Distributed photovoltaic cluster dynamic division method and device

Abstract

The application provides a distributed photovoltaic cluster dynamic division method and device, and belongs to the field of distributed photovoltaic regulation. The method comprises the following steps: constructing a first division similarity matrix based on a comprehensive index of electrical characteristics, and constructing a Jaccard similarity matrix based on the first division similarity matrix; performing first layer division on distributed photovoltaic nodes based on the Jaccard similarity matrix to obtain at least one cluster; each cluster comprises two distributed photovoltaic nodes, one of which is the node with the maximum Jaccard similarity to the other node, and the number of common neighbor nodes owned by the two nodes is greater than or equal to 2; constructing a cluster network taking the cluster as a network node to obtain a second division similarity matrix; combining the improved module degree gain of the reactive power balance degree, using the Louvain algorithm to perform distributed cluster division on the plurality of distributed photovoltaic nodes, and obtaining the division result of the plurality of distributed photovoltaic nodes.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of distributed photovoltaic regulation, and particularly relates to a distributed photovoltaic cluster dynamic division method and device. BACKGROUND

[0002] The distributed photovoltaic power generation is connected to the distribution network, and the cluster technology is used for scheduling management to divide the distribution network into sub-networks for scheduling control, so that the real-time monitoring and management of the sub-network operation state are realized, the management efficiency is improved, the operation and maintenance cost and the demand for human resources are reduced, and the overall operation of the system is more stable and efficient.

[0003] If the distributed photovoltaic cluster is divided unreasonably, the divided cluster is not conducive to coordinated scheduling, and the stability of the power generation work of the power grid is reduced. SUMMARY

[0004] The present application provides a distributed photovoltaic cluster dynamic division method and device to improve the stability of the power generation work of the power grid.

[0005] The present application provides a distributed photovoltaic cluster dynamic division method, comprising:

[0006] Based on the comprehensive index of electrical characteristics, a first division similarity matrix of a plurality of distributed photovoltaic nodes is constructed, and the comprehensive index of electrical characteristics includes at least one of the following: electrical distance matrix based on reactive power sensitivity, active power regulation capacity, reactive power regulation capacity, load rate and net load;

[0007] Based on the first division similarity matrix, the adjacency relationship of the distributed photovoltaic nodes is obtained;

[0008] Based on the adjacency relationship of the distributed photovoltaic nodes, a Jaccard similarity matrix of the distributed photovoltaic nodes is constructed;

[0009] Based on the Jaccard similarity matrix, the distributed photovoltaic nodes are divided in the first layer to obtain at least one cluster; wherein the distributed photovoltaic nodes included in different clusters are different, each cluster includes two distributed photovoltaic nodes, one of the two distributed photovoltaic nodes is the distributed photovoltaic node with the largest Jaccard similarity of the other distributed photovoltaic node, and the number of common neighbor nodes possessed by the two distributed photovoltaic nodes is greater than or equal to 2;

[0010] A cluster network taking the at least one cluster as at least one network node is constructed, and a second division similarity matrix of the at least one network node is obtained;

[0011] Based on the second partition similarity matrix, combined with the improved module gain of reactive power balance, the Louvain algorithm is used for distributed cluster partition of the plurality of distributed photovoltaic nodes, and a partition result of the plurality of distributed photovoltaic nodes is obtained.

[0012] According to the distributed photovoltaic cluster dynamic partition method provided by the application, the partition similarity matrix of the distributed photovoltaic node is constructed based on the comprehensive index of electrical characteristics, and the method comprises the following steps:

[0013] The active regulation capacity, the reactive regulation capacity, the load rate and the net load of each photovoltaic node are normalized to obtain a regulation capacity normalized index;

[0014] Based on the regulation capacity normalized index, the electrical distance matrix based on the reactive sensitivity is fused to construct the partition similarity matrix;

[0015] The electrical distance matrix based on the reactive sensitivity is represented as: d ij represents the electrical distance between node i and node j, and the element S ij in the reactive sensitivity matrix S represents the change value of the voltage of node i corresponding to the injection of unit reactive power by node j, and the element S jj in the reactive sensitivity matrix S represents the change value of the voltage of node j corresponding to the injection of unit reactive power by node j, and the related expression of the reactive sensitivity matrix S is ΔV=SΔQ, wherein ΔV represents the change amount of the voltage amplitude, and ΔQ represents the reactive change amount.

