Flexible control method of substation area based on Pareto frontier perception optimization

Through the flexible control method of the station area based on Pareto frontier perception optimization, a multi-objective source-load commutation optimization model was constructed, which solved the problem of three-phase imbalance in the low-voltage distribution network, realized effective regulation of the power grid, and reduced imbalance improves system efficiency and stability.

CN118713127BActive Publication Date: 2024-11-26WILLFAR INFORMATION TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411179212.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-11-26
Estimated Expiration
2044-08-27

AI Technical Summary

Technical Problem

The existing technology is difficult to fully solve the problem of three-phase imbalance in low-voltage distribution networks, especially in the aggravated volatility and randomness after distributed photovoltaic access, resulting in reduced system efficiency, equipment damage and waste of electricity.

Method used

Using the flexible control method of the station area based on Pareto frontier perception optimization, the multi-objective source-load commutation optimization control model is constructed by obtaining the topology structure and node phase of the low-voltage station area, and using the Pareto frontier perception optimization algorithm to solve it step by step Regulation, generate flexible regulation strategies in the table area to reduce the three-phase imbalance.

Benefits of technology

It effectively reduces the overall imbalance of the platform area, improves the reliability of the flexible regulation strategy, extends the balanced area to the end of the platform area, solves the problem of three-phase imbalance in the power grid, and improves the operating efficiency and stability of the power system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118713127B_ABST
    Figure CN118713127B_ABST
Patent Text Reader

Abstract

The present application is applicable to the field of power system optimization and control technology, and specifically relates to a flexible control method for substations based on Pareto frontier sensing optimization. First, when the source-load phase-changing result does not meet the first preset condition, the imbalance calculation level is gradually reduced and the node imbalance is recalculated, thereby realizing a step-by-step decreasing control scheme for the three-phase imbalance of the power grid. This step-by-step control method can extend the balance area to the end of the substation as much as possible, thereby reducing the overall imbalance of the substation, and thus can effectively and comprehensively solve the problem of three-phase imbalance of the power grid. Secondly, by constructing a multi-objective source-load phase-changing optimization control model based on the low-voltage substation topology and the phase to which each node in the substation belongs, and solving it using the Pareto frontier sensing optimization algorithm, a source-load phase-changing result that achieves a balance between multiple objectives is obtained, further improving the reliability of the flexible control strategy of the substation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of power system optimization and control, and in particular to a flexible control method for substations based on Pareto frontier perception optimization. Background Art

[0002] Currently, low-voltage distribution networks widely utilize a three-phase, four-wire system, with a predominantly single-phase load structure. This structure can easily lead to three-phase imbalance in the grid. With the vigorous promotion of clean energy and the widespread integration of distributed photovoltaics, low-voltage distribution networks are becoming the new normal. The resulting volatility and randomness further exacerbate imbalances in the network. Three-phase imbalance not only impacts system efficiency and safety but can also lead to equipment damage and energy waste.

[0003] In order to deal with this problem, low-voltage substations generally install static VAR compensators, active filters and other load compensation devices on the load side or the grid side to suppress the three-phase imbalance of the power grid. However, these methods are not only costly, but also difficult to fully solve the three-phase imbalance problem of the entire distribution line. The existing patent with publication number CN108776869B provides a method for controlling the three-phase imbalance of substation transformers based on smart meter big data, including first extracting the voltage sequence data and current sequence data of the substation transformer and its users' smart meters from the electricity information collection system; then, calculating the correlation coefficient between each user and the three-phase voltage sequence data of the substation transformer A, B, and C, and identifying the phase sequence of each user; then, searching for an optimal substation user phase sequence combination based on a genetic algorithm to minimize the three-phase imbalance of the substation transformer. However, this control scheme mainly focuses on the imbalance optimization of the transformer node and lacks multi-level control capabilities. Therefore, it cannot effectively and comprehensively solve the three-phase imbalance problem of the entire power grid. Summary of the Invention

[0004] Based on this, it is necessary to provide a flexible control method based on Pareto frontier perception optimization that can achieve multi-level control and effectively and comprehensively solve the three-phase imbalance problem of the power grid.

[0005] This application provides a flexible control method for substations based on Pareto frontier perception optimization, including:

[0006] Step S1, obtaining the low voltage area topology and the phase of each node in the area;

[0007] Step S2: constructing a multi-objective source-load switching optimization control model based on the low-voltage substation topology and the phase of each node in the substation;

[0008] Step S3, using the Pareto frontier perception optimization algorithm to solve the multi-objective source-load commutation optimization control model to obtain the source-load commutation result;

[0009] Step S4, determining whether the source-load commutation result meets the first preset condition; if not, executing step S41; if so, executing step S5;

[0010] Step S41, shrinking the imbalance calculation level and recalculating the node imbalance;

[0011] Step S42, determining whether the node imbalance is greater than a preset node imbalance threshold; if so, executing step S43; if not, executing step S5;

[0012] Step S43, determining whether the unbalance calculation level before contraction is the transformer level; if not, returning to step S2; if yes, executing step S44;

[0013] Step S44, generating a photovoltaic inverter output power regulation strategy;

[0014] Step S5: generating a flexible control strategy for the substation based on the source-load commutation result and the photovoltaic inverter output power adjustment strategy.

[0015] In one embodiment, the multi-objective source-load commutation optimization control model includes an objective function of minimizing the overall power consumption imbalance in the substation area and an objective function of minimizing the number of commutation switch operations. Step S2 includes:

[0016] Step S200: Construct the R node three-phase outlet power consumption matrix for a sampling period ; R node three-phase outlet power consumption matrix The expression is as follows:

[0017]

[0018] Where, represents the power consumption of phase A outlet obtained by the i-th sampling of node R, represents the B-phase outlet power consumption obtained by the i-th sampling of the R node, represents the C-phase outlet power consumption obtained by the i-th sampling at the R node, and N represents the total number of sampling times during the operation cycle;

[0019] Step S201, according to the formula

[0020]

[0021] Calculate the power consumption of phase A outlet of R node within sampling period T respectively , Power consumption of phase B outlet and the power consumption of phase C outlet ;

[0022] Step S202: The sum of the power consumption of the three-phase outlets A, B, and C of the R node is taken as the total power consumption of the outlet of the R node; the total power consumption of the outlet of the R node The expression is as follows:

[0023]

[0024] Step S203, according to the formula

[0025]

[0026] Calculate the optimal power consumption of phase A to adjust the source and load separately , Phase B needs to adjust the optimal power consumption of the source and load The optimal power consumption of the source and load needs to be adjusted for phase C ;

[0027] Step S204: Obtain the power consumption vector of the adjustable source load connected to the lower end of the R node within the sampling period T. Phase sequence matrix of single-phase source and load of each phase of the lower node before R node adjustment ;in, Indicates the phase sequence of the nth single-phase source and load before adjustment. The expression is:

[0028]

[0029] Step S205: The power consumption vector of the adjustable source load connected to the lower end of the R node is Phase sequence matrix of single-phase source and load of each phase of the lower node before R node adjustment The difference is multiplied as the sum of the single-phase source and load power consumption of each phase of the lower level before the R node is adjusted; the expression for the sum of the single-phase source and load power consumption of each phase of the lower level before the R node is adjusted is as follows:

[0030]

[0031] Where, It represents the sum of the power consumption of the lower-level A-phase single-phase source and load before the R node is adjusted. It represents the sum of the power consumption of the lower-level B-phase single-phase source and load before the R node is adjusted. It represents the sum of the power consumption of the lower-level C-phase single-phase source and load before the R node is adjusted;

[0032] Step S206: Obtain the phase sequence matrix of the lower nodes after the phase sequence of the single-phase load of the R node is adjusted. And adjust the phase sequence of the single-phase load of the R node to the phase sequence matrix of the lower node The power consumption vector of the adjustable source load connected to the lower end of the R node The difference is multiplied as the sum of the single-phase source and load power consumption of each phase in the lower level after the R node is adjusted; the expression for the sum of the single-phase source and load power consumption of each phase in the lower level after the R node is adjusted is as follows:

[0033]

