Power distribution network harmonic control method and device, computer device

Abstract

This application relates to a method, apparatus, and computer equipment for harmonic mitigation in distribution networks. The method includes: first, for each node in the distribution network, obtaining a risk factor characterizing the degree of harmonic exceedance risk at each node, calculating a risk weighting value; based on this risk weighting value, the worst-case node risk penalty term, and the fixed investment cost of voltage-sensing active power filters at rated capacity and the operation and maintenance cost corresponding to the actual compensation capacity, constructing a comprehensive objective function while setting constraints; using the virtual harmonic conductance of each voltage-sensing active power filter at each node of the distribution network as optimization variables, iteratively solving the constructed comprehensive objective function using an improved sine and cosine algorithm to obtain the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active power filter; and finally, configuring the harmonic conductance parameters of the voltage-sensing active power filters in the distribution network based on these optimal values. This method can accurately and efficiently achieve harmonic suppression and safe and stable operation of the distribution network.

Description

Technical Field

[0001] This application relates to the field of power distribution network harmonic mitigation technology, and in particular to a power distribution network harmonic mitigation method, device, and computer equipment. Background Technology

[0002] With the accelerated integration of new power elements such as controllable loads, distributed generation, energy storage systems, and electric vehicles into the distribution network, and the large-scale grid connection of renewable energy sources such as wind and photovoltaics, as well as the widespread adoption of power electronic interfaces in terminal load equipment, the distribution network is showing a trend of deep integration of power electronic equipment in all aspects of the source-grid-load chain. It is rapidly evolving towards "high proportion of renewable energy access" and "high proportion of power electronic equipment application". In order to adapt to the construction of new power systems, it is urgent to build a large-capacity distribution network with strong adaptability.

[0003] Traditional harmonic mitigation methods for distribution networks primarily target centralized harmonic sources, employing point-to-point compensation. However, in today's electronic power distribution networks, background harmonic sources exhibit new characteristics such as high density, multi-point distribution, network-wide propagation, and weakened single-source power. Traditional point-to-point compensation methods are insufficient to effectively manage dispersed and minute harmonics. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, device, computer equipment, computer-readable storage medium, and computer program product for harmonic control in power distribution networks that combines a voltage-sensing active filter to effectively suppress harmonics, in order to address the aforementioned technical problems.

[0005] Firstly, this application provides a method for harmonic mitigation in power distribution networks, including:

[0006] For nodes in the distribution network, based on the node's voltage harmonic distortion rate and upper limit of voltage harmonic distortion rate, an over-limit risk factor is obtained to characterize the harmonic over-limit risk level of the node; based on the over-limit risk factor, the over-limit risk weight coefficient of the node is obtained.

[0007] Based on the node's over-limit risk weight coefficient and voltage harmonic distortion rate, a risk weighting value is obtained; based on the risk weighting value, the worst node risk penalty term, the fixed investment cost of the rated capacity voltage-sensing active filter, and the operation and maintenance cost of the actual compensation capacity voltage-sensing active filter, a comprehensive objective function for harmonic control of the distribution network is constructed, and fundamental power flow constraints, harmonic power flow constraints, and compensation capacity limit constraints of the voltage-sensing active filter are set.

[0008] The virtual harmonic conductance of each voltage-sensing active filter at each node of the distribution network is used as the optimization variable. The preset distribution network parameters and harmonic source parameters are input, and the comprehensive objective function of harmonic control is iteratively solved by the improved sine and cosine algorithm to obtain the optimal value of each virtual harmonic conductance of the voltage-sensing active filter.

[0009] Configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

[0010] In one embodiment, the process of obtaining the voltage harmonic distortion rate of a node includes:

[0011] The sum of squares of the voltage amplitudes of each harmonic that the node needs to control is calculated, and then the arithmetic square root of the sum of squares is calculated. The ratio between the arithmetic square root and the fundamental voltage amplitude of the node is calculated to obtain the voltage harmonic distortion rate of the node.

[0012] In one embodiment, the excess risk weight coefficient of a node is obtained based on the excess risk factor, including:

[0013] The risk sensitivity index is raised to the power of the risk factor of each node to obtain the calculation result of each node; the risk sensitivity index is used to adjust the strength of the weighting coefficient of the risk factor of high-risk nodes.

[0014] Obtain the first sum of the calculation results of each node; calculate the ratio between the calculation result of each node and the first sum to obtain the over-limit risk weight coefficient of the corresponding node.

[0015] In one embodiment, the process of obtaining the fixed investment cost of the rated capacity voltage-sensing active filter and the operation and maintenance cost of the actual compensation capacity voltage-sensing active filter includes:

[0016] Based on the over-limit risk weight coefficient of each node, nodes with over-limit risk weight coefficients greater than a preset value are selected from all nodes to obtain a set of installation nodes for voltage-sensing active filters. The rated capacity of the voltage-sensing active filter of each node in the set of installation nodes is multiplied by the investment cost coefficient per unit capacity to obtain the first product result of each node. The first product results of each node are summed to obtain the fixed investment cost of the rated capacity voltage-sensing active filter.

[0017] The actual compensation capacity of the voltage-sensing active filter at each node in the set of voltage-sensing active filter installation nodes is multiplied sequentially by the percentage coefficient of operation and maintenance costs to fixed investment costs and the investment cost coefficient per unit capacity to obtain the second product result for each node; the second product results of each node are summed to obtain the operation and maintenance cost of the voltage-sensing active filter with actual compensation capacity.

[0018] In one embodiment, a comprehensive objective function for harmonic mitigation of the distribution network is constructed based on risk weighting, worst-case node risk penalty, fixed investment cost of voltage-sensing active power filters with rated capacity, and operation and maintenance cost of voltage-sensing active power filters with actual compensation capacity. This function includes:

[0019] Set a governance effect ratio coefficient for the risk weighting value, set a penalty item ratio coefficient for the worst node risk penalty item, and set an economic cost ratio coefficient for the second sum between the fixed investment cost of the voltage detection active filter with rated capacity and the operation and maintenance cost of the voltage detection active filter with actual compensation capacity.

[0020] The comprehensive objective function for harmonic control of the distribution network is obtained by weighting and summing the risk weight, the worst-case node risk penalty, and the second sum based on the governance effect ratio coefficient, the penalty coefficient, and the economic cost ratio coefficient.

[0021] In one embodiment, an improved sine and cosine algorithm is used to incorporate a stride parameter adjustment strategy based on a population diversity index; the implementation process of the stride parameter adjustment strategy includes:

[0022] For nodes whose risk weight coefficient exceeds the preset value, the virtual harmonic conductance of the voltage-sensing active filter of the node is used as the algorithm variable dimension. The sum of squares of the deviations between the individual positions and the population mean under the algorithm variable dimension is calculated. The arithmetic mean of the sum of squares of the deviations is calculated and then the corresponding arithmetic square root is calculated. The difference between the preset adjustable minimum value and the adjustable maximum value under the algorithm variable dimension is obtained, and the ratio between the arithmetic square root and the difference is calculated.

[0023] Calculate the arithmetic mean of the ratios under each virtual harmonic conductance as the solution value of the population diversity index in the current iteration; calculate the ratio between the solution value of the population diversity index in the current iteration and the maximum solution value of the population diversity index in the entire iteration; obtain the adaptive function value based on the ratio.

[0024] The stride parameter is adjusted based on the adaptive function value; the stride parameter is positively correlated with the adaptive function value.

[0025] In one embodiment, the improved sine and cosine algorithms are integrated with a population initialization strategy based on back learning; the implementation process of the population initialization strategy includes:

[0026] For nodes whose risk weight coefficient exceeds the preset value, the virtual harmonic conductance of the voltage-sensing active filter of the node is used as the algorithm variable dimension. The difference between the preset adjustable minimum value and the adjustable maximum value under the algorithm variable dimension is obtained. The difference is multiplied by a random number from 0 to 1 and then added to the adjustable minimum value to obtain a single initial value under the algorithm variable dimension.

[0027] Combine the individual initial values ​​corresponding to each virtual harmonic conductance to form a single initial solution, and construct a random initial population composed of multiple initial solutions; subtract the initial value of a single initial solution in the random initial population under the corresponding algorithm variable dimension from the sum of the adjustable minimum value and the adjustable maximum value to obtain a single inverse value under the algorithm variable dimension.

[0028] Combine the single inverse values ​​corresponding to each virtual harmonic conductance to form a single inverse solution, and construct an inverse population composed of multiple inverse solutions; merge the random initial population and the inverse population to obtain a candidate population; based on the comprehensive objective function of harmonic governance, calculate the fitness values ​​of all solutions in the candidate population, and select the top preset number of solutions with the best fitness values ​​as the initial population for algorithm iteration, with the preset number being the preset initial population size.

[0029] In one embodiment, the improved sine and cosine algorithm incorporates an elite solution mutation strategy that combines Cauchy and Gaussian mixed perturbations; the implementation process of the elite solution mutation strategy includes:

[0030] In the iterative process of the improved sine and cosine algorithm, the fitness values ​​of all solutions in the current iteration population are calculated based on the comprehensive objective function of harmonic control. The solution with the best fitness value is selected as the current elite solution, which is a combination of the corresponding values ​​of each virtual harmonic conductance.

[0031] Obtain the random perturbation terms of the standard Cauchy distribution and the standard Gaussian distribution, calculate the ratio between the current iteration number and the maximum iteration number of the algorithm, subtract the ratio from 1 as the weight coefficient of the Cauchy distribution perturbation term, and use the ratio as the weight coefficient of the Gaussian distribution perturbation term.