[0016] The active regulation capacity P i of the photovoltaic node i is represented as: P i ∈[0,P max ], P max represents the active output of the photovoltaic power generation system under the control of the maximum power point tracking system.

[0017] The reactive regulation capacity Q j of the photovoltaic node i is represented as:

[0018]

[0019] S max is the grid-connected capacity of the photovoltaic cluster corresponding to the photovoltaic inverter, P max represents the active output of the photovoltaic power generation system under the control of the maximum power point tracking system, λ max represents the upper limit value of the power factor, λ min represents the lower limit value of the power factor.

[0020] According to the distributed photovoltaic cluster dynamic division method provided by the application, the active regulation capacity, the reactive regulation capacity, the load rate and the net load of each photovoltaic node are normalized to obtain a regulation capacity normalized index, and the method comprises the following steps:

[0021] The active regulation capacity, the reactive regulation capacity, the load rate and the net load of each photovoltaic node are normalized by the first formula to obtain the regulation capacity normalized index of each photovoltaic node.

[0022] The first formula is as follows:

[0023]

[0024] Wherein, x i,m ′ represents the mth regulation capacity normalized index of the photovoltaic node i, x i,m represents the mth index of the photovoltaic node i, and the index comprises the active regulation capacity, the reactive regulation capacity, the load rate and the net load, x 1 / 2,m represents the median value of the mth index, and x 1 / 4,m represents the quarter value of the mth index.

[0025] According to the distributed photovoltaic cluster dynamic division method provided by the application, the active regulation capacity, the reactive regulation capacity, the load rate and the net load of each photovoltaic node are normalized to obtain a regulation capacity normalized index, and the method comprises the following steps:

[0026] The active regulation capacity, the reactive regulation capacity, the load rate and the net load of each photovoltaic node are normalized by the first formula to obtain the regulation capacity normalized index of each photovoltaic node.

[0027] The second formula is as follows:

[0028]

[0029] Wherein, s ij represents the element of the i row and the j column of the similarity matrix, α m represents the weight of the mth index, x i,m ′ represents the mth regulation capacity normalized index of the photovoltaic node i, x j,m ′ represents the mth regulation capacity normalized index of the photovoltaic node j, d ij represents the electrical distance between the node i and the node j based on the reactive sensitivity, d max represents the maximum value in the electrical distance matrix based on the reactive sensitivity, and λ1 and λ2 represent the weights of the comprehensive index and the electrical distance, respectively.

[0030] According to the present invention, a dynamic partitioning method for distributed photovoltaic (PV) clusters is provided, wherein the Jaccard similarity matrix of the distributed PV nodes is constructed based on the adjacency relationships of the distributed PV nodes, including...

[0031] Based on the adjacency relationship of the distributed photovoltaic nodes, the Jaccard similarity matrix of the distributed photovoltaic nodes is constructed using the third formula.

[0032] The third formula is expressed as follows:

[0033]

[0034] Among them, Sim jaccard，ij Sim represents the element in the i-th row and j-th column of the Jaccard similarity matrix. jaccard，ij Y represents the Jaccard similarity between photovoltaic node i and photovoltaic node j. i Y represents the set of neighboring nodes of distributed photovoltaic node i. j Let Ψ(Y) represent the set of neighboring nodes of distributed photovoltaic node j. i Y j ) indicates when Y i With Y j If the intersection is not empty, take the modulus of the intersection; otherwise, take the value 1.