[0034] Where, It represents the sum of the power consumption of the lower-level A-phase single-phase source and load after the R node is adjusted. It represents the sum of the power consumption of the lower-level B-phase single-phase source and load after the R node is adjusted. It represents the sum of the power consumption of the lower-level C-phase single-phase source and load after the R node is adjusted;

[0035] Step S207: The difference between the sum of the single-phase source-load power consumption of each phase of the lower level after the R node is adjusted and the sum of the single-phase source-load power consumption of each phase of the lower level before the R node is adjusted is used as the change in the power consumption of each phase before and after the three-phase load adjustment of the R node. The specific expression is as follows:

[0036]

[0037] Where, Indicates the change in power consumption of phase A before and after the three-phase load adjustment of node R. Indicates the change in power consumption of phase B before and after the three-phase load adjustment of node R. Indicates the change in C-phase power consumption before and after the three-phase load adjustment at R node;

[0038] Step S208: Calculate the difference between the change in power consumption of each phase before and after the three-phase load adjustment at the R node and the optimal power consumption of each phase that needs to be adjusted. The expression for the difference in optimal power consumption of each phase that needs to be adjusted is as follows:

[0039]

[0040] Where, It represents the difference between the change in the power consumption of phase A before and after the three-phase load adjustment at node R and the optimal power consumption of phase A that needs to adjust the source load. It represents the difference between the change in the power consumption of phase B before and after the three-phase load adjustment of node R and the optimal power consumption of phase B that needs to adjust the source load. It represents the difference between the change in the power consumption of phase C before and after the three-phase load adjustment at node R and the optimal power consumption of phase C that needs to adjust the source load;

[0041] Step S209: Obtain the level weight of the level m where the R node is located. The calculation formula of the level weight is as follows:

[0042]

[0043] Where, represents the level weight of the mth layer, and s represents the current three-phase unbalance calculation level;

[0044] Step S210, according to the formula

[0045]

[0046] Calculate the maximum power imbalance before and after the three-phase load adjustment of the R node ;

[0047] Step S211, according to the formula

[0048]

[0049] Calculate the average power consumption imbalance of all three-phase nodes in the mth layer separately ; where j represents the total number of three-phase nodes in the mth layer, represents the maximum power consumption imbalance of the i-th node in the m-th layer;

[0050] Step S212: constructing a minimum objective function for the overall power consumption imbalance of the substation based on the average power consumption imbalance of all three-phase nodes in each layer; the expression of the minimum objective function for the overall power consumption imbalance of the substation is as follows:

[0051]

[0052] Where, represents the average power consumption imbalance of all three-phase nodes in the kth layer;

[0053] Step S213: The phase sequence matrix of the lower node is adjusted according to the phase sequence of the single-phase load of the R node. Get the number of commutation switch actions of n adjustable sources and loads in the substation; the number of commutation switch actions The expression is as follows:

[0054]

[0055] Where, Indicates whether the commutation switch of the i-th adjustable source load is in operation;

[0056] Step S214: constructing a minimum number of phase-changing switching operations objective function based on the number of phase-changing switching operations; the expression of the minimum number of phase-changing switching operations objective function is: .

[0057] In one embodiment, the multi-objective source-load commutation optimization control model also includes a transformer node power consumption imbalance constraint, an overall area power consumption imbalance constraint, and a commutation switch action logic constraint; wherein the transformer node power consumption imbalance constraint is expressed as:

[0058]

[0059] Where, Indicates the imbalance of power consumption at transformer nodes, Indicates the transformer node imbalance threshold;

[0060] The expression for the overall power consumption imbalance constraint in the substation is:

[0061]

[0062] Where, Indicates the overall imbalance of electricity consumption in the substation area. Indicates the overall power consumption imbalance threshold of the substation;

[0063] The expression of the reversing switch action logic constraint is:

[0064]

[0065] Where, Indicates the single-phase source load Q connected to the lower end of the R node m and Connection status of the phase busbar.

[0066] In one embodiment, the multi-objective source-load commutation optimization control model further includes a commutation switch action number constraint, and the expression for the commutation switch action number constraint is:

[0067]

[0068] Where, Indicates the maximum number of commutation times.

[0069] In one embodiment, step S3 includes:

[0070] Step S31, generating an initial population according to the access phases of the single-phase source and load;

[0071] Step S32, calculating the power imbalance and commutation times of the initial population;

[0072] Step S33, performing fast non-dominated sorting and crowding distance comparison on the initial population;

[0073] Step S34, selecting, crossover and mutation to generate offspring population;

[0074] Step S35, calculating the power imbalance and commutation times of the offspring population;

[0075] Step S36, merging the offspring population with the initial population, and performing fast non-dominated sorting and crowding distance comparison;

[0076] Step S37, determine whether the stop condition is met; if so, execute step S38, if not, return to step S34;

[0077] Step S38: output the optimal phase difference.

[0078] In one embodiment, step S33 includes:

[0079] Step S330, initializing the value of variable e to 1;

[0080] Step S331, select a Pareto solution set from the initial population, and record the genes in the Pareto solution set as the e-th dominance level;

[0081] Step S332: remove genes in the e-th dominance level from the initial population, reselect a Pareto solution set based on the remaining genes, and record the genes in the newly selected Pareto solution set as the e+1-th dominance level;

[0082] Step S333: update the value of e to e+1, and repeat step S332 until all genes in the initial population are ranked, obtaining a non-dominated ranking result for all genes in the initial population;

[0083] Step S334, according to the formula

[0084]

[0085] Calculate the crowding distance of each gene in the same dominance level; where, represents the crowding distance of gene i, represents the j-th objective function value of gene i+1, represents the j-th objective function value of gene i-1, M represents the number of objective functions, and d represents the number of genes in the current dominance level.

[0086] In one embodiment, step S34 includes:

[0087] Step S340: Based on the non-dominated sorting result and the crowding distance, genes that meet the iteration condition are selected to enter the next iteration; the iteration condition is expressed as follows:

[0088]

[0089] Where, represents the non-dominated level of gene i, represents the non-dominated level of gene j, represents the crowding distance of gene j;

[0090] Step S341, randomly select two genes from the parent population and , calculate the two offspring chromosomes Q1 and Q2 according to the crossover operation calculation formula; the crossover operation calculation formula is as follows:

[0091]

[0092] in, The calculation formula is as follows:

[0093]

[0094] Where, is the cross-distribution index, is a random number;

[0095] Step S342: randomly select a chromosome from the parent population and generate a daughter chromosome Q3 according to the mutation operation calculation formula; the mutation operation calculation formula is expressed as follows:

[0096]

[0097] Where, represents the upper limit of the variable, represents the lower limit of the variable, is the variation distribution index, is a random number.

[0098] In one embodiment, before step S1, the method further includes:

[0099] Step S01, obtaining the operation data of the distribution station area;

[0100] Step S02, calculating the load rate of the distribution area according to the operating data of the distribution area;

[0101] Step S03, determining whether the load rate of the substation is greater than a preset load rate threshold; if so, executing step S04; if not, executing step S5;

[0102] Step S04, determining whether the area imbalance is greater than a preset area imbalance threshold; if so, executing step S1, if not, executing step S5.

[0103] In one embodiment, after step S01 and before step S02, the method further includes:

[0104] Step S011, calculating the reverse power of the distribution area according to the operation data of the distribution area;

[0105] Step S012, determining whether the reverse power of the substation is greater than a preset reverse power threshold; if so, executing step S1; if not, executing step S02.

[0106] In one embodiment, step S44 includes:

[0107] Step S441: Generate a photovoltaic inverter output power regulation strategy based on the user's power generation capacity, load demand, and historical regulation records.

