Impact equipment waveform parameter adjusting method and device suitable for different distribution network equipment

Abstract

The invention provides an impact equipment waveform parameter adjusting method and device suitable for different distribution network equipment, and relates to the technical field of impact test wave modulation. The method comprises the following steps: determining a target waveform parameter and an adjustable parameter; carrying out linear weighting on the two and a waveform index between current parameters measured in real time, and constructing a fitness function; based on the function, an adaptive hybrid particle swarm-differential evolution algorithm is adopted to complete candidate parameter initialization, and updated parameters are obtained through particle swarm global search; if the global search is stagnated, differential evolution local refinement is started to obtain refinement parameters; and judging the fitness, if convergence is satisfied, outputting the optimal adjustment parameter, otherwise, returning to global search until convergence. According to the scheme, real-time monitoring, high-precision adjustment and self-adaptive optimization are integrated, the multi-modal impact waveform can be quickly generated, and the universality and the intelligent level of a distribution network equipment test are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of impulse test wave modulation, and in particular to a method and device for adjusting waveform parameters of an impulse device applicable to different distribution network devices. Background Art

[0002] In modern power systems, the reliability and tolerance of distribution network equipment are key to ensuring the stable operation of the power grid. In order to evaluate the performance of these devices under extreme conditions such as lightning impulses or operational shocks, impact testing has become an indispensable testing method. The core of the impact test is to generate a waveform that meets the standards, and the precise solution and adjustment of the waveform parameters directly determine the accuracy and reliability of the test results. However, the existing methods for solving the modulation parameters in impact tests have significant deficiencies in adaptability, accuracy, efficiency and safety, and are difficult to meet the diversified, efficient and intelligent testing requirements of different distribution network equipment. Therefore, the development of an innovative waveform parameter adjustment device has important practical significance. The present invention proposes an intelligent parameter adjustment device based on a particle swarm algorithm, which realizes fast and accurate parameter adjustment of different distribution network equipment in an automated and intelligent manner, so as to improve the overall level of the impact test.

[0003] Existing shock test wave modulation methods mainly include the empirical formula method, the trial and error method, and the numerical simulation method. Although they meet the basic test requirements to a certain extent, they have exposed many problems in actual application, as shown below:

[0004] The empirical formula method is a parameter-solving approach based on historical experimental data and empirical experience. During impulse testing, researchers analyzed the performance of a large number of distribution network devices under various test conditions and developed a series of mathematical formulas or empirical relationships for quickly estimating modulation parameters. This method's advantages lie in its simplicity and speed, making it suitable for rapid estimation and preliminary design. However, the empirical formula method has significant drawbacks. First, its adaptability is limited. Because the formulas are derived based on specific equipment and test conditions, they lack the flexibility to adapt to different types or specifications of distribution network equipment. For example, a formula for a transformer may not be applicable to a circuit breaker, requiring frequent formula changes when testing different devices and increasing operational complexity. Second, limited accuracy is another major drawback. The calculated results often contain significant errors, making them difficult to meet the requirements of high-precision testing. Furthermore, with advances in power equipment technology, the electrical characteristics of new equipment may exceed the applicable range of empirical formulas, limiting their application in modern testing.

[0005] The trial-and-error method is a traditional approach to determining modulation parameters through repeated trials and adjustments. The basic principle is that the tester first selects a set of initial parameters based on experience or preliminary calculations, configures a Marx generator, and performs a shock test. The resulting waveform is observed using an instrument such as an oscilloscope. Parameters such as the peak value, rise time, and fall time of the waveform are then measured and compared with the target waveform. Based on the comparison results, the tester uses experience to determine how to adjust the parameters. However, the trial-and-error method has significant shortcomings. First, it is extremely inefficient. Each test requires reconfiguring parameters and performing tests, which is time-consuming. Especially when testing multiple devices in batches, the adjustment process can take hours, significantly impacting test progress. Furthermore, this method is highly subjective, and the results are highly dependent on the tester's experience. Different testers may produce different results, lacking consistency and repeatability. Finally, safety risks are not negligible. Manually adjusting parameters in a high-voltage environment exposes testers to the risk of electric shock or equipment failure, potentially threatening personal safety and equipment integrity.

