Intelligent scheduling management system and method for power grid communication maintenance plan

Abstract

The invention discloses an intelligent scheduling management system and method for a power grid communication maintenance plan, and belongs to the field of intelligent scheduling management, and the method comprises the steps: collecting the historical operation data of power grid communication, and carrying out the preprocessing, and obtaining a sample data set; performing feature screening on the sample data set based on an LASSO regression model to obtain key features; constructing a prediction model by using the screened key features, and training and evaluating the prediction model by using the divided training sample set and verification sample set to obtain a fault prediction model; inputting the real-time operation data into the fault prediction model, obtaining the prediction probability of the equipment fault, and comparing the prediction probability with a preset probability threshold to judge whether the equipment has a fault or not; the method comprises the following steps: establishing a maintenance plan scheduling optimization model by setting an optimization target and a constraint condition, and solving by adopting a heuristic algorithm to generate an optimal maintenance plan scheduling scheme; and outputting the optimal maintenance plan scheduling scheme to the system, and executing the optimal maintenance plan scheduling scheme after the optimal maintenance plan scheduling scheme is audited and adjusted by a manager.

Description

Technical Field

[0001] The present invention relates to the field of intelligent scheduling management, and in particular to an intelligent scheduling management system and method for a power grid communication maintenance plan. Background Art

[0002] With the continuous development of power systems and the expansion of power grid scale, the stable operation of power grid communication equipment is crucial to ensure the reliability and safety of power systems. Traditional power grid communication equipment maintenance mainly relies on manual experience and regular inspections, which has the following problems:

[0003] 1. Insufficient data-driven decision support: Traditional methods often lack in-depth analysis and utilization of historical operation data, and cannot fully utilize the information contained in the data to guide maintenance decisions;

[0004] 2. Low fault prediction accuracy: Due to the lack of an effective prediction model, it is difficult to accurately predict when equipment will fail, resulting in an inability to take preventive measures or respond in a timely manner;

[0005] 3. Lack of flexibility: Traditional maintenance plans are usually fixed and difficult to adapt to the needs of emergencies or special circumstances. Summary of the invention

[0006] The purpose of the present invention is to provide an intelligent scheduling management system for power grid communication maintenance plan to solve the deficiencies in the background technology.

[0007] The technical solution of the present invention is: an intelligent scheduling management system for power grid communication maintenance plan, comprising: a data collection and preprocessing module, which collects historical operation data of power grid communication and performs preprocessing to obtain a sample data set;

[0008] The feature extraction module performs feature screening on the sample data set based on the LASSO regression model to obtain key features;

[0009] The fault prediction module divides the sample data set to obtain a training sample set and a verification sample set, builds a prediction model using the selected key features, and uses the training sample set and the verification sample set to train and evaluate the prediction model to obtain a fault prediction model;

[0010] Input the real-time operation data into the fault prediction model to obtain the predicted probability of equipment failure, and compare it with the preset probability threshold to determine whether the equipment will fail;

[0011] Intelligent scheduling optimization module sets optimization goals and constraints, establishes a maintenance plan scheduling optimization model, and uses heuristic algorithms to solve and generate the optimal maintenance plan scheduling solution;

[0012] The output module outputs the optimal maintenance plan scheduling plan to the system, which is reviewed and adjusted by the management personnel before execution.

[0013] Preferably, preprocessing the historical operation data includes: using statistical methods to clean the historical operation data and remove outliers; and then performing standardization to map the data to the [0,1] interval to obtain a sample data set.

[0014] Preferably, the objective function of the LASSO regression model in the feature extraction module is:

[0015]

[0016] In the formula, y i is the target variable of the i-th sample, indicating whether the device corresponding to the i-th sample fails; x ij is the jth feature of the i-th sample; β j is the parameter vector of the model; i = 1, 2, ..., n, n is the total number of samples; j = 1, 2, ..., h, h is the total number of features; λ is a non-negative regularization parameter;

[0017] The coordinate descent method is used to solve the objective function of the LASSO regression model. In each iteration, the parameter β is updated in turn, and the iteration is continued until the parameter converges, that is, |||β is satisfied. new -β old |||<ε, select the non-zero coefficient features after linear regression as the key features for final screening; among them, β new is the updated parameter, β old is the parameter before updating, and ε is a preset convergence threshold.