[0032] The Cauchy distribution perturbation term and the Gaussian distribution perturbation term are multiplied by their corresponding weight coefficients and summed. The summation result is then multiplied by the value of the current elite solution in each dimension of the algorithm variable to obtain the variation value in each dimension of the algorithm variable. The variation values ​​are then combined to form the candidate elite solution.

[0033] The fitness values ​​of candidate elite solutions are calculated based on the comprehensive objective function of harmonic control. The fitness values ​​of candidate elite solutions are compared with the fitness values ​​of the current elite solutions, and the solution with the smallest fitness value is selected as the elite solution for the next iteration.

[0034] Secondly, this application also provides a power distribution network harmonic mitigation device, comprising:

[0035] The acquisition module is used to acquire, for nodes in the distribution network, an over-limit risk factor that characterizes the harmonic over-limit risk level of the node based on the node's voltage harmonic distortion rate and voltage harmonic distortion rate upper limit; and acquire the over-limit risk weight coefficient of the node based on the over-limit risk factor.

[0036] A module is constructed to obtain the risk weighting value based on the node's over-limit risk weighting coefficient and voltage harmonic distortion rate; based on the risk weighting value, the worst node risk penalty term, the fixed investment cost of the rated capacity voltage sensing active filter, and the operation and maintenance cost of the actual compensation capacity voltage sensing active filter, a comprehensive objective function for harmonic control of the distribution network is constructed.

[0037] The solution module is used to take the virtual harmonic conductance of each voltage-sensing active filter at each node of the distribution network as optimization variables, input the preset distribution network parameters and harmonic source parameters, and iteratively solve the comprehensive objective function of harmonic control through an improved sine and cosine algorithm to obtain the optimal value of each virtual harmonic conductance of the voltage-sensing active filter.

[0038] The configuration module is used to configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

[0039] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0040] For nodes in the distribution network, based on the node's voltage harmonic distortion rate and upper limit of voltage harmonic distortion rate, an over-limit risk factor is obtained to characterize the harmonic over-limit risk level of the node; based on the over-limit risk factor, the over-limit risk weight coefficient of the node is obtained.

[0041] Based on the node's over-limit risk weight coefficient and voltage harmonic distortion rate, a risk weighting value is obtained; based on the risk weighting value, the worst-case node risk penalty term, the fixed investment cost of the rated capacity voltage-sensing active filter, and the operation and maintenance cost of the actual compensation capacity voltage-sensing active filter, a comprehensive objective function for harmonic control of the distribution network is constructed.

[0042] The virtual harmonic conductance of each voltage-sensing active filter at each node of the distribution network is used as the optimization variable. The preset distribution network parameters and harmonic source parameters are input, and the comprehensive objective function of harmonic control is iteratively solved by the improved sine and cosine algorithm to obtain the optimal value of each virtual harmonic conductance of the voltage-sensing active filter.

[0043] Configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

[0044] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0045] For nodes in the distribution network, based on the node's voltage harmonic distortion rate and upper limit of voltage harmonic distortion rate, an over-limit risk factor is obtained to characterize the harmonic over-limit risk level of the node; based on the over-limit risk factor, the over-limit risk weight coefficient of the node is obtained.

[0046] Based on the node's over-limit risk weight coefficient and voltage harmonic distortion rate, a risk weighting value is obtained; based on the risk weighting value, the worst-case node risk penalty term, the fixed investment cost of the rated capacity voltage-sensing active filter, and the operation and maintenance cost of the actual compensation capacity voltage-sensing active filter, a comprehensive objective function for harmonic control of the distribution network is constructed.

[0047] The virtual harmonic conductance of each voltage-sensing active filter at each node of the distribution network is used as the optimization variable. The preset distribution network parameters and harmonic source parameters are input, and the comprehensive objective function of harmonic control is iteratively solved by the improved sine and cosine algorithm to obtain the optimal value of each virtual harmonic conductance of the voltage-sensing active filter.

[0048] Configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

[0049] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0050] For nodes in the distribution network, based on the node's voltage harmonic distortion rate and upper limit of voltage harmonic distortion rate, an over-limit risk factor is obtained to characterize the harmonic over-limit risk level of the node; based on the over-limit risk factor, the over-limit risk weight coefficient of the node is obtained.

[0051] Based on the node's over-limit risk weight coefficient and voltage harmonic distortion rate, a risk weighting value is obtained; based on the risk weighting value, the worst-case node risk penalty term, the fixed investment cost of the rated capacity voltage-sensing active filter, and the operation and maintenance cost of the actual compensation capacity voltage-sensing active filter, a comprehensive objective function for harmonic control of the distribution network is constructed.

[0052] The virtual harmonic conductance of each voltage-sensing active filter at each node of the distribution network is used as the optimization variable. The preset distribution network parameters and harmonic source parameters are input, and the comprehensive objective function of harmonic control is iteratively solved by the improved sine and cosine algorithm to obtain the optimal value of each virtual harmonic conductance of the voltage-sensing active filter.

[0053] Configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

[0054] The aforementioned methods, devices, computer equipment, computer-readable storage media, and computer program products for harmonic mitigation in distribution networks first determine the over-limit risk factor, characterizing the degree of harmonic over-limit risk at each node of the distribution network based on the node voltage harmonic distortion rate and its upper limit, and accordingly determine the node over-limit risk weight coefficient. By introducing the over-limit risk factor and risk weight coefficient, the focus is on mitigating nodes with higher harmonic over-limit risks, balancing the overall harmonic mitigation effect with the risk control of key nodes. Then, a risk weighted value is calculated by combining the node over-limit risk weight coefficient and the voltage harmonic distortion rate. Based on this risk weighted value, the worst-case node risk penalty term, and the fixed investment cost of the voltage-sensing active filter at rated capacity and the operation and maintenance cost corresponding to the actual compensation capacity, a comprehensive objective function for distribution network harmonic mitigation is constructed. The addition of the worst-case node risk penalty term effectively avoids extreme harmonic over-limit situations. Integrating the mitigation effect with the investment and operation and maintenance costs of the voltage-sensing active filter to construct the comprehensive objective function optimizes equipment investment and operating costs while ensuring harmonic mitigation performance, thus improving the economic efficiency of the solution. Using the virtual harmonic conductances of voltage-sensing active power filters at each node of the distribution network as optimization variables, and inputting preset distribution network parameters and harmonic source parameters, an improved sine and cosine algorithm is used to iteratively solve the constructed comprehensive objective function to obtain the optimal values ​​of the virtual harmonic conductances of the voltage-sensing active power filters. Finally, the harmonic conductance parameters of the voltage-sensing active power filters in the distribution network are configured based on these optimal values. The improved sine and cosine algorithm is used for iterative solution, which has stronger solution accuracy and global optimization capability. It can quickly obtain the optimal parameters of the virtual harmonic conductances of the voltage-sensing active power filters. Based on the optimal values, the equipment parameters can be configured to accurately and efficiently achieve harmonic suppression and safe and stable operation of the distribution network, and improve the rationality and reliability of harmonic management in the distribution network. Attached Figure Description

[0055] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 This is an application environment diagram of the power distribution network harmonic mitigation method in one embodiment;

[0057] Figure 2 This is a flowchart illustrating a method for controlling harmonics in a power distribution network in one embodiment;

[0058] Figure 3 Here is a flowchart of the improved SCA algorithm solution in another embodiment;

[0059] Figure 4 This is a structural block diagram of a power distribution network harmonic mitigation device in one embodiment;

[0060] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0062] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0063] The power distribution network harmonic mitigation method provided in this application embodiment can be applied to, for example... Figure 1The application environment shown can involve only terminal 102, only server 104, or both terminal 102 and server 104. Terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or located in the cloud or on other network servers. Specifically, terminal 102 or server 104 completes a method for harmonic mitigation in a distribution network. This method includes: first, for each node in the distribution network, based on the node voltage harmonic distortion rate and its upper limit, obtaining an exceedance risk factor characterizing the degree of harmonic exceedance risk at the node, and determining the node exceedance risk weight coefficient accordingly; then, combining the node exceedance risk weight coefficient with the voltage harmonic distortion rate to calculate a risk weighted value; based on this risk weighted value, the worst-case node risk penalty term, and the fixed investment cost of the voltage-sensing active filter at rated capacity and the operation and maintenance cost corresponding to the actual compensation capacity, constructing a comprehensive objective function for harmonic mitigation in the distribution network; using the virtual harmonic conductance of the voltage-sensing active filter at each node in the distribution network as optimization variables, inputting preset distribution network parameters and harmonic source parameters, and using an improved sine / cosine algorithm to iteratively solve the constructed comprehensive objective function to obtain the optimal value of each virtual harmonic conductance of the voltage-sensing active filter; finally, configuring the harmonic conductance parameters of the voltage-sensing active filter in the distribution network based on this optimal value.

[0064] Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, power monitoring host computers, distribution automation terminals, substation monitoring equipment, and power grid edge computing terminals. Distribution automation terminals can be fault recording devices, harmonic monitoring terminals, distribution network data acquisition terminals, smart distribution terminals, power grid status monitoring devices, etc. Power grid edge computing terminals can be local protection devices, smart gateway terminals, distribution IoT controllers, etc. Smart distribution terminals can be voltage monitoring terminals, harmonic mitigation controllers, voltage detection active filter control devices, active filter controllers, etc. Server 104 can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0065] In one exemplary embodiment, such as Figure 2 As shown, a method for harmonic mitigation in power distribution networks is provided, which is then applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps 202 to 208. Wherein:

[0066] Step 202: For nodes in the distribution network, based on the node's voltage harmonic distortion rate and the upper limit of voltage harmonic distortion rate, obtain the over-limit risk factor used to characterize the harmonic over-limit risk level of the node; based on the over-limit risk factor, obtain the over-limit risk weight coefficient of the node.

