A distribution network fault emergency repair dispatching system and method

Abstract

The invention discloses a distribution network fault emergency repair dispatching system and method, which relates to the field of emergency repair dispatching technology, and comprises the following steps: collecting distribution network operation data, constructing an equipment health evaluation index set, and outputting an equipment fault probability distribution map through a fault probability prediction model; establishing an equipment space topology database based on a geographic information system, locating the fault point through a Dijkstra algorithm and generating a three-dimensional geographic coordinate label; constructing a multi-dimensional emergency repair resource portrait library, recording the skill level of the emergency repair team, the inventory status of materials and the real-time location information of the vehicle; establishing a dynamic priority evaluation model, and generating an emergency repair task priority sequence according to the urgency of the fault, the level of the affected user and the criticality of the equipment; using a multi-objective grey wolf optimization algorithm to solve the emergency repair resource scheduling plan, and setting the objective function in the direction of minimizing the response time, the path cost and the resource waste rate; issuing the optimal scheduling instruction through a collaborative scheduling platform, and continuously monitoring the task execution status.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid dispatching, and more specifically, to a distribution network fault emergency repair dispatching system and method. Background Art

[0002] In modern society, power supply is the basic guarantee for maintaining the normal operation of production and life. As the terminal link of power transmission, the distribution network directly faces the vast number of users, and its operation stability and reliability are crucial. Once the distribution network fails, it will bring great inconvenience to users and even cause serious economic losses.

[0003] At present, there are many technical means for emergency repair and dispatching of distribution network faults. The traditional method mainly relies on manual experience for fault judgment and emergency repair dispatching. After receiving the fault report, the emergency repair personnel roughly judge the type and location of the fault based on their own accumulated experience and understanding of the regional power grid, and then organize emergency repair work. With the development of technology, some information-based auxiliary means have also emerged, such as using SCADA (data acquisition and monitoring system) to monitor the operating data of the distribution network in real time. Once abnormal data appears, the system can issue an alarm to help operation and maintenance personnel find the fault. There is also fault location technology based on geographic information system (GIS), which improves the accuracy of fault location to a certain extent by combining information such as lines and equipment of the distribution network with geographic information.

[0004] However, the existing technology still has obvious limitations. Although manual experience judgment has a certain effect on the handling of some common faults, it is highly subjective and lacks systematicity and accuracy. When faced with complex faults, misjudgment is prone to occur, resulting in low repair efficiency. Although the SCADA system can monitor data in real time, it can only detect abnormalities in operating parameters and cannot accurately locate the fault point. Moreover, for some intermittent faults, the monitoring effect is not good. Although the GIS-based fault location technology can combine geographic information, the positioning accuracy will be greatly reduced when the distribution network structure is complex, there are multiple similar lines or the equipment information is not updated in time. In addition, the existing technology generally has deficiencies in the optimal configuration of fault repair resources. It cannot quickly and reasonably allocate personnel, materials and equipment according to factors such as the urgency of the fault, the type and quantity of repair resources required, resulting in waste or unreasonable allocation of repair resources, affecting the overall repair efficiency. At the same time, there is a lack of effective integration between the various technologies, data sharing is difficult, and it is impossible to form an efficient and coordinated fault repair scheduling system.

[0005] For example, the invention patent announcement with announcement number: CN111539566B discloses a distribution network multi-fault emergency repair and recovery method and system considering pre-disaster pre-scheduling, including: determining the probability of component outage, generating an expected accident set; calculating the system DC current; exiting the expected accident concentrated load nodes in turn to determine whether the system has a fault, if so, merging the faulty node into the accident determination set; determining whether all nodes in the system have been judged to be finished, if so, obtaining the accident determination set; if not, obtaining the next node information; determining the system fault location and the partition to which the faulty component belongs based on the accident determination set and constructing a path scheduling plan; using a DG output prediction method based on nearest neighbor clustering to determine the DG output prediction curve; and determining the distribution network emergency repair and recovery plan by constructing a multi-source collaborative optimization model based on the DG output prediction curve. The present invention can effectively improve the efficiency of fault emergency repair, optimize the emergency repair sequence, adjust the emergency repair plan in real time, and optimize the power supply capacity of the available resources of the distribution network.

[0006] For example, the invention patent with announcement number: CN113627733B announces a method and system for dynamic repair of a post-disaster distribution network, wherein the method includes: obtaining the state information of the post-disaster distribution network to be dynamically repaired through the environment side; inputting the state information of the post-disaster distribution network to be dynamically repaired into the intelligent body side including the reinforcement learning model, and obtaining the dynamic repair result of the post-disaster distribution network output by the intelligent body; wherein the intelligent body side obtains the reinforcement learning state from the environment side, acts on the distribution network based on the strategy selection action, and receives the corresponding reward value and the next state for iterative training, and obtains the reinforcement learning model after continuously updating the network parameters. The embodiment of the present invention realizes the high efficiency and high accuracy of the post-disaster repair of the distribution network under the conditions of coordinated deployment of multiple repair teams, deep coupling of repair and restoration, and power transfer and restoration, and uncertainty of the disaster situation.

