An intelligent monitoring method for power reactive power compensation controller

Abstract

The present invention discloses an intelligent monitoring method for a reactive power compensation controller, the method includes data collection, data preprocessing, real-time monitoring of reactive power compensation controller, hyperparameter optimization and intelligent early warning. The present invention relates to the field of electric power information technology, and specifically refers to an intelligent monitoring method for reactive power compensation controller. This solution innovatively proposes an intelligent compensation strategy generation and remote monitoring adjustment mechanism, which effectively solves the intelligent monitoring and strategy execution problems of reactive power compensation controllers in various scenarios; a dynamic compensation monitoring activation function and a reactive power compensation monitoring loss function are designed, abnormal data are processed in detail, and the accuracy, robustness and complex scenario processing capabilities of the model are improved; the algorithm for obtaining the optimal parameters of the monitoring model is improved through individual mutations of survival factors and periodic fluctuation strategies, thereby obtaining the optimal parameter combination and improving the accuracy of the model results.

Description

Technical Field

[0001] The invention relates to the technical field of electric power informatization, and in particular to an intelligent monitoring method for an electric power reactive power compensation controller. Background Art

[0002] With the rapid development of power systems, reactive power compensation technology plays an important role in improving the stability of power systems and improving the quality of electric energy. The efficient and stable operation of power systems requires real-time and accurate reactive power compensation and power monitoring, so intelligent reactive power compensation controller monitoring methods have emerged; however, traditional power reactive power compensation controller monitoring methods have the problems of single compensation strategy and inability to adapt to diverse fault scenarios, which may result in the compensation scheme may not be able to effectively solve the actual power grid needs; the limitations of activation functions and loss functions in existing monitoring models for power reactive power compensation controllers lead to technical problems such as inaccurate monitoring results and unclear identification of abnormal types; the existing monitoring models have the problem of inappropriate built-in parameter settings, which leads to inaccurate model monitoring results. Summary of the invention

[0003] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides an intelligent monitoring method for reactive power compensation controller. In view of the problems in the traditional monitoring method for reactive power compensation controller that the compensation strategy is single and cannot adapt to diversified fault scenarios, resulting in the compensation scheme may not be able to effectively solve the actual power grid needs. This scheme innovatively proposes an intelligent compensation strategy generation and remote monitoring and adjustment mechanism. By real-time monitoring of the power system, flexible compensation rules and early warning mechanisms are established. It can automatically generate corresponding compensation measures according to different abnormal types, and support maintenance personnel to remotely adjust in real time to ensure the flexibility and effectiveness of the response measures, thereby effectively solving the intelligent monitoring and strategy execution problems of reactive power compensation controllers in various scenarios; in view of the activation function and loss function in the existing monitoring model for reactive power compensation controllers In order to solve the technical problems of inaccurate monitoring results and unclear identification of abnormal types due to the limitations of the existing numbers, this solution innovatively designs the dynamic compensation monitoring activation function and the reactive power compensation monitoring loss function, which can more accurately identify the deviation between the monitoring model output and the actual power monitoring value, and carefully process the abnormal data, achieving a good balance between calculation accuracy and efficiency, thereby significantly improving the robustness, accuracy and complex scenario processing capabilities of the model, and effectively solving the problem of insufficient accuracy of monitoring results; in view of the technical problem that the existing monitoring models have inappropriate built-in parameter settings, resulting in inaccurate model monitoring results, this solution improves the algorithm for obtaining the optimal parameters of the monitoring model through individual mutation of the survival factor and periodic fluctuation strategy, so as to obtain the optimal parameter combination and improve the accuracy of the monitoring model results.

[0004] The technical solution adopted by the present invention is as follows: The present invention provides an intelligent power reactive power compensation controller monitoring method, the method comprising the following steps:

[0005] Step S1: data collection;

[0006] Step S2: data preprocessing;

[0007] Step S3: Real-time monitoring by the power reactive power compensation controller;

[0008] Step S4: hyperparameter optimization;

[0009] Step S5: Intelligent early warning.

[0010] Furthermore, in step S1, the data collection is specifically to obtain the original data monitored by the controller from the power system by collection;

[0011] The controller monitoring raw data includes historical monitoring data and real-time monitoring data; the historical monitoring data and real-time monitoring data both include voltage data, current data, power factor data, active power data, reactive power data and harmonic data; the historical monitoring data also includes the power reactive power compensation controller monitoring abnormality type and the compensation treatment measures data corresponding to the abnormality type; the power reactive power compensation controller monitoring abnormality type and monitoring normality are used as data labels.

