Power distribution network adaptive overcurrent protection method and system based on dynamic multi-objective optimization algorithm

Abstract

The invention discloses a power distribution network adaptive overcurrent protection method and system based on a dynamic multi-objective optimization algorithm. The method comprises the following steps: establishing an initial parent population according to protection constant value historical data in basic parameters of a power distribution network; generating a filial population according to the parent population, and combining the filial population with the parent population to form an extended population; comparing the new energy power fluctuation amplitude with a preset threshold value to determine whether the new energy output is remarkably changed or not; comparing the fluctuation amplitude with the new energy output fluctuation ratio to determine a solution injection strategy; generating corresponding solutions to replace individuals in the population according to a solution injection strategy; and calculating a corresponding Pareto frontier and a new parent population for the population after replacement, reserving an individual with the minimum value of the comprehensive objective function in the parent population as a current optimal solution, and performing protection configuration according to a protection parameter vector of the current optimal solution after the maximum number of iterations is reached or the Pareto frontier converges. According to the method, the maloperation rate can be effectively reduced, the action time is shortened, and the reliability and anti-disturbance capability of the power distribution network are remarkably improved.

Description

Technical Field

[0001] This invention relates to distribution network protection technology, specifically to an adaptive overcurrent protection method and system for distribution networks based on a dynamic multi-objective optimization algorithm. Background Technology

[0002] With the large-scale integration of new energy sources into the distribution network, traditional overcurrent protection faces significant challenges. Specifically, traditional distribution network overcurrent protection employs a fixed-value strategy, which cannot adapt to the rapid changes in power output from new energy sources (such as wind and solar power). When new energy power surges, the protection device may fail to operate due to insufficient sensitivity; conversely, when new energy power changes gradually, the protection range may overlap due to fixed parameters, leading to false tripping. While existing dynamic optimization algorithms (such as NSGA-II) can handle multi-objective optimization problems, they suffer from drawbacks in dynamic environments, including slow convergence speed, insufficient diversity, and limited real-time performance. Summary of the Invention

[0003] The technical problem to be solved by this invention is to provide a method and system for adaptive overcurrent protection of distribution networks based on a dynamic multi-objective optimization algorithm, which dynamically detects power fluctuations of new energy sources and generates adaptive solutions to achieve online optimization of protection parameters.

[0004] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: An adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm includes the following steps: S101) Obtain basic parameters of the distribution network, including real-time power of new energy sources, load power, power output fluctuation rate of new energy sources and historical data of protection settings. Establish different protection parameter vectors based on the historical data of protection settings and use them as individuals in the initial parent population. S102) Generate offspring individuals from individuals in the parent population, and merge the parent and offspring populations into an extended population; S103) Calculate the power fluctuation amplitude of new energy sources based on the real-time power and load power of new energy sources, and compare the power fluctuation amplitude of new energy sources with the preset threshold to determine whether the output of new energy sources has changed significantly; if the output of new energy sources has changed significantly, determine the solution injection strategy for this iteration based on the comparison result of the power fluctuation amplitude of new energy sources and the power output fluctuation rate of new energy sources. S104) Generate the corresponding solution according to the solution injection strategy of this iteration and replace the individuals in the extended population; S105) Calculate the corresponding Pareto front and the new parent population for the replaced extended population, calculate the comprehensive objective function of each individual in the parent population, retain the individual with the smallest comprehensive objective function value as the current optimal solution, and jump to step S102) until the maximum number of iterations is reached or the Pareto front converges, and select the optimal solution with the smallest comprehensive objective function value from the optimal solutions of each iteration as the new current optimal solution; S106) Configure protection according to the protection parameter vector of the current optimal solution.

