Harmonic compensation capacity distribution algorithm and system based on feeder harmonic responsibility

Abstract

The invention provides a harmonic compensation capacity distribution method based on feeder harmonic responsibility. The harmonic compensation capacity distribution method comprises the steps of S1, collecting operation data of each feeder in a power distribution network; s2, extracting a voltage amplitude and a current amplitude of each harmonic wave of each feeder line; s3, calculating a harmonic wave responsibility coefficient of each feeder line under each harmonic wave; s4, acquiring a voltage allowable limit value of the power distribution network system under each harmonic wave; s5, a harmonic compensation capacity distribution model is established on the basis of the harmonic responsibility coefficient and the voltage allowable limit value, and the model takes the minimum total compensation capacity as the target and comprises the constraint condition that the compensated harmonic voltage does not exceed the voltage allowable limit value; and S6, solving the harmonic compensation capacity distribution model by adopting an optimization algorithm to obtain the compensation capacity of each feeder line under each harmonic. Through multi-algorithm collaborative optimization and full-process dynamic adaptation, scientific distribution of harmonic compensation capacity is realized, and accurate, economic and adaptive technical support is provided for harmonic suppression of the power distribution network.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system automation and power quality management, in particular to a harmonic compensation capacity distribution algorithm and system based on feeder harmonic responsibility. BACKGROUND

[0002] With the development of power electronic technology, the proportion of nonlinear loads (such as frequency converters, charging piles, photovoltaic inverters, etc.) in distribution networks is increasing year by year, leading to increasingly serious harmonic pollution. Harmonics not only increase line loss and reduce equipment life, but also may cause protection misoperation, metering deviation and other problems, seriously affecting power quality.

[0003] Existing harmonic compensation technologies mostly adopt the "centralized compensation" or "empirical capacity distribution" mode, which has the following limitations: The harmonic responsibility of each feeder is not quantified, resulting in a mismatch between compensation capacity distribution and actual harmonic contribution, leading to "overcompensation" (increasing cost) or "insufficient compensation" (harmonics still exceeding the standard); It relies on a fixed compensation scheme and cannot adapt to dynamic scenarios such as load fluctuations and topology changes in distribution networks, and is prone to failure after long-term operation; The responsibility coefficient calculation is based on a single algorithm (such as the current superposition method), without considering factors such as system impedance fluctuations and voltage coupling, resulting in low accuracy in responsibility division and affecting the rationality of capacity distribution.

[0004] Therefore, there is an urgent need for a technical solution that can accurately quantify the harmonic responsibility of each feeder and dynamically optimize the distribution of compensation capacity to improve the scientificity and economy of harmonic management. SUMMARY

[0005] The present application provides a harmonic compensation capacity distribution algorithm and system based on feeder harmonic responsibility, which realizes scientific distribution of harmonic compensation capacity through multi-algorithm collaborative optimization and full-process dynamic adaptation, providing precise, economic and adaptive technical support for distribution network harmonic management.

[0006] To achieve the above purpose, the present application adopts the following technical solutions: The harmonic compensation capacity distribution method based on feeder harmonic responsibility comprises: S1: Collecting the operating data of each feeder in the distribution network, including the voltage time domain signal, current time domain signal, impedance parameter and system topology structure of each feeder; S2: Processing the voltage time domain signal and current time domain signal to extract the voltage amplitude and current amplitude of each harmonic of each feeder; S3: Calculating the harmonic responsibility coefficient of each feeder under each harmonic based on the current amplitude, voltage amplitude and impedance parameter of each harmonic; S4: obtaining voltage allowable limits of the power distribution system under each harmonic; S5: establishing a harmonic compensation capacity distribution model based on the harmonic responsibility coefficient and the voltage allowable limits, the model taking the minimum total compensation capacity as the target and containing a constraint condition that the harmonic voltage after compensation does not exceed the voltage allowable limits; S6: solving the harmonic compensation capacity distribution model by using an optimization algorithm to obtain the compensation capacity of each feeder under each harmonic.

[0007] In the specification, the harmonic compensation capacity distribution method based on the feeder harmonic responsibility further includes S7: verifying the compensation effect corresponding to the compensation capacity based on the system topological structure and impedance parameters, judging whether the harmonic voltage after compensation meets the voltage allowable limits; if not, feeding back the deviation to step S5 to adjust the model constraint condition, and repeating steps S5 to S7.

[0008] In the specification, the harmonic compensation capacity distribution method based on the feeder harmonic responsibility further includes S8: if the compensation effect verification passes, summarizing the compensation capacity of each feeder under each harmonic to obtain the total harmonic compensation capacity of each feeder.

[0009] In the specification, the harmonic compensation capacity distribution method based on the feeder harmonic responsibility further includes S9: periodically repeating steps S1 to S8 to update the total harmonic compensation capacity according to newly collected operation data.

[0010] In the specification, in step S3, the harmonic responsibility coefficient is obtained by weighted fusion of an initial coefficient and a correction coefficient, the initial coefficient is calculated by the current superposition method, and specifically, the product of the initial coefficient of a certain feeder under a certain harmonic and the amplitude of the harmonic current of the feeder and the impedance module value is proportional to the sum of the products of the amplitudes of the harmonic currents of all feeders and the corresponding impedance module values; the correction coefficient is obtained by correcting the initial coefficient by using the least square method, and the least square method takes the minimum error square sum of the calculated value and the measured value of the harmonic voltage at each time as the target.

[0011] In the specification, in step S5, the constraint condition is dynamically corrected by using a neural network, the input of the neural network includes the harmonic responsibility coefficient, the harmonic voltage amplitude before compensation, and the voltage allowable limits, and the output is a correction coefficient, which is used to adjust the weight of the harmonic responsibility coefficient in the constraint condition.

[0012] In this specification, in step S6, the optimization algorithm is a fusion algorithm of genetic algorithm and particle swarm optimization algorithm; the genetic algorithm is used to generate an initial compensation capacity solution that satisfies the constraints, and the particle swarm optimization algorithm is used to refine the initial compensation capacity solution. During the refinement process, the particle velocity and position update formulas are iteratively optimized until the solution with the minimum total compensation capacity is obtained.

[0013] In this specification, the weights of the weighted fusion are determined based on the stability of the feeder load. The smaller the load fluctuation, the greater the weight of the correction coefficient; the greater the load fluctuation, the greater the weight of the initial coefficient. The value range of the weights is 0.6 to 0.9.

