Power distribution network fault self-healing strategy optimization method and system

Abstract

The invention discloses a power distribution network fault self-healing strategy optimization method and system, and the method comprises the steps: collecting the data of a power distribution network, recognizing a fault, and obtaining a self-healing strategy set. And classifying the automatic switches according to the average response speed of the automatic switches to obtain an automatic switch set and corresponding response speed labels. And analyzing the self-healing strategy set according to the response speed label to obtain an evaluation index of the self-healing strategy, and outputting a target self-healing strategy. And optimizing the target self-healing strategy, and outputting an optimized self-healing strategy. According to the power distribution network fault self-healing strategy optimization method and system provided by the invention, effective evaluation of different self-healing strategies is realized, and the optimal performance of the selected strategy in the aspects of recovery time, system stability and response efficiency is ensured. And optimizing a target self-healing strategy by adopting a multi-element heuristic method fusing a genetic algorithm and a simulated annealing algorithm. The global optimization of the self-healing strategy is realized, and the efficient and stable fault processing capability in a complex and changeable power distribution network environment is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network fault self-healing, and in particular to a distribution network fault self-healing strategy optimization method and system. Background Art

[0002] With the continuous development of modern power systems, the safety and reliability of distribution networks, as an important part of the power system, have attracted much attention. Traditional distribution network fault handling mainly relies on manual intervention and preset protection devices. This method not only has a slow response speed, but also has difficulty in coping with complex and changeable power fault scenarios. In recent years, with the rapid development of smart grids and information technology, fault self-healing technology based on big data and artificial intelligence has become a research hotspot. By real-time monitoring of the operating status of the distribution network, using advanced data processing and analysis technology, automatically identifying faults and taking corresponding self-healing measures, the reliability and stability of the distribution network have been significantly improved.

[0003] However, the existing distribution network fault self-healing technology still has some shortcomings in practical applications. First, the accuracy and timeliness of fault identification are insufficient, which often leads to misjudgment and missed judgment. Secondly, the existing self-healing strategies are mostly statically preset, lack dynamic optimization capabilities, and are difficult to adapt to the complex and changeable power system operating environment. Finally, when dealing with multiple fault points occurring simultaneously in large-scale distribution networks, the existing technology often exhibits problems such as untimely response and insufficient stability. In response to these problems, our invention proposes a new distribution network fault self-healing strategy optimization method, which significantly improves the self-healing capability of the distribution network by comprehensively utilizing multiple technologies such as data collection, fault identification, strategy analysis and optimization. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is that the existing technology often exhibits problems of untimely response and insufficient stability when dealing with multiple fault points occurring simultaneously in a large-scale distribution network.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a method for optimizing a distribution network fault self-healing strategy, comprising: collecting distribution network data, identifying faults, and obtaining a set of self-healing strategies.

[0007] According to the average response speed of the automatic switches, the automatic switches are classified to obtain an automatic switch set and a corresponding response speed label.

[0008] The self-healing strategy set is analyzed according to the response speed label, the evaluation index of the self-healing strategy is obtained, and the target self-healing strategy is output.

[0009] Optimize the target self-healing strategy and output the optimized self-healing strategy.

[0010] As a preferred solution of the method for optimizing the distribution network fault self-healing strategy described in the present invention, the collected distribution network data includes distribution network topology data, electrical parameters, and fault records.

[0011] Distribution network topology data includes lines, transformers, switch locations and connections.

[0012] Electrical parameters include voltage, current, and power.

[0013] The fault record includes the fault type and fault handling method.

[0014] The data is normalized and expressed as:

[0015]

[0016] Among them, X represents the original data, X min Indicates the minimum value of the data, X max represents the maximum value of the data, and X' represents the normalized data.

[0017] As a preferred solution of the distribution network fault self-healing strategy optimization method described in the present invention, wherein: the identifying fault and obtaining the self-healing strategy set includes using an LSTM network for fault identification, including an input layer, an LSTM layer, a fully connected layer and an output layer.

