Intelligent power distribution network first-aid repair scheduling system

By integrating environmental perception, fault detection, optimized scheduling and self-learning modules in the distribution network emergency repair scheduling system, the shortcomings of the existing system in fault location and resource allocation are solved, efficient and reliable emergency repair scheduling are achieved, and the power grid is improved stability and user experience.

CN119047615APending Publication Date: 2024-11-29GONGYI POWER SUPPLY CO OF STATE GRID HENAN ELECTRIC POWER CO +1
View PDF 0 Cites 8 Cited by

Patent Information

Application Number
CN202410977417.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

The existing distribution network emergency repair scheduling system has shortcomings in fault positioning accuracy, emergency repair scheduling optimization and resource allocation, and lacks automation and self-learning mechanisms, resulting in long response time, unreasonable resource allocation and frequent misoperation.

Method used

An intelligent distribution network emergency repair scheduling system was designed, integrating environmental perception module, fault detection and positioning module, evaluation and warning module, optimization scheduling decision-making module, emergency resource allocation module and feedback self-learning module, and adopting advanced fault detection technology, multiple Target optimization scheduling and dynamic resource allocation, combined with multi-objective evolution algorithm and reinforcement learning algorithm, realize system self-optimization.

Benefits of technology

It significantly improves emergency repair efficiency and grid stability, shortens fault diagnosis time, optimizes resource allocation, reduces power supply interruption time, improves grid reliability and user experience, and reduces operating costs through self-optimization capabilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119047615A_ABST
    Figure CN119047615A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of power system management, in particular to an intelligent power distribution network first-aid repair scheduling system which comprises an environment sensing module, a fault detection and positioning module, an evaluation and early warning module, an optimization scheduling decision module, an emergency resource allocation module and a feedback self-learning module. Wherein the environment sensing module is used for collecting environment information of a power distribution network in real time; the fault detecting and positioning module detects faults of the power distribution network in real time; the evaluation and early warning module is used for intelligently evaluating the fault information; the optimal scheduling decision module is used for generating an optimal first-aid repair scheduling scheme; the emergency resource allocation module is used for dynamically allocating first-aid repair personnel, equipment and materials; and the feedback self-learning module is used for carrying out self-learning, optimization and improvement. According to the invention, by integrating efficient fault detection, multi-objective optimization scheduling, dynamic resource allocation and feedback and self-learning mechanisms, the first-aid repair response speed and the resource use efficiency are remarkably improved, so that the reliability of a power grid is enhanced, and the operation and maintenance cost is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of power system management, and in particular to an intelligent power distribution network emergency repair dispatching system. Background Art

[0002] In modern power systems, stable and efficient operation of distribution networks is crucial to ensuring the reliability of power supply. With the acceleration of urbanization and the continuous increase in grid load, the challenges faced by distribution networks are also increasing, especially in emergency repair and dispatching. Traditional distribution network emergency repair and dispatching systems mainly rely on manual experience and static planning, which often seem inadequate when dealing with sudden, complex and changeable faults. For example, traditional systems often lack flexibility and efficiency in fault detection, resource scheduling, and fault repair, resulting in long emergency response times, unreasonable resource allocation, and even frequent misoperations.

[0003] In view of the above problems, the present invention aims to solve several defects and shortcomings in the prior art. The prior art fails to provide a distribution network emergency repair and dispatching system that fully utilizes modern information technology and realizes a high degree of automation and intelligence. Specifically, the prior system has obvious deficiencies in the accuracy of fault location, optimization of emergency repair dispatching, and dynamic adjustment of resource allocation according to real-time data. In addition, the prior system lacks an effective feedback and self-learning mechanism, making it difficult to accumulate experience from historical emergency repair activities and optimize future emergency repair strategies.

[0004] Therefore, developing an intelligent distribution network emergency repair and dispatching system that integrates advanced fault detection technology, multi-objective optimization scheduling, dynamic resource allocation, and self-learning and feedback mechanism has important practical significance and application value for improving the emergency repair efficiency and reliability of the power grid. Summary of the invention

[0005] Based on the above objectives, the present invention provides an intelligent distribution network emergency repair dispatching system.

[0006] The intelligent distribution network emergency repair and dispatching system includes an environmental perception module, a fault detection and positioning module, an evaluation and early warning module, an optimization and dispatching decision module, an emergency resource allocation module, and a feedback self-learning module; among which:

[0007] Environmental perception module: used to collect distribution network environmental information in real time, including temperature, humidity, wind speed, voltage, and current parameters;

[0008] Fault detection and positioning module: connected to the environment perception module, used to detect distribution network faults in real time according to environmental information and power grid operation data, and locate the fault point;

[0009] Evaluation and early warning module: connected to the fault detection and positioning module, intelligently evaluates fault information, predicts fault development trends, and issues early warning signals when necessary;

[0010] Optimization scheduling decision module: connected to the evaluation and early warning module, using a multi-objective evolutionary algorithm to generate an optimal emergency repair scheduling plan based on fault information, environmental information and historical data;

[0011] Emergency resource allocation module: connected to the optimization scheduling decision module, dynamically allocates emergency repair personnel, equipment and materials according to the optimal emergency repair scheduling plan to ensure efficient completion of the emergency repair task;

[0012] Feedback self-learning module: connected to the emergency resource allocation module, used to record the actual execution status during the emergency repair process, analyze the execution effect of the scheduling plan, perform self-learning and optimization improvements, and improve the overall performance of the system.

[0013] Furthermore, the environment sensing module includes a temperature sensor unit, a humidity sensor unit, a wind speed sensor unit, a voltage sensor unit and a current sensor unit; wherein:

[0014] Temperature sensor unit: used to measure the temperature around the distribution network. The temperature sensor unit uses the thermistor principle to calculate the specific temperature value through the change of resistance value;

[0015] Humidity sensor unit: used to measure the humidity around the distribution network. The humidity sensor unit uses a capacitive humidity sensor to calculate the specific humidity value through the change of capacitance value;

[0016] Wind speed sensor unit: used to measure the wind speed around the distribution network. The wind speed sensor unit uses an ultrasonic wind speed sensor to calculate the wind speed value through the propagation speed of ultrasonic waves in the air;

[0017] Voltage sensor unit: used to measure the voltage in the distribution network. The voltage sensor unit uses the principle of electromagnetic induction to calculate the voltage value by measuring the changes in the electromagnetic field generated around the conductor;

[0018] Current Sensor Unit: Used to measure the current in the power distribution network, this unit uses a Hall Effect sensor to calculate the current value by measuring the change in the magnetic field generated by the current flowing through the conductor.