[0108] The above-mentioned flexible control method of the substation based on Pareto frontier perception optimization first realizes a step-by-step reduction control scheme for the three-phase imbalance of the power grid by gradually reducing the imbalance calculation level when the source-load phase-changing result does not meet the first preset condition and recalculating the node imbalance. This step-by-step control method can extend the balance area to the end of the substation as much as possible, thereby reducing the overall imbalance of the substation, and thus can effectively and comprehensively solve the problem of three-phase imbalance of the power grid. Secondly, by constructing a multi-objective source-load phase-changing optimization control model based on the low-voltage substation topology and the phase to which each node in the substation belongs, and solving it using the Pareto frontier perception optimization algorithm, a source-load phase-changing result that achieves a balance between multiple objectives is obtained, further improving the reliability of the flexible control strategy of the substation. BRIEF DESCRIPTION OF THE DRAWINGS

[0109] Figure 1 1. A flow chart of a flexible control method for a substation based on Pareto frontier perception optimization in one embodiment;

[0110] Figure 2 Schematic diagram of the connection between adjustable source and load and low-voltage power distribution system in one embodiment;

[0111] Figure 3 1. A flowchart illustrating steps for solving a multi-objective source-load commutation optimization control model using a fast non-dominated sorting genetic algorithm in one embodiment;

[0112] Figure 4 A schematic flow chart of a flexible control method for a substation based on Pareto frontier perception optimization in another embodiment;

[0113] Figure 5 In one embodiment, the imbalance change curve of the front and rear branch nodes is adjusted;

[0114] Figure 6 For adjusting the front and rear meter box node change curve in one embodiment;

[0115] Figure 7 A transformer node change curve before and after adjustment in one embodiment;

[0116] Figure 8 In another embodiment, the imbalance change curve of the front and rear branch nodes is adjusted;

[0117] Figure 9 For adjusting the front and rear meter box node change curves in another embodiment;

[0118] Figure 10 For another embodiment, the transformer node change curve before and after adjustment;

[0119] FIG11( a ) is a curve showing the change of imbalance before and after photovoltaic regulation in one embodiment;

[0120] FIG11( b ) is a photovoltaic power generation change curve before and after regulation in one embodiment;

[0121] FIG11( c ) is a graph showing the load and photovoltaic output of phase A before and after regulation in one embodiment;

[0122] FIG11( d ) shows the B-phase load, photovoltaic output before regulation, and photovoltaic output curve after regulation in one embodiment;

[0123] Figure 11 (e) shows the C-phase load, photovoltaic output before regulation, and photovoltaic output curve after regulation in one embodiment;

[0124] FIG11( f ) is a curve showing the change of the A supply ratio before and after regulation in one embodiment;

[0125] FIG11( g ) is a curve showing the change of the B supply ratio before and after regulation in one embodiment;

[0126] FIG11( h ) is a curve showing the change of the C-phase supply ratio before and after regulation in one embodiment. DETAILED DESCRIPTION

[0127] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0128] In one embodiment, Figure 1 As shown, a flexible control method for substations based on Pareto front perception optimization is provided. This embodiment uses the method applied to a terminal as an example. It is understandable that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0129] Step S1: Obtain the low voltage substation topology and the phase to which each node in the substation belongs.

[0130] Specifically, when the unbalance degree of the substation is greater than the preset unbalance degree threshold of the substation, the low voltage substation topology and the phase of each node in the substation are obtained. Among them, the unbalance degree of the substation is the overall unbalance degree of the substation. The overall unbalance degree of the substation can be obtained by summing or weighted summing the three-phase current unbalance degrees of each node in the substation. The three-phase current unbalance degree of each node in the substation The calculation formula is shown in formula (1):

[0131] (1)

[0132] Where, A phase output current , B phase output current and C phase output current The maximum current value in A phase output current , B phase output current and C phase output current The minimum current value in .

[0133] Step S2: construct a multi-objective source-load switching optimization control model based on the low-voltage substation topology and the phase of each node in the substation.

[0134] The multi-objective source-load switching optimization control model can include multiple objective functions and constraints. By conducting a detailed analysis of the low-voltage substation topology, understanding the connection relationships between nodes, and identifying key nodes and paths, corresponding constraints can be established by analyzing node voltage ranges, branch current limits, equipment capacity limits, and other factors. For example, the objective function can include a substation imbalance minimization objective function. Substation imbalance is an important indicator for measuring the degree of three-phase imbalance in a substation. By optimizing source-load switching, the substation imbalance can be minimized. The objective function can also include a power loss minimization objective function. By optimizing source-load switching, power flow within the substation can be reduced, thereby reducing power loss. Constraints can include node voltage constraints and source-load switching constraints. Node voltages within the substation must remain within a certain range to ensure the normal operation of power equipment. Source-load switching operations must meet actual operational constraints, such as the capacity of the switching equipment and the switching time.

[0135] Step S3: using the Pareto frontier perception optimization algorithm to solve the multi-objective source-load commutation optimization control model to obtain the source-load commutation result.

[0136] Among them, the Pareto frontier-aware optimization algorithm is an algorithm for multi-objective optimization problems. It aims to find a set of solutions that cannot be improved and that achieve the best trade-off between multiple objectives. The specific solution process includes: (1) generating an initial solution set, that is, possible source-load switching schemes; (2) for each initial solution, calculating its value on multiple objective functions; (3) based on the calculated objective function values, the solutions in the solution set are screened according to the definition of Pareto optimal solutions to obtain the Pareto frontier, that is, a set of non-dominated solution sets; (4) selecting a part of the solutions from the Pareto frontier as parent solutions through selection strategies (such as roulette selection, tournament selection, etc.); (5) performing crossover and mutation operations on the parent solutions to generate new solutions; (6) adding the newly generated solutions to the population and deleting some poor solutions; repeating the above steps (2) to (6) until the termination condition is met, such as reaching the maximum number of iterations or the quality of the solution has not been significantly improved within a certain number of times. At the end of the algorithm, the obtained Pareto frontier is the source-load switching result.

[0137] Step S4, determining whether the source-load commutation result meets the first preset condition; if not, executing step S41; if so, executing step S5.

[0138] The first preset condition may include whether the number of commutation times, transformer imbalance, and total station imbalance are all less than corresponding thresholds. If so, the source-load commutation structure obtained in step S3 is the optimal solution. If not, further search for the optimal source-load commutation solution is required.

[0139] Step S41: shrink the imbalance calculation level and recalculate the node imbalance.

[0140] Among them, the imbalance calculation level can be divided according to the low-voltage substation topology structure. Specifically, if the topological level of the low-voltage substation topology structure is T, the imbalance calculation level may include the T-1 layer. For example, when the low-voltage substation topology structure is a four-layer substation structure, and from top to bottom are transformers, meter boxes, branches and users, the three-level imbalance calculation level may include transformers, meter boxes and branches, the two-level imbalance calculation level may include transformers and meter boxes, and the one-level imbalance calculation level may include transformers. Exemplarily, if the current level is a three-level imbalance calculation level, the imbalance calculation level is shrunk to a two-level imbalance calculation level, and then the three-phase current imbalance of the transformer node and the meter box node in the two-level imbalance calculation level is recalculated to obtain the imbalance of each node.

[0141] Step S42, determining whether the node imbalance is greater than a preset node imbalance threshold; if so, executing step S43; if not, executing step S5.

[0142] Specifically, the calculated imbalances for each node are weighted and summed to obtain the total imbalance for the current imbalance calculation level. If the total imbalance for the current imbalance calculation level is still greater than the preset imbalance threshold, the process returns to step S2 and re-performs the source-load switching optimization. If the total imbalance for the current imbalance calculation level is less than or equal to the preset node imbalance threshold, step S5 is executed. In a more detailed embodiment, the calculated imbalances for each node can be weighted and summed with the level weights obtained in step S209 to obtain the total imbalance for the current imbalance calculation level.

[0143] Step S43, determining whether the unbalance calculation level before contraction is the transformer level; if not, returning to step S2; if so, executing step S44.

[0144] The transformer level only includes transformer nodes. If the imbalance calculation level before contraction is not the transformer level, the imbalance calculation level needs to be contracted to a smaller range, for example, from a three-level imbalance calculation level to a two-level imbalance calculation level. The system then returns to step S2, reconstructing the multi-objective source-load commutation optimization control model based on each node and the phase to which each node belongs in the two-level imbalance calculation level. The Pareto front-aware optimization algorithm is then used to solve the multi-objective source-load commutation optimization control model corresponding to the three-level imbalance calculation level to obtain an updated source-load commutation result. If the imbalance calculation level before contraction is already the transformer level, step S44 is executed.

[0145] Step S44: generating a photovoltaic inverter output power regulation strategy.

[0146] If three-phase imbalance still exists after the imbalance calculation level is reduced to the transformer level, the transformer level is balanced according to the generated PV inverter output power regulation strategy. The PV inverter output power regulation strategy is based on the principle of local consumption of generated power and rotation adjustment.

