Optimization method for reducing vulnerability of power distribution network based on demand side resources

Abstract

The invention discloses an optimization method for reducing the vulnerability of a power distribution network based on demand side resources, and relates to the technical field of power distribution network safety maintenance, and the method comprises the steps: constructing a demand side resource pool, building a resource adjustable capability model, and outputting resource adjustable capability data; based on the output resource adjustable capability data, constructing a vulnerability evaluation index system and calculating an index weight, and outputting a power distribution network vulnerability comprehensive evaluation value through a fuzzy comprehensive evaluation method; a multi-objective optimization modeling technology is adopted to construct an optimization scheduling model; and based on the scheduling model containing the constraint condition, solving by using an improved intelligent optimization algorithm, and outputting a demand side resource optimal scheduling strategy. Through the big data analysis and machine learning technology, the operation state and load demand data of each node of the power distribution network are acquired in real time, resource scheduling and vulnerability assessment are performed by using a multi-objective optimization algorithm, and the problems of inaccurate scheduling and lagging vulnerability assessment of a traditional power distribution network in a high-load and complex environment are solved.

Description

Technical Field

[0001] This invention relates to the field of distribution network security maintenance technology, specifically to an optimization method for reducing the vulnerability of distribution networks based on demand-side resources. Background Technology

[0002] The distribution network is a core component of the power system, responsible for providing electricity to end users. However, due to its complex topology, wide power supply area, and large load fluctuations, the security and stability of the distribution network are easily affected by various factors such as natural disasters, equipment failures, and sudden load changes, leading to increased vulnerability and even large-scale power outages. To improve the distribution network's resistance to disturbances and its recovery capabilities, current improvement methods mostly rely on enhancing physical infrastructure, such as adding power equipment and expanding the network, resulting in high costs and wasted resources.

[0003] With the development of smart grid technology, demand-side management, as an emerging power regulation method, has gradually gained attention. Through intelligent scheduling, demand-side resources can alleviate the vulnerability of the distribution network to a certain extent and improve its operational reliability and stability. These resources have strong flexibility and can be adjusted according to real-time load changes, thereby effectively reducing the risk of the distribution network and mitigating the impact of potential disturbances on its operation. However, how to rationally schedule these resources and build an efficient optimization scheduling model remains an urgent problem to be solved. Summary of the Invention

[0004] To address the aforementioned technical issues and provide an optimization method for reducing the vulnerability of distribution networks based on demand-side resources, this technical solution resolves the aforementioned problems.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: Optimization methods for reducing distribution network vulnerability based on demand-side resources include: The hierarchical classification method is used to sort out the demand-side resources within the distribution network coverage area, collect key resource parameters, construct a demand-side resource pool, establish a resource adjustability model, and output resource adjustability data. Based on the output resource adjustability data, combined with the distribution network topology, operating conditions and potential disturbance scenarios, a vulnerability assessment index system is constructed and the index weights are calculated. The comprehensive evaluation value of the distribution network vulnerability is output through the fuzzy comprehensive evaluation method. Based on the comprehensive vulnerability assessment value, and with demand-side resources and the safe operation of the distribution network as constraints, an optimized scheduling model is constructed using multi-objective optimization modeling technology, and the output scheduling model with constraints is generated. Based on the scheduling model with constraints, an improved intelligent optimization algorithm is used to solve the problem. After adjustment by feasibility verification technology, the optimal scheduling strategy for demand-side resources is output.

[0006] Preferably, a hierarchical classification method is used to analyze demand-side resources within the distribution network coverage area, collect key resource parameters, construct a demand-side resource pool, establish a resource adjustability model, and output resource adjustability data, including: The demand-side resources are divided into four core types using a hierarchical classification method: distributed power sources, energy storage devices, flexible loads, and adjustable charging piles. Key parameters of each type of resource are obtained by combining real-time sensor data collection with ledger statistics. The parameters of distributed power sources include output characteristics and fluctuation range; the parameters of energy storage devices include charging and discharging capacity, efficiency, and SOC threshold; the parameters of flexible loads include transferable capacity and response time; and the parameters of adjustable charging piles include charging power and user demand elasticity. Based on resource types and key parameters, a demand-side resource pool is constructed, and a resource information ledger is established to record resource location, capacity, schedulable time periods, and technical constraints. Differentiated adjustable capability models are established for different types of resources. Distributed power sources use output probability models to describe uncertainties, energy storage devices use charge and discharge state equations to characterize adjustment characteristics, flexible loads use load transfer matrices to quantify adjustment potential, and adjustable charging piles use user response coefficients to reflect adjustment intentions. Historical operational data is used to verify the error of various resource adjustability models. The least squares method is used to correct the model parameters. The error of the resource adjustability data output by the model is within the preset threshold. Finally, the verified resource adjustability data is output.