[0019] Furthermore, the fault detection and locating module includes a data acquisition unit, a fault detection unit and a fault locating unit; wherein:

[0020] Data acquisition unit: used to receive and process environmental information from the environmental perception module and operating data from the distribution network operation system, and realize real-time synchronization of environmental information and operating data through a preset data interface;

[0021] Fault detection unit: used to analyze the environmental information and operating data received by the data acquisition unit and identify abnormal conditions in the distribution network. The fault detection unit uses a pattern recognition algorithm to determine the time point when the fault occurs by comparing the normal operating data with the abnormal changes in the current data;

[0022] Fault location unit: used to locate the specific fault point according to the abnormal state identified by the fault detection unit. The fault location unit adopts a fault location algorithm based on power lines to locate the geographical location of the fault point by analyzing the traveling waves and impedance changes generated when the fault occurs.

[0023] Furthermore, the fault detection unit includes:

[0024] Data preprocessing: preprocess the received environmental information and operation data, including denoising, normalization and data smoothing;

[0025] Feature extraction: Extract features from preprocessed data. Specifically, fast Fourier transform is used to convert time domain data into frequency domain data for subsequent pattern recognition. The specific extraction formula is:

[0026] Among them, x n ′ is the nth normalized time domain data point, X(f j ) is the jth frequency component, N is the total number of data points, f j is the jth frequency;

[0027] Normal mode establishment: Use historical normal operation data to establish a normal mode, use principal component analysis to reduce the dimension, and extract characteristic components. The formula is: Among them, X m is the mth eigenvalue, W km is the element in the PCA weight matrix, Z k is the kth principal component, M is the total number of eigenvalues;

[0028] Real-time data monitoring: Perform the same feature extraction algorithm as in the feature extraction step on the current real-time data to obtain the real-time feature value. The formula is: Among them, x current,p is the current p-th time domain data point, X current (f l ) is the current lth frequency component, P is the total number of current data points, f l is the current lth frequency;

[0029] Anomaly detection: Compare the real-time feature value with the normal mode and calculate the feature difference. Specifically, the Mahalanobis distance is used to evaluate the degree of anomaly. The formula is:

[0030] Among them, X current,t is the current eigenvector at time point t, μ N is the mean vector of the normal mode, ∑ N is the covariance matrix of the normal mode, D M,t is the Mahalanobis distance at time point t;

[0031] Fault time point determination: When the Mahalanobis distance exceeds the preset threshold, a fault is determined to have occurred, and the time point t at which the fault occurred is recorded. f , the formula is: f =min{t:D M,t >Threshold}, where t f is the time point when the fault occurs, and Threshold is the preset threshold.

[0032] Furthermore, the fault location unit includes:

[0033] Traveling wave detection: When a fault occurs, the voltage and current waveforms on the power line are collected, and the arrival time of the traveling wave is detected. The formula is: Among them, t w1 is the arrival time of the traveling wave, d is the distance from the fault point to the detection point, and v is the propagation speed of the traveling wave;

[0034] Traveling wave time difference calculation: detect the arrival time of traveling waves at multiple locations and calculate the traveling wave time difference. The formula is: Δt w =t w1 -t w2 , where Δt w is the traveling wave time difference between the two detection points, t w2 is the arrival time of the traveling wave at the second detection point;

[0035] Traveling wave fault distance calculation: According to the traveling wave time difference, calculate the distance from the fault point to the detection point. The formula is: Among them, d fault is the distance from the fault point to the midpoint of the two detection points;

[0036] Impedance measurement: When a fault occurs, measure the impedance change of the power line and calculate the impedance of the fault point. The formula is: Among them, Z fault is the impedance of the fault point, V fault is the voltage at the fault point, I foult is the current at the fault point;

[0037] Impedance fault distance calculation: According to the impedance of the fault point, calculate the distance from the fault point to the detection point. The formula is: Among them, d imp is the distance to the fault point obtained by impedance measurement, Z lineis the line impedance per unit length, L is the total length of the line;

[0038] Comprehensive positioning: Combine the calculation results of the traveling wave method and the impedance method to determine the final geographical location of the fault point. The formula is: Among them, d final is the distance to the fault point after comprehensive positioning.

[0039] Furthermore, the evaluation and warning module includes a data analysis unit, a trend prediction unit, a warning judgment unit and a warning signal unit; wherein:

[0040] Data analysis unit: used to receive and process fault information from the fault detection and positioning module, compare and analyze historical data and real-time data, and identify potential fault modes. The data analysis unit specifically uses the support vector machine algorithm for classification and prediction. The formula is: Among them, x is the input data, α i and i is the model parameter, K(x i ,x) is the kernel function, b is the bias;

[0041] Trend prediction unit: It is used to predict fault trends based on the output results of the data analysis unit using a long short-term memory network to determine the potential development direction and severity of the fault; the formula is:

[0042] h t =σ(W h ·[h t-1 ,x t ]+b h ), where h t is the hidden state at time t, W h is the weight matrix, x t is the input at time t, b h is bias;

[0043] Early warning judgment unit: It is used to judge whether it is necessary to issue an early warning signal according to the prediction results of the trend prediction unit. When the prediction results show that the fault will reach a serious level within the preset time, the early warning signal will be triggered. The judgment formula is:

[0044] Among them, P(fault t+k ) is the predicted value of the failure probability at time t+k, and Threshold is the warning threshold;

[0045] Early warning signal unit: used to send out early warning signals when the early warning judgment unit triggers the early warning. The early warning signals will be sent out through system interface display, sound alarm and SMS notification.