[0012] Furthermore, in step S2, the data preprocessing is used to perform data cleaning, standardization and feature selection on the original data monitored by the controller to obtain preliminary controller monitoring data; the data cleaning is to process missing values, abnormal values ​​and duplicate values; the standardization is to standardize the data based on the maximum and minimum normalization method; the feature selection is to use the correlation analysis method to screen out features related to reactive power compensation control from the original data monitored by the controller.

[0013] Further, in step S3, the power reactive power compensation controller is monitored in real time, specifically by establishing a real-time monitoring model of the controller and training the monitoring model with historical data, using the real-time monitoring data in the preliminary monitoring data of the controller as input data of the trained monitoring model, and obtaining the real-time monitoring result of the power reactive power compensation controller; comprising the following steps:

[0014] Step S31: Establishing a controller real-time monitoring model, specifically including the following steps:

[0015] Step S311: Initialize the model internal structure LM, the formula used is as follows:

[0016] ;

[0017] In the formula, f t represents the output of the forget gate, X t It indicates input data. Indicates the hidden state at the previous moment, W xf and W hf They represent the weight matrices corresponding to the forget gate processing input data and processing hidden states, respectively, and b f represents the bias parameter of the forget gate, represents the Sigmoid function, i t represents the output of the input gate, W xi and W hi They represent the weight matrices corresponding to the input gate processing input data and processing hidden states, respectively, and b i represents the bias parameter of the input gate, represents the candidate value of the new cell state, W xc and W hc They represent the weight matrices for processing input data and hidden states of corresponding cell states, and b c The bias parameter representing the cell state, c t Indicates the current cell state, c t-1 Indicates the cell state at the previous moment, o t represents the output of the output gate, W xo and W ho They represent the weight matrices corresponding to the output gate processing input data and processing hidden states, respectively, and b o Represents the bias parameter of the output gate, h t Indicates the hidden state at the current moment, It is the hyperbolic tangent function;

[0018] Step S312: Calculate the bidirectional model output using the following formula:

[0019] ;

[0020] ;

[0021] ;

[0022] Where h represents the hidden state of the combined forward and backward LM, Indicates the hidden state in the positive direction at the current moment, Indicates the hidden state in the positive direction at the previous moment, Indicates the reverse hidden state at the current moment. Indicates the hidden state in the reverse direction at the next moment;

[0023] Step S313: Design a dynamic compensation monitoring activation function; the formula used is as follows:

[0024] ;

[0025] In the formula, represents the dynamic compensation monitoring activation function, x represents the input variable of the allocation activation function, Represents the parameters that adjust the shape of the activation function and control the sensitivity of the nonlinear mapping. Represents the scaling factor, which is used to adjust the magnitude of the output value. Used to compress the absolute value of the input. Represents a dynamically adjusted parameter used to control the nonlinear response of the activation function;

[0026] Step S314: Calculate the output layer prediction value, the formula used is as follows:

[0027] ;

[0028] In the formula, represents the predicted value of the monitoring model, represents the weight matrix connecting the forward state to the output, represents the weight matrix connecting the reverse state to the output, b y Represents the bias parameter of the output layer;

[0029] Step S315: Design reactive power compensation monitoring loss function; the formula used is as follows:

[0030] ;

[0031] In the formula, represents the reactive power compensation monitoring loss function value, N represents the total number of samples used to calculate the loss function, represents the time weight parameter, which is used to control the impact of data at different time points on the loss. i represents the sample index. Represents the true value at the current time point, Represents the conformity function, which is used to detect the direction of the true value and the predicted value. Represents the abnormal penalty coefficient, which is used to impose additional penalties on abnormal samples when the prediction results exceed a specific threshold; Represents the trend term weight coefficient, which is used to adjust the impact of the trend error term on the overall loss. Represents the trend error term, which is used to measure whether the trends of the predicted value and the true value are consistent over a period of time;

[0032] Step S32: training the controller real-time monitoring model, specifically using the historical monitoring data to train the controller real-time monitoring model to obtain a trained controller real-time monitoring model;

[0033] Step S33: Real-time monitoring of reactive power compensation, specifically, using the real-time monitoring data as input data of the trained controller real-time monitoring model to obtain real-time monitoring results of the reactive power compensation controller.