[0018] Preferably, the specific implementation content of the fault prediction module includes: obtaining the key features of the training sample set and the corresponding target variables, denoted as and y;

[0019] Where k is the number of key features; the target variable y indicates whether the device fails, which can be expressed as: y = 1, indicating that the device fails; y = 0, indicating that the device does not fail;

[0020] Constructing a prediction model, the predicted probability of equipment failure is:

[0021]

[0022] In the formula, Represents a given key feature X * When , the probability that the target variable y takes the value of 1 is, represents the vth key feature, β vrepresents the parameter corresponding to the vth key feature, v = 1, 2, ..., k, k is a positive integer, representing the total number of key features; σ is the sigmoid function;

[0023] When in use, the prediction model outputs a prediction probability range of (0,1). If the output prediction probability is greater than or equal to the preset probability threshold, the corresponding y=1, indicating that the device has failed; otherwise, the corresponding y=0, indicating that the device has not failed.

[0024] Preferably, the method for evaluating the performance of the prediction model using the validation sample set is as follows: statistically analyzing the prediction probability and actual results of the validation sample data to calculate the accuracy of the prediction model;

[0025] The calculation formula is:

[0026]

[0027] Where ZQ is the accuracy of the prediction model, TP is a true positive example, TN is a true negative example, FP is a false positive example, and FN is a false negative example.

[0028] The accuracy of the prediction model is obtained. If the accuracy is greater than or equal to the preset standard threshold, it means that the performance of the prediction model meets the standard and a fault prediction model is obtained; otherwise, the model needs to be further adjusted and optimized.

[0029] Preferably, the operating data of the power grid communication equipment is collected in real time through sensors, and the real-time operating data is input into the fault prediction model to obtain the prediction result, that is, the predicted probability of equipment failure; if the predicted probability is greater than or equal to the preset probability threshold, the system automatically issues an early warning prompt to remind maintenance personnel to perform further maintenance in a timely manner.

[0030] Preferably, the optimization objectives include:

[0031] By formula Minimize the total maintenance cost of all equipment, where JX is the total maintenance cost of all equipment; c m A represents the maintenance cost of the mth equipment; m is a binary variable indicating whether the mth equipment is overhauled, and M is the total number of equipment overhauls;

[0032] By formula Minimize the total maintenance time of all equipment; where JS is the total maintenance time of all equipment; t m is the maintenance time of the mth equipment;

[0033] By formula Minimize the impact on grid operation; where FX is the total impact on grid operation; mis the power outage loss caused during the maintenance of the mth equipment;

[0034] Combining the above three optimization objectives, the objective function can be defined as:

[0035] Z=JX·θ1+JS·θ2+FX·θ3;

[0036] Where Z is the objective function value, and θ1, θ2, and θ3 are the weights corresponding to the total maintenance cost, the total maintenance time of all equipment, and the total impact on the operation of the power grid, respectively.

[0037] Preferably, the constraints include:

[0038] Get the earliest maintenance time T preset for different equipment min and the latest maintenance time T max , set the maintenance time window to T min ≤T m ≤T max ; Among them, T m The time when the maintenance of equipment m begins;

[0039] At any given time t, the number of devices under maintenance must not exceed the maximum resource capacity L max ,Right now Among them, A mt Indicates whether equipment m is under maintenance within time t.

[0040] Preferably, the heuristic algorithm is a simulated annealing algorithm, and the specific steps include:

[0041] Set the initial temperature WD0, randomly generate an initial maintenance plan and record it as the initial solution S;

[0042] Starting from the current solution S, randomly exchange the maintenance time of two devices to generate a new solution S';

[0043] Calculate the difference ΔZ between the objective function values ​​of S and S'. According to the Metropolis criterion, if ΔZ < 0, the new state S' is accepted unconditionally; if ΔZ ≥ 0, the new state S' is accepted with a certain probability; the calculation formula for a certain probability is P' is a certain probability; ΔZ is the difference between the objective function values ​​of S and S'; WD is the current temperature; exp represents the natural exponential function, which is an exponential expression with the natural constant e as the base;

[0044] The temperature is updated by the formula WD'=α·WD; where α is a coefficient, α∈(0,1); WD' is the updated temperature;

[0045] Repeat the above steps until the stop condition is met;

[0046] Output the current solution as the optimal maintenance plan scheduling solution.