[0046] Furthermore, the optimization scheduling decision module includes a data integration unit, a target definition unit, an individual generation unit, a fitness evaluation unit, an evolutionary operation unit and an optimal solution selection unit; wherein:

[0047] Data fusion unit: used to receive and integrate multi-source data from the environment perception module, fault detection and positioning module and historical database, including real-time environment information, current fault information and historical fault data; and ensure data accuracy and consistency through data cleaning and feature extraction;

[0048] Target definition unit: used to define multiple optimization targets according to the preset scheduling targets, including minimizing the repair time, minimizing the repair cost and optimizing resource allocation; the formula is:

[0049]

[0050] Among them, T i is the repair time of the i-th task, C i is the repair cost of the ith task, R i is the resource utilization efficiency of the i-th task;

[0051] Individual generation unit: used to generate the initial solution in the multi-objective evolutionary algorithm, that is, multiple emergency repair scheduling plans, each of which contains the specific content of the allocation of emergency repair personnel, equipment and material scheduling; the formula is:

[0052] X={x1,x2,…,x n}, where x i is the i-th initial solution, representing a specific scheduling plan;

[0053] The fitness evaluation unit is used to evaluate the fitness of each scheduling scheme and calculate the comprehensive score of each scheme according to the definition of the objective function. The calculation formula is: F(X) = w1f1(X) + w2f2(X) + w3f3(X), where F(X) is the comprehensive fitness score and w1, w2, w3 are the weights of the objective function;

[0054] Evolutionary operation unit: used to perform operations of multi-objective evolutionary algorithms, including selection, crossover and mutation;

[0055] The selection operation formula is: selected =Select(P population ,F(X)), where P selected is the population after selection, P population is the current population, F(X) is the fitness score;

[0056] The crossover operation formula is: X new =Crossover(Xparent1 ,X parent2 ), where X new is the newly generated individual, X parent11 ,X parent2 is the selected parent individual;

[0057] The mutation operation formula is: mutated =Mutate(X individual ), where X mutated is the individual after mutation, X individual For the individual who undergoes the mutation;

[0058] The optimal solution selection unit is used to select the final optimal emergency repair scheduling solution from the evolved solution set.

[0059] Furthermore, the optimal solution selection unit includes:

[0060] Non-dominated sorting: Perform non-dominated sorting on the evolved solution set and divide the solutions into different levels. The goal of non-dominated sorting is to find the maximum non-dominated solution. The formula is expressed as:

[0061] S i ={X j |X j DominateX i};

[0062] n i =|{X j |X i X j Dominance}; Among them, S i is the set of solutions dominated by the ith solution, n i is the number of solutions that dominate the ith solution;

[0063] Crowding calculation: Calculate the crowding of each solution to measure the density of the solution in the target space. The formula is: Among them, I i is the congestion degree of the ith solution, and are the target values ​​of the previous and next neighbor solutions of the ith solution on the mth objective function, and are the maximum and minimum values ​​of the mth objective function respectively;

[0064] Solution selection: According to the non-dominated sorting and congestion calculation results, the optimal solution is selected from the evolved solution set, with priority given to solutions with high non-dominated levels, and solutions with high congestion levels are selected within the same non-dominated level. The specific formula is: Among them, X optimalis the optimal solution, Rank(X) is the non-dominated rank of solution X, and I X is the congestion degree of scheme X, P final The final solution set.

[0065] Furthermore, the emergency resource allocation module includes a personnel allocation unit, an equipment allocation unit, a material allocation unit and a scheduling execution unit; wherein:

[0066] Personnel dispatching unit: used to dynamically allocate emergency repair personnel according to the optimal emergency repair scheduling plan. The personnel dispatching unit matches the corresponding emergency repair personnel to the designated fault location according to the personnel's skills, experience and current availability;

[0067] Equipment dispatching unit: used to dynamically allocate emergency repair equipment according to the optimal emergency repair scheduling plan, and match the corresponding equipment to the designated fault location according to the type, performance and current availability of the equipment;

[0068] Material dispatching unit: used to dynamically allocate emergency repair materials according to the optimal emergency repair dispatch plan; according to the type, quantity and inventory status of the materials, the corresponding materials are matched to the designated fault location;

[0069] Scheduling execution unit: used to integrate the scheduling results of the personnel deployment unit, equipment deployment unit and material deployment unit, generate the final comprehensive scheduling plan and execute it.

[0070] Furthermore, the feedback self-learning module includes a data recording unit, an execution effect analysis unit, a self-learning unit and an optimization and improvement unit; wherein:

[0071] Data recording unit: used to record the actual execution status of the repair process, including the arrival time of personnel, equipment usage, material consumption and fault repair time;

[0072] Execution effect analysis unit: Based on the recorded data, evaluate the deviation between the actual execution results and the planned solution, and identify the factors affecting the execution effect. The calculation formula is:

[0073] Among them, E deviation For execution deviation, A i is the actual execution value of item i, P i is the planned value of the i-th item, and n is the total number of items;

[0074] Self-learning unit: Based on the results of the execution effect analysis, self-learning and optimization are carried out. Specifically, the reinforcement learning algorithm is used to continuously update the model parameters to improve the decision-making ability and response speed of the system. The formula is: Q(s,a)=Q(s,a)+α[r+γmax a′Q(s′,a′)-Q(s,a)], where Q(s,a) is the value of performing action a in state s, α is the learning rate, r is the immediate reward, γ is the discount factor, s′ is the next state, and a′ is the next action;

[0075] Optimization and improvement unit: used to adjust the scheduling strategy and parameter settings according to the output results of the self-learning unit.

[0076] Beneficial effects of the present invention:

[0077] The present invention significantly improves the efficiency and effectiveness of power distribution network emergency repair by integrating and optimizing multiple key technologies. First, the system adopts advanced fault detection and positioning technology, which can quickly and accurately identify and locate fault points in the distribution network. This efficient fault response mechanism greatly shortens the fault diagnosis time and provides reliable data support for subsequent emergency repair scheduling, thereby reducing the duration of power supply interruptions and significantly improving the stability of the power grid and the user's power supply experience.

[0078] The present invention optimizes the configuration of manpower, equipment and materials by adopting a multi-objective evolutionary algorithm to generate an emergency repair scheduling plan. This intelligent scheduling strategy not only takes into account the minimization of emergency repair costs and time, but also ensures the optimal utilization of resources. By dynamically allocating resources, the system can adapt to the ever-changing emergency repair needs and on-site conditions, effectively avoiding waste of resources and delays in the emergency repair process, and further enhancing the economy of emergency repair operations and the flexibility of operations.