[0005] Furthermore, the real-time power of the new energy source includes wind power and / or photovoltaic power, and the power output volatility of the new energy source includes the standard deviation of wind power output and / or the standard deviation of photovoltaic power output. When the power output of the new energy source changes significantly, the power fluctuation amplitude of the new energy source exceeds the corresponding threshold, as shown in the following formula:

[0006] in, Indicates wind power output rate of change, photovoltaic power Rate of change and load power The maximum value of the rate of change It is the rate of change threshold, and the formula is as follows:

[0007] Where k is the adjustment coefficient. Indicates the standard deviation of wind power output Standard deviation of photovoltaic output The maximum value in.

[0008] Furthermore, when determining the solution injection strategy for this iteration based on the comparison results of the fluctuation amplitude of renewable energy power and the fluctuation rate of renewable energy output, it includes: If the relationship between the magnitude of the fluctuation in renewable energy power and the fluctuation rate of renewable energy output satisfies If the solution is positive, the random solution injection strategy will be used for this iteration; otherwise, the mutated solution strategy will be used.

[0009] Furthermore, when generating the corresponding solution and replacing individuals in the expanded population based on the solution injection strategy of this iteration, the process includes: If the solution injection strategy for this iteration is a random solution injection strategy, then select a specified proportion of individuals in the current population, generate corresponding random solutions in a uniform distribution, and then replace the worst-performing individual in the current population with the random solution. Specifically, the worst-performing individual is the one with the worst comprehensive objective function value. The formula for the random solution is as follows:

[0010] in, For the selected individuals, k is the adjustment coefficient. It is a function for generating uniformly distributed random numbers; If the solution injection strategy for this iteration is a mutation strategy, then Gaussian mutation is performed on selected individuals in the current population to obtain a mutated solution. The mutated solution then replaces the individual with the lowest crowding density in the current population. Specifically, the individual with the lowest crowding density is one surrounded by solutions with high redundancy. The formula for the mutated solution is as follows:

[0011] in , Here, t is the scaling factor, and t is the current iteration number. It is a function for generating normally distributed random numbers.

[0012] Furthermore, the formula for calculating the comprehensive objective function is as follows:

[0013] in, , , , , These are the weight coefficients for each optimization objective. It is to minimize the action time. It maximizes selectivity, and the formula is as follows:

[0014] in, Main protection operating current, For backup protection operating current, For each individual, the protection setting current value is in the protection parameter vector. It is to minimize the sensitivity coefficient, and the formula is as follows:

[0015] in, This refers to the fault current under the minimum operating conditions of the system. This is to protect reliability, and the formula is as follows:

[0016] in, To correctly remove the fault, This represents the total number of failure events. It is the minimum false alarm rate, and the formula is as follows:

[0017] in, This is the theoretically optimal value for the current.

[0018] Furthermore, before selecting the optimal solution with the smallest comprehensive objective function value from the optimal solutions of each iteration as the new current optimal solution, the process includes: checking whether the protection parameter vector of the optimal solution in each iteration satisfies the constraints, and eliminating optimal solutions whose protection parameter vectors do not satisfy the constraints, wherein the constraints include: Sensitivity constraints:

[0019] Selectivity constraints:

[0020] Error rate constraint:

[0021] in, It is to minimize the sensitivity coefficient. It is the minimum false alarm rate. Indicates the timing of the primary protection action. Indicates the timing of the backup protection action. It is the time difference in the protection parameter vector, used to ensure that the main protection takes priority and the backup protection only operates when the main protection fails.

[0022] Furthermore, step S103 also includes: if the output of new energy sources does not change significantly, then jump to step S106 to configure protection according to the protection parameter vector of the current optimal solution.

[0023] The present invention also proposes an adaptive overcurrent protection system for distribution networks based on a dynamic multi-objective optimization algorithm, comprising a processor and a computer-readable storage medium interconnected with each other. The computer-readable storage medium stores a computer program, which is executed by the processor to implement the steps of the adaptive overcurrent protection method for distribution networks based on the dynamic multi-objective optimization algorithm.

[0024] The present invention also proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm.

[0025] The present invention also proposes a computer program product, including a computer program, which, when executed by a processor, implements the steps of the adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm.