[0006] Numerical simulation is a modern method that uses computers to simulate the process of generating shock waveforms to solve the modulation parameters. By solving the differential equations of the circuit or using professional simulation software, the waveform output under different modulation parameter combinations is simulated. Researchers set a series of parameter combinations, run simulations, obtain waveform data, and compare them with the target waveform. By analyzing the differences, the parameters are adjusted to optimize the waveform. However, the numerical simulation method also has many shortcomings. First, the complexity of the model is its main obstacle. Establishing an accurate mathematical model involves complex circuit analysis, and the modeling process is difficult and time-consuming. Secondly, the large amount of calculation limits its efficiency. Simulation and optimization require a lot of computing resources, which makes it difficult to meet the needs of fast testing. In addition, the method is not versatile enough. Different models need to be established for different devices. When faced with new equipment or non-standard tests, the cost of model development is high, which limits its widespread application.

[0007] In view of the many shortcomings of the existing methods for solving the modulation parameters of impact tests, various methods cannot simultaneously meet the requirements of reliability, accuracy, controllable range and efficiency. Therefore, it is of great technical necessity and practical significance to develop a new technical solution that can integrate high-precision dynamic adjustment, multi-modal waveform generation and intelligent adaptive decision-making. Summary of the Invention

[0008] In order to overcome the shortcomings of the existing technology, the purpose of the present invention is to provide a method and device for adjusting the waveform parameters of impact equipment suitable for different distribution network equipment, which can integrate high-precision dynamic adjustment, multi-modal waveform generation and intelligent adaptive decision-making, solve the shortcomings of existing methods in terms of versatility, real-time performance and intelligence, and meet the diversified testing needs of complex distribution network equipment.

[0009] To achieve the above object, the present invention provides the following solutions:

[0010] A method for adjusting waveform parameters of impact equipment applicable to different distribution network equipment, comprising:

[0011] determining target waveform parameters and adjustable parameters for reflecting target impact waveform characteristics;

[0012] Obtaining a fitness function by linearly weighting waveform indices between the target waveform parameter, the adjustable parameter, and the measured current waveform parameter;

[0013] Based on the fitness function, the candidate adjustment parameters are initialized using an adaptive hybrid particle swarm-differential evolution algorithm to obtain initialization parameters;

[0014] Based on the initialization parameters, executing a global search strategy of the particle swarm algorithm, iteratively updating the candidate adjustment parameters, and obtaining updated candidate adjustment parameters;

[0015] When the iterative result of the particle swarm algorithm meets the stagnation judgment condition, the local refinement strategy of the differential evolution algorithm is triggered to mutate, crossover and select the updated candidate adjustment parameters to obtain refined parameters;

[0016] The fitness value corresponding to the refined parameter is determined according to a preset convergence criterion. If the fitness value meets the convergence condition, the refined parameter is determined as the optimal adjustment parameter; if not, the global search strategy of the particle swarm algorithm is re-executed until convergence.

[0017] Preferably, it also includes:

[0018] The optimal adjustment parameters are loaded into the impulse generator control system to output an impulse voltage waveform that meets the target waveform requirements.

[0019] Preferably, it also includes:

[0020] monitoring an actual waveform output by the impulse generator control system in real time, and comparing the actual waveform with a target waveform;

[0021] When the deviation of any of the waveform indicators exceeds the allowable range, the current optimal adjustment parameter is used as a new candidate adjustment parameter, and the global search strategy of the particle swarm algorithm and the local refinement strategy of the differential evolution algorithm are re-executed to complete rapid readjustment;

[0022] When all the waveform indicators are within the allowable range, the existing adjustment parameters are maintained and continuously monitored.

[0023] Preferably, the target waveform parameters include: wave front time, wave tail time and peak voltage; the adjustable parameters include wave front resistance, wave tail resistance and wave modulation inductance.

[0024] Preferably, the waveform indicators include: error and energy loss ratio.

[0025] Preferably, based on the fitness function, the candidate adjustment parameters are initialized using an adaptive hybrid particle swarm-differential evolution algorithm to obtain the initialization parameters, including:

[0026] Randomly generating a plurality of groups of candidate adjustment parameters within the allowable range of the adjustable parameters to form an initial population;

[0027] respectively calling the fitness function to calculate the fitness value for each group of candidate adjustment parameters;

[0028] Recording the fitness value of each group of candidate adjustment parameters as the individual initial optimal value of the corresponding group of candidate adjustment parameters, and selecting the group with the best fitness value from all candidate adjustment parameters as the initial global optimal value of the population;

[0029] Set the inertia weight and learning coefficient of the particle swarm algorithm, as well as the mutation factor and crossover probability of the differential evolution algorithm as algorithm control variables;

[0030] The judgment conditions of the stagnation window width, the improvement threshold and the maximum number of iterations are set, and the algorithm control amount is output together with the initial population to obtain the initialization parameters.