[0035] According to a distributed photovoltaic cluster dynamic partitioning method provided by the present invention, the step of constructing a cluster network with the at least one cluster as at least one network node and obtaining a second partitioning similarity matrix of the at least one network node includes:

[0036] The cluster network is constructed using the at least one cluster as at least one network node;

[0037] The zero matrix is ​​used as the initial similarity matrix of the cluster network;

[0038] Update the diagonal elements of the initial similarity matrix based on the sum of pairwise similarities of nodes within the at least one cluster;

[0039] Based on the fourth formula, the pairwise partition similarity of at least one cluster is calculated, the initial similarity matrix is ​​updated, and the second partition similarity matrix is ​​obtained.

[0040] The fourth formula is expressed as:

[0041]

[0042] in, S represents the partition similarity between cluster t and cluster c in at least one cluster, where distributed photovoltaic node i and distributed photovoltaic node j are nodes in cluster t and cluster c, respectively.i,j represents the similarity degree between the distributed photovoltaic nodes i and j.

[0043] According to the distributed photovoltaic cluster dynamic division method provided by the application, the first division similarity matrix is constructed based on the comprehensive index of electrical characteristics, the electrical characteristics include at least one of the following: electrical distance matrix based on reactive sensitivity, active regulation capacity, reactive regulation capacity, load rate and net load.

[0044] The first process is performed at least once until the division results of the plurality of distributed photovoltaic nodes obtained in the last two first processes are unchanged, and the division result of the plurality of distributed photovoltaic nodes obtained in the last first process is taken as the division result of the plurality of distributed photovoltaic nodes.

[0045] The first process includes:

[0046] The plurality of distributed photovoltaic nodes are traversed until the second process is performed once for each distributed photovoltaic node.

[0047] The second process includes:

[0048] For photovoltaic node i, the improvement of the reactive balance degree of the photovoltaic node i to each photovoltaic node adjacent to the photovoltaic node i is calculated.

[0049] The maximum value of the improvement of the reactive balance degree of the photovoltaic node i to each photovoltaic node adjacent to the photovoltaic node i is selected, and the photovoltaic node j corresponding to the maximum value is selected.

[0050] The improvement of the reactive balance degree is represented as: ΔM new = β1ΔM + β2Δω; ΔM represents the improvement of the module degree gain of the moving node, Δω represents the improvement of the reactive balance degree of the moving node, and β1 and β2 represent the weights of the improvement of the module degree gain and the improvement of the reactive balance degree, respectively.

[0051] The application further provides a distributed photovoltaic cluster dynamic division device, comprising:

[0052] The first construction module is configured to construct a first division similarity matrix of a plurality of distributed photovoltaic nodes based on a comprehensive index of electrical characteristics, wherein the comprehensive index of electrical characteristics includes at least one of the following: an electrical distance matrix based on reactive sensitivity, active regulation capacity, reactive regulation capacity, load rate and net load.

[0053] The adjacent relationship acquisition module is configured to obtain the adjacent relationship of the distributed photovoltaic nodes based on the first division similarity matrix.

[0054] a second construction module, configured to construct a Jaccard similarity matrix of the distributed photovoltaic nodes based on the adjacency relationship of the distributed photovoltaic nodes;

[0055] a first division module, configured to perform first layer division on the distributed photovoltaic nodes based on the Jaccard similarity matrix to obtain at least one cluster; wherein different clusters include different distributed photovoltaic nodes, each cluster includes two distributed photovoltaic nodes, one of the two distributed photovoltaic nodes is the distributed photovoltaic node with the largest Jaccard similarity of the other distributed photovoltaic node, and the two distributed photovoltaic nodes have a number of common neighbor nodes greater than or equal to 2;

[0056] a third construction module, configured to construct a cluster network with the at least one cluster as at least one network node, and obtain a second division similarity matrix of the at least one network node;

[0057] a second division module, configured to perform distributed cluster division on the plurality of distributed photovoltaic nodes based on the second division similarity matrix and combined with improved module gain of reactive power balance degree, and obtain a division result of the plurality of distributed photovoltaic nodes.

[0058] The application further provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the distributed photovoltaic cluster dynamic division method according to any one of the above.

[0059] The application further provides a non-transitory computer readable storage medium, which stores a computer program, wherein the computer program is executable on a processor to implement the steps of the distributed photovoltaic cluster dynamic division method according to any one of the above.

