Dynamic partitioning method and device for distributed power distribution network

Abstract

The invention belongs to the technical field of power distribution network distributed voltage optimization control, and provides a distributed power distribution network dynamic partitioning method and device. The difference value between the modularity, the voltage regulation capability and the membership degree of the current moment of the distributed power distribution network and the modularity, the voltage regulation capability and the membership degree of the previous moment of the current moment is not greater than a preset threshold value; determining the clustering center of the distributed power distribution network according to the degree of membership and a pre-generated clustering center determination model; and dividing nodes in the distributed power distribution network to corresponding clustering centers according to the pre-generated K-means model so as to partition the distributed power distribution network. The invention provides the regional division method which covers more indexes, improves the algorithm capability and dynamically changes according to the characteristics of the power distribution network at different moments.

Description

Technical Field

[0001] The present application belongs to the technical field of distributed voltage optimization control in distribution networks, and in particular to a method and device for dynamic partitioning of distributed distribution networks. Background Art

[0002] At present, with the access of large-scale distributed power sources, the problems of communication delay, large amount of calculation, and too many control devices in the centralized control mode of distributed distribution networks (such as distribution networks in rural areas) have become increasingly prominent. The distributed control method has relatively low communication requirements, good scalability, and strong robustness. It is an effective way to improve the voltage control effect of distributed distribution networks. The basis of distributed control is to reasonably divide the distribution network into small areas for parallel calculation.

[0003] First, in existing research, regional division is mainly based on heuristic algorithms and clustering methods. However, the optimization process based on heuristic algorithms converges slowly, and the global shrinkage ability of clustering algorithms is insufficient.

[0004] Secondly, the existing evaluation indicators such as electrical distance and power regulation capability lack a comprehensive summary of the zoning effect, and the division method cannot adapt to various scenarios under the dynamic changes of distributed distribution networks. Therefore, a regional division method that covers more indicators, improves algorithm capabilities, and dynamically changes according to the characteristics of the distribution network at different times is needed. Summary of the invention

[0005] The distributed distribution network dynamic partitioning method and device provided by the present invention aims to at least solve the following problems in the prior art:

[0006] (1) The technical problems existing in the distributed distribution network partitioning process include slow convergence of the algorithm optimization process and the global shrinkage ability of the clustering algorithm.

[0007] (2) The existing evaluation indicators for distribution network regional division, such as electrical distance and power regulation capability, lack a comprehensive summary of the zoning effect, and the division method cannot adapt to various scenarios under the dynamic changes of distributed distribution networks.

[0008] Based on this, and in order to solve the technical problems in the background technology of this application, the present invention provides the following technical solutions:

[0009] In a first aspect, the present invention provides a method for dynamic partitioning of a distributed power distribution network, the method comprising:

[0010] Perform the following iterative operations until the difference between the modularity, voltage regulation capability, and membership of the distributed distribution network at the current moment and the modularity, voltage regulation capability, and membership at the previous moment of the current moment is not greater than a preset threshold:

[0011] Determine the cluster center of the distributed distribution network according to the modularity, voltage regulation capability, membership degree and pre-generated cluster center determination model of the distributed distribution network; wherein the cluster center determination model is based on an adaptive learning method, and its inertia weight decreases nonlinearly in the process of determining the cluster center;

[0012] According to a pre-generated K-means model, the nodes in the distributed distribution network are divided into corresponding cluster centers to partition the distributed distribution network; wherein, in the K-means model, the fitness function generated by the modularity, the voltage regulation capability, and the membership is the objective function.

[0013] In some embodiments of the present application, the step of generating the cluster center determination model includes:

[0014] Generate an initial model of the cluster center generation model based on a particle swarm algorithm; wherein the nodes of the distributed distribution network correspond to particles in the particle swarm algorithm, and all nodes correspond to populations in the particle swarm algorithm;

[0015] Determining the evolution speed of the population according to the ratio of the current global optimal position to the previous global optimal position of the current global optimal position;

[0016] Determine the population premature convergence factor based on the current population optimal fitness and the current population average fitness;

[0017] Nonlinearly reducing the inertia weight according to the evolution speed, the population premature convergence factor, the current population optimal fitness and the current population average fitness;

[0018] Each particle is made to search in the multidimensional space at a preset speed according to the inertia weight, and the particle position and speed in the search process are updated with the optimal position of each particle and the global optimal position of the population.

[0019] In some embodiments of the present application, the step of determining the modularity includes:

[0020] The modularity is determined according to the weight of the edge connected to the current node, the active voltage and reactive voltage between the current node and the nodes connected to it, and the influence of the nodes connected to it on the current node.

[0021] In some embodiments of the present application, the voltage regulation capability includes: the regulation capability of the adjustable active power of the photovoltaic resources in the distributed distribution network to the maximum voltage deviation, and the regulation capability of the adjustable reactive power of the photovoltaic resources to the maximum voltage deviation;

[0022] The ability of the energy storage resources in the distributed distribution network to adjust the maximum voltage deviation by the adjustable active power, and the ability of the energy storage resources to adjust the maximum voltage deviation by the adjustable reactive power; and

[0023] The adjustable active power of the adjustable load resources in the distributed distribution network has the ability to adjust the maximum voltage deviation, and the adjustable reactive power of the adjustable load resources has the ability to adjust the maximum voltage deviation.

[0024] In some embodiments of the present application, the step of determining the degree of membership includes:

[0025] Determine the sum of the first edges between all nodes in the partition where the current node is located;

[0026] In the partition, determining a first sub-membership between the current node and a node connected to the current node;

[0027] Determine other partitions in the distributed distribution network that do not include the current node;

[0028] Determine the sum of the second edge numbers between all nodes in the other partitions;

[0029] In the other partitions, determining a second sub-membership between the current node and the nodes connected thereto;

[0030] The membership is determined according to the first sum of edge numbers, the second sum of edge numbers, the first sub-membership, and the second sub-membership.

[0031] In a second aspect, the present invention provides a dynamic partitioning device for a distributed power distribution network, the device comprising:

[0032] An iterative operation module is used to perform the following iterative operations until the difference between the modularity, voltage regulation capability, and membership of the distributed distribution network at a current moment and the modularity, voltage regulation capability, and membership at a moment before the current moment is not greater than a preset threshold:

[0033] A cluster center determination module, used to determine the cluster center of the distributed distribution network according to the modularity, voltage regulation capability, membership degree and pre-generated cluster center determination model of the distributed distribution network; wherein the cluster center determination model is based on an adaptive learning method, and its inertia weight decreases nonlinearly in the process of determining the cluster center;

[0034] A power grid partitioning module is used to divide the nodes in the distributed distribution network into corresponding cluster centers according to a pre-generated K-means model to partition the distributed distribution network; wherein, in the K-means model, the fitness function generated by the modularity, the voltage regulation capability, and the membership degree is the objective function.