[0079] The present invention records the data during the emergency repair process in real time and analyzes the execution effect of the scheduling plan. The system can continuously adjust and improve the emergency repair strategy. This continuous self-optimization capability makes the system more efficient and accurate in future emergency repair tasks, effectively reduces long-term operating costs and improves service quality, creating greater economic and social value for users and power companies. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0081] Figure 1 A schematic diagram of a smart distribution network emergency repair and dispatching system according to an embodiment of the present invention;

[0082] Figure 2 Schematic diagram of an optimization scheduling decision module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0083] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0084] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. The words "include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0085] like Figure 1-Figure 2 As shown, the intelligent distribution network emergency repair and dispatching system includes an environmental perception module, a fault detection and positioning module, an evaluation and early warning module, an optimization and dispatching decision module, an emergency resource allocation module, and a feedback self-learning module; among which:

[0086] Environmental perception module: used to collect distribution network environmental information in real time, including temperature, humidity, wind speed, voltage, and current parameters;

[0087] Fault detection and positioning module: connected to the environment perception module, used to detect distribution network faults in real time based on environmental information and power grid operation data, and locate fault points;

[0088] Evaluation and early warning module: connected to the fault detection and positioning module, it conducts intelligent evaluation of fault information, predicts the fault development trend, and issues early warning signals when necessary;

[0089] Optimization and scheduling decision module: connected to the evaluation and early warning module, it uses a multi-objective evolutionary algorithm (MOEA) to combine fault information, environmental data and historical data to generate the optimal emergency repair scheduling plan. The multi-objective evolutionary algorithm ensures efficient and economical emergency repair execution by optimizing multiple goals such as emergency repair time, cost and resource allocation;

[0090] Emergency resource allocation module: connected to the optimization scheduling decision module, dynamically allocates emergency repair personnel, equipment and materials according to the generated emergency repair scheduling plan to ensure the efficient completion of the emergency repair task;

[0091] Feedback self-learning module: connected to the emergency resource allocation module, used to record the actual execution status of the emergency repair process, analyze the execution effect of the scheduling plan, perform self-learning and optimization, and improve the overall performance of the system.

[0092] The environment sensing module includes a temperature sensor unit, a humidity sensor unit, a wind speed sensor unit, a voltage sensor unit and a current sensor unit; wherein:

[0093] Temperature sensor unit: used to measure the temperature around the distribution network. The temperature sensor unit adopts the principle of thermistor, where temperature changes lead to changes in resistance value, and the specific temperature value is calculated through the change in resistance value;

[0094] Humidity sensor unit: used to measure the humidity around the distribution network. The humidity sensor unit adopts a capacitive humidity sensor, in which changes in ambient humidity lead to changes in capacitance value, and the specific humidity value is calculated through the change in capacitance value;

[0095] Wind speed sensor unit: used to measure the wind speed around the distribution network. The wind speed sensor unit uses an ultrasonic wind speed sensor to calculate the wind speed value through the propagation speed of ultrasonic waves in the air;

[0096] Voltage sensor unit: used to measure the voltage in the distribution network. The voltage sensor unit uses the principle of electromagnetic induction to calculate the voltage value by measuring the changes in the electromagnetic field generated around the conductor;

[0097] Current sensor unit: used to measure the current in the distribution network. This unit uses a Hall effect sensor to calculate the current value by measuring the change in the magnetic field generated by the current flowing through the conductor. Through the close cooperation of the above units, the environmental perception module can collect environmental information of the distribution network in real time and accurately, providing reliable data support for subsequent fault detection and scheduling decisions.

[0098] The fault detection and locating module includes a data acquisition unit, a fault detection unit and a fault locating unit; wherein:

[0099] Data acquisition unit: used to receive and process environmental information from the environmental perception module and operating data from the distribution network operation system, and realize real-time synchronization of environmental information and operating data through a preset data interface;

[0100] Fault detection unit: used to analyze the environmental information and operating data received by the data acquisition unit and identify abnormal conditions in the distribution network. The fault detection unit uses a pattern recognition algorithm to determine the time point when the fault occurs by comparing the normal operating data with the abnormal changes in the current data;

[0101] Fault location unit: used to locate the specific fault point according to the abnormal state identified by the fault detection unit. The fault location unit adopts a fault location algorithm based on power lines to locate the geographical location of the fault point by analyzing the traveling waves and impedance changes generated when the fault occurs. Through the detailed description of the fault detection and location module and the specific functions of each unit, it is clarified how to detect distribution network faults and locate fault points in real time according to environmental information and power grid operation data. The combination of data acquisition unit, fault detection unit and fault location unit enables the fault detection and location module to provide high-precision and high-real-time fault detection and location services, which provides a solid foundation for subsequent evaluation, early warning and optimized scheduling decisions.

[0102] The fault detection unit includes:

[0103] Data preprocessing: preprocess the received environmental information and operation data, including denoising, normalization and data smoothing;

[0104] Feature extraction: Extract features from preprocessed data. Specifically, fast Fourier transform (FFT) is used to convert time domain data into frequency domain data for subsequent pattern recognition. The specific extraction formula is:

[0105] Among them, x n ′ is the nth normalized time domain data point, X(f j ) is the jth frequency component, N is the total number of data points, f j is the jth frequency;

[0106] Normal mode establishment: Use historical normal operation data to establish a normal mode, use principal component analysis (PCA) to reduce the dimension, and extract characteristic components. The formula is: Among them, X m is the mth eigenvalue, W km is the element in the PCA weight matrix, Z k is the kth principal component, M is the total number of eigenvalues;

[0107] Real-time data monitoring: Perform the same feature extraction algorithm as in the feature extraction step on the current real-time data to obtain the real-time feature value. The formula is: Among them, x current,p is the current p-th time domain data point, X current (f l ) is the current lth frequency component, P is the total number of current data points, f l is the current lth frequency;

[0108] Anomaly detection: Compare the real-time feature value with the normal mode and calculate the feature difference. Specifically, the Mahalanobis distance is used to evaluate the degree of anomaly. The formula is:

[0109] Among them, X current,t is the current eigenvector at time point t, μ N is the mean vector of the normal mode, ∑ N is the covariance matrix of the normal mode, D M,t is the Mahalanobis distance at time point t;

[0110] Fault time point determination: When the Mahalanobis distance exceeds the preset threshold, a fault is determined to have occurred, and the time point t at which the fault occurred is recorded. f , the formula is: f =min{t:D M,t >Threshold}, where t f is the time point when the fault occurs, and Threshold is the preset threshold. Through data preprocessing, feature extraction, normal mode establishment, real-time data monitoring, anomaly detection and fault time point determination, the accuracy and timeliness of fault detection are ensured, the overall response speed and fault handling capability of the system are improved, thereby effectively ensuring the stability and reliability of power supply.

