Power distribution network pre-disaster investment scheme optimization system and method based on genetic algorithm

Abstract

The invention discloses a power distribution network pre-disaster investment scheme optimization system and method based on a genetic algorithm, and the method comprises the steps: building a two-stage optimization model containing pre-disaster investment and post-disaster recovery after obtaining basic data; the expected value of post-disaster load recovery can be maximized under the conditions of budget constraint, node capacity, load recovery, network connectivity and the like. A disaster scene is independently modeled according to representative intensity and spatial distribution, a pre-disaster investment scheme is coded through a genetic algorithm, the fitness is evaluated based on a multi-scene expected recovery load, global optimization is achieved through selection, intersection and variation, and a distributed power supply deployment, line reinforcement and controllable switch configuration scheme meeting budget is output. And rapid recovery is realized in cooperation with post-disaster fault isolation and network reconstruction. The load recovery level and the resource utilization efficiency of the power distribution network after disasters are remarkably improved under limited resources, economy and reliability are considered, and the method can be used for disaster-resistant planning and transformation of the power distribution network.

Description

Technical Field

[0001] This invention relates to the field of disaster mitigation planning for power system distribution networks, specifically to a system and method for optimizing pre-disaster investment schemes for distribution networks based on genetic algorithms. Background Technology

[0002] In recent years, the frequency and intensity of extreme weather events have been increasing, posing a severe challenge to the safe and stable operation of power distribution networks. In such disasters, critical infrastructure such as distribution lines, transformers, and switches may be damaged, leading to power outages and seriously affecting social life and economic operations. In the face of these risks, improving the resilience of power distribution networks to disaster impacts and shortening post-disaster recovery time is of great significance for ensuring the safety of the power system.

[0003] To enhance the disaster resilience of distribution networks, extensive research has been conducted both domestically and internationally, proposing various disaster prevention and mitigation measures such as reinforcing critical lines, optimizing distributed power supply layout, and improving emergency dispatch strategies. These measures have improved the operational reliability of distribution networks to some extent under disaster conditions. However, shortcomings remain in areas such as the overall planning of pre-disaster investment and post-disaster recovery, improving resource allocation efficiency, and optimizing computational methods. Especially under limited budgets and resources, how to formulate distribution network disaster resilience enhancement schemes that balance economic feasibility and practicality remains a critical issue that urgently needs to be addressed. Summary of the Invention

[0004] The purpose of this invention is to provide a system and method for optimizing pre-disaster investment schemes in distribution networks based on genetic algorithms. By introducing genetic algorithms into the disaster mitigation planning of distribution networks, this invention achieves coordinated optimization throughout the entire process before and after a disaster, effectively improving the power supply recovery capability of distribution networks under extreme weather conditions, and has good practical value.

[0005] To achieve this objective, the present invention provides a distribution network pre-disaster investment scheme optimization system based on genetic algorithms, comprising: The basic operation data acquisition module is used to acquire basic operation data of the power distribution network; The disaster scenario generation module is used to generate multiple disaster scenarios based on historical extreme weather data and obtain the occurrence probability of each disaster scenario; The two-stage optimization model building module is used to obtain pre-disaster investment constraints based on the basic operation data of the distribution network; and to obtain the post-disaster maximum load recovery expectation based on various disaster scenarios and the probability of occurrence of the corresponding disaster scenarios. The genetic algorithm optimization module is used to solve the pre-disaster investment constraints using a genetic algorithm, obtain multiple initial pre-disaster investment schemes, and use the genetic algorithm to perform multiple rounds of iterative screening of all initial pre-disaster investment schemes under various disaster scenarios with the fitness of maximizing the expected load recovery after the disaster, to obtain the optimal pre-disaster investment scheme that can meet the expectation of maximizing the load recovery after the disaster.

[0006] Preferably, the basic operating data of the distribution network includes the distribution network topology, load demand of each load node, location and capacity limit of alternative distributed power sources, location of alternative controllable switches, and disaster vulnerability parameters of each transmission line of the distribution network.

[0007] Preferably, the method for generating multiple disaster scenarios based on historical extreme weather data and obtaining the occurrence probability of each disaster scenario includes: generating multiple disaster scenarios based on historical extreme weather data and disaster vulnerability parameters of each transmission line of the distribution network through Monte Carlo simulation or historical statistical methods, and determining the fault status and occurrence probability of each transmission line of the distribution network in each disaster scenario.