[0026] Compared with the prior art, the advantages of the present invention are as follows: This invention uses the rate of change of renewable energy power as the core criterion. By monitoring the fluctuations in wind power and photovoltaic output and load changes in real time, it dynamically triggers overcurrent protection. In each iteration, different solution injection strategies are dynamically selected based on the amplitude of renewable energy power fluctuations to optimize the population. Then, the Pareto front is calculated for the population, thereby dynamically adjusting the overcurrent protection setting and delay parameters in the protection parameter vector to ensure that the setting always matches the real-time grid state. By embedding hard constraints such as sensitivity, selectivity, and false trip rate into the comprehensive objective function, it strictly ensures that the optimization results meet the requirements of the relay protection regulations. At the same time, the continuous convergence of the Pareto front rate of change is used as the termination condition, which reduces computational redundancy and avoids local optimum traps. Attached Figure Description

[0027] Figure 1 This is a flowchart of a method according to an embodiment of the present invention.

[0028] Figure 2 This is a schematic diagram of a typical 10kV distribution network connected to a wind farm. Detailed Implementation

[0029] The present invention will be further described below with reference to the accompanying drawings and specific preferred embodiments, but this does not limit the scope of protection of the present invention.

[0030] Example 1 This embodiment proposes an adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm. It solves the problems of false tripping and failure to tripping of traditional fixed-value protection under dynamic access of new energy sources. The optimized protection system can effectively reduce the false tripping rate, shorten the action time, and significantly improve the reliability and anti-disturbance capability of the distribution network.

[0031] like Figure 1 As shown, the method includes the following steps: S101) Obtain basic parameters of the distribution network, including real-time power of new energy sources, load power, power output fluctuation rate of new energy sources and historical data of protection settings. Establish different protection parameter vectors based on the historical data of protection settings and use them as individuals in the initial parent population. S102) Generate offspring individuals from individuals in the parent population, and merge the parent and offspring populations into an extended population; S103) Calculate the power fluctuation amplitude of new energy sources based on real-time power and load power, and compare the power fluctuation amplitude with a preset threshold to determine whether the power output of new energy sources has changed significantly. If the power output of new energy sources has changed significantly, determine the solution injection strategy for this iteration based on the comparison result of the power fluctuation amplitude and the power output fluctuation rate of new energy sources. The solution injection strategy includes a random solution injection strategy (DNSGA-II-A) or a Gaussian mutation strategy (DNSGA-II-B). For scenarios with drastic fluctuations in new energy sources, DNSGA-II-A rapidly expands the search range through uniformly distributed random perturbations. For scenarios with gradual changes, DNSGA-II-B uses adaptive Gaussian mutation to achieve fine parameter tuning. S104) Generate the corresponding solution according to the solution injection strategy of this iteration and replace the individuals in the extended population; S105) Calculate the corresponding Pareto front and the new parent population for the replaced extended population, calculate the comprehensive objective function of each individual in the parent population, retain the individual with the smallest comprehensive objective function value as the current optimal solution, thereby realizing online optimization of the protection parameters, and jump to step S102) until the maximum number of iterations is reached or the Pareto front converges, and select the optimal solution with the smallest comprehensive objective function value from the optimal solutions of each iteration as the new current optimal solution; S106) Configure protection according to the protection parameter vector of the current optimal solution.

[0032] Through the above steps, the method in this embodiment dynamically triggers the improved NSGA-II optimization algorithm by real-time monitoring of wind power and photovoltaic output fluctuations and load changes. Based on the magnitude of new energy power fluctuations, it selects either a random solution injection strategy (DNSGA-II-A) or a Gaussian mutation strategy (DNSGA-II-B) to optimize overcurrent protection settings and timing parameters online. A comprehensive objective function is introduced to integrate multi-objective conflict optimization considering action time, selectivity, reliability, sensitivity coefficient, and false trip rate. Combined with hard constraints on sensitivity and false trip rate, this ensures that the protection parameters comply with relay protection regulations.

