Microgrid group dynamic partitioning method and system based on graph theory and non-dominated sorting

Abstract

The present invention discloses a method and system for dynamically partitioning a microgrid group based on graph theory and non-dominated sorting. The method comprises: generating a graph theory model of the microgrid group using a weighted undirected graph based on the corresponding data of the microgrid group; defining an improved modularity index, a power matching index, and a partition supply and demand coordination index based on the graph theory model; partitioning the microgrid group using a second-generation non-dominated sorting genetic algorithm; calculating the difference in partitioning indexes at different times and dynamically updating a weighted adjacency matrix; selecting a solution with the most balanced trade-offs between objectives by calculating the target gradient change rate of the solution; outputting a visualized partitioning result of the microgrid group based on the updated weighted adjacency matrix; iteratively updating the partitioning result and outputting the final partitioning result of the microgrid group. The present invention solves the problems of difficulty in characterizing electrical network parameter characteristics, the number of partitions of a microgrid group and the partition to which each microgrid belongs, and the problem of nonlinear conflicts between objectives.

Description

Technical Field

[0001] The present invention belongs to the field of power grids and relates to microgrid optimization and control technology, and specifically to a method and system for dynamically partitioning microgrid groups based on graph theory and non-dominated sorting. Background Art

[0002] Driven by the development of new power systems and power electronics technologies, the number of adjustable resources in microgrid clusters (e.g., flexible, adjustable loads such as air conditioners and electric vehicles, distributed, controllable power generation equipment such as diesel generators and fuel cells, and various energy storage devices) has grown exponentially, transforming adjustable parameters from a single dimension to multiple dimensions. Existing centralized and distributed optimization methods struggle to achieve optimal scheduling strategies within a limited timeframe. Consequently, current research often employs a "dimensionality reduction control" approach to address this problem. This involves dividing a microgrid cluster into multiple self-sufficient partitions, each of which achieves energy management within the microgrid cluster through both internal optimization and coordinated optimization between microgrids. This approach requires only the microgrids within a partition to exchange information, significantly reducing communication requirements and the amount of information transmitted, simplifying the complexity of controlling massive amounts of equipment.

[0003] The rational allocation, scheduling, and energy management of various internal resources are based on a rational regional division. The design of appropriate division indicators and the adoption of efficient division algorithms are fundamental and crucial to regional division. Division indicators can be broadly categorized into structural and functional indicators. Structural indicators aim to ensure close connectivity between nodes within subregions and loose connectivity between nodes across regions, and primarily include metrics such as modularity, node membership, and stability. Functional indicators aim to optimize regional regulation. Depending on the optimization requirements, functional indicators in different studies will exhibit varying degrees of effectiveness. These indicators primarily include power balance, cluster absorption capacity, and source-load-storage matching. Division algorithms primarily include clustering algorithms, intelligent optimization algorithms, and complex network-based community detection algorithms. It is noteworthy that existing research primarily focuses on typical operating scenarios and fixed network topologies, often using fixed division methods to divide microgrid clusters, resulting in results that do not change over time. However, fixed partitioning methods fail to account for the temporal variations and spatial migration of renewable energy output and loads. This can result in fixed microgrid group partitioning results being unable to adapt to the dynamic spatiotemporal power variations of renewable energy and loads, or changes in network topology, making their optimization strategies unsuitable for energy management in different operating scenarios. Building on this, some studies have proposed dynamic partitioning methods. These methods can update the microgrid group partitioning results based on the system's real-time operating status, helping to improve the applicability of optimization strategies across different operating scenarios. Therefore, constructing reasonable partitioning metrics and using efficient partitioning algorithms to dynamically partition microgrid groups can help improve the applicability of optimization strategies across different operating scenarios.

[0004] The existing technical solution mainly combines the reactive balance and active balance indicators with the modularity indicator as the partitioning indicator. Then, the adjacency matrix in the undirected graph is used to encode the chromosomes in the genetic algorithm, and new individuals are generated by performing crossover and mutation operations on the adjacency matrix. The modified adjacency matrix is ​​composed of multiple block matrices, each of which represents a partition in the distribution network partition result. The non-zero square matrix row or column number in the block matrix represents the distribution network node number in the partition. However, this solution has the following disadvantages:

[0005] 1. Since the adjacency matrix in an undirected graph contains only two elements, 0 and 1, it can only indicate whether there is a line connection between microgrids, but cannot represent the electrical parameter characteristics such as line resistance and transmission capacity.