[0008] Preferred methods for obtaining pre-disaster investment constraints based on basic operational data of the distribution network include: Pre-disaster investment constraints include budget constraints, capacity constraints, load constraints, and connectivity or recovery status constraints. Budget constraints: ; Capacity constraints: ; Load constraints: ; Connectivity or recovery state constraint: When load 𝑗 is restored in scenario 𝑠, there must be an effective power supply path from the main network or distributed power source to 𝑗 in the reconstructed topology. For this purpose, a binary variable is introduced. and Two load switch constraints; in, For a set of nodes, For the set of load nodes; For nodes The newly added distributed power supply installation capacity, The upper limit is , This is a binary variable indicating whether the line needs to be reinforced. This is a binary variable indicating whether a controllable switch is installed. For the unit cost of distributed power sources, The unit cost of the line, The unit cost of a controllable switch. This represents the upper limit of the total pre-disaster investment budget; For load nodes The rated demand; In the reconstructed network of scenario A, there exists an effective power supply path from the power source to A.

[0009] Preferably, methods for maximizing post-disaster load recovery expectation based on various disaster scenarios and their occurrence probabilities include: ; Where S is the selected set of disaster scenarios, and Ps is the probability of scenario s occurring. For the scene The total load of power restoration in the process.

[0010] Preferably, the method of using a genetic algorithm to solve pre-disaster investment constraints and obtain multiple initial pre-disaster investment schemes includes: representing the location and cost of power distribution line reinforcement, the capacity, node location and cost of installing distributed power sources, and the node location and cost of installing controllable switches in the pre-disaster investment schemes limited by the pre-disaster investment constraints using binary and integer codes respectively, as chromosome coding segments of the genetic algorithm; multiple chromosome coding segments form a complete chromosome; one chromosome represents a pre-disaster investment scheme that satisfies the pre-disaster investment constraints; the genetic algorithm randomly generates multiple initial pre-disaster investment schemes; and the multiple initial pre-disaster investment schemes serve as the initial population of the genetic algorithm.

[0011] Preferably, the method of using a genetic algorithm to iteratively screen all initial pre-disaster investment schemes under various disaster scenarios, with the fitness value being the expected maximum load recovery after a disaster, to obtain the optimal pre-disaster investment scheme that can satisfy the expected maximum load recovery after a disaster includes: for each initial pre-disaster investment scheme in the initial population, simulating the damage and recovery process of the distribution network under each disaster scenario, calculating the average recovery load or the probability-weighted recovery load of all disaster scenarios, using the average recovery load or the probability-weighted recovery load of all disaster scenarios as the expected recovery load value, using the expected recovery load as the fitness value of the initial pre-disaster investment scheme, and using a tournament selection or roulette wheel selection mechanism based on the fitness value to select from the initial... The genetic algorithm selects pre-disaster investment schemes that meet preset screening criteria from the population and introduces them into the next generation of pre-disaster investment scheme population. A crossover operation is performed on the selected pre-disaster investment schemes that meet the preset screening criteria, exchanging some chromosome segments to generate new offspring pre-disaster investment schemes. The chromosomes of these offspring pre-disaster investment schemes are then randomly mutated with preset probabilities, randomly flipping the binary bits of certain chromosome coding segments or adjusting the parameter values ​​of certain chromosome coding segments to obtain a new generation of pre-disaster investment scheme population. The fitness of this new generation of pre-disaster investment scheme population is calculated again, and the above selection, crossover, and mutation operations are repeated. After iterating and updating the pre-disaster investment scheme population for multiple generations, the genetic algorithm stops when a preset termination condition is met, yielding the optimal pre-disaster investment scheme for the distribution network. Furthermore, a distribution network disaster resistance enhancement system based on a distribution network pre-disaster investment scheme optimization system is characterized in that the distribution network disaster resistance enhancement system is used to formulate corresponding post-disaster power supply restoration strategies based on the optimal pre-disaster investment scheme and the distribution network fault situation caused by the disaster. The specific methods for formulating corresponding post-disaster power restoration strategies are as follows: For each disaster scenario, perform fault isolation and network reconstruction operations, activate alternative distributed power sources deployed with the best pre-disaster investment plan to support power supply, control switches to switch power supply paths, prioritize the restoration of power supply to important load nodes, and maximize the amount of load restored in that scenario.