[0033] The relevant steps are explained in detail below.

[0034] In this embodiment, the population is initialized in step S101. The basic parameters in the distribution network include: real-time power of new energy sources and load power. New energy power output volatility, line impedance Transformer turns ratio n, historical data of protection settings (overcurrent stage I setting) Section II fixed value ),in: Real-time power of new energy sources includes (Wind power) and / or (Photovoltaic power), if the new energy source is wind power, then select the appropriate parameters. If the new energy source is photovoltaic power generation, then select the appropriate parameters. If both are included, then choose the one that best represents the desired outcome. and ; New energy output volatility includes the standard deviation of wind power output. and / or standard deviation of photovoltaic output If the new energy source is wind power, then select the appropriate parameters. If the new energy source is photovoltaic power generation, then select the appropriate parameters. If both are included, then choose the one that best represents the desired outcome. and .

[0035] Thus, in the initial parent population, each individual's encoding includes a pair of overcurrent I-segment settings based on the historical data of protection settings. With Section II fixed value The constructed protection parameter vector represents a candidate solution, in the form of: ,in For time delay, For time limit increments, time delay and time limit difference Based on the overcurrent protection principle in relay protection, its settings must meet the three-stage overcurrent protection stepped time delay characteristics. Furthermore, the individual codes in the initial parent population also include the real-time power of new energy sources (wind power) corresponding to the historical data of the protection settings. and / or photovoltaic power ), load power and the volatility of new energy output (standard deviation of wind power output) and / or standard deviation of photovoltaic output ).

[0036] In this embodiment, step S102 specifically follows the NSGA-II standard procedure (crossover, mutation). Crossover is performed on the selected parent individuals to generate a offspring population Q. The offspring individuals are then perturbed, and the parent population P and the offspring population Q are merged into an extended population. R The specific implementation process is well known to those skilled in the art and is not the focus of this application, so it will not be described in detail here.

[0037] In this embodiment, the fluctuation range of renewable energy power is detected in step S103. When the fluctuation rate of renewable energy power or load exceeds a preset threshold, the optimization condition is triggered. After the optimization condition is triggered, the optimization strategy is selected according to the fluctuation range of renewable energy power.

[0038] Specifically, when there is a significant change in the output of new energy sources, the fluctuation amplitude of the new energy power exceeds the corresponding threshold, as shown in the following formula:

[0039] in, It is the objective function for the fluctuation range of new energy power, representing wind power. rate of change Photovoltaic power rate of change With load power rate of change The maximum value in the middle. This is the rate of change threshold, which is correlated with the standard deviation of new energy output, using the following formula:

[0040] Where k is the adjustment coefficient. Indicates the standard deviation of wind power output Standard deviation of photovoltaic output The maximum value in.

[0041] In this embodiment, since the individual encoding includes the wind power required to calculate the objective function and the threshold, Photovoltaic power Load power Standard deviation of wind power output and the standard deviation of photovoltaic output Since parameters are such that the objective function and threshold can be directly calculated from the parameters contained in the extracted individuals, the specific operation to determine whether the adaptive overcurrent protection setting optimization trigger condition is met is as follows: at the beginning of each iteration, 10% of the parent individuals are randomly selected, and their objective function is recalculated. and threshold If the fluctuation range of any individual's renewable energy power or load exceeds the threshold, it is determined that the renewable energy output has changed significantly.

[0042] In step S103 of this embodiment, when determining the de-injection strategy for this iteration based on the comparison result of the new energy power fluctuation amplitude and the new energy output fluctuation rate, specifically, the applicable scenarios of the two strategies included in the adaptive overcurrent protection are judged based on the changes in the new energy output state, including: If the relationship between the magnitude of the fluctuation in renewable energy power and the fluctuation rate of renewable energy output satisfies If the output of new energy sources fluctuates drastically or the grid topology changes rapidly, the random solution injection strategy DNSGA-II-A will be adopted for this iteration. Otherwise, it indicates that the state of new energy sources in the distribution network changes slowly, and the mutation solution strategy DNSGA-II-B will be adopted for this iteration.