[0006] 2. Nodes in the same block matrix in the adjacency matrix are grouped into the same partition, and the partition results of the microgrid cluster can be intuitively obtained from the block matrix. However, when the node numbers within the same partition in the adjacency matrix are discontinuous, the same partition in the adjacency matrix will have multiple block matrices. The partition results of the microgrid cluster cannot be directly obtained from the block matrix in the adjacency matrix. The adjacency matrix must be converted into an undirected graph to obtain the number of partitions of the microgrid cluster and the partition to which each microgrid belongs. The modularity index, power matching index, and partition supply and demand coordination index are then calculated, resulting in low algorithm efficiency.

[0007] 3. Existing technologies often use a linear weighting approach to normalize multiple partitioning metrics, multiply them by weight coefficients, and then perform a weighted sum. First, the weight setting in linear weighting often relies on subjective judgment or trial and error, lacking theoretical basis. Second, linear weighting assumes that the partitioning metrics are independent of each other, but in reality, nonlinear conflicts exist between different partitioning metrics. Summary of the Invention

[0008] Objectives of the invention: To overcome the shortcomings of the prior art, a method and system for dynamic partitioning of microgrid clusters based on graph theory and non-dominated sorting are provided. A graph theory model of a microgrid cluster is constructed using a weighted undirected graph, thereby obtaining a weighted adjacency matrix containing electrical parameters such as line impedance and transmission capacity, thereby solving the problem that the adjacency matrix encoding based on the undirected graph is difficult to characterize the characteristics of electrical network parameters. A depth-first search (DFS) algorithm is used to identify the connected components in the weighted adjacency matrix to obtain the number of partitions of the microgrid cluster and the partition number to which each microgrid belongs, thereby solving the problem that the number of partitions of the microgrid cluster and the partition to which each microgrid belongs cannot be directly obtained from the block matrix in the adjacency matrix when the microgrid numbering sequence in the microgrid cluster is disordered. A second-generation non-dominated sorting genetic algorithm (NSGA-II) is used to simultaneously optimize different partitioning indicators to solve the problem that the existing linear weighting method of partitioning indicators relies on subjective weight setting and there is a nonlinear conflict between multiple optimization objectives.

[0009] Technical solution: To achieve the above objectives, the present invention provides a method for dynamically partitioning a microgrid group based on graph theory and non-dominated sorting, comprising the following steps:

[0010] S1: Based on the corresponding data of the microgrid group, a graph theory model of the microgrid group is generated using a weighted undirected graph;

[0011] S2: Based on the graph theory model, the improved modularity index, power matching index and partition supply and demand coordination index are defined respectively;

[0012] S3: Based on the indicators defined in step S2, the second generation non-dominated sorting genetic algorithm is used to divide the microgrid group;

[0013] S4: Calculate the difference of partitioning indexes at different times and dynamically update the weighted adjacency matrix;

[0014] S5: Select the solution with the most balanced trade-off between objectives by calculating the target gradient change rate of the solution;

[0015] S6: Output the visualization division result of the microgrid group according to the updated weighted adjacency matrix;

[0016] S7: Iteratively update the division results and output the final division results of the microgrid group.

[0017] Furthermore, the data corresponding to the microgrid group in step S1 includes the microgrid group topology, node number, line resistance and transmission capacity, photovoltaic output of each microgrid, wind turbine output and load forecast power data.

[0018] Furthermore, in step S1, a weighted undirected graph is used. Generate a graph model of microgrid cluster, the microgrid cluster is constructed as a vertex set , Representative microgrids, edge collection Represents the transmission lines connecting adjacent microgrids, and each edge in the edge concentration , weighted adjacency matrix Used to store the network topology and weight information of the microgrid group.