[0018] The normalized data is converted into an input form that can be processed by the LSTM network. The input time window is T, and the input of each data sample is expressed as:

[0019] X=[x t ,x t+1 ,…,x t+T-1 ]

[0020] Among them, x t Represents the feature vector of the data at time t.

[0021] The output layer is represented as:

[0022]

[0023] in, represents the failure probability, W y represents the weight of the fully connected layer, h T represents the hidden state, b y represents the bias of the fully connected layer.

[0024] When a fault point is identified, the network structure of the power transmission of the distribution network is reconstructed to obtain a reconstructed topology. Treatment measures are taken for the fault point to obtain a set of self-healing strategies, including the reconstructed topology of each fault point and the treatment measures for the fault point.

[0025] As a preferred solution of the distribution network fault self-healing strategy optimization method described in the present invention, wherein: the automatic switches are classified according to their average response speeds to obtain automatic switch sets and corresponding response speed labels, including collecting historical response speed data of the automatic switches, including response times and timestamps.

[0026] The response time is the time interval from when the automatic switch detects a fault to when it takes action.

[0027] The timestamp is the time point when each response is recorded.

[0028] In order to calculate the average response speed of the automatic switch, the historical response time data of each automatic switch is statistically analyzed. The historical response speed data of the automatic switch is represented by {t1, t2, …, t n}, where t i Represents the time of the i-th response.

[0029] Calculate the average response speed of the automatic switch, expressed as:

[0030]

[0031] in, It represents the average response time of the automatic switch, and n represents the total number of historical response speed data.

[0032] The automatic switches are classified according to the average response speed to obtain the automatic switch set and the corresponding response speed label. , the response speed label is high speed response.

[0033] When the average response time , the response speed label is medium speed response.

[0034] When the average response time , the response speed label is low speed response.

[0035] As a preferred solution of the distribution network fault self-healing strategy optimization method described in the present invention, wherein: the self-healing strategy set is analyzed according to the response speed label to obtain the evaluation index of the self-healing strategy, including extracting the reconstructed topology structure of each fault point and the treatment measures for the fault point from the self-healing strategy set, and for each self-healing strategy, the fault point location, reconstructed topology structure, treatment measures and response speed label are recorded.

[0036] Evaluation indicators are set, including recovery time, system stability integral and response efficiency, and the evaluation indicators are calculated.

[0037] The recovery time measures the average recovery time of the self-healing strategy under different response speed labels, expressed as:

[0038]

[0039] Where N represents the number of self-healing strategies, T i represents the recovery time of the i-th strategy, and λ represents the decay factor.

[0040] The system stability of the self-healing strategy under different response speed labels is measured as:

[0041]

[0042] Where T is the total time, M is the number of switches, and V j (t) represents the voltage of the jth switch at time t, R j represents the response speed label of the jth switch, and β represents the stability coefficient.

[0043] The response efficiency measures the response efficiency of the self-healing strategy under different response speed labels, which is expressed as:

[0044]

[0045] Where K represents the number of fault points, γ k represents the response coefficient of the kth fault point, represents the average response time of the kth fault point, and τ represents the adjustment parameter.

[0046] As a preferred solution of the method for optimizing the self-healing strategy of distribution network faults described in the present invention, the output target self-healing strategy includes taking the self-healing strategy with the best evaluation index as the target self-healing strategy, satisfying the conditions of shortest recovery time, highest system stability and highest response efficiency, and the objective function is expressed as:

[0047]

[0048] The constraints include response time distribution variance constraint, power loss constraint, system stability constraint and response coverage constraint.

[0049] The time distribution variance constraint is expressed as:

[0050]

[0051] Where ∈1 represents the volatility constant that limits the response time.

[0052] The power loss constraint is expressed as:

[0053]

[0054] in, represents the power loss of the jth switch at time t, and ∈2 represents the upper limit of the power loss.

[0055] The system stability constraint is expressed as:

[0056]

[0057] in, represents the rate of change of the jth switch voltage, ∈3 represents the maximum allowed rate of change, and κ represents the stability attenuation coefficient.