[0043] In this embodiment, It compares the wind power of the individual samples. rate of change With photovoltaic power rate of change The maximum value is obtained. It is based on the standard deviation of wind power output of the selected individuals. The calculated wind power is the power of any individual unit. rate of change Or photovoltaic power rate of change Exceeding the standard deviation of wind power output for that individual If the value is twice that of the target value, the random solution injection strategy DNSGA-II-A is adopted; otherwise, the mutation solution strategy DNSGA-II-B is adopted.

[0044] In this embodiment, when the selected strategy is executed in step S104, and the corresponding solution is generated and individuals in the expanded population are replaced according to the solution injection strategy of this iteration, the following steps are included: If the solution injection strategy for this iteration is the random solution injection strategy DNSGA-II-A, then a specified proportion of individuals in the current population are selected, and corresponding random solutions are generated in a uniform distribution. Then, the worst-performing individual in the current population is replaced by a random solution. Specifically, the worst-performing individual is the one with the worst comprehensive objective function value. The formula for the random solution is as follows:

[0045] in, For the selected individuals, k is the adjustment coefficient. It is a function for generating uniformly distributed random numbers; If the solution injection strategy for this iteration is the mutation solution strategy DNSGA-II-B, then Gaussian mutation is performed on selected individuals in the current population to obtain the mutated solution. Then, the mutated solution replaces the individual with the lowest crowding density in the current population. Specifically, the individual with the lowest crowding density is the one surrounded by solutions that are densely packed and have high redundancy. The formula for the mutated solution is as follows:

[0046] in , Here, t is the scaling factor, and t is the current iteration number. It is a function for generating normally distributed random numbers.

[0047] by Figure 2 For example, Figure 2In the diagram, 101, 201, 202, 301, 302, 401, 501, and 601 are the installation points for the head-end protection of each line. The calculation steps for handling sudden changes in wind power output using DNSGA-II-A (random solution injection strategy) are as follows: Step 1: The scenario is set as a 10kV distribution network connected to a wind farm, where a sudden increase in initial wind speed leads to a decrease in wind power output. The volatility jumped from 1MW to 3MW in 5 seconds. (Standard deviation); Step 2: Set initial parameters: for a certain individual in the current population , The adjustment coefficient k = 0.5 (set based on experience), and the replacement ratio is set to 50%. Step 3: Generate random solutions: Calculate the perturbation range That is, the range of disturbance

[0048] Step 4: Generate new fixed values Assuming uniformly distributed random numbers ,but: , , Step 5: Replace the worst individual with the objective function value F in the current population with the new solution described above.

[0049] After the above steps, the optimized output parameters are as follows: , , ,

[0050] Compared with the original protection setting, the optimized setting is increased by 10%, which effectively avoids false tripping due to the short circuit current exceeding the original setting caused by a sudden increase in wind power output.

[0051] exist Figure 2 In the typical 10kV distribution network connected to a wind farm scenario shown, assuming changes in wind power output, the calculation steps for handling sudden changes in wind power output using DNSGA-II-B (mutation solution strategy) are as follows: Step 1: The scenario is set as a 10kV distribution network connected to a wind farm, with wind power output changing gradually, i.e. Volatility The number of iterations is t=10; Step 2: Set initial parameters: for a certain individual in the current population , scaling factor ; Step 3: Perform Gaussian mutation: Calculate the mutation strength ; Step 4: Generate new fixed values Assuming normally distributed random numbers ,but

[0052]

[0053] Step 5: Replace the individual with the lowest crowding in the current population (surrounded by dense solutions, high redundancy) with the new solution mentioned above.