[0031] Preferably, based on the initialization parameters, a global search strategy of the particle swarm algorithm is executed to iteratively update the candidate adjustment parameters to obtain updated candidate adjustment parameters, including:

[0032] Calculating a search step size for each set of candidate adjustment parameters based on the inertia weight and the learning coefficient;

[0033] Synchronously updating the positions of all the candidate adjustment parameters using the search step to form updated candidate adjustment parameters;

[0034] Calling the fitness function to calculate the fitness value corresponding to each group of updated candidate adjustment parameters;

[0035] Record the individual optimal fitness of each set of candidate adjustment parameters separately, and refresh the global optimal fitness of the entire group;

[0036] Determine the improvement of the global optimal fitness within the preset algebraic window. If the improvement is lower than the stagnation threshold or the number of iterations reaches the upper limit, end the step; otherwise, return to the step of "calculating the search step size of each set of candidate adjustment parameters based on the inertia weight and the learning coefficient."

[0037] Preferably, when the iterative result of the particle swarm algorithm meets the stagnation judgment condition, the local refinement strategy of the differential evolution algorithm is triggered to mutate, crossover and select the updated candidate adjustment parameters to obtain refined parameters, including:

[0038] Select some parameters whose fitness values ​​rank within a preset range from the current updated candidate adjustment parameter population as the parent population;

[0039] performing a mutation operation on the parent population according to a differential vector strategy to generate intermediate candidate adjustment parameters;

[0040] Performing a crossover operation on the intermediate candidate adjustment parameters and the parent population to form a child population;

[0041] Calling the fitness function to calculate the fitness value of the offspring population, and comparing it with the corresponding parent population, retaining the best to update the candidate adjustment parameters;

[0042] Determine whether the global optimal fitness of the offspring population meets the convergence condition or has reached the preset number of iterations; if so, output the refined parameters; otherwise, return to the step of "selecting some parameters whose fitness values ​​rank within the preset range from the current updated candidate adjustment parameter population as the parent population."

[0043] Preferably, the preset range is the top 5% to 20%.

[0044] A device for adjusting waveform parameters of impact equipment applicable to different distribution network equipment, comprising:

[0045] a target waveform and adjustable parameter determination sub-device, for determining target waveform parameters and adjustable parameters for reflecting target impact waveform characteristics;

[0046] a fitness function calculation sub-device, configured to perform linear weighting on waveform indices between the target waveform parameter, the adjustable parameter, and the measured current waveform parameter to obtain a fitness function;

[0047] An adaptive hybrid algorithm initialization sub-device, configured to initialize candidate adjustment parameters based on the fitness function using an adaptive hybrid particle swarm-differential evolution algorithm to obtain initialization parameters;

[0048] A particle swarm global search sub-device, configured to execute a global search strategy of the particle swarm algorithm based on the initialization parameters, iteratively update the candidate adjustment parameters, and obtain updated candidate adjustment parameters;

[0049] A differential evolution local refinement sub-device, configured to trigger a local refinement strategy of the differential evolution algorithm when the iterative result of the particle swarm algorithm satisfies a stagnation judgment condition, and perform mutation, crossover, and selection on the updated candidate adjustment parameters to obtain refined parameters;

[0050] The optimal parameter determination and output sub-device is used to determine the fitness value corresponding to the refined parameter according to a preset convergence criterion. If the fitness value meets the convergence condition, the refined parameter is determined as the optimal adjustment parameter; if not, the global search strategy of the particle swarm algorithm is re-executed until convergence.