[0007] The above disclosed technical solutions have at least the following technical problems: the existing technologies generally have deficiencies in the optimal allocation of fault repair resources, and cannot quickly and reasonably allocate personnel, materials and equipment according to factors such as the urgency of the fault, the type and quantity of repair resources required, resulting in waste of repair resources or unreasonable allocation, affecting the overall repair efficiency. At the same time, there is a lack of effective integration between the various technologies, data sharing is difficult, and it is impossible to form an efficient and coordinated fault repair scheduling system.

[0008] In view of the above problems, the present invention proposes a solution. Summary of the invention

[0009] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a distribution network fault repair dispatching system and method, which, through a multi-source data fusion architecture, solves the common deficiencies in the prior art in the optimal configuration of fault repair resources, and the problem that personnel, materials and equipment cannot be quickly and reasonably deployed according to factors such as the urgency of the fault, the type and quantity of repair resources required, etc.

[0010] To achieve the above object, the present invention provides the following technical solutions:

[0011] A distribution network fault emergency repair dispatching method comprises the following steps: real-time collection of distribution network equipment operation data, construction of equipment health evaluation index set, and output of equipment fault probability distribution map through fault probability prediction model; establishment of equipment spatial topology database based on geographic information system, and when equipment abnormality is detected, locating the fault point through Dijkstra algorithm and generating three-dimensional geographic coordinate label; building a multi-dimensional emergency repair resource portrait library, dynamically recording the skill level of the emergency repair team, material inventory status and real-time vehicle location information; establishing a dynamic priority evaluation model, and generating an emergency repair task priority sequence according to the emergency urgency, the level of affected users and the criticality of equipment; using a multi-objective grey wolf optimization algorithm to solve the emergency repair resource dispatching scheme, and setting the objective function in the direction of minimizing response time, path cost and resource waste rate; issuing the optimal dispatching instruction through a collaborative dispatching platform, and continuously monitoring the task execution status.

[0012] In a preferred embodiment, the real-time collection of distribution network equipment operation data, the construction of equipment health evaluation index set, and the output of equipment failure probability distribution map through the fault probability prediction model are specifically as follows: based on the sensor, the distribution network equipment operation data is obtained, and the data acquisition terminal receives the operation data transmitted by the sensor, performs filtering, amplification, and analog-to-digital conversion preprocessing to remove noise interference; based on the physical structure and working principle of the distribution network equipment, the parameters closely related to the health status of the equipment are obtained based on regression analysis as candidate indicators; the candidate indicators are divided into electrical performance indicators, mechanical performance indicators, and chemical performance indicators, and a multi-level indicator system is constructed, the overall health of the equipment is used as the top-level indicator, the top-level indicator is divided into several first-level indicators, and each first-level indicator is further decomposed into several second-level indicators; based on the multi-level indicator system, the XGBoost-LSTM hybrid model is used to obtain the fault probability prediction model; the geographical location information of the distribution network equipment is associated with the equipment failure probability prediction result, and the equipment location-fault probability mapping relationship is established; the geographic information system software is selected as the map drawing platform, and the equipment is marked on the map according to the actual location of the equipment, and different colors are used for visual display according to the size of the failure probability, and the equipment with high failure probability is marked with red, and the equipment with low failure probability is marked with green.

[0013] In a preferred embodiment, the establishment of a device spatial topology database based on a geographic information system is specifically as follows: collecting models, specifications, and performance parameters of transformers, circuit breakers, poles, and cables in the distribution network; obtaining the geographical location information of the equipment, including latitude and longitude coordinates and elevation information; abstracting the equipment into nodes, and the connection relationships between the equipment into edges, and constructing a node-edge topological structure; establishing a database structure, including the definition of data tables, the setting of fields, and the establishment of data relationships, storing device information, geographic information, and topological relationships; importing the preprocessed data into the database of the geographic information system to establish a device spatial topology database.

[0014] In a preferred embodiment, the fault point is located by the Dijkstra algorithm and a three-dimensional geographic coordinate label is generated, specifically: based on the device space topology database, each device node is regarded as a vertex in the graph, and the connection between devices is regarded as an edge; taking the monitoring center or a known power supply node as the starting point, the Dijkstra algorithm is used to calculate the shortest path to each node, and when the abnormal device node is traversed, the path information of the node is recorded; the shortest path obtained by the calculation is analyzed, and the range of the fault point is obtained in combination with the distribution of abnormal equipment and the law of fault propagation; the latitude and longitude coordinates and elevation information of the fault point are extracted from the database of the geographic information system to generate three-dimensional geographic coordinates; the three-dimensional geographic coordinates of the fault point, the fault type, and the occurrence time are combined into a label, and visualized on the geographic information system map.