[0060] The application further provides a computer program product, which includes a computer program, wherein the computer program is executable on a processor to implement the steps of the distributed photovoltaic cluster dynamic division method according to any one of the above.

[0061] The application provides a distributed photovoltaic cluster dynamic division method and device, which is divided in layers by using an integrated index system based on electrical characteristics and constructing two similarity matrices based on location characteristics; the timeliness of the Louvain algorithm is improved by improving the Louvain algorithm in layers, wherein the first layer division is based on location characteristics, and the division is allowed to have only two nodes in a cluster and each node is allowed to be divided only once, so that the timeliness is improved and the effectiveness of the division is ensured, and the influence of the first layer division result on subsequent division is reduced; the second layer division is an improved Louvain algorithm division based on electrical characteristics, the improved module gain is combined with the reactive power balance degree, and the reactive power balance degree is introduced for optimization, so that the division effect is improved, the effectiveness and rationality of the distributed photovoltaic cluster dynamic division are ensured, the timeliness of the distributed photovoltaic cluster dynamic division algorithm is improved, and the stability of power generation work of the power grid is improved. BRIEF DESCRIPTION OF DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative effort.

[0063] Figure 1 It is a flowchart of the distributed photovoltaic cluster dynamic division method provided by the application;

[0064] Figure 2 It is a schematic diagram of the first layer division provided by the application;

[0065] Figure 3 It is a flowchart of the cluster network construction method provided by the application;

[0066] Figure 4 It is a flowchart of the improved module gain combined with the reactive power balance degree using the Louvain algorithm for the new network in the application;

[0067] Figure 5 It is a schematic diagram of the data set network structure provided by the application;

[0068] Figure 6 It is a schematic diagram of the Louvain algorithm division network structure provided by the application;

[0069] Figure 7 It is a schematic diagram of the improved layered Louvain algorithm division network structure provided by the application;

[0070] Figure 8It is a structural schematic diagram of the distributed photovoltaic cluster dynamic division device provided by the application.

[0071] Figure 9 It is a structural schematic diagram of the electronic device provided by the application. DETAILED DESCRIPTION

[0072] To make the objectives, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described below in connection with the drawings in the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0073] The distributed photovoltaic cluster dynamic division method and device of the present application will be described below in connection with the drawings.

[0074] Figure 1 It is a flowchart of the distributed photovoltaic cluster dynamic division method provided by the application, as shown in Figure 1 , comprising:

[0075] In step 100, a first division similarity matrix of a plurality of distributed photovoltaic nodes is constructed based on a comprehensive index of electrical characteristics, and the comprehensive index of electrical characteristics comprises at least one of the following: an electrical distance matrix based on reactive sensitivity, active regulation capacity, reactive regulation capacity, load rate and net load.

[0076] Specifically, the comprehensive index system of electrical characteristics in the present application mainly consists of 5 parts, i.e., the electrical distance matrix based on reactive sensitivity, active regulation capacity, reactive regulation capacity, load rate and net load.

[0077] Specifically, the electrical distance is used to characterize the electrical coupling degree between two distributed photovoltaic nodes.

[0078] Specifically, the existence of reactive regulation capacity can adjust the reactive output of the generator to maintain the voltage stability of the power system; the existence of active regulation capacity can adjust the active output of the generator to maintain the power balance and frequency stability of the power system; the regulation capacity can adjust the output power of the generator according to the demand of load change to meet the load demand of the power system, and when the system fault occurs, the regulation capacity can provide additional power support to maintain the stable operation of the power system; the load rate can help to ensure that each distributed photovoltaic node in the cluster is allocated with similar load demand, which helps to ensure the balance, reliability and sustainability of the load distribution between the distributed photovoltaic nodes in the distributed photovoltaic cluster. Through reasonable load balancing, the efficiency and stability of the power system can be maximized.

[0079] Specifically, the net load reflects the actual load demand of the power system, and is crucial for maintaining the stability of the system. Similar net loads mean that the power demand patterns of the two nodes are similar; in a distributed photovoltaic cluster, it can help achieve load balancing, i.e., it is easier to match the photovoltaic power generation capacity with the load demand; it helps to reduce excess or insufficient power supply and improve the stability of the power system.