[0058] The response coverage constraint is expressed as:

[0059]

[0060] Among them, R k represents the response coverage of the kth fault point, ρ represents the threshold of the response coverage, η represents the adjustment parameter of the response coverage, and ∈4 represents the minimum response coverage.

[0061] As a preferred solution of the distribution network fault self-healing strategy optimization method described in the present invention, wherein: the optimization of the target self-healing strategy and output of the optimized self-healing strategy include using a genetic algorithm to optimize the target self-healing strategy, incorporating a simulated annealing algorithm into the selection and mutation stages of the genetic algorithm, and enhancing the diversity and global search capability of the offspring through the acceptance criteria of the simulated annealing.

[0062] First, the initial population is randomly generated, each individual represents a self-healing strategy, and P initial individuals are randomly generated. Each individual is represented by a vector x i Indicates that i=1,2,…,P.

[0063] Calculate the fitness of each individual f(x i ), expressed as:

[0064]

[0065] Using the roulette wheel selection method, the selection probability is expressed as:

[0066]

[0067] Combined with the acceptance criterion of simulated annealing, it is expressed as:

[0068] P accept =e -ΔE / T

[0069] Among them, ΔE represents the fitness difference between the new individual and the current individual, and T represents the current temperature.

[0070] Single-point crossover is used to generate new individuals, and random mutation is used to decide whether to mutate based on the acceptance criteria of simulated annealing, which is expressed as:

[0071]

[0072] Where, ΔE mut Represents the fitness difference before and after mutation.

[0073] Update the population according to the fitness value and retain the best individual. Gradually reduce the simulated annealing temperature, converge when the maximum number of iterations is reached, and output the optimized self-healing strategy.

[0074] A distribution network fault self-healing strategy optimization system, characterized by: comprising:

[0075] The data collection module collects distribution network data, identifies faults, and obtains a set of self-healing strategies.

[0076] The classification module classifies the automatic switches according to their average response speeds, and obtains an automatic switch set and corresponding response speed labels.

[0077] The output target module analyzes the self-healing strategy set according to the response speed label, obtains the evaluation index of the self-healing strategy, and outputs the target self-healing strategy.

[0078] The optimization module optimizes the target self-healing strategy and outputs the optimized self-healing strategy.

[0079] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0080] A computer-readable storage medium stores a computer program, which implements the steps of the method described above when executed by a processor.

[0081] The beneficial effects of the present invention are as follows: the self-healing strategy set is analyzed according to the response speed label, the evaluation index of the self-healing strategy is obtained, and the target self-healing strategy is output, thereby realizing the effective evaluation of different self-healing strategies and ensuring that the selected strategy performs best in terms of recovery time, system stability and response efficiency.

[0082] The multivariate heuristic method integrating genetic algorithm and simulated annealing algorithm is used to further optimize the target self-healing strategy. The global optimization of the self-healing strategy is achieved to ensure efficient and stable fault handling capabilities in the complex and changeable distribution network environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0084] Figure 1 An overall flow chart of a method for optimizing a distribution network fault self-healing strategy provided by the first embodiment of the present invention. DETAILED DESCRIPTION

[0085] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0086] Example 1, reference Figure 1 , is an embodiment of the present invention, and provides a method for optimizing a distribution network fault self-healing strategy, comprising:

[0087] S1: Collect distribution network data, identify faults, and obtain a set of self-healing strategies.

[0088] Distribution network data includes distribution network topology data, electrical parameters, and fault records.

[0089] Distribution network topology data includes lines, transformers, switch locations and connections.

[0090] Electrical parameters include voltage, current, and power.

[0091] The fault record includes the fault type and fault handling method.

[0092] The data is normalized and expressed as:

[0093]

[0094] Among them, X represents the original data, X min Indicates the minimum value of the data, X max represents the maximum value of the data, and X' represents the normalized data.

[0095] The LSTM network is used for fault identification, including input layer, LSTM layer, fully connected layer and output layer.

[0096] The normalized data is converted into an input form that can be processed by the LSTM network. The input time window is T, and the input of each data sample is expressed as:

[0097] X=[x t ,x t+1 ,…,x t+T-1 ]

[0098] Among them, x t Represents the feature vector of the data at time t.