[0015] In a preferred embodiment, the multi-dimensional emergency repair resource portrait library is constructed to dynamically record the skill level of the emergency repair team, the inventory status of materials and the real-time location information of the vehicle, specifically: the main dimensions for constructing the portrait library are obtained, and the main dimensions include the skill level of the emergency repair team, the inventory status of materials and the real-time location information of the vehicle; at the same time, specific information fields are planned under each dimension, and the information fields include the skill type of the emergency repair team, the personnel qualification certificate, the name, specification, quantity, storage location of the materials, the license plate number, model, and real-time coordinates of the vehicle; through the fusion of the on-board global positioning system and traffic big data, the current location, driving speed, estimated arrival time and available status of the emergency repair vehicle are continuously tracked; a multi-dimensional emergency repair resource portrait library is constructed according to the main dimensions and data fields.

[0016] In a preferred embodiment, the dynamic priority evaluation model is established, and the principal component analysis method is used to assign weights to each indicator by combining the fault type, power outage scope, power outage time, user level and equipment criticality factors. The evaluation model is constructed by a linear weighting method to calculate the priority score of the emergency repair task, thereby generating a priority sequence of the emergency repair task, specifically: the urgency level is divided according to the fault type, power outage scope and power outage time factors, and large-scale power outages or faults affecting important users are set to high urgency; small-scale power outages or faults that can be restored in a short time are set to low urgency; users are divided into important users and ordinary users, and the impact level of important user faults is high, and the impact level of ordinary user faults is low; equipment criticality is divided into ordinary equipment of branch lines, general transmission line equipment, and key equipment of core substations; based on the principal component analysis method, weights are assigned to the fault urgency, the level of user impact and the criticality of equipment; an evaluation model is constructed by a linear weighting method, and each indicator value is multiplied by the corresponding weight and then added to obtain the priority score of each emergency repair task; all emergency repair tasks are sorted according to the priority score calculated by the evaluation model, and tasks with high scores have high priorities, thereby generating a priority sequence of emergency repair tasks.

[0017] In a preferred embodiment, the multi-objective grey wolf optimization algorithm is used to solve the emergency repair resource scheduling scheme, and the objective function is set in the direction of minimizing the response time, path cost and resource waste rate, specifically:

[0018] A set of candidate solutions is randomly generated as the initial wolf pack; the fitness value of each candidate solution is calculated according to the objective function; the optimal solution is selected as the leader wolf according to the fitness value, and the positions of wolves at other levels are updated; by simulating the social hierarchy and hunting behavior of the gray wolf group, the candidate solutions are continuously iterated and updated until the stopping condition is met; a set of Pareto optimal solutions are selected from the final wolf pack as the emergency repair resource scheduling plan.

[0019] The technical effects and advantages of a distribution network fault emergency repair dispatching system and method of the present invention are as follows:

[0020] 1. The present invention collects equipment operation data in real time, constructs a health assessment index set, and uses the XGBoost-LSTM hybrid model to predict the probability of failure, thereby achieving accurate control of the equipment status. At the same time, a multi-objective gray wolf optimization algorithm is used to solve the emergency repair resource scheduling plan, comprehensively considering the response time, path cost and resource waste rate to ensure the scientificity and efficiency of resource scheduling. This accurate and efficient resource scheduling method can quickly locate the fault point, quickly deploy emergency repair resources, effectively shorten the fault recovery time, and improve power supply reliability.

[0021] 2. The present invention uses a collaborative dispatching platform, and the method realizes information sharing and collaborative work among multiple parties such as power companies, emergency repair teams, and material suppliers. The automatic triggering and execution of smart contracts ensures the accurate transmission of dispatching instructions and the smooth execution of tasks. The distributed ledger characteristics of blockchain enable real-time query of task execution status and automatic triggering of early warning rules, which improves the transparency and credibility of dispatching. This collaborative dispatching method helps to improve emergency repair efficiency, reduce communication costs, and enhance trust and cooperation among all parties. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The present invention is a flow chart of a method for dispatching power distribution network fault emergency repair.

[0023] Figure 2 The present invention is a schematic diagram of the structure of a distribution network fault emergency repair and dispatching system.

[0024] Figure 3 It is the failure probability distribution diagram. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0026] Embodiment 1, Figure 1 The present invention provides a distribution network fault emergency repair dispatching method, comprising the following steps:

[0027] Collect distribution network equipment operation data in real time, build equipment health evaluation indicator set, and output equipment failure probability distribution map through failure probability prediction model;

[0028] Establish a spatial topology database of equipment based on the geographic information system. When an equipment abnormality is detected, locate the fault point through the Dijkstra algorithm and generate a three-dimensional geographic coordinate label;

[0029] Build a multi-dimensional emergency repair resource portrait library to dynamically record the skill level of the emergency repair team, material inventory status and real-time vehicle location information;

[0030] Establish a dynamic priority assessment model to generate a priority sequence for emergency repair tasks based on the urgency of the fault, the level of affected users, and the criticality of the equipment;

[0031] The multi-objective grey wolf optimization algorithm is used to solve the emergency repair resource scheduling scheme, and the objective function is set in the direction of minimizing the response time, path cost and resource waste rate;

[0032] The optimal scheduling instructions are issued through the collaborative scheduling platform, and the task execution status is continuously monitored.