[0007] Preferably, based on the output resource adjustability data, combined with the distribution network topology, operating conditions, and potential disturbance scenarios, a vulnerability assessment index system is constructed and index weights are calculated. The comprehensive vulnerability assessment value of the distribution network is then output using the fuzzy comprehensive evaluation method, including: Based on verified resource adjustability data, topology parsing technology is used to extract the topology information of distribution network node connection relationships and line parameters, and operating condition data is collected. The scenario enumeration method is used to set four potential disturbance scenarios: extreme weather, equipment failure, load change, and new energy output fluctuation. A vulnerability assessment index system is constructed from three dimensions: structural vulnerability, operational vulnerability, and recovery vulnerability. The structural vulnerability index includes node degree centrality, line betweenness, and network connectivity; the operational vulnerability index includes node voltage deviation rate, line load rate, and power deficit rate; and the recovery vulnerability index includes fault recovery time, load recovery rate, and resource scheduling response speed. The weights of each evaluation index are calculated using the entropy-weighted analytic hierarchy process (AHP). Expert subjective weights are determined through AHP, and objective data weights are calculated based on information entropy theory. These are then integrated according to a preset ratio to obtain the comprehensive weight. Resource adjustability data, topology information, operating condition data, and scenario information are substituted into the index system. The fuzzy comprehensive evaluation method is used to calculate the distribution network vulnerability evaluation value under each scenario. The value is then weighted and summed based on the scenario occurrence probability to output the comprehensive distribution network vulnerability evaluation value. The formula for calculating the comprehensive distribution network vulnerability evaluation value is as follows:

[0008] In the formula, For comprehensive evaluation, As an indicator The weight, As an indicator membership degree This represents the number of samples.

[0009] Preferably, based on the comprehensive vulnerability assessment value, and constrained by demand-side resources and the safe operation of the distribution network, an optimized scheduling model is constructed using multi-objective optimization modeling techniques, outputting a scheduling model with constraints, including: A multi-objective optimization objective function is constructed, with the core objective being to minimize the comprehensive evaluation value of the distribution network vulnerability. The auxiliary objectives include minimizing the demand-side resource scheduling cost and maximizing user electricity satisfaction. The weighting coefficient method is used to transform the multi-objective into a single objective. Set demand-side resource constraints: distributed power output constraints, energy storage device charging and discharging power and SOC constraints, flexible load regulation capacity and time constraints, adjustable charging pile charging power and user demand constraints. Set the following constraints for safe operation of the distribution network: node voltage upper and lower limits, line transmission power constraints, power balance constraints, and frequency stability constraints. Uncertainty constraints are introduced into the scenario, and a robust optimization method is used to handle the uncertainty of distributed power output and load demand. An uncertainty budget is set to represent the degree of disturbance. The nonlinear constraints are transformed into linear constraints by linearization methods, forming a solvable mixed-integer linear programming model.

[0010] Preferably, based on a scheduling model with constraints, an improved intelligent optimization algorithm is used to solve the problem. After adjustment using feasibility verification technology, the optimal scheduling strategy for demand-side resources is output, including: Particle swarm optimization algorithm was selected as the basic algorithm, and it was improved by combining crossover and mutation operations of genetic algorithm to construct an improved particle swarm optimization algorithm. The population size, maximum number of iterations, learning factor, crossover probability and mutation probability parameters were determined by trial and error. The demand-side resource scheduling scheme is encoded as individual particles, and the optimization objective function value is used as the particle fitness value. After initializing the population, the fitness value of each particle is calculated, the particle velocity and position are updated according to preset rules, and crossover and mutation operations are performed to increase population diversity. Determine if the iteration termination condition is met. If it is, output the scheduling scheme corresponding to the current optimal particle; otherwise, continue iterating. Use constraint verification technology to verify whether the optimal scheduling scheme meets all the constraints set in step 3. If there is a constraint violation, use the local adjustment method to correct the violation parameters until the scheme meets all constraints. Finally, output the optimal scheduling strategy for demand-side resources.

[0011] Preferably, based on the output demand-side resource optimal scheduling strategy, combined with real-time monitoring data of the distribution network, the scheduling strategy is dynamically corrected using deviation analysis and rolling optimization techniques, and the dynamically corrected optimal scheduling strategy is output, including: Establish a real-time monitoring platform for the power distribution network, collect power distribution network operation parameters, demand-side resource status parameters, and environmental parameters, with a data sampling frequency of no less than 5 minutes / time; calculate the deviation between the real-time monitoring data and the input parameters of the optimized scheduling model, set a deviation threshold, and trigger a strategy correction mechanism if the deviation exceeds the preset threshold; The initial parameters of the scheduling model are updated and optimized based on real-time deviation data. The model is then re-solved using a rolling optimization method to obtain a preliminary revised scheduling strategy. The constraint satisfaction of the revised strategy is verified again using feasibility verification technology. For the parts that do not meet the constraints, a secondary adjustment is made to output the dynamically revised optimal scheduling strategy.