[0035] In some embodiments of the present application, a distributed power distribution network dynamic partitioning device further includes:

[0036] A cluster center determination module is used to generate the cluster center determination model; the cluster center determination module includes:

[0037] An initial model generation unit, used to generate an initial model of the cluster center generation model based on a particle swarm algorithm; wherein the nodes of the distributed distribution network correspond to particles in the particle swarm algorithm, and all nodes correspond to populations in the particle swarm algorithm;

[0038] An evolution speed determination unit, used to determine the evolution speed of the population according to the ratio of the current global optimal position to the previous global optimal position of the current global optimal position;

[0039] A premature convergence factor determination unit, used to determine a population premature convergence factor according to the current population optimal fitness and the current population average fitness;

[0040] An inertia weight reduction unit, used for nonlinearly reducing the inertia weight according to the evolution speed, the population premature convergence factor, the current population optimal fitness and the current population average fitness;

[0041] The position and speed updating unit is used to make each particle search in the multidimensional space at a preset speed according to the inertia weight, and to update the particle position and speed during the search process with the optimal position of each particle and the global optimal position of the population.

[0042] In some embodiments of the present application, a distributed power distribution network dynamic partitioning device further includes:

[0043] A modularity determination module is used to determine the modularity; the modularity determination module includes:

[0044] The modularity determination unit is used to determine the modularity according to the weight of the edge connected to the current node, the active voltage and reactive voltage between the current node and the nodes connected to it, and the influence degree of the nodes connected to it on the current node.

[0045] In some embodiments of the present application, the voltage regulation capability includes: the regulation capability of the adjustable active power of the photovoltaic resources in the distributed distribution network to the maximum voltage deviation, and the regulation capability of the adjustable reactive power of the photovoltaic resources to the maximum voltage deviation;

[0046] The ability of the energy storage resources in the distributed distribution network to adjust the maximum voltage deviation by the adjustable active power, and the ability of the energy storage resources to adjust the maximum voltage deviation by the adjustable reactive power; and

[0047] The adjustable active power of the adjustable load resources in the distributed distribution network has the ability to adjust the maximum voltage deviation, and the adjustable reactive power of the adjustable load resources has the ability to adjust the maximum voltage deviation.

[0048] In some embodiments of the present application, a distributed power distribution network dynamic partitioning device further includes:

[0049] A membership determination module is used to determine the membership; the membership determination module includes:

[0050] A first edge number sum determination unit, used to determine the first edge number sum between all nodes in the partition where the current node is located;

[0051] a first sub-membership determination unit, configured to determine, within the partition, a first sub-membership between the current node and a node connected thereto;

[0052] Another partition determination unit, used for determining another partition in the distributed distribution network that does not include the current node;

[0053] A second edge number sum determination unit, used to determine the second edge number sum between all nodes in the other partitions;

[0054] a second sub-membership determination unit, configured to determine, in the other partitions, a second sub-membership between the current node and a node connected thereto;

[0055] A membership determination unit is used to determine the membership according to the first sum of edge numbers, the second sum of edge numbers, the first sub-membership, and the second sub-membership.

[0056] In a third aspect, the present invention provides a computer program product, comprising a computer program / instruction, which, when executed by a processor, implements the steps of a method for dynamic partitioning of a distributed distribution network.

[0057] In a fourth aspect, the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, steps of a method for dynamic partitioning of a distributed distribution network are implemented.

[0058] In a fifth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a method for dynamic partitioning of a distributed distribution network.

[0059] From the above description, it can be seen that an embodiment of the present invention provides a method and device for dynamic partitioning of a distributed distribution network, and the corresponding method for dynamic partitioning of a distributed distribution network includes: performing the following iterative operations until the difference between the modularity, voltage regulation capability, and membership of the distributed distribution network at the current moment and the modularity, voltage regulation capability, and membership at the previous moment of the current moment is not greater than a preset threshold: determining the clustering center of the distributed distribution network according to the modularity, voltage regulation capability, and membership of the distributed distribution network and a pre-generated clustering center determination model; wherein the clustering center determination model is based on an adaptive learning method, and its inertia weight decreases nonlinearly in the process of determining the clustering center; dividing the nodes in the distributed distribution network into corresponding clustering centers according to the pre-generated K-means model to partition the distributed distribution network; wherein, in the K-means model, the fitness function generated by the modularity, voltage regulation capability, and membership is the objective function.

[0060] The present invention considers structural indicators, functional indicators and scale indicators to establish comprehensive indicators to provide evaluation indicators for distributed distribution network partitioning; proposes an adaptive learning particle swarm algorithm to solve the problem that the current particle swarm algorithm is prone to fall into local optimality, and improves the optimization accuracy; uses an adaptive particle swarm optimization algorithm to optimize the K-means clustering center, improves the problem that the algorithm depends on the clustering center, and uses the improved K-means algorithm to divide the distribution network area; in order to realize that the partition result changes with the characteristics of the rural distribution network, a comprehensive indicator difference trigger mechanism is set, and the comprehensive indicators at this moment and the previous moment are compared and further updated, thereby realizing dynamic partitioning of the distribution network according to the state on the 24-hour time axis. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0062] Figure 1 A schematic diagram of a distributed power distribution network dynamic partitioning method in an embodiment of the present invention Figure 1 .

[0063] Figure 2 A schematic diagram of a distributed power distribution network dynamic partitioning method in an embodiment of the present invention Figure 2 .

[0064] Figure 3 It is a flowchart diagram of step 400 of a method for dynamic partitioning of a distributed power distribution network in an embodiment of the present invention.

[0065] Figure 4 A schematic diagram of a distributed power distribution network dynamic partitioning method in an embodiment of the present invention Figure 3 .

[0066] Figure 5 It is a flowchart diagram of step 500 of a method for dynamic partitioning of a distributed power distribution network in an embodiment of the present invention.

[0067] Figure 6 A schematic diagram of a distributed power distribution network dynamic partitioning method in an embodiment of the present invention Figure 4 .

[0068] Figure 7 600 is a flow chart of a method for dynamic partitioning of a distributed power distribution network according to an embodiment of the present invention.

[0069] Figure 8 It is a schematic diagram of the distribution network structure in a specific implementation manner of the present invention.

[0070] Fig. 9 It is a flow chart of a method for dynamic partitioning of a distributed power distribution network in a specific implementation manner of the present invention.

[0071] Fig.10 This is a flowchart of improving Kmeans by adaptive learning particle swarm in a specific implementation manner of the present invention.

[0072] Fig.11 It is a flowchart of step S3 in a specific implementation manner of the present invention.

[0073] Fig.12 It is a flowchart of step S4 in a specific implementation manner of the present invention.

[0074] Fig.13 It is a schematic diagram of a distance cost function descent curve in a specific implementation manner of the present invention.

[0075] Fig.14 It is a schematic diagram of distribution network division results under different weight coefficients in a specific implementation manner of the present invention.

[0076] Fig.15 It is a schematic diagram of the distribution network zoning result in a specific implementation manner of the present invention.