[0033] S1, collects distribution network equipment operation data in real time, builds a set of equipment health evaluation indicators, and outputs equipment failure probability distribution diagrams through failure probability prediction models;

[0034] The distribution network equipment operation data includes real-time monitoring data, equipment historical maintenance records and meteorological environment parameters. For transformers, real-time monitoring data includes oil temperature, winding temperature, dissolved gas content in oil, and load current; for circuit breakers, contact wear, number of operations, and opening and closing time; for transmission lines, current, voltage, conductor temperature, and insulator leakage current are collected;

[0035] Historical maintenance records include the equipment’s past maintenance time, maintenance content, and fault handling. Meteorological environmental parameters include temperature, humidity, wind speed, and rainfall.

[0036] The real-time collection of distribution network equipment operation data, the construction of equipment health evaluation index set, and the output of equipment failure probability distribution map through the failure probability prediction model are specifically as follows:

[0037] The data acquisition terminal receives the operating data from the sensor and performs filtering, amplification, analog-to-digital conversion preprocessing to remove noise interference.

[0038] According to the physical structure and working principle of the distribution network equipment, the parameters closely related to the health status of the equipment are obtained as candidate indicators based on regression analysis;

[0039] The candidate indicators are divided into electrical performance indicators, mechanical performance indicators, and chemical performance indicators, and a multi-level indicator system is constructed. The overall health of the equipment is taken as the top-level indicator, and the top-level indicator is divided into several first-level indicators, and each first-level indicator is further decomposed into several second-level indicators;

[0040] Based on the multi-level indicator system, the XGBoost-LSTM hybrid model is used to obtain the fault probability prediction model;

[0041] The geographical location information of the distribution network equipment is associated with the equipment failure probability prediction results to establish a mapping relationship between equipment location and failure probability;

[0042] Geographic Information System (GIS) software is used as the mapping platform, and the equipment is marked on the map according to its actual location. Different colors or symbols are used for visual display according to the probability of failure. Equipment with a high probability of failure is marked in red, and equipment with a low probability of failure is marked in green.

[0043] Electrical performance indicators (such as voltage, current, power factor), mechanical performance indicators (such as vibration and wear of equipment), and chemical performance indicators (such as the chemical composition of transformer oil).

[0044] The XGBoost-LSTM hybrid model is used to obtain the fault probability prediction model, specifically:

[0045] Extract features related to equipment failure from a multi-level indicator system;

[0046] Use the cross-validation method to evaluate and screen the extracted features, and finally obtain the core features as the input of the model;

[0047] Use the XGBoost algorithm to perform preliminary feature conversion and modeling on the input data to extract the nonlinear features of the data;

[0048] The output of the XGBoost layer is used as input to capture the temporal characteristics of the data through the LSTM network;

[0049] The XGBoost-LSTM hybrid model is trained with time series features. By continuously adjusting the model parameters, the error between the predicted results and the actual fault labels is minimized to obtain a trained XGBoost-LSTM hybrid model.

[0050] Use the trained XGBoost-LSTM hybrid model to predict the failure probability of the equipment in the next 2 hours, generate a heat map of the failure probability of the equipment in the next 2 hours, and trigger an early warning when the probability threshold exceeds 0.7.

[0051] S2, establish a spatial topology database of equipment based on the geographic information system. When an equipment abnormality is detected, the fault point is located through the Dijkstra algorithm and a three-dimensional geographic coordinate label is generated;

[0052] The establishment of a device spatial topology database based on a geographic information system is specifically as follows:

[0053] Collect the models, specifications and performance parameters of transformers, circuit breakers, towers and cables in the distribution network;

[0054] Get the device's geographic location information, including latitude and longitude coordinates and elevation information;

[0055] Devices are abstracted as nodes, and the connection relationship between devices is abstracted as edges, to build a node-edge topology structure;

[0056] Establish database structure, including definition of data tables, setting of fields and establishment of data relationships, storage of device information, geographic information and topological relationships;

[0057] Import the preprocessed data into the database of the geographic information system to establish the equipment spatial topology database.

[0058] The method of locating the fault point by using the Dijkstra algorithm and generating a three-dimensional geographic coordinate label is specifically as follows:

[0059] Based on the device space topology database, each device node is regarded as a vertex in the graph, and the connections between devices are regarded as edges;

[0060] Taking the monitoring center or known power node as the starting point, the Dijkstra algorithm is used to calculate the shortest path to each node. When traversing to the abnormal device node, the path information of the node is recorded;

[0061] Analyze and calculate the shortest path, combine the distribution of abnormal equipment and the law of fault propagation, and obtain the range of the fault point;

[0062] Extract the latitude and longitude coordinates and elevation information of the fault point from the database of the geographic information system to generate three-dimensional geographic coordinates;

[0063] The three-dimensional geographic coordinates of the fault point, the fault type and the occurrence time are combined into a label and visualized on the GIS map.