[0047] In order to solve the above problems, the present invention also provides a method for intelligent scheduling management of power grid communication maintenance plan, comprising the following steps:

[0048] Collect historical operation data of power grid communication and perform preprocessing to obtain a sample data set;

[0049] Based on the LASSO regression model, the sample data set is screened to obtain key features;

[0050] The sample data set is divided into a training sample set and a verification sample set, a prediction model is constructed using the screened key features, and the prediction model is trained and evaluated and optimized using the training sample set and the verification sample set to obtain a fault prediction model;

[0051] Input the real-time operation data into the fault prediction model to obtain the predicted probability of equipment failure, and compare it with the preset probability threshold to determine whether the equipment will fail;

[0052] Set optimization goals and constraints, establish a maintenance plan scheduling optimization model, and use heuristic algorithms to solve and generate the optimal maintenance plan scheduling solution;

[0053] The optimal maintenance plan scheduling scheme is output to the system and executed after review and adjustment by management personnel.

[0054] Compared with the existing solutions, the present invention achieves the following beneficial effects:

[0055] The present invention reduces the noise and inconsistency in the data by collecting the historical operation data of the power grid communication and performing preprocessing, improves the data quality, and lays a solid foundation for establishing an accurate prediction model;

[0056] The present invention uses the feature selection function of LASSO regression to identify the key features that have the greatest impact on fault prediction, simplifies the complexity of the model, reduces unnecessary computational costs, and improves the interpretability of the model, making the key factors clearer.

[0057] The present invention builds a prediction model and uses the divided training sample set and verification sample set to train and evaluate the prediction model, thereby ensuring the effectiveness and generalization ability of model training, avoiding overfitting, and enabling the model to better adapt to new data; inputs real-time operation data into the fault prediction model to determine whether the equipment will fail, thereby achieving real-time monitoring of the equipment status and early fault warning, helping operation and maintenance personnel to take timely measures to prevent potential equipment failures;

[0058] The present invention establishes a maintenance plan scheduling optimization model by setting optimization objectives and constraints, and uses a heuristic algorithm to solve it and generate an optimal maintenance plan scheduling scheme. It can improve maintenance work efficiency, reduce operating costs, and enhance the stability and reliability of the power grid, thereby enhancing market competitiveness and promoting the sustainable development of the power industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The present invention will be further described below in conjunction with the accompanying drawings.

[0060] Figure 1 This is a module structure diagram of an intelligent scheduling management system for power grid communication maintenance plan proposed by the present invention.

[0061] Figure 2 This is a flowchart of an intelligent scheduling management method for power grid communication maintenance plan proposed by the present invention. DETAILED DESCRIPTION

[0062] The technical solutions in the embodiments of the present invention will be described clearly and completely 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.

[0063] Embodiment 1, as Figure 1 As shown, a power grid communication maintenance plan intelligent scheduling management system includes a data collection and preprocessing module, a feature extraction module, a fault prediction module, an intelligent scheduling optimization module, and an output module;

[0064] Data collection and preprocessing module, which collects historical operation data of power grid communication equipment, including but not limited to voltage, current, power, and temperature;

[0065] Preprocessing the collected historical operation data includes: using statistical methods to clean the historical operation data and remove outliers; then standardizing the data to map it to the [0,1] interval so that different data have the same dimension and obtain a sample data set;

[0066] For example, a statistical method is used to remove outliers: a current data set D is obtained, and the mean of the obtained set is calculated. and standard deviation If any current data a satisfies It is considered as an outlier and removed.

[0067] In the embodiment of the present invention, by collecting and processing historical operation data, the historical data is sorted and cleaned, the noise and inconsistency in the data are reduced, the data quality is improved, and a solid foundation is laid for establishing an accurate prediction model.