[0077] Fig.16 It is a comparison chart of various indicators in the specific implementation mode of the present invention.

[0078] Fig.17 A block diagram of a distributed power distribution network dynamic partitioning device in an embodiment of the present invention Figure 1 .

[0079] Fig.18 A block diagram of a distributed power distribution network dynamic partitioning device in an embodiment of the present invention Figure 2 .

[0080] Fig.19 FIG. 4 is a block diagram of the cluster center determination module 40 in an embodiment of the present invention.

[0081] Fig. 20 A block diagram of a distributed power distribution network dynamic partitioning device in an embodiment of the present invention Figure 3 .

[0082] Fig.21 FIG. 5 is a block diagram of a modularity determination module 50 in an embodiment of the present invention.

[0083] Fig. 22 A block diagram of a distributed power distribution network dynamic partitioning device in an embodiment of the present invention Figure 4 .

[0084] Fig.23 is a block diagram of the membership determination module 60 in an embodiment of the present invention.

[0085] Fig.24 It is a schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0086] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0087] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0088] It should be noted that the terms "including" and "having" in the specification and claims of the present application and the above-mentioned drawings and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. In the absence of conflict, the embodiments in the present application and the features in the embodiments may be combined with each other. The present application will be described in detail below with reference to the drawings and in combination with the embodiments.

[0089] The embodiment of the present invention provides a specific implementation method of a distributed power distribution network dynamic partitioning method, see Figure 1 , the method comprising:

[0090] Step 100: Perform the following iterative operations until the difference between the modularity, voltage regulation capability, and membership of the distributed distribution network at the current moment and the modularity, voltage regulation capability, and membership at the previous moment is not greater than a preset threshold:

[0091] Step 200: Determine the cluster center of the distributed distribution network according to the modularity, voltage regulation capability, membership degree and pre-generated cluster center determination model of the distributed distribution network; wherein the cluster center determination model is based on an adaptive learning method, and its inertia weight decreases nonlinearly in the process of determining the cluster center;

[0092] Step 300: Divide the nodes in the distributed distribution network into corresponding cluster centers according to a pre-generated K-means model to partition the distributed distribution network; wherein, in the K-means model, the fitness function generated by the modularity, the voltage regulation capability, and the membership is the objective function.

[0093] From the above description, it can be seen that an embodiment of the present invention provides a method for dynamic partitioning of a distributed distribution network, including: performing the following iterative operations until the difference between the modularity, voltage regulation capability, and membership of the distributed distribution network at the current moment and the modularity, voltage regulation capability, and membership at the previous moment of the current moment is not greater than a preset threshold: determining the clustering center of the distributed distribution network according to the modularity, voltage regulation capability, and membership of the distributed distribution network and a pre-generated clustering center determination model; wherein the clustering center determination model is based on an adaptive learning method, and its inertia weight decreases nonlinearly in the process of determining the clustering center; according to a pre-generated K-means model, the nodes in the distributed distribution network are divided into corresponding clustering centers to partition the distributed distribution network; wherein, in the K-means model, the fitness function generated by the modularity, voltage regulation capability, and membership is the objective function.

[0094] In summary, the method proposed in the present invention, on the basis of considering structural indicators, functional indicators and scale indicators, establishes comprehensive indicators to provide evaluation indicators for rural distribution network partitioning; proposes an adaptive learning particle swarm algorithm to solve the problem that the current particle swarm algorithm is prone to fall into local optimality, and improves the optimization accuracy; uses an adaptive particle swarm optimization algorithm to optimize the K-means clustering center, improves the problem of algorithm dependence on the clustering center, and uses the improved K-means algorithm to divide the distribution network area; in order to realize that the partitioning result changes with the characteristics of the rural distribution network, a comprehensive indicator difference trigger mechanism is set, and the comprehensive indicators at this moment and the previous moment are compared and further updated, thereby realizing the dynamic partitioning of the distribution network according to the state on the 24-hour time axis, obtaining a reasonable and adaptable method for the rural distribution network, and providing a basis for distributed computing.

[0095] For step 100, in order to realize that the partitioning results change with the characteristics of the distributed distribution network, a comprehensive index difference trigger mechanism is set to compare the comprehensive indicators of the current moment and the previous moment for further update, thereby realizing dynamic partitioning of the distribution network according to the state on the 24-hour time axis, obtaining a reasonable method that adapts to the distributed distribution network, and providing a basis for distributed computing.

[0096] For step 200, it considers structural indicators, functional indicators and scale indicators to establish comprehensive indicators to provide evaluation indicators for distributed distribution network partitions, and proposes an adaptive learning particle swarm algorithm to solve the problem that the current particle swarm algorithm is prone to fall into local optimality, thereby improving the optimization accuracy.

[0097] For step 300, an adaptive particle swarm optimization algorithm is used to optimize the K-means cluster center, improve the problem of the algorithm relying on the cluster center, and use the improved K-means algorithm to divide the power grid area.

[0098] In some embodiments of the present invention, see Figure 2 , a distributed distribution network dynamic partitioning method, further comprising:

[0099] Step 400: Generate the cluster center determination model; then, see Figure 3 , step 400 comprises:

[0100] Step 401: Generate an initial model of the cluster center generation model based on a particle swarm algorithm; wherein the nodes of the distributed distribution network correspond to particles in the particle swarm algorithm, and all nodes correspond to populations in the particle swarm algorithm;

[0101] Step 402: determining the evolution speed of the population according to the ratio of the current global optimal position to the previous global optimal position of the current global optimal position;

[0102] Step 403: determining a population premature convergence factor according to the current population optimal fitness and the current population average fitness;

[0103] Step 404: nonlinearly reducing the inertia weight according to the evolution speed, the population premature convergence factor, the current population optimal fitness and the current population average fitness;

[0104] Step 405: each particle is made to search in the multidimensional space at a preset speed according to the inertia weight, and the particle position and speed during the search process are updated with the optimal position of each particle and the global optimal position of the population.

[0105] For step 401 to step 405, first, the particle swarm algorithm process is as follows:

[0106]

[0107] Where M is the population size, and each particle searches in the d-dimensional space at a certain speed; the i-th particle (x i ) is a d-dimensional vector; the velocity of the i-th example (v i ) is a d-dimensional vector; the individual optimal position of the i-th particle is the individual optimal (pbest); the global optimal position of the population is the global optimal (gbest). After determining the optimal, update the position and speed:

[0108]

[0109] In the formula, v i dt is the velocity of the dth dimension of the i-th particle at the t-th iteration. i dt The position of the ith particle in the dth dimension at the tth iteration. w is the inertia weight, which affects the global and local search capabilities of the algorithm. c1 and c2 are learning factors. r1 and r2 are random numbers between 0 and 1.