[0099] The output layer is represented as:

[0100]

[0101] in, represents the failure probability, W y represents the weight of the fully connected layer, h T represents the hidden state, b y represents the bias of the fully connected layer.

[0102] When a fault point is identified, the network structure of the power transmission of the distribution network is reconstructed to obtain a reconstructed topology. Treatment measures are taken for the fault point to obtain a set of self-healing strategies, including the reconstructed topology of each fault point and the treatment measures for the fault point.

[0103] It should be noted that the LSTM unit is represented as:

[0104] f t =σ ( W f h t-1 +U f x t +b f )

[0105] i t =σ(W i h t-1 +U i x t +b i )

[0106] c=f t ⊙c t-1 +i t ⊙tanh(W c h t-1 +U c x t +b c )

[0107] 0 t =σ(W o h t-1 +Uo x t +b o )

[0108] h t =o t ⊙tanh(c t )

[0109] Among them, σ(`) represents the sigmoid activation function, ⊙ represents element multiplication, and h t represents the hidden state, c t Indicates the memory status.

[0110] S2: Classify the automatic switches according to their average response speeds, and obtain an automatic switch set and corresponding response speed labels.

[0111] Collect historical response speed data of automatic switches, including response time and timestamp.

[0112] The response time is the time interval from when the automatic switch detects a fault to when it takes action.

[0113] The timestamp is the time point when each response is recorded.

[0114] In order to calculate the average response speed of the automatic switch, the historical response time data of each automatic switch is statistically analyzed. The historical response speed data of the automatic switch is represented by {t1, t2, …, t n}, where t i Represents the time of the i-th response.

[0115] Calculate the average response speed of the automatic switch, expressed as:

[0116]

[0117] in, It represents the average response time of the automatic switch, and n represents the total number of historical response speed data.

[0118] The automatic switches are classified according to the average response speed to obtain the automatic switch set and the corresponding response speed label. , the response speed label is high speed response.

[0119] When the average response time , the response speed label is medium speed response.

[0120] When the average response time , the response speed label is low speed response.

[0121] It should be noted that by analyzing historical data, the automatic switches are accurately grouped, so that the subsequent self-healing strategy can be optimized according to the characteristics of different types of switches, improving the execution efficiency and accuracy of the strategy. Developing targeted self-healing strategies based on different response speed labels can give priority to calling high-speed response switches when faults occur, shorten the recovery time, and improve the overall fault handling efficiency. Understanding the response characteristics of each automatic switch helps to reasonably allocate resources. When resources are limited, faster response switches are given priority to ensure the rapid recovery of key nodes.

[0122] S3: Analyze the self-healing strategy set according to the response speed label, obtain the evaluation index of the self-healing strategy, and output the target self-healing strategy.

[0123] The reconstructed topology structure of each fault point and the treatment measures for the fault point are extracted from the self-healing strategy set. For each self-healing strategy, the fault point location, reconstructed topology structure, treatment measures and response speed label are recorded.

[0124] Evaluation indicators are set, including recovery time, system stability integral and response efficiency, and the evaluation indicators are calculated.

[0125] It should be noted that recovery time is a key indicator for evaluating the effectiveness of self-healing strategies. Short recovery time means that the system can quickly return to normal, reducing power outage time and economic losses.

[0126] The recovery time measures the average recovery time of the self-healing strategy under different response speed labels, expressed as:

[0127]

[0128] Where N represents the number of self-healing strategies, T i represents the recovery time of the i-th strategy, and λ represents the decay factor.

[0129] Furthermore, by calculating the exponential decay of the recovery time, the weighted average recovery time is obtained, highlighting the importance of short-time strategies.

[0130] It should be noted that system stability is the core indicator to ensure the safe operation of the distribution network. Stable voltage changes help avoid secondary failures and equipment damage and improve the overall reliability of the system.

[0131] The system stability of the self-healing strategy under different response speed labels is measured as:

[0132]

[0133] Where T is the total time, M is the number of switches, and V j (t) represents the voltage of the jth switch at time t, Rj represents the response speed label of the jth switch, and β represents the stability coefficient.