[0034] Furthermore, in step S4, the hyperparameter optimization specifically includes the following steps:

[0035] Step S41: Initializing parameters, specifically by constructing initial algorithm parameters; the initial algorithm parameters include population size N, dimension d of search space and maximum number of iterations T;

[0036] Step S42: Initialize the population, specifically, all search individuals are randomly initialized in the feasible search space; the formula used is as follows:

[0037] ;

[0038] ;

[0039] ;

[0040] In the formula, represents the position of the jth dimension of the i-th individual, and r represents the position between A random number uniformly distributed in the range, and Respectively represent the search lower and upper limits of dimension j;

[0041] Step S43: Population division, calculation of individual fitness values ​​f in the population i ; The controller based on the individual position monitors the model performance in real time as the individual fitness value;

[0042] Step S44: Design the survival factor, the formula used is as follows:

[0043] ;

[0044] In the formula, represents the individual survival factor of the population, e represents the natural base number, represents a dynamic parameter used to control the growth rate of the survival factor. Indicates the current iteration number;

[0045] Step S45: Individual survival stage, specifically using the periodic fluctuation strategy to adjust the individual position, the formula used is as follows:

[0046] ;

[0047] In the formula, represents the j-th dimension position of the i-th individual in the t+1-th generation population, represents the j-th dimension position of the i-th individual in the t-th generation population, and Indicates between A random number uniformly distributed in the range, Indicates between A random number uniformly distributed in the range, represents the current global best individual position, Represents the current global optimal fitness value;

[0048] Step S46: Escape from the arrest phase, the formula used is as follows:

[0049] ;

[0050] ;

[0051] In the formula, represents the j-th dimension escape position of the i-th individual in the t+1-th generation population, and Indicates between A random number uniformly distributed in the range, represents the fitness function;

[0052] Step S47: Individual mutation stage, specifically, individual mutation is performed through survival factors, and the formula used is as follows:

[0053] ;

[0054] Step S48: search determination, specifically, by constructing a search termination condition, performing search determination on the optimal individual position, and obtaining the optimal individual position data setting;

[0055] The search termination conditions include threshold termination and iteration termination;

[0056] The threshold termination is specifically to set the fitness threshold, when the individual fitness value f i When it is above the fitness threshold, the hyperparameter search is completed;

[0057] The iteration termination specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached;

[0058] The optimal individual position includes a controller monitoring model parameters in real time.

[0059] Further, in step S5, the intelligent early warning is specifically to remind the maintenance personnel according to the output result of the controller real-time monitoring model; including the following steps:

[0060] Step S51: the monitor performs real-time supplementation, specifically, judging according to the output result of the real-time monitoring model of the controller, and automatically triggering the intelligent early warning mechanism if the output result is not a normal monitoring state, reminding the maintenance personnel to check the state; and generating a supplementary strategy according to the compensation treatment measure data implemented corresponding to the abnormal type; the supplementary strategy includes adjusting the parameters of the reactive power compensation equipment, reconfiguring the load distribution of the power grid and starting the backup power supply content;

[0061] Step S52: remote monitoring and adjustment, specifically, maintenance personnel remotely view the current controller status via mobile phone, and manually review and adjust the automatically generated compensation strategy based on the output results of the model combined with real-time data to ensure the accuracy and effectiveness of the supplementary strategy.

[0062] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0063] (1) In view of the problems in traditional power reactive power compensation controller monitoring methods, such as a single compensation strategy and inability to adapt to diverse fault scenarios, which may result in the compensation scheme being unable to effectively meet the actual power grid needs, this scheme innovatively proposes an intelligent compensation strategy generation and remote monitoring and adjustment mechanism. By real-time monitoring of the power system, flexible compensation rules and early warning mechanisms are established. It can automatically generate corresponding compensation measures according to different abnormal types, and support maintenance personnel to make remote real-time adjustments to ensure the flexibility and effectiveness of the response measures, thereby effectively solving the intelligent monitoring and strategy execution problems of power reactive power compensation controllers in various scenarios.