[0012] Furthermore, a method for optimizing pre-disaster investment schemes for distribution networks based on a genetic algorithm, according to the aforementioned distribution network pre-disaster investment scheme optimization system, includes: Obtain basic operational data of the power distribution network; Multiple disaster scenarios are generated based on historical extreme weather data, and the probability of occurrence of each disaster scenario is obtained; Pre-disaster investment constraints are derived from the basic operational data of the distribution network; post-disaster maximum load recovery expectations are derived from various disaster scenarios and their corresponding probabilities of occurrence. The genetic algorithm optimization module is used to solve the pre-disaster investment constraints using a genetic algorithm, obtain multiple initial pre-disaster investment schemes, and use the genetic algorithm to perform multiple rounds of iterative screening of all initial pre-disaster investment schemes under various disaster scenarios with the fitness of maximizing the expected load recovery after the disaster, to obtain the optimal pre-disaster investment scheme that can meet the expectation of maximizing the load recovery after the disaster.

[0013] The beneficial effects of this invention are as follows: This method improves the overall disaster resistance and resource allocation efficiency of the distribution network under limited budget constraints through two-stage collaborative optimization of pre-disaster reinforcement and post-disaster recovery; it can efficiently find the best solution and generate high-quality disaster relief decision schemes under multiple scenarios and complex disaster conditions by leveraging the global search advantage of genetic algorithms; and it can also combine rapid power restoration strategies such as pre-deployment of distributed power sources and network reconfiguration to maximize the priority restoration of important loads, shorten power outage time, and reduce power outage losses, thereby significantly enhancing the power restoration capability of the distribution network under extreme weather conditions. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating a method for enhancing the disaster resistance capability of a distribution network based on a genetic algorithm, according to the present invention. Figure 2 This is a schematic diagram of the structure of the present invention. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to represent selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0016] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments: Example 1 like Figure 2 As shown, a distribution network pre-disaster investment optimization system based on genetic algorithm includes: The basic operation data acquisition module is used to acquire basic operation data of the power distribution network; The disaster scenario generation module is used to generate multiple disaster scenarios based on historical extreme weather data and obtain the occurrence probability of each disaster scenario; The two-stage optimization model building module is used to obtain pre-disaster investment constraints based on the basic operation data of the distribution network; and to obtain the post-disaster maximum load recovery expectation based on various disaster scenarios and the probability of occurrence of the corresponding disaster scenarios. The genetic algorithm optimization module is used to solve the pre-disaster investment constraints using a genetic algorithm, obtain multiple initial pre-disaster investment schemes, and use the genetic algorithm to perform multiple rounds of iterative screening of all initial pre-disaster investment schemes under various disaster scenarios with the fitness of maximizing the expected load recovery after the disaster, to obtain the optimal pre-disaster investment scheme that can meet the expectation of maximizing the load recovery after the disaster.

[0017] In some technical solutions, the basic operating data of the distribution network includes the distribution network topology, load demand of each load node, location and capacity limit of alternative distributed power sources, location of alternative controllable switches, and disaster vulnerability parameters of each transmission line of the distribution network.

[0018] In some embodiments, basic data for power grid planning is obtained from the power company to which the distribution network belongs. Distribution network topology, load node requirements, configuration parameters of alternative distributed power sources and switches, and data on line disaster vulnerability are used to construct an accurate foundation for optimization models, providing a factual data basis for the coordinated optimization of pre-disaster investment and post-disaster recovery.

[0019] In some technical solutions, the methods for generating multiple disaster scenarios based on historical extreme weather data and obtaining the occurrence probability of each disaster scenario include: generating multiple disaster scenarios through Monte Carlo simulation or historical statistical methods based on historical extreme weather data and disaster vulnerability parameters of each transmission line of the distribution network, and determining the fault status and occurrence probability of each transmission line of the distribution network in each disaster scenario.