[0067] Among them, voltage harmonic distortion rate (THD) is the ratio of the root mean square value of each harmonic voltage RMS value to the RMS value of the fundamental voltage in the bus voltage of a distribution network node, expressed as a percentage. It is a core indicator for measuring the degree of harmonic pollution of node voltage. The over-limit risk factor characterizes the... The quantitative factor for the harmonic exceedance risk level of each node reflects the degree to which the current voltage harmonic distortion rate of the node deviates from the threshold, and is the core basis for calculating the exceedance risk weight coefficient. The exceedance risk weight coefficient is a coefficient obtained by further normalization / weighting based on the exceedance risk factor, and is used to distinguish the harmonic mitigation priority of different nodes. The larger the coefficient, the higher the harmonic exceedance risk of the node, and the more likely it is to be equipped with voltage detection active filters.

[0068] Step 204: Obtain the risk weighting value based on the node's over-limit risk weighting coefficient and voltage harmonic distortion rate; construct the comprehensive objective function for harmonic control of the distribution network based on the risk weighting value, the worst node risk penalty term, the fixed investment cost of the rated capacity voltage detection active filter, and the operation and maintenance cost of the actual compensation capacity voltage detection active filter.

[0069] Among them, the worst-case node risk penalty is a quantitative penalty measure set to prevent individual nodes from exceeding harmonic limits for a long period of time, even though the overall harmonic risk of the distribution network has decreased. Nodes with more severe exceedances will trigger higher penalty values ​​to ensure the balance of harmonic control across the entire network. The fixed investment cost of voltage-sensing active power filters is the one-time purchase and deployment cost based on the rated capacity of the equipment when configuring voltage-sensing active power filters in the distribution network. It is the basic economic cost indicator for configuring voltage-sensing active power filters. The operation and maintenance cost of voltage-sensing active power filters is the operation and maintenance cost linked to the actual harmonic compensation capacity of the voltage-sensing active power filters. The compensation capacity changes dynamically with the harmonic conditions of the power grid, and the operation and maintenance cost is also adjusted accordingly. The comprehensive objective function for harmonic control is a mathematical function constructed with the core objectives of achieving the optimal harmonic control effect of the distribution network and minimizing the economic cost of the voltage-sensing active power filters throughout their entire life cycle. It is the core basis for subsequent algorithms to solve for the optimal configuration scheme of voltage-sensing active power filters.

[0070] Specifically, constructing the objective function requires setting fundamental power flow constraints, harmonic power flow constraints, and compensation capacity limit constraints for voltage-sensing active filters.

[0071] Fundamental power flow constraints refer to the active and reactive power balance rules that the fundamental power transmission in a distribution network must meet. These are fundamental constraints to ensure the safe and stable operation of the distribution network's fundamental power. Harmonic power flow constraints refer to the admittance matrix operation rules that the transmission of harmonic voltages and currents in a distribution network must meet. These rules ensure that the distribution of harmonic power flow in the grid conforms to electrical operating principles during harmonic compensation. Compensation capacity limit constraints are the maximum harmonic compensation capacity thresholds set for voltage-sensing active filters. These limits prevent damage caused by overcapacity operation of equipment while reserving a reasonable capacity margin to cope with harmonic disturbances in the grid.

[0072] Step 206: Using the virtual harmonic conductance of each voltage-sensing active filter at each node of the distribution network as optimization variables, inputting the preset distribution network parameters and harmonic source parameters, and iteratively solving the comprehensive objective function of harmonic control through the improved sine and cosine algorithm to obtain the optimal value of each virtual harmonic conductance of the voltage-sensing active filter.

[0073] Among them, virtual harmonic conductance refers to the equivalent conductance parameter of a voltage-sensing active power filter (VPS) used to compensate for harmonics at distribution network nodes. It is the core quantitative indicator for harmonic compensation control of VPS, with independent virtual harmonic conductance values ​​corresponding to different harmonic orders. Distribution network parameters are fundamental parameters characterizing the topology and electrical characteristics of the distribution network, including core data such as the number of nodes, bus admittance matrix, fundamental voltage / power, and electrical connections between nodes. Harmonic source parameters refer to parameters such as the location of each harmonic source in the distribution network, the amplitude of injected harmonic current / voltage, and the harmonic order distribution. These are key bases for calculating the harmonic power flow and compensation requirements of the VPS.

[0074] Optionally, the upper-level system for harmonic mitigation in the distribution network first uses the virtual harmonic conductances corresponding to the 5th, 7th, 11th, and 13th orders of key harmonics from the voltage-sensing active filters at each node as core optimization variables. Simultaneously, it retrieves basic parameters of the distribution network, such as the network topology, bus admittance, and fundamental power, as well as harmonic source parameters, including the location of each harmonic source, harmonic injection amplitude, and harmonic order distribution, as algorithm inputs. Subsequently, the upper-level system calls an improved sine and cosine algorithm, using the previously constructed constrained comprehensive objective function for harmonic mitigation as the algorithm's solution objective. The algorithm initializes the population with optimization variables and performs multiple iterations to update population positions, adjust population diversity step parameters, and perform elite solution mutation and selection retention for Cauchy and Gaussian mixed perturbations. In each iteration, the solution is checked for fundamental power flow, harmonic power flow, and voltage-sensing active filter compensation capacity limit constraints, and invalid solutions are eliminated. When the algorithm meets the iteration termination condition, the host system outputs the solution that optimizes the comprehensive objective function, which is the optimal value of each virtual harmonic conductance of the voltage-sensing active filter at each node of the distribution network.

[0075] Step 208: Configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

[0076] Among them, the harmonic conductance parameter refers to the core operating control parameter of the voltage-sensing active filter. It directly determines the amplitude and phase of the compensation current injected into the distribution network by the voltage-sensing active filter. It is a key hardware configuration parameter for achieving accurate harmonic compensation and corresponds one-to-one with the optimal value of virtual harmonic conductance.

[0077] The aforementioned method for harmonic mitigation in distribution networks firstly identifies a risk factor for each node, based on its voltage harmonic distortion rate and upper limit. This risk factor characterizes the degree of harmonic exceedance risk at each node, and a node exceedance risk weighting coefficient is determined accordingly. By introducing the exceedance risk factor and risk weighting coefficient, the method prioritizes the mitigation of nodes with higher harmonic exceedance risks, balancing overall harmonic mitigation effectiveness with risk control at critical nodes. Next, a risk weighting value is calculated by combining the node exceedance risk weighting coefficient and the voltage harmonic distortion rate. Based on this risk weighting value, the worst-case node risk penalty term, and the fixed investment cost of voltage-sensing active power filters at rated capacity and the operation and maintenance cost corresponding to the actual compensation capacity, a comprehensive objective function for distribution network harmonic mitigation is constructed. The inclusion of the worst-case node risk penalty term effectively avoids extreme harmonic exceedance situations. Integrating the mitigation effect with the investment and operation and maintenance costs of voltage-sensing active power filters to construct a comprehensive objective function optimizes equipment investment and operating costs while ensuring harmonic mitigation performance, thereby improving the economic efficiency of the solution. Using the virtual harmonic conductances of voltage-sensing active power filters at each node of the distribution network as optimization variables, and inputting preset distribution network parameters and harmonic source parameters, an improved sine and cosine algorithm is used to iteratively solve the constructed comprehensive objective function to obtain the optimal values ​​of the virtual harmonic conductances of the voltage-sensing active power filters. Finally, the harmonic conductance parameters of the voltage-sensing active power filters in the distribution network are configured based on these optimal values. The improved sine and cosine algorithm is used for iterative solution, which has stronger solution accuracy and global optimization capability. It can quickly obtain the optimal parameters of the virtual harmonic conductances of the voltage-sensing active power filters. Based on the optimal values, the equipment parameters can be configured to accurately and efficiently achieve harmonic suppression and safe and stable operation of the distribution network, and improve the rationality and reliability of harmonic management in the distribution network.

[0078] In one embodiment, the process of obtaining the voltage harmonic distortion rate of a node includes:

[0079] The sum of squares of the voltage amplitudes of each harmonic that the node needs to control is calculated, and then the arithmetic square root of the sum of squares is calculated. The ratio between the arithmetic square root and the fundamental voltage amplitude of the node is calculated to obtain the voltage harmonic distortion rate of the node.

[0080] Specifically, nodes Voltage harmonic distortion rate is defined as:

[0081]

[0082] In the formula, For nodes Voltage harmonic distortion rate, For nodes of Second harmonic voltage amplitude; Represents a node The amplitude of the fundamental voltage; This refers to the set of harmonics that need to be controlled.

[0083] In the above embodiments, the calculation process focuses only on the harmonic orders that need to be addressed, eliminating harmonic interference that does not require addressing, thus significantly improving calculation efficiency and reducing resource consumption for distribution network harmonic monitoring and calculation. Furthermore, through the standardized calculation process of sum of squares, arithmetic square roots, and ratios, the comprehensive distortion impact of each harmonic order on node voltage can be accurately quantified, avoiding the one-sidedness of single harmonic analysis and making the obtained voltage harmonic distortion rate more in line with the actual harmonic addressing needs.