[0110] In addition, in order to prevent the population from falling into the local optimum during the optimization process, the fixed parameters of the Particle Swarm Optimization (PSO) performance are improved. First, the population evolution speed, population aggregation degree, and population premature convergence are defined:

[0111] 1) Population evolution speed: the ratio of the current global optimum to the previous global optimum:

[0112]

[0113] In the formula, α tis the evolution speed factor of the population at the tth iteration, between 0 and 1. The smaller the value of α, the faster the population evolves. w should be increased to improve the global search performance of the population, and vice versa.

[0114] 2) Population aggregation degree: the ratio of the current global optimal fitness to the current population average fitness:

[0115]

[0116] In the formula, β t is the population aggregation speed at the tth iteration, between 0 and 1. The larger the β value, the easier it is for the population to fall into the local optimum. In order to make the particles diverge, w should be increased, so that the population can easily jump out of the local optimum, and vice versa.

[0117] 3) Premature population convergence: The population converges prematurely. During the evolution process, the population diversity is gradually lost and shows strong convergence.

[0118]

[0119] In the formula, γ t is the population convergence factor at the tth iteration, f avg t’ is the optimal average. The smaller γ is, the earlier the population converges.

[0120] The weights of the particle swarm algorithm are improved based on nonlinear reduction of inertia weights. Specifically, first, the mapping relationship between w and γ is set:

[0121]

[0122] When the number of iterations is less than the set number of iterations Trec, the adjustment strategy is as follows to make the population maintain a larger w for a longer time.

[0123]

[0124] In order to prevent the population from converging prematurely in the late stage of evolution, w should be adaptively increased, and the adjustment strategy is as follows:

[0125]

[0126] In the formula, w max and w min are the upper and lower limits of the inertia weight respectively. q1 and q2 are empirical constants. max is the maximum number of iterations.

[0127] Then, by analyzing the stability of the particle swarm algorithm, an improvement method for the adaptive learning factor is proposed, which specifically includes the following steps:

[0128] Convert the problem space into 1 dimension, assuming that pbesti and gbest are constants in the process of population movement, counted as p i and p g .

[0129] The velocity update and position update strategies for each particle are as follows:

[0130]

[0131] Fallback time:

[0132]

[0133] get:

[0134]

[0135] Further substituting into:

[0136]

[0137] Therefore, the particle velocity change is a second-order difference equation. Eliminate v t id The particle position is obtained, which is a second-order difference equation:

[0138]

[0139] Assume that the change of particle velocity and position is a continuous process, and s1 and s2 are the solutions of the following equations:

[0140]

[0141] If |s 1,2 |<1, then the solution is stable, let Consider two cases:

[0142] a) And|s 1,2 |<1 article or and or

[0143] b) And|s 1,2 |<1, then and

[0144] The learning factors are as follows:

[0145] Finally, in order to solve the problem of low particle swarm optimization accuracy caused by poor diversity of particle swarm, a smoothing filter operator is used to process the position of the particle swarm, which controls the influence of adjacent dimensions on the new position, reduces the blindness of position selection during the search process, and improves the optimization accuracy of the particle swarm algorithm. For the particle swarm algorithm, the position of particles outside the boundary is usually set to an extreme value, which causes the position of the particle to stay on the boundary for a long time. The position of the particles is readjusted so that the escaped particles are pulled back into the population, thereby ensuring the diversity of the new population:

[0146]

[0147] In some embodiments of the present invention, see Figure 4 , a distributed distribution network dynamic partitioning method, further comprising:

[0148] Step 500: Determine the modularity, then refer to Figure 5 , step 500 comprises:

[0149] Step 501: Determine the modularity according to the weight of the edge connected to the current node, the active voltage and reactive voltage between the current node and the nodes connected to it, and the influence of the nodes connected to it on the current node.

[0150] Specifically, in terms of structure, the electrical connections between nodes within the region are close, and the connections between regions are loose, which is convenient for regional operation and management. The modularity index of the distribution network is expressed as:

[0151]

[0152] In the formula, e ij It is expressed as the weight of the edge between points i and j, which is represented by the electrical distance between the two nodes:

[0153] e ij =1-L ij / max(L)

[0154]

[0155] In the formula, m represents half of the sum of the weights of all edges in the region, and k represents the sum of the weights of all edges connected to node i. When nodes i and j are in the same region, δ(i,j) = 1, otherwise it is 0. S VP and S VQ Represent active, reactive and voltage sensitivity matrices respectively. ij Measures the influence of node j on node i, d ij The larger the value, the smaller the impact, and the larger the distance between the two nodes. ijIt is the degree of association between two nodes when considering the influence of other nodes in the area. The larger the modularity index, the better the distribution network structure.

[0156] In some embodiments of the present invention, the voltage regulation capability includes: the regulation capability of the adjustable active power of the photovoltaic resources in the distributed distribution network to the maximum voltage deviation, and the regulation capability of the adjustable reactive power of the photovoltaic resources to the maximum voltage deviation;

[0157] The ability of the energy storage resources in the distributed distribution network to adjust the maximum voltage deviation by the adjustable active power, and the ability of the energy storage resources to adjust the maximum voltage deviation by the adjustable reactive power; and

[0158] The adjustable active power of the adjustable load resources in the distributed distribution network has the ability to adjust the maximum voltage deviation, and the adjustable reactive power of the adjustable load resources has the ability to adjust the maximum voltage deviation.

[0159] It is understandable that voltage over-limit is the main problem faced by high-proportion distributed photovoltaics after they are connected to the distribution network. The voltage regulation capability index is defined as the ability of the adjustable active and reactive power of the photovoltaic resources, energy storage resources, and adjustable load resources in the region to regulate the maximum voltage deviation, expressed as:

[0160]

[0161] The region should adjust the over-limit voltage by adjusting the active and reactive power of internal adjustable resources as much as possible to reduce the power loss between regions. The voltage regulation capacity of the distribution network is the average value of the regulation capacity of each region, expressed as:

[0162]

[0163] Where: n,t V and ψ i,t V They represent the voltage regulation capability index of each node i in region n, and A represents the total number of nodes in the region. ΔV t i Indicates the voltage deviation of the node i with the largest node voltage deviation in the region, ΔV i,t max It represents the maximum voltage regulation of node i when considering the active and reactive power margin in the area. Both are directional, and N represents the number of divided areas.

[0164] In some embodiments of the present invention, see Figure 6 , a distributed distribution network dynamic partitioning method, further comprising:

[0165] Step 600: Determine the membership degree, then refer to Figure 7, step 600 includes:

[0166] Step 601: Determine the sum of the first edge numbers between all nodes in the partition where the current node is located;

[0167] Step 602: Determine, within the partition, a first sub-membership between the current node and the nodes connected thereto;

[0168] Step 603: Determine other partitions in the distributed distribution network that do not include the current node;

[0169] Step 604: Determine the sum of the second edge numbers between all nodes in the other partitions;

[0170] Step 605: Determine, in the other partitions, a second sub-membership between the current node and the nodes connected thereto;

[0171] Step 606: Determine the membership degree according to the first sum of edge numbers, the second sum of edge numbers, the first sub-membership degree, and the second sub-membership degree.