[0012] Preferably, before constructing the vulnerability assessment index system, the method further includes identifying key nodes in the distribution network to provide a basis for setting index weights, including: Based on the distribution network topology data, a node importance assessment method is adopted, which combines the node load scale, power supply range, and connection relationship with other nodes to construct a node importance calculation model and calculate the importance value of each node. Set an importance threshold, mark nodes with importance values ​​higher than the threshold as key nodes, focus on statistically analyzing the resource distribution of key nodes, and output a list of key nodes and surrounding resource distribution data; substitute the relevant information of key nodes into the indicator weight calculation process, set weight tilt coefficients for the evaluation indicators corresponding to key nodes, and increase the weight ratio of key node-related indicators.

[0013] Preferably, the step of calculating the vulnerability evaluation value of the distribution network under each scenario using the fuzzy comprehensive evaluation method includes: Construct a vulnerability assessment index system and calculate index weights. Standardize the actual values ​​of each assessment index extracted in step 2 to eliminate the difference in index dimensions and obtain standardized index values. Construct a fuzzy evaluation matrix, determine the fuzzy membership function of each index, substitute the standardized index values ​​into the membership function, and calculate the membership degree of each index to different vulnerability levels. The calculated comprehensive weights of the indicators are multiplied by the fuzzy evaluation matrix to obtain the fuzzy comprehensive evaluation value of the distribution network vulnerability under each scenario. The fuzzy comprehensive evaluation value is mapped to a specific vulnerability level using the maximum membership principle. Combined with the probability of scenario occurrence, the weighted comprehensive evaluation value of the distribution network vulnerability is obtained.

[0014] Preferably, the improved intelligent optimization algorithm solution process also includes algorithm convergence optimization, including: During the iteration process, the changing trend of the particle swarm fitness value is monitored in real time, and the difference in fitness value between adjacent iterations is calculated. A convergence judgment threshold is set. If the difference in fitness value within a preset number of consecutive iterations is less than the convergence judgment threshold, the algorithm is determined to have converged, the iteration is terminated in advance, and the current optimal solution is output. If convergence is not achieved after the maximum number of iterations, the particle update speed is optimized by dynamically adjusting the learning factor to improve the algorithm's local search capability until the convergence condition is met and the maximum number of iterations is reached, thus ensuring the solution efficiency and the quality of the optimal solution.

[0015] Preferably, the method further includes scheduling effect verification to ensure the effectiveness of the optimal scheduling strategy, including: Set up a baseline scenario and an optimized scenario, and use scenario simulation technology to reproduce various potential disturbance scenarios. Calculate the core indicators of the distribution network vulnerability comprehensive evaluation value, fault occurrence rate and load loss rate under the two scenarios respectively, and use statistical hypothesis testing methods to verify the significance of the differences between the two types of scenario indicators. If the core metrics in the optimized scenario are significantly better than those in the baseline scenario, the scheduling strategy is deemed effective; if the expected results are not achieved, the target weights of the optimized scheduling model are adjusted, and the optimized scheduling strategy is obtained again.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention uses big data analysis and machine learning technology to acquire real-time data on the operating status and load demand of each node in the distribution network. It uses a multi-objective optimization algorithm for resource scheduling and vulnerability assessment, which effectively solves the problems of inaccurate scheduling and lagging vulnerability assessment in traditional distribution networks under high load and complex environments, and realizes dynamic optimization and efficient operation of the distribution network. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the steps of the present invention. Detailed Implementation

[0018] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0019] Reference Figure 1 As shown, optimization methods for reducing the vulnerability of distribution networks based on demand-side resources include: The hierarchical classification method is used to sort out the demand-side resources within the distribution network coverage area, collect key resource parameters, construct a demand-side resource pool, establish a resource adjustability model, and output resource adjustability data. Based on the output resource adjustability data, combined with the distribution network topology, operating conditions and potential disturbance scenarios, a vulnerability assessment index system is constructed and the index weights are calculated. The comprehensive evaluation value of the distribution network vulnerability is output through the fuzzy comprehensive evaluation method. Based on the comprehensive vulnerability assessment value, and with demand-side resources and the safe operation of the distribution network as constraints, an optimized scheduling model is constructed using multi-objective optimization modeling technology, and the output scheduling model with constraints is generated. Based on the scheduling model with constraints, an improved intelligent optimization algorithm is used to solve the problem. After adjustment by feasibility verification technology, the optimal scheduling strategy for demand-side resources is output.