[0134] Furthermore, the comprehensive integral calculation of the voltage change rate and response speed labels is used to highlight the stability of the system in different time periods.

[0135] It should be noted that high response efficiency indicates that the self-healing strategy can quickly respond to and handle fault points, improve the timeliness and effectiveness of fault handling, and ensure rapid recovery of the system.

[0136] The response efficiency measures the response efficiency of the self-healing strategy under different response speed labels, which is expressed as:

[0137]

[0138] Where K represents the number of fault points, γ k represents the response coefficient of the kth fault point, represents the average response time of the kth fault point, and τ represents the adjustment parameter.

[0139] It should be noted that, through the function calculation of the fault point response coefficient and the average response time, the importance of the fault points with short response time is highlighted.

[0140] The self-healing strategy with the best evaluation index is taken as the target self-healing strategy, which meets the conditions of shortest recovery time, highest system stability and highest response efficiency. The objective function is expressed as:

[0141]

[0142] The constraints include response time distribution variance constraint, power loss constraint, system stability constraint and response coverage constraint.

[0143] The time distribution variance constraint is expressed as:

[0144]

[0145] Where ∈1 represents the volatility constant that limits the response time.

[0146] It should be noted that by calculating the variance of the response time distribution, the volatility of the response time is limited, ensuring that each response time is as consistent as possible. This improves the stability of the response time, avoids some response times being too long, which leads to a longer overall recovery time, and ensures that the system can handle faults stably and efficiently.

[0147] The power loss constraint is expressed as:

[0148]

[0149] in, represents the power loss of the jth switch at time t, and ∈2 represents the upper limit of the power loss.

[0150] The system stability constraint is expressed as:

[0151]

[0152] in, represents the rate of change of the jth switch voltage, ∈3 represents the maximum allowed rate of change, and κ represents the stability attenuation coefficient.

[0153] The response coverage constraint is expressed as:

[0154]

[0155] Among them, R k represents the response coverage of the kth fault point, ρ represents the threshold of the response coverage, η represents the adjustment parameter of the response coverage, and ∈4 represents the minimum response coverage.

[0156] It should be noted that by calculating the response coverage, we ensure that each fault point has sufficient response coverage to avoid missed reports and false reports. This improves the comprehensiveness and effectiveness of fault handling, ensures that all fault points can be handled in a timely manner, and improves the overall reliability of the system.

[0157] S4: Optimize the target self-healing strategy and output the optimized self-healing strategy.

[0158] Genetic algorithm is used to optimize the target self-healing strategy. Simulated annealing algorithm is integrated into the selection and mutation stages of genetic algorithm. The acceptance criterion of simulated annealing is used to enhance the diversity and global search ability of offspring.

[0159] First, the initial population is randomly generated, each individual represents a self-healing strategy, and P initial individuals are randomly generated. Each individual is represented by a vector x i Indicates that i=1,2,…,P.

[0160] Calculate the fitness of each individual f(x i ), expressed as:

[0161]

[0162] Using the roulette wheel selection method, the selection probability is expressed as:

[0163]

[0164] Combined with the acceptance criterion of simulated annealing, it is expressed as:

[0165] P accept =e -ΔE / T

[0166] Among them, ΔE represents the fitness difference between the new individual and the current individual, and T represents the current temperature.

[0167] Single-point crossover is used to generate new individuals, and random mutation is used to decide whether to mutate based on the acceptance criteria of simulated annealing, which is expressed as:

[0168]

[0169] Where, ΔE mut Represents the fitness difference before and after mutation.

[0170] Update the population according to the fitness value and retain the best individual. Gradually reduce the simulated annealing temperature, converge when the maximum number of iterations is reached, and output the optimized self-healing strategy.