[0084] In one embodiment, the excess risk weight coefficient of a node is obtained based on the excess risk factor, including:

[0085] The risk sensitivity index is raised to the power of the risk factor of each node to obtain the calculation result of each node; the risk sensitivity index is used to adjust the strength of the weighting coefficient of the risk factor of high-risk nodes.

[0086] Obtain the first sum of the calculation results of each node; calculate the ratio between the calculation result of each node and the first sum to obtain the over-limit risk weight coefficient of the corresponding node.

[0087] Specifically, nodes The normalized out-of-limit risk factor is:

[0088]

[0089] In the formula, For nodes The upper limit of THD.

[0090] The risk weighting coefficient for exceeding the limit is:

[0091]

[0092] In the formula, The number of system nodes, parameter It is a risk sensitivity index used to adjust the model's preference for high-risk nodes.

[0093] In the above embodiments, the introduction of exponentiation of the risk sensitivity index allows for flexible adjustment of the assignment intensity of high-risk nodes according to the actual operating needs of the distribution network. This not only highlights the governance priority of high-risk nodes and achieves precise allocation of governance resources, but also adjusts the assignment intensity according to the power grid operating conditions to adapt to different harmonic governance scenarios. At the same time, by performing network-wide summation and normalization on the results of the exponentiation, the weight coefficients of each node are kept within a unified quantification range, ensuring the comparability and overall coordination of the weight coefficients and avoiding resource allocation imbalances caused by excessively high or low weights of a single node.

[0094] In one embodiment, the process of obtaining the fixed investment cost of the rated capacity voltage-sensing active filter and the operation and maintenance cost of the actual compensation capacity voltage-sensing active filter includes:

[0095] Based on the over-limit risk weight coefficient of each node, nodes with over-limit risk weight coefficients greater than a preset value are selected from all nodes to obtain a set of installation nodes for voltage-sensing active filters. The rated capacity of the voltage-sensing active filter of each node in the set of installation nodes is multiplied by the investment cost coefficient per unit capacity to obtain the first product result of each node. The first product results of each node are summed to obtain the fixed investment cost of the rated capacity voltage-sensing active filter.

[0096] The actual compensation capacity of the voltage-sensing active filter at each node in the set of voltage-sensing active filter installation nodes is multiplied sequentially by the percentage coefficient of operation and maintenance costs to fixed investment costs and the investment cost coefficient per unit capacity to obtain the second product result for each node; the second product results of each node are summed to obtain the operation and maintenance cost of the voltage-sensing active filter with actual compensation capacity.

[0097] Specifically, the set of installation nodes is as follows:

[0098]

[0099] In the formula, As the risk threshold; when The closer to or greater than 1, the stronger the node. The closer a node is to or has exceeded the THD limit, the better. It has a higher governance priority.

[0100] The fixed investment cost of a voltage-sensing active filter is:

[0101]

[0102] In the formula, This refers to the investment cost coefficient per unit capacity of a voltage-sensing active filter. This represents the number of nodes where the voltage-sensing active filter is installed, and the rated capacity of the voltage-sensing active filter at each node.

[0103] Considering that the maintenance cost of a voltage-sensing active filter is related to its actual compensation capacity, the maintenance cost of a voltage-sensing active filter with actual compensation capacity is defined as follows:

[0104]

[0105] In the formula, It is the percentage coefficient of the operation and maintenance cost of voltage-sensing active filters in the fixed investment cost; This is the investment cost coefficient per unit capacity of a voltage-sensing active filter; It is a voltage-sensing active filter in the first... The actual compensation amount for each node.

[0106] Installed on node The actual compensation capacity of a voltage-sensing active filter can be calculated by combining the conductance of each harmonic order and the corresponding harmonic voltage:

[0107]

[0108] In the formula, For voltage sensing active filters Sub-virtual harmonic conductance; For voltage-sensing active filter nodes Secondary harmonic voltage amplitude.

[0109] In the above embodiments, the installation nodes of voltage-sensing active power filters are selected based on the risk weighting coefficient of exceeding limits. This ensures that cost accounting is focused solely on high-risk governance nodes, eliminating invalid calculations for nodes with no installation requirements. This approach ensures that cost investment is highly matched with harmonic governance needs, avoiding resource waste. Simultaneously, fixed investment costs are calculated based on rated capacity and unit capacity investment coefficients, while operation and maintenance costs are calculated by multiplying the actual compensation capacity, operation and maintenance cost ratio coefficient, and unit capacity investment coefficient. Both types of cost calculations rely on standardized coefficient multiplication and summation operations, resulting in clear logic and precise quantification. This accurately reflects the actual economic investment in the configuration and operation of voltage-sensing active power filters. Both types of costs are obtained by summing the calculation results for the installation nodes, achieving comprehensive integrated accounting of the cost of voltage-sensing active power filters at high-risk nodes across the entire network. This provides an accurate and reliable economic quantitative basis for subsequently constructing a comprehensive objective function that balances governance effectiveness and economic cost, helping to achieve the dual goals of optimal harmonic governance effectiveness and lowest economic cost.

[0110] In one embodiment, a comprehensive objective function for harmonic mitigation of the distribution network is constructed based on risk weighting, worst-case node risk penalty, fixed investment cost of voltage-sensing active power filters with rated capacity, and operation and maintenance cost of voltage-sensing active power filters with actual compensation capacity. This function includes:

[0111] Set a governance effect ratio coefficient for the risk weighting value, set a penalty item ratio coefficient for the worst node risk penalty item, and set an economic cost ratio coefficient for the second sum between the fixed investment cost of the voltage detection active filter with rated capacity and the operation and maintenance cost of the voltage detection active filter with actual compensation capacity.

[0112] The comprehensive objective function for harmonic control of the distribution network is obtained by weighting and summing the risk weight, the worst-case node risk penalty, and the second sum based on the governance effect ratio coefficient, the penalty coefficient, and the economic cost ratio coefficient.

[0113] Specifically, corresponding governance effect ratio coefficients, penalty ratio coefficients, and economic cost ratio coefficients are set for the risk weighting value, the worst-case node risk penalty term, and the sum of the fixed investment cost and operation and maintenance cost of the voltage detection active filter. Each term is multiplied by its corresponding ratio coefficient and then weighted and summed to construct a comprehensive objective function for distribution network harmonic governance that takes into account governance effect, risk prevention and control, and economic cost.

[0114] The overall objective function is expressed as:

[0115]

[0116] In the formula, This is a coefficient representing the proportion of treatment effectiveness. This is the penalty ratio coefficient. This is the economic cost ratio coefficient.

[0117] In the above embodiments, independent and adjustable proportional coefficients are configured for the three core items of governance effect, risk penalty, and economic cost. This allows the objective function to focus on different governance directions according to the actual operation needs of the power grid, greatly improving the versatility and scenario adaptability of the solution. At the same time, the three independent governance objectives are integrated into a unified function, avoiding the governance imbalance caused by pursuing a single objective. This ensures that the harmonic governance of the distribution network achieves good governance results, avoids extreme over-limit problems at individual nodes, and achieves reasonable control of economic costs.

[0118] In one embodiment, an improved sine and cosine algorithm is used to incorporate a stride parameter adjustment strategy based on a population diversity index; the implementation process of the stride parameter adjustment strategy includes:

[0119] For nodes whose risk weight coefficient exceeds the preset value, the virtual harmonic conductance of the voltage-sensing active filter of the node is used as the algorithm variable dimension. The sum of squares of the deviations between the individual positions and the population mean under the algorithm variable dimension is calculated. The arithmetic mean of the sum of squares of the deviations is calculated and then the corresponding arithmetic square root is calculated. The difference between the preset adjustable minimum value and the adjustable maximum value under the algorithm variable dimension is obtained, and the ratio between the arithmetic square root and the difference is calculated.

[0120] Calculate the arithmetic mean of the ratios under each virtual harmonic conductance, and use it as the solution value of the population diversity index generated in the current iteration; calculate the ratio between the solution value of the population diversity index generated in the current iteration and the maximum solution value generated by the population diversity index in the entire iteration; based on the ratio, obtain the adaptive function value;

[0121] The stride parameter is adjusted based on the adaptive function value; the stride parameter is positively correlated with the adaptive function value.

[0122] Specifically, for high-risk nodes whose risk weighting coefficient exceeds a preset value, the virtual harmonic conductance of each order of its voltage-sensing active filter is used as the algorithm variable dimension, assuming the first... The population size of the generation is The variable dimension is , No. The lower and upper bounds of the dimension are respectively and The normalized population diversity index is defined as:

[0123]

[0124] In the formula, For the first The mean of the population.

[0125] Then, the adaptive function value is calculated by comparing the current solution value with the global maximum solution value. Finally, the step parameter is adaptively adjusted based on the adaptive function value. When the value is large, the adaptive function value automatically increases to enhance global exploration; when When the value is reduced, the adaptive function value decreases to improve local expansion and convergence stability.

[0126] In the above embodiments, by introducing a step size parameter adjustment strategy based on population diversity index into the improved sine and cosine algorithm, the step size can be dynamically adjusted according to the actual distribution state of the population during the algorithm iteration process. This effectively balances the algorithm's global search capability and local exploitation capability, avoids premature convergence or getting trapped in local optima, significantly improves the solution accuracy and convergence speed for the optimal value of virtual harmonic conductance of voltage-sensing active filters, and only calculates for high-risk nodes, which not only ensures the optimization accuracy of key areas of harmonic mitigation but also reduces the computational load of the algorithm, making the overall optimization process more efficient and stable, and ultimately making the distribution network harmonic mitigation scheme more in line with actual operation requirements.