[0020] A hierarchical classification method is used to analyze demand-side resources within the distribution network coverage area, collect key resource parameters, construct a demand-side resource pool, establish a resource adjustability model, and output resource adjustability data, including: The demand-side resources are divided into four core types using a hierarchical classification method: distributed power sources, energy storage devices, flexible loads, and adjustable charging piles. Key parameters of each type of resource are obtained by combining real-time sensor data collection with ledger statistics. The parameters of distributed power sources include output characteristics and fluctuation range; the parameters of energy storage devices include charging and discharging capacity, efficiency, and SOC threshold; the parameters of flexible loads include transferable capacity and response time; and the parameters of adjustable charging piles include charging power and user demand elasticity. Based on resource types and key parameters, a demand-side resource pool is constructed, and a resource information ledger is established to record resource location, capacity, schedulable time periods, and technical constraints. Differentiated adjustable capability models are established for different types of resources. Distributed power sources use output probability models to describe uncertainties, energy storage devices use charge and discharge state equations to characterize adjustment characteristics, flexible loads use load transfer matrices to quantify adjustment potential, and adjustable charging piles use user response coefficients to reflect adjustment intentions. Historical operational data is used to verify the error of various resource adjustability models. The least squares method is used to correct the model parameters. The error of the resource adjustability data output by the model is within the preset threshold. Finally, the verified resource adjustability data is output. The adjustable capacity of resources is reflected by establishing differentiated adjustable capacity models for different resource types. For example, distributed power generation uses an output probability model to describe uncertainty, while energy storage devices use charge-discharge state equations to characterize adjustment characteristics. The model is verified for error using historical operating data, and the least squares method is used to correct the model parameters to ensure that the output data error is within a preset threshold.

[0021] Based on the output resource adjustability data, combined with the distribution network topology, operating conditions, and potential disturbance scenarios, a vulnerability assessment index system is constructed and index weights are calculated. The comprehensive vulnerability assessment value of the distribution network is then output using the fuzzy comprehensive evaluation method, including: Based on verified resource adjustability data, topology parsing technology is used to extract the topology information of distribution network node connection relationships and line parameters, and operating condition data is collected. The scenario enumeration method is used to set four potential disturbance scenarios: extreme weather, equipment failure, load change, and new energy output fluctuation. A vulnerability assessment index system is constructed from three dimensions: structural vulnerability, operational vulnerability, and recovery vulnerability. The structural vulnerability index includes node degree centrality, line betweenness, and network connectivity; the operational vulnerability index includes node voltage deviation rate, line load rate, and power deficit rate; and the recovery vulnerability index includes fault recovery time, load recovery rate, and resource scheduling response speed. The weights of each evaluation index are calculated using the entropy-weighted analytic hierarchy process (AHP). Expert subjective weights are determined through AHP, and objective data weights are calculated based on information entropy theory. These are then integrated according to a preset ratio to obtain the comprehensive weight. Resource adjustability data, topology information, operating condition data, and scenario information are substituted into the index system. The fuzzy comprehensive evaluation method is used to calculate the distribution network vulnerability evaluation value under each scenario. The value is then weighted and summed based on the scenario occurrence probability to output the comprehensive distribution network vulnerability evaluation value. The formula for calculating the comprehensive distribution network vulnerability evaluation value is as follows:

[0022] In the formula, For comprehensive evaluation, As an indicator The weight, As an indicator membership degree The number of samples; The topological structure information of the distribution network node connection relationships and line parameters is extracted by topology analysis technology. Simultaneously, operational data is collected, and potential disturbance scenarios are defined. A vulnerability assessment index system is constructed, including structural, operational, and recovery vulnerability dimensions. The fuzzy comprehensive evaluation method is used for calculation, ultimately outputting a comprehensive vulnerability assessment value for the distribution network.

[0023] Based on the comprehensive vulnerability assessment value, and constrained by demand-side resources and the safe operation of the distribution network, an optimized scheduling model is constructed using multi-objective optimization modeling techniques. The output scheduling model with constraints includes: A multi-objective optimization objective function is constructed, with the core objective being to minimize the comprehensive evaluation value of the distribution network vulnerability. The auxiliary objectives include minimizing the demand-side resource scheduling cost and maximizing user electricity satisfaction. The weighting coefficient method is used to transform the multi-objective into a single objective. Set demand-side resource constraints: distributed power output constraints, energy storage device charging and discharging power and SOC constraints, flexible load regulation capacity and time constraints, adjustable charging pile charging power and user demand constraints. Set the following constraints for safe operation of the distribution network: node voltage upper and lower limits, line transmission power constraints, power balance constraints, and frequency stability constraints. Uncertainty constraints are introduced into the scenario, and a robust optimization method is used to handle the uncertainty of distributed power output and load demand. An uncertainty budget is set to represent the degree of disturbance. Based on the linearization method, nonlinear constraints are transformed into linear constraints, forming a solvable mixed-integer linear programming model; The optimization objective function includes: the core objective is to minimize the comprehensive evaluation value of the distribution network vulnerability, while the auxiliary objectives include minimizing resource scheduling costs and maximizing user satisfaction. The multi-objective is transformed into a single objective through the weighting coefficient method, and demand-side resource constraints and distribution network safety operation constraints are introduced. Scenario uncertainty constraints are handled through robust optimization methods to cope with the uncertainty of distributed power output and load demand.