[0111] The fault location unit includes:

[0112] Traveling wave detection: When a fault occurs, the voltage and current waveforms on the power line are collected, and the arrival time of the traveling wave is detected. The formula is: Among them, t w1 is the arrival time of the traveling wave, d is the distance from the fault point to the detection point, and v is the propagation speed of the traveling wave;

[0113] Traveling wave time difference calculation: detect the arrival time of traveling waves at multiple locations and calculate the traveling wave time difference. The formula is: Δt w =t w1 -t w2 , where Δt w is the traveling wave time difference between the two detection points, t w2 is the arrival time of the traveling wave at the second detection point;

[0114] Traveling wave fault distance calculation: According to the traveling wave time difference, calculate the distance from the fault point to the detection point. The formula is: Among them, d fault is the distance from the fault point to the midpoint of the two detection points;

[0115] Impedance measurement: When a fault occurs, measure the impedance change of the power line and calculate the impedance of the fault point. The formula is: Among them, Z fault is the impedance of the fault point, V fault is the voltage at the fault point, I foult is the current at the fault point;

[0116] Impedance fault distance calculation: According to the impedance of the fault point, calculate the distance from the fault point to the detection point. The formula is: Among them, d imp is the distance to the fault point obtained by impedance measurement, Z line is the line impedance per unit length, L is the total length of the line;

[0117] Comprehensive positioning: Combine the calculation results of the traveling wave method and the impedance method to determine the final geographical location of the fault point. The formula is: Among them, d final The distance to the fault point after comprehensive positioning; by analyzing the traveling waves and impedance changes generated when the fault occurs, the geographical location of the fault point is accurately located. The steps of traveling wave detection, time difference calculation, impedance measurement and comprehensive positioning ensure the accuracy and reliability of fault positioning, and provide key data support for subsequent emergency repair scheduling.

[0118] The evaluation and early warning module includes a data analysis unit, a trend prediction unit, an early warning judgment unit and an early warning signal unit; wherein:

[0119] Data analysis unit: used to receive and process fault information from the fault detection and positioning module, compare and analyze historical data and real-time data, and identify potential fault modes. The data analysis unit specifically uses the support vector machine algorithm for classification and prediction. The formula is: Among them, x is the input data, α i and i is the model parameter, K(x i ,x) is the kernel function, b is the bias;

[0120] Trend prediction unit: It is used to predict fault trends based on the output results of the data analysis unit using a long short-term memory network to determine the potential development direction and severity of the fault; the formula is:

[0121] h t =σ(W h ·[h t-1 ,x t ]+b h ), where h t is the hidden state at time t, W h is the weight matrix, x t is the input at time t, b h is bias;

[0122] Early warning judgment unit: It is used to judge whether it is necessary to issue an early warning signal according to the prediction results of the trend prediction unit. When the prediction results show that the fault will reach a serious level within the preset time, the early warning signal will be triggered. The judgment formula is:

[0123] Among them, P(fault t+k ) is the predicted value of the failure probability at time t+k, and Threshold is the warning threshold;

[0124] Early warning signal unit: used to send out an early warning signal when the early warning judgment unit triggers the early warning. The early warning signal will be sent out through the system interface display, sound alarm and SMS notification; through the close cooperation of the above units, the evaluation and early warning module can intelligently evaluate the fault information, predict the development trend of the fault, and send out an early warning signal when the prediction result shows that the fault may reach a serious level within the preset time.

[0125] The optimization scheduling decision module includes a data integration unit, a target definition unit, an individual generation unit, a fitness evaluation unit, an evolutionary operation unit, and an optimal solution selection unit; among which:

[0126] Data fusion unit: used to receive and integrate multi-source data from the environment perception module, fault detection and positioning module and historical database, including real-time environment information, current fault information and historical fault data; and ensure data accuracy and consistency through data cleaning and feature extraction;

[0127] Objective definition unit: used to define multiple optimization objectives according to the preset scheduling objectives, including minimizing the repair time, minimizing the repair cost and optimizing resource allocation, where each objective is clearly described by specific parameters and constraints; the formula is:

[0128]

[0129] Among them, T i is the repair time of the i-th task, C i is the repair cost of the ith task, R i is the resource utilization efficiency of the i-th task;

[0130] Individual generation unit: used to generate the initial solution in the multi-objective evolutionary algorithm, that is, multiple emergency repair scheduling plans, each of which contains the specific content of the allocation of emergency repair personnel, equipment and material scheduling; the formula is:

[0131] X={x1,x2,…,x n}, where x i is the i-th initial solution, representing a specific scheduling plan;

[0132] The fitness evaluation unit is used to evaluate the fitness of each scheduling scheme and calculate the comprehensive score of each scheme according to the definition of the objective function. The calculation formula is: F(X) = w1f1(X) + w2f2(X) + w3f3(X), where F(X) is the comprehensive fitness score and w1, w2, w3 are the weights of the objective function;

[0133] Evolutionary operation unit: used to perform operations of multi-objective evolutionary algorithms, including selection, crossover and mutation;

[0134] The selection operation formula is: selected =Select(P population ,F(X)), where P selected is the population after selection, P population is the current population, F(X) is the fitness score;

[0135] The crossover operation formula is: X new =Crossover(X parent1 ,X parent2 ), where X new is the newly generated individual, X parent11 ,X parent2 is the selected parent individual;

[0136] The mutation operation formula is: mutated =Mutate(X individual ), where X mutated is the individual after mutation, X individual For the individual who undergoes the mutation;

[0137] The optimal solution selection unit is used to select the final optimal emergency repair scheduling solution from the evolved solution set; through the close cooperation of the above units, the optimization scheduling decision module can use the multi-objective evolutionary algorithm, combined with fault information, environmental data and historical data, to generate a scientific, reasonable and efficient emergency repair scheduling solution, providing a basis for subsequent emergency resource allocation.