[0064] (2) In order to address the technical problems of inaccurate monitoring results and unclear identification of abnormal types due to the limitations of activation functions and loss functions in the existing monitoring models for reactive power compensation controllers, this scheme innovatively designs dynamic compensation monitoring activation functions and reactive power compensation monitoring loss functions, which can more accurately identify the deviation between the monitoring model output and the actual power monitoring value, and carefully process abnormal data, achieving a good balance between calculation accuracy and efficiency, thereby significantly improving the model's robustness, accuracy and ability to handle complex scenarios, and effectively solving the problem of insufficient accuracy of monitoring results.

[0065] (3) In order to address the technical problem that the existing monitoring models have inappropriate built-in parameter settings, which leads to inaccurate model monitoring results, this scheme improves the algorithm for obtaining the optimal parameters of the monitoring model through individual mutation and periodic fluctuation strategies of survival factors, thereby obtaining the optimal parameter combination and improving the accuracy of the monitoring model results. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 A schematic flow chart of an intelligent power reactive power compensation controller monitoring method provided by the present invention;

[0067] Figure 2 is a schematic flow chart of step S3;

[0068] Figure 3 is a schematic flow chart of step S4;

[0069] Figure 4 is a schematic flow chart of step S5;

[0070] Figure 5 is a flow chart of step S31;

[0071] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0072] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0073] In the description of the present invention, it is necessary to understand that terms such as “upper”, “lower”, “front”, “back”, “left”, “right”, “top”, “bottom”, “inside” and “outside” indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred system or element must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as a limitation on the present invention.

[0074] Example 1, see Figure 1 The technical solution adopted by the present invention is as follows: The present invention provides an intelligent power reactive power compensation controller monitoring method, the method comprising the following steps:

[0075] Step S1: data collection, by collecting historical monitoring data and real-time monitoring data from the power system to obtain the controller monitoring raw data;

[0076] Step S2: data preprocessing, by performing data cleaning, standardization and feature selection on the original data monitored by the controller, to obtain preliminary data of the controller monitoring;

[0077] Step S3: Real-time monitoring of the reactive power compensation controller, by establishing a real-time monitoring model of the controller and training the monitoring model with historical data, using the real-time monitoring data as input data of the trained monitoring model, and obtaining the real-time monitoring result of the reactive power compensation controller;

[0078] Step S4: Hyperparameter optimization, by using the survival factor to perform individual mutation and periodic fluctuation strategies, optimize the hyperparameters of the monitoring model and obtain the optimal parameter combination of the model;

[0079] Step S5: Intelligent early warning, the output results of the real-time monitoring model are automatically triggered by the controller to trigger the early warning mechanism and generate a supplementary strategy. The maintenance personnel remotely monitor the controller status through the mobile phone and manually review and adjust the automatically generated supplementary strategy based on the real-time data.

[0080] Example 2, see Figure 1 , This embodiment is based on the above embodiment. In step S1, the data collection is specifically to obtain the original data monitored by the controller from the power system through collection;

[0081] The controller monitoring raw data includes historical monitoring data and real-time monitoring data; the historical monitoring data and real-time monitoring data both include voltage data, current data, power factor data, active power data, reactive power data and harmonic data; the historical monitoring data also includes the power reactive power compensation controller monitoring abnormality type and the compensation treatment measures data corresponding to the abnormality type; the power reactive power compensation controller monitoring abnormality type and monitoring normality are used as data labels.

[0082] Example 3, see Figure 1 This embodiment is based on the above embodiment. In step S2, the data preprocessing is used to perform data cleaning, standardization processing and feature selection on the original data monitored by the controller to obtain preliminary controller monitoring data; the data cleaning is to process missing values, abnormal values ​​and repeated values; the standardization processing is to standardize the data based on the maximum and minimum normalization method; the feature selection is to use the correlation analysis method to screen out the features related to reactive power compensation control from the original data monitored by the controller.

[0083] Example 4, see Figure 1 , Figure 2 and Figure 5 This embodiment is based on the above embodiment. In step S3, the power reactive power compensation controller performs real-time monitoring, specifically by establishing a real-time monitoring model for the controller and training the monitoring model with historical data, and using the real-time monitoring data in the preliminary monitoring data of the controller as input data of the trained monitoring model to obtain the real-time monitoring result of the power reactive power compensation controller; the embodiment includes the following steps:

[0084] Step S31: Establishing a controller real-time monitoring model, specifically including the following steps:

[0085] Step S311: Initialize the model internal structure LM, the formula used is as follows:

[0086] ;