[0147] Specifically, historical control records and other relevant parameters (such as the current output power of the PV inverter, node voltage, current, and other key parameters) are collected to assess the user's current power generation and real-time consumption. If a user's power generation exceeds its consumption, the excess power is consumed preferentially within that user or its immediate surroundings. If the user's local load is high, the user's power generation is not adjusted. Nodes with strong power generation capacity and low load demand are ranked by power, with lower rankings giving them higher priority. PV control is then adjusted in a rotating order of priority, from high to low, until equilibrium is achieved.

[0148] Step S5: generating a flexible control strategy for the substation based on the source-load commutation result and the photovoltaic inverter output power adjustment strategy.

[0149] Specifically, the source-load commutation result obtained in step S4 is combined with the photovoltaic inverter output power regulation strategy obtained in step S44 to form a complete flexible control strategy for the substation. According to the flexible control strategy for the substation, source-load commutation is performed to adjust the output power of the photovoltaic inverter, enabling dynamic adjustment of the source-load equipment within the substation to optimize the operating status of each node within the substation.

[0150] This embodiment firstly reduces the imbalance calculation level step by step when the source-load phase-changing result does not meet the first preset condition, updates the imbalance of the substation, and realizes a step-by-step decreasing control scheme for the three-phase imbalance of the power grid. This step-by-step control method can extend the balance area as far as possible to the end of the substation, thereby reducing the overall imbalance of the substation, and thus can effectively and comprehensively solve the problem of three-phase imbalance of the power grid. Secondly, by constructing a multi-objective source-load phase-changing optimization control model based on the low-voltage substation topology and the phase to which each node in the substation belongs, and solving it using the Pareto frontier perception optimization algorithm, a source-load phase-changing result that achieves a balance between multiple objectives is obtained, further improving the reliability of the substation flexible control strategy.

[0151] In one embodiment, the multi-objective source-load commutation optimization control model includes an objective function of minimizing the overall power consumption imbalance in the substation area and an objective function of minimizing the number of commutation switch operations. Step S2 includes:

[0152] Step S200: Construct the R node three-phase outlet power consumption matrix for a sampling period . R node three-phase export power consumption matrix The expression of is shown in formula (2):

[0153] (2)

[0154] Wherein, represents the A-phase export power consumption obtained by the i-th sampling of the R node, represents the B-phase export power consumption obtained by the i-th sampling of the R node, represents the C-phase export power consumption obtained by the i-th sampling of the R node, and N represents the total number of sampling times during the operation cycle.

[0155] Step S201: Calculate the power consumption of the A phase outlet of the R node within the sampling period T according to the formula shown in formula (3): , Power consumption of phase B outlet and the power consumption of phase C outlet ;

[0156] ( A, B, C}) (3)

[0157] Step S202: The sum of the power consumption of the three-phase outlets A, B, and C of the R node is taken as the total power consumption of the outlet of the R node; the total power consumption of the outlet of the R node The expression of is shown in formula (4):

[0158] (4)

[0159] Step S203: To make the power consumption of phases A, B, and C of the R node as evenly distributed as possible, the optimal power consumption of phase A that needs to adjust the source and load is calculated according to the formula shown in formula (5). , Phase B needs to adjust the optimal power consumption of the source and load The optimal power consumption of the source and load needs to be adjusted for phase C ;

[0160] (5)

[0161] Step S204: Obtain the power consumption vector of the adjustable source load connected to the lower end of the R node within the sampling period T. Phase sequence matrix of single-phase source and load of each phase of the lower node before R node adjustment ;in, Indicates the phase sequence of the nth single-phase source and load before adjustment. The expression of is shown in formula (6):

[0162] (6)

[0163] The phase connected to the single-phase source load is usually single-phase. The schematic diagram of the adjustable source load connected to the lower end of the R node is connected to the low-voltage distribution system as shown in the figure. Figure 2 As shown in the figure, They represent the n load values ​​connected to the lower end of the R node.

[0164] Step S205: The power consumption vector of the adjustable source load connected to the lower end of the R node is Phase sequence matrix of single-phase source and load of each phase of the lower node before R node adjustment The difference is multiplied as the sum of the single-phase source and load power consumption of each phase in the lower level before the R node is adjusted. Specifically, the expression of the sum of the single-phase source and load power consumption of each phase in the lower level before the R node is adjusted is shown in formula (7):

[0165] (7)

[0166] Where, It represents the sum of the power consumption of the lower-level A-phase single-phase source and load before the R node is adjusted. It represents the sum of the power consumption of the lower-level B-phase single-phase source and load before the R node is adjusted. It represents the sum of the power consumption of the lower-level C-phase single-phase source and load before the R node is adjusted.

[0167] Step S206: Obtain the phase sequence matrix of the lower nodes after the phase sequence of the single-phase load of the R node is adjusted. And adjust the phase sequence of the single-phase load of the R node to the phase sequence matrix of the lower node The power consumption vector of the adjustable source load connected to the lower end of the R node The difference is multiplied as the sum of the single-phase source and load power consumption of each phase in the lower level after the R node is adjusted. Specifically, the expression of the sum of the single-phase source and load power consumption of each phase in the lower level after the R node is adjusted is shown in formula (8):

[0168] (8)

[0169] Where, It represents the sum of the power consumption of the lower-level A-phase single-phase source and load after the R node is adjusted. It represents the sum of the power consumption of the lower-level B-phase single-phase source and load after the R node is adjusted. It represents the sum of the power consumption of the lower-level C-phase single-phase source and load after the R node is adjusted.

[0170] In step S207, the difference between the sum of the single-phase source-load power consumption of each phase of the lower level after the R node is adjusted and the sum of the single-phase source-load power consumption of each phase of the lower level before the R node is adjusted is used as the change in the power consumption of each phase before and after the three-phase load adjustment of the R node. The specific expression is shown in formula (9):

[0171] (9)

[0172] Where, Indicates the change in power consumption of phase A before and after the three-phase load adjustment of node R. Indicates the change in power consumption of phase B before and after the three-phase load adjustment of node R. Indicates the change in C-phase power consumption before and after the three-phase load adjustment at the R node.

[0173] Step S208: Calculate the difference between the change in power consumption of each phase before and after the three-phase load adjustment at the R node and the optimal power consumption of each phase that needs to be adjusted. The expression for the difference in optimal power consumption of each phase that needs to be adjusted is shown in formula (10):

[0174] (10)

[0175] Where, It represents the difference between the change in the power consumption of phase A before and after the three-phase load adjustment at node R and the optimal power consumption of phase A that needs to adjust the source load. It represents the difference between the change in the power consumption of phase B before and after the three-phase load adjustment of node R and the optimal power consumption of phase B that needs to adjust the source load. It represents the difference between the change in power consumption of phase C before and after the three-phase load adjustment at node R and the optimal power consumption of phase C that requires source load adjustment.

[0176] Step S209: Obtain the hierarchical weight of level m where R node is located. For the substation area structure with topological level T, the calculation formula of the hierarchical weight of the m-th level is shown in Equation (11):

[0177] (11)

[0178] In the formula, represents the hierarchical weight of the m-th layer, and m > 0, indicating that the calculation level is at least 1 layer; s represents the current three-phase unbalance degree calculation level, 0 < s < T. The minimum three-phase unbalance degree calculation level is 1 layer, and the maximum is one layer less than the topological level. For example, for a four-level topological level, the maximum unbalance degree calculation level is three layers.

[0179] Step S210: Calculate the maximum power consumption unbalance degree before and after the three-phase load adjustment of the R node according to the formula shown in Equation (12) ;

[0180] (12)

[0181] Assume that the R node belongs to the (s - 1)-th layer. If the three-phase power consumption unbalance degree of the R node is the smallest, there exists a phase sequence matrix E such that the maximum value in the weighted is the smallest, that is, is the smallest, where represents the maximum difference before and after the adjustment of the R node in the (s - 1)-th layer.

[0182] Step S211: Calculate the average power consumption unbalance degree of all three-phase nodes in the m-th layer respectively according to the formula shown in Equation (13) :

[0183] (13)

[0184] In the formula, j represents the total number of three-phase nodes in the m-th layer, represents the maximum power consumption unbalance degree of node i in the m-th layer.