[0054] After the above steps, the optimized output parameters are as follows: , , ,

[0055] Compared to the original protection settings, the optimized settings are slightly adjusted by 0.5%, which is suitable for gradually changing scenarios and can reduce the impact on the original protection logic.

[0056] The above calculations show that DNSGA-II-A expands the search range through uniformly distributed random perturbations, quickly adapting to severe fluctuations and avoiding false alarms; while DNSGA-II-B finely adjusts the setpoints through Gaussian variation, balancing protection performance and stability. The cooperation between these two strategies ensures that distribution networks with renewable energy access can achieve dynamic optimization under different fluctuation intensities, improving the robustness of the protection system.

[0057] In step S105 of this embodiment, when calculating the corresponding Pareto front and the new parent population for the replaced extended population, the individuals in the extended population are sequentially subjected to non-dominated sorting, crowding calculation, and elite retention to obtain the corresponding Pareto front and the new generation of parent population. The specific implementation process is well known to those skilled in the art and is not the focus of this application; therefore, it will not be described in detail here.

[0058] In step S105 of this embodiment, the comprehensive objective function quantifies multiple conflicting optimization objectives into a comparable index. In this embodiment, the optimization objective in distribution network overcurrent protection is set as: minimizing the action time. Maximize selectivity Minimize the sensitivity coefficient To ensure reliability R and minimize the false alarm rate E, where: Minimize action time It includes the fault current threshold trigger time, protection algorithm calculation time, and circuit breaker tripping time, which are specifically determined by hardware response time and algorithm delay, and need to be measured in engineering. Maximize selectivity The calculation formula is as follows:

[0059] in, Main protection operating current, For backup protection operating current, In this embodiment, the protection setting current value is the protection parameter vector for an individual device. To protect the overcurrent I-section setpoint in the parameter vector With Section II fixed value The protection setting current value, i.e. ; Minimize sensitivity coefficient The calculation formula is as follows:

[0060] in, This refers to the fault current under the minimum operating conditions of the system. Protect reliability The calculation formula is as follows:

[0061] in, To correctly remove the fault, This represents the total number of failure events. Minimum error rate The calculation formula is as follows:

[0062] in, This is the theoretically optimal value for the current.

[0063] The formula for calculating the comprehensive objective function is:

[0064] in, , , , , These are the weight coefficients for each optimization objective, determined based on expert experience, and satisfying the following conditions: .

[0065] In this embodiment, the maximum number of iterations is set to 100, and the Pareto front convergence condition is that the rate of change of the Pareto front is less than 1% for five consecutive generations. After the iteration is terminated, it is checked whether the protection parameter vector of the optimal solution in each iteration satisfies the constraint condition, and the optimal solution whose protection parameter vector does not satisfy the constraint condition is eliminated.

[0066] Specifically, the constraints include: Sensitivity constraints:

[0067] Selectivity constraints:

[0068] Error rate constraint:

[0069] in, It is to minimize the sensitivity coefficient. It is the minimum false alarm rate. Indicates the timing of the action of the superior protection (main protection). Indicates the action time of the lower-level protection (backup protection). It is the time difference in the protection parameter vector, used to ensure that the main protection takes priority and the backup protection only operates when the main protection fails.

[0070] For the optimal solution that satisfies the constraints, the set of protection parameters (overcurrent protection operating current threshold) of the optimal solution with the smallest comprehensive objective function value is selected. Time parameters The output is sent to the protection device to ensure that it can adapt to the changes in short-circuit current caused by the fluctuation of new energy sources in real time, thereby improving the reliability, selectivity and speed of distribution network protection.

[0071] In this embodiment, step S103 also includes a processing step when the output of new energy sources does not change significantly. Specifically, if the output of new energy sources does not change significantly, it means that the current protection parameters are already the optimal solution required by the distribution network environment. The current optimal solution can be directly output for protection configuration, and then the process jumps to step S106 to perform protection configuration according to the protection parameter vector of the current optimal solution.