[0087] In the formula, f t represents the output of the forget gate, X t It indicates input data. Indicates the hidden state at the previous moment, W xf and W hf They represent the weight matrices corresponding to the forget gate processing input data and processing hidden states, respectively, and b f represents the bias parameter of the forget gate, represents the Sigmoid function, i t represents the output of the input gate, W xi and W hi They represent the weight matrices corresponding to the input gate processing input data and processing hidden states, respectively, and b i represents the bias parameter of the input gate, represents the candidate value of the new cell state, W xc and W hc They represent the weight matrices for processing input data and hidden states of corresponding cell states, and b c The bias parameter representing the cell state, c t Indicates the current cell state, c t-1 Indicates the cell state at the previous moment, o t represents the output of the output gate, W xo and W ho They represent the weight matrices corresponding to the output gate processing input data and processing hidden states, respectively, and b o Represents the bias parameter of the output gate, h t Indicates the hidden state at the current moment, It is the hyperbolic tangent function;

[0088] Step S312: Calculate the bidirectional model output using the following formula:

[0089] ;

[0090] ;

[0091] ;

[0092] Where h represents the hidden state of the combined forward and backward LM, Indicates the hidden state in the positive direction at the current moment, Indicates the hidden state in the positive direction at the previous moment, Indicates the reverse hidden state at the current moment. Indicates the hidden state in the reverse direction at the next moment;

[0093] Step S313: Design a dynamic compensation monitoring activation function; the formula used is as follows:

[0094] ;

[0095] In the formula, represents the dynamic compensation monitoring activation function, x represents the input variable of the allocation activation function, Represents the parameters that adjust the shape of the activation function and control the sensitivity of the nonlinear mapping. Represents the scaling factor, which is used to adjust the magnitude of the output value. Used to compress the absolute value of the input. Represents a dynamically adjusted parameter used to control the nonlinear response of the activation function;

[0096] Step S314: Calculate the output layer prediction value, the formula used is as follows:

[0097] ;

[0098] In the formula, represents the predicted value of the monitoring model, represents the weight matrix connecting the forward state to the output, represents the weight matrix connecting the reverse state to the output, b y Represents the bias parameter of the output layer;

[0099] Step S315: Design reactive power compensation monitoring loss function; the formula used is as follows:

[0100] ;

[0101] In the formula, represents the reactive power compensation monitoring loss function value, N represents the total number of samples used to calculate the loss function, represents the time weight parameter, which is used to control the impact of data at different time points on the loss. i represents the sample index. Represents the true value at the current time point, Represents the conformity function, which is used to detect the direction of the true value and the predicted value. Represents the abnormal penalty coefficient, which is used to impose additional penalties on abnormal samples when the prediction results exceed a specific threshold; Represents the trend term weight coefficient, which is used to adjust the impact of the trend error term on the overall loss. Represents the trend error term, which is used to measure whether the trends of the predicted value and the true value are consistent over a period of time;

[0102] Step S32: training the controller real-time monitoring model, specifically using the historical monitoring data to train the controller real-time monitoring model to obtain a trained controller real-time monitoring model;

[0103] Step S33: Real-time monitoring of reactive power compensation, specifically, using the real-time monitoring data as input data of the trained controller real-time monitoring model to obtain real-time monitoring results of the reactive power compensation controller.

[0104] By performing the above operations, this solution innovatively designs dynamic compensation monitoring activation functions and reactive power compensation monitoring loss functions to address the technical problems of inaccurate monitoring results and unclear identification of abnormal types caused by the limitations of activation functions and loss functions in the existing monitoring models for reactive power compensation controllers. These functions can more accurately identify the deviation between the monitoring model output and the actual power monitoring value, and process abnormal data in detail, achieving a good balance between calculation accuracy and efficiency, thereby significantly improving the model's robustness, accuracy, and ability to handle complex scenarios, and effectively solving the problem of insufficient accuracy of monitoring results.