[0020] In some embodiments, the method for generating multiple typical disaster scenarios through Monte Carlo simulation includes: establishing an intensity model of extreme weather events (such as typhoons and hail). Taking typhoons as an example, key parameters of typhoon weather (such as central pressure, movement speed, radius, etc.) are determined, and a physical wind field model (such as a gradient wind field model) is used to calculate the disaster intensity (such as wind speed) at the location of each transmission line in the power distribution network, thereby transforming a typhoon scenario into a spatially uneven intensity field acting on the power distribution network. Vulnerability parameters are usually presented in the form of curves or tables relating failure probability and disaster intensity. For example, the failure probability of a transmission line is 10% at wind speed of 33 m / s, 50% at 38 m / s, and 90% at 43 m / s. Sampling and judgment are performed: A segment of transmission line i is randomly selected. The probability of transmission line i failing under the current disaster intensity is set as Pi. A random number Ri, uniformly distributed in the interval [0,1], is generated for each transmission line i. If Ri ≤ Pi, then transmission line i fails in the current scenario; if Ri > Pi, then transmission line i is normal in the current scenario. This random sampling process is repeated for all transmission lines in the distribution network. The state of each transmission line segment under typhoon weather constitutes a complete disaster scenario. Since the failure state of each transmission line is independently and randomly sampled, the probability of the corresponding disaster scenario can be estimated by the joint probability of the states of all its transmission lines. In practice, tens of thousands of simulations are usually performed, with each scenario considered as a sample, and its probability of occurrence is approximately equal, 1 / N (N is the total number of simulations). Through statistical analysis of a large number of samples, representative typical scenarios can be selected, and their probability is estimated by the frequency of occurrence of this type of scenario.

[0021] Historical statistical methods directly analyze recorded historical disaster events and their consequences. From each type of historical weather disaster, one or more typical disaster scenarios are extracted, and the typical fault modes of power distribution lines in those events are directly adopted. For example, for typhoons characterized by strong winds and coastal areas, a typical pattern is summarized as "approximately 60% fault rate for 10kV lines in coastal areas and approximately 20% in inland areas." Values ​​are then assigned based on the frequency of these historical weather disasters. For instance, if a certain type of typhoon occurred 5 times in the past 50 years, the annual probability of this typhoon scenario can be estimated as 5 / 50 = or 0.1.

[0022] Monte Carlo simulations or historical statistical methods transform the uncertainty caused by severe weather into a quantifiable set of random events, providing a real-world data foundation for optimizing pre-disaster investment plans.

[0023] Some technical solutions involve methods for deriving pre-disaster investment constraints based on basic operational data of the distribution network, including: Pre-disaster investment constraints include budget constraints, capacity constraints, load constraints, and connectivity or recovery status constraints; Budget constraints: ; Capacity constraints: ; Load constraints: ; Connectivity or recovery state constraint: When load 𝑗 is restored in scenario 𝑠, there must be an effective power supply path from the main network or distributed power source to 𝑗 in the reconstructed topology. For this purpose, a binary variable is introduced. and Two load switch constraints; in, For a set of nodes, For the set of load nodes; For nodes The newly added distributed power supply installation capacity, The upper limit is , This is a binary variable indicating whether the line needs to be reinforced. This is a binary variable indicating whether a controllable switch is installed. For the unit cost of distributed power sources, The unit cost of the line, The unit cost of a controllable switch. This represents the upper limit of the total pre-disaster investment budget; For load nodes The rated demand; To ensure a valid power supply path exists from the power source to the network in scenario A, the node set is defined. Load node set ,node New distributed power installation capacity A binary variable indicating whether the line has been reinforced. A binary variable indicating whether a controllable switch is installed. Load nodes Rated requirements All of these are basic data for the power distribution network.

[0024] Budget constraints require that the total cost of pre-disaster investment does not exceed the predetermined budget limit; distributed power capacity constraints require that the output of newly deployed distributed power sources at each node does not exceed the maximum capacity allowed for that node; load recovery constraints stipulate that the amount of power restored to each load node in any scenario does not exceed the maximum load demand of that node under normal conditions; and network connectivity constraints require that the distribution network after switching operations should remain connected after the disaster to ensure that each power supply branch has an effective path from the power source (main substation or distributed power source) to the supplied load.

[0025] Some technical solutions involve methods for maximizing post-disaster load recovery expectations based on various disaster scenarios and their probabilities of occurrence. ; Where S is the selected set of disaster scenarios, and Ps is the probability of scenario s occurring. For the scene The total load of power restoration in the process.

[0026] By formally defining the recovery effects across multiple scenarios mathematically, a single optimization objective is aggregated, and an expected value function with probability weights is constructed. This transforms the optimization decision-making process from addressing a single scenario to minimizing the overall risk across all possible scenarios. Uncertain disaster impacts are converted into definite quantitative indicators, providing a calculable fitness evaluation basis for optimization methods such as genetic algorithms. A probability-weighted approach for disaster scenarios balances high-frequency, low-impact events with low-frequency, high-impact events, avoiding oversensitivity to extreme low-probability disaster scenarios or neglect of common disaster scenarios. In some embodiments, the pre-disaster investment constraints and the post-disaster maximization of load recovery expectations constitute a two-stage optimization model encompassing pre-disaster investment and post-disaster recovery. This model comprehensively considers multiple representative disaster scenarios, calculating the expected total post-disaster recovery load by introducing scenario occurrence probabilities.