[0051] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0052] The present invention provides a method and device for adjusting waveform parameters of impact equipment applicable to different distribution network equipment. The method comprises: determining target waveform parameters and adjustable parameters for reflecting target impact waveform characteristics; performing linear weighting according to waveform indicators between the target waveform parameters, the adjustable parameters and the measured current waveform parameters to obtain a fitness function; initializing candidate adjustment parameters based on the fitness function using an adaptive hybrid particle swarm-differential evolution algorithm to obtain initialization parameters; executing a global search strategy of the particle swarm algorithm based on the initialization parameters to iteratively update the candidate adjustment parameters to obtain updated candidate adjustment parameters; when the iterative result of the particle swarm algorithm meets a stagnation judgment condition, triggering a local refinement strategy of the differential evolution algorithm to mutate, crossover and select the updated candidate adjustment parameters to obtain refined parameters; judging the fitness value corresponding to the refined parameter according to a preset convergence criterion, and if the fitness value meets the convergence condition, determining the refined parameter as the optimal adjustment parameter; if not, re-executing the global search strategy of the particle swarm algorithm until convergence. The present invention can integrate high-precision dynamic adjustment, multi-modal waveform generation and intelligent adaptive decision-making, solving the shortcomings of existing methods in terms of versatility, real-time performance and intelligence, and meeting the diversified testing needs of complex distribution network equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0054] Figure 1 A flow chart of a method provided by an embodiment of the present invention;

[0055] Figure 2A schematic diagram of a technical route for adjusting waveform parameters of a Gröning loop impact device based on a hybrid PSO-DE algorithm provided in an embodiment of the present invention;

[0056] Figure 3 This is a flow chart of an adaptive hybrid PSO-DE algorithm based on stagnant state switching provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0058] The purpose of the present invention is to provide a method and device for adjusting the waveform parameters of an impact device suitable for different distribution network equipment, which can be used to quickly and accurately adjust the impact waveforms of different distribution network equipment.

[0059] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0060] Figure 1 A flow chart of the method provided in the embodiment of the present invention is shown in FIG. Figure 1 As shown, the present invention provides a method for adjusting waveform parameters of impact equipment applicable to different distribution network equipment, including:

[0061] Step 100: determining target waveform parameters and adjustable parameters for reflecting target impact waveform characteristics;

[0062] Step 200: linearly weighting the waveform indices between the target waveform parameters, the adjustable parameters, and the measured current waveform parameters to obtain a fitness function;

[0063] Step 300: Based on the fitness function, the candidate adjustment parameters are initialized using the adaptive hybrid particle swarm-differential evolution algorithm to obtain the initialization parameters;

[0064] Step 400: Based on the initialization parameters, execute the global search strategy of the particle swarm algorithm to iteratively update the candidate adjustment parameters to obtain updated candidate adjustment parameters;

[0065] Step 500: When the iteration result of the particle swarm algorithm meets the stagnation judgment condition, the local refinement strategy of the differential evolution algorithm is triggered to mutate, crossover and select the updated candidate adjustment parameters to obtain refined parameters;

[0066] Step 600: The fitness value corresponding to the refined parameter is determined according to the preset convergence criterion. If the fitness value meets the convergence condition, the refined parameter is determined as the optimal adjustment parameter; if not, the global search strategy of the particle swarm algorithm is re-executed until convergence.

[0067] like Figure 2 As shown, this embodiment invents a Gröning circuit wave modulation device based on an adaptive hybrid PSO-DE algorithm with stagnant state switching. The Gröning circuit is a common impulse waveform generator, widely used to generate high-voltage pulses such as lightning impulse waveforms to test the withstand voltage capability of power equipment. Its wave modulation parameters, such as the wave head resistance, wave tail resistance, and wave modulation inductance, directly determine key characteristics of the waveform, such as the rise time, fall time, and peak value. The technical route of this embodiment includes:

[0068] Step 1: Target waveform setting. Define the desired impulse waveform characteristics, such as the standard lightning impulse waveform (1.2 / 50μs, peak voltage 75kV). Specific parameters include: wave head time, wave tail time, and peak voltage (e.g., 75kV).

[0069] Step 2: Identify adjustable parameters. In a Gröning circuit impact device, the adjustable parameters that affect the waveform typically include: front resistor (controls the wavefront time); tail resistor (controls the wavetail time); and modulation inductor (adjusts the waveform oscillation and stability). The ranges of these parameters must be determined based on the device design. By optimizing these parameters, the device output waveform can be as close to the target waveform as possible, while also balancing energy efficiency and system stability.

[0070] Step 3: Define the fitness function. The fitness function is used to measure the matching degree between the current parameter combination and the target waveform. The design is as follows:

[0071]

[0072] Among them, R front is the wave head resistance, R tail is the wave tail resistance, L adjust is the modulation inductor. X i =[R front , R tail , L adjust ] is the i-th candidate solution; is the actual waveform parameter under the current parameters; is the target waveform parameter; E loss is the energy loss, E total is the total energy; w1, w2, w3, w4 are weight coefficients.