[0064] The Dijkstra algorithm used in the present invention is still based on a greedy strategy, starting from the starting node and gradually expanding the search path outward. It will select the unvisited node closest to the starting node (with the smallest weight) as the next expansion node at each step. This closest distance is measured by the weight of the edge. In the traditional Dijkstra algorithm, a single factor such as path length is usually used as the weight. The Dijkstra algorithm used in the present invention redefines the weight according to the specific scenario.

[0065] The algorithm records an estimate of the shortest distance from the starting node to each node, and continuously updates this estimate as the search progresses until the true shortest path is found. For example, when building a distribution network graph model, it is initially assumed that the distance from all non-starting nodes to the starting node is infinite, and these distance estimates are continuously updated as the algorithm traverses the graph.

[0066] Improvements to traditional algorithms:

[0067] Consider multi-dimensional weight factors: The traditional Dijkstra algorithm may only consider the path length as the weight, while the Dijkstra algorithm used in the present invention will comprehensively consider more factors to determine the weight of the edge in the distribution network fault point location scenario. For example, in addition to the line length, the elevation factor in three-dimensional space will also be considered. If there is terrain undulation, the path with large elevation changes may make fault propagation more difficult, and its weight needs to be adjusted accordingly; the fault propagation characteristics will also be considered. For example, if the fault propagation speed of some lines is fast, then the possibility of reaching the fault point through these lines is relatively low, and its weight will also be different.

[0068] Heuristic information introduction: The algorithm used in the present invention may introduce heuristic information to guide the search direction and improve the search efficiency. For example, based on the topological structure of the distribution network and historical fault data, it is known in advance that certain areas are more prone to failures. In this case, when searching, these areas will be explored first, rather than blindly performing breadth-first search in the manner of the traditional Dijkstra algorithm.

[0069] Handling special situations and constraints: There may be some special situations and constraints in the distribution network, such as some lines are unavailable at certain times or under certain conditions, or different types of faults may propagate differently on different types of lines. The Dijkstra algorithm used in the present invention will adjust the algorithm for these situations, dynamically consider these constraints when searching for paths, and avoid selecting unavailable lines or paths that do not conform to the fault propagation law.

[0070] S3, build a multi-dimensional emergency repair resource portrait library to dynamically record the skill level of the emergency repair team, the inventory status of materials and the real-time location information of the vehicle;

[0071] The multi-dimensional emergency repair resource portrait library is constructed to dynamically record the skill level of the emergency repair team, the inventory status of materials and the real-time location information of the vehicle, specifically:

[0072] Obtain the main dimensions for building the portrait library, including the repair team's skill level, material inventory status, and vehicle real-time location information;

[0073] At the same time, the specific information fields under each dimension are planned, including the skill type and personnel qualification certificates of the repair team, the name, specification, quantity, storage location of the materials, the license plate number, model, and real-time coordinates of the vehicle;

[0074] By integrating the vehicle-mounted GPS with traffic big data, the current location, driving speed, estimated arrival time and availability of the repair vehicle can be continuously tracked;

[0075] Build a multi-dimensional emergency repair resource portrait library based on main dimensions and data fields.

[0076] Portrait generation algorithm: For the skill level portrait of the repair team, a weighted calculation is performed based on the skill assessment results of the personnel to obtain the overall skill level of the team; for the material inventory status portrait, classification and labeling are performed based on inventory quantity and in-and-out frequency indicators.

[0077] S4, establish a dynamic priority evaluation model to generate a priority sequence of emergency repair tasks based on the urgency of the fault, the level of affected users, and the criticality of the equipment;

[0078] The dynamic priority evaluation model is established to generate a priority sequence of emergency repair tasks according to the urgency of the fault, the level of affected users and the criticality of the equipment, specifically:

[0079] The emergency levels are divided according to the fault type, power outage scope and power outage time factors. Faults that cause large-scale power outages or affect important users are set to high emergency levels; small-scale power outages or faults that can be restored in a short time are set to low emergency levels.

[0080] Users are divided into important users and ordinary users. The impact of failures on important users is high, while the impact of failures on ordinary users is low.

[0081] The criticality of equipment is divided into ordinary equipment for branch lines, general transmission line equipment, and critical equipment for core substations;

[0082] Based on the principal component analysis method, weights are assigned to the fault urgency, user impact level, and equipment criticality;

[0083] The evaluation model is constructed using a linear weighted approach, and each indicator value is multiplied by the corresponding weight and then added together to obtain the priority score of each repair task;

[0084] According to the priority scores calculated by the evaluation model, all emergency repair tasks are sorted, and tasks with higher scores have higher priorities, thus generating an emergency repair task priority sequence.

