetic algorithm.

[0080] S204: Filtering out an optimal sample feature set from different sample feature sets based on the fitness value, performing feature matching between the features in the preprocessed multivariate data and the optimal sample feature set, and obtaining an optimal feature set consisting of a plurality of features contained in the preprocessed multivariate data.

[0081] In this embodiment, finding the optimal combination from a large number of feature sets is the core of improving the generalization ability of the model and the accuracy of anomaly detection.

[0082] In the specific implementation, based on the fitness value of the genetic algorithm, the feature set with the best performance is selected as the optimal feature set. Then, these optimal features are matched with the actual features in the preprocessed multivariate data to ensure that the features selected in actual application are consistent with the optimization process, thereby ensuring that the feature set of the final application is optimal, improving the efficiency and accuracy of anomaly detection, and providing a reliable decision-making basis for real-time control of power equipment.

[0083] In this embodiment, reference Figure 4 As shown, step S3 includes the following steps:

[0084] S301: Design the architecture and parameters of the neural network model based on the genetic algorithm, and use the optimal feature set of the sample to train the initial neural network model to obtain an anomaly detection model.

[0085] In this embodiment, the traditional manual parameter adjustment method is time-consuming and inefficient. The genetic algorithm can automatically search the structure and parameter space of the neural network to find the optimal or approximately optimal model configuration, thereby improving the prediction performance and generalization ability of the model.

[0086] In the specific implementation, first, the structure of the neural network (such as the number of layers, the number of nodes in each layer) and parameters (learning rate, regularization coefficient, etc.) are encoded as chromosomes of the genetic algorithm. Then, through the evolutionary process of the genetic algorithm, the performance of the model under different structures and parameter combinations on the training set is continuously evaluated, and the solutions with high fitness are selected for reproduction and mutation, and the model design is gradually optimized. Finally, the best architecture and parameter settings obtained by genetic algorithm optimization are combined with the optimal feature set of samples extracted from historical data to train the neural network model specifically for anomaly detection tasks. Then, it can efficiently explore complex network structures and parameter spaces, reduce manual intervention, obtain high-performance anomaly detection models, and improve detection speed and accuracy.

[0087] S302: Inputting the optimal feature set of the preprocessed multivariate data obtained in real time into the anomaly detection model for forward propagation, and outputting the detection result corresponding to the optimal feature set.

[0088] In this embodiment, through forward propagation, the neural network model can process the input feature data and calculate the abnormality possibility score or category prediction, which is a key step in real-time monitoring of the status of power equipment and identifying abnormalities.

[0089] In the specific implementation, the multivariate data collected in real time and preprocessed are screened according to the optimal feature set obtained in step S2, and then these features are input into the neural network model trained in S301. The model performs forward propagation operations, transmits and transforms data layer by layer, and finally generates one or a group of values ​​in the output layer, representing the probability or score of the data point being normal or abnormal, so as to evaluate the status of the power equipment in real time and quickly, provide abnormal warning signals, and provide a basis for subsequent control decisions.

[0090] S303: Performing an abnormality determination on the detection result based on a preset abnormality threshold, and obtaining an abnormal result.

[0091] In this embodiment, the abnormality threshold is set to convert the continuous output of the model into discrete abnormality / normality judgments to facilitate actual operation and decision-making.

[0092] In specific implementation, according to business needs and model characteristics, an abnormal threshold is set in advance, and the abnormal score or probability output by the model in S302 is compared with the abnormal threshold. If it exceeds the abnormal threshold, it is judged as an abnormal state, otherwise it is normal. Then, a clear and actionable abnormal indication is provided to facilitate triggering the corresponding control strategy and responding to the abnormal conditions of the power equipment in a timely manner.

[0093] In this embodiment, reference Figure 5 As shown, step S4 includes the following steps:

[0094] S401: constructing a control strategy mapping library using a genetic algorithm, wherein the control strategy mapping library includes sample abnormal results and control strategies of a mapping relationship.