[0171] It should be noted that in the selection stage, the optimization process of the distribution network fault self-healing strategy needs to evaluate the fitness of multiple self-healing strategies and determine the optimal strategy. Due to the randomness of the initial population, some high-quality strategies may be mistaken for low-quality strategies and eliminated. By introducing the acceptance criterion of simulated annealing, some strategies with lower fitness can be allowed to continue to participate in the iteration, increasing the diversity of the population and the global search capability, making it more likely to find the global optimal solution.

[0172] The distribution network system is complex and changeable. Relying solely on genetic algorithms is prone to falling into local optimality and failing to fully explore the entire strategy space. The simulated annealing algorithm can avoid premature convergence in the selection stage by controlling the mechanism of gradually lowering the temperature, improve the breadth and depth of the search, and better adapt to the complexity of the optimization of the distribution network self-healing strategy.

[0173] Furthermore, in the mutation phase, the diversity of self-healing strategies is the key to improving fault handling. Traditional mutation methods easily lead to repeated searches in local areas, but fail to fully explore new strategy spaces. The introduction of the acceptance criterion of simulated annealing can retain individuals with poor fitness in the mutation operation, increase the diversity of the population, further broaden the search scope, and ensure that more potential optimal self-healing strategies are explored.

[0174] Since the optimization of the self-healing strategy of the distribution network involves multiple evaluation indicators (such as recovery time, system stability and response efficiency), the mutation stage needs to generate offspring with different characteristics. The simulated annealing algorithm accepts the mechanism of poor individuals, so that the mutated offspring still have the opportunity to be retained even if the fitness is low, which increases the flexibility and diversity of strategy optimization and helps to find a better comprehensive self-healing strategy.

[0175] The above embodiments also include a distribution network fault self-healing strategy optimization system, specifically:

[0176] The classification module classifies the automatic switches according to their average response speeds, and obtains an automatic switch set and corresponding response speed labels.

[0177] The output target module analyzes the self-healing strategy set according to the response speed label, obtains the evaluation index of the self-healing strategy, and outputs the target self-healing strategy.

[0178] The optimization module optimizes the target self-healing strategy and outputs the optimized self-healing strategy.

[0179] The computer device may be a server. The computer device includes a processor, a memory, an input / output interface (I / O for short) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data cluster data of a power monitoring system. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method is implemented.

[0180] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0181] Embodiment 2, an embodiment of the present invention, provides a method and system for optimizing a distribution network fault self-healing strategy. In order to verify the beneficial effects of the present invention, a scientific demonstration is carried out through simulation experiments.

[0182] The existing technology adopts the traditional static self-healing strategy, while our invention uses a dynamic self-healing strategy based on response speed classification and multivariate heuristic optimization.

[0183] Existing technologies only collect voltage and current data of the distribution network and ignore environmental factors.

[0184] Our invention comprehensively collects distribution network topology data, electrical parameters, and fault records and performs data cleaning and normalization.

[0185] The existing technology uses a simple threshold determination method to identify faults, which is prone to misjudgment and missed judgment.

[0186] We invented the use of LSTM network for fault identification.

[0187] The prior art generates a static self-healing strategy according to preset rules.

[0188] After identifying the fault point, our invention reconstructs the distribution network topology and generates a set of self-healing strategies, including the reconstructed topology and processing measures.

[0189] The prior art does not take into account the difference in response speed of the automatic switch.

[0190] Our invention calculates the average response speed based on the historical response speed data of the automatic switch, and classifies it into high-speed response, medium-speed response and low-speed response.

[0191] Some technologies were directly applied without evaluating the self-healing strategy.

[0192] Our invention evaluates the self-healing strategy based on three indicators: recovery time, system stability score and response efficiency.

[0193] The existing technology does not perform strategy optimization.

[0194] We invented the optimization by combining genetic algorithm with simulated annealing algorithm, and introduced the acceptance criteria of simulated annealing algorithm in the selection and mutation stages.

[0195] The experimental results are shown in Table 1.

[0196] Table 1 Experimental results

[0197]

[0198] By comparing the prior art with our invention through the tabular data, we can clearly see that our invention has significant advantages in all indicators.

[0199] The average recovery time of the prior art is 47.4 minutes, while the average recovery time of our invention is 30.0 minutes. Our invention effectively shortens the recovery time and improves the efficiency of fault handling by using LSTM network for fault identification and optimizing the self-healing strategy according to the response speed label.