[0172] In step 601 to step 606, whether the partition size is balanced will affect the difficulty of distributed solution. Improving the balance of partition size can reduce the complexity of the optimization model. The node membership index is proposed to evaluate the rationality of the size of each partition:

[0173]

[0174] Where λ is the node membership index, the larger its value is, the more balanced the sizes of the partitions are; V[i] is the partition where node i is located; |V[i]| is the sum of the number of edges between all nodes in the partition; j∈V[i] represents the node that is in the same partition as node i and is connected to it; μ∈(i,V[i]) is the degree of membership of the nodes that are in the same partition as node i and are connected to it in partition V[i]; VV[i] is the partition that does not contain node i; |VV[i]|| is the sum of the number of edges between nodes in partitions other than partition V[i]; j∈VV[i] represents the node that is not in the same partition as node i but is connected to it; μ(i,VV[i]) is the degree of membership of the nodes that are not in the same partition as node i but are connected to it in partition VV[i].

[0175] Finally, a comprehensive classification index is proposed based on the above indicators, which is expressed as:

[0176]

[0177] In the formula, W1, W2, and W3 are the weight coefficients of each indicator, and the sum is 1. Different weight coefficients may correspond to different division results.

[0178] In order to further illustrate the scheme, the present invention also takes an actual 35kV / 10kV rural distribution network (distributed distribution network) in a certain place as an example to provide a specific implementation method of a distributed distribution network dynamic partitioning method, which includes the following contents:

[0179] The above distribution network is a typical radial network with a total of 33 nodes, including 20 load nodes and 9 distributed photovoltaic nodes. The installed capacity of distributed photovoltaic is 4.308MW. The network structure of the system is as follows: Figure 8 Here, the K-means algorithm improved by the adaptive learning particle swarm optimization algorithm (ALPSO) mentioned above is used to divide the rural distribution network area mentioned above.

[0180] See also Fig. 9 The specific implementation of a distributed distribution network dynamic partitioning method provided by the present invention includes the following steps:

[0181] S1: Construct comprehensive evaluation indicators.

[0182] Specifically, a comprehensive indicator system for distribution network regional division is established, including modularity index, voltage regulation capability index and membership index;

[0183] S2: Determine the cluster center based on the improved particle swarm algorithm-adaptive learning particle swarm algorithm.

[0184] Specifically, the particle swarm algorithm is improved based on the adaptive learning mechanism, its inertia weight is changed, an adaptive weight method based on nonlinear reduction of inertia weight is proposed, and an adaptive learning factor improvement method is proposed. At the same time, a smoothing filter operator is used to process the particle swarm position.

[0185] S3: Use the adaptive learning particle swarm algorithm to improve the K-means algorithm to optimize the clustering center.

[0186] The global optimal position of the improved particle swarm algorithm is the cluster center of the K-means algorithm. Fig.10 as well as Fig.11 , step S3 comprises the following steps:

[0187] Step S31: improving the K-means algorithm based on the improved particle swarm algorithm:

[0188] Dataset D = {o1,o2,...,o n} consists of n sample data, where o iis a p-dimensional vector. The purpose of the K-means clustering algorithm is to find a partition to minimize the sum of distances. After determining the cluster center, the nearest neighbor method is used to divide the data objects into corresponding clusters as follows:

[0189] min j＝1,2,...,k dist(o,e j )

[0190] Cluster center e j That is the solution of the improved particle swarm algorithm.

[0191] Next, perform particle encoding:

[0192] Based on real number coding, particle position, velocity and fitness are expressed as:

[0193] X 11 X 12... X 1p X 21 X 22... X 2p ...X k1 X k2 ...X kp

[0194] V 11 V 12 ...V 1p V 21 V 22 ...V 2p ......V k1 V k2 ...V kp

[0195] fitness(x)

[0196] Where X jp represents the p-th dimension attribute of the j-th cluster center. The first part shows that the particle position is composed of k cluster centers and is a k×p dimensional vector. The middle part represents the particle velocity, which is a k×p dimensional vector. The last part gives the fitness to evaluate the performance of the improved clustering algorithm.

[0197] The fitness function is used as the objective function. The fitness function is a multiple of the sum of distances, where a is a positive constant. The lower the fitness, the better the clustering effect.

[0198] fitness=a×SD

[0199] Step S32: Optimizing K-means clustering centers by improving particle swarm algorithm:

[0200] According to the nearest neighbor principle of K-means clustering algorithm, the cluster center is optimized by self-pair. For the adaptive learning particle swarm algorithm, the inertia weight and learning factor are redesigned, and a particle position update method is proposed. The adaptive learning particle swarm algorithm shows strong randomness and global search capabilities in the next generation population, and is not easy to fall into local minima. The adaptive learning particle swarm algorithm also shows relatively fast convergence. The global optimal position obtained by the adaptive learning particle swarm algorithm is the cluster center of the K-means clustering algorithm.

[0201] First, input the number of samples n, the number of partitions k, and the feature dimension p. Then, initialize the particle swarm. Randomly select k sample data points as the initial position of the particle, repeat this process M times, and generate a total of M particles. Calculate the initial population code. The main parameters include the dimension of the particle D (D = k × p), the maximum number of iterations Tmax, the particle position boundary [Xmin, Xmax], and the random position of all particles in the D-dimensional space.

[0202] Calculate the current population distribution status, including population evolution speed, population aggregation, and population premature convergence.

[0203] Calculate the fitness value of the current particle. According to the fitness function, calculate the current fitness value f(x i ). Set the fitness value of the current particle f(x i ) and individual extreme value f(p i ) and the global extreme value f(p g ) to compare. Update the individual optimum of the particle and the global optimum of the population.

[0204] Update the inertia weight and learning factor of the current population. Compare the premature convergence factor γ, set the threshold τ, and update the inertia weight w of the current population. Based on the analysis of the stability of the particle swarm algorithm, determine the relationship between the learning factors c1 and c2 and the inertia weight w.

[0205] Update the particle's position. Use a smoothing filter operator to update the particle's position. If the position value of a certain dimension of the particle exceeds the population search space, the particle's position is reset.

[0206] Check whether the maximum number of iterations has been reached. If so, output the global optimal position; otherwise, jump to step S323 for the next iteration.

[0207] Calculate the final clustering results. For each sample data, determine the clustering division of the object according to the nearest neighbor principle. Finally, assign all sample data to k cluster centers to complete clustering.

[0208] S4: Calculate the comprehensive evaluation index after partitioning and compare it with the previous moment to determine whether dynamic partitioning is needed.

[0209] By designing a comprehensive index difference trigger mechanism, the comprehensive index at time t and time t-1 is judged based on the source and load and it is decided whether to update, so as to obtain the dynamic area division result as the distribution network status changes;

[0210] Specifically, see Fig.12 , step S4 includes the following steps (where the dynamic division of step S4 is reflected):

[0211] Step S41: Based on the operation status of the distribution network, the comprehensive indicators of the partitioning results at the current moment and the previous moment are compared, dynamic partitioning is triggered by thresholds, and the regional division results are updated.