[0095] In this embodiment, in order to achieve fast and accurate mapping of abnormal detection results to specific control strategies, improve the system's response speed and processing efficiency to abnormal events, and build an efficient and adaptable control strategy mapping mechanism, genetic algorithms, due to their powerful optimization capabilities, have become an ideal choice for building this control strategy mapping library.

[0096] The specific steps for constructing the control strategy mapping library are as follows: first, according to the working principle and safety specifications of the power equipment, a series of control strategies are preset, and initial parameter sets are defined for these strategies; then, the historical multivariate data of the power equipment is used, combined with the predefined control strategies, and iterative optimization is performed through genetic algorithms. In each generation of genetic algorithms, the fitness of the control strategy is evaluated according to the simulation results of the strategy execution (such as improved system stability, shortened response time, etc.), and the control strategy parameters are gradually optimized through selection, crossover and mutation operations. Then, a mapping relationship is established between the optimized control strategies and the abnormal situations (abnormal results) that they successfully coped with during the simulation process to form a control strategy mapping library. Each abnormal situation has the most effective control strategy or strategy combination that matches it. Finally, the constructed control strategy mapping library ensures that when facing different abnormal scenarios, the most appropriate control strategy or strategy combination can be quickly called, thereby improving the flexibility and accuracy of the system in coping with abnormalities.

[0097] S402: Based on the established mapping relationship, the determined abnormal result is matched with the sample abnormal result of the control strategy mapping library, and the matched sample abnormal result is output.

[0098] In this embodiment, when the real-time monitoring system determines an abnormal situation, it is necessary to immediately find a corresponding control strategy to respond quickly and avoid the spread of the fault.

[0099] In specific implementation, the abnormal results monitored in real time are compared with the sample abnormal results in the control strategy mapping library, and the most similar or consistent sample abnormalities are quickly located through a matching algorithm, where the matching algorithm includes but is not limited to the nearest neighbor matching algorithm, support vector machine matching, and K-nearest neighbor algorithm. Ultimately, rapid retrieval of abnormal detection results to the control strategy mapping library is achieved, providing immediate guidance for subsequent control strategy execution.

[0100] S403: extracting the control strategy corresponding to the matched sample abnormal result, and determining the preset operation for the power equipment according to the extracted control strategy.

[0101] In this embodiment, according to the abnormal results of the successfully matched samples, the corresponding control strategy needs to be directly called so as to quickly take the correct preset operation on the power equipment to ensure the safety of the equipment.

[0102] During the specific implementation, the corresponding control strategy is extracted from the mapping library according to the matched sample abnormal results, and further refined into specific preset operation instructions. Then the specific operation plan for the current sample abnormal results is clarified, providing direct instructions for the automatic control of power equipment.

[0103] S404: Control the electrical equipment according to preset operations.

[0104] In this embodiment, executing the preset operation is a key step to directly intervene in the state of the power equipment and prevent the occurrence of faults.

[0105] During specific implementation, the determined preset operation instructions will be sent to the relevant control unit through the communication network, and the control unit will perform specific operations, such as load adjustment, voltage and frequency regulation, line switching, power factor correction, harmonic suppression, emergency shutdown and reclosing, etc. Real-time equipment control ensures that abnormal situations are properly handled and maintains the stable operation of the power system.

[0106] The working principle of a remote control method for electric equipment provided by the present invention is as follows: the present invention obtains multivariate data of electric equipment in real time and preprocesses the multivariate data to obtain preprocessed multivariate data; then, based on the preprocessed multivariate data, feature selection is performed according to a predefined feature screening rule to obtain an optimal feature set; then the optimal feature set is input into a pre-constructed abnormality detection model, and the abnormal result is determined; finally, based on a control strategy mapping library self-constructed by a genetic algorithm, the determined abnormal result is matched with the control strategy in the control strategy mapping library, the corresponding preset operation of the electric equipment when the abnormal result occurs is determined, and the electric equipment is regulated according to the determined preset operation to ensure the safety of the electric equipment. The present invention forms an efficient and intelligent remote monitoring and control system by integrating real-time data acquisition, feature selection assisted by a genetic algorithm, a neural network abnormality detection model and a control strategy mapping library constructed based on a genetic algorithm, which can not only monitor the multivariate data of the electric equipment in real time, realize fast and accurate identification of abnormalities by using a feature set optimized by a genetic algorithm and a neural network model, but also intelligently match and execute preset operation instructions, effectively ensure the safe operation of the electric equipment, and improve the stability and efficiency of the entire power system.