[0014] In this specification, the process of generating the initial compensation capacity solution using the genetic algorithm includes: encoding the compensation capacity of each harmonic of each feeder into chromosomes, selecting the parent generation by roulette wheel selection, generating offspring generation using arithmetic crossover and Gaussian mutation operations, iterating through a preset number of generations, and selecting the solution with the highest fitness as the initial solution. The fitness function is inversely proportional to the total compensation capacity.

[0015] The harmonic compensation capacity allocation system based on feeder harmonic responsibility, applying any one of the above-described harmonic compensation capacity allocation methods based on feeder harmonic responsibility, comprises: The data acquisition module is used to collect the operating data of each feeder in the distribution network. The operating data includes the voltage time-domain signal, current time-domain signal, impedance parameters and system topology of each feeder. The harmonic parameter extraction module is used to process the voltage time-domain signal and the current time-domain signal to extract the voltage amplitude and current amplitude of each harmonic of each feeder. The feeder harmonic responsibility factor calculation module is used to calculate the harmonic responsibility factor of each feeder under each harmonic based on the current amplitude, voltage amplitude and impedance parameters of each harmonic. The system harmonic permissible limit determination module is used to obtain the voltage permissible limit of the distribution network system under each harmonic. The harmonic compensation capacity allocation model construction module is used to establish a harmonic compensation capacity allocation model based on the harmonic responsibility coefficient and the voltage allowable limit. The model aims to minimize the total compensation capacity and includes the constraint that the harmonic voltage after compensation does not exceed the voltage allowable limit. The compensation capacity allocation model solution module is used to solve the harmonic compensation capacity allocation model using an optimization algorithm to obtain the compensation capacity of each feeder under each harmonic.

[0016] In summary, the present invention has at least the following beneficial effects: Improve compensation accuracy: By combining the current superposition method and the least squares method to optimize the harmonic responsibility coefficient of the feeder, and combining the neural network to dynamically correct the constraints, the compensation capacity and the actual harmonic contribution of each feeder are accurately matched, avoiding overcompensation or undercompensation.

[0017] Enhanced dynamic adaptability: The introduction of a periodic dynamic adjustment mechanism can respond in real time to changes in operating conditions such as distribution network load fluctuations and equipment switching, ensuring that the harmonic compensation effect continues to meet the standards under different operating conditions.

[0018] Optimize resource allocation: With the goal of minimizing the total compensation capacity, a solution is found by combining genetic algorithms and particle swarm optimization. This reduces equipment investment costs and improves resource utilization efficiency while meeting harmonic limits.

[0019] Simplify project implementation: The solution covers the entire process of data collection, parameter extraction, responsibility quantification, capacity allocation, and effect verification. It has a clear logic and strong operability, making it easy to integrate with the existing power distribution network dispatching system and reducing the difficulty of project implementation. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the harmonic compensation capacity allocation method based on feeder harmonic responsibility involved in this invention.

[0022] Figure 2 This is a schematic diagram of the harmonic liability coefficient calculation process involved in this invention.

[0023] Figure 3 This is a schematic diagram of the compensation capacity allocation model and solution process involved in this invention.

[0024] Figure 4 This is a schematic diagram of the reinforcement learning and PSO collaborative optimization process involved in this invention. Detailed Implementation

[0025] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the embodiments of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0026] The following disclosure provides many different implementations or examples for carrying out different structures of the embodiments of the present invention. To simplify the disclosure of the embodiments of the present invention, specific examples of components and arrangements are described below. Of course, these are merely examples and are not intended to limit the embodiments of the present invention. Furthermore, reference numerals and / or reference letters may be repeated in different examples of the embodiments of the present invention; such repetition is for simplification and clarity and does not in itself indicate a relationship between the various implementations and / or arrangements discussed.

[0027] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0028] like Figure 1 As shown, this embodiment provides a harmonic compensation capacity allocation method based on feeder harmonic responsibility, including: S1: Collect the operating data of each feeder in the distribution network. The operating data includes the voltage time-domain signal, current time-domain signal, impedance parameters and system topology of each feeder. S2: Process the voltage time-domain signal and the current time-domain signal to extract the voltage amplitude and current amplitude of each harmonic of each feeder; S3: Based on the current amplitude, voltage amplitude, and impedance parameters of each harmonic, calculate the harmonic responsibility factor of each feeder under each harmonic. S4: Obtain the permissible voltage limits of the distribution network system under each harmonic; S5: Based on the harmonic responsibility coefficient and voltage allowable limit, establish a harmonic compensation capacity allocation model. The model aims to minimize the total compensation capacity and includes the constraint that the harmonic voltage after compensation does not exceed the voltage allowable limit. S6: The harmonic compensation capacity allocation model is solved using an optimization algorithm to obtain the compensation capacity of each feeder under each harmonic.

[0029] In some embodiments, the harmonic compensation capacity allocation method based on feeder harmonic responsibility further includes S7: based on the system topology and impedance parameters, verify the compensation effect corresponding to the compensation capacity, and determine whether the voltage of each harmonic after compensation meets the voltage allowable limit; if not, feed the deviation back to step S5 to adjust the model constraint conditions, and repeat steps S5 to S7.

[0030] In some embodiments, the harmonic compensation capacity allocation method based on feeder harmonic responsibility further includes S8: if the compensation effect is verified, the compensation capacity of each feeder under each harmonic is summarized to obtain the total harmonic compensation capacity of each feeder.

[0031] In some embodiments, the harmonic compensation capacity allocation method based on feeder harmonic responsibility further includes S9: periodically repeating steps S1 to S8 to update the total harmonic compensation capacity according to newly acquired operating data.

[0032] In some embodiments, in step S3, the harmonic responsibility coefficient is obtained by weighted fusion of initial coefficient and correction coefficient. The initial coefficient is calculated by the current superposition method, specifically: the initial coefficient of a feeder at a certain harmonic order is proportional to the product of the amplitude of the harmonic current and the impedance modulus of that feeder, and inversely proportional to the sum of the products of the amplitude of the harmonic current and the corresponding impedance modulus of all feeders; the correction coefficient is obtained by correcting the initial coefficient using the least squares method, which aims to minimize the sum of the squared errors between the calculated and measured values ​​of the harmonic voltage at each moment.