[0024] Based on a scheduling model with constraints, an improved intelligent optimization algorithm is used to solve the problem. After adjustment using feasibility verification techniques, the optimal scheduling strategy for demand-side resources is output, including: Particle swarm optimization algorithm was selected as the basic algorithm, and it was improved by combining crossover and mutation operations of genetic algorithm to construct an improved particle swarm optimization algorithm. The population size, maximum number of iterations, learning factor, crossover probability and mutation probability parameters were determined by trial and error. The demand-side resource scheduling scheme is encoded as individual particles, and the optimization objective function value is used as the particle fitness value. After initializing the population, the fitness value of each particle is calculated, the particle velocity and position are updated according to preset rules, and crossover and mutation operations are performed to increase population diversity. Determine whether the iteration termination condition is met. If it is met, output the scheduling scheme corresponding to the current optimal particle. Otherwise, continue iterating. Use constraint verification technology to verify whether the optimal scheduling scheme meets all the constraints set in step 3. If there is a constraint violation, use the local adjustment method to correct the violation parameters until the scheme meets all constraints. Finally, output the optimal scheduling strategy for demand-side resources. An improved particle swarm optimization algorithm is used, which is combined with a genetic algorithm to enhance its global search capability. Appropriate algorithm parameters are determined through trial and error, and population diversity is increased through crossover and mutation operations. The optimized scheduling strategy is adjusted through feasibility verification techniques to ensure that all constraints are met.

[0025] Based on the output demand-side resource optimal scheduling strategy, combined with real-time monitoring data of the distribution network, the scheduling strategy is dynamically corrected using deviation analysis and rolling optimization techniques, and the dynamically corrected optimal scheduling strategy is output, including: Establish a real-time monitoring platform for the power distribution network, collect power distribution network operation parameters, demand-side resource status parameters, and environmental parameters, with a data sampling frequency of no less than 5 minutes / time; calculate the deviation between the real-time monitoring data and the input parameters of the optimized scheduling model, set a deviation threshold, and trigger a strategy correction mechanism if the deviation exceeds the preset threshold; The initial parameters of the scheduling model are updated and optimized based on real-time deviation data. The model is then solved again using the rolling optimization method to obtain the preliminary revised scheduling strategy. The constraint satisfaction of the revised strategy is verified again by the feasibility verification technique. The parts that do not meet the constraints are adjusted a second time, and the optimal scheduling strategy after dynamic revision is output. Based on the real-time monitoring platform of the distribution network, the status data of the distribution network and demand-side resources are collected. If the deviation between the real-time monitoring data and the input parameters of the optimization model exceeds the preset threshold, the strategy correction mechanism will be triggered. The scheduling strategy will be dynamically adjusted using rolling optimization technology, and the feasibility will be verified to ensure that the corrected scheduling strategy meets all constraints.

[0026] Before constructing the vulnerability assessment index system, the method also includes the identification of key nodes in the distribution network to provide a basis for setting index weights, including: Based on the distribution network topology data, a node importance assessment method is adopted, which combines the node load scale, power supply range, and connection relationship with other nodes to construct a node importance calculation model and calculate the importance value of each node. Set an importance threshold, mark nodes with importance values ​​higher than the threshold as key nodes, focus on statistically analyzing the resource distribution of key nodes, and output a list of key nodes and surrounding resource distribution data; substitute the relevant information of key nodes into the indicator weight calculation process, set weight tilt coefficients for the evaluation indicators corresponding to key nodes, and increase the weight ratio of indicators related to key nodes. Before constructing a vulnerability assessment index system, the key nodes of the distribution network are first identified through a node importance assessment method. These key nodes are assessed based on load size, power supply range, and their connection relationship with other nodes. The assessment indexes of key nodes are given higher weights to ensure that the operating status of these nodes is given priority consideration during the optimization process.

[0027] The calculation of the distribution network vulnerability evaluation value under each scenario using the fuzzy comprehensive evaluation method includes: A vulnerability assessment index system is constructed and index weights are calculated. The actual values ​​of each assessment index are standardized to eliminate the differences in index dimensions and obtain standardized index values. A fuzzy evaluation matrix is ​​constructed to determine the fuzzy membership function of each index. The standardized index values ​​are substituted into the membership function to calculate the membership degree of each index to different vulnerability levels. The calculated comprehensive weights of the indicators are multiplied by the fuzzy evaluation matrix to obtain the fuzzy comprehensive evaluation value of the distribution network vulnerability under each scenario. The fuzzy comprehensive evaluation value is mapped to a clear vulnerability level using the maximum membership principle. Combined with the probability of scenario occurrence, the weighted comprehensive evaluation value of the distribution network vulnerability is obtained. The evaluation index values ​​are processed using a standardized method to eliminate differences in different dimensions. The membership degree of each index to different vulnerability levels is calculated by constructing a fuzzy evaluation matrix. Finally, based on the principle of maximum membership degree, the fuzzy comprehensive evaluation value is mapped to a specific vulnerability level.