[0072] Example 2 This embodiment proposes an adaptive overcurrent protection system for distribution networks based on a dynamic multi-objective optimization algorithm, including a processor and a computer-readable storage medium connected to each other. The computer-readable storage medium stores a computer program, which is executed by the processor to implement the steps of the adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm described in Embodiment 1.

[0073] This embodiment also proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm described in Embodiment 1.

[0074] This embodiment also proposes a computer program product, including a computer program that, when executed by a processor, implements the steps of the adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm as described in Embodiment 1.

[0075] In summary, this invention addresses the shortcomings of traditional overcurrent protection in renewable energy integration scenarios. Through the deep integration of a dynamic triggering mechanism and a multi-objective optimization algorithm, it achieves adaptive real-time adjustment of protection settings, significantly improving the reliability and disturbance rejection capability of the distribution network. Traditional methods use fixed settings, which are difficult to adapt to the dynamic changes in short-circuit current caused by fluctuations in renewable energy output, easily leading to false trips or failures to trip. Furthermore, single-objective optimization and empirical parameter adjustments lack a systematic balance for multiple conflicting objectives such as speed, selectivity, and sensitivity. This invention uses the renewable energy power change rate as the core criterion, combined with the improved DNSGA-II algorithm to inject random solutions in different scenarios (DNSGA-II-A) or perform Gaussian mutation (DNSGA-II-B), dynamically adjusting the overcurrent protection settings and delay parameters to ensure that the settings always match the real-time grid state. By embedding hard constraints such as sensitivity, selectivity, and false trip rate, it strictly ensures that the optimization results comply with the requirements of relay protection regulations. Simultaneously, it uses continuous convergence of the Pareto front rate of change as an intelligent termination condition, reducing computational redundancy while avoiding local optimum traps. The final output set of protection parameters can quickly suppress the risk of false tripping through a wide range of searches when the wind power output changes abruptly, and can maintain the accuracy of action through fine parameter adjustment in the case of slow output change. In practical applications, it can reduce the false tripping rate by more than 30% and shorten the action time by 20%, providing a smart protection solution for high-proportion new energy distribution networks that takes into account dynamic adaptability, engineering compliance and optimized efficiency.

[0076] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the process. Figure 1One or more processes and / or boxes Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0077] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for adaptive overcurrent protection of distribution networks based on a dynamic multi-objective optimization algorithm, characterized in that, Includes the following steps: S101) Obtain basic parameters of the distribution network, including real-time power of new energy sources, load power, power output fluctuation rate of new energy sources and historical data of protection settings. Establish different protection parameter vectors based on the historical data of protection settings and use them as individuals in the initial parent population. S102) Generate offspring individuals based on individuals in the parent population, and merge the parent and offspring populations into an extended population; S103) Calculate the power fluctuation amplitude of new energy sources based on the real-time power and load power of new energy sources, and compare the power fluctuation amplitude of new energy sources with the preset threshold to determine whether the output of new energy sources has changed significantly; if the output of new energy sources has changed significantly, determine the solution injection strategy for this iteration based on the comparison result of the power fluctuation amplitude of new energy sources and the power output fluctuation rate of new energy sources. S104) Generate the corresponding solution according to the solution injection strategy of this iteration and replace the individuals in the extended population; S105) Calculate the corresponding Pareto front and the new parent population for the replaced extended population, calculate the comprehensive objective function of each individual in the parent population, retain the individual with the smallest comprehensive objective function value as the current optimal solution, and jump to step S102) until the maximum number of iterations is reached or the Pareto front converges, and select the optimal solution with the smallest comprehensive objective function value from the optimal solutions of each iteration as the new current optimal solution; S106) Configure protection according to the protection parameter vector of the current optimal solution.