[0033] In some embodiments, in step S5, the constraint conditions are dynamically corrected by a neural network. The input of the neural network includes the harmonic responsibility coefficient, the harmonic voltage amplitude before compensation, and the voltage allowable limit. The output is a correction coefficient, which is used to adjust the weight of the harmonic responsibility coefficient in the constraint conditions.

[0034] In some embodiments, in step S6, the optimization algorithm is a fusion algorithm of genetic algorithm and particle swarm optimization algorithm; the genetic algorithm is used to generate an initial compensation capacity solution that satisfies the constraints, and the particle swarm optimization algorithm is used to refine the initial compensation capacity solution. During the refinement process, the particle velocity and position update formulas are iteratively optimized until the solution with the minimum total compensation capacity is obtained.

[0035] In some embodiments, the weights of the weighted fusion are determined based on the feeder load stability. The smaller the load fluctuation, the greater the weight of the correction coefficient; the greater the load fluctuation, the greater the weight of the initial coefficient. The value of the weights ranges from 0.6 to 0.9.

[0036] In some embodiments, the process of generating the initial compensation capacity solution by the genetic algorithm includes: encoding the compensation capacity of each harmonic of each feeder into chromosomes, selecting the parent generation by roulette wheel selection, generating offspring generation by arithmetic crossover and Gaussian mutation operations, iterating a preset number of generations, and selecting the solution with the highest fitness as the initial solution, wherein the fitness function is inversely proportional to the total compensation capacity.

[0037] The harmonic compensation capacity allocation system based on feeder harmonic responsibility, applying any one of the above-described harmonic compensation capacity allocation methods based on feeder harmonic responsibility, comprises: The data acquisition module is used to collect the operating data of each feeder in the distribution network. The operating data includes the voltage time-domain signal, current time-domain signal, impedance parameters and system topology of each feeder. The harmonic parameter extraction module is used to process the voltage time-domain signal and the current time-domain signal to extract the voltage amplitude and current amplitude of each harmonic of each feeder. The feeder harmonic responsibility factor calculation module is used to calculate the harmonic responsibility factor of each feeder under each harmonic based on the current amplitude, voltage amplitude and impedance parameters of each harmonic. The system harmonic permissible limit determination module is used to obtain the voltage permissible limit of the distribution network system under each harmonic. The harmonic compensation capacity allocation model construction module is used to establish a harmonic compensation capacity allocation model based on the harmonic responsibility coefficient and the voltage allowable limit. The model aims to minimize the total compensation capacity and includes the constraint that the harmonic voltage after compensation does not exceed the voltage allowable limit. The compensation capacity allocation model solution module is used to solve the harmonic compensation capacity allocation model using an optimization algorithm to obtain the compensation capacity of each feeder under each harmonic. The specific ideas behind this invention are as follows: The harmonic compensation capacity allocation algorithm based on feeder harmonic responsibility is implemented through the following steps: Collect voltage, current, impedance, and system topology data for each feeder, and extract key harmonic parameters; The feeder harmonic responsibility factor is calculated and corrected by combining the current superposition method and the least squares method, and the contribution of each feeder to harmonics is accurately quantified. Based on the national standard harmonic limits, an allocation model is constructed with the goal of minimizing the total compensation capacity, and a neural network is introduced to dynamically correct the constraints. The model is solved by combining genetic algorithm and particle swarm optimization to obtain the compensation capacity of each harmonic of each feeder; The compensation effect is verified through simulation. When the standard is exceeded, the model is adjusted based on feedback. Finally, the total compensation capacity scheme is formed and updated dynamically on a regular basis.

[0038] S1: Data Acquisition Data acquisition is the foundation of the entire harmonic compensation capacity allocation algorithm. It is crucial to ensure the accuracy, real-time performance, and completeness of the data to provide reliable input for subsequent steps such as harmonic parameter extraction and liability factor calculation. The specific steps are as follows: 1. Data collection objects and locations Voltage and current harmonic data: Harmonic monitoring devices (such as high-precision harmonic analyzers) are installed at the input and output ends of each feeder to collect the time-domain signals of three-phase voltage and current in real time. The acquisition frequency is set to 20 milliseconds / time (i.e., 100 points are collected per cycle in a 50Hz system) to ensure coverage of the main frequency components of harmonics (3rd to 25th harmonics).

[0039] Feeder impedance parameters: including the resistance of each feeder. (Unit: Ω) and reactance (Unit: Ω), frequency characteristics need to be considered (impedance values ​​differ at different harmonic orders, such as the impedance magnitude of the h-th harmonic). The data is retrieved from the distribution network system database, which is updated daily and records the nameplate parameters and actual operation and maintenance values ​​of each feeder.

[0040] System topology: including the connection relationship between feeders (such as radial or ring network structure), transformer location and capacity, load type (industrial, commercial, residential), etc., obtained from the distribution network GIS system to ensure that the topology relationship is consistent with the actual operating status.

[0041] 2. Data Preprocessing The collected voltage and current signals are initially screened to remove outliers caused by sensor malfunctions (such as data that exceed the normal range by 3 times the standard deviation). The monitoring data of each feeder are synchronized by timestamp to ensure that the harmonic parameters at the same time can be used for subsequent superposition analysis.

[0042] 3. Core Role This step provides the raw data support for all subsequent calculations, and the data quality directly affects the accuracy of the liability coefficient calculation and compensation capacity allocation. For example, if there is a 2% error in the current signal acquisition of a feeder, it may cause the harmonic liability coefficient of that feeder to deviate by more than 5%, thereby affecting the rationality of the compensation scheme.

[0043] S2: Harmonic parameter extraction Based on the voltage and current time-domain signals acquired by S1, key parameters of each harmonic are extracted using signal processing techniques, providing fundamental data for calculating the harmonic liability coefficient by S3. The specific process is as follows: 1. Preprocessing The original time-domain signal is denoised: wavelet transform denoising method is used, the db4 wavelet basis is selected, decomposed into 5 layers, and high-frequency noise (such as pulse interference) is filtered out by threshold function (soft threshold) to improve the signal-to-noise ratio to more than 30dB.

[0044] 2. Harmonic parameter calculation The denoised signal is analyzed in the frequency domain using Fast Fourier Transform (FFT): Window settings: Select a 1024-point FFT window with a corresponding time length of 20.48 milliseconds (covering one complete 50Hz cycle) to ensure a frequency resolution of 0.0488Hz, which can accurately distinguish adjacent harmonics (such as the 5th 250Hz and the 7th 350Hz).