[0107] Embodiment 2

[0108] The present invention also provides a remote control system for electric equipment, which is applied to the remote control method for electric equipment, referring to Figure 6 As shown, the control system includes:

[0109] The preprocessing module is used to acquire multivariate data of the power equipment in real time and preprocess the multivariate data to obtain preprocessed multivariate data, wherein the multivariate data includes sensor data and environmental data corresponding to the power equipment.

[0110] The optimal feature set determination module is used to select features based on the preprocessed multivariate data according to predefined feature screening rules to obtain the optimal feature set, wherein the feature screening rules are set according to the genetic algorithm in the processing of historical multivariate data of the power equipment.

[0111] The anomaly detection module is used to input the optimal feature set into a pre-built anomaly detection model and determine the abnormal result, wherein the anomaly detection model is a neural network model trained by extracting the optimal feature set of samples from the historical multivariate data of the power equipment based on a genetic algorithm.

[0112] The control strategy matching module is used to match the determined abnormal result with the control strategy in the control strategy mapping library based on the control strategy mapping library self-constructed by using the genetic algorithm, determine the corresponding preset operation of the power equipment when the abnormal result occurs, and regulate the power equipment according to the determined preset operation to ensure the safety of the power equipment.

[0113] The present application is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0114] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, the storage medium including a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0115] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device that includes a series of elements includes not only those elements, but also other elements that are not explicitly listed, or includes elements that are inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device that includes the elements.

Claims

1. A remote control method for electric power equipment, characterized in that: The control method comprises the following steps: Acquire multivariate data of the power equipment in real time, and preprocess the multivariate data to obtain preprocessed multivariate data, wherein the multivariate data includes sensor data and environmental data corresponding to the power equipment; Based on the preprocessed multivariate data, feature selection is performed according to a predefined feature screening rule to obtain an optimal feature set, wherein the feature screening rule is set in the processing of historical multivariate data of the power equipment according to a genetic algorithm; The process of obtaining the optimal feature set specifically includes: Extract sample features from historical multivariate data and use genetic algorithms to encode sample features; Initializing the sample features after feature encoding to obtain a sample initial solution set, wherein each solution in the sample initial solution set represents a sample feature set; Using different sample feature sets in the sample initial solution set to train an anomaly detection model, and obtaining a fitness value according to the anomaly detection model training result, wherein the historical multivariate data includes normal data and abnormal data; Based on the fitness value, a sample optimal feature set is selected from different sample feature sets, and the features in the preprocessed multivariate data are matched with the sample optimal feature set to obtain an optimal feature set consisting of multiple features contained in the preprocessed multivariate data; Inputting the optimal feature set into a pre-built anomaly detection model and determining an abnormal result, wherein the anomaly detection model is a neural network model trained by extracting the optimal feature set of samples from historical multivariate data of the power equipment based on a genetic algorithm; Based on a control strategy mapping library self-constructed using a genetic algorithm, the determined abnormal result is matched with the control strategy in the control strategy mapping library, the corresponding preset operation of the power equipment when the abnormal result occurs is determined, and the power equipment is regulated according to the determined preset operation to ensure the safety of the power equipment.