[0073] Step 4: If Figure 3 As shown, this embodiment provides an initialized hybrid PSO-DE algorithm.

[0074] (1) Particle swarm initialization: Generate an initial particle swarm; Define the number of particles; The initial position X i = [R front , R tail， L adjust and the initial velocity [v front , v tail， v adjust are randomly generated within the parameter range. v front , v tail， v adjust represents the initial velocities of the wave head resistance, wave tail resistance, and tuning wave inductor of the three particles.

[0075] (2) PSO parameter setting: Inertia weight w: The initial value is 0.9, and it linearly decreases to 0.4 with iteration; Learning factors c1, c2: Set to 2.0.

[0076] (3) DE parameter setting: Mutation factor F: 0.5 - 0.8; Crossover probability CR: 0.7 - 0.9.

[0077] (4) Adaptive switching parameter initialization:

[0078] 1) Stagnation window (SW): Set how many consecutive generations without significant improvement are considered stagnant. SW = 10

[0079] 2) Improvement threshold (Thresh): Set the minimum effective improvement amount of the global optimal fitness. Thresh = 1e -5 [[ID=3⑧]].

[0080] 3) Maximum number of PSO iterations (MaxIter_PSO): Set an upper limit for the PSO operation to prevent infinite running due to the inability to trigger stagnation for some reason. MaxIter_PSO = 100.

[0081] 4) Initialize the stagnation counter: stagnation_counter = 0.

[0082] 5) Initialize the global optimal fitness record: Gbest_fitness = infinity. Gbest = None.

[0083] Step Five: Execute the adaptive PSO stage

[0084] (1) Set the current PSO iteration number t_pso = 0.

[0085] (2) Start the loop (the termination condition is: t_pso < MaxIter_PSO and stagnation_counter < SW)

[0086] 1) Calculate the fitness: For each particle Xi , measure the waveform generated by simulation or experiment, and calculate F(X i ).

[0087] 2) Update individual and global optimal solutions: individual optimal solution (P besti ): Each particle records its best historical position; the global optimal solution (G best ): The position with the best fitness among all particles.

[0088] 3) Update the global optimal solution (Gbest): Find the optimal particle in the current population and its fitness current_best_fitness; record the global optimal fitness before the update Gbest_fitness_prev = Gbest_fitness; if current_best_fitness is better than Gbest_fitness, update Gbest to the position of the current optimal particle and update Gbest_fitness = current_best_fitness.

[0089] 4) Stagnation detection:

[0090] Calculate the improvement: improvement = abs(Gbest_fitness_prev - Gbest_fitness) and determine whether it is stagnant:

[0091] If improvement <Thresh:stagnation_counter=stagnation_counter+1。

[0092] Otherwise: stagnation_counter=0.

[0093] 5) Speed ​​and position update:

[0094]

[0095] Where w is the inertia weight, c1 and c2 are learning factors, r1 and r2 are random numbers (0 to 1), Pbesti is the optimal solution for the individual particle, and Gbest is the global optimal solution.

[0096] 6) Boundary processing: If If it exceeds the range, it will be clamped within the boundary value.

[0097] 7) Iteration count: t_pso=t_pso+1.

[0098] End of loop (automatically exits when MaxIter_PSO is reached or stagnation_counter reaches SW)

[0099] Step 6: Transition to the DE phase. After the loop of Step 5 completes, record the number of PSO generations actually run (t_pso_final = t_pso). Perform initial population selection: From the final PSO particle swarm, select the top K particles with the best fitness (K = ceil(N_PSO * 0.1)). Use the positions of these particles as the initial DE population, Pop_DE.

[0100] Step 7: DE stage (local refinement)

[0101] (1) Set the DE population size NP_DE = K.

[0102] (2) Set the fixed number of iterations of DE operation MaxIter_DE. MaxIter_DE = 50

[0103] (3) Set the current number of DE iterations t_de=0.

[0104] (4) Loop starts (termination condition: t_de <MaxIter_DE)

[0105] 1) Mutation operation: for each particle X i Generate mutation vector:

[0106] V i =X r1 +F·(X r2 -X r3 )

[0107] X r1 , X r1 , X r1 are three different particles selected randomly; F is the variation factor.