[0068] The feature extraction module builds a LASSO regression model and uses L1 regularization to screen the sample data set to obtain key features;

[0069] In some preferred embodiments, the objective function of the LASSO regression model is:

[0070]

[0071] In the formula, y i is the target variable of the i-th sample, indicating whether the device corresponding to the i-th sample fails; x ij is the jth feature of the i-th sample; β j is the parameter vector of the model; i = 1, 2, ..., n, n is the total number of samples; j = 1, 2, ..., h, h is the total number of features; λ is a non-negative regularization parameter;

[0072] In some preferred embodiments, the objective function of the LASSO regression model is solved using the coordinate descent method. In each iteration, the parameter β is updated in turn, and the iteration is continued until the parameter converges, that is, |||β is satisfied. new -β old |||<ε, select the non-zero coefficient features after linear regression as the key features for final screening; among them, β new is the updated parameter, β old is the parameter before updating, ε is a preset convergence threshold;

[0073] It should be noted that the coordinate descent method is an optimization algorithm that minimizes the objective function by updating each variable one by one. In each iteration, a variable, namely parameter β, is selected, and then a one-dimensional search is performed along this direction to find the optimal value of the variable, so as to minimize the objective function value.

[0074] In the implementation of the steps of the present invention, through the regularization effect of the LASSO regression model, key features that have a significant impact on the occurrence of faults are automatically selected, irrelevant or redundant features are removed, the data dimension is reduced, and the interpretability and computational efficiency of the model are improved.

[0075] The fault prediction module divides the sample data set into a training sample set and a validation sample set according to a certain ratio; obtains the key features of the training sample set and the corresponding target variables, recorded as and y;

[0076] Among them, the certain ratio is usually set to 7:3; k is the number of key features; the target variable y indicates whether the equipment fails, which can be expressed as: y = 1, indicating that the equipment fails; y = 0, indicating that the equipment does not fail;

[0077] Constructing a prediction model, the predicted probability of equipment failure is:

[0078]

[0079] In the formula, Represents a given key feature X * When , the probability that the target variable y takes the value of 1 is, represents the vth key feature, β v represents the parameter corresponding to the vth key feature, v = 1, 2, ..., k, k is a positive integer, representing the total number of key features; σ is the sigmoid function;

[0080] When used, the prediction model outputs a prediction probability range of (0,1). If the output prediction probability is greater than or equal to the preset probability threshold, the corresponding y=1, indicating that the device has failed; otherwise, the corresponding y=0, indicating that the device has not failed; the probability threshold is a reasonable threshold set according to the specific requirements of the model and the experience of experts in the field;

[0081] Use the validation sample set to evaluate the performance of the prediction model;

[0082] In some preferred embodiments, the predicted results and actual results of the sample data are statistically verified to calculate the accuracy of the prediction model;

[0083] The calculation formula is:

[0084]

[0085] Where ZQ is the accuracy of the prediction model; TP stands for true positive examples, that is, the number of samples whose actual results are positive and whose predicted results are positive; TN stands for true negative examples, that is, the number of samples whose actual results are negative and whose predicted results are negative; FP stands for false positive examples, that is, the number of samples whose actual results are negative but whose predicted results are positive; FN stands for false negative examples, that is, the number of samples whose actual results are positive but whose predicted results are negative.

[0086] Obtain the accuracy of the prediction model. If the accuracy is greater than or equal to the preset standard threshold, it means that the performance of the prediction model meets the standard and a fault prediction model is obtained. Otherwise, it is necessary to continue to adjust and optimize the model. The standard threshold can be set to 90%.

[0087] The operating data of the power grid communication equipment is collected in real time through sensors, and the real-time operating data is input into the fault prediction model to obtain the prediction result, that is, the predicted probability of equipment failure; if the predicted probability is greater than or equal to the preset probability threshold, the system automatically issues an early warning to remind maintenance personnel to perform further maintenance in a timely manner.

[0088] In the embodiment of the present invention, a model is established that can accurately predict equipment failures, discover potential failure risks in advance, and help take preventive measures to reduce unexpected downtime and maintenance costs.