[0080] Specifically, the electrical distance matrix based on reactive sensitivity mainly focuses on considering the impact of reactive power on the stability of the power system. Photovoltaic power generation systems mainly generate active power, while reactive power is relatively small, especially under standard conditions. Therefore, the electrical distance matrix based on reactive sensitivity can better reflect the reactive voltage stability problem of the photovoltaic cluster. The electrical distance matrix based on reactive sensitivity can be used to evaluate the interaction of reactive power between the photovoltaic cluster and other elements in the power system (such as reactive power compensation devices, transformers, etc.) to ensure voltage stability. This is very important for voltage control and reactive power regulation in the cluster to avoid voltage collapse or voltage peak problems. Although active power also affects the stability of the power system to some extent, the management of reactive power is more urgent in photovoltaic clusters. The electrical distance matrix based on reactive sensitivity provides information about reactive power flow, which helps to accurately assess reactive voltage stability and solve related problems. The first partition similarity matrix constructed by the above five indicators well reflects the electrical characteristics between the distributed photovoltaic nodes, and can improve the rationality of the partition result.

[0081] Step 110, based on the first partition similarity matrix, obtaining the adjacency relationship of the distributed photovoltaic nodes;

[0082] Specifically, after obtaining the first partition similarity matrix constructed by the above five indicators, the adjacency relationship of the distributed photovoltaic nodes can be further determined;

[0083] It should be noted that as long as the method based on the first partition similarity matrix can obtain the adjacency relationship of the distributed photovoltaic nodes, it is applicable to the present application, and is not limited herein.

[0084] Step 120, based on the adjacency relationship of the distributed photovoltaic nodes, constructing a Jaccard similarity matrix of the distributed photovoltaic nodes;

[0085] Specifically, in order to partition the distributed photovoltaic nodes, a Jaccard similarity matrix of the distributed photovoltaic nodes can be constructed based on the adjacency relationship of the distributed photovoltaic nodes;

[0086] It should be noted that as long as the adjacent relationship based on the distributed photovoltaic node can be realized, the manner of constructing the Jaccard similarity matrix of the distributed photovoltaic node is suitable for the present application, and is not limited here.

[0087] Step 130, based on the Jaccard similarity matrix, the first layer of the distributed photovoltaic node is divided to obtain at least one cluster; wherein the different clusters include different distributed photovoltaic nodes, each cluster includes two distributed photovoltaic nodes, one of the two distributed photovoltaic nodes is the Jaccard similarity of the other distributed photovoltaic node is the largest distributed photovoltaic node, and the number of common neighbor nodes of the two distributed photovoltaic nodes is greater than or equal to 2;

[0088] Figure 2 It is the schematic diagram of the first layer division provided by the present application, as shown in Figure 2 The manner of using the Jaccard similarity matrix of the distributed photovoltaic node to divide the distributed photovoltaic node of the distributed photovoltaic cluster in the first layer is as follows:

[0089] (1) Each individual distributed photovoltaic node is a cluster, and all distributed photovoltaic nodes are traversed, when traversing to the distributed photovoltaic node i, if the number of nodes in the cluster where node i is located is not 1, then skip to traverse the next node, otherwise enter step 2);

[0090] (2) Find the Jaccard similarity of the largest distributed photovoltaic node j from the Jaccard similarity matrix of the distributed photovoltaic node i, if the number of common neighbor nodes of the largest Jaccard similarity node j and node i is greater than or equal to 2, then node i is added to the cluster of the largest Jaccard similarity node j, otherwise skip to traverse the next node;

[0091] (3) After the traversal is completed, the first layer division is ended, and the first layer distributed photovoltaic cluster division result is saved.

[0092] In the present application, the first layer division only allows one cluster to have two nodes and the two nodes have more than two common neighbor nodes, and each node only performs one division. In order to prevent excessive division based on geographical characteristics, which will seriously affect the effectiveness of subsequent division, this division repeatedly uses geographical characteristics while greatly reducing the impact on subsequent division, which ensures the effectiveness of the division while effectively improving the timeliness of the subsequent division.