[0127] In one embodiment, the improved sine and cosine algorithms are integrated with a population initialization strategy based on back learning; the implementation process of the population initialization strategy includes:

[0128] For nodes whose risk weight coefficient exceeds the preset value, the virtual harmonic conductance of the voltage-sensing active filter of the node is used as the algorithm variable dimension. The difference between the preset adjustable minimum value and the adjustable maximum value under the algorithm variable dimension is obtained. The difference is multiplied by a random number from 0 to 1 and then added to the adjustable minimum value to obtain a single initial value under the algorithm variable dimension.

[0129] Combine the individual initial values ​​corresponding to each virtual harmonic conductance to form a single initial solution, and construct a random initial population composed of multiple initial solutions; subtract the initial value of a single initial solution in the random initial population under the corresponding algorithm variable dimension from the sum of the adjustable minimum value and the adjustable maximum value to obtain a single inverse value under the algorithm variable dimension.

[0130] Combine the single inverse values ​​corresponding to each virtual harmonic conductance to form a single inverse solution, and construct an inverse population composed of multiple inverse solutions; merge the random initial population and the inverse population to obtain a candidate population; based on the comprehensive objective function of harmonic governance, calculate the fitness values ​​of all solutions in the candidate population, and select the top preset number of solutions with the best fitness values ​​as the initial population for algorithm iteration, with the preset number being the preset initial population size.

[0131] Specifically, for high-risk nodes where the risk weight coefficient exceeds the preset value, the sub-virtual harmonic conductance of the voltage-sensing active filter is used as the algorithm variable dimension. A single initial value is generated using the upper and lower limits of the variable and a random number between 0 and 1. The initial values ​​of each dimension are combined to form an initial solution and construct a random initial population. Then, the corresponding initial value is obtained by subtracting the sum of the upper and lower limits of the variable. The reverse values ​​of each dimension are combined to form a reverse solution and construct a reverse population. The random initial population and the reverse population are merged to obtain a candidate population. Finally, the fitness value of each solution is calculated based on the comprehensive objective function of harmonic governance, and the top preset number of solutions with the best fitness are selected as the final initial population.

[0132] In the above embodiments, by introducing a population initialization strategy based on reverse learning, the search coverage of the initial solution can be significantly expanded, the quality of the initial solution can be improved, and the initial population can be effectively prevented from gathering in a local area too early, thus providing a better search starting point for subsequent algorithm iterations.

[0133] In one embodiment, the improved sine and cosine algorithm incorporates an elite solution mutation strategy that combines Cauchy and Gaussian mixed perturbations; the implementation process of the elite solution mutation strategy includes:

[0134] In the iterative process of the improved sine and cosine algorithm, the fitness values ​​of all solutions in the current iteration population are calculated based on the comprehensive objective function of harmonic control. The solution with the best fitness value is selected as the current elite solution, which is a combination of the corresponding values ​​of each virtual harmonic conductance.

[0135] Obtain the random perturbation terms of the standard Cauchy distribution and the standard Gaussian distribution, calculate the ratio between the current iteration number and the maximum iteration number of the algorithm, subtract the ratio from 1 as the weight coefficient of the Cauchy distribution perturbation term, and use the ratio as the weight coefficient of the Gaussian distribution perturbation term.

[0136] The Cauchy distribution perturbation term and the Gaussian distribution perturbation term are multiplied by their corresponding weight coefficients and summed. The summation result is then multiplied by the value of the current elite solution in each dimension of the algorithm variable to obtain the variation value in each dimension of the algorithm variable. The variation values ​​are then combined to form the candidate elite solution.

[0137] The fitness values ​​of candidate elite solutions are calculated based on the comprehensive objective function of harmonic control. The fitness values ​​of candidate elite solutions are compared with the fitness values ​​of the current elite solutions, and the solution with the smallest fitness value is selected as the elite solution for the next iteration.

[0138] Specifically, let the first The current optimal position obtained in the next iteration is: By applying a hybrid mutation, candidate elite solutions are obtained. :

[0139]

[0140] In the formula, This indicates multiplication by dimension; and These are the random perturbation terms of the standard Cauchy distribution and the standard Gaussian distribution, respectively.

[0141] The weighting coefficients are adaptively adjusted with the number of iterations as follows:

[0142]

[0143]

[0144] In the formula, T represents the maximum number of iterations.

[0145] The weighted mixed perturbation term is multiplied dimension-wise with the elite solution to obtain candidate elite solutions. Finally, the fitness values ​​of the two are compared and the better one is selected as the next generation elite solution, thus completing the adaptive mutation update of the elite solution.

[0146] In the above embodiments, by adopting an elite solution mutation strategy with mixed Cauchy and Gaussian perturbations, the search range can be expanded and local optima can be escaped in the early stage of iteration by leveraging the strong global perturbation capability of the Cauchy distribution. In the later stage of iteration, the fine perturbation capability of the Gaussian distribution can be used to achieve accurate local optimization. The exploration and development performance of the dynamic balancing algorithm is improved, the optimization accuracy and convergence stability are effectively enhanced, and the optimized voltage detection active filter parameters are closer to the global optimum, further improving the overall effect of harmonic control in the distribution network.

[0147] In one embodiment, a method for controlling harmonics in a distribution network is provided, comprising:

[0148] Taking the virtual harmonic conductances of voltage-sensing active filters connected in parallel on system nodes as the optimization object, a node over-limit risk factor is introduced to quantitatively characterize the harmonic over-limit risk level of each node, and a node weight coefficient is constructed accordingly. On this basis, considering the capacity investment cost and the operation and maintenance cost of the compensation capacity of the voltage-sensing active filter equipment, a distribution network harmonic governance objective function that takes into account both governance effectiveness and economy is constructed.

[0149] node The normalized out-of-limit risk factor is:

[0150] (1)

[0151] In the formula, For nodes Total Harmonic Distortion (THD); For nodes The upper limit of THD.

[0152] To ensure metric consistency, nodes Voltage harmonic distortion rate is defined as:

[0153] (2)

[0154] In the formula, For nodes of Second harmonic voltage amplitude; Represents a node The amplitude of the fundamental voltage; For the set of harmonics that need to be controlled, this invention takes... .

[0155] To achieve risk-oriented resource prioritization within limited governance capacity, node importance weights are defined. The weighting coefficient for the risk of exceeding the limit is:

[0156] (3)

[0157] In the formula, The number of system nodes, parameter It is a risk sensitivity index used to adjust the model's preference for high-risk nodes.

[0158] Furthermore, the set of installation nodes is as follows:

[0159] (4)

[0160] In the formula, As the risk threshold; when The closer to or greater than 1, the stronger the node. The closer a node is to or has exceeded the THD limit, the better. It has a higher governance priority.

[0161] Therefore, the risk-weighted value of the objective function can be obtained as follows:

[0162] (5)

[0163] By assigning higher optimization weights to nodes with high risk factors that exceed limits, governance resources are prioritized for high-risk areas, thereby improving the precise suppression of harmonic pollution. After weight normalization, coordinated optimization is achieved across the entire network, ensuring balanced resource allocation and synergistic improvement in overall system performance under risk-oriented guidance.

[0164] To prevent individual critical nodes from remaining in an over-limit state for an extended period while the overall system risk indicators decline, a worst-case node risk penalty term is introduced into the risk weighting value:

[0165] (6)

[0166] In the formula, This is the penalty coefficient. When the set of key nodes... When it is an empty set, we can let .

[0167] set up This is a set of installation nodes for voltage-sensing active filters. This refers to the number of nodes installed in a voltage-sensing active filter. For nodes The rated capacity of the voltage-sensing active filter is given. Therefore, the fixed investment cost of the voltage-sensing active filter is:

[0168] (7)

[0169] In the formula, This is the investment cost coefficient per unit capacity of a voltage-sensing active filter.

[0170] Considering that the maintenance cost of a voltage-sensing active filter is related to its actual compensation capacity, the maintenance cost of a voltage-sensing active filter with actual compensation capacity is defined as follows:

[0171] (8)

[0172] In the formula, It is the percentage coefficient of the operation and maintenance cost of voltage-sensing active filters in the fixed investment cost; This is the investment cost coefficient per unit capacity of a voltage-sensing active filter; It is a voltage-sensing active filter in the first... The actual compensation amount for each node.

[0173] Installed on node The actual compensation capacity of a voltage-sensing active filter can be calculated by combining the conductance of each harmonic order and the corresponding harmonic voltage:

[0174] (9)

[0175] In the formula, For voltage sensing active filters Sub-virtual harmonic conductance; For voltage-sensing active filter nodes Secondary harmonic voltage amplitude.

[0176] With the objective functions of minimizing the risk-weighted total harmonic distortion (THD) of all nodes in the distribution network and minimizing the investment, compensation capacity, and operation and maintenance costs of voltage sensing active power filters, the comprehensive objective function is expressed as follows:

[0177] (10)

[0178] In the formula, This is a coefficient representing the proportion of treatment effectiveness. This is the penalty ratio coefficient. This is the economic cost ratio coefficient.