[0045] Parameter extraction: The voltage amplitudes of each harmonic (h=3,5,7,...,25) were obtained by FFT calculation. (Unit: V) Voltage Phase (Unit: °), Current Amplitude (Unit: A) Current Phase (Unit: °). For example, the result of processing feeder 1 might be: 3rd harmonic voltage amplitude. =4.2V, phase =30°; Current amplitude =6A, phase =-15°.

[0046] 3. Data Validation The extracted harmonic parameters are compared with those of a standard signal source (such as a known signal output from a harmonic generator) to ensure that the amplitude error is ≤2% and the phase error is ≤5°. The results are output after the verification is passed.

[0047] 4. Core Role This step transforms the original time-domain signal into harmonic parameters that can be directly used for calculation, and is a crucial link connecting data acquisition and liability factor calculation. Accurate harmonic amplitude and phase ensure the calculation accuracy of the current superposition method and least squares method in S3, laying the foundation for the rationality of the liability factor.

[0048] S3: Feeder Harmonic Liability Factor Calculation (Harmonic liability factor calculation process is as follows) Figure 2 (As shown) 3.1 Calculation of Initial Liability Factor (Current Superposition Method) In distribution network harmonic analysis, the contribution of each feeder to system harmonics can be initially measured by the interaction between its injected harmonic current and its own impedance. The core logic of the current superposition method is that the harmonic responsibility of a feeder is proportional to the product of the amplitude of its injected harmonic current and the magnitude of its impedance—this is because harmonic currents generate harmonic voltage drops when flowing through the feeder impedance, and these voltage drops directly reflect the feeder's "active contribution" to system harmonics. Based on this, the amplitude of each harmonic current of each feeder is extracted using S2. (k is the feeder number, h is the harmonic order) and the feeder impedance magnitude obtained from S1. The formula for calculating the initial liability coefficient is: ; The superscript “(0)” in the middle indicates the initial value to distinguish it from subsequent correction values; n is the total number of all feeders in the distribution network (e.g., if a distribution network contains 4 feeders, then n=4). and These correspond to the h-th harmonic current amplitude and impedance magnitude of the i-th feeder, respectively.

[0049] Core Function: This formula quickly quantifies the initial harmonic responsibility ratio of each feeder, providing a basic framework for subsequent compensation allocation. For example, if under the 3rd harmonic, feeder 1... =6A、 =1.5Ω, the other 3 feeders If the values ​​are 10, 8, and 6 respectively, then... =(6×1.5) / (9+10+8+6)=9 / 33≈0.273, that is, the initial responsibility of feeder 1 for the 3rd harmonic is about 27.3%.

[0050] 3.2 Responsibility Coefficient Correction (Least Squares Optimization) In the operation of a distribution network, the system impedance may dynamically change due to factors such as transformer switching and load fluctuations, leading to deviations in the initial responsibility factor obtained solely through the current superposition method. To correct this deviation, the least squares method is introduced: using the actual harmonic voltage generated by each feeder as a benchmark, the responsibility factor is optimized by minimizing the error between the theoretically calculated and measured values. The error function is constructed as follows: ; is the sum of squares of voltage error under the hth harmonic; m is the total number of sampling times (usually 200 consecutive times are taken to ensure coverage of a complete load fluctuation cycle). The measured value of the h-th harmonic voltage of feeder k at time t; The measured value of the h-th harmonic voltage of feeder i at time t; This represents the responsibility coefficient to be corrected.

[0051] Model training process: 1. Training data: Select the harmonic voltage data collected by S1 for the first 72 hours (one sampling point every 10 minutes, a total of 432 data points) to ensure coverage of typical operating conditions such as peak, flat and low periods.

[0052] 2. Optimization process: The correction amount is iteratively adjusted using the gradient descent method. Each iteration calculates right The partial derivatives are updated along the negative gradient direction. until (Convergence threshold).

[0053] 3. Correction Result: The final corrected responsibility coefficient is: .

[0054] Core contribution: By minimizing voltage error, the influence of system impedance fluctuations is eliminated, making the liability factor closer to the actual harmonic contribution. For example, if the initial... =0.273, after correction =0.025, then =0.298, which more accurately reflects the actual responsibility of feeder 1.

[0055] 3.3 Fusion Output To balance the stability of the initial value and the accuracy of the correction value, a weighted fusion is performed on both. Weights

[0056] Determined based on feeder load type: Industrial feeders (load fluctuation ≤5%): =0.85 (The correction value has a higher weight because the voltage data is more reliable due to the stable load). Commercial feeders (5% < load fluctuation ≤ 15%): =0.75; Residential feeder (load fluctuation >15%): =0.65 (The initial value has a higher weight because the voltage data fluctuates greatly, and the correction is easily affected by interference).

[0057] The fusion formula is: ; The superscript "*" in the upper right corner indicates the final fusion value, which will be used as input for subsequent steps.

[0058] Output example: If feeder 1 is an industrial feeder ( =0.85), then =0.85×0.298+0.15×0.273≈0.294. This value will directly affect the allocation weight of compensation capacity in S5.

[0059] S4: Determine the permissible limits for system harmonics The system harmonic allowable limit is the basis for judging whether the compensation effect meets the standard. It must be determined strictly according to national standards and actual system requirements, providing a benchmark for constructing constraints in S5 and verifying the compensation effect in S7. The specific operation is as follows: 1. Standard Basis According to the regulations on harmonics in the public power grid: Total harmonic distortion (THD) limits at the point of common coupling: 4% for 10kV systems and 5% for 380V systems.

[0060] The voltage content limits for each harmonic are as follows: 2.4% for the 3rd harmonic in a 10kV system, 1.9% for the 5th harmonic, 1.5% for the 7th harmonic, and so on (the specific values ​​need to be determined by referring to the table according to the system voltage level).

[0061] 2. Limit value conversion Converting harmonic content to a specific voltage value: Let the system nominal voltage be... (Line voltage RMS value), then the permissible limit for the h-th harmonic voltage. (Regarding phase voltage). For example, a 10kV system ( The third harmonic limit for (e.g., 10000V) is calculated as follows: ≈138.56V (phase voltage).

[0062] 3. Adjustments for special scenarios If there are sensitive loads (such as precision machine tools or medical equipment) in the power distribution network, the limits need to be tightened according to the load requirements. For example, the third harmonic limit of a hospital feeder can be adjusted to 100V to avoid abnormal equipment operation.