[0089] Intelligent scheduling optimization module sets optimization goals and constraints, establishes a maintenance plan scheduling optimization model, and uses a heuristic algorithm to solve the maintenance plan scheduling optimization model to generate the optimal maintenance plan scheduling solution;

[0090] In some preferred embodiments, the optimization objectives include:

[0091] By formula Minimize the total maintenance cost of all equipment, where JX is the total maintenance cost of all equipment; c m A represents the maintenance cost of the mth equipment; m is a binary variable indicating whether the mth equipment is overhauled (1 for overhaul, 0 for no overhaul), and M is the total number of equipment overhauls;

[0092] By formula Minimize the total maintenance time of all equipment; where JS is the total maintenance time of all equipment; t m is the maintenance time of the mth equipment;

[0093] By formula Minimize the impact on grid operation; where FX is the total impact on grid operation; m is the power outage loss caused during the maintenance of the mth equipment;

[0094] Combining the above three optimization objectives, the objective function can be defined as:

[0095] Z=JX·θ1+JS·θ2+FX·θ3;

[0096] Where Z is the objective function value, θ1, θ2, θ3 are the weights corresponding to the total maintenance cost, the total maintenance time of all equipment, and the total impact on the operation of the power grid, which are determined according to the actual needs and importance of the power grid communication system;

[0097] In some preferred embodiments, the constraints include:

[0098] Get the earliest maintenance time T preset for different equipment min and the latest maintenance time T max , set the maintenance time window to T min ≤T m ≤T max ; Among them, T m The time when the maintenance of equipment m begins;

[0099] At any given time t, the number of devices under maintenance must not exceed the maximum resource capacity L max ,Right now Among them, A mt Indicates whether equipment m is overhauled within time t;

[0100] In some preferred embodiments, the heuristic algorithm is a simulated annealing algorithm, which avoids falling into a local optimal solution by randomly searching and accepting inferior solutions with a certain probability. The specific steps include:

[0101] Set the initial temperature WD0, randomly generate an initial maintenance plan and record it as the initial solution S;

[0102] Starting from the current solution S, randomly exchange the maintenance time of two devices to generate a new solution S';

[0103] Calculate the difference ΔZ between the objective function values ​​of S and S'. According to the Metropolis criterion, if ΔZ < 0, the new state S' is accepted unconditionally; if ΔZ ≥ 0, the new state S' is accepted with a certain probability; the calculation formula for a certain probability is P' is a certain probability; ΔZ is the difference between the objective function values ​​of S and S'; WD is the current temperature; exp represents the natural exponential function, which is an exponential expression with the natural constant e (approximately equal to 2.71828) as the base;

[0104] The temperature is updated by the formula WD'=α·WD; where α is a coefficient, α∈(0,1), which is usually close to 1, such as 0.95, 0.9; WD' is the updated temperature;

[0105] Repeat the above steps until the stop condition is met; wherein the stop condition is that the algorithm stops when the number of iterations reaches the preset maximum number of iterations or the temperature is lower than a certain temperature threshold;

[0106] Output the current solution as the optimal maintenance plan scheduling solution.

[0107] In the embodiment of the present invention, a near-optimal maintenance plan scheduling scheme is found through a simulated annealing algorithm, which not only demonstrates its powerful global search capability and adaptability, but also demonstrates its flexibility and effectiveness in solving practical problems.

[0108] The output module outputs the optimal maintenance plan to the system in the form of charts or reports. The management personnel will conduct necessary review and adjustment of the maintenance plan based on the actual situation, arrange personnel and resources according to the finalized maintenance plan, and perform maintenance work.