[0138] The optimal solution selection unit includes:

[0139] Non-dominated sorting: Perform non-dominated sorting on the evolved solution set and divide the solutions into different levels. The goal of non-dominated sorting is to find the maximum non-dominated solution. The formula is expressed as:

[0140] S i ={X j |X j DominateX i};

[0141] n i =|{X j |X iX j Dominance}; Among them, S i is the set of solutions dominated by the ith solution, n i is the number of solutions that dominate the ith solution;

[0142] Crowding calculation: Calculate the crowding of each solution to measure the density of the solution in the target space. The formula is: Among them, I i is the congestion degree of the ith solution, and are the target values ​​of the previous and next neighbor solutions of the ith solution on the mth objective function, and are the maximum and minimum values ​​of the mth objective function respectively;

[0143] Solution selection: According to the non-dominated sorting and congestion calculation results, the optimal solution is selected from the evolved solution set, with priority given to solutions with high non-dominated levels, and solutions with high congestion levels are selected within the same non-dominated level. The specific formula is: Among them, X optimal is the optimal solution, Rank(X) is the non-dominated rank of solution X, and I X is the congestion degree of scheme X, P final is the final set of solutions; through a detailed description of the optimal solution selection unit, the method of selecting the final optimal emergency repair scheduling solution from the evolved solution set is clarified. The steps of non-dominated sorting, congestion calculation and solution selection ensure that the selected optimal solution has high fitness and diversity based on multi-objective optimization, which provides key support for the efficient execution of the system and improves the reliability and stability of the entire smart distribution network emergency repair scheduling system.

[0144] The emergency resource allocation module includes a personnel allocation unit, an equipment allocation unit, a material allocation unit, and a dispatch execution unit; among which:

[0145] Personnel dispatching unit: used to dynamically allocate emergency repair personnel according to the optimal emergency repair scheduling plan. The personnel dispatching unit matches the corresponding emergency repair personnel to the designated fault location according to the personnel's skills, experience and current availability. The specific steps are as follows:

[0146] Step 1: Retrieve information of all available repair personnel from the personnel database, including skills, experience, and current status;

[0147] Step 2: According to the task requirements in the optimal scheduling plan, use the personnel matching algorithm to select the most suitable personnel for task allocation;

[0148] Step 3: Generate a dispatch order and notify relevant personnel to go to the designated location.

[0149] Equipment dispatching unit: used to dynamically allocate emergency repair equipment according to the optimal emergency repair scheduling plan, specifically matching the corresponding equipment to the designated fault location according to the type, performance and current availability of the equipment; the specific steps are as follows:

[0150] Step 1: Retrieve information about all available devices from the device database, including type, performance, and current status;

[0151] Step 2: According to the task requirements in the optimal scheduling plan, use the device matching algorithm to select the most suitable device for task allocation;

[0152] Step 3: Generate a dispatch order and notify relevant departments to prepare equipment.

[0153] Material dispatching unit: used to dynamically allocate emergency repair materials according to the optimal emergency repair dispatch plan; according to the type, quantity and inventory status of the materials, the corresponding materials are matched to the designated fault location; the specific steps are as follows:

[0154] Step 1: Retrieve information about all available materials from the material database, including type, quantity, and inventory status;

[0155] Step 2: Based on the task requirements in the optimal scheduling plan, use the material matching algorithm to select the most needed materials for task allocation;

[0156] Step 3: Generate a dispatch order and notify the warehouse department to prepare materials.

[0157] Dispatching execution unit: used to integrate the dispatching results of personnel deployment unit, equipment deployment unit and material deployment unit, generate the final comprehensive dispatching plan and execute it; through the combination of personnel deployment unit, equipment deployment unit, material deployment unit and dispatching execution unit, the emergency resource allocation module can provide efficient, accurate and dynamically optimized resource allocation services, which provides key support for the system's rapid response and efficient execution, and improves the reliability and stability of the entire smart distribution network emergency repair dispatching system.

[0158] The feedback self-learning module includes a data recording unit, an execution effect analysis unit, a self-learning unit, and an optimization and improvement unit; wherein:

[0159] Data recording unit: used to record the actual execution status of the repair process, including the arrival time of personnel, equipment usage, material consumption and fault repair time;

[0160] Execution effect analysis unit: Based on the recorded data, evaluate the deviation between the actual execution results and the planned solution, and identify the factors affecting the execution effect. The calculation formula is:

[0161] Among them, E deviationFor execution deviation, A i is the actual execution value of item i, P i is the planned value of the ith item, and n is the total number of items;

[0162] Self-learning unit: Based on the results of the execution effect analysis, self-learning and optimization are carried out. Specifically, the reinforcement learning algorithm is used to continuously update the model parameters to improve the decision-making ability and response speed of the system. The formula is: Q(s,a)=Q(s,a)+α[r+γmax a′ Q(s′,a′)-Q(s,a)], where Q(s,a) is the value of performing action a in state s, α is the learning rate, r is the immediate reward, γ is the discount factor, s′ is the next state, and a′ is the next action;

[0163] Optimization and improvement unit: used to adjust the dispatching strategy and parameter settings according to the output results of the self-learning unit; through the close cooperation of the above units, the feedback self-learning module can record the actual execution status of the emergency repair process in real time, analyze the execution effect of the dispatching plan, and continuously optimize the system's dispatching strategy through self-learning algorithms. This provides key support for the continuous improvement and efficient operation of the system, and improves the reliability and stability of the entire smart distribution network emergency repair dispatching system.

[0164] The present invention is intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. Intelligent distribution network emergency repair dispatching system, characterized in that: It includes environment perception module, fault detection and positioning module, assessment and early warning module, optimization and scheduling decision module, emergency resource allocation module and feedback self-learning module; among which: Environmental perception module: used to collect distribution network environmental information in real time, including temperature, humidity, wind speed, voltage, and current parameters; Fault detection and positioning module: connected to the environment perception module, used to detect distribution network faults in real time according to environmental information and power grid operation data, and locate the fault point; Evaluation and early warning module: connected to the fault detection and positioning module, intelligently evaluates fault information, predicts fault development trends, and issues early warning signals when necessary; Optimization scheduling decision module: connected to the evaluation and early warning module, using a multi-objective evolutionary algorithm to generate an optimal emergency repair scheduling plan based on fault information, environmental information and historical data; Emergency resource allocation module: connected to the optimization scheduling decision module, dynamically allocates emergency repair personnel, equipment and materials according to the optimal emergency repair scheduling plan to ensure efficient completion of the emergency repair task; Feedback self-learning module: connected to the emergency resource allocation module, used to record the actual execution status during the emergency repair process, analyze the execution effect of the scheduling plan, perform self-learning and optimization improvements, and improve the overall performance of the system.