[0063] 4. Core Role This step determines It is the core parameter of the S5 constraint, which directly determines the lower limit of the compensation capacity (the voltage after compensation must be ≤ the limit value). It is also the judgment standard for S7 to verify the compensation effect, ensuring that the final solution meets the power quality requirements.

[0064] S5: Harmonic Compensation Capacity Allocation Model (Integrating Neural Networks and Constrained Optimization) 5.1 Model Construction (Objective Function and Constraints) The core objective of the model is to minimize the total compensation capacity (reducing equipment costs) while ensuring harmonic compliance. This is achieved by combining the output of S3. and S4 (H-th harmonic voltage allowable limit), construct the objective function: ; Total system compensation capacity (unit: var); The h-th harmonic compensation capacity of feeder k; m is the harmonic order to be controlled (e.g., 3rd, 5th, 7th, 11th, a total of 4 orders, then m=4).

[0065] Objective and significance: To achieve economic optimization while meeting power quality standards by minimizing total capacity.

[0066] The constraints must ensure that the harmonic voltage does not exceed the limit after compensation, as follows: 1. Voltage limit constraints: ; To compensate for the h-th harmonic voltage of the feeder k; To compensate for the voltage before (extracted from S2); This represents the rated capacity of feeder k (unit: kVA, from the S1 system database). The physical meaning of the formula is: compensation capacity. Through its responsibility coefficient The reduction in voltage, as converted to rated capacity Related (the larger the capacity, the smaller the voltage drop for the same compensation).

[0067] Example: Feeder 1 =1000kVA, =4.2V, =3V, =0.294, then the constraint condition is 4.2- Solving for ≈4082var.

[0068] 2. Non-negativity constraint: (The compensation capacity cannot be negative to avoid reverse harmonic injection).

[0069] 5.2 Neural Network Weight Optimization (Improving Constraint Adaptability) Since the relationship between the responsibility factor and voltage drop may change under different operating conditions (such as peak / valley load), a BP neural network is introduced to dynamically correct the weight of the responsibility factor.

[0070] Network structure: Input layer has 3 nodes ( , , The hidden layer has 2 layers (12 neurons per layer, with ReLU activation function), and the output layer has 1 node (with correction coefficient). ).

[0071] Model training process: 1. Training set: Historical data from the past 6 months, containing 1200 samples (each sample contains...) , , After actual compensation ).

[0072] 2. Loss function: ,in This is the voltage value calculated according to the original constraints.

[0073] 3. Training process: Using the Adam optimizer (learning rate 0.001), iterate for 800 rounds until... (Ensure that the error between the calculated value and the measured value is small enough).

[0074] Revised constraints: ; Core function: Dynamically adjust according to real-time operating conditions (such as during peak load periods). (Increasing and strengthening the influence of the responsibility factor on voltage drop) makes the constraints more realistic. For example, during peak hours, feeder 1... =1.1, then the corrected constraint is 4.2- Solving for var (more suitable for peak operating conditions).

[0075] S6: Solving the Compensation Capacity Allocation Model (Integrating Genetic Algorithm and Particle Swarm Optimization) 6.1 Initial Solution Generation of Genetic Algorithm Genetic algorithms find the global optimal solution by simulating the biological evolution process (selection, crossover, mutation), and are suitable for quickly generating initial solutions that satisfy constraints.

[0076] Population initialization: Generate 60 feasible solutions (per group) The modified constraints and non-negativity constraints must be met. For example, for the third harmonic of feeder 1, randomly generated... Values ​​within the range [3711var, 5000var].

[0077] Fitness function: F = 1 / (The smaller the total capacity, the higher the fitness), which is directly related to the objective function of S5.

[0078] Selection operation: The roulette wheel method selects 30 pairs of parents (individuals with high fitness have a higher probability of being selected).

[0079] Crossover operation: Arithmetic crossover (probability 0.7), such as parent generation. =4000var =4500var, offspring =0.6×4000+0.4×4500=4200var.

[0080] Mutation operation: Gaussian mutation (probability 0.06), for... Add a mean of 0 and a standard deviation of 0.08× A random number (e.g., 4200var may become 4150var after mutation).

[0081] Iteration Termination: After 60 iterations, the solution with the highest fitness is selected as the initial optimal solution. (The superscript "(g)" indicates the result of the genetic algorithm). For example, feeder 1's... =3850var.

[0082] 6.2 Particle Swarm Optimization (PSO) Refinement Particle swarm optimization improves the accuracy of solutions by simulating the foraging behavior of bird flocks and performing a localized, refined search around the initial solution.

[0083] Parameter settings: Number of particles = 60 (consistent with the genetic algorithm population), initial position is Inertial weight =0.7 (balancing global and local search), learning factor = =1.8 (Guiding particles to move towards the individual / global optimal).

[0084] Velocity and position update formulas: ; ; For the particle at time t Dimensional speed; This is the particle's own historical optimal position; The globally optimal position (satisfying the constraints and) Minimum); , It is a random number in the range [0,1].

[0085] Iterative process: Each iteration updates the velocity and position, and re-evaluates. and After 120 iterations, a refined solution was obtained. (The superscript "(pso)" indicates the particle swarm optimization result). For example, feeder 1's... =3780var.

[0086] 6.3 Fusion Output The final compensation capacity is the result of PSO refinement: ; Output correlation: This result is directly derived from S3. and S5 Decide-- The larger the initial solution The larger; The larger it is, the more refined it will be. The closer to the lower limit of the constraint, the smaller the total capacity.

[0087] Example: Feeder 1 =3780var, which satisfies the corrected constraint (after compensation). (≈3V), which also minimizes the total capacity.

[0088] Summary of Algorithm Interaction and Enumeration 1. Interaction Logic: S3 Directly inputting S5 affects the calculation of voltage drop in the constraint conditions; S5's Through neural network training and The connections form a dynamic correction loop.

[0089] The objective function and constraints of S5 provide the optimization direction for S6. The initial solution of the genetic algorithm and the refined solution of PSO in S6 must satisfy the constraints of S5 to ensure that the results are compliant.

[0090] 2. Complete enumeration (taking the 3rd harmonic h=3 and feeder k=1 as an example): S3: Initial =0.273→Correction =0.298→ Fusion =0.294 (industrial feeder) =0.85).