[0109] Embodiment 2, as Figure 2 As shown, a method for intelligent scheduling management of power grid communication maintenance plan includes the following steps:

[0110] Collect historical operation data of power grid communication and perform preprocessing to obtain a sample data set;

[0111] Based on the LASSO regression model, the sample data set is screened to obtain key features;

[0112] The sample data set is divided into a training sample set and a verification sample set, a prediction model is constructed using the screened key features, and the prediction model is trained and evaluated and optimized using the training sample set and the verification sample set to obtain a fault prediction model;

[0113] Input the real-time operation data into the fault prediction model to obtain the predicted probability of equipment failure, and compare it with the preset probability threshold to determine whether the equipment will fail;

[0114] Set optimization goals and constraints, establish a maintenance plan scheduling optimization model, and use heuristic algorithms to solve and generate the optimal maintenance plan scheduling solution;

[0115] The optimal maintenance plan scheduling scheme is output to the system and executed after review and adjustment by management personnel.

[0116] In the embodiments provided by the present invention, it should be understood that the disclosed system can be implemented in other ways. For example, the above-described embodiments of the invention are only illustrative, for example, the division of modules is only a logical function division, and there may be other division methods in actual implementation.

[0117] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0118] In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0119] It is obvious to a person skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the essential characteristics of the present invention.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. An intelligent scheduling management system for power grid communication maintenance plan, characterized in that: include: The data collection and preprocessing module is used to collect the historical operation data of the power grid communication and perform preprocessing to obtain a sample data set; The feature extraction module is used to screen the sample data set based on the LASSO regression model to obtain key features; The fault prediction module is used to divide the sample data set to obtain a training sample set and a verification sample set, build a prediction model using the screened key features, and use the training sample set and the verification sample set to train and evaluate the prediction model to obtain a fault prediction model; Input the real-time operation data into the fault prediction model to obtain the predicted probability of equipment failure, and compare it with the preset probability threshold to determine whether the equipment will fail; Intelligent scheduling optimization module is used to set optimization goals and constraints, establish a maintenance plan scheduling optimization model, and use heuristic algorithms to solve and generate the optimal maintenance plan scheduling solution; The output module is used to output the optimal maintenance plan scheduling scheme to the system, which will be reviewed and adjusted by the management personnel before execution.

2. The intelligent scheduling management system for power grid communication maintenance plan according to claim 1 is characterized in that: The preprocessing of historical operation data includes: using statistical methods to clean the historical operation data and remove outliers; then performing standardization to map the data to the [0,1] interval to obtain a sample data set.

3. The intelligent scheduling management system for power grid communication maintenance plan according to claim 2 is characterized in that: The objective function of the LASSO regression model described in the feature extraction module is: In the formula, y i is the target variable of the i-th sample, indicating whether the device corresponding to the i-th sample fails; x ij is the jth feature of the i-th sample; β j is the parameter vector of the model; i = 1, 2, ..., n, n is the total number of samples; j = 1, 2, ..., h, h is the total number of features; λ is a non-negative regularization parameter; The coordinate descent method is used to solve the objective function of the LASSO regression model. In each iteration, the parameter β is updated in turn, and the iteration is continued until the parameter converges, that is, |||β is satisfied. new -β old |||<ε, select the non-zero coefficient features after linear regression as the key features for final screening; among them, β new is the updated parameter, β old is the parameter before updating, and ε is the preset convergence threshold.

4. The intelligent scheduling management system for power grid communication maintenance plan according to claim 1, characterized in that: The specific implementation of the fault prediction module includes: obtaining the key features of the training sample set and the corresponding target variables, denoted as and y; Where k is the number of key features; the target variable y indicates whether the device fails, expressed as: y = 1, indicating that the device fails; y = 0, indicating that the device does not fail; Constructing a prediction model, the predicted probability of equipment failure is: In the formula, Represents a given key feature X * When , the probability that the target variable y takes the value of 1 is, represents the vth key feature, β v represents the parameter corresponding to the vth key feature, v = 1, 2, ..., k, k is a positive integer, representing the total number of key features; σ is the sigmoid function; When in use, the prediction model outputs a prediction probability range of (0,1). If the output prediction probability is greater than or equal to the preset probability threshold, the corresponding y=1, indicating that the device has failed; otherwise, the corresponding y=0, indicating that the device has not failed.