[0093] Step 140, constructing a cluster network with the at least one cluster as at least one network node, and obtaining a second division similarity matrix of the at least one network node;

[0094] Specifically, after the first layer division of the distributed photovoltaic nodes of the distributed photovoltaic cluster, at least one cluster is obtained, and one cluster is taken as one network node to divide the at least one cluster,

[0095] In step 150, based on the second division similarity matrix, the Louvain algorithm is used to divide the plurality of distributed photovoltaic nodes in a distributed cluster manner by combining the improved modularity gain of the reactive power balance degree, and a division result of the plurality of distributed photovoltaic nodes is obtained.

[0096] Optionally, in the present application, the distributed photovoltaic cluster dynamic division algorithm based on the improved modularity gain hierarchical Louvain algorithm comprises the following steps:

[0097] (1) a division similarity matrix of the distributed photovoltaic nodes is constructed by combining a comprehensive index system based on electrical characteristics;

[0098] (2) a Jaccard similarity matrix of the distributed photovoltaic nodes is constructed by combining the adjacency relationship of the distributed photovoltaic nodes, i.e. the location characteristics;

[0099] (3) the first layer division of the distributed photovoltaic cluster is performed by using the Jaccard similarity matrix of the distributed photovoltaic nodes;

[0100] (4) the network reconstruction of the distributed photovoltaic nodes is performed on the first layer division result, i.e. a cluster network is constructed and a second division similarity matrix is obtained, and each node in the cluster network is one cluster after the first layer division;

[0101] (5) the Louvain algorithm is used to divide the distributed photovoltaic cluster by using the second division similarity matrix and the improved modularity gain combined with the reactive power balance degree.

[0102] The commonly used distributed photovoltaic cluster dynamic division algorithm in the related art mainly includes a clustering algorithm, an intelligent optimization algorithm and a complex community discovery algorithm; in the clustering algorithm, the K-means clustering algorithm is the most common, but the K-means clustering needs to determine the number of clusters and the center point, which reduces the effect and timeliness of division; the intelligent optimization algorithm constructs a comprehensive evaluation index system and constructs a corresponding objective function to optimize to obtain the final cluster division result, and compared with the K-means clustering, the intelligent optimization algorithm does not need to determine the number of clusters and the center point, but it is difficult to construct the comprehensive evaluation index system and the optimization problem is easy to fall into local optimum; the complex community discovery algorithm constructs a similarity matrix based on an electrical index system, and then performs community discovery algorithm division based on the optimization modularity on the basis of the similarity matrix, without considering the problems such as constructing an objective function or determining the number of clusters, and the Louvain algorithm is one of the most efficient complex community discovery algorithms. The present application is also designed based on the application idea of the complex community discovery algorithm in the distributed photovoltaic cluster division, which is theoretically feasible.

[0103] For large-scale distributed photovoltaic cluster division, the Louvain algorithm has poor timeliness when performing division, and the present application improves the Louvain algorithm in layers to improve the timeliness of the Louvain algorithm. The first layer division is based on location characteristics and considers that only one cluster is allowed to have two nodes and each node is only divided once, which improves the timeliness and ensures the effectiveness of the division, and reduces the influence of the first layer division result on subsequent division.

[0104] The second layer division is an improved Louvain algorithm division based on electrical characteristics, which improves the modularity gain by combining the reactive power balance degree, and the traditional modularity gain does not calculate the modularity gain brought by node movement when the node moves, which may result in poor subsequent division effect. The present application improves this and introduces the reactive power balance degree for optimization, and the reactive power balance degree is an index for measuring the balance degree of the reactive power in the power system, which indicates whether the distribution of the reactive power in the power system is balanced, so that the division effect is improved.

[0105] The present application uses a comprehensive index system based on electrical characteristics and constructs two similarity matrices based on location characteristics respectively to perform layered division. The present application can also ensure the effectiveness and rationality of the distributed photovoltaic cluster dynamic division and improve the timeliness of the distributed photovoltaic cluster dynamic division algorithm.