[0091] S5: =4.2V, =3V, =1000kVA, =1.1 → Constraint ≥3711var.

[0092] S6: Initial genetic solution =3850var→PSO Refinement =3780var (satisfies the constraints and has the minimum total capacity).

[0093] The compensation capacity allocation model and solution process are as follows: Figure 3 As shown in the diagram, the above steps achieve full-process coordination of harmonic responsibility quantification, compensation model optimization, and solution, ensuring that the allocation of compensation capacity is both scientific and reasonable as well as economical and efficient.

[0094] In some embodiments, S7: Compensation effect verification The simulation verifies whether the compensation capacity output by S6 can reduce the system harmonic voltage to within the allowable limit. If it does not meet the requirement, feedback adjustments are made to ensure the feasibility of the solution. The specific process is as follows: 1. Simulation Model Construction Based on the system topology and impedance parameters of S1, a distribution network simulation model was built in PSCAD / EMTDC: Each feeder is equivalent to a resistor-inductor series circuit (parameters taken from S1). and The load is set according to the actual type (e.g., industrial load is inductive, and residential load is resistive-capacitive).

[0095] Compensation device (such as SVG) model: adopts a current source control strategy, based on the output of S6. Calculate the required injected harmonic compensation current (Simplified calculation ignoring phase effects), and set the response time to ≤10 milliseconds.

[0096] 2. Simulation and Data Acquisition The simulation time was set to 10 seconds (including a 2-second transient process and an 8-second steady-state process), and the harmonic voltages of each feeder after compensation were collected. and the total harmonic distortion (THD) of the system. For example, the voltage after compensation for the third harmonic of feeder 1. =130V, needs to be connected to S4 =138.56V comparison.

[0097] 3. Verification Standards and Feedback If all subharmonics satisfy If THD ≤ standard limit, then the verification passes and proceeds to S8.

[0098] If there is an excess (such as the 5th harmonic) Then calculate the deviation from the standard. =10V, feedback to S5: In the compensation capacity allocation model, add the constraint weight of this harmonic (e.g., adjust the original weight 1.0 to 1.2) so that the model prioritizes satisfying the limit of this harmonic when solving, and then re-execute S5 to S7.

[0099] 4. Core Role This step is a "trial and error" phase before the solution is implemented. Simulations help identify deficiencies in the compensation capacity allocation in advance, preventing harmonic exceedances after actual operation. The feedback mechanism ensures the model dynamically adapts to system characteristics, improving the robustness of the solution.

[0100] In some embodiments, S8: Determine the final compensation capacity After S7 verification is passed, the harmonic compensation capacity of each feeder is summarized to form the final allocation scheme, providing a basis for the selection and commissioning of the compensation device. The specific operation is as follows: 1. Capacity Summary The harmonic compensation capacity of the S6 output Summing these values ​​yields the total harmonic compensation capacity of each feeder: ; For example, the third compensation capacity of feeder 1 =5000var, 5 times =3000var, 7 times =2000var, then the total capacity =5000+3000+2000=10000var=10kvar.

[0101] 2. Redundancy considerations To cope with sudden harmonics in the system (such as transient harmonics generated during motor startup), a 10% redundancy is added to the total capacity: For example, the final operational capacity of feeder 1 is 10 × 1.1 = 11 kvar.

[0102] 3. Solution Output This will generate a list including the total compensation capacity of each feeder, details of harmonic distribution, and recommendations for compensation device models (e.g., based on...). Select the final solution document (15kvar SVG) and submit it to the power distribution network operation and maintenance department.

[0103] 4. Core Role This step integrates the dispersed harmonic compensation capacities into an executable total capacity scheme, clarifies the compensation requirements of each feeder, and serves as a bridge connecting algorithm calculations with practical engineering applications.

[0104] In some embodiments, S9: Dynamic adjustment The operating status of the distribution network changes dynamically with load variations (such as differences in load between morning and evening peak hours), requiring regular updates to the compensation capacity allocation scheme to ensure long-term compliance with harmonic mitigation requirements. Specific procedures are as follows: 1. Adjustment cycle and triggering conditions Regular execution: Steps S1 to S8 are automatically started at the 10th minute of each hour to collect the latest voltage and current data (only real-time data of changing feeders are collected to reduce the amount of calculation).

[0105] Trigger update: When the newly calculated total compensation capacity With the currently executing relative difference When this happens, a scheme update is triggered. For example, if the current capacity of feeder 2 is 8kvar, and the new calculation shows 8.5kvar, the difference is 6.25% > 5%, requiring an update.

[0106] 2. Update Process After the new scheme is generated, the update command is sent to the compensation device through the distribution network dispatch system, and the device completes the parameter adjustment (such as the current command update of SVG) within 1 minute.

[0107] Record the solution adjustment log, including the adjustment time, reason, and before and after capacity comparison, for subsequent operation and maintenance analysis.

[0108] 3. Core Role This step enables the compensation scheme to be dynamically adaptable, which can cope with harmonic changes caused by load fluctuations and equipment switching, ensuring that the distribution network can maintain qualified power quality under different operating conditions and avoiding the risk of harmonic exceedance caused by long-term use of fixed schemes.

[0109] Summary of data flow at each step Voltage, current, impedance, and topology data from S1 → Harmonic parameters extracted from S2 ( → S3 calculates the responsibility coefficient ( ).

[0110] S4 With S3 →S5 constructs the compensation model (including neural network correction) →S6 solution .

[0111] S6 →S7 Simulation Verification: If the result exceeds the limit, feedback is sent to S5 to adjust the model. →After successful verification, S8 summarizes the results. →S9 updates its plans regularly.

[0112] In some embodiments, a reinforcement learning (RL) dynamic adjustment factor is introduced between the neural network correction in S5 and the particle swarm optimization (PSO) in S6. This allows the inertia weights of PSO to change in real time with the correction coefficients output by the neural network, achieving deep synergy between "constraint correction → optimization parameter adjustment → solution accuracy improvement," thus solving the problem that fixed inertia weights are difficult to adapt to different working conditions. The reinforcement learning and PSO co-optimization process is as follows: Figure 4 As shown.

[0113] 1. Reinforcement Learning Model Construction 1.1 State Space Define state The key parameter combination at time t directly relates the output of S5 to the input of S6: ; This is the final harmonic liability factor output by S3; These are the correction coefficients for the output of the S5 neural network; To compensate for the preceding harmonic voltage; This is the permissible limit for S4; Let t be the total compensation capacity for t iterations.