[0185] Step S212: Construct an objective function for minimizing the overall power consumption unbalance degree of the substation area based on the average power consumption unbalance degree of all three-phase nodes in each layer. For a T-layer distribution network, the expression of the objective function for minimizing the overall power consumption unbalance degree of the substation area is shown in Equation (14):

[0186] (14)

[0187] In the formula, represents the average power consumption unbalance degree of all three-phase nodes in the k-th layer.

[0188] Step S213: The phase sequence matrix of the lower node is adjusted according to the phase sequence of the single-phase load of the R node. Get the number of commutation switch actions of n adjustable sources and loads in the substation. The expression of is shown in formula (15):

[0189] (15)

[0190] Where, Indicates whether the phase-changing switch of the i-th adjustable source load is in operation. If the phase sequence of the load changes before and after the phase adjustment, K=1; if the phase sequence of the load does not change before and after the phase adjustment, K=0.

[0191] Step S214: construct an objective function for minimizing the number of commutation switching operations based on the number of commutation switching operations. The expression of the objective function for minimizing the number of commutation switching operations is shown in formula (16):

[0192] (16)

[0193] This embodiment can improve the reliability of the source-load phase-changing results by constructing the objective function of minimizing the overall power consumption imbalance in the substation and the objective function of minimizing the number of phase-changing switch actions, so that the flexible control strategy of the substation finally generated is conducive to improving the operating efficiency of the power network, while reducing energy loss and extending the service life of equipment.

[0194] In one embodiment, the multi-objective source-load commutation optimization control model also includes constraints on transformer node power consumption imbalance, overall power consumption imbalance in the substation, and commutation switch action logic. Influenced by factors such as the transformer's operating life, the transformer node power consumption imbalance should be less than a certain value. The specific expression is shown in Equation (17):

[0195] (17)

[0196] Where, Indicates the imbalance of power consumption at transformer nodes, Indicates the transformer node imbalance threshold. For example, the transformer node imbalance threshold may be 5%.

[0197] Affected by factors such as power quality and line loss, the overall power consumption imbalance in the substation should be less than a certain value. The specific expression is shown in formula (18):

[0198] (18)

[0199] Where, Indicates the overall imbalance of electricity consumption in the area. Indicates the overall power consumption imbalance threshold of the substation. For example, the overall power consumption imbalance threshold of the substation can be 30%.

[0200] The expression of the reversing switch action logic constraint is shown in formula (19):

[0201] (19)

[0202] Where, Indicates the single-phase source load Q connected to the lower end of the R node m and Connection status of the phase busbar.

[0203] like Figure 2 As shown, taking the single-phase source load Q1 connected to the lower end of the R node as an example, its corresponding Indicates that Q1 is connected to The switch status on the phase bus, =0 means Q1 and Phase busbar is not connected, =1 means Q1 and Phase bus connection. If there are m adjustable source loads in the substation, since the phase-changing switch of each load has only one switching state at the same time, the mth source load must meet the constraint conditions shown in formula (19).

[0204] In one embodiment, the multi-objective source-load commutation optimization control model also includes a commutation switch operation number constraint. Considering that the number of phase modulations in the distribution network should be controlled within a reasonable range, the number of commutation switch operations can be constrained according to demand. The expression for the commutation switch operation number constraint is shown in Equation (20):

[0205] (20)

[0206] Where, Indicates the maximum number of commutation times. For example, the maximum number of commutation times may be 1 / 5 of the number of users that can commutate.

[0207] In one embodiment, a fast non-dominated sorting genetic algorithm is selected to solve the above multi-objective source-load commutation optimization control model. Specifically, Figure 3 As shown, step S3 includes:

[0208] Step S31: generating an initial population according to the access phases of the single-phase source and load.

[0209] The initial population is a set of possible phase adjustment schemes. Each phase adjustment scheme is a chromosome. Each chromosome corresponds to the phase sequence matrix of the lower node after the phase sequence of the single-phase load of the R node is adjusted. .

[0210] Step S32, calculating the power imbalance and commutation times of the initial population.

[0211] Specifically, the average power consumption imbalance corresponding to each chromosome in the initial population is calculated according to formula (13), and the number of phase changes corresponding to each chromosome in the initial population is calculated according to formula (15).

[0212] Step S33: Perform fast non-dominated sorting and crowding distance comparison on the initial population.

[0213] Specifically, the individuals in the initial population are sorted according to their non-dominated relationships, and the crowding distance between two adjacent individuals is calculated. The larger the crowding distance, the more evenly the individuals are distributed in the population.

[0214] Step S34: selection, crossover and mutation to generate offspring population.

[0215] The selection operation is used to select excellent individuals as parents based on non-dominated sorting and crowding distance. The crossover operation simulates the genetic recombination process in biological evolution, exchanging parts of the chromosomes of two parents to generate new offspring chromosomes. The mutation operation simulates the genetic mutation process in biological evolution, changing certain elements in the offspring chromosomes with a certain probability to increase population diversity.

[0216] Step S35, calculating the power imbalance and the number of commutation times of the offspring population.

[0217] Similarly, the average power consumption imbalance corresponding to each chromosome in the offspring population is calculated according to formula (13), and the number of phase changes corresponding to each chromosome in the offspring population is calculated according to formula (15).

[0218] Step S36: Merge the offspring population with the parent population, and perform fast non-dominated sorting and crowding distance comparison.

[0219] The offspring population is merged with the parent population to form a larger population. This larger population is then subjected to fast non-dominated sorting and crowding distance comparison to select the best individuals for the next generation.

[0220] Step S37, determine whether the stop condition is met; if so, execute step S38, if not, return to step S34.

[0221] The stopping condition can be reaching a preset number of iterations, meeting a certain performance indicator, or the population not showing significant improvement over several generations. If the stopping condition is met, the iteration stops; otherwise, the next step is executed.

[0222] Step S38: output the optimal phase difference.

[0223] Specifically, the optimal solution in the final population, namely the optimal phase difference, is output. This optimal phase difference corresponds to the optimal phase difference for each node's commutation switch. The corresponding imbalance value and commutation times for each node are also output, providing guidance for the development of flexible control strategies for the substation.

[0224] This embodiment uses a fast non-dominated sorting genetic algorithm to solve a multi-objective source-load commutation optimization control model to obtain the optimal access phase of each node commutation switch, thereby improving the reliability of the flexible control strategy of the substation.

[0225] Based on the above embodiment, in one embodiment, step S33 includes:

[0226] Step S330: Initialize the value of the variable e to 1.

[0227] Step S331: Select a Pareto solution set from the initial population, and record the genes in the Pareto solution set as the e-th dominance level.

[0228] Among them, the Pareto solution set is defined as: when the decision variable The corresponding objective function When the conditions shown in formula (21) are met, it is called Dominate ;

[0229] (twenty one)

[0230] If and only if within the feasible region of the decision variables, there is no Make Dominate , is called a non-dominated optimal solution, and the set of non-dominated optimal solutions is called a Pareto solution set (Pareto). The graph of the objective function corresponding to the Pareto solution set is presented as a boundary line or boundary surface in space. The decision variable in this embodiment is the phase sequence matrix of the lower-level node after the phase sequence of the single-phase load at the R node is adjusted. The objective functions include the objective function of minimizing the overall power consumption imbalance in the substation as shown in formula (14) and the objective function of minimizing the number of commutation switch actions as shown in formula (16).

[0231] Step S332: remove genes in the e-th dominance level from the initial population, reselect a Pareto solution set based on the remaining genes, and record the genes in the newly selected Pareto solution set as the e+1-th dominance level.

[0232] In step S333, the value of e is updated to e+1, and step S332 is repeated until all genes in the initial population are ranked, thereby obtaining a non-dominated ranking result for all genes in the initial population.

[0233] For example, after removing genes from the first dominance level, a Pareto solution set is selected from the remaining genes, and the genes in this solution set are recorded as the second dominance level. After removing genes from the second dominance level, a new Pareto solution set is selected from the remaining genes, and the genes in this solution set are recorded as the third dominance level. This process continues in this way until all genes in the population have been ranked, resulting in a non-dominated ranking of all genes in the population.