[0108] 2) Crossover operation: Cross the mutation vector Vi with the original particle Xi to generate a candidate solution U i，j :

[0109]

[0110] 3) Calculate fitness: Calculate U i，j The fitness F(U i，j ).

[0111] 4) Select operation: If F(U i，j ) is better than F(Xi), then the next generation will use U i，j Replace Xi.

[0112] 5) Update global optimality: After each DE iteration, check whether the optimal individual in Pop_DE is better than the currently recorded global optimal Gbest. If so, update Gbest and Gbest_fitness.

[0113] 6) Iteration count: t_de=t_de+1.

[0114] 7) The loop ends (ends after running MaxIter_DE times)

[0115] Step 8: Verification and output. Take the optimal solution of the DE stage:

[0116]

[0117] The optimized wave head resistance, wave tail resistance, and wave modulation inductance values ​​are verified through simulation or experiment to see whether the generated waveform meets the target.

[0118] Step 9: Dynamic Adjustment: If the load or environment changes during device operation, the PSO-DE optimization can be retriggered: the particle swarm is initialized with the current optimal solution, and simplified iterations are run to quickly adjust parameters.

[0119] Compared to the original hybrid PSO-DE algorithm, which switched based on a fixed number of iterations, the adaptive hybrid PSO-DE algorithm proposed in this revision differs significantly in key steps. The most significant change lies in the refinement of the PSO phase's operating mechanism and termination conditions in steps 4 through 7: While the original algorithm forces the PSO to run for a fixed number of iterations before switching to DE, the adaptive algorithm dynamically monitors the improvement of the PSO global optimal solution. The number of iterations in the PSO phase becomes variable, terminating when the global optimal solution improves by less than a threshold Thresh over consecutive SW generations or when a preset maximum number of iterations is reached. Therefore, the switch from PSO to DE is no longer a fixed point in time, but is triggered dynamically based on the algorithm's actual search performance. This introduces a new control parameter, SW,Thresh, and alters the algorithm's internal logic, transforming it from a simple sequential execution process to a more intelligent process that includes state monitoring and decision-making.

[0120] This adaptive switching mechanism addresses several key technical deficiencies of the original fixed-iteration switching algorithm when applied to complex optimization problems such as Graning loop wave modulation. First, it addresses the issue of irrational computational resource allocation. The original algorithm may have improperly set a fixed number of PSO iterations, leading to premature switching or redundant computation. The adaptive algorithm avoids inefficient computation by promptly terminating the PSO and switching to DE when the PSO stagnates, improving overall computational efficiency. Second, it overcomes the shortcomings of rigid switching timing and a lack of intelligence. Adaptive switching bases algorithm phase transitions on actual performance rather than arbitrary iteration counts, making it more adaptable to the dynamic nature of the optimization problem. Third, by allowing the PSO to explore until efficiency decreases before introducing DE for further development, the adaptive algorithm helps achieve a better balance between exploration and development, avoiding the imbalance that can result from fixed partitioning. Finally, although new parameters are introduced, performance-based switching reduces the algorithm's absolute dependence on a single fixed number of iterations, potentially improving the algorithm's robustness to parameter settings. In summary, the adaptive hybrid PSO-DE algorithm, through its dynamic switching strategy, can more intelligently and efficiently solve the wave modulation parameter optimization task of Graning loops.

[0121] Corresponding to the above method, this embodiment further provides a device for adjusting waveform parameters of an impact device applicable to different distribution network devices, including:

[0122] a target waveform and adjustable parameter determination sub-device, for determining target waveform parameters and adjustable parameters for reflecting target impact waveform characteristics;

[0123] a fitness function calculation sub-device, configured to perform linear weighting on waveform indices between the target waveform parameter, the adjustable parameter, and the measured current waveform parameter to obtain a fitness function;

[0124] An adaptive hybrid algorithm initialization sub-device, configured to initialize candidate adjustment parameters based on the fitness function using an adaptive hybrid particle swarm-differential evolution algorithm to obtain initialization parameters;

[0125] A particle swarm global search sub-device, configured to execute a global search strategy of the particle swarm algorithm based on the initialization parameters, iteratively update the candidate adjustment parameters, and obtain updated candidate adjustment parameters;

[0126] A differential evolution local refinement sub-device, configured to trigger a local refinement strategy of the differential evolution algorithm when the iterative result of the particle swarm algorithm satisfies a stagnation judgment condition, and perform mutation, crossover, and selection on the updated candidate adjustment parameters to obtain refined parameters;

[0127] The optimal parameter determination and output sub-device is used to determine the fitness value corresponding to the refined parameter according to a preset convergence criterion. If the fitness value meets the convergence condition, the refined parameter is determined as the optimal adjustment parameter; if not, the global search strategy of the particle swarm algorithm is re-executed until convergence.