[0027] In some technical solutions, the method of using genetic algorithms to solve pre-disaster investment constraints and obtain multiple initial pre-disaster investment schemes includes: representing the location and cost of power distribution line reinforcement, the capacity, node location and cost of installing distributed power sources, and the node location and cost of installing controllable switches in the pre-disaster investment schemes limited by the pre-disaster investment constraints using binary and integer codes respectively, as chromosome coding segments of the genetic algorithm; multiple chromosome coding segments form a complete chromosome; one chromosome represents a pre-disaster investment scheme that satisfies the pre-disaster investment constraints; the genetic algorithm randomly generates multiple initial pre-disaster investment schemes; and multiple initial pre-disaster investment schemes serve as the initial population of the genetic algorithm.

[0028] The aforementioned two-stage optimization model involves discrete investment decision variables and considers multiple scenarios, resulting in a large problem scale and high complexity. To efficiently solve this optimization problem, a genetic algorithm is employed to solve the two-stage optimization model. First, the pre-disaster investment decisions are encoded, with each individual (chromosome) containing key information about the distribution network reinforcement scheme, such as the location and capacity of distributed power sources, the selection of reinforcement lines, and the location of newly installed switches. Then, for each individual, its comprehensive power supply restoration effect under all disaster scenarios (i.e., maximizing the expected load restoration after a disaster) is evaluated, and this value is used as the fitness. For schemes that violate budget, capacity, or other constraints, a large penalty is added to the fitness function to decrease their fitness. The genetic algorithm iteratively searches the candidate scheme space through selection, crossover, and mutation operations, continuously generating new distribution network reinforcement schemes and evaluating them. After multiple generations of evolution, the algorithm converges to obtain the optimal or near-optimal distribution network disaster resistance enhancement scheme.

[0029] In some technical solutions, a genetic algorithm is used to iteratively screen all initial pre-disaster investment schemes under various disaster scenarios, using the expected post-disaster load recovery as the fitness, to obtain the optimal pre-disaster investment scheme that can meet the expected post-disaster load recovery. This method includes: for each initial pre-disaster investment scheme in the initial population, simulating the damage and recovery process of the distribution network under each disaster scenario, calculating the average recovery load or the probability-weighted recovery load of all disaster scenarios, using the average recovery load or the probability-weighted recovery load of all disaster scenarios as the expected recovery load value, and using the expected recovery load as the fitness of the initial pre-disaster investment scheme. Based on the fitness value, a tournament selection or roulette wheel selection mechanism is used to select from... In the initial population, pre-disaster investment schemes that meet preset screening conditions are selected to enter the next generation of pre-disaster investment scheme population. Crossover operation is performed on the selected pre-disaster investment schemes that meet the preset screening conditions, exchanging some chromosome segments to generate new offspring pre-disaster investment schemes. The chromosomes of the offspring pre-disaster investment schemes are randomly mutated with preset probability, randomly flipping the binary bits of some chromosome coding segments or adjusting the parameter values ​​of some chromosome coding segments to obtain a new generation of pre-disaster investment scheme population. The fitness of the new generation of pre-disaster investment scheme population is calculated again, and the above selection, crossover, and mutation operations are repeated. After iterating and updating the pre-disaster investment scheme population for multiple generations, the genetic algorithm stops when the preset termination condition is reached, and the optimal distribution network pre-disaster investment scheme is obtained.

[0030] By quantifying the expected recovery load under multiple disaster scenarios into the fitness function of a genetic algorithm, and using tournament or roulette wheel selection mechanisms to retain high-quality solutions, the advantages of different solutions are integrated through crossover operations, and innovative solutions are introduced through mutation operations, forming an iterative mechanism for continuous optimization. This global search strategy based on the principle of natural evolution can efficiently explore the solution space under complex constraints, and finally output the optimal investment plan that maximizes the post-disaster recovery capacity of the distribution network under limited budget, significantly improving the economic efficiency of disaster relief planning.