[0019] Furthermore, the weighted adjacency matrix of step S1 is middle:

[0020]

[0021] In the formula, the edge weight set For storage of transmission lines Resistance and transport capacity .

[0022] Furthermore, the definition of the improved modularity index in step S2 includes:

[0023] Introducing modularity indicators The indicators are quantified and expressed as follows:

[0024]

[0025] Where, for The weighted adjacency matrix at the moment Rank Elements of the column; express 1 / 2 of the sum of the weights of all edges in the partition at that moment; express All nodes at the moment The sum of the weights of the connected edges; express All nodes at the moment The sum of the weights of the connected edges; Used to indicate Time Node and Are they in the same partition and connected? ,express Time Node and In the same partition and connected, express Time Node and Not in the same partition and not connected.

[0026] Furthermore, the definition of the power matching index in step S2 includes:

[0027] The power matching index is introduced to quantify the index, which is expressed as follows:

[0028]

[0029]

[0030]

[0031]

[0032] Where, and Represents the power matching index values ​​of all partitions and Moment The average value of the power matching index of each partition; and Respectively represent The number of zones and microgrids in a zone; and Respectively Moment Adjustable active power and net load power in each microgrid; 、 、 、 and Respectively Moment The adjustable active power of photovoltaic, wind turbines, adjustable loads, energy storage and diesel generators in each microgrid; 、 and Respectively Moment The real-time active power of photovoltaic, wind turbine and load in a microgrid.

[0033] Furthermore, the definition of the regional supply and demand coordination index in step S2 includes:

[0034] The regional supply and demand coordination index is introduced for quantification, which is expressed as follows:

[0035]

[0036]

[0037]

[0038] Where, and Respectively The average value and the first Supply and demand coordination index value of each zone; express The adjustable capacity of the electrical equipment that can be controlled at any time The difference between the net load powers within the two partitions.

[0039] Furthermore, the process of using the second generation non-dominated sorting genetic algorithm to divide the microgrid group in step S3 includes:

[0040] A1: Initialization parameters, including population size, maximum number of iterations, crossover probability range, mutation probability range, and dynamic update threshold;

[0041] A2: According to the weighted adjacency matrix Encodes chromosomes;

[0042] A3: Initialize the population, including: the upper triangular part of the initial weight matrix Randomly change to 0 or remain unchanged, and then synchronously copy the upper triangular elements to the lower triangular part to ensure the symmetry of the weighted matrix, and randomly generate the population;

[0043] A4: Weight threshold filtering, including: partition detection of the generated weighted adjacency matrix, retaining only those that meet the , construct a binary adjacency matrix , as shown below:

[0044]

[0045] A5: Traverse and identify connected components, including: using the depth-first search algorithm to traverse and identify connected components; marking all nodes as unvisited, and starting from any unvisited node Start by following the valid edge ( ) recursively visit its adjacent nodes; all visited nodes are marked as the same partition and assigned a unique partition label number, and this process is repeated until all nodes are visited; after the traversal is completed, the number of different labels is the number of partitions, and the label number of each node indicates the partition number to which it belongs;

[0046] A6: Based on the initial partition results of each individual in the population obtained by partition detection, multi-threaded parallel calculation is used to calculate the fitness function values ​​of the modularity index, power matching index, and partition supply and demand coordination index of each individual in the population;

[0047] A7: Non-dominated sorting, including: fast non-dominated sorting of populations and classification of Pareto frontier levels;

[0048] A8: Target congestion calculation, including: Calculating the congestion of each solution in the target space:

[0049]

[0050] Where, express Moment The degree of crowding of individuals; For the objective function values; and Represents all individuals in objective function The maximum and minimum values ​​under ;

[0051] A9: Elite selection, including: selecting the two individuals with the highest non-dominant rank and the largest crowding degree from the population;

[0052] A10: Crossover operation, including: the first triangle of the upper triangle of the binary adjacency matrix of the two selected parents Rank Column Elements Perform crossover, and then synchronously copy the elements of the crossover in the upper triangle of the parent's binary adjacency matrix to the lower triangle to maintain the symmetry of the parent's binary adjacency matrix;