[0085] The priority score is specifically:

[0086] ,

[0087] in, is the fault urgency weight, To influence the user level weight, is the device criticality weight, is the time factor weight, Score the urgency of the fault, To influence the user's rating score, Score the criticality of the device. is the time when the fault occurred, is the maximum fault duration.

[0088] S5, a multi-objective grey wolf optimization algorithm is used to solve the emergency repair resource scheduling scheme, and the objective function is set in the direction of minimizing the response time, path cost and resource waste rate;

[0089] The multi-objective grey wolf optimization algorithm is used to solve the emergency repair resource scheduling scheme, and the objective function minimizes the response time, path cost and resource waste rate at the same time, specifically:

[0090] Randomly generate a set of candidate solutions as the initial wolf pack;

[0091] Calculate the fitness value of each candidate solution according to the objective function;

[0092] Select the optimal solution as the leader wolf (i.e., α wolf) according to the fitness value, and update the positions of other level wolves (β wolf, δ wolf, and ω wolf);

[0093] By simulating the social hierarchy and hunting behavior of gray wolf groups, candidate solutions are continuously updated iteratively until the stopping condition is met;

[0094] A set of Pareto optimal solutions is selected from the final wolf pack as the emergency repair resource scheduling solution.

[0095] The objective function is specifically:

[0096] ,

[0097] ,

[0098] ,

[0099] ,

[0100] in, is the response time objective function, is the number of fault points, is the number of repair resource points, is the time from the repair resource point j to the fault point i, is a decision variable, which is used to indicate whether to allocate the emergency repair resource point j to the fault point i. is the path cost objective function, is the path cost from the repair resource point j to the fault point i, is the resource waste rate objective function, is the number of the kth resource owned by the repair resource point j, is the number of resource types, is the demand quantity of the kth resource at the fault point i, is the objective function.

[0101] The multi-objective gray wolf optimization algorithm is an optimization algorithm based on swarm intelligence. It is developed on the basis of the gray wolf optimization algorithm and is used to solve multi-objective optimization problems.

[0102] The multi-objective gray wolf optimization algorithm mainly simulates the social hierarchy structure and predation behavior of gray wolf groups. In the wolf pack, there are wolves of different levels, such as α, β, δ and ω. α wolf is at the highest level and is responsible for leading the wolf pack to make decisions; β wolf assists α wolf in decision-making and management; δ wolf obeys the command of α and β wolves and manages ω wolf at the same time; ω wolf is at the lowest level and needs to obey the orders of other wolves.

[0103] In the algorithm, the solution space of the problem is regarded as the search space of the prey. The wolf pack approaches the prey by constantly searching and updating its position. The position of each wolf represents a potential solution. The wolf pack searches for the optimal solution through mutual cooperation and information sharing. During the search process, the wolf pack constantly adjusts its search direction and step length according to its current position and the position of the prey to gradually approach the prey.

[0104] S6, issues the optimal scheduling instructions through the collaborative scheduling platform and continuously monitors the task execution status.

[0105] The optimal scheduling instructions are issued through the collaborative scheduling platform, and the task execution status is continuously monitored, specifically:

[0106] According to the distributed ledger characteristics of blockchain, power companies, emergency repair teams, and material suppliers are organized as nodes to build a consortium chain;

[0107] Write a smart contract. When the optimal dispatching plan is obtained, the smart contract is automatically triggered and the dispatching instruction is sent to the corresponding repair team node. After the repair team completes the task and confirms it, the smart contract automatically records the result.

[0108] The dispatch center obtains the optimal dispatch plan based on the multi-objective grey wolf optimization algorithm, and broadcasts the dispatch instructions containing the repair task allocation and resource allocation to the blockchain network after signing with the encryption algorithm and storing them in the block;

[0109] Each repair team node receives dispatch instructions from the blockchain, uses the public key to verify the integrity and authenticity of the instructions, and prepares to carry out repair work according to the contents of the instructions;

[0110] During the execution of the task, the repair team collects task progress and resource usage data in real time, encrypts it and uploads it to the blockchain. The task progress data includes the time of arrival at the fault point, the time of starting repair, and the time of completion of repair.

[0111] The dispatch center and related party nodes query the task execution status through the blockchain. At the same time, the smart contract sets early warning rules. When abnormal situations such as task progress delays and resource shortages occur, early warnings are automatically triggered to notify related parties for timely processing.