[0106] The application aims to realize reasonable division of distributed photovoltaic clusters, improve the efficiency of cluster control scheduling, increase the timeliness of the distributed photovoltaic cluster dynamic division algorithm, and improve the effectiveness of short-time scale control scheduling.

[0107] The distributed photovoltaic cluster dynamic division method provided by the application performs hierarchical division by using an electrical characteristic-based comprehensive index system and a position characteristic-based similarity matrix; the Louvain algorithm is improved in layers to improve the timeliness of the Louvain algorithm, wherein the first layer division is based on position characteristics, and the division is allowed to have only two nodes in a cluster and each node is only divided once, which improves the timeliness and ensures the effectiveness of the division, reduces the influence of the first layer division result on subsequent division; the second layer division is an improved Louvain algorithm based on electrical characteristics, which combines the improved modularity gain of the reactive power balance degree, introduces the reactive power balance degree for optimization, improves the division effect, ensures the effectiveness and rationality of the distributed photovoltaic cluster dynamic division, improves the timeliness of the distributed photovoltaic cluster dynamic division algorithm, and improves the stability of power generation work.

[0108] In some optional embodiments, the electrical characteristic-based comprehensive index is used to construct a division similarity matrix of the distributed photovoltaic nodes, including:

[0109] The active regulation capacity, the reactive regulation capacity, the load rate and the net load of each photovoltaic node are normalized to obtain a regulation capacity normalized index;

[0110] Based on the regulation capacity normalized index, the electrical distance matrix based on the reactive sensitivity is fused to construct the division similarity matrix;

[0111] The electrical distance matrix based on the reactive sensitivity is represented as: d ij The electrical distance between node i and node j, the element S ij The change value of the voltage of node i corresponding to the injection of unit reactive power by node j, the element S jjThe reactive power sensitivity matrix S represents the change in voltage at node j corresponding to the unit reactive power injected at node j. The relevant expression for S is ΔV = SΔQ, where ΔV represents the change in voltage amplitude and ΔQ represents the change in reactive power.

[0112] Active power regulation capacity P of photovoltaic node i i Represented as: P i ∈[0,P max ], P max This indicates the active power output of a photovoltaic power generation system under the control of a maximum power point tracking system.

[0113] Reactive power regulation capacity Q of photovoltaic node i j Represented as:

[0114]

[0115] S max P represents the grid-connected capacity of the photovoltaic inverters corresponding to the photovoltaic cluster. max λ represents the active power output of the photovoltaic power generation system containing the photovoltaic cluster under the control of the maximum power point tracking system. max λ represents the upper limit of the power factor. min This indicates the lower limit of the power factor.

[0116] Specifically, the steps for constructing a similarity matrix for distributed photovoltaic nodes based on a comprehensive index system of electrical characteristics are as follows:

[0117] 1) The comprehensive indicators based on electrical characteristics mainly include five parts, namely, the electrical distance matrix based on reactive power sensitivity, active power regulation capacity, reactive power regulation capacity, load factor and net load.

[0118] 2) The reactive power sensitivity matrix is ​​obtained by the following formula:

[0119] ΔV=SΔQ (1)

[0120] Where ΔV represents the change in voltage amplitude; ΔQ represents the change in reactive power; S is the sensitivity matrix, where S... ij Let represent the change in voltage at node i corresponding to the unit reactive power injected at node j. The expression for the electrical distance matrix based on reactive power sensitivity is as follows:

[0121]

[0122] Where, d ij This represents the electrical distance between node i and node j. The electrical distance is used to characterize the degree of electrical coupling between the two nodes. The closer the distance, the more similar the electrical characteristics of the two nodes are; the farther the distance, the greater the difference in the electrical characteristics of the two nodes.

[0123] 3) Specifically, the presence of active regulation capacity can adjust the active output of the generator to maintain the power balance and frequency stability of the power system. The active regulation capacity is calculated by the following formula:

[0124] P i ∈[0,P max ](3)

[0125] where P i ∈[0,P max ] represents the active regulation capacity of photovoltaic node i, P max represents the active output of the photovoltaic power system under the control of the maximum power point tracking system.