[0128] The present invention proposes a new solution to the limitations of traditional impact test wave modulation methods when applied to different distribution network equipment, as well as the shortcomings in wave modulation efficiency and accuracy. Its core lies in the design of a Graning loop wave modulation device based on an adaptive hybrid PSO-DE algorithm. The device can automatically and intelligently adjust the key parameters of the Graning loop impact equipment, including the wave head resistance, wave tail resistance and wave modulation inductance, so as to accurately control and optimize the key characteristics of the impact waveform, such as the wave head time, wave tail time and peak voltage. Compared with traditional empirical formulas and trial-and-error methods, the invention uses a hybrid optimization algorithm with a dynamic switching strategy based on the search state (PSO first explores and automatically switches to DE for refinement when its efficiency decreases) to improve the efficiency and accuracy of the wave modulation process, and can quickly and accurately generate impact waveforms that meet the needs of specific distribution network equipment. In addition, the invention also has dynamic adjustment capabilities, and can adjust parameters in real time according to load or environmental changes during equipment operation to ensure the stability and reliability of the impact waveform. In summary, the key to this invention lies in applying the adaptive hybrid optimization algorithm to the automated parameter adjustment of the Graning loop impact device. Through more intelligent algorithm stage switching, it solves the efficiency and accuracy problems of traditional wave modulation methods and achieves good adaptability to different distribution network equipment.

[0129] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0130] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A method for adjusting waveform parameters of impact equipment applicable to different distribution network equipment, characterized in that: include: determining target waveform parameters and adjustable parameters for reflecting target impact waveform characteristics; Obtaining a fitness function by linearly weighting waveform indices between the target waveform parameter, the adjustable parameter, and the measured current waveform parameter; Based on the fitness function, the candidate adjustment parameters are initialized using an adaptive hybrid particle swarm-differential evolution algorithm to obtain initialization parameters; Based on the initialization parameters, executing a global search strategy of the particle swarm algorithm, iteratively updating the candidate adjustment parameters, and obtaining updated candidate adjustment parameters; When the iterative result of the particle swarm algorithm meets the stagnation judgment condition, the local refinement strategy of the differential evolution algorithm is triggered to mutate, crossover and select the updated candidate adjustment parameters to obtain refined parameters; The fitness value corresponding to the refined parameter is determined according to a preset convergence criterion. If the fitness value meets the convergence condition, the refined parameter is determined as the optimal adjustment parameter; if not, the global search strategy of the particle swarm algorithm is re-executed until convergence.

2. The method for adjusting waveform parameters of impact equipment applicable to different distribution network equipment according to claim 1, characterized in that: Also includes: The optimal adjustment parameters are loaded into the impulse generator control system to output an impulse voltage waveform that meets the target waveform requirements.

3. The method for adjusting waveform parameters of impact equipment applicable to different distribution network equipment according to claim 2, characterized in that: Also includes: monitoring an actual waveform output by the impulse generator control system in real time, and comparing the actual waveform with a target waveform; When the deviation of any of the waveform indicators exceeds the allowable range, the current optimal adjustment parameter is used as a new candidate adjustment parameter, and the global search strategy of the particle swarm algorithm and the local refinement strategy of the differential evolution algorithm are re-executed to complete rapid readjustment; When all the waveform indicators are within the allowable range, the existing adjustment parameters are maintained and continuously monitored.

4. The method for adjusting waveform parameters of impact equipment applicable to different distribution network equipment according to claim 1, characterized in that: The target waveform parameters include: wave head time, wave tail time and peak voltage; the adjustable parameters include wave head resistance, wave tail resistance and wave modulation inductance.

5. The method for adjusting waveform parameters of impact equipment applicable to different distribution network equipment according to claim 1, characterized in that: The waveform indicators include: error and energy loss ratio.