[0234] Step S334, calculating the crowding distance of each gene in the same dominance level according to the formula shown in formula (22);

[0235] (twenty two)

[0236] For chromosomes in the same dominance level, crowding distance is used for sorting. represents the crowding distance of gene i, represents the j-th objective function value of gene i+1, represents the jth objective function value of gene i-1, M represents the number of objective functions, and d represents the number of genes in the current dominant level. and The crowding distance of is recorded as infinite. For a solution, the larger its crowding distance is, the sparser the solution is. The solution with a large crowding distance is selected to ensure the diversity of the population.

[0237] Based on the above embodiment, in one embodiment, step S34 includes:

[0238] In step S340, genes that meet the iteration conditions are selected based on the non-dominated sorting results and the crowding distance to enter the next iteration. The expression of the iteration condition is shown in formula (23):

[0239] (twenty three)

[0240] Where, represents the non-dominated level of gene i, represents the non-dominated level of gene j, represents the crowding distance of gene j. Specifically, when ranking genes in a population, the non-dominated hierarchy is compared first, followed by the crowding distance. For any two genes i and j in a population, if the conditions shown in Equation (23) are met, gene i is said to have a priority in being retained and entering the next iteration due to gene j.

[0241] Step S341, randomly select two genes from the parent population and , calculate the two offspring chromosomes Q1 and Q2 according to the crossover operation calculation formula. The expression of the crossover operation calculation formula is shown in formula (24):

[0242] (twenty four)

[0243] in, The calculation formula is shown in formula (25):

[0244] (25)

[0245] Where, is the cross-distribution index, is a random number, . The size of indicates the similarity between offspring and parent. The larger the value, the more similar the offspring is to the parent.

[0246] In step S342, a chromosome is randomly selected from the parent population, and a daughter chromosome Q3 is generated according to the mutation operation calculation formula. The mutation operation calculation formula is shown in formula (26):

[0247] (26)

[0248] Where, represents the upper limit of the decision variable, represents the lower limit of the decision variable, is the variation distribution index, It is a random number ranging from 0 to 1.

[0249] In one embodiment, before step S1, the method further includes:

[0250] Step S01: Acquire the operation data of the distribution station area.

[0251] The operating data of the distribution substation can include voltage, current, power consumption, and PV inverter output power at each node outlet. Specifically, the distribution substation operating data is collected using electricity meters, and outlier screening and missing value filling are performed on the collected data to ensure data accuracy and reliability. The sampling interval can be 15 minutes.

[0252] Step S02: Calculate the load rate of the distribution area based on the operating data of the distribution area.

[0253] The substation load rate is the ratio of the substation's actual load to its rated load. The actual load can be determined by collecting the power of the substation's transformer. The rated load is typically determined based on the rated power and configuration of the substation's transformer.

[0254] Step S03, determining whether the load rate of the substation is greater than a preset load rate threshold; if so, executing step S04; if not, executing step S5.

[0255] The preset load rate threshold may be 20%, that is, if the load rate of the transformer area exceeds 20%, the three-phase imbalance of the transformer area will be adjusted when there is an imbalance.

[0256] Step S04, determining whether the area imbalance is greater than a preset area imbalance threshold; if so, executing step S1, if not, executing step S5.

[0257] This embodiment can timely discover and handle potential overload risks by judging whether the load rate of the substation is greater than a preset load rate threshold, thereby avoiding equipment damage and power outages and improving the efficiency of flexible regulation of the substation.

[0258] In one embodiment, after step S01 and before step S02, the method further includes:

[0259] Step S011, calculating the reverse power of the distribution area according to the operation data of the distribution area.

[0260] The reverse power of the substation may include the power output by the photovoltaic inverter to the grid.

[0261] Step S012, determining whether the reverse power of the substation is greater than a preset reverse power threshold; if so, executing step S1; if not, executing step S02.

[0262] The preset reverse power threshold may be when the reverse power exceeds 60% of the rated power of the transformer. For example, if the power output from the photovoltaic inverter to the grid exceeds 60% of the rated power of the transformer, step S1 is executed.

[0263] This embodiment enables real-time monitoring and early warning of the substation's operating status by determining whether the reverse power supply exceeds a preset reverse power threshold. Once the reverse power exceeds the threshold, the system can quickly respond and initiate a flexible control strategy to avoid safety hazards caused by reverse overload in the substation.

[0264] In one embodiment, step S44 includes:

[0265] Step S441: Generate a photovoltaic inverter output power regulation strategy based on the user's power generation capacity, load demand, and historical regulation records.

[0266] Specifically, the photovoltaic inverter output power regulation strategy is based on the basic principles of local consumption and rotation adjustment of power generation. It receives historical regulation records and other relevant parameters, and evaluates the user's current power generation and real-time consumption. If the user's power generation exceeds its consumption, priority is given to consuming excess power in the user or its neighboring areas. If the user's local load is large, the user's power generation power will not be adjusted.

[0267] This embodiment generates a photovoltaic inverter output power regulation strategy based on the user's power generation capacity, load demand, and historical regulation records, thereby improving the reliability of the flexible regulation strategy of the substation.

[0268] In order to facilitate a better understanding of the flexible control method of the substation based on Pareto front perception optimization in the above embodiment, a more detailed embodiment is provided below for explanation.

[0269] In one embodiment, Figure 4 As shown in FIG, the flexible control method of the substation based on Pareto frontier perception optimization specifically includes the following steps:

[0270] 1. When the substation flexible control reaches a specific operating cycle, the current and voltage data of each node in the substation and the generated power of the photovoltaic inverter are obtained in real time. The system first determines whether the substation reverse power supply is within the limit. If so, it further determines whether the substation load factor is greater than a threshold. If not, it obtains the low-voltage substation topology and the phases of each node. When the reverse power exceeds 60% of the transformer's rated power, the substation reverse power supply is considered to be exceeded. When the substation load factor is greater than the threshold, that is, when the ratio of the substation's actual load factor to the rated load factor is greater than 20%, it is determined whether the substation imbalance is greater than a threshold, i.e., the preset substation imbalance threshold in the above embodiment. The preset substation imbalance threshold in this embodiment includes the transformer imbalance threshold (5%) and the substation total imbalance threshold (30%). When the substation imbalance is less than or equal to the threshold, a phase adjustment strategy is generated based on the substation's current actual load, phase status, and other factors, combined with grid operation requirements and stability requirements. When the substation imbalance is greater than the threshold, the low-voltage substation topology and the phases of each node are obtained.

[0271] 2. Construct a multi-objective optimal commutation mathematical model based on Pareto frontier perception optimization to find the optimal solution. The multi-objective optimal commutation mathematical model includes the objective functions shown in Equations (14) and (16), and the constraints shown in Equations (17)-(20). Specifically, if the topology of the distribution network is a four-level structure, from top to bottom, it is transformer, meter box, branch, and user, with the transformer being the top level and the user being the bottom level, and including 490 user nodes (450 load nodes and 40 photovoltaic nodes), 75 branch nodes, and 10 meter box nodes. The user nodes are randomly connected to 75 branch nodes, the branch nodes are randomly connected to 10 meter box nodes, and the meter box nodes are connected to the transformer nodes. Then the topology level is 4, and the imbalance calculation level can be three-level imbalance calculation level: transformer, meter box, and branch, two-level imbalance calculation level: transformer and meter box, and one-level imbalance calculation level: transformer.