[0053] A11: Mutation operation, including: when the upper triangle of the offspring binary adjacency matrix Rank Column Elements When , random mutation is performed or ; When the first Rank Column Elements Only when the first Rank Column Elements When the random variation is or , when the elements in the upper triangle of the original binary adjacency matrix When the elements in the upper triangle of the offspring binary adjacency matrix remain unchanged; then, the mutated elements of the upper triangle of the offspring binary adjacency matrix are synchronously copied to the lower triangle to maintain the symmetry of the offspring binary adjacency matrix;

[0054] A12: Elite retention, including: retaining high-quality solutions in the non-dominated frontier to avoid losing potential optimal solutions, and merging parent and child populations.

[0055] Furthermore, the step S4 specifically includes:

[0056] judge Is the modularity index function value obtained by binarizing the adjacency matrix at time t consistent with The modularity index function value obtained by the binary adjacency matrix at the time is the same; if they are the same, the binary adjacency matrix is ​​not updated; if they are not the same, the binary adjacency matrix is ​​judged Does the modularity index function value obtained by binarizing the adjacency matrix at the moment satisfy ,like And in The power matching index and partition supply and demand coordination index function values ​​obtained by the binary adjacency matrix at time The power matching index difference obtained by the binary adjacency matrix at time The difference between the partition supply and demand coordination index function value When the threshold is exceeded, the binary adjacency matrix is ​​updated; the expression is as follows:

[0057]

[0058] Where: and They are Moment and The binary adjacency matrix at time t; is a dynamically updated threshold;

[0059] when The binary adjacency matrix at time When there is an update, The binary adjacency matrix at time Get the corresponding weighted adjacency matrix .

[0060] The present invention also provides a microgrid group dynamic partitioning system based on graph theory and non-dominated sorting, comprising:

[0061] A graph theory model generation module, used for generating a graph theory model of a microgrid group using a weighted undirected graph;

[0062] The indicator definition module is used to define the improved modularity indicator, power matching indicator and partition supply and demand coordination indicator respectively;

[0063] A partitioning module, used to partition the microgrid group;

[0064] Update module, used to dynamically update the weighted adjacency matrix;

[0065] The iterative output module is used to iteratively update the division results and output the final division results of the microgrid group.

[0066] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0067] 1. A graph theory model of the microgrid cluster is constructed by using a weighted undirected graph. On this basis, a weighted adjacency matrix containing electrical parameters such as line impedance and transmission capacity is used to encode the chromosomes, realizing dual encoding of network topology and electrical characteristics, so that the partitioning results are more in line with actual physical constraints.

[0068] 2. The depth-first search algorithm is used to identify the connected components in the weighted adjacency matrix to intuitively obtain the division results of the microgrid group. This avoids the steps of converting the adjacency matrix into an undirected graph and then reading the division results from the undirected graph, effectively improving the operation efficiency of the algorithm.

[0069] 3. Use the second-generation non-dominated sorting genetic algorithm to optimize multiple objectives separately, and generate a uniformly distributed Pareto optimal solution set through non-dominated sorting and crowding calculation, avoiding defects such as subjectivity in weight selection, a single solution set, and poor adaptability caused by dynamically modifying weights through conditional judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 Schematic diagram of the process of the present invention;

[0071] Figure 2 Schematic diagram of the graph theory model of the microgrid group;

[0072] Figure 3 A schematic diagram of chromosome encoding;

[0073] Figure 4 This is the result diagram of dynamic partitioning of microgrid groups. DETAILED DESCRIPTION

[0074] The present invention is further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, modifications of various equivalent forms of the present invention made by those skilled in the art all fall within the scope defined by the claims attached to this application.

[0075] Example 1:

[0076] like Figure 1 As shown, this embodiment provides a method for dynamically partitioning a microgrid group based on graph theory and non-dominated sorting, comprising the following steps:

[0077] Step 1: Input the microgrid group topology, node number, line resistance and transmission capacity, each microgrid's photovoltaic output, wind turbine output, and load forecast power data.