[0031] In some embodiments, tournament selection, roulette wheel selection, or other mechanisms can be employed, and an elite strategy can be introduced to select pre-disaster investment plans. Tournament selection mechanism: A small number of individuals (initial pre-disaster investment plans) are randomly selected from the current population to form a temporary competitive group. Within the group, the fitness values ​​of each individual are compared, and the individual with the highest fitness in the group is selected as the winner and placed into the next generation population. This competitive process is repeated multiple times, with individuals randomly selected each time, until the next generation population reaches the expected size. Roulette wheel selection mechanism: The fitness value of each individual (initial pre-disaster investment plan) in the population is calculated, and the sum of the fitness values ​​of all individuals is calculated. Based on the fitness of each individual... The proportion of the fitness value to the total fitness value determines the probability of each individual being selected (higher fitness values ​​result in a higher probability). A roulette wheel is constructed based on the probability of each individual being selected, with the size of the region on the wheel proportional to the selection probability of each individual. Spinning the wheel generates a random number, and the region where the random number lands is used to select the corresponding individual. This process of spinning the wheel is repeated until a sufficient number of individuals are selected to form the next generation population. Elite strategy: Before the selection operation, one or more individuals with the highest fitness are selected from the current population (called elite individuals, representing the initial pre-disaster investment plan). These elite individuals are directly retained in the next generation population without undergoing selection, crossover, or mutation operations. The remaining individuals are then subjected to selection (e.g., tournament or roulette wheel), crossover, and mutation operations to generate the other individuals in the next generation population. This ensures that the optimal solution of this generation is not lost due to genetic operations, thereby increasing the probability of convergence to the global optimum or a near-global optimum. The preset probability values ​​include, but are not limited to, 0.1% to 5%. The preset termination conditions include when the number of iterations reaches the upper limit (specifically set according to the scale of the distribution network) or when the improvement in optimal fitness is lower than a threshold for several consecutive generations (including but not limited to 0.01%-0.1%, which can be specifically set based on actual accuracy requirements), the genetic algorithm stops. The average recovery load can be obtained by averaging the recovery loads in all disaster scenarios. The probability-weighted recovery load of the disaster scenarios can be obtained by multiplying the recovery loads in all disaster scenarios by the probability of occurrence of the corresponding disaster scenario and summing them.

[0032] The optimal chromosome output by the genetic algorithm corresponds to a pre-disaster investment plan, which is the final result of the distribution network disaster resilience enhancement method described in this invention. This plan includes the optimal distributed power deployment plan, line reinforcement list, and switch installation scheme within the budget. Power operators can then formulate implementation plans based on this, strengthening key links in the distribution network in advance and configuring backup power resources to enable the distribution network to restore power more quickly in the event of extreme weather disasters, reducing losses caused by power outages.

[0033] Example 2 A distribution network disaster resistance enhancement system based on the aforementioned distribution network pre-disaster investment scheme optimization system includes: the distribution network disaster resistance enhancement system is used to formulate corresponding post-disaster power supply restoration strategies based on the optimal pre-disaster investment scheme and the distribution network fault situation caused by the disaster; The specific methods for formulating corresponding post-disaster power restoration strategies are as follows: For each disaster scenario, perform fault isolation and network reconstruction operations, activate alternative distributed power sources deployed with the best pre-disaster investment plan to support power supply, control switches to switch power supply paths, prioritize the restoration of power supply to important load nodes, and maximize the amount of load restored in that scenario.

[0034] like Figure 1 The diagram illustrates a flowchart of a method for enhancing the disaster resilience of power distribution networks based on genetic algorithms. It includes two stages: pre-disaster investment planning and post-disaster recovery optimization. The genetic algorithm is used to globally optimize the decisions made in both stages. First, in the data preparation stage, parameters related to the distribution network topology, load distribution, and alternative reinforcement measures are collected, and several extreme weather disaster scenarios are generated based on historical meteorological data. Then, in the pre-disaster investment decision-making stage, considering a pre-set investment budget, several optional reinforcement measures are optimized and selected. These include measures such as configuring distributed power sources and their capacity at which nodes, reinforcing existing distribution lines over a specific area, and adding controllable switches at certain locations. Through reasonable pre-disaster reinforcement, the probability of damage to critical equipment can be reduced during a disaster, and favorable conditions can be created for proactive post-disaster recovery.