[0114] 1.2 Action Space Define Action Adjustment amount for PSO inertia weight: ; The PSO inertial weight at time t (originally a fixed value of 0.7); To adjust the step size, the value range is [-0.1, 0.1], meaning that each adjustment does not exceed 0.1.

[0115] 1.3 Reward Function award Used to evaluate the quality of a movement, taking into account both compensation effectiveness and cost-effectiveness: ; Let be the voltage after compensation following t iterations; =0.6 is the voltage compliance weight; =0.4 is the capacity cost weight; if ,but =-10 (Severe penalty).

[0116] 2. Model Training Process The reinforcement learning model is trained using a deep Q-network (DQN): Experience replay pool: stores historical state-action-reward data. , , , A total of 5,000 samples were collected (from compensation cases over the past three months).

[0117] 2. Network Structure: Input layer has 5 nodes (corresponding to...) The output layer has 5 parameters, 2 hidden layers (16 neurons / layer, ReLU activation), and 1 output node (Q value, corresponding to the action). (value).

[0118] 3. Training iteration: Each time, 32 samples are randomly selected from the replay pool, and the target Q value is calculated: ( =0.9 is the discount factor. (Target network parameters).

[0119] Loss function: ( (These are the current network parameters).

[0120] Optimizer: Adam, learning rate 0.0005, 10000 iterations, synchronizing target network and current network parameters every 100 steps until the loss function converges to < .

[0121] 3. Algorithm interaction and formula integration 3.1 Formula for Dynamic Update of Inertia Weight The action of reinforcement learning output Substituting the inertial weight update of PSO, a closed-loop interaction with S5 is formed: ; ; Correction coefficients for S5 output The normalization function (to Mapped to [0,1]); =0.5、 =1.5 represents the historical extreme value of the neural network output.

[0122] Interaction logic: The larger the value (the stronger the constraint). The closer to 1, The stronger the influence, the more significant the adjustment of inertia weights, accelerating the convergence of PSO to a solution that satisfies strong constraints.

[0123] 3.2 Collaborative Solving of PSO and Reinforcement Learning Revised PSO velocity update formula (incorporating reinforcement learning results): ; Core function: Through Related to S5 The solution process of S6, when Increase (e.g., during peak load periods). Reduce (enhance local search) to make the solution closer to the lower constraint limit; when Reduce (e.g., when the load is stable). Increase (enhance global search) to explore smaller total capacity.

[0124] 4. Examples of enumeration (continuation of the 3rd harmonic h=3, feeder 1k=1) 1. S5 Output: =0.294, =1.1 (higher than the mean of 1.0, indicating stronger constraints). =4.2V, =3V.

[0125] 2. Reinforcement learning input state: =[0.294,1.1,4.2,3,12000var] (Initial total capacity).

[0126] 3. Action output: DQN evaluation results =-0.05 (Reduce inertia weight to enhance local search).

[0127] 4. Inertia weight update: , .

[0128] 5. PSO Refinement: The new velocity formula causes particles to move closer to the lower constraint limit after iteration. =3780var (a decrease of 50var compared to before optimization), total capacity reduced to 11800var, while =2.98V≤3V, reward ≈0.012-0.472=-0.46 (Although negative, it is close to the constraint, avoiding penalty).

[0129] 5. Core Contributions 1. Cross-step collaboration: The S5 neural network is modified through reinforcement learning ( The PSO solution for S6 () This direct correlation solves the limitation of fixed parameters in traditional optimization algorithms.

[0130] 2. Dynamic adaptability: The system automatically adjusts and optimizes strategies based on real-time operating conditions, prioritizing compliance when constraints are tightened and cost reduction when constraints are relaxed, thus balancing power quality and economy.

[0131] 3. Improved accuracy: In the example, the total compensation capacity is reduced while ensuring that the voltage meets the standard, which verifies the effectiveness of the fusion algorithm.

[0132] This optimization creates a closed loop between S5 and S6, consisting of "constraint correction → parameter adjustment → solution optimization," further improving the accuracy and adaptability of compensation capacity allocation.

[0133] The embodiments described above are for illustrative purposes only and are not intended to limit the invention. Therefore, any changes in numerical values ​​or substitutions of equivalent elements should still fall within the scope of this invention.

[0134] The above detailed description will enable those skilled in the art to understand that the present invention can indeed achieve the aforementioned objectives and has complied with the provisions of the Patent Law.

[0135] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention. The above descriptions are merely preferred embodiments of the invention and are not intended to limit the invention. It should be noted that any modifications, equivalent substitutions, and improvements made within the spirit and principles of the invention should be included within the scope of protection of the invention.

[0136] It should be noted that the above description of the process is for illustrative purposes only and does not limit the scope of this specification. Those skilled in the art can make various modifications and changes to the process under the guidance of this specification. However, these modifications and changes remain within the scope of this specification.

[0137] The basic concepts have been described above. Obviously, for those skilled in the art who have read this application, the above disclosure is merely illustrative and does not constitute a limitation of this application. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this application. Such modifications, improvements, and corrections are suggested in this application, and therefore, such modifications, improvements, and corrections still fall within the spirit and scope of the exemplary embodiments of this application.

[0138] Furthermore, this application uses specific terms to describe its embodiments. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic related to at least one embodiment of this application. Therefore, it should be emphasized and noted that "an embodiment," "one embodiment," or "an alternative embodiment" mentioned twice or more in different positions in this specification do not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of this application can be appropriately combined.

[0139] Furthermore, those skilled in the art will understand that aspects of this application can be described and illustrated through several patentable types or situations, including any new and useful combination of processes, machines, products, or substances, or any new and useful improvements thereof. Therefore, aspects of this application can be implemented entirely in hardware, entirely in software (including firmware, resident software, microcode, etc.), or a combination of hardware and software. All of the above hardware or software can be referred to as a “unit,” “module,” or “system.” Furthermore, aspects of this application can take the form of a computer program product embodied in one or more computer-readable media, wherein computer-readable program code is contained therein.

[0140] Furthermore, unless expressly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or other names described in this application are not intended to limit the order of the processes and methods of this application. Although some currently considered useful embodiments of the invention have been discussed in the foregoing disclosure by way of various examples, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments; rather, the claims are intended to cover all modifications and equivalent combinations that conform to the substance and scope of the embodiments of this application. For example, although the implementation of the various components described above can be embodied in a hardware device, it can also be implemented as a purely software solution, such as an installation on an existing server or mobile device.