[0200] The average system stability score of the prior art is 145, while the average system stability score of our invention is 171.57. Our invention reconstructs the distribution network topology during the self-healing strategy generation process, significantly improving the stability of the system and avoiding the impact of voltage fluctuations on system stability.

[0201] The average response efficiency of the prior art is 0.73, while the average response efficiency of our invention is 0.86. Our invention classifies the response speed of automatic switches and gives priority to high-speed response switches for fault handling, thereby improving the response efficiency and ensuring that the fault point can be handled in a timely manner.

[0202] The average variance of the response time of the prior art is 14 seconds, while the average variance of the response time of our invention is 7.29 seconds. Our invention introduces the simulated annealing algorithm to retain individuals with lower fitness during the mutation phase, thereby increasing population diversity, ensuring the stability of the response time, and reducing the fluctuation of the response time.

[0203] The response coverage of the existing technology is 85% on average, while the response coverage of our invention is 96.57% on average. Our invention optimizes the self-healing strategy to ensure that all fault points can receive timely and effective responses, thereby improving the reliability of the system.

[0204] In summary, through data analysis, it can be seen that our invention is superior to the existing technology in all key indicators, showing significant innovation and advantages.

[0205] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A distribution network fault self-healing strategy optimization method, characterized in that: include: Collect distribution network data, identify faults, and obtain a set of self-healing strategies; According to the average response speed of the automatic switches, the automatic switches are classified to obtain the automatic switch set and the corresponding response speed labels; Analyze the self-healing strategy set according to the response speed label, obtain the evaluation index of the self-healing strategy, and output the target self-healing strategy; Optimize the target self-healing strategy and output the optimized self-healing strategy.

2. The method for optimizing the distribution network fault self-healing strategy according to claim 1, characterized in that: The collected distribution network data includes distribution network topology data, electrical parameters, and fault records; Distribution network topology data including lines, transformers, switch locations and connections; Electrical parameters include voltage, current, and power; Fault records include fault type and fault handling method; The data is normalized and expressed as: Among them, X represents the original data, X min Indicates the minimum value of the data, X max represents the maximum value of the data, and X' represents the normalized data.

3. The method for optimizing the distribution network fault self-healing strategy according to claim 2, characterized in that: The identifying fault and obtaining the self-healing strategy set includes using an LSTM network to identify the fault, including an input layer, an LSTM layer, a fully connected layer, and an output layer; The normalized data is converted into an input form that can be processed by the LSTM network. The input time window is T, and the input of each data sample is expressed as: X=[x t ,x t+1 ,…,x t+T-1 ] Among them, x t Represents the data feature vector at time t; The output layer is represented as: in, represents the failure probability, W y represents the weight of the fully connected layer, h T represents the hidden state, b y represents the bias of the fully connected layer; When a fault point is detected, the network structure of the power transmission in the distribution network is reconstructed to obtain a reconstructed topology. Processing measures are taken for the fault point to obtain a set of self-healing strategies, including the reconstructed topology of each fault point and the processing measures for the fault point.

4. The method for optimizing the distribution network fault self-healing strategy according to claim 3, characterized in that: The automatic switches are classified according to their average response speeds to obtain a set of automatic switches and corresponding response speed labels, including collecting historical response speed data of the automatic switches, including response time and timestamp; The response time is the time interval from the detection of a fault to the action taken by each automatic switch; The timestamp is the time point at which each response is recorded; In order to calculate the average response speed of the automatic switch, the historical response time data of each automatic switch is statistically analyzed. The historical response speed data of the automatic switch is represented by {t1, t2, …, t n }, where t i represents the time of the i-th response; Calculate the average response speed of the automatic switch, expressed as: in, represents the average response time of the automatic switch, and n represents the total number of historical response speed data; The automatic switches are classified according to the average response speed to obtain the automatic switch set and the corresponding response speed label. When , the response speed label is high speed response; When the average response time When , the response speed label is medium speed response; When the average response time , the response speed label is low speed response.