[0078] Step 2: Use a weighted undirected graph Generate a graph theory model of the microgrid group, such as Figure 2 As shown. The microgrid group is constructed as a vertex set , Representative Microgrid. Represents the transmission lines connecting adjacent microgrids, and each edge in the edge concentration . Weighted adjacency matrix It is used to store the network topology and weight information of the microgrid group, as shown below:

[0079]

[0080] In the formula, the edge weight set For storage of transmission lines Resistance and transport capacity .

[0081] Step 3: Define the improved modularity metric.

[0082] In order to ensure that the partitions obtained after the microgrid group is divided maintain sufficient structural strength, the modularity index is introduced. It is quantized as follows:

[0083]

[0084] Where, for The weighted adjacency matrix at the moment Rank Elements of the column; express 1 / 2 of the sum of the weights of all edges in the partition at that moment; express All nodes at the moment The sum of the weights of the connected edges; express All nodes at the moment The sum of the weights of the connected edges; Used to indicate Time Node and Are they in the same partition and connected? ,express Time Node and In the same partition and connected, express Time Node and Not in the same partition and not connected.

[0085] Step 4: Define the power matching index.

[0086] The randomness and volatility of renewable energy and load power increase net load fluctuations. In order to utilize the adjustable resources within the microgrid cluster to smooth out net load fluctuations, a power matching index is introduced to quantify it, which is expressed as follows:

[0087]

[0088]

[0089]

[0090]

[0091] Where, and Represents the power matching index values ​​of all partitions and Moment The average value of the power matching index of each partition; and Respectively represent The number of zones and microgrids in a zone; and Respectively Moment Adjustable active power and net load power in each microgrid; 、 、 、 and Respectively Moment The adjustable active power of photovoltaic, wind turbines, adjustable loads, energy storage and diesel generators in each microgrid; 、 and Respectively Moment The real-time active power of photovoltaic, wind turbine and load in a microgrid.

[0092] Step 5: Define the zone supply and demand coordination indicators.

[0093] To reflect the supply and demand coordination relationship between distributed energy storage and net load in the microgrid group division, and to show the coordination degree between the adjustable resources and system operation requirements in each partition, a partition supply and demand coordination index is introduced for quantification, which is defined as follows:

[0094]

[0095]

[0096]

[0097] Where, and Respectively The average value and the first Supply and demand coordination index value of each zone; express The adjustable capacity of the electrical equipment that can be controlled at any time The difference between the net load powers within the two partitions.

[0098] Step 6: Use the second-generation non-dominated sorting genetic algorithm to divide the microgrid group. The specific process includes the following steps:

[0099] Step 6.1: Initialize parameters. Set the population size to 200, the maximum number of iterations to 50, the crossover probability range to [0.2, 0.5], the mutation probability range to [0.1, 0.3], and the dynamic update threshold to .

[0100] Step 6.2: The weighted adjacency matrix generated according to step 2 Encode the chromosomes, such as Figure 3 shown.

[0101] Step 6.3: Initialize the population. Put the upper triangular part of the initial weight matrix into Randomly change it to 0 or keep it unchanged, and then synchronously copy the upper triangular elements to the lower triangular part to ensure the symmetry of the weighted matrix, and randomly generate 200 populations.

[0102] Step 6.4: Weight threshold filtering. Perform partition detection on the generated weighted adjacency matrix and only retain those that meet , construct a binary adjacency matrix , as shown below:

[0103]

[0104] Step 6.5: Traverse and identify connected components.

[0105] Use the depth-first search algorithm to traverse and identify connected components; mark all nodes as unvisited, and Start by following the valid edge ( ) recursively visit its adjacent nodes; all visited nodes are marked as the same partition and assigned a unique partition label number, and this process is repeated until all nodes are visited; after the traversal is completed, the number of different labels is the number of partitions, and the label number of each node indicates the partition number to which it belongs.

[0106] Step 6.6: Based on the initial partition results of each individual in the population obtained by partition detection, use multi-threaded parallel calculation of the fitness function values ​​of the modularity index, power matching index and partition supply and demand coordination index of each individual in the population.

[0107] Step 6.7: Non-dominated sorting.

[0108] Perform fast non-dominated sorting on the population and divide the Pareto front levels.