2. The intelligent distribution network emergency repair dispatching system according to claim 1 is characterized in that: The environment sensing module includes a temperature sensor unit, a humidity sensor unit, a wind speed sensor unit, a voltage sensor unit and a current sensor unit; wherein: Temperature sensor unit: used to measure the temperature around the distribution network. The temperature sensor unit uses a thermistor to calculate the specific temperature value through the change of resistance value; Humidity sensor unit: used to measure the humidity around the distribution network. The humidity sensor unit uses a capacitive humidity sensor to calculate the specific humidity value through the change of capacitance value; Wind speed sensor unit: used to measure the wind speed around the distribution network. The wind speed sensor unit uses an ultrasonic wind speed sensor to calculate the wind speed value through the propagation speed of ultrasonic waves in the air; Voltage sensor unit: used to measure the voltage in the distribution network. The voltage sensor unit uses the principle of electromagnetic induction to calculate the voltage value by measuring the changes in the electromagnetic field generated around the conductor; Current Sensor Unit: Used to measure the current in the power distribution network, this unit uses a Hall Effect sensor to calculate the current value by measuring the change in the magnetic field generated by the current flowing through the conductor.

3. The intelligent distribution network emergency repair dispatching system according to claim 1 is characterized in that: The fault detection and locating module includes a data acquisition unit, a fault detection unit and a fault locating unit; wherein: Data acquisition unit: used to receive and process environmental information from the environmental perception module and operating data from the distribution network operation system, and realize real-time synchronization of environmental information and operating data through a preset data interface; Fault detection unit: used to analyze the environmental information and operation data received by the data acquisition unit and identify abnormal conditions in the distribution network. The fault detection unit uses a pattern recognition algorithm to determine the time point when the fault occurs by comparing the normal operating data with the abnormal changes in the current data; Fault location unit: used to locate the specific fault point according to the abnormal state identified by the fault detection unit. The fault location unit adopts a fault location algorithm based on power lines to locate the geographical location of the fault point by analyzing the traveling waves and impedance changes generated when the fault occurs.

4. The intelligent distribution network emergency repair dispatching system according to claim 3 is characterized in that: The fault detection unit comprises: Data preprocessing: preprocess the received environmental information and operation data, including denoising, normalization and data smoothing; Feature extraction: Extract features from preprocessed data. Specifically, fast Fourier transform is used to convert time domain data into frequency domain data for subsequent pattern recognition. The specific extraction formula is: Among them, x n ′ is the nth normalized time domain data point, X(f j ) is the jth frequency component, N is the total number of data points, f j is the jth frequency; Normal mode establishment: Use historical normal operation data to establish a normal mode, use principal component analysis to reduce the dimension, and extract characteristic components. The formula is: Among them, X m is the mth eigenvalue, W km is the element in the PCA weight matrix, Z k is the kth principal component, M is the total number of eigenvalues; Real-time data monitoring: Perform the same feature extraction algorithm as in the feature extraction step on the current real-time data to obtain the real-time feature value. The formula is: Among them, x current,p is the current p-th time domain data point, X current (f l ) is the current lth frequency component, P is the total number of current data points, f l is the current lth frequency; Anomaly detection: Compare the real-time feature value with the normal mode and calculate the feature difference. Specifically, the Mahalanobis distance is used to evaluate the degree of anomaly. The formula is: Among them, X current,t is the current eigenvector at time point t, μ N is the mean vector of the normal mode, ∑ N is the covariance matrix of the normal mode, D M,t is the Mahalanobis distance at time point t; Fault time point determination: When the Mahalanobis distance exceeds the preset threshold, a fault is determined to have occurred and the time point t of the fault occurrence is recorded. f , the formula is: f =min{t:D M,t >Threshold}, where t f is the time point when the fault occurs, and Threshold is the preset threshold.

5. The intelligent distribution network emergency repair dispatching system according to claim 4 is characterized in that: The fault location unit comprises: Traveling wave detection: When a fault occurs, the voltage and current waveforms on the power line are collected, and the arrival time of the traveling wave is detected. The formula is: Among them, t w1 is the arrival time of the traveling wave, d is the distance from the fault point to the detection point, and v is the propagation speed of the traveling wave; Traveling wave time difference calculation: detect the arrival time of traveling waves at multiple locations and calculate the traveling wave time difference. The formula is: Δt w =t w1 -t w2 , where Δt w is the traveling wave time difference between the two detection points, t w2 is the arrival time of the traveling wave at the second detection point; Traveling wave fault distance calculation: According to the traveling wave time difference, calculate the distance from the fault point to the detection point. The formula is: Among them, d fault is the distance from the fault point to the midpoint of the two detection points; Impedance measurement: When a fault occurs, measure the impedance change of the power line and calculate the impedance of the fault point. The formula is: Among them, Z fault is the impedance of the fault point, V fault is the voltage at the fault point, I foult is the current at the fault point; Impedance fault distance calculation: According to the impedance of the fault point, calculate the distance from the fault point to the detection point. The formula is: Among them, d imp is the distance to the fault point obtained by impedance measurement, Z line is the line impedance per unit length, L is the total length of the line; Comprehensive positioning: Combine the calculation results of the traveling wave method and the impedance method to determine the final geographical location of the fault point. The formula is: Among them, d final is the distance to the fault point after comprehensive positioning.

6. The intelligent distribution network emergency repair dispatching system according to claim 1, characterized in that: The evaluation and early warning module includes a data analysis unit, a trend prediction unit, an early warning judgment unit and an early warning signal unit; wherein: Data analysis unit: used to receive and process fault information from the fault detection and positioning module, compare and analyze historical data and real-time data, and identify potential fault modes. The data analysis unit specifically uses the support vector machine algorithm for classification and prediction. The formula is: Among them, x is the input data, α i and i is the model parameter, K(x i ,x) is the kernel function, b is the bias; Trend prediction unit: It is used to predict fault trends based on the output results of the data analysis unit using a long short-term memory network to determine the potential development direction and severity of the fault; the formula is: h t =σ(W h ·[h t-1 ,x t ]+b h ), where h t is the hidden state at time t, W h is the weight matrix, x t is the input at time t, b h is bias; Early warning judgment unit: It is used to judge whether it is necessary to issue an early warning signal according to the prediction results of the trend prediction unit. When the prediction results show that the fault will reach a serious level within the preset time, the early warning signal will be triggered. The judgment formula is: Among them, P(fault t+k ) is the predicted value of the failure probability at time t+k, and Threshold is the warning threshold; Early warning signal unit: used to send out early warning signals when the early warning judgment unit triggers the early warning. The early warning signals will be sent out through system interface display, sound alarm and SMS notification.