5. The method for optimizing the distribution network fault self-healing strategy according to claim 4, characterized in that: The self-healing strategy set is analyzed according to the response speed label to obtain the evaluation index of the self-healing strategy, including extracting the reconstructed topology structure of each fault point and the treatment measures for the fault point from the self-healing strategy set, and for each self-healing strategy, recording the fault point location, the reconstructed topology structure, the treatment measures and the response speed label; Setting evaluation indicators, including recovery time, system stability integral and response efficiency, and calculating the evaluation indicators; The recovery time measures the average recovery time of the self-healing strategy under different response speed labels, expressed as: Where N represents the number of self-healing strategies, T i represents the recovery time of the i-th strategy, and λ represents the decay factor; The system stability of the self-healing strategy under different response speed labels is measured as: Where T is the total time, M is the number of switches, and V j (t) represents the voltage of the jth switch at time t, R j represents the response speed label of the jth switch, and β represents the stability coefficient; The response efficiency measures the response efficiency of the self-healing strategy under different response speed labels, which is expressed as: Where K represents the number of fault points, γ k represents the response coefficient of the kth fault point, represents the average response time of the kth fault point, and τ represents the adjustment parameter.

6. The method for optimizing the distribution network fault self-healing strategy according to claim 5, characterized in that: The output target self-healing strategy includes taking the self-healing strategy with the best evaluation index as the target self-healing strategy, satisfying the conditions of shortest recovery time, highest system stability and highest response efficiency. The objective function is expressed as: The constraints include response time distribution variance constraint, power loss constraint, system stability constraint and response coverage constraint; The time distribution variance constraint is expressed as: Among them, ∈1 represents the volatility constant that limits the response time; The power loss constraint is expressed as: in, represents the power loss of the jth switch at time t, ∈2 represents the upper limit of the power loss; The system stability constraint is expressed as: in, represents the change rate of the jth switch voltage, ∈3 represents the maximum allowed change rate, and κ represents the stability attenuation coefficient; The response coverage constraint is expressed as: Among them, R k represents the response coverage of the kth fault point, ρ represents the threshold of the response coverage, η represents the adjustment parameter of the response coverage, and ∈4 represents the minimum response coverage.

7. The method for optimizing the distribution network fault self-healing strategy according to claim 6, characterized in that: The step of optimizing the target self-healing strategy and outputting the optimized self-healing strategy includes optimizing the target self-healing strategy using a genetic algorithm, integrating a simulated annealing algorithm into the selection and mutation phases of the genetic algorithm, and enhancing the diversity and global search capability of offspring through the acceptance criteria of the simulated annealing; First, the initial population is randomly generated, each individual represents a self-healing strategy, and P initial individuals are randomly generated. Each individual is represented by a vector x i Indicates, i = 1, 2, ..., P; Calculate the fitness of each individual f(x i ), expressed as: Using the roulette wheel selection method, the selection probability is expressed as: Combined with the acceptance criterion of simulated annealing, it is expressed as: P accept =and -ΔE / T Among them, ΔE represents the fitness difference between the new individual and the current individual, and T represents the current temperature; Single-point crossover is used to generate new individuals, and random mutation is used to decide whether to mutate based on the acceptance criteria of simulated annealing, which is expressed as: Where, ΔE mut Indicates the fitness difference before and after mutation; Update the population according to the fitness value and retain the best individuals; gradually reduce the simulated annealing temperature, converge when the maximum number of iterations is reached, and output the optimized self-healing strategy.

8. A distribution network fault self-healing strategy optimization system using the method according to any one of claims 1 to 7, characterized in that: The data collection module collects distribution network data, identifies faults, and obtains a set of self-healing strategies; The classification module classifies the automatic switches according to their average response speeds, and obtains a set of automatic switches and corresponding response speed labels; The output target module analyzes the self-healing strategy set according to the response speed label, obtains the evaluation index of the self-healing strategy, and outputs the target self-healing strategy; The optimization module optimizes the target self-healing strategy and outputs the optimized self-healing strategy.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images