[0109] Step 6.8: Calculate target congestion.

[0110] Calculate the crowding degree of each solution in the target space to ensure the diversity of the frontier solutions. The calculation formula is as follows:

[0111]

[0112] Where, express Moment The degree of crowding of individuals; For the objective function values; and Represents all individuals in objective function The maximum and minimum values ​​under .

[0113] Step 6.9: Elite selection. Select the two individuals with the highest non-dominated rank and the highest crowding degree from the population.

[0114] Step 6.10: Crossover operation. The first triangle of the upper triangle of the binary adjacency matrix of the two selected parents is Rank Column Elements Then, the elements of the upper triangle crossover of the parent binary adjacency matrix are synchronously copied to the lower triangle to maintain the symmetry of the parent binary adjacency matrix.

[0115] Step 6.11: Mutation operation. When the upper triangle of the offspring binary adjacency matrix Rank Column Elements When , random mutation is performed or ; When the first Rank Column Elements Only when the first Rank Column Elements When the random variation is or , when the elements in the upper triangle of the original binary adjacency matrix When the elements in the upper triangle of the offspring binary adjacency matrix remain unchanged; then, the mutated elements in the upper triangle of the offspring binary adjacency matrix are synchronously copied to the lower triangle to maintain the symmetry of the offspring binary adjacency matrix.

[0116] Step 6.12: Elite Retention. Retain the best solutions in the non-dominated frontier to avoid losing potential optimal solutions, and merge the parent and child populations.

[0117] Step 7: Calculate the difference of the partition index at different times and dynamically update the binary adjacency matrix. Is the modularity index function value obtained by binarizing the adjacency matrix at time t consistent with The modularity index function value obtained by the binary adjacency matrix at the time is the same. If they are the same, the binary adjacency matrix is ​​not updated; if they are not the same, the Does the modularity index function value obtained by binarizing the adjacency matrix at the moment satisfy ,like And in The power matching index and partition supply and demand coordination index function values ​​obtained by the binary adjacency matrix at time The power matching index difference obtained by the binary adjacency matrix at time The difference between the partition supply and demand coordination index function value When the threshold is exceeded, the binary adjacency matrix is ​​updated. The details are as follows:

[0118]

[0119] Where: and They are Moment and The binary adjacency matrix at time t; Dynamically update the threshold.

[0120] Step 8: Solution set screening: by calculating the target gradient change rate of the solution, select the solution with the most balanced trade-off between the targets.

[0121] Step 9: When The binary adjacency matrix at time When there is an update, The binary adjacency matrix at time Get the corresponding weighted adjacency matrix , and based on the weighted adjacency matrix Update output Visualization of the partitioning results of the microgrid group at each moment.

[0122] Step 10: Determine whether the current moment is the termination moment. If not, perform time series rolling optimization, using the current Pareto solution set as the initial population, and The microgrid is divided according to the operating status at the time, and steps 6 to 9 are repeated; if it is the termination time, the operation is stopped. The dynamic division result of the microgrid group finally obtained in this embodiment is as follows: Figure 4 shown.

[0123] Example 2:

[0124] Based on the method of Example 1, this embodiment provides a microgrid group dynamic partitioning system based on graph theory and non-dominated sorting, including:

[0125] A graph theory model generation module, used for generating a graph theory model of a microgrid group using a weighted undirected graph;

[0126] The indicator definition module is used to define the improved modularity indicator, power matching indicator and partition supply and demand coordination indicator respectively;

[0127] A partitioning module, used to partition the microgrid group;

[0128] Update module, used to dynamically update the weighted adjacency matrix;

[0129] The iterative output module is used to iteratively update the division results and output the final division results of the microgrid group.