[0028] The process of improving intelligent optimization algorithms also includes optimizing algorithm convergence, including: During the iteration process, the changing trend of the particle swarm fitness value is monitored in real time, and the difference in fitness value between adjacent iterations is calculated. A convergence judgment threshold is set. If the difference in fitness value within a preset number of consecutive iterations is less than the convergence judgment threshold, the algorithm is determined to have converged, the iteration is terminated in advance, and the current optimal solution is output. If convergence is not achieved after the maximum number of iterations, the particle update speed is optimized by dynamically adjusting the learning factor to improve the local search capability of the algorithm until the convergence condition is met and the maximum number of iterations is reached, thus ensuring the solution efficiency and the quality of the optimal solution. During the solution process of the particle swarm optimization algorithm, the changing trend of the particle swarm fitness value is monitored in real time. If the difference in fitness value in several consecutive iterations is less than the set convergence threshold, the iteration is terminated in advance and the current optimal solution is output. If convergence is not achieved after reaching the maximum number of iterations, the algorithm will dynamically adjust the learning factor to improve the local search capability until the convergence condition is met.

[0029] It also includes scheduling effect verification to ensure the effectiveness of the optimal scheduling strategy, including: Set up a baseline scenario and an optimized scenario, and use scenario simulation technology to reproduce various potential disturbance scenarios. Calculate the core indicators of the distribution network vulnerability comprehensive evaluation value, fault occurrence rate and load loss rate under the two scenarios respectively, and use statistical hypothesis testing methods to verify the significance of the differences between the two types of scenario indicators. If the core metrics in the optimized scenario are significantly better than those in the baseline scenario, the scheduling strategy is deemed effective; if the expected results are not achieved, the target weights of the optimized scheduling model are adjusted, and the optimized scheduling strategy is obtained by solving the problem again. To ensure the effectiveness of the optimized scheduling strategy, scenario simulation technology is used to reproduce potential disturbance scenarios, and indicators such as vulnerability comprehensive evaluation value, failure rate, and load loss rate are calculated under the baseline scenario and the optimized scenario. The differences between the indicators under the two scenarios are verified by statistical hypothesis testing. If the indicators under the optimized scenario are better than those under the baseline scenario, the optimized scheduling strategy is determined to be effective; otherwise, the optimization target weights are readjusted and the solution is performed again.

[0030] In summary, the advantages of this invention are: By constructing a demand-side resource pool and combining it with the topology, operating conditions, and potential disturbance scenarios of the distribution network, the vulnerability of the distribution network is comprehensively assessed, providing a basis for subsequent optimized scheduling. The fuzzy comprehensive evaluation method is used to comprehensively assess the vulnerability of the distribution network, helping to quickly identify potential vulnerable links and risk points. Based on multi-objective optimization modeling technology, considering the schedulability of demand-side resources and the constraints of safe operation of the distribution network, an optimized scheduling model is constructed. By improving the intelligent optimization algorithm, it can not only effectively solve the problems of low solution efficiency and difficulty in meeting constraints in traditional scheduling methods, but also improve the accuracy and practicality of the optimized scheduling scheme.

[0031] By using real-time monitoring data of the distribution network and employing deviation analysis and rolling optimization techniques for dynamic correction, the scheduling strategy is continuously adjusted to ensure the continuous and stable operation of the distribution network under complex disturbance scenarios. This dynamic optimization method can quickly respond to emergencies, ensure stable power supply to the distribution network, and reduce the probability of faults. Based on the characteristics of different types of demand-side resources, establish differentiated adjustable capacity models, rationally allocate various resources, improve resource utilization, and minimize scheduling costs and improve user satisfaction with electricity use.

[0032] By employing fuzzy comprehensive evaluation, weighting coefficient method, and multi-objective optimization techniques, a complete decision support system is formed to help power operators better formulate distribution network dispatch strategies, thereby ensuring the safe and efficient operation of the distribution network under different environments. By assessing resilience indicators and considering factors such as fault recovery time, load recovery rate, and resource dispatch response speed, the normal operation of the distribution network can be restored quickly after a fault occurs, thereby reducing power outage duration and economic losses.

[0033] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. An optimization method for reducing the vulnerability of distribution networks based on demand-side resources, characterized in that, include: The hierarchical classification method is used to sort out the demand-side resources within the distribution network coverage area, collect key resource parameters, construct a demand-side resource pool, establish a resource adjustability model, and output resource adjustability data. Based on the output resource adjustability data, combined with the distribution network topology, operating conditions and potential disturbance scenarios, a vulnerability assessment index system is constructed and the index weights are calculated. The comprehensive evaluation value of the distribution network vulnerability is output through the fuzzy comprehensive evaluation method. Based on the comprehensive vulnerability assessment value, and with demand-side resources and the safe operation of the distribution network as constraints, an optimized scheduling model is constructed using multi-objective optimization modeling technology, and the output scheduling model with constraints is generated. Based on the scheduling model with constraints, an improved intelligent optimization algorithm is used to solve the problem. After adjustment by feasibility verification technology, the optimal scheduling strategy for demand-side resources is output.