[0035] During the post-disaster recovery decision-making phase, when a disaster scenario occurs, some lines or equipment in the distribution network may trip due to faults. In this case, the distribution automation system is used to quickly isolate the affected lines and prevent the fault from spreading. Simultaneously, previously installed controllable switches are used to restructure the network structure, ensuring that undamaged network segments reconnect to power supply paths. By activating pre-installed distributed power sources to supply power to nearby isolated networks and adjusting the network topology to prioritize power supply to critical load nodes, the power supply service of the distribution network can be restored to the greatest extent possible. Under each disaster scenario, the above measures are used to meet load demand as much as possible, and the total load successfully restored under that scenario is calculated.

[0036] Example 3 A method for optimizing pre-disaster investment schemes for distribution networks based on a genetic algorithm, according to the aforementioned distribution network pre-disaster investment scheme optimization system, includes: Obtain basic operational data of the power distribution network; Multiple disaster scenarios are generated based on historical extreme weather data, and the probability of occurrence of each disaster scenario is obtained; Pre-disaster investment constraints are derived from the basic operational data of the distribution network; post-disaster maximum load recovery expectations are derived from various disaster scenarios and their corresponding probabilities of occurrence. The genetic algorithm optimization module is used to solve the pre-disaster investment constraints using a genetic algorithm, obtain multiple initial pre-disaster investment schemes, and use the genetic algorithm to perform multiple rounds of iterative screening of all initial pre-disaster investment schemes under various disaster scenarios with the fitness of maximizing the expected load recovery after the disaster, to obtain the optimal pre-disaster investment scheme that can meet the expectation of maximizing the load recovery after the disaster.

[0037] Example 4 The present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the method described in Embodiment 3.

[0038] This invention can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented in whole or in part as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0039] It will be readily understood by those skilled in the art that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, combinations, substitutions, improvements, etc., made under the spirit and principles of the present invention are included within the protection scope of the present invention.

[0040] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

Claims

1. A pre-disaster investment scheme optimization system for a power distribution network based on a genetic algorithm, characterized by, It comprises: The basic operation data acquisition module is used for acquiring basic operation data of the power distribution network; The disaster scenario generation module is used for generating a plurality of disaster scenarios based on historical extreme weather data, and obtaining occurrence probabilities of the disaster scenarios; The two-stage optimization model establishment module is used for obtaining pre-disaster investment constraints according to the basic operation data of the power distribution network, and obtaining post-disaster maximum load recovery expectations according to the disaster scenarios and the occurrence probabilities of the disaster scenarios; The genetic algorithm optimization module is used for solving the pre-disaster investment constraints by using a genetic algorithm, obtaining a plurality of initial pre-disaster investment schemes, and performing multiple rounds of iteration screening on all the initial pre-disaster investment schemes by using the genetic algorithm under the plurality of disaster scenarios and taking the post-disaster maximum load recovery expectations as fitness, to obtain an optimal pre-disaster investment scheme capable of meeting the post-disaster maximum load recovery expectations.

2. The system of claim 1, wherein The basic operation data of the power distribution network comprises a power distribution network topology, load demands of each load node, positions and capacity upper limits of alternative distributed power sources, positions of alternative controllable switches, and disaster vulnerability parameters of each power transmission line of the power distribution network.

3. The system of claim 2, wherein The method for generating a plurality of disaster scenarios based on historical extreme weather data and obtaining occurrence probabilities of the disaster scenarios comprises: generating a plurality of disaster scenarios and determining failure conditions and occurrence probabilities of each power transmission line of the power distribution network in each disaster scenario by using Monte Carlo simulation or historical statistical methods based on historical extreme weather data and disaster vulnerability parameters of each power transmission line of the power distribution network.

4. The system of claim 1, wherein The method for obtaining pre-disaster investment constraints according to the basic operation data of the power distribution network comprises: The pre-disaster investment constraint function comprises a budget constraint, a capacity constraint, a load constraint, and a connectivity or recovery state constraint; Budget constraints: ; Capacity constraints: ; Load constraints: ; Connectivity or restoration state constraints: when a load j is restored in a scenario s, there must exist an active power supply path from the main grid or distributed generation to j in the reconfigured topology, for which a binary variable and Two load switch constraints; wherein, is a set of nodes, is a set of load nodes; is a node of the newly installed distributed power supply capacity, the upper limit of which is , is a binary variable of whether the line is reinforced, is a binary variable of whether the controllable switch is installed; is the unit cost of the distributed power supply, is the unit cost of the line, is the unit cost of the controllable switch, is the upper limit of the total pre-disaster investment budget; is the rated demand of the load node ; is that there is an effective power supply path from the power supply to j in the reconstructed network of the scene s.