[0212] Step S42: Initialize the partitioning of distribution network parameters according to the source-load operation data of the distribution network, and use the comprehensive index as the fitness function of the improved K-means algorithm. The problem to be solved is to find k optimal initial clustering centers.

[0213] Step S43: Obtain the optimal partitioning result at time t, and compare the comprehensive index at this time with that at time t-1. If the difference is greater than the set threshold, it is updated, otherwise it remains unchanged to obtain the partitioning result at time 24.

[0214] Fig.13 The following is a comparison chart of each index of the partition under different weight coefficients. The electrical modularity index is positively correlated with the node membership index. When the power balance index is constant, the node membership index increases with the increase of the electrical modularity index; in the actual partitioning process, it is difficult for each index to reach the optimal value at the same time. When the comprehensive partitioning index is selected to reach the maximum value, the corresponding optimal weight coefficients are: w1 = 0.36, w2 = 0.35, w3 = 0.29. This weight combination is used to analyze the partitioning results later.

[0215] The ideal clustering result is that the distance between data points in the partition is as small as possible, and the distance between cluster centers between partitions is as large as possible. The distance cost function is used to determine the optimal number of clusters k. In the partitioning process, the relationship between the distance cost function value and the number of partitions is as follows: Fig.14 As shown, the optimal number of clusters k=5.

[0216] The distribution network area division results are as follows: Fig.15 The proposed comprehensive partitioning index that considers the partition size and power balance index can effectively balance the partition size, avoid the situation where isolated nodes form groups, and facilitate the subsequent regulation of the partitions.

[0217] Comparison of results of different partitioning indicators Fig.16As shown. The internal nodes of the partitions divided by the two partition division indicators have a good degree of coupling tightness. For the power balance index, the partition divided based on the comprehensive partition division index has a higher degree of power balance and better voltage regulation capability. When the voltage exceeds the limit, the reactive power of the photovoltaic inverter is first adjusted to restore the voltage level. When the reactive power regulation margin is exceeded, the active power of the distributed photovoltaic is cut to maintain the stable operation of the distribution network, thereby effectively suppressing the voltage limit problem caused by power imbalance. In general, when the electrical modularity index is not much different, the partitions divided by the comprehensive partition division index have a certain improvement in other indicators except the electrical modularity index.

[0218] Based on the same inventive concept, the embodiments of the present application also provide a distributed distribution network dynamic partitioning device, which can be used to implement the method described in the above embodiments, such as the following embodiments. Since the principle of solving the problem by the distributed distribution network dynamic partitioning device is similar to that of the distributed distribution network dynamic partitioning method, the implementation of the distributed distribution network dynamic partitioning device can refer to the implementation of the distributed distribution network dynamic partitioning method, and the repeated parts will not be repeated. As used below, the term "unit" or "module" can be a combination of software and / or hardware that implements predetermined functions. Although the system described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0219] The embodiment of the present invention provides a specific implementation of a distributed distribution network dynamic partitioning device capable of implementing a distributed distribution network dynamic partitioning method, see Fig.17 , a distributed distribution network dynamic partitioning device specifically includes the following contents:

[0220] The iterative operation module 10 is used to perform the following iterative operations until the difference between the modularity, voltage regulation capability, and membership of the distributed distribution network at the current moment and the modularity, voltage regulation capability, and membership at the previous moment of the current moment is not greater than a preset threshold:

[0221] A cluster center determination module 20 is used to determine the cluster center of the distributed distribution network according to the modularity, voltage regulation capability, membership degree and a pre-generated cluster center determination model of the distributed distribution network; wherein the cluster center determination model is based on an adaptive learning method, and its inertia weight decreases nonlinearly in the process of determining the cluster center;

[0222] The power grid partitioning module 30 is used to divide the nodes in the distributed distribution network into corresponding cluster centers according to a pre-generated K-means model to partition the distributed distribution network; wherein, in the K-means model, the fitness function generated by the modularity, the voltage regulation capability, and the membership degree is the objective function.

[0223] In some embodiments of the present application, see Fig.18 , a distributed power distribution network dynamic partitioning device, further comprising:

[0224] The cluster center determination module 40 is used to generate the cluster center determination model; see Fig.19 , the cluster center determination module 40 includes:

[0225] An initial model generating unit 40a is used to generate an initial model of the cluster center generation model based on a particle swarm algorithm; wherein the nodes of the distributed distribution network correspond to particles in the particle swarm algorithm, and all nodes correspond to populations in the particle swarm algorithm;

[0226] An evolution speed determination unit 40b, used to determine the evolution speed of the population according to the ratio of the current global optimal position to the previous global optimal position of the current global optimal position;

[0227] A premature convergence factor determination unit 40c, used to determine a population premature convergence factor according to the current population optimal fitness and the current population average fitness;

[0228] An inertia weight reduction unit 40d, configured to perform nonlinear reduction on the inertia weight according to the evolution speed, the population premature convergence factor, the current population optimal fitness, and the current population average fitness;

[0229] The position and speed updating unit 40e is used to make each particle search in the multidimensional space at a preset speed according to the inertia weight, and to update the particle position and speed during the search process with the optimal position of each particle and the global optimal position of the population.

[0230] In some embodiments of the present application, see Fig. 20 , a distributed power distribution network dynamic partitioning device, further comprising:

[0231] The modularity determination module 50 is used to determine the modularity; see Fig.21 , the modularity determination module 50 comprises:

[0232] The modularity determination unit 50a is used to determine the modularity according to the weight of the edge connected to the current node, the active voltage and reactive voltage between the current node and the nodes connected to it, and the influence of the nodes connected to it on the current node.

[0233] In some embodiments of the present application, the voltage regulation capability includes: the regulation capability of the adjustable active power of the photovoltaic resources in the distributed distribution network to the maximum voltage deviation, and the regulation capability of the adjustable reactive power of the photovoltaic resources to the maximum voltage deviation;

[0234] The ability of the energy storage resources in the distributed distribution network to adjust the maximum voltage deviation by the adjustable active power, and the ability of the energy storage resources to adjust the maximum voltage deviation by the adjustable reactive power; and

[0235] The adjustable active power of the adjustable load resources in the distributed distribution network has the ability to adjust the maximum voltage deviation, and the adjustable reactive power of the adjustable load resources has the ability to adjust the maximum voltage deviation.