7. The intelligent distribution network emergency repair dispatching system according to claim 1, characterized in that: The optimization scheduling decision module includes a data integration unit, a target definition unit, an individual generation unit, a fitness evaluation unit, an evolutionary operation unit and an optimal solution selection unit; wherein: Data fusion unit: used to receive and integrate multi-source data from the environment perception module, fault detection and positioning module and historical database, including real-time environment information, current fault information and historical fault data; and ensure data accuracy and consistency through data cleaning and feature extraction; Target definition unit: used to define multiple optimization targets according to the preset scheduling targets, including minimizing the repair time, minimizing the repair cost and optimizing resource allocation; the formula is: Among them, T i is the repair time of the i-th task, C i is the repair cost of the ith task, R i is the resource utilization efficiency of the i-th task; Individual generation unit: used to generate the initial solution in the multi-objective evolutionary algorithm, that is, multiple emergency repair scheduling plans, each of which contains the specific content of the allocation of emergency repair personnel, equipment and material scheduling; the formula is: X={x1,x2,…,x n }, where x i is the i-th initial solution, representing a specific scheduling plan; The fitness evaluation unit is used to evaluate the fitness of each scheduling scheme and calculate the comprehensive score of each scheme according to the definition of the objective function. The calculation formula is: F(X) = w1f1(X) + w2f2(X) + w3f3(X), where F(X) is the comprehensive fitness score and w1, w2, w3 are the weights of the objective function; Evolutionary operation unit: used to perform operations of multi-objective evolutionary algorithms, including selection, crossover and mutation; The selection operation formula is: selected =Select(P population ,F(X)), where P selected is the population after selection, P population is the current population, F(X) is the fitness score; The crossover operation formula is: X new =Crossover(X parent1 ,X parent2 ), where X new is the newly generated individual, X parent11 ,X parent2 is the selected parent individual; The mutation operation formula is: mutated =Mutate(X individual ), where X mutated is the individual after mutation, X individual For the individual who undergoes the mutation; The optimal solution selection unit is used to select the final optimal emergency repair scheduling solution from the evolved solution set.

8. The intelligent distribution network emergency repair dispatching system according to claim 7, characterized in that: The optimal solution selection unit comprises: Non-dominated sorting: Perform non-dominated sorting on the evolved solution set and divide the solutions into different levels. The goal of non-dominated sorting is to find the maximum non-dominated solution. The formula is expressed as: S i ={X j |X j DominateX i }; n i =|{X j |X i X j Dominance}; Among them, S i is the set of solutions dominated by the ith solution, n i is the number of solutions that dominate the ith solution; Crowding calculation: Calculate the crowding of each solution to measure the density of the solution in the target space. The formula is: Among them, I i is the congestion degree of the ith solution, and are the target values ​​of the previous and next neighbor solutions of the ith solution on the mth objective function, and are the maximum and minimum values ​​of the mth objective function respectively; Scheme selection: According to the non-dominated sorting and crowding calculation results, the optimal scheme is selected from the evolved scheme set, with the scheme with high non-dominated level being given priority, and the scheme with high crowding level being selected within the same non-dominated level. The specific formula is: X optimal = argmax X∈Pfinal (Rank(X),I X ), where X optimal is the optimal solution, Rank(X) is the non-dominated rank of solution X, and I X is the congestion degree of plan X, P final The final solution set.

9. The intelligent distribution network emergency repair dispatching system according to claim 1, characterized in that: The emergency resource allocation module includes a personnel allocation unit, an equipment allocation unit, a material allocation unit and a scheduling execution unit; wherein: Personnel dispatching unit: used to dynamically allocate emergency repair personnel according to the optimal emergency repair dispatching plan. The personnel dispatching unit matches the corresponding emergency repair personnel to the designated fault location according to the personnel's skills, experience and current availability; Equipment dispatching unit: used to dynamically allocate emergency repair equipment according to the optimal emergency repair scheduling plan, and match the corresponding equipment to the designated fault location according to the type, performance and current availability of the equipment; Material dispatching unit: used to dynamically allocate emergency repair materials according to the optimal emergency repair dispatch plan; according to the type, quantity and inventory status of the materials, the corresponding materials are matched to the designated fault location; Scheduling execution unit: used to integrate the scheduling results of the personnel deployment unit, equipment deployment unit and material deployment unit, generate the final comprehensive scheduling plan and execute it.

10. The intelligent distribution network emergency repair dispatching system according to claim 1, characterized in that: The feedback self-learning module includes a data recording unit, an execution effect analysis unit, a self-learning unit and an optimization and improvement unit; wherein: Data recording unit: used to record the actual execution status of the repair process, including the arrival time of personnel, equipment usage, material consumption and fault repair time; Execution effect analysis unit: Based on the recorded data, evaluate the deviation between the actual execution results and the planned solution, and identify the factors affecting the execution effect. The calculation formula is: Among them, E deviation For execution deviation, A i is the actual execution value of item i, P i is the planned value of the ith item, and n is the total number of items; Self-learning unit: Based on the results of the execution effect analysis, self-learning and optimization are carried out. Specifically, the reinforcement learning algorithm is used to continuously update the model parameters to improve the decision-making ability and response speed of the system. The formula is: Q(s,a)=Q(s,a)+α[r+γmax a′ Q(s′,a′)-Q(s,a)], where Q(s,a) is the value of performing action a in state s, α is the learning rate, r is the immediate reward, γ is the discount factor, s′ is the next state, and a′ is the next action; Optimization and improvement unit: used to adjust the scheduling strategy and parameter settings according to the output results of the self-learning unit.

Citation Information

Cited By

  • Ground fault troubleshooting system for ten-thousand-volt power distribution network

    CN119689158A

  • Power distribution automatic control system based on fault positioning

    CN120073717A

  • Intelligent jacking system for bridge construction

    CN120197279A

  • Automatic alarm method and system based on distribution line on-line fault detection

    CN120507608A

  • Power distribution network line fault intelligent inspection and real-time early warning system and method based on AIoT

    CN120542703A