[0112] Embodiment 2, a system for a distribution network fault emergency repair dispatching method, comprising the following modules:

[0113] Data acquisition module: used to collect distribution network equipment operation data in real time, build equipment health evaluation index set, and output equipment failure probability distribution map through failure probability prediction model;

[0114] Coordinate label generation module: used to establish a spatial topology database of equipment based on the geographic information system. When an equipment abnormality is detected, the fault point is located through the Dijkstra algorithm and a three-dimensional geographic coordinate label is generated;

[0115] The multi-dimensional portrait construction module of emergency repair resources is used to build a multi-dimensional emergency repair resource portrait library, dynamically recording the skill level of the emergency repair team, the inventory status of materials, and the real-time location information of vehicles;

[0116] Dynamic priority assessment module for emergency repair tasks: used to establish a dynamic priority assessment model to generate an emergency repair task priority sequence based on the urgency of the fault, the level of affected users, and the criticality of the equipment;

[0117] Multi-objective Grey Wolf Optimization Algorithm Resource Scheduling Module: It is used to solve the emergency repair resource scheduling scheme by using the multi-objective Grey Wolf Optimization Algorithm, and set the objective function in the direction of minimizing the response time, path cost and resource waste rate;

[0118] Blockchain collaborative scheduling instruction issuing module: used to issue optimal scheduling instructions through the collaborative scheduling platform and continuously monitor the task execution status.

[0119] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0120] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0121] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0122] In addition, each functional module in each embodiment of the present application 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.

[0123] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0124] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A distribution network fault repair dispatching method, characterized in that: The following steps are involved: Collect distribution network equipment operation data in real time, build equipment health evaluation indicator set, and output equipment failure probability distribution map through failure probability prediction model; Establish a spatial topology database of equipment based on the geographic information system. When an equipment abnormality is detected, locate the fault point through the Dijkstra algorithm and generate a three-dimensional geographic coordinate label; Build a multi-dimensional emergency repair resource portrait library to dynamically record the skill level of the emergency repair team, material inventory status and real-time vehicle location information; A dynamic priority evaluation model is established. By combining the fault type, outage scope, outage time, user level and equipment criticality factors, the principal component analysis method is used to assign weights to each indicator. A linear weighted method is used to construct an evaluation model to calculate the priority score of the emergency repair task, thereby generating a priority sequence for the emergency repair task. The multi-objective grey wolf optimization algorithm is used to solve the emergency repair resource scheduling scheme, and the objective function is set in the direction of minimizing the response time, path cost and resource waste rate; Issue the optimal scheduling instructions through the collaborative scheduling platform and continuously monitor the task execution status; The multi-objective grey wolf optimization algorithm is used to solve the emergency repair resource scheduling scheme, and the objective function is set in the direction of minimizing the response time, path cost and resource waste rate, specifically: Randomly generate a set of candidate solutions as the initial wolf pack; Calculate the fitness value of each candidate solution according to the objective function; Select the optimal solution as the leader wolf according to the fitness value, and update the positions of wolves at other levels; By simulating the social hierarchy and hunting behavior of gray wolf groups, candidate solutions are continuously updated iteratively until the stopping condition is met; Select a set of Pareto optimal solutions from the final wolf pack as the emergency repair resource scheduling solution; The priority score is specifically: ， in, is the fault urgency weight, To influence the user level weight, is the device criticality weight, is the time factor weight, Score the urgency of the fault, To influence the user's rating score, Score the criticality of the device. is the time when the fault occurred, is the maximum fault duration; The objective function is specifically: ， ， ， ， in, is the response time objective function, is the number of fault points, is the number of repair resource points, is the time from the repair resource point j to the fault point i, is a decision variable, which is used to indicate whether to allocate the emergency repair resource point j to the fault point i. is the path cost objective function, is the path cost from the repair resource point j to the fault point i, is the resource waste rate objective function, is the number of the kth resource owned by the repair resource point j, is the number of resource types, is the demand quantity of the kth resource at the fault point i, is the objective function.

2. A distribution network fault repair dispatching method according to claim 1, characterized in that: The real-time collection of distribution network equipment operation data, the construction of equipment health evaluation index set, and the output of equipment failure probability distribution map through the failure probability prediction model are specifically as follows: The data acquisition terminal receives the operating data from the sensor and performs filtering, amplification, analog-to-digital conversion preprocessing to remove noise interference. According to the physical structure and working principle of the distribution network equipment, the parameters closely related to the health status of the equipment are obtained as candidate indicators based on regression analysis; The candidate indicators are divided into electrical performance indicators, mechanical performance indicators, and chemical performance indicators, and a multi-level indicator system is constructed. The overall health of the equipment is taken as the top-level indicator, and the top-level indicator is divided into several first-level indicators, and each first-level indicator is further decomposed into several second-level indicators; Based on the multi-level indicator system, the XGBoost-LSTM hybrid model is used to obtain the fault probability prediction model; The geographical location information of the distribution network equipment is associated with the equipment failure probability prediction results to establish a mapping relationship between equipment location and failure probability; Geographic information system software is selected as the mapping platform, and the equipment is marked on the map according to its actual location. Different colors are used for visualization according to the probability of failure. Equipment with a high probability of failure is marked in red, and equipment with a low probability of failure is marked in green.