[0179] The fundamental wave power flow equations are shown below, for each node. The fundamental active and reactive power balance is as follows:

[0180] (11)

[0181] In the formula, and These are the busbars Active power and reactive power; node Voltage at the point; and These are the bus in the system bus admittance matrix. and The electrical conductance and susceptance between them; For nodes and nodes The voltage phase difference.

[0182] No. The node voltage and current under subharmonics satisfy the following relationship:

[0183] (12)

[0184] In the formula, , and For each system The harmonic voltage, harmonic current, and harmonic admittance matrix.

[0185] The harmonic conductance of the voltage-sensing active filter is compensated as follows: The specific harmonic power flow constraints are as follows:

[0186] (13)

[0187] During harmonic compensation of a voltage-sensing active filter, the maximum compensation amount cannot exceed the capacity limit. The constraint formula is as follows:

[0188] (14)

[0189] In the formula, For the first The voltage-sensing active filter has a reserved capacity factor to ensure that the capacity of the voltage-sensing active filter is sufficient to compensate for harmonic disturbances.

[0190] Next, the optimal configuration model is solved based on the improved sine and cosine algorithm. An improved mechanism integrating multiple strategies is proposed, which introduces adaptive parameter adjustment based on population diversity, population initialization strategy based on back learning, and elite solution mutation strategy based on Cauchy and Gaussian mixed perturbation into the sine and cosine algorithm.

[0191] The comprehensive optimization configuration of voltage-sensing active power filters proposed in this invention is a nonlinear and nonconvex problem. The essence of solving this problem is to determine the optimal installation capacity of the voltage-sensing active power filters based on the installation nodes, in order to obtain an economical and long-term feasible power quality management scheme. To this end, this invention uses a sine-cosine algorithm to solve the constructed optimal configuration model.

[0192] (1) Basic sine and cosine algorithm

[0193] The Sine Cosine Algorithm (SCA) is a novel metaheuristic optimization algorithm proposed by Mirjalili. This algorithm leverages the periodic fluctuations of sine and cosine functions, combined with a stochastic parameter adjustment mechanism, to effectively balance global exploration and local development processes, thereby efficiently finding the optimal solution globally. Specifically, SCA first generates... A random initialization process is initiated, and the fitness values ​​of each solution are evaluated. Then, the current optimal solution is selected by comparing fitness values. Finally, an oscillating convergence mechanism constructed using sine and cosine functions is used to dynamically update the positions of all individuals, gradually approaching the global optimum. The individual update formula is as follows:

[0194] (15)

[0195] (16)

[0196] In the formula, Indicates the first In the nth iteration The individual in the first The position of the dimension ; for The optimal solution after the second iteration The position of the dimension; Given 3 random numbers that follow a uniform distribution, , , , For adaptive values, adaptive adjustment is achieved through formulas; It is a constant, usually taken as 2; This represents the current iteration number; This represents the maximum number of iterations.

[0197] (2) Improved sine and cosine algorithms

[0198] 1) Adaptive stride adjustment strategy based on population diversity. To avoid the patterned search behavior caused by relying solely on the fixed annealing rule of the number of iterations, an adaptive stride control mechanism based on population diversity is introduced to adjust key parameters in SCA. Dynamic adjustment is performed. Let the first... The population size of the generation is The variable dimension is , No. The lower and upper bounds of the dimension are respectively and The normalized population diversity index is defined as:

[0199] (17)

[0200] In the formula, For the first The mean of the population.

[0201] Further Designed as an adaptive function driven by population diversity feedback:

[0202] (18)

[0203] When the population is dispersed, Larger Automatically increase in size to enhance global exploration; as the population gradually converges... reduce, Adaptive reduction is used to improve local exploitation and convergence stability, thereby achieving a balance between exploration and exploitation at different search stages.

[0204] 2) Population Initialization Strategy for Opposition-Based Learning (OBL). Sine and cosine algorithms typically use a uniform random method to generate the initial population during the initialization phase. This can easily lead to uneven distribution of individuals and insufficient coverage, thereby reducing early global exploration efficiency and inducing convergence instability. To improve the spatial coverage and diversity of the initial population, this invention introduces an opposition-based learning strategy. While generating random initial solutions, it constructs their inverse solutions and obtains a more representative initial population through merging and selection mechanisms.

[0205] First, generate a random initial population. :

[0206] (19)

[0207] Subsequently, an inverse population is constructed using interval symmetric mapping. :

[0208] (20)

[0209] right Perform boundary consistency checks to ensure the populations are within the upper and lower bounds. Merge the two populations to obtain the candidate set. For all candidates Individuals are calculated for fitness values, then sorted from best to worst fitness, and the top performers are selected. Each individual is used as the final initial population. This strategy, by introducing reverse samples at the same random sampling cost, helps to form a more uniform initial coverage in the solution space, providing a higher quality starting point for global search and local exploitation in subsequent iterations.

[0210] 3) Elite Solution Mutation Strategy with Cauchy and Gaussian Mixed Perturbation. In the later stages of SCA, the position update process tends to over-rely on the current best individual, leading to decreased population diversity and search stagnation, thus increasing the probability of getting trapped in local optima. Therefore, this study introduces a "mixed Cauchy-Gaussian mutation" to adaptively perturb the elite solution after obtaining the current best individual in each iteration, achieving a dynamic balance between enhancing global exploration in the early stages and strengthening local refinement in the later stages. The core idea of ​​this strategy is to utilize the heavy-tailed characteristic of the Cauchy distribution to generate a higher probability of cross-domain jumps in the early stages of iteration, and gradually increase the proportion of Gaussian perturbation in the later stages of iteration, thereby performing fine-grained search within the optimal neighborhood. This mixed mechanism can effectively suppress premature convergence and improve the algorithm's adaptability to different initial values ​​and complex non-convex targets.

[0211] Let the first The current optimal position obtained in the next iteration is: By applying a hybrid mutation, candidate elite solutions are obtained. :

[0212] (twenty one)

[0213] In the formula, This indicates multiplication by dimension; and These are the random perturbation terms of the standard Cauchy distribution and the standard Gaussian distribution, respectively.

[0214] The weighting coefficients are adaptively adjusted with the number of iterations as follows:

[0215] (twenty two)

[0216] (twenty three)

[0217] In the formula, T represents the maximum number of iterations.

[0218] This shows that in the early stages of iteration Larger The mutations are relatively small, with Cauchy perturbations dominating to expand the search range; as iterations progress, Gradually reduce As the mutation mechanism gradually increases, it smoothly transitions to being dominated by Gaussian perturbations to improve the accuracy of local optimization.

[0219] 4) Optimization steps for voltage-sensing active filters based on improved sine and cosine algorithms, such as... Figure 3 As shown, specifically:

[0220] Based on the installation location of the voltage-sensing active power filter, an improved SCA (Self-Optimizing Algorithm) is applied to solve the established collaborative optimization configuration model, yielding the overall optimal compensation capacity of the voltage-sensing active power filter. The specific steps of solving the comprehensive optimization configuration model based on the improved SCA can be summarized as follows:

[0221] S1: Input distribution network parameters, harmonic source parameters, and algorithm-related parameters;

[0222] S2: A population initialization strategy based on reverse learning, where each individual is encoded to represent the compensated virtual conductance parameters of the voltage-sensing active filter under each harmonic order;

[0223] S3: Solve the harmonic power flow equations and calculate indicators such as node voltage harmonic distortion rate; impose penalties on individuals that do not meet the constraints;

[0224] S4: Calculate the fitness of each individual based on the objective function, and record the global optimal solution of the current iteration for subsequent position updates;

[0225] S5: Based on the basic improvement of the SCA update mechanism, an elite solution mutation strategy of mixed Cauchy and Gaussian perturbation is introduced to perturb elite individuals and select the best to retain them in order to enhance the ability to escape local optima.

[0226] S6: Determine if the termination condition is met. If not, return to S3 to continue iteration. Otherwise, output the optimal solution and the optimal configuration result of the voltage sensing active filter, and configure the harmonic conductance parameters of the voltage sensing active filter in the distribution network according to the optimal value.

[0227] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0228] Based on the same inventive concept, this application also provides a power grid harmonic mitigation device for implementing the aforementioned power grid harmonic mitigation method. The solution provided by this device is similar to the implementation scheme described in the above method; therefore, the specific limitations in one or more power grid harmonic mitigation device embodiments provided below can be found in the limitations of the power grid harmonic mitigation method described above, and will not be repeated here.

[0229] In one exemplary embodiment, such as Figure 4 As shown, a power distribution network harmonic mitigation device is provided, comprising: an acquisition module 402, a construction module 404, a solution module 406, and a configuration module 408, wherein:

[0230] The acquisition module 402 is used to acquire, for nodes in the distribution network, an over-limit risk factor that characterizes the harmonic over-limit risk level of the node based on the node's voltage harmonic distortion rate and voltage harmonic distortion rate upper limit; and to acquire the over-limit risk weight coefficient of the node based on the over-limit risk factor.

[0231] Module 404 is constructed to obtain the risk weighting value based on the node's over-limit risk weighting coefficient and voltage harmonic distortion rate; based on the risk weighting value, the worst node risk penalty term, the fixed investment cost of the rated capacity voltage detection active filter, and the operation and maintenance cost of the actual compensation capacity voltage detection active filter, a comprehensive objective function for harmonic control of the distribution network is constructed.

[0232] The solution module 406 is used to take the virtual harmonic conductance of each voltage-sensing active filter at each node of the distribution network as optimization variables, input the preset distribution network parameters and harmonic source parameters, and iteratively solve the comprehensive objective function of harmonic control through the improved sine and cosine algorithm to obtain the optimal value of each virtual harmonic conductance of the voltage-sensing active filter.