[0272] 3. Starting from the three-level imbalance calculation level, the Pareto frontier perception optimization algorithm is used to solve the multi-objective source-load phase-changing optimization control model to obtain the optimal solution, that is, the optimal phase sequence matrix. If the current number of phase changes, transformer imbalance and overall imbalance are all less than the corresponding thresholds, the single-phase source and load are regulated according to the optimal phase sequence matrix currently solved. If one of the current number of phase changes, transformer imbalance and overall imbalance does not meet the corresponding threshold, the imbalance calculation level is contracted to the second-level imbalance calculation level, and the three-phase nodes in the user level are removed and the multi-objective source-load phase-changing optimization control model is reused to obtain the optimal solution, that is, the imbalance of the removed three-phase nodes is not considered when solving the objective function. If the optimal solution obtained after contraction to the optimal level, that is, the transformer level, still does not meet the requirements, the photovoltaic inverter connected at the highest phase is searched to generate the photovoltaic inverter output power adjustment strategy, and the power generation power is further adjusted based on the optimal result of the transformer so that the transformer level reaches a balanced state, and finally a flexible control scheme is generated. Among them, Figure 5-Figure 7 The following plot shows the unbalance curves for each branch node before and after adjusting the unbalance in the substation using the generated flexible control scheme, when all user nodes are adjustable (all equipped with phase commutators). The figure shows that before optimization, the transformer node unbalance was 21.39%, the average unbalance at the meter box node was 44.49%, the average unbalance at the branch node was 66.73%, and the total unbalance in the substation was 36.65%. After optimization, the transformer node unbalance was 0.23%, the average unbalance at the meter box node was 7.83%, the average unbalance at the branch node was 64.26%, and the total unbalance in the substation was 13.44%. The number of phase changes was 99. Figures 8-10 The graph shows the change in imbalance at each branch node before and after adjustment, considering a three-level optimization hierarchy (transformer, meter box, and branch) with 45 randomly selected nodes (indicating only 45 phase converters installed). The graph shows that after optimization, the transformer node imbalance is 14.62%, the average imbalance at the meter box node is 39.22%, the average imbalance at the branch node is 65.75%, the total imbalance at the substation is 31.34%, and the number of phase changes is 14. The optimal result does not meet the specified requirements (the total imbalance exceeds the threshold), so the calculation hierarchy is reduced to two levels (transformer and meter box). After reduction, the number of phase changes is 21, and the optimized transformer node imbalance is 11.72%, the average imbalance at the meter box node is 40.10%, the average imbalance at the branch node is 66.96%, and the total imbalance at the substation is 19.22%.

[0273] Furthermore, transformer node data was randomly generated in a laboratory environment, with the maximum PV power set at 5000, the maximum transformer power set at 50000, the transformer reverse power adjustment factor set at -0.2, the imbalance threshold set at 0.3, and the adjusted imbalance threshold set at 0.1. Three phases were assigned to the transformer node, each with two loads and one PV module. The load generated power was randomly set between 0 and 2000, and the PV generated power was randomly set between 0 and 5000. A data point was generated (in this case, the PV module typically exceeds the load, resulting in reverse power). A total of 96 data points were randomly generated. The test results of the PV control algorithm are shown in Figure 11. Figure 11(a) shows the imbalance curve before and after PV control, which shows that the imbalance can be greatly reduced after control. Figure 11(b) shows the PV power generation curve before and after control, which shows that PV output increases after control, improving energy utilization. Figure 11(c) shows the load and PV output curves of phase A before control, and after control. Figure 11(d) shows the load and PV output curves of phase B before control, and after control. Figure 11(e) shows the load and PV output curves of phase C before control, and after control. It can be seen that the control algorithm maximizes the PV output of phases A / B / C while ensuring balance. Figure 11(f) shows the phase A supply ratio curve before and after control. Figure 11(g) shows the phase B supply ratio curve before and after control. Figure 11(h) shows the phase C supply ratio curve before and after control. It can be seen that the control algorithm minimizes the reverse supply rate while ensuring balance. These results show that this method can maximize PV output while ensuring three-phase balance and preventing reverse supply, significantly improving clean energy utilization efficiency.

[0274] The Pareto front-aware, optimized flexible control method for substations in this embodiment combines phase adjustment with power generation adjustment strategies to optimize source-load distribution. Real-time online adjustment using photovoltaic inverters and commutators optimizes the source-load phase sequence and photovoltaic inverter output power, mitigating the risk of three-phase imbalance. A multi-objective optimization method and hierarchical contraction strategy are used to extend the balanced region to the end of the substation. A multi-objective function is designed to establish a multi-objective source-load commutation optimization control model for the low-voltage distribution network. A step-by-step, hierarchical contraction strategy is designed to maximize overall substation balance, improve the operating efficiency and stability of the distribution network, and reduce line losses. A photovoltaic inverter output power adjustment strategy is used to achieve localized power consumption and rotational adjustment. Based on the principles of localized power consumption and rotational adjustment, the photovoltaic inverter output power adjustment strategy maximizes the utilization of local power generation and reduces energy waste. By combining user power generation capacity, load demand, and historical control records, photovoltaic power generation is dynamically adjusted, improving the stability and efficiency of the overall power system.

[0275] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned 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 specification.

[0276] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A flexible control method for substations based on Pareto frontier perception optimization, characterized in that: The method comprises: Step S1, obtaining a low-voltage area topology structure and a phase of each node in the area; the low-voltage area topology structure is a four-layer area structure, and from top to bottom are transformers, meter boxes, branches and users; Step S2, constructing a multi-objective source-load phase-changing optimization control model based on the low-voltage area topology structure and the phase to which each node in the area belongs; Step S3, using the Pareto frontier perception optimization algorithm to solve the multi-objective source-load commutation optimization control model to obtain a source-load commutation result; Step S4, judging whether the source-load commutation result satisfies the first preset condition; if not, executing step S41; if yes, executing step S5; the first preset condition is whether the commutation times, transformer unbalance and total unbalance of the substation are all less than the corresponding threshold values ​​respectively; Step S41, shrinking the imbalance calculation level, and recalculating the node imbalance; the imbalance calculation level is divided according to the low voltage area topology structure, including a three-level imbalance calculation level, a two-level imbalance calculation level and a one-level imbalance calculation level; wherein the three-level imbalance calculation level includes a transformer, a meter box and a branch, the two-level imbalance calculation level includes a transformer and a meter box, and the one-level imbalance calculation level includes a transformer; Step S42, determining whether the node imbalance is greater than a preset node imbalance threshold; if so, executing step S43; if not, executing step S5; Step S43, determining whether the unbalance calculation level before contraction is the transformer level; if not, returning to step S2; if yes, executing step S44; Step S44, generating a photovoltaic inverter output power regulation strategy; Step S5, generating a flexible control strategy for the substation according to the source-load switching result and the photovoltaic inverter output power adjustment strategy; The multi-objective source-load commutation optimization control model includes an objective function of minimizing the overall power consumption imbalance in the substation area and an objective function of minimizing the number of commutation switch actions. Step S2 includes: Step S200: construct a three-phase outlet power consumption matrix of the R node in a sampling period ; The R node three-phase export power consumption matrix The expression is as follows: ; In the formula, represents the A-phase outlet power consumption obtained by the i-th sampling of the R node, represents the B-phase outlet power consumption obtained by the i-th sampling of the R node, represents the C-phase outlet power consumption obtained by the i-th sampling of the R node, and N represents the total number of sampling times in the operation cycle; Step S201, according to the formula ; Calculate the power consumption of phase A outlet of node R within sampling period T , Power consumption of phase B outlet And the power consumption of phase C outlet ; Step S202: The sum of the power consumption of the three-phase outlets A, B, and C of the R node is taken as the total power consumption of the outlet of the R node; the total power consumption of the outlet of the R node The expression is as follows: ; Step S203, according to the formula ; Calculate the optimal power consumption of phase A to adjust the source and load 、B phase needs to adjust the optimal power consumption of the source load The optimal power consumption of the source and load needs to be adjusted for phase C ; Step S204, obtaining the power consumption vector of the adjustable source load connected to the lower end of the R node within the sampling period T The phase sequence matrix of the single-phase source and load of each phase of the lower node before the R node is adjusted ;in, Indicates the phase sequence of the nth single-phase source load before adjustment. The expression is: ; Step S205: The power consumption vector of the adjustable source load connected to the lower end of the R node is The phase sequence matrix of the single-phase source and load of each phase of the lower node before the R node adjustment The difference is multiplied as the sum of the single-phase source and load power consumption of each phase of the lower level before the R node is adjusted; the expression of the sum of the single-phase source and load power consumption of each phase of the lower level before the R node is adjusted is as follows: ; In the formula, It represents the sum of the power consumption of the lower-level A-phase single-phase source and load before the R node is adjusted. It represents the sum of the power consumption of the lower-level B-phase single-phase source and load before the R node is adjusted. It represents the sum of the power consumption of the lower-level C-phase single-phase source and load before the R node is adjusted; Step S206, obtaining the phase sequence matrix of the lower node after the phase sequence of the single-phase load of the R node is adjusted , and adjust the phase sequence of the single-phase load at the R node to the phase sequence matrix of the lower node and the power consumption vector of the adjustable source load connected to the lower end of the R node The difference is multiplied as the sum of the single-phase source and load power consumption of each phase of the lower level after the R node is adjusted; the expression of the sum of the single-phase source and load power consumption of each phase of the lower level after the R node is adjusted is as follows: ; In the formula, It represents the sum of the power consumption of the lower-level A-phase single-phase source and load after the R node is adjusted. It represents the sum of the power consumption of the lower-level B-phase single-phase source and load after the R node is adjusted. It represents the sum of the power consumption of the lower-level C-phase single-phase source and load after the R node is adjusted; Step S207, the difference between the sum of the single-phase source load power consumption of each phase of the lower level after the R node is adjusted and the sum of the single-phase source load power consumption of each phase of the lower level before the R node is adjusted is used as the change in the power consumption of each phase before and after the three-phase load adjustment of the R node. The specific expression is as follows: ; In the formula, It indicates the change of power consumption of phase A before and after the three-phase load adjustment of node R. Indicates the change in power consumption of phase B before and after the three-phase load adjustment at node R. It indicates the change of power consumption of phase C before and after the three-phase load adjustment of node R; Step S208, calculating the difference between the change in the power consumption of each phase before and after the three-phase load adjustment of the R node and the optimal power consumption of each phase source load that needs to be adjusted. The expression of the difference in the optimal power consumption of each phase source load that needs to be adjusted is as follows: ; In the formula, It represents the difference between the change in the power consumption of phase A before and after the three-phase load adjustment at node R and the optimal power consumption of phase A that needs to adjust the source load. It represents the difference between the change in the power consumption of phase B before and after the three-phase load adjustment at the R node and the optimal power consumption of phase B that needs to adjust the source load. It represents the difference between the change in the power consumption of phase C before and after the three-phase load adjustment at the R node and the optimal power consumption of phase C that needs to adjust the source load; Step S209, obtaining the level weight of the level m where the R node is located; the calculation formula of the level weight is as follows: ; In the formula, represents the level weight of the mth layer, and s represents the current three-phase unbalance calculation level; Step S210, according to the formula ; Calculate the maximum power consumption imbalance before and after the three-phase load adjustment of the R node ; Step S211, according to the formula ; Calculate the average power consumption imbalance of all three-phase nodes in the mth layer ; In the formula, j represents the total number of three-phase nodes in the mth layer, represents the maximum power consumption imbalance of the i-th node in the m-th layer; Step S212: constructing the minimum objective function of the overall power consumption imbalance of the substation based on the average power consumption imbalance of all three-phase nodes in each layer; the expression of the minimum objective function of the overall power consumption imbalance of the substation is as follows: ; In the formula, represents the average power consumption imbalance of all three-phase nodes in the kth layer; Step S213, the phase sequence matrix of the lower node after the phase sequence of the single-phase load of the R node is adjusted The number of phase-changing switch actions of n adjustable sources and loads in the station area is obtained; the number of phase-changing switch actions The expression is as follows: ; In the formula, Indicates whether the commutation switch of the i-th adjustable source load is in operation; Step S214, constructing the objective function of minimizing the number of phase-changing switch actions based on the number of phase-changing switch actions; the expression of the objective function of minimizing the number of phase-changing switch actions is: .