[0236] In some embodiments of the present application, see Fig. 22 , a distributed power distribution network dynamic partitioning device, further comprising:

[0237] The membership determination module 60 is used to determine the membership; see Fig.23 , the membership determination module 60 includes:

[0238] A first edge number sum determination unit 60a, configured to determine the first edge number sum between all nodes in the partition where the current node is located;

[0239] A first sub-membership determination unit 60b, configured to determine, within the partition, a first sub-membership between the current node and a node connected thereto;

[0240] Other partition determination unit 60c, used to determine other partitions in the distributed distribution network that do not include the current node;

[0241] A second edge number sum determination unit 60d, configured to determine the second edge number sum between all nodes in the other partitions;

[0242] A second sub-membership determination unit 60e, configured to determine, in the other partitions, a second sub-membership between the current node and the nodes connected thereto;

[0243] The membership determination unit 60f is configured to determine the membership according to the first sum of edge numbers, the second sum of edge numbers, the first sub-membership, and the second sub-membership.

[0244] From the above description, it can be seen that an embodiment of the present invention provides a distributed distribution network dynamic partitioning device, including: an iterative operation module, which is used to perform the following iterative operations until the difference between the modularity, voltage regulation capability, and membership of the distributed distribution network at the current moment and the modularity, voltage regulation capability, and membership at the previous moment of the current moment is not greater than a preset threshold value: a cluster center determination module, which is used to determine the cluster center of the distributed distribution network according to the modularity, voltage regulation capability, and membership of the distributed distribution network and a pre-generated cluster center determination model; wherein the cluster center determination model is based on an adaptive learning method, and its inertia weight decreases nonlinearly in the process of determining the cluster center; a power grid partitioning module, which is used to divide the nodes in the distributed distribution network into corresponding cluster centers according to a pre-generated K-means model to partition the distributed distribution network; wherein, in the K-means model, the fitness function generated by the modularity, voltage regulation capability, and membership is the objective function.

[0245] The present invention considers structural indicators, functional indicators and scale indicators to establish comprehensive indicators to provide evaluation indicators for distributed distribution network partitioning; proposes an adaptive learning particle swarm algorithm to solve the problem that the current particle swarm algorithm is prone to fall into local optimality, and improves the optimization accuracy; uses an adaptive particle swarm optimization algorithm to optimize the K-means clustering center, improves the problem that the algorithm depends on the clustering center, and uses the improved K-means algorithm to divide the distribution network area; in order to realize that the partition result changes with the characteristics of the rural distribution network, a comprehensive indicator difference trigger mechanism is set, and the comprehensive indicators at this moment and the previous moment are compared and further updated, thereby realizing dynamic partitioning of the distribution network according to the state on the 24-hour time axis.

[0246] The embodiment of the present application also provides a specific implementation of an electronic device capable of implementing all steps in the distributed distribution network dynamic partitioning method in the above embodiment, see Fig.24 , electronic equipment specifically includes the following:

[0247] Processor (processor) 1201, memory (memory) 1202, communication interface (CommunicationsInterface) 1203 and bus 1204;

[0248] The processor 1201, the memory 1202, and the communication interface 1203 communicate with each other through the bus 1204; the communication interface 1203 is used to realize information transmission between the server device and the client device and other related devices;

[0249] The processor 1201 is used to call the computer program in the memory 1202. When the processor executes the computer program, all the steps in the distributed distribution network dynamic partitioning method in the above embodiment are implemented. For example, when the processor executes the computer program, the following steps are implemented:

[0250] Step 100: Perform the following iterative operations until the difference between the modularity, voltage regulation capability, and membership of the distributed distribution network at the current moment and the modularity, voltage regulation capability, and membership at the previous moment is not greater than a preset threshold:

[0251] Step 200: Determine the cluster center of the distributed distribution network according to the modularity, voltage regulation capability, membership degree and pre-generated cluster center determination model of the distributed distribution network; wherein the cluster center determination model is based on an adaptive learning method, and its inertia weight decreases nonlinearly in the process of determining the cluster center;

[0252] Step 300: Divide the nodes in the distributed distribution network into corresponding cluster centers according to a pre-generated K-means model to partition the distributed distribution network; wherein, in the K-means model, the fitness function generated by the modularity, the voltage regulation capability, and the membership is the objective function.

[0253] The embodiments of the present application also provide a computer-readable storage medium capable of implementing all the steps in the method for dynamic partitioning of a distributed distribution network in the above embodiments. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, all the steps in the method for dynamic partitioning of a distributed distribution network in the above embodiments are implemented. For example, when the processor executes the computer program, the following steps are implemented:

[0254] Step 100: Perform the following iterative operations until the difference between the modularity, voltage regulation capability, and membership of the distributed distribution network at the current moment and the modularity, voltage regulation capability, and membership at the previous moment is not greater than a preset threshold:

[0255] Step 200: Determine the cluster center of the distributed distribution network according to the modularity, voltage regulation capability, membership degree and pre-generated cluster center determination model of the distributed distribution network; wherein the cluster center determination model is based on an adaptive learning method, and its inertia weight decreases nonlinearly in the process of determining the cluster center;

[0256] Step 300: Divide the nodes in the distributed distribution network into corresponding cluster centers according to a pre-generated K-means model to partition the distributed distribution network; wherein, in the K-means model, the fitness function generated by the modularity, the voltage regulation capability, and the membership is the objective function.

[0257] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the hardware + program embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0258] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0259] Although the present application provides method operation steps such as embodiments or flow charts, more or fewer operation steps may be included based on conventional or non-creative labor. The order of steps listed in the embodiments is only one way of executing the order of many steps and does not represent the only execution order. When the actual device or client product is executed, it can be executed in the order of the method shown in the embodiments or the drawings or in parallel (for example, in a parallel processor or multi-threaded processing environment).

[0260] For the convenience of description, the above devices are described in various modules according to their functions. Of course, when implementing the embodiments of this specification, the functions of each module can be implemented in the same or more software and / or hardware, or the module implementing the same function can be implemented by a combination of multiple sub-modules or sub-units. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0261] Those skilled in the art also know that, in addition to implementing the controller in a purely computer-readable program code, the controller can be made to implement the same function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, such a controller can be considered as a hardware component, and the devices for implementing various functions included therein can also be considered as structures within the hardware component. Or even, the devices for implementing various functions can be considered as both software modules for implementing the method and structures within the hardware component.

[0262] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0263] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0264] The present specification embodiments may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present specification embodiments may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0265] Each embodiment in this specification is described in a progressive manner, and the same and similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. In the description of this specification, the description of the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the embodiment of this specification. In this specification, the schematic representation of the above terms does not necessarily target the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, in the absence of contradiction, a person skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0266] The above is only an example of the embodiment of the present specification and is not intended to limit the embodiment of the present specification. For those skilled in the art, the embodiment of the present specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiment of the present specification shall be included in the scope of the claims of the embodiment of the present specification.

Claims

1. A distributed distribution network dynamic partitioning method, characterized in that: include: Perform the following iterative operations until the difference between the modularity, voltage regulation capability, and membership of the distributed distribution network at the current moment and the modularity, voltage regulation capability, and membership at the previous moment of the current moment is not greater than a preset threshold: Determine the cluster center of the distributed distribution network according to the modularity, voltage regulation capability, membership degree and pre-generated cluster center determination model of the distributed distribution network; wherein the cluster center determination model is based on an adaptive learning method, and its inertia weight decreases nonlinearly in the process of determining the cluster center; According to a pre-generated K-means model, the nodes in the distributed distribution network are divided into corresponding cluster centers to partition the distributed distribution network; wherein, in the K-means model, the fitness function generated by the modularity, the voltage regulation capability, and the membership is the objective function.