5. The pre-disaster investment scheme optimization system for power distribution network based on genetic algorithm of claim 1, wherein: The method for obtaining post-disaster maximum load recovery expectations according to each disaster scenario and the occurrence probability thereof comprises: ； Wherein, S is a selected disaster scenario set, Ps is the occurrence probability of scenario s, is the total load amount of the power supply recovery in scenario .

6. The pre-disaster investment scheme optimization system for power distribution network based on genetic algorithm of claim 1 or 4, characterized in that: The method for solving the pre-disaster investment constraints by using a genetic algorithm to obtain a plurality of initial pre-disaster investment schemes comprises: representing positions and costs of power distribution line reinforcements, capacities and node positions and costs of installed distributed power sources, and node positions and costs of installed controllable switches in a pre-disaster investment scheme defined in the pre-disaster investment constraints by using binary and integer coding respectively as chromosome coding segments of the genetic algorithm, combining a plurality of chromosome coding segments to form a complete chromosome, taking one chromosome as representing one pre-disaster investment scheme meeting the pre-disaster investment constraints, randomly generating a plurality of initial pre-disaster investment schemes by the genetic algorithm, and taking the plurality of initial pre-disaster investment schemes as an initial population of the genetic algorithm.

7. The system of claim 1, wherein the system is characterized by: The method for obtaining the optimal pre-disaster investment scheme capable of meeting the post-disaster maximum load recovery expectation comprises the following steps: for each initial pre-disaster investment scheme in the initial population, simulating the damage and recovery process of the power distribution network under each disaster scenario, calculating the average recovery load in all disaster scenarios or the disaster scenario probability weighted recovery load, taking the average recovery load in all disaster scenarios or the disaster scenario probability weighted recovery load as the expected recovery load value, taking the expected recovery load as the fitness of the initial pre-disaster investment scheme, selecting the pre-disaster investment schemes meeting the preset screening condition from the initial population into the next generation of pre-disaster investment scheme population according to the fitness value, performing the crossover operation on the selected pre-disaster investment schemes meeting the preset screening condition to exchange part of the chromosome fragments, generating new offspring pre-disaster investment schemes, randomly mutating the chromosomes of the offspring pre-disaster investment schemes at a preset probability, randomly flipping the binary bits of some chromosome coding fragments or adjusting the parameter values of some chromosome coding fragments, obtaining the new generation of pre-disaster investment scheme population, calculating the fitness of the new generation of pre-disaster investment scheme population again, and repeatedly performing the selection, crossover and mutation operations, so as to iteratively update the multiple generations of pre-disaster investment scheme population, and when the preset termination condition is reached, the genetic algorithm stops, and the optimal pre-disaster investment scheme of the power distribution network is obtained.

8. A power grid disaster resistance enhancement system based on the power grid pre-disaster investment scheme optimization system of claim 1, characterized in that, The power distribution network disaster resistance enhancement system is used for formulating a corresponding post-disaster power supply recovery strategy according to the power distribution network fault caused by disasters based on the optimal pre-disaster investment scheme. The specific method for formulating the corresponding post-disaster power supply recovery strategy comprises the following steps: Performing the fault isolation and network reconstruction operation for each disaster scenario, enabling the standby distributed power supply supported by the optimal pre-disaster investment scheme to supply power, controlling the switch to switch the power supply path, and preferentially recovering the power supply of important load nodes to maximize the amount of recovered loads under the scenario.

9. The power distribution network pre-disaster investment scheme optimization method based on the genetic algorithm of the power distribution network pre-disaster investment scheme optimization system according to claim 1, characterized in that, The method comprises the following steps: Obtaining the basic operation data of the power distribution network; Generating multiple disaster scenarios based on historical extreme weather data and obtaining the occurrence probability of each disaster scenario; Obtaining the pre-disaster investment constraint based on the basic operation data of the power distribution network, and obtaining the post-disaster maximum load recovery expectation based on each disaster scenario and the occurrence probability of the corresponding disaster scenario; The genetic algorithm optimization module is used for solving the pre-disaster investment constraint by using the genetic algorithm, obtaining multiple initial pre-disaster investment schemes, and performing multiple rounds of iterative screening on all initial pre-disaster investment schemes by using the genetic algorithm under multiple disaster scenarios with the post-disaster maximum load recovery expectation as the fitness, so as to obtain the optimal pre-disaster investment scheme capable of meeting the post-disaster maximum load recovery expectation.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the method in claim 9.

Images