[0233] Configuration module 408 is used to configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

[0234] In one embodiment, the acquisition module 402 is further configured to:

[0235] The sum of squares of the voltage amplitudes of each harmonic that the node needs to control is calculated, and then the arithmetic square root of the sum of squares is calculated. The ratio between the arithmetic square root and the fundamental voltage amplitude of the node is calculated to obtain the voltage harmonic distortion rate of the node.

[0236] In one embodiment, the excess risk weight coefficient of a node is obtained based on the excess risk factor, including:

[0237] The risk sensitivity index is raised to the power of the risk factor of each node to obtain the calculation result of each node; the risk sensitivity index is used to adjust the strength of the weighting coefficient of the risk factor of high-risk nodes.

[0238] Obtain the sum of the operation results of each node; calculate the ratio between the operation result of each node and the sum, and obtain the over-limit risk weight coefficient of the corresponding node.

[0239] In one embodiment, the construction module 404 is further configured to:

[0240] Based on the over-limit risk weight coefficient of each node, nodes with over-limit risk weight coefficients greater than a preset value are selected from all nodes to obtain a set of installation nodes for voltage-sensing active filters. The rated capacity of the voltage-sensing active filter of each node in the set of installation nodes is multiplied by the investment cost coefficient per unit capacity to obtain the first product result of each node. The first product results of each node are summed to obtain the fixed investment cost of the rated capacity voltage-sensing active filter.

[0241] The actual compensation capacity of the voltage-sensing active filter at each node in the set of voltage-sensing active filter installation nodes is multiplied sequentially by the percentage coefficient of operation and maintenance costs to fixed investment costs and the investment cost coefficient per unit capacity to obtain the second product result for each node; the second product results of each node are summed to obtain the operation and maintenance cost of the voltage-sensing active filter with actual compensation capacity.

[0242] In one embodiment, the construction module 404 is further configured to:

[0243] Set a governance effect ratio coefficient for the risk weighting value, set a penalty item ratio coefficient for the worst node risk penalty item, and set an economic cost ratio coefficient for the sum of the fixed investment cost of the voltage detection active filter with rated capacity and the operation and maintenance cost of the voltage detection active filter with actual compensation capacity.

[0244] The comprehensive objective function for harmonic control of the distribution network is obtained by weighting and summing the risk weight, the worst-case node risk penalty, and the second sum based on the governance effect ratio coefficient, the penalty coefficient, and the economic cost ratio coefficient.

[0245] In one embodiment, the solving module 406 is further configured to implement the stride parameter adjustment strategy, including:

[0246] For nodes whose risk weight coefficient exceeds the preset value, the virtual harmonic conductance of the voltage-sensing active filter of the node is used as the algorithm variable dimension. The sum of squares of the deviations between the individual positions and the population mean under the algorithm variable dimension is calculated. The arithmetic mean of the sum of squares of the deviations is calculated and then the corresponding arithmetic square root is calculated. The difference between the preset adjustable minimum value and the adjustable maximum value under the algorithm variable dimension is obtained, and the ratio between the arithmetic square root and the difference is calculated.

[0247] Calculate the arithmetic mean of the ratios under each virtual harmonic conductance, and use it as the solution value of the population diversity index generated in the current iteration; calculate the ratio between the solution value of the population diversity index generated in the current iteration and the maximum solution value generated by the population diversity index in the entire iteration; based on the ratio, obtain the adaptive function value;

[0248] The stride parameter is adjusted based on the adaptive function value; the stride parameter is positively correlated with the adaptive function value.

[0249] In one embodiment, the solving module 406 is further used for the implementation process of the population initialization strategy, including:

[0250] For nodes whose risk weight coefficient exceeds the preset value, the virtual harmonic conductance of the voltage-sensing active filter of the node is used as the algorithm variable dimension. The difference between the preset adjustable minimum value and the adjustable maximum value under the algorithm variable dimension is obtained. The difference is multiplied by a random number from 0 to 1 and then added to the adjustable minimum value to obtain a single initial value under the algorithm variable dimension.

[0251] Combine the individual initial values ​​corresponding to each virtual harmonic conductance to form a single initial solution, and construct a random initial population composed of multiple initial solutions; subtract the initial value of a single initial solution in the random initial population under the corresponding algorithm variable dimension from the sum of the adjustable minimum value and the adjustable maximum value to obtain a single inverse value under the algorithm variable dimension.

[0252] Combine the single inverse values ​​corresponding to each virtual harmonic conductance to form a single inverse solution, and construct an inverse population composed of multiple inverse solutions; merge the random initial population and the inverse population to obtain a candidate population; based on the comprehensive objective function of harmonic governance, calculate the fitness values ​​of all solutions in the candidate population, and select the top preset number of solutions with the best fitness values ​​as the initial population for algorithm iteration, with the preset number being the preset initial population size.

[0253] In one embodiment, the solver module 406 is further configured to implement the elite solution mutation strategy, including:

[0254] In the iterative process of the improved sine and cosine algorithm, the fitness values ​​of all solutions in the current iteration population are calculated based on the comprehensive objective function of harmonic control. The solution with the best fitness value is selected as the current elite solution, which is a combination of the corresponding values ​​of each virtual harmonic conductance.

[0255] Obtain the random perturbation terms of the standard Cauchy distribution and the standard Gaussian distribution, calculate the ratio between the current iteration number and the maximum iteration number of the algorithm, subtract the ratio from 1 as the weight coefficient of the Cauchy distribution perturbation term, and use the ratio as the weight coefficient of the Gaussian distribution perturbation term.

[0256] The Cauchy distribution perturbation term and the Gaussian distribution perturbation term are multiplied by their corresponding weight coefficients and summed. The summation result is then multiplied by the value of the current elite solution in each dimension of the algorithm variable to obtain the variation value in each dimension of the algorithm variable. The variation values ​​are then combined to form the candidate elite solution.

[0257] The fitness values ​​of candidate elite solutions are calculated based on the comprehensive objective function of harmonic control. The fitness values ​​of candidate elite solutions are compared with the fitness values ​​of the current elite solutions, and the solution with the smallest fitness value is selected as the elite solution for the next iteration.

[0258] Each module in the aforementioned power distribution network harmonic mitigation device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0259] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a method for harmonic mitigation in a power distribution network. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0260] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0261] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0262] For nodes in the distribution network, based on the node's voltage harmonic distortion rate and upper limit of voltage harmonic distortion rate, an over-limit risk factor is obtained to characterize the harmonic over-limit risk level of the node; based on the over-limit risk factor, the over-limit risk weight coefficient of the node is obtained.

[0263] Based on the node's over-limit risk weight coefficient and voltage harmonic distortion rate, a risk weighting value is obtained; based on the risk weighting value, the worst node risk penalty term, the fixed investment cost of the rated capacity voltage-sensing active filter, and the operation and maintenance cost of the actual compensation capacity voltage-sensing active filter, a comprehensive objective function for harmonic control of the distribution network is constructed, and fundamental power flow constraints, harmonic power flow constraints, and compensation capacity limit constraints of the voltage-sensing active filter are set.

[0264] The virtual harmonic conductances of the voltage-sensing active filters at each node of the distribution network are used as optimization variables. The distribution network parameters and harmonic source parameters are input, and the comprehensive objective function for harmonic control is iteratively solved by the improved sine and cosine algorithm to obtain the optimal values ​​of the virtual harmonic conductances of the voltage-sensing active filters.

[0265] Configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

[0266] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0267] For nodes in the distribution network, based on the node's voltage harmonic distortion rate and upper limit of voltage harmonic distortion rate, an over-limit risk factor is obtained to characterize the harmonic over-limit risk level of the node; based on the over-limit risk factor, the over-limit risk weight coefficient of the node is obtained.

[0268] Based on the node's over-limit risk weight coefficient and voltage harmonic distortion rate, a risk weighting value is obtained; based on the risk weighting value, the worst node risk penalty term, the fixed investment cost of the rated capacity voltage-sensing active filter, and the operation and maintenance cost of the actual compensation capacity voltage-sensing active filter, a comprehensive objective function for harmonic control of the distribution network is constructed, and fundamental power flow constraints, harmonic power flow constraints, and compensation capacity limit constraints of the voltage-sensing active filter are set.

[0269] The virtual harmonic conductances of the voltage-sensing active filters at each node of the distribution network are used as optimization variables. The distribution network parameters and harmonic source parameters are input, and the comprehensive objective function for harmonic control is iteratively solved by the improved sine and cosine algorithm to obtain the optimal values ​​of the virtual harmonic conductances of the voltage-sensing active filters.

[0270] Configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

[0271] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0272] For nodes in the distribution network, based on the node's voltage harmonic distortion rate and upper limit of voltage harmonic distortion rate, an over-limit risk factor is obtained to characterize the harmonic over-limit risk level of the node; based on the over-limit risk factor, the over-limit risk weight coefficient of the node is obtained.

[0273] Based on the node's over-limit risk weight coefficient and voltage harmonic distortion rate, a risk weighting value is obtained; based on the risk weighting value, the worst node risk penalty term, the fixed investment cost of the rated capacity voltage-sensing active filter, and the operation and maintenance cost of the actual compensation capacity voltage-sensing active filter, a comprehensive objective function for harmonic control of the distribution network is constructed, and fundamental power flow constraints, harmonic power flow constraints, and compensation capacity limit constraints of the voltage-sensing active filter are set.