Claims

1. A method for dynamic partitioning of microgrid groups based on graph theory and non-dominated sorting, characterized by: The steps include: S1: Based on the corresponding data of the microgrid group, a graph theory model of the microgrid group is generated using a weighted undirected graph; S2: Based on the graph theory model, the improved modularity index, power matching index and partition supply and demand coordination index are defined respectively; S3: Based on the indicators defined in step S2, the second generation non-dominated sorting genetic algorithm is used to divide the microgrid group; S4: Calculate the difference of partitioning indexes at different times and dynamically update the weighted adjacency matrix; S5: Select the solution with the most balanced trade-off between objectives by calculating the target gradient change rate of the solution; S6: Output the visualization division result of the microgrid group according to the updated weighted adjacency matrix; S7: Iteratively update the division results and output the final division results of the microgrid group; The definition of the power matching index in step S2 includes: The power matching index is introduced to quantify the index, which is expressed as follows: Where, and They represent the power matching index values ​​of all partitions and the average value of the power matching index of the nth partition at time t, respectively; N and I represent the number of partitions and the number of microgrids in the nth partition, respectively; and They represent the adjustable active power and net load power in the i-th microgrid at time t respectively; and They represent the adjustable active power of photovoltaic, wind turbine, adjustable load, energy storage and diesel generator in the i-th microgrid at time t respectively; and They represent the real-time active power of photovoltaic, wind turbine and load in the i-th microgrid at time t respectively; The definition of the regional supply and demand coordination index in step S2 includes: The regional supply and demand coordination index is introduced for quantification, which is expressed as follows: Where, and They represent the average value of the partition supply and demand coordination index at time t and the supply and demand coordination index value of the nth partition respectively; It represents the difference between the adjustable capacity of the controllable electrical equipment at time t and the net load power in the nth partition.

2. The method for dynamic partitioning of microgrid groups based on graph theory and non-dominated sorting according to claim 1, characterized in that: The data corresponding to the microgrid group in step S1 includes the microgrid group topology, node number, line resistance and transmission capacity, photovoltaic output of each microgrid, wind turbine output and load forecast power data.

3. The method for dynamic partitioning of microgrid groups based on graph theory and non-dominated sorting according to claim 1, characterized in that: In step S1, a weighted undirected graph G=(V, E, A) is used to generate a graph theory model of a microgrid group. The microgrid group is constructed as a vertex set V=(v1, v2, ..., v i ,v j ,…,v n ), v n represents the nth microgrid, edge set Represents the transmission line connecting adjacent microgrids, and each edge L(v i ,v j )∈E, weighted adjacency matrix A=[a ij ] n×n Used to store the network topology and weight information of the microgrid group.

4. The method for dynamic partitioning of microgrid groups based on graph theory and non-dominated sorting according to claim 3 is characterized in that: The weighted adjacency matrix A in step S1 is [a ij ] n×n middle: In the formula, the edge weight set For storing transmission line L(v i ,v j ) resistance and transport capacity 5. The method for dynamic partitioning of microgrid groups based on graph theory and non-dominated sorting according to claim 4 is characterized in that: The definition of the improved modularity index in step S2 includes: Introducing modularity index ρ t The indicators are quantified and expressed as follows: Where a ij,t is the element in the i-th row and j-th column of the weighted adjacency matrix at time t; It represents 1 / 2 of the sum of the weights of all edges in the partition at time t; represents the sum of the weights of all edges connected to node i at time t; represents the sum of the weights of all edges connected to node j at time t; δ t (i, j) is used to indicate whether nodes i and j are in the same partition and connected at time t, δ t (i, j) = 1, indicating that nodes i and j are in the same partition and connected at time t, δ t (i, j) = 0 means that at time t, nodes i and j are not in the same partition and are not connected.