[0105] Example 5, see Figure 1 and Figure 3 This embodiment is based on the above embodiment. In step S4, the hyperparameter optimization specifically includes the following steps:

[0106] Step S41: Initializing parameters, specifically by constructing initial algorithm parameters; the initial algorithm parameters include population size N, dimension d of search space and maximum number of iterations T;

[0107] Step S42: Initialize the population, specifically, all search individuals are randomly initialized in the feasible search space; the formula used is as follows:

[0108] ;

[0109] ;

[0110] ;

[0111] In the formula, represents the position of the jth dimension of the i-th individual, and r represents the position between A random number uniformly distributed in the range, and Respectively represent the search lower and upper limits of dimension j;

[0112] Step S43: Population division, calculation of individual fitness values ​​f in the population i ; The controller based on the individual position monitors the model performance in real time as the individual fitness value;

[0113] Step S44: Design the survival factor, the formula used is as follows:

[0114] ;

[0115] In the formula, represents the individual survival factor of the population, e represents the natural base number, represents a dynamic parameter used to control the growth rate of the survival factor. Indicates the current iteration number;

[0116] Step S45: Individual survival stage, specifically using the periodic fluctuation strategy to adjust the individual position, the formula used is as follows:

[0117] ;

[0118] In the formula, represents the j-th dimension position of the i-th individual in the t+1-th generation population, represents the j-th dimension position of the i-th individual in the t-th generation population, and Indicates between A random number uniformly distributed in the range, Indicates between A random number uniformly distributed in the range, represents the current global best individual position, Represents the current global optimal fitness value;

[0119] Step S46: Escape from the arrest phase, the formula used is as follows:

[0120] ;

[0121] ;

[0122] In the formula, represents the j-th dimension escape position of the i-th individual in the t+1-th generation population, and Indicates between A random number uniformly distributed in the range, represents the fitness function;

[0123] Step S47: Individual mutation stage, specifically, individual mutation is performed through survival factors, and the formula used is as follows:

[0124] ;

[0125] Step S48: search determination, specifically, by constructing a search termination condition, performing search determination on the optimal individual position, and obtaining the optimal individual position data setting;

[0126] The search termination conditions include threshold termination and iteration termination;

[0127] The threshold termination is specifically to set the fitness threshold, when the individual fitness value f i When it is above the fitness threshold, the hyperparameter search is completed;

[0128] The iteration termination specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached;

[0129] The optimal individual position includes a controller monitoring model parameters in real time.

[0130] By performing the above operations, in order to address the technical problem that the existing monitoring models have inappropriate built-in parameter settings, resulting in inaccurate model monitoring results, this solution improves the algorithm for obtaining the optimal parameters of the monitoring model through individual mutation of survival factors and periodic fluctuation strategies, thereby obtaining the optimal parameter combination and improving the accuracy of the monitoring model results.

[0131] Example 6, see Figure 1 and Figure 4 This embodiment is based on the above embodiment. In step S5, the intelligent early warning is specifically to remind the maintenance personnel according to the output result of the controller real-time monitoring model; it includes the following steps:

[0132] Step S51: the monitor performs real-time supplementation, specifically, judging according to the output result of the real-time monitoring model of the controller, and automatically triggering the intelligent early warning mechanism if the output result is not a normal monitoring state, reminding the maintenance personnel to check the state; and generating a supplementary strategy according to the compensation treatment measure data implemented corresponding to the abnormal type; the supplementary strategy includes adjusting the parameters of the reactive power compensation equipment, reconfiguring the load distribution of the power grid and starting the backup power supply content;

[0133] Step S52: remote monitoring and adjustment, specifically, maintenance personnel remotely view the current controller status via mobile phone, and manually review and adjust the automatically generated compensation strategy based on the output results of the model combined with real-time data to ensure the accuracy and effectiveness of the supplementary strategy.

[0134] By performing the above operations, in view of the problems in the traditional power reactive power compensation controller monitoring method that the compensation strategy is single and cannot adapt to various fault scenarios, which may lead to the compensation scheme may not be able to effectively solve the actual power grid needs, this scheme innovatively proposes an intelligent compensation strategy generation and remote monitoring and adjustment mechanism. Through real-time monitoring of the power system, flexible compensation rules and early warning mechanisms are established. It can automatically generate corresponding compensation measures according to different abnormal types, and support maintenance personnel to make remote real-time adjustments to ensure the flexibility and effectiveness of the response measures, thereby effectively solving the intelligent monitoring and strategy execution problems of the power reactive power compensation controller in various scenarios.