[0274] The virtual harmonic conductances of the voltage-sensing active filters at each node of the distribution network are used as optimization variables. The distribution network parameters and harmonic source parameters are input, and the comprehensive objective function for harmonic control is iteratively solved by the improved sine and cosine algorithm to obtain the optimal values ​​of the virtual harmonic conductances of the voltage-sensing active filters.

[0275] Configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

[0276] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0277] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0278] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0279] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for controlling harmonics in a power distribution network, characterized in that, The method includes: For a node in the distribution network, based on the voltage harmonic distortion rate and the upper limit of the voltage harmonic distortion rate of the node, an over-limit risk factor is obtained to characterize the harmonic over-limit risk level of the node; based on the over-limit risk factor, the over-limit risk weight coefficient of the node is obtained. Based on the over-limit risk weight coefficient and voltage harmonic distortion rate of the node, a risk weighting value is obtained; based on the risk weighting value, the worst node risk penalty term, the fixed investment cost of the rated capacity voltage detection active filter and the operation and maintenance cost of the actual compensation capacity voltage detection active filter, a comprehensive objective function for harmonic control of the distribution network is constructed. The virtual harmonic conductance of each voltage-sensing active filter at each node of the distribution network is used as the optimization variable. The preset distribution network parameters and harmonic source parameters are input, and the comprehensive objective function for harmonic control is iteratively solved by the improved sine and cosine algorithm to obtain the optimal value of each virtual harmonic conductance of the voltage-sensing active filter. Configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

2. The method according to claim 1, characterized in that, The process of obtaining the voltage harmonic distortion rate of the node includes: Calculate the sum of squares of the voltage amplitudes of each harmonic that the node needs to control, and then calculate the arithmetic square root of the sum of squares; calculate the ratio between the arithmetic square root and the fundamental voltage amplitude of the node to obtain the voltage harmonic distortion rate of the node.

3. The method according to claim 1, characterized in that, The step of obtaining the limit-breaking risk weight coefficient of the node based on the limit-breaking risk factor includes: The risk sensitivity index is raised to the power of the risk factor of each node to obtain the calculation result of each node; the risk sensitivity index is used to adjust the strength of the weighting coefficient of the risk factor of high-risk nodes in the model. Obtain the first sum of the calculation results of each node; calculate the ratio between the calculation result of each node and the first sum to obtain the over-limit risk weight coefficient of the corresponding node.

4. The method according to claim 1, characterized in that, The process of obtaining the fixed investment cost of the rated capacity voltage-sensing active filter and the operation and maintenance cost of the actual compensation capacity voltage-sensing active filter includes: Based on the over-limit risk weight coefficient of each node, nodes with an over-limit risk weight coefficient greater than a preset value are selected from all nodes to obtain a set of installation nodes for voltage-sensing active filters. The rated capacity of the voltage-sensing active filter of each node in the set of installation nodes is multiplied by the investment cost coefficient per unit capacity to obtain the first product result of each node. The first product results of each node are summed to obtain the fixed investment cost of the rated capacity voltage-sensing active filter. The actual compensation capacity of the voltage-sensing active filter at each node in the set of voltage-sensing active filter installation nodes is multiplied sequentially by the percentage coefficient of operation and maintenance costs to fixed investment costs and the investment cost coefficient per unit capacity to obtain the second product result of each node; the second product results of each node are summed to obtain the operation and maintenance cost of the voltage-sensing active filter with actual compensation capacity.

5. The method according to claim 1, characterized in that, The comprehensive objective function for harmonic mitigation of the distribution network is constructed based on the risk weighting value, the worst-case node risk penalty term, the fixed investment cost of the rated capacity voltage-sensing active power filter, and the operation and maintenance cost of the actual compensation capacity voltage-sensing active power filter, including: A governance effect ratio coefficient is set for the risk weighting value, a penalty item ratio coefficient is set for the worst node risk penalty item, and an economic cost ratio coefficient is set for the second sum between the fixed investment cost of the voltage detection active filter with rated capacity and the operation and maintenance cost of the voltage detection active filter with actual compensation capacity. The comprehensive objective function for harmonic control of the distribution network is obtained by weighting and summing the risk weight, the worst-case node risk penalty, and the second sum based on the governance effect ratio coefficient, the penalty coefficient, and the economic cost ratio coefficient.

6. The method according to claim 1, characterized in that, The improved sine and cosine algorithm incorporates a stride parameter adjustment strategy based on population diversity indicators. The implementation process of the stride parameter adjustment strategy includes: For nodes whose risk weighting coefficient exceeds a preset value, the virtual harmonic conductance of the voltage-sensing active filter of the node is used as the algorithm variable dimension. The sum of squared deviations between the individual positions and the population mean under the algorithm variable dimension is calculated. The arithmetic mean of the sum of squared deviations is calculated, and then the corresponding arithmetic square root is calculated. The difference between the preset adjustable minimum value and the adjustable maximum value under the algorithm variable dimension is obtained, and the ratio between the arithmetic square root and the difference is calculated. Calculate the arithmetic mean of the ratios under each virtual harmonic conductance, and use it as the solution value generated by the population diversity index in the current iteration; calculate the ratio between the solution value generated by the population diversity index in the current iteration and the maximum solution value generated by the population diversity index in the entire iteration; obtain the adaptive function value based on the ratio. The stride parameter is adjusted based on the adaptive function value; wherein the stride parameter is positively correlated with the adaptive function value.

7. The method according to claim 1, characterized in that, The improved sine and cosine algorithms incorporate a population initialization strategy based on reverse learning. The implementation process of the population initialization strategy includes: For nodes whose risk weight coefficient exceeds the preset value, the virtual harmonic conductance of the voltage-sensing active filter of the node is used as the algorithm variable dimension. The difference between the preset adjustable minimum value and the adjustable maximum value under the algorithm variable dimension is obtained. The difference is multiplied by a random number from 0 to 1 and then added to the adjustable minimum value to obtain a single initial value under the algorithm variable dimension. Combine the individual initial values ​​corresponding to each virtual harmonic conductance to form a single initial solution, and construct a random initial population composed of multiple initial solutions; subtract the initial value of the single initial solution in the random initial population under the corresponding algorithm variable dimension from the sum of the adjustable minimum value and the adjustable maximum value to obtain a single inverse value under the algorithm variable dimension. Combine the single inverse values ​​corresponding to each virtual harmonic conductance to form a single inverse solution, and construct an inverse population composed of multiple inverse solutions; merge the random initial population and the inverse population to obtain a candidate population; based on the comprehensive objective function of harmonic governance, calculate the fitness values ​​of all solutions in the candidate population, and select the top preset number of solutions with the best fitness values ​​as the initial population for algorithm iteration, where the preset number is the preset initial population size.

8. The method according to claim 1, characterized in that, The improved sine and cosine algorithms incorporate an elite solution mutation strategy that combines Cauchy and Gaussian mixed perturbations. The implementation process of the elite solution mutation strategy includes: In the iterative process of the improved sine and cosine algorithm, the fitness values ​​of all solutions in the current iteration population are calculated based on the comprehensive objective function of harmonic control, and the solution with the best fitness value is selected as the current elite solution. The current elite solution is a combination of the corresponding values ​​of each virtual harmonic conductance. Obtain the standard Cauchy distribution random perturbation term and the standard Gaussian distribution random perturbation term, calculate the ratio between the current iteration number and the maximum iteration number of the algorithm, subtract the ratio from 1 to obtain the weight coefficient of the Cauchy distribution perturbation term, and use the ratio as the weight coefficient of the Gaussian distribution perturbation term; The Cauchy distribution perturbation term and the Gaussian distribution perturbation term are multiplied by their corresponding weight coefficients and then summed. The summation result is multiplied by the value of the current elite solution under each algorithm variable dimension one by one to obtain the variation value under each algorithm variable dimension. The variation values ​​are combined to form a candidate elite solution. The fitness values ​​of the candidate elite solutions are calculated based on the comprehensive objective function for harmonic control. The fitness values ​​of the candidate elite solutions are compared with the fitness values ​​of the current elite solutions to select the solution with the smallest fitness value as the elite solution for the next iteration.

9. A power distribution network harmonic mitigation device, characterized in that, The device includes: The acquisition module is used to acquire, for nodes in the distribution network, an over-limit risk factor characterizing the harmonic over-limit risk level of the node based on the voltage harmonic distortion rate and the upper limit of the voltage harmonic distortion rate of the node; and acquire the over-limit risk weight coefficient of the node based on the over-limit risk factor. A construction module is used to obtain a risk weighting value based on the over-limit risk weighting coefficient and voltage harmonic distortion rate of the node; and to construct a comprehensive objective function for harmonic control of the distribution network based on the risk weighting value, the worst node risk penalty term, the fixed investment cost of the rated capacity voltage detection active filter and the operation and maintenance cost of the actual compensation capacity voltage detection active filter. The solution module is used to take the virtual harmonic conductance of each voltage-sensing active filter at each node of the distribution network as optimization variables, input the preset distribution network parameters and harmonic source parameters, and iteratively solve the comprehensive objective function of harmonic control through an improved sine and cosine algorithm to obtain the optimal value of each virtual harmonic conductance of the voltage-sensing active filter. The configuration module is used to configure the harmonic conductance parameters of the voltage-sensing active filter in the distribution network according to the optimal values ​​of each virtual harmonic conductance of the voltage-sensing active filter.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

Images