2. A method for remote control of electric power equipment according to claim 1, characterized in that: The real-time acquisition of multivariate data of the power equipment and preprocessing of the multivariate data to obtain preprocessed multivariate data include: Using data acquisition equipment to collect sensor data and environmental data of the power equipment in real time, wherein the sensor data includes current, voltage, temperature and humidity, and the environmental data includes external climate; Performing data cleaning on sensor data and environmental data collected from power equipment in real time, wherein the data cleaning includes missing value processing, outlier processing, duplicate data processing and consistency checking; The cleaned sensor data and environmental data are combined to obtain preprocessed multivariate data.

3. A method for remote control of electric power equipment according to claim 2, characterized in that: The optimal feature set is input into a pre-built anomaly detection model, and an abnormal result is determined, wherein the anomaly detection model is a neural network model trained by extracting the optimal feature set of samples from the historical multivariate data of the power equipment based on a genetic algorithm, and includes: Design the architecture and parameters of the neural network model based on the genetic algorithm, and use the optimal feature set of the sample to train the initial neural network model to obtain the anomaly detection model; The optimal feature set of the preprocessed multivariate data obtained in real time is input into the anomaly detection model for forward propagation, and the detection result corresponding to the optimal feature set is output; The detection result is judged as abnormal based on a preset abnormal threshold, and an abnormal result is obtained.

4. A method for remote control of electric power equipment according to claim 3, characterized in that: The control strategy mapping library self-constructed based on the genetic algorithm matches the determined abnormal result with the control strategy in the control strategy mapping library, determines the preset operation corresponding to the power equipment when the abnormal result occurs, and regulates the power equipment according to the determined preset operation, including: A control strategy mapping library is constructed using a genetic algorithm, wherein the control strategy mapping library includes sample abnormal results and control strategies of a mapping relationship; Based on the established mapping relationship, the determined abnormal result is matched with the sample abnormal result of the control strategy mapping library, and the matched sample abnormal result is output; Extracting the control strategy corresponding to the matched sample abnormal result, and determining the preset operation for the power equipment according to the extracted control strategy; Control electrical equipment according to preset operations.

5. A method for remote control of electric power equipment according to claim 4, characterized in that: The control strategy mapping library is constructed by using a genetic algorithm, wherein the control strategy mapping library includes sample abnormal results and control strategies of a mapping relationship, including: Predefine control strategies and perform initialization operations; Based on the control strategy, the genetic algorithm is iteratively optimized in combination with the historical multivariate data of the power equipment to obtain the sample abnormality results; A mapping relationship is established between the iteratively optimized control strategy and the sample abnormality results to form a control strategy mapping library, where each sample abnormality corresponds to a control strategy or a combination of multiple control strategies.

6. A method for remote control of electric power equipment according to claim 5, characterized in that: The preset operation includes one or a combination of load adjustment, voltage and frequency regulation, line switching, power factor correction, harmonic suppression, emergency shutdown and reclosing.

7. A remote control system for electric power equipment, applied to a remote control method for electric power equipment as claimed in any one of claims 1 to 6, characterized in that: The control system comprises: A preprocessing module, used for acquiring multivariate data of the power equipment in real time, and preprocessing the multivariate data to obtain preprocessed multivariate data, wherein the multivariate data includes sensor data and environmental data corresponding to the power equipment; An optimal feature set determination module, used for selecting features based on the preprocessed multivariate data according to predefined feature screening rules to obtain an optimal feature set, wherein the feature screening rules are set according to a genetic algorithm in processing historical multivariate data of the power equipment; An anomaly detection module, used for inputting the optimal feature set into a pre-built anomaly detection model and determining an abnormal result, wherein the anomaly detection model is a neural network model trained by extracting the optimal feature set of samples from the historical multivariate data of the power equipment based on a genetic algorithm; The control strategy matching module is used to match the determined abnormal result with the control strategy in the control strategy mapping library based on the control strategy mapping library self-constructed by using the genetic algorithm, determine the corresponding preset operation of the power equipment when the abnormal result occurs, and regulate the power equipment according to the determined preset operation to ensure the safety of the power equipment.

Images