2. According to the method of flexible control of substations based on Pareto frontier perception optimization according to claim 1, it is characterized in that: The multi-objective source-load commutation optimization control model also includes transformer node power consumption imbalance constraint, area overall power consumption imbalance constraint and commutation switch action logic constraint; wherein, the transformer node power consumption imbalance constraint is expressed as: ; In the formula, Indicates the imbalance of power consumption at the transformer node. Indicates the transformer node unbalance threshold; The expression of the overall power consumption imbalance constraint in the substation area is: ; In the formula, Indicates the overall imbalance of power consumption in the area. Indicates the overall power consumption imbalance threshold of the substation; The expression of the reversing switch action logic constraint is: ; In the formula, Indicates the single-phase source load Q connected to the lower end of the R node m and Connection status of the phase bus.

3. The method for flexible control of substations based on Pareto frontier perception optimization according to claim 2 is characterized in that: The multi-objective source-load commutation optimization control model also includes a commutation switch action times constraint, and the expression of the commutation switch action times constraint is: ; In the formula, Indicates the maximum number of commutation times.

4. The flexible control method for substations based on Pareto frontier perception optimization according to claim 1 is characterized in that: The step S3 comprises: Step S31, generating an initial population according to the access phases of the single-phase source and load; Step S32, calculating the power imbalance and the number of commutation times of the initial population; Step S33, performing fast non-dominated sorting and crowding distance comparison on the initial population; Step S34, selecting, crossovering and mutating to generate a progeny population; Step S35, calculating the power imbalance and the number of commutation times of the offspring population; Step S36, merging the offspring population with the parent population, and performing fast non-dominated sorting and crowding distance comparison; Step S37, determine whether the stop condition is met; if so, execute step S38, if not, return to step S34; Step S38, output the optimal phase difference.

5. The flexible control method for substations based on Pareto frontier perception optimization according to claim 4 is characterized in that: The step S33 comprises: Step S330, initializing the value of variable e to 1; Step S331, selecting a Pareto solution set from the initial population, and recording the genes in the Pareto solution set as the e-th dominance level; Step S332, removing genes in the e-th dominance level from the initial population, reselecting the Pareto solution set based on the remaining genes, and recording the genes in the newly selected Pareto solution set as the e+1-th dominance level; Step S333, updating the value of e to e+1, and repeating step S332 until the genes in the initial population are completely classified, thereby obtaining a non-dominated ranking result of all genes in the initial population; Step S334, according to the formula ; Calculate the crowding distance of each gene in the same dominance level; where, represents the crowding distance of gene i, represents the j-th objective function value of gene i+1, represents the j-th objective function value of gene i-1, M represents the number of objective functions, and d represents the number of genes in the current dominating level.

6. The flexible control method for substations based on Pareto frontier perception optimization according to claim 5 is characterized in that: The step S34 comprises: Step S340, based on the non-dominated sorting result and the crowding distance, select genes that meet the iteration condition to enter the next iteration; the expression of the iteration condition is as follows: ; In the formula, represents the non-dominated level of gene i, represents the non-dominated level of gene j, represents the crowding distance of gene j; Step S341, randomly select two genes from the parent population and , calculate the two offspring chromosomes Q1 and Q2 according to the crossover operation calculation formula; the expression of the crossover operation calculation formula is as follows: ; in, The calculation formula is as follows: ; In the formula, is the cross-distribution index, is a random number; Step S342, randomly select a chromosome from the parent population, and generate a child chromosome Q3 according to the mutation operation calculation formula; the expression of the mutation operation calculation formula is as follows: ; In the formula, represents the upper limit of the variable, represents the lower limit of the variable, is the variation distribution index, Is a random number.

7. The flexible control method for substations based on Pareto frontier perception optimization according to claim 1 is characterized in that: Before step S1, the method further includes: Step S01, obtaining the operation data of the distribution station area; Step S02, calculating the load rate of the distribution area according to the operation data of the distribution area; Step S03, determining whether the load rate of the station area is greater than a preset load rate threshold; if so, executing step S04; if not, executing step S5; Step S04, determining whether the area imbalance is greater than a preset area imbalance threshold; if so, executing step S1, if not, executing step S5.

8. The flexible control method for substations based on Pareto frontier perception optimization according to claim 7 is characterized in that: After step S01 and before step S02, the following further comprises: Step S011, calculating the reverse power of the distribution area according to the operation data of the distribution area; Step S012, determining whether the reverse power of the station area is greater than a preset reverse power threshold; if so, executing step S1; if not, executing step S02.

9. The flexible control method for substations based on Pareto frontier perception optimization according to claim 1 is characterized in that: The step S44 comprises: Step S441, generating the photovoltaic inverter output power regulation strategy according to the user's power generation capacity, load demand and historical regulation records.