2. The optimization method for reducing the vulnerability of distribution networks based on demand-side resources according to claim 1, characterized in that, A hierarchical classification method is used to analyze demand-side resources within the distribution network coverage area, collect key resource parameters, construct a demand-side resource pool, establish a resource adjustability model, and output resource adjustability data, including: The demand-side resources are divided into four core types using a hierarchical classification method: distributed power sources, energy storage devices, flexible loads, and adjustable charging piles. Key parameters of each type of resource are obtained by combining real-time sensor data collection with ledger statistics. The parameters of distributed power sources include output characteristics and fluctuation range; the parameters of energy storage devices include charging and discharging capacity, efficiency, and SOC threshold; the parameters of flexible loads include transferable capacity and response time; and the parameters of adjustable charging piles include charging power and user demand elasticity. Based on resource types and key parameters, a demand-side resource pool is constructed, and a resource information ledger is established to record resource location, capacity, schedulable time periods, and technical constraints. Differentiated adjustable capability models are established for different types of resources. Distributed power sources use output probability models to describe uncertainties, energy storage devices use charge and discharge state equations to characterize adjustment characteristics, flexible loads use load transfer matrices to quantify adjustment potential, and adjustable charging piles use user response coefficients to reflect adjustment intentions. Historical operational data is used to verify the error of various resource adjustability models. The least squares method is used to correct the model parameters. The error of the resource adjustability data output by the model is within the preset threshold. Finally, the verified resource adjustability data is output.

3. The optimization method for reducing the vulnerability of distribution networks based on demand-side resources according to claim 2, characterized in that, Based on the output resource adjustability data, combined with the distribution network topology, operating conditions, and potential disturbance scenarios, a vulnerability assessment index system is constructed and index weights are calculated. The comprehensive vulnerability assessment value of the distribution network is then output using the fuzzy comprehensive evaluation method, including: Based on verified resource adjustability data, topology parsing technology is used to extract the topology information of distribution network node connection relationships and line parameters, and operating condition data is collected. The scenario enumeration method is used to set four potential disturbance scenarios: extreme weather, equipment failure, load change, and new energy output fluctuation. A vulnerability assessment index system is constructed from three dimensions: structural vulnerability, operational vulnerability, and recovery vulnerability. The structural vulnerability index includes node degree centrality, line betweenness, and network connectivity; the operational vulnerability index includes node voltage deviation rate, line load rate, and power deficit rate; and the recovery vulnerability index includes fault recovery time, load recovery rate, and resource scheduling response speed. The weights of each evaluation index are calculated using the entropy-weighted analytic hierarchy process (AHP). Expert subjective weights are determined through AHP, and objective data weights are calculated based on information entropy theory. These are then integrated according to a preset ratio to obtain the comprehensive weight. Resource adjustability data, topology information, operating condition data, and scenario information are substituted into the index system. The fuzzy comprehensive evaluation method is used to calculate the distribution network vulnerability evaluation value under each scenario. The value is then weighted and summed based on the scenario occurrence probability to output the comprehensive distribution network vulnerability evaluation value. The formula for calculating the comprehensive distribution network vulnerability evaluation value is as follows: ； In the formula, For comprehensive evaluation, As an indicator The weight, As an indicator membership degree This represents the number of samples.

4. The optimization method for reducing the vulnerability of distribution networks based on demand-side resources according to claim 3, characterized in that, Based on the comprehensive vulnerability assessment value, and constrained by demand-side resources and the safe operation of the distribution network, an optimized scheduling model is constructed using multi-objective optimization modeling techniques. The output scheduling model with constraints includes: A multi-objective optimization objective function is constructed, with the core objective being to minimize the comprehensive evaluation value of the distribution network vulnerability. The auxiliary objectives include minimizing the demand-side resource scheduling cost and maximizing user electricity satisfaction. The weighting coefficient method is used to transform the multi-objective into a single objective. Set demand-side resource constraints: distributed power output constraints, energy storage device charging and discharging power and SOC constraints, flexible load regulation capacity and time constraints, adjustable charging pile charging power and user demand constraints. Set the following constraints for safe operation of the distribution network: node voltage upper and lower limits, line transmission power constraints, power balance constraints, and frequency stability constraints. Uncertainty constraints are introduced into the scenario, and a robust optimization method is used to handle the uncertainty of distributed power output and load demand. An uncertainty budget is set to represent the degree of disturbance. The nonlinear constraints are transformed into linear constraints by linearization methods, forming a solvable mixed-integer linear programming model.

5. The optimization method for reducing the vulnerability of distribution networks based on demand-side resources according to claim 4, characterized in that, Based on a scheduling model with constraints, an improved intelligent optimization algorithm is used to solve the problem. After adjustment using feasibility verification techniques, the optimal scheduling strategy for demand-side resources is output, including: Particle swarm optimization algorithm was selected as the basic algorithm, and it was improved by combining crossover and mutation operations of genetic algorithm to construct an improved particle swarm optimization algorithm. The population size, maximum number of iterations, learning factor, crossover probability and mutation probability parameters were determined by trial and error. The demand-side resource scheduling scheme is encoded as individual particles, and the optimization objective function value is used as the particle fitness value. After initializing the population, the fitness value of each particle is calculated, the particle velocity and position are updated according to preset rules, and crossover and mutation operations are performed to increase population diversity. Determine if the iteration termination condition is met. If it is, output the scheduling scheme corresponding to the current optimal particle; otherwise, continue iterating. Use constraint verification technology to verify whether the optimal scheduling scheme meets all the constraints set in step 3. If there is a constraint violation, use the local adjustment method to correct the violation parameters until the scheme meets all constraints. Finally, output the optimal scheduling strategy for demand-side resources.