3. A distribution network fault repair dispatching method according to claim 2, characterized in that: The establishment of a device spatial topology database based on a geographic information system is specifically as follows: Collect the models, specifications and performance parameters of transformers, circuit breakers, towers and cables in the distribution network; Get the device's geographic location information, including latitude and longitude coordinates and elevation information; Devices are abstracted as nodes, and the connection relationship between devices is abstracted as edges, to build a node-edge topology structure; Establish database structure, including definition of data tables, setting of fields and establishment of data relationships, storage of device information, geographic information and topological relationships; Import the preprocessed data into the database of the geographic information system to establish the equipment spatial topology database.

4. A distribution network fault repair dispatching method according to claim 3, characterized in that: The method of locating the fault point by using the Dijkstra algorithm and generating a three-dimensional geographic coordinate label is specifically as follows: Based on the device space topology database, each device node is regarded as a vertex in the graph, and the connections between devices are regarded as edges; Taking the monitoring center or known power node as the starting point, the Dijkstra algorithm is used to calculate the shortest path to each node. When traversing to the abnormal device node, the path information of the node is recorded; Analyze and calculate the shortest path, combine the distribution of abnormal equipment and the law of fault propagation, and obtain the range of the fault point; Extract the latitude and longitude coordinates and elevation information of the fault point from the database of the geographic information system to generate three-dimensional geographic coordinates; The three-dimensional geographic coordinates of the fault point, the fault type and the occurrence time are combined into a label and visualized on the GIS map.

5. A distribution network fault repair and dispatching method according to claim 4, characterized in that: The multi-dimensional emergency repair resource portrait library is constructed to dynamically record the skill level of the emergency repair team, the inventory status of materials and the real-time location information of the vehicle, specifically: Obtain the main dimensions for building the portrait library, including the repair team's skill level, material inventory status, and vehicle real-time location information; At the same time, the specific information fields under each dimension are planned, including the skill type and personnel qualification certificates of the repair team, the name, specification, quantity, storage location of the materials, the license plate number, model, and real-time coordinates of the vehicle; By integrating the vehicle's global positioning system with traffic big data, the current location, driving speed, estimated arrival time and availability of the repair vehicle can be continuously tracked; Build a multi-dimensional emergency repair resource portrait library based on main dimensions and data fields.

6. A distribution network fault repair dispatching method according to claim 5, characterized in that: The dynamic priority evaluation model is established by combining the fault type, power outage scope, power outage time, user level and equipment criticality factors, using the principal component analysis method to assign weights to each indicator, and using a linear weighted method to construct an evaluation model to calculate the priority score of the emergency repair task, thereby generating a priority sequence of the emergency repair task, specifically: The emergency levels are divided according to the fault type, power outage scope and power outage time factors. Faults that cause large-scale power outages or affect important users are set to high emergency levels; small-scale power outages or faults that can be restored in a short time are set to low emergency levels. Users are divided into important users and ordinary users. The impact of failures on important users is high, while the impact of failures on ordinary users is low. The criticality of equipment is divided into ordinary equipment for branch lines, general transmission line equipment, and critical equipment for core substations; Based on the principal component analysis method, weights are assigned to the fault urgency, user impact level, and equipment criticality; The evaluation model is constructed using a linear weighted approach, and each indicator value is multiplied by the corresponding weight and then added together to obtain the priority score of each repair task; According to the priority scores calculated by the evaluation model, all emergency repair tasks are sorted, and tasks with higher scores have higher priorities, thus generating an emergency repair task priority sequence.

7. A system using a distribution network fault emergency repair dispatching method as claimed in any one of claims 1 to 6, characterized in that: Includes the following modules: Data acquisition module: used to collect distribution network equipment operation data in real time, build equipment health evaluation index set, and output equipment failure probability distribution map through failure probability prediction model; Coordinate label generation module: used to establish a spatial topology database of equipment based on the geographic information system. When an equipment abnormality is detected, the fault point is located through the Dijkstra algorithm and a three-dimensional geographic coordinate label is generated; The multi-dimensional portrait construction module of emergency repair resources is used to build a multi-dimensional emergency repair resource portrait library, dynamically recording the skill level of the emergency repair team, the inventory status of materials, and the real-time location information of vehicles; Dynamic priority assessment module for emergency repair tasks: used to establish a dynamic priority assessment model to generate an emergency repair task priority sequence based on the urgency of the fault, the level of affected users, and the criticality of the equipment; Multi-objective Grey Wolf Optimization Algorithm Resource Scheduling Module: It is used to solve the emergency repair resource scheduling scheme by using the multi-objective Grey Wolf Optimization Algorithm, and set the objective function in the direction of minimizing the response time, path cost and resource waste rate; Blockchain collaborative scheduling instruction issuing module: used to issue optimal scheduling instructions through the collaborative scheduling platform and continuously monitor the task execution status.

Images