6. The method for adjusting waveform parameters of impact equipment applicable to different distribution network equipment according to claim 1, characterized in that: Based on the fitness function, the candidate adjustment parameters are initialized using the adaptive hybrid particle swarm-differential evolution algorithm to obtain the initialization parameters, including: Randomly generating a plurality of groups of candidate adjustment parameters within the allowable range of the adjustable parameters to form an initial population; respectively calling the fitness function to calculate the fitness value for each group of candidate adjustment parameters; Recording the fitness value of each group of candidate adjustment parameters as the individual initial optimal value of the corresponding group of candidate adjustment parameters, and selecting the group with the best fitness value from all candidate adjustment parameters as the initial global optimal value of the population; Set the inertia weight and learning coefficient of the particle swarm algorithm, as well as the mutation factor and crossover probability of the differential evolution algorithm as algorithm control variables; The judgment conditions of the stagnation window width, the improvement threshold and the maximum number of iterations are set, and the algorithm control amount is output together with the initial population to obtain the initialization parameters.

7. The method for adjusting waveform parameters of impact equipment applicable to different distribution network equipment according to claim 6, characterized in that: Based on the initialization parameters, a global search strategy of the particle swarm algorithm is executed to iteratively update the candidate adjustment parameters to obtain updated candidate adjustment parameters, including: Calculating a search step size for each set of candidate adjustment parameters based on the inertia weight and the learning coefficient; Synchronously updating the positions of all the candidate adjustment parameters using the search step to form updated candidate adjustment parameters; Calling the fitness function to calculate the fitness value corresponding to each group of updated candidate adjustment parameters; Record the individual optimal fitness of each set of candidate adjustment parameters separately, and refresh the global optimal fitness of the entire group; Determine the improvement of the global optimal fitness within a preset algebraic window. If the improvement is lower than a stagnation threshold or the number of iterations reaches an upper limit, end the step; otherwise, return to step "calculating the search step size of each set of candidate adjustment parameters based on the inertia weight and the learning coefficient." 8. The method for adjusting waveform parameters of impact equipment applicable to different distribution network equipment according to claim 7, characterized in that: When the iterative result of the particle swarm algorithm meets the stagnation judgment condition, the local refinement strategy of the differential evolution algorithm is triggered to mutate, crossover and select the updated candidate adjustment parameters to obtain refined parameters, including: Select some parameters whose fitness values ​​rank within a preset range from the current updated candidate adjustment parameter population as the parent population; performing a mutation operation on the parent population according to a differential vector strategy to generate intermediate candidate adjustment parameters; Performing a crossover operation on the intermediate candidate adjustment parameters and the parent population to form a child population; Calling the fitness function to calculate the fitness value of the offspring population, and comparing it with the corresponding parent population, retaining the best to update the candidate adjustment parameters; Determine whether the global optimal fitness of the offspring population meets the convergence condition or has reached the preset number of iterations; if so, output the refined parameters; otherwise, return to step "selecting some parameters whose fitness values ​​rank within the preset range from the current updated candidate adjustment parameter population as the parent population." 9. The method for adjusting waveform parameters of impact equipment applicable to different distribution network equipment according to claim 7, characterized in that: The preset range is the top 5% to 20%.

10. A device for adjusting waveform parameters of impact equipment applicable to different distribution network equipment, characterized in that: include: a target waveform and adjustable parameter determination sub-device, for determining target waveform parameters and adjustable parameters for reflecting target impact waveform characteristics; a fitness function calculation sub-device, configured to perform linear weighting on waveform indices between the target waveform parameter, the adjustable parameter, and the measured current waveform parameter to obtain a fitness function; An adaptive hybrid algorithm initialization sub-device, configured to initialize candidate adjustment parameters based on the fitness function using an adaptive hybrid particle swarm-differential evolution algorithm to obtain initialization parameters; A particle swarm global search sub-device, configured to execute a global search strategy of the particle swarm algorithm based on the initialization parameters, iteratively update the candidate adjustment parameters, and obtain updated candidate adjustment parameters; A differential evolution local refinement sub-device, configured to trigger a local refinement strategy of the differential evolution algorithm when the iterative result of the particle swarm algorithm satisfies a stagnation judgment condition, and perform mutation, crossover, and selection on the updated candidate adjustment parameters to obtain refined parameters; The optimal parameter determination and output sub-device is used to determine the fitness value corresponding to the refined parameter according to a preset convergence criterion. If the fitness value meets the convergence condition, the refined parameter is determined as the optimal adjustment parameter; if not, the global search strategy of the particle swarm algorithm is re-executed until convergence.

Images