6. The optimization method for reducing the vulnerability of distribution networks based on demand-side resources according to claim 5, characterized in that, Also includes: Based on the optimal scheduling strategy for demand-side resources output, combined with real-time monitoring data of the distribution network, the scheduling strategy is dynamically corrected using deviation analysis and rolling optimization techniques, and the dynamically corrected optimal scheduling strategy is output. Establish a real-time monitoring platform for the power distribution network, collect power distribution network operation parameters, demand-side resource status parameters, and environmental parameters, with a data sampling frequency of no less than 5 minutes / time; calculate the deviation between the real-time monitoring data and the input parameters of the optimized scheduling model, set a deviation threshold, and trigger a strategy correction mechanism if the deviation exceeds the preset threshold; The initial parameters of the scheduling model are updated and optimized based on real-time deviation data. The model is then solved again using the rolling optimization method to obtain the preliminary revised scheduling strategy. The constraint satisfaction of the modified strategy is verified again by the feasibility verification technique. The parts that do not meet the constraints are adjusted a second time, and the optimal scheduling strategy after dynamic correction is output.

7. The optimization method for reducing the vulnerability of distribution networks based on demand-side resources according to claim 6, characterized in that, Before constructing the vulnerability assessment index system, the method also includes: Key nodes in the distribution network are identified to provide a basis for setting indicator weights. Based on the distribution network topology data, a node importance assessment method is adopted. Combining the node load scale, power supply range, and connection relationship with other nodes, a node importance calculation model is constructed to calculate the importance value of each node. Set an importance threshold, mark nodes with importance values ​​higher than the threshold as key nodes, focus on statistically analyzing the resource distribution of key nodes, and output a list of key nodes and surrounding resource distribution data; substitute the relevant information of key nodes into the indicator weight calculation process, set weight tilt coefficients for the evaluation indicators corresponding to key nodes, and increase the weight ratio of key node-related indicators.

8. The optimization method for reducing the vulnerability of distribution networks based on demand-side resources according to claim 7, characterized in that, The fuzzy comprehensive evaluation method is used to calculate the vulnerability evaluation values ​​of the distribution network under various scenarios, including: Construct a vulnerability assessment index system and calculate index weights. Standardize the actual values ​​of each assessment index extracted in step 2 to eliminate the difference in index dimensions and obtain standardized index values. Construct a fuzzy evaluation matrix, determine the fuzzy membership function of each index, substitute the standardized index values ​​into the membership function, and calculate the membership degree of each index to different vulnerability levels. The calculated comprehensive weights of the indicators are multiplied by the fuzzy evaluation matrix to obtain the fuzzy comprehensive evaluation value of the distribution network vulnerability under each scenario. The fuzzy comprehensive evaluation value is mapped to a specific vulnerability level using the maximum membership principle. Combined with the probability of scenario occurrence, the weighted comprehensive evaluation value of the distribution network vulnerability is obtained.

9. The optimization method for reducing the vulnerability of distribution networks based on demand-side resources according to claim 8, characterized in that, The process of improving the intelligent optimization algorithm also includes: During the iteration process, the changing trend of the particle swarm fitness value is monitored in real time, and the difference in fitness value between adjacent iterations is calculated. A convergence judgment threshold is set. If the difference in fitness value within a preset number of consecutive iterations is less than the convergence judgment threshold, the algorithm is determined to have converged, the iteration is terminated in advance, and the current optimal solution is output. If convergence is not achieved after the maximum number of iterations, the particle update speed is optimized by dynamically adjusting the learning factor to improve the algorithm's local search capability until the convergence condition is met and the maximum number of iterations is reached, thus ensuring the solution efficiency and the quality of the optimal solution.

10. The optimization method for reducing the vulnerability of distribution networks based on demand-side resources according to claim 9, characterized in that, The method also includes scheduling effect verification to ensure the effectiveness of the optimal scheduling strategy, including: Set up a baseline scenario and an optimized scenario, and use scenario simulation technology to reproduce various potential disturbance scenarios. Calculate the core indicators of the distribution network vulnerability comprehensive evaluation value, fault occurrence rate and load loss rate under the two scenarios respectively, and use statistical hypothesis testing methods to verify the significance of the differences between the two types of scenario indicators. If the core metrics in the optimized scenario are significantly better than those in the baseline scenario, the scheduling strategy is deemed effective; if the expected results are not achieved, the target weights of the optimized scheduling model are adjusted, and the optimized scheduling strategy is obtained again.

Images