[0141] Similarly, it should be noted that, in order to simplify the description of the present application and thus aid in the understanding of one or more embodiments of the invention, the foregoing description of the embodiments of the present application sometimes combines multiple features into a single embodiment, drawing, or description thereof. However, this approach of the present application should not be construed as reflecting an intention that the claimed subject matter requires more features than expressly recited in each claim. Rather, the subject of the invention should possess fewer features than in any single embodiment described above.

Claims

1. A harmonic compensation capacity allocation method based on feeder harmonic responsibility, characterized in that, include: S1: Collect the operating data of each feeder in the distribution network. The operating data includes the voltage time-domain signal, current time-domain signal, impedance parameters and system topology of each feeder. S2: Process the voltage time-domain signal and the current time-domain signal to extract the voltage amplitude and current amplitude of each harmonic of each feeder; S3: Based on the current amplitude, voltage amplitude, and impedance parameters of each harmonic, calculate the harmonic responsibility factor of each feeder under each harmonic. S4: Obtain the permissible voltage limits of the distribution network system under each harmonic; S5: Based on the harmonic responsibility coefficient and voltage allowable limit, establish a harmonic compensation capacity allocation model. The model aims to minimize the total compensation capacity and includes the constraint that the harmonic voltage after compensation does not exceed the voltage allowable limit. S6: The harmonic compensation capacity allocation model is solved by using an optimization algorithm to obtain the compensation capacity of each feeder under each harmonic.

2. The harmonic compensation capacity allocation method based on feeder harmonic responsibility according to claim 1, characterized in that, It also includes S7: Based on the system topology and impedance parameters, verify the compensation effect corresponding to the compensation capacity, and determine whether the voltage of each harmonic after compensation meets the voltage allowable limit; if not, feed the deviation back to step S5 to adjust the model constraint conditions, and repeat steps S5 to S7.

3. The harmonic compensation capacity allocation method based on feeder harmonic responsibility according to claim 2, characterized in that, It also includes S8: If the compensation effect is verified, the compensation capacity of each feeder under each harmonic is summarized to obtain the total harmonic compensation capacity of each feeder.

4. The harmonic compensation capacity allocation method based on feeder harmonic responsibility according to claim 3, characterized in that, It also includes S9: periodically repeating steps S1 to S8 to update the total harmonic compensation capacity based on the newly collected operating data.

5. The harmonic compensation capacity allocation method based on feeder harmonic responsibility according to claim 1, characterized in that, In step S3, the harmonic responsibility coefficient is obtained by weighted fusion of the initial coefficient and the correction coefficient. The initial coefficient is calculated by the current superposition method, specifically: the initial coefficient of a feeder at a certain harmonic order is proportional to the product of the amplitude of the harmonic current and the impedance modulus of that feeder, and inversely proportional to the sum of the products of the amplitude of the harmonic current and the corresponding impedance modulus of all feeders. The correction coefficient is obtained by correcting the initial coefficient using the least squares method, which aims to minimize the sum of the squared errors between the calculated and measured values ​​of the harmonic voltage at each moment.

6. The harmonic compensation capacity allocation method based on feeder harmonic responsibility according to claim 1, characterized in that, In step S5, the constraint conditions are dynamically corrected through a neural network. The input of the neural network includes the harmonic responsibility coefficient, the harmonic voltage amplitude before compensation, and the voltage allowable limit. The output is a correction coefficient, which is used to adjust the weight of the harmonic responsibility coefficient in the constraint conditions.

7. The harmonic compensation capacity allocation method based on feeder harmonic responsibility according to claim 1, characterized in that, In step S6, the optimization algorithm is a fusion algorithm of genetic algorithm and particle swarm optimization algorithm; the genetic algorithm is used to generate an initial compensation capacity solution that satisfies the constraints, and the particle swarm optimization algorithm is used to refine the initial compensation capacity solution. During the refinement process, the particle velocity and position update formulas are iteratively optimized until the solution with the minimum total compensation capacity is obtained.

8. The harmonic compensation capacity allocation method based on feeder harmonic responsibility according to claim 5, characterized in that, The weights of the weighted fusion are determined based on the stability of the feeder load. The smaller the load fluctuation, the greater the weight of the correction coefficient; the greater the load fluctuation, the greater the weight of the initial coefficient. The value of the weights ranges from 0.6 to 0.

9.

9. The harmonic compensation capacity allocation method based on feeder harmonic responsibility according to claim 7, characterized in that, The process of generating the initial compensation capacity solution by the genetic algorithm includes: encoding the compensation capacity of each harmonic of each feeder into chromosomes, selecting the parent generation by roulette wheel selection, generating offspring generation by arithmetic crossover and Gaussian mutation operations, iterating through a preset number of generations, and selecting the solution with the highest fitness as the initial solution. The fitness function is inversely proportional to the total compensation capacity.

10. A harmonic compensation capacity allocation system based on feeder harmonic responsibility, characterized in that, The harmonic compensation capacity allocation method based on feeder harmonic responsibility according to any one of claims 1 to 9, wherein the harmonic compensation capacity allocation system based on feeder harmonic responsibility comprises: The data acquisition module is used to collect the operating data of each feeder in the distribution network. The operating data includes the voltage time-domain signal, current time-domain signal, impedance parameters and system topology of each feeder. The harmonic parameter extraction module is used to process the voltage time-domain signal and the current time-domain signal to extract the voltage amplitude and current amplitude of each harmonic of each feeder. The feeder harmonic responsibility factor calculation module is used to calculate the harmonic responsibility factor of each feeder under each harmonic based on the current amplitude, voltage amplitude and impedance parameters of each harmonic. The system harmonic permissible limit determination module is used to obtain the voltage permissible limit of the distribution network system under each harmonic. The harmonic compensation capacity allocation model construction module is used to establish a harmonic compensation capacity allocation model based on the harmonic responsibility coefficient and the voltage allowable limit. The model aims to minimize the total compensation capacity and includes the constraint that the harmonic voltage after compensation does not exceed the voltage allowable limit. The compensation capacity allocation model solution module is used to solve the harmonic compensation capacity allocation model using an optimization algorithm to obtain the compensation capacity of each feeder under each harmonic.

Images