2. The distributed power distribution network dynamic partitioning method according to claim 1, characterized in that: The steps of generating the cluster center determination model include: Generate an initial model of the cluster center generation model based on a particle swarm algorithm; wherein the nodes of the distributed distribution network correspond to particles in the particle swarm algorithm, and all nodes correspond to populations in the particle swarm algorithm; Determining the evolution speed of the population according to the ratio of the current global optimal position to the previous global optimal position of the current global optimal position; Determine the population premature convergence factor based on the current population optimal fitness and the current population average fitness; Nonlinearly reducing the inertia weight according to the evolution speed, the population premature convergence factor, the current population optimal fitness and the current population average fitness; Each particle is made to search in the multidimensional space at a preset speed according to the inertia weight, and the particle position and speed in the search process are updated with the optimal position of each particle and the global optimal position of the population.

3. The distributed power distribution network dynamic partitioning method according to claim 1, characterized in that: The step of determining the modularity comprises: The modularity is determined according to the weight of the edge connected to the current node, the active voltage and reactive voltage between the current node and the nodes connected to it, and the influence of the nodes connected to it on the current node.

4. The distributed power distribution network dynamic partitioning method according to claim 1, characterized in that: The voltage regulation capability includes: the regulation capability of the adjustable active power of the photovoltaic resources in the distributed distribution network to the maximum voltage deviation, and the regulation capability of the adjustable reactive power of the photovoltaic resources to the maximum voltage deviation; The ability of the energy storage resources in the distributed distribution network to adjust the maximum voltage deviation by the adjustable active power, and the ability of the energy storage resources to adjust the maximum voltage deviation by the adjustable reactive power; and The adjustable active power of the adjustable load resources in the distributed distribution network has the ability to adjust the maximum voltage deviation, and the adjustable reactive power of the adjustable load resources has the ability to adjust the maximum voltage deviation.

5. The distributed power distribution network dynamic partitioning method according to claim 1, characterized in that: The step of determining the degree of membership comprises: Determine the sum of the first edges between all nodes in the partition where the current node is located; In the partition, determining a first sub-membership between the current node and a node connected to the current node; Determine other partitions in the distributed distribution network that do not include the current node; Determine the sum of the second edge numbers between all nodes in the other partitions; In the other partitions, determining a second sub-membership between the current node and the nodes connected thereto; The membership is determined according to the first sum of edge numbers, the second sum of edge numbers, the first sub-membership, and the second sub-membership.

6. A distributed distribution network dynamic partitioning device, characterized in that: include: An iterative operation module is used to perform the following iterative operations until the difference between the modularity, voltage regulation capability, and membership of the distributed distribution network at a current moment and the modularity, voltage regulation capability, and membership at a moment before the current moment is not greater than a preset threshold: A cluster center determination module, used to determine the cluster center of the distributed distribution network according to the modularity, voltage regulation capability, membership degree and pre-generated cluster center determination model of the distributed distribution network; wherein the cluster center determination model is based on an adaptive learning method, and its inertia weight decreases nonlinearly in the process of determining the cluster center; A power grid partitioning module is used to divide the nodes in the distributed distribution network into corresponding cluster centers according to a pre-generated K-means model to partition the distributed distribution network; wherein, in the K-means model, the fitness function generated by the modularity, the voltage regulation capability, and the membership degree is the objective function.

7. The distributed power distribution network dynamic partitioning device according to claim 6, characterized in that: Also includes: A cluster center determination module, used to generate the cluster center determination model; The cluster center determination module includes: An initial model generation unit, used to generate an initial model of the cluster center generation model based on a particle swarm algorithm; wherein the nodes of the distributed distribution network correspond to particles in the particle swarm algorithm, and all nodes correspond to populations in the particle swarm algorithm; An evolution speed determination unit, used to determine the evolution speed of the population according to the ratio of the current global optimal position to the previous global optimal position of the current global optimal position; A premature convergence factor determination unit, used to determine a population premature convergence factor according to the current population optimal fitness and the current population average fitness; An inertia weight reduction unit, used for nonlinearly reducing the inertia weight according to the evolution speed, the population premature convergence factor, the current population optimal fitness and the current population average fitness; The position and speed updating unit is used to make each particle search in the multidimensional space at a preset speed according to the inertia weight, and to update the particle position and speed during the search process with the optimal position of each particle and the global optimal position of the population.

8. The distributed power distribution network dynamic partitioning device according to claim 6, characterized in that: Also includes: A modularity determination module, used to determine the modularity; The modularity determination module comprises: The modularity determination unit is used to determine the modularity according to the weight of the edge connected to the current node, the active voltage and reactive voltage between the current node and the nodes connected to it, and the influence degree of the nodes connected to it on the current node.

9. The distributed power distribution network dynamic partitioning device according to claim 6, characterized in that: The voltage regulation capability includes: the regulation capability of the adjustable active power of the photovoltaic resources in the distributed distribution network to the maximum voltage deviation, and the regulation capability of the adjustable reactive power of the photovoltaic resources to the maximum voltage deviation; The ability of the energy storage resources in the distributed distribution network to adjust the maximum voltage deviation by the adjustable active power, and the ability of the energy storage resources to adjust the maximum voltage deviation by the adjustable reactive power; and The adjustable active power of the adjustable load resources in the distributed distribution network has the ability to adjust the maximum voltage deviation, and the adjustable reactive power of the adjustable load resources has the ability to adjust the maximum voltage deviation.

10. The distributed power distribution network dynamic partitioning device according to claim 6, characterized in that: Also includes: A membership determination module, used to determine the membership; The membership degree determination module comprises: A first edge number sum determination unit, used to determine the first edge number sum between all nodes in the partition where the current node is located; a first sub-membership determination unit, configured to determine, within the partition, a first sub-membership between the current node and a node connected thereto; Another partition determination unit, used for determining another partition in the distributed distribution network that does not include the current node; A second edge number sum determination unit, used to determine the second edge number sum between all nodes in the other partitions; a second sub-membership determination unit, configured to determine, in the other partitions, a second sub-membership between the current node and a node connected thereto; A membership determination unit is used to determine the membership according to the first sum of edge numbers, the second sum of edge numbers, the first sub-membership, and the second sub-membership.

11. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the steps of the method for dynamic partitioning of a distributed distribution network as described in any one of claims 1 to 5 are implemented.

12. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the distributed distribution network dynamic partitioning method described in any one of claims 1 to 5 are implemented.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for dynamic partitioning of a distributed distribution network as described in any one of claims 1 to 5 are implemented.

Images