5. The intelligent scheduling management system for power grid communication maintenance plan according to claim 4 is characterized in that: The method of evaluating the performance of the prediction model using the validation sample set is as follows: statistically analyze the prediction probability and actual results of the validation sample data and calculate the accuracy of the prediction model; The calculation formula is: Where ZQ is the accuracy of the prediction model, TP is a true positive example, TN is a true negative example, and FP is a false positive example. FN stands for false negative; The accuracy of the prediction model is obtained. If the accuracy is greater than or equal to the preset standard threshold, it means that the performance of the prediction model meets the standard and a fault prediction model is obtained; otherwise, the model needs to be further adjusted and optimized.

6. The intelligent scheduling management system for power grid communication maintenance plan according to claim 5, characterized in that: The operating data of the power grid communication equipment is collected in real time through sensors, and the real-time operating data is input into the fault prediction model to obtain the prediction result, that is, the predicted probability of equipment failure; if the predicted probability is greater than or equal to the preset probability threshold, the system automatically issues an early warning to remind maintenance personnel to perform further maintenance in a timely manner.

7. The intelligent scheduling management system for power grid communication maintenance plan according to claim 1, characterized in that: The optimization objectives include: By formula Minimize the total maintenance cost of all equipment, where JX is the total maintenance cost of all equipment; c m A represents the maintenance cost of the mth equipment; m is a binary variable indicating whether the mth equipment is overhauled, and M is the total number of equipment overhauls; By formula Minimize the total maintenance time of all equipment; where JS is the total maintenance time of all equipment; t m is the maintenance time of the mth equipment; By formula Minimize the impact on grid operation; where FX is the total impact on grid operation; m is the power outage loss caused during the maintenance of the mth equipment; Combining the above three optimization objectives, the objective function is defined as: Z=JX·θ1+JS·θ2+FX·θ3; Where Z is the objective function value, and θ1, θ2, and θ3 are the weights corresponding to the total maintenance cost, the total maintenance time of all equipment, and the total impact on the operation of the power grid, respectively.

8. The intelligent scheduling management system for power grid communication maintenance plan according to claim 7, characterized in that: The constraints include: Get the earliest maintenance time T preset for different equipment min and the latest maintenance time T max , set the maintenance time window to T min ≤T m ≤T max ; Among them, T m The time when the maintenance of equipment m begins; At any given time t, the number of devices under maintenance must not exceed the maximum resource capacity L max ,Right now Among them, A mt Indicates whether equipment m is under maintenance within time t.

9. The intelligent scheduling management system for power grid communication maintenance plan according to claim 8, characterized in that: The heuristic algorithm is a simulated annealing algorithm, and the specific steps include: Set the initial temperature WD0, randomly generate an initial maintenance plan and record it as the initial solution S; Starting from the current solution S, randomly exchange the maintenance time of two devices to generate a new solution S'; Calculate the difference ΔZ between the objective function values ​​of S and S'. According to the Metropolis criterion, if ΔZ < 0, the new state S' is accepted unconditionally; if ΔZ ≥ 0, the new state S' is accepted with a certain probability; the calculation formula for a certain probability is P' is a certain probability; ΔZ is the difference between the objective function values ​​of S and S'; WD is the current temperature; exp represents the natural exponential function; The temperature is updated by the formula WD'=α·WD; where α is a coefficient, α∈(0,1); WD' is the updated temperature; Repeat the above steps until the stop condition is met; Output the current solution as the optimal maintenance plan scheduling solution.

10. A method for intelligent scheduling management of power grid communication maintenance plan, comprising the following steps: Collect historical operation data of power grid communication and perform preprocessing to obtain a sample data set; Based on the LASSO regression model, the sample data set is screened to obtain key features; The sample data set is divided into a training sample set and a verification sample set, a prediction model is constructed using the screened key features, and the prediction model is trained and evaluated and optimized using the training sample set and the verification sample set to obtain a fault prediction model; Input the real-time operation data into the fault prediction model to obtain the predicted probability of equipment failure, and compare it with the preset probability threshold to determine whether the equipment will fail; Set optimization goals and constraints, establish a maintenance plan scheduling optimization model, and use heuristic algorithms to solve and generate the optimal maintenance plan scheduling solution; The optimal maintenance plan scheduling scheme is output to the system and executed after review and adjustment by management personnel.

Images