6. The method for dynamic partitioning of microgrid groups based on graph theory and non-dominated sorting according to claim 5, characterized in that: The process of using the second generation non-dominated sorting genetic algorithm to divide the microgrid group in step S3 includes: A1: Initialization parameters, including population size, maximum number of iterations, crossover probability range, mutation probability range, and dynamic update threshold; A2: According to the weighted adjacency matrix A = [a ij ] n×n Encodes chromosomes; A3: Initialize the population, including: the upper triangular part of the initial weight matrix Randomly change to 0 or remain unchanged, and then synchronously copy the upper triangular elements to the lower triangular part to ensure the symmetry of the weighted matrix, and randomly generate the population; A4: Weight threshold filtering, including: partition detection of the generated weighted adjacency matrix, retaining only those that meet the The edge of the binary adjacency matrix B[b ij ] n×n , as shown below: A5: Traverse and identify connected components, including: using depth-first search algorithm to traverse and identify connected components; marking all nodes as unvisited, and starting from any unvisited node v i First, recursively visit the adjacent nodes along the valid edges; all visited nodes are marked as the same partition and assigned a unique partition label number, and this process is repeated until all nodes are visited; after the traversal is completed, the number of different labels is the number of partitions, and the label number of each node indicates the partition number to which it belongs; A6: Based on the initial partition results of each individual in the population obtained by partition detection, multi-threaded parallel calculation is used to calculate the fitness function values ​​of the modularity index, power matching index, and partition supply and demand coordination index of each individual in the population; A7: Non-dominated sorting, including: fast non-dominated sorting of populations and classification of Pareto frontier levels; A8: Target congestion calculation, including: Calculating the congestion of each solution in the target space: In the formula, CD i,t represents the crowding degree of the i-th individual at time t; f m is the mth objective function value; and Respectively represent all individuals in m objective functions f m The maximum and minimum values ​​under ; A9: Elite selection, including: selecting the two individuals with the highest non-dominant rank and the largest crowding degree from the population; A10: Crossover operation, including: the element b in the i-th row and j-th column of the upper triangle of the binary adjacency matrix of the two selected parents ij Perform crossover, and then synchronously copy the elements of the crossover in the upper triangle of the parent's binary adjacency matrix to the lower triangle to maintain the symmetry of the parent's binary adjacency matrix; A11: Mutation operation, including: when the element b in the upper triangle of the i-th row and j-th column of the offspring binary adjacency matrix ij =1, random mutation to b ij =0 or b ij =1; when the element b in the upper triangle of the offspring binary adjacency matrix is ​​in the i-th row and j-th column ij = 0, only when the element b in the i-th row and j-th column of the upper triangle of the original binary adjacency matrix ij =1, random variation is b ij =0 or b ij =1, when the element b in the upper triangle of the original binary adjacency matrix ij = 0, the elements in the upper triangle of the offspring binary adjacency matrix remain b ij =0 unchanged; then, the mutated elements of the upper triangle of the offspring binary adjacency matrix are synchronously copied to the lower triangle to maintain the symmetry of the offspring binary adjacency matrix; A12: Elite retention, including: retaining high-quality solutions in the non-dominated frontier to avoid losing potential optimal solutions, and merging parent and child populations.

7. The method for dynamic partitioning of microgrid groups based on graph theory and non-dominated sorting according to claim 6, characterized in that: The step S4 specifically includes: Determine whether the modularity index function value obtained by the binary adjacency matrix at time t is the same as the modularity index function value obtained by the binary adjacency matrix at time t-1; if they are the same, do not update the binary adjacency matrix; if they are not the same, determine whether the modularity index function value obtained by the binary adjacency matrix at time t satisfies ρ t ≥0.5, if ρ t ≥0.5 and the difference between the power matching index and the partition supply and demand coordination index function value obtained by the binary adjacency matrix at time t and the power matching index obtained by the binary adjacency matrix at time t-1 The difference between the partition supply and demand coordination index function value When the threshold is exceeded, the binary adjacency matrix is ​​updated; the expression is as follows: Where: B t-1 and B t are the binary adjacency matrices at time t-1 and time t respectively; ε is the dynamic update threshold; At time t, the binary adjacency matrix B t When there is an update, the binary adjacency matrix B at time t is used t Get the corresponding weighted adjacency matrix A t .

8. A microgrid group dynamic partitioning system based on graph theory and non-dominated sorting according to the method of claim 1, characterized in that: include: A graph theory model generation module, used for generating a graph theory model of a microgrid group using a weighted undirected graph; The indicator definition module is used to define the improved modularity indicator, power matching indicator and partition supply and demand coordination indicator respectively; A partitioning module, used to partition the microgrid group; Update module, used to dynamically update the weighted adjacency matrix; The iterative output module is used to iteratively update the division results and output the final division results of the microgrid group.

Images