[0135] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0136] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0137] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. An intelligent monitoring method for a reactive power compensation controller, characterized in that: The method comprises the following steps: Step S1: data collection, by collecting historical monitoring data and real-time monitoring data from the power system to obtain the controller monitoring raw data; Step S2: data preprocessing, by performing data cleaning, standardization and feature selection on the original data monitored by the controller, to obtain preliminary data of the controller monitoring; Step S3: Real-time monitoring of the reactive power compensation controller, establishing a real-time monitoring model of the controller by designing a dynamic compensation monitoring activation function and a reactive power compensation monitoring loss function, and training the monitoring model with historical data, using the real-time monitoring data as input data of the trained monitoring model, and obtaining the real-time monitoring result of the reactive power compensation controller; Step S4: Hyperparameter optimization, by using the survival factor to perform individual mutation and periodic fluctuation strategies, optimize the hyperparameters of the monitoring model and obtain the optimal parameter combination of the model; Step S5: Intelligent early warning, the output results of the real-time monitoring model are automatically triggered by the controller to trigger the early warning mechanism and generate a supplementary strategy. The maintenance personnel remotely monitor the controller status through the mobile phone and manually review and adjust the automatically generated supplementary strategy based on the real-time data.

2. The intelligent power reactive power compensation controller monitoring method according to claim 1 is characterized in that: In step S3, the power reactive power compensation controller performs real-time monitoring, which specifically includes the following steps: Step S31: Establishing a controller real-time monitoring model, specifically including the following steps: Step S311: Initialize the model internal structure LM, the formula used is as follows: ； In the formula, f t represents the output of the forget gate, X t It indicates input data. Indicates the hidden state at the previous moment, W xf and W hf They represent the weight matrices corresponding to the forget gate processing input data and processing hidden states, respectively, and b f represents the bias parameter of the forget gate, represents the Sigmoid function, i t represents the output of the input gate, W xi and W hi They represent the weight matrices corresponding to the input gate processing input data and processing hidden states, respectively, and b i represents the bias parameter of the input gate, represents the candidate value of the new cell state, W xc and W hc They represent the weight matrices for processing input data and hidden states of corresponding cell states, and b c The bias parameter representing the cell state, c t Indicates the current cell state, c t-1 Indicates the cell state at the previous moment, o t represents the output of the output gate, W xo and W ho They represent the weight matrices corresponding to the output gate processing input data and processing hidden states, respectively, and b o Represents the bias parameter of the output gate, h t Indicates the hidden state at the current moment, It is the hyperbolic tangent function; Step S312: Calculate the bidirectional model output using the following formula: ； ； ； Where h represents the hidden state of the combined forward and backward LM, Indicates the hidden state in the positive direction at the current moment, Indicates the hidden state in the positive direction at the previous moment, Indicates the reverse hidden state at the current moment. Indicates the hidden state in the reverse direction at the next moment; Step S313: Design a dynamic compensation monitoring activation function; the formula used is as follows: ； In the formula, represents the dynamic compensation monitoring activation function, x represents the input variable of the allocation activation function, Represents the parameters that adjust the shape of the activation function and control the sensitivity of the nonlinear mapping. Represents the scaling factor, which is used to adjust the magnitude of the output value. Used to compress the absolute value of the input. Represents a dynamically adjusted parameter used to control the nonlinear response of the activation function; Step S314: Calculate the output layer prediction value, the formula used is as follows: ； In the formula, represents the predicted value of the monitoring model, represents the weight matrix connecting the forward state to the output, represents the weight matrix connecting the reverse state to the output, b y Represents the bias parameter of the output layer; Step S315: Design reactive power compensation monitoring loss function; the formula used is as follows: ； In the formula, represents the reactive power compensation monitoring loss function value, N represents the total number of samples used to calculate the loss function, represents the time weight parameter, which is used to control the impact of data at different time points on the loss. i represents the sample index. Represents the true value at the current time point, Represents the conformity function, which is used to detect the direction of the true value and the predicted value. Represents the abnormal penalty coefficient, which is used to impose additional penalties on abnormal samples when the prediction results exceed a specific threshold; Represents the trend term weight coefficient, which is used to adjust the impact of the trend error term on the overall loss. Represents the trend error term, which is used to measure whether the trends of the predicted value and the true value are consistent over a period of time; Step S32: training the controller real-time monitoring model, specifically using the historical monitoring data to train the controller real-time monitoring model to obtain a trained controller real-time monitoring model; Step S33: Real-time monitoring of reactive power compensation, specifically, using the real-time monitoring data as input data of the trained controller real-time monitoring model to obtain real-time monitoring results of the reactive power compensation controller.