[0126] 4) The presence of reactive regulation capacity can adjust the reactive output of the generator to maintain the power balance and frequency stability of the power system. The reactive regulation capacity is calculated by the following formula:

[0127]

[0128] where Q j represents the reactive regulation capacity of photovoltaic node j, S max is the grid-connected capacity of the photovoltaic inverter, P max represents the active output of the photovoltaic power system under the control of the maximum power point tracking system, λ max and λ min represent the upper and lower limit values of the power factor, respectively.

[0129] The regulation capacity can adjust the output power of the generator according to the demand of load change to meet the load demand of the power system. In the event of system failure, the regulation capacity can provide additional power support to maintain the stable operation of the power system.

[0130] 5) The load rate of the photovoltaic node reflects the generation efficiency of the photovoltaic node within a certain time range, which is used to describe the satisfaction degree of the photovoltaic system to the node load. The formula for calculating the load rate is as follows:

[0131] L r,i =L d,i / L f,i (5)

[0132] where L r,i represents the load rate of photovoltaic node i, L d,i represents the actual power generation of photovoltaic node i, and L f,i represents the maximum power generation of photovoltaic node i.

[0133] 6) The photovoltaic node net load takes into account the contribution of photovoltaic power generation, involves the power balance of the photovoltaic power generation system and its influence on the overall power system, and the formula for calculating the net load is as follows:

[0134] H i = H o,i -H l,i (6)

[0135] wherein H i represents the net load of the photovoltaic node i, H o,i represents the photovoltaic output of the photovoltaic node i, and H l,i represents the load of the photovoltaic node i.

[0136] In some optional embodiments, the active regulation capacity, the reactive regulation capacity, the load rate and the net load of each photovoltaic node are normalized to obtain a regulation capacity normalization index, including:

[0137] The active regulation capacity, the reactive regulation capacity, the load rate and the net load of each photovoltaic node are normalized by the first formula to obtain the regulation capacity normalization index of each photovoltaic node;

[0138] The first formula is represented as:

[0139]

[0140] wherein x i,m ′ represents the mth regulation capacity normalization index of the photovoltaic node i, x i,m represents the mth index of the photovoltaic node i, the index including the active regulation capacity, the reactive regulation capacity, the load rate and the net load, x 1 / 2,m represents the median value of the mth index, and x 1 / 4,m represents the quarter value of the mth index.

[0141] Specifically, when constructing the partition similarity matrix of the distributed photovoltaic node, the active regulation capacity, the reactive regulation capacity, the load rate and the net load can be normalized first:

[0142]

[0143] wherein x i,m ′ represents the normalized mth index of the photovoltaic node i, x i,m represents the mth index of the photovoltaic node i, x 1 / 2,m represents the median value of the mth index, and x 1 / 4,m represents the quarter value of the mth index.

[0144] In some optional embodiments, based on the normalized regulation capacity index, the reactive power sensitivity-based electrical distance matrix is fused to construct the partition similarity matrix, including:

[0145] Based on the normalized regulation capacity index, the reactive power sensitivity-based electrical distance matrix is fused by a second formula to obtain the partition similarity matrix.

[0146] The second formula is represented as:

[0147]

[0148] Wherein, x i,m represents the mth index of photovoltaic node i, which can be active regulation capacity, reactive regulation capacity, load rate and net load, s ij represents the element of the i-th row and j-th column of the similarity matrix, a m represents the weight of the mth index, x i,m ' represents the mth normalized regulation capacity index of photovoltaic node i, x j,m ' represents the mth normalized regulation capacity index of photovoltaic node j, d ij represents the electrical distance between node i and node j based on reactive power sensitivity, d max represents the maximum value in the reactive power sensitivity-based electrical distance matrix, and λ1 and λ2 represent the weights of the comprehensive index and the electrical distance, respectively.

[0149] Specifically, when constructing the partition similarity matrix of the distributed photovoltaic node, after normalizing