2. The adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm according to claim 1, characterized in that, The real-time power of the new energy sources includes wind power and / or photovoltaic power, and the power output volatility of the new energy sources includes the standard deviation of wind power output and / or the standard deviation of photovoltaic power output. When the power output of the new energy sources changes significantly, the power fluctuation amplitude of the new energy sources exceeds the corresponding threshold, as shown in the following formula: in, Indicates wind power output rate of change, photovoltaic power Rate of change and load power The maximum value of the rate of change It is the rate of change threshold, and the formula is as follows: Where k is the adjustment coefficient. Indicates the standard deviation of wind power output Standard deviation of photovoltaic output The maximum value in.

3. The adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm according to claim 2, characterized in that, When determining the solution injection strategy for this iteration based on the comparison between the fluctuation range of renewable energy power and the fluctuation rate of renewable energy output, the following are included: If the relationship between the magnitude of the fluctuation in renewable energy power and the fluctuation rate of renewable energy output satisfies If the solution is positive, the random solution injection strategy will be used for this iteration; otherwise, the mutated solution strategy will be used.

4. The adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm according to claim 3, characterized in that, When generating the corresponding solution and replacing individuals in the expanded population according to the solution injection strategy of this iteration, the following are included: If the solution injection strategy for this iteration is a random solution injection strategy, then select a specified proportion of individuals in the current population, generate corresponding random solutions in a uniform distribution, and then replace the worst-performing individual in the current population with the random solution. Specifically, the worst-performing individual is the one with the worst comprehensive objective function value. The formula for the random solution is as follows: in, For the selected individuals, k is the adjustment coefficient. It is a function for generating uniformly distributed random numbers; If the solution injection strategy for this iteration is a mutation strategy, then Gaussian mutation is performed on selected individuals in the current population to obtain a mutated solution. The mutated solution then replaces the individual with the lowest crowding density in the current population. Specifically, the individual with the lowest crowding density is one surrounded by solutions with high redundancy. The formula for the mutated solution is as follows: in , The scaling factor is t, where t is the current iteration number. It is a function for generating normally distributed random numbers.

5. The adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm according to claim 1, characterized in that, The formula for calculating the comprehensive objective function is as follows: in, , , , , These are the weight coefficients for each optimization objective. It is to minimize the action time. It maximizes selectivity, and the formula is as follows: in, Main protection operating current, For backup protection operating current, For each individual, the protection setting current value is in the protection parameter vector. It is to minimize the sensitivity coefficient, and the formula is as follows: in, This refers to the fault current under the minimum operating conditions of the system. This is to protect reliability, and the formula is as follows: in, To correctly remove the fault, This represents the total number of failure events. It is the minimum false alarm rate, and the formula is as follows: in, This is the theoretically optimal value for the current.

6. The adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm according to claim 5, characterized in that, Before selecting the optimal solution with the smallest comprehensive objective function value from the optimal solutions of each iteration as the new current optimal solution, the process includes: checking whether the protection parameter vector of the optimal solution in each iteration satisfies the constraints, and eliminating optimal solutions whose protection parameter vectors do not satisfy the constraints, wherein the constraints include: Sensitivity constraints: Selectivity constraints: Error rate constraint: in, It is to minimize the sensitivity coefficient. It is the minimum false alarm rate. Indicates the timing of the primary protection action. Indicates the timing of the backup protection action. It is the time difference in the protection parameter vector, used to ensure that the main protection takes priority and the backup protection only operates when the main protection fails.

7. The adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm according to claim 1, characterized in that, Step S103 also includes: if the output of new energy sources does not change significantly, then jump to step S106 to configure protection according to the protection parameter vector of the current optimal solution.

8. A distribution network adaptive overcurrent protection system based on a dynamic multi-objective optimization algorithm, characterized in that, The device includes an interconnected processor and a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program that is executed by the processor to implement the steps of the adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, The method includes a computer program that, when executed by a processor, implements the steps of the adaptive overcurrent protection method for distribution networks based on a dynamic multi-objective optimization algorithm as described in any one of claims 1 to 7.

Images