3. The intelligent power reactive power compensation controller monitoring method according to claim 1 is characterized in that: In step S4, the hyperparameter optimization specifically includes the following steps: Step S41: Initializing parameters, specifically by constructing initial algorithm parameters; the initial algorithm parameters include population size N, dimension d of search space and maximum number of iterations T; Step S42: Initialize the population, specifically, all search individuals are randomly initialized in the feasible search space; the formula used is as follows: ； ； ； In the formula, represents the position of the jth dimension of the i-th individual, and r represents the position between A random number uniformly distributed in the range, and Respectively represent the search lower and upper limits of dimension j; Step S43: Population division, calculation of individual fitness values ​​f in the population i ; The controller based on the individual position monitors the model performance in real time as the individual fitness value; Step S44: Design the survival factor, the formula used is as follows: ； In the formula, represents the individual survival factor of the population, e represents the natural base number, represents a dynamic parameter used to control the growth rate of the survival factor. Indicates the current iteration number; Step S45: Individual survival stage, specifically using the periodic fluctuation strategy to adjust the individual position, the formula used is as follows: ； In the formula, represents the j-th dimension position of the i-th individual in the t+1-th generation population, represents the j-th dimension position of the i-th individual in the t-th generation population, and Indicates between A random number uniformly distributed in the range, Indicates between A random number uniformly distributed in the range, represents the current global best individual position, Represents the current global optimal fitness value; Step S46: Escape from the arrest phase, the formula used is as follows: ； ； In the formula, represents the j-th dimension escape position of the i-th individual in the t+1-th generation population, and Indicates between A random number uniformly distributed in the range, represents the fitness function; Step S47: Individual mutation stage, specifically, individual mutation is performed through survival factors, and the formula used is as follows: ； Step S48: search determination, specifically, by constructing a search termination condition, performing search determination on the optimal individual position, and obtaining the optimal individual position data setting; The search termination conditions include threshold termination and iteration termination; The threshold termination is specifically to set the fitness threshold, when the individual fitness value f i When it is above the fitness threshold, the hyperparameter search is completed; The iteration termination specifically refers to terminating the iteration and obtaining the optimal individual position when the maximum number of iterations is reached; The optimal individual position includes a controller monitoring model parameters in real time.

4. The intelligent power reactive power compensation controller monitoring method according to claim 1 is characterized in that: In step S5, the intelligent early warning is specifically to remind the maintenance personnel according to the output result of the controller real-time monitoring model; including the following steps: Step S51: The monitor performs real-time supplementation, specifically, judging according to the output result of the real-time monitoring model of the controller. If the output result is not a normal monitoring state, the intelligent early warning mechanism is automatically triggered to remind the maintenance personnel to check the state; and a supplementary strategy is generated according to the compensation treatment measure data implemented corresponding to the abnormal type; the supplementary strategy includes adjusting the parameters of the reactive power compensation equipment, reconfiguring the load distribution of the power grid and starting the backup power supply content; Step S52: remote monitoring and adjustment, specifically, maintenance personnel remotely view the current controller status via mobile phone, and manually review and adjust the automatically generated compensation strategy based on the output results of the model combined with real-time data to ensure the accuracy and effectiveness of the supplementary strategy.

5. The intelligent power reactive power compensation controller monitoring method according to claim 1 is characterized in that: In step S1, the data collection is specifically to obtain the original data monitored by the controller from the power system through collection; The controller monitoring raw data includes historical monitoring data and real-time monitoring data; the historical monitoring data and real-time monitoring data both include voltage data, current data, power factor data, active power data, reactive power data and harmonic data; the historical monitoring data also includes the power reactive power compensation controller monitoring abnormality type and the compensation treatment measures data corresponding to the abnormality type; the power reactive power compensation controller monitoring abnormality type and monitoring normality are used as data labels.

6. The intelligent power reactive power compensation controller monitoring method according to claim 1 is characterized in that: In step S2, the data preprocessing is used to perform data cleaning, standardization processing and feature selection on the original data monitored by the controller to obtain preliminary data for controller monitoring; the data cleaning is to process missing values, abnormal values ​​and duplicate values; the standardization processing is to standardize the data based on the maximum and minimum normalization method; The feature selection is to use the correlation analysis method to screen out the features related to reactive power compensation control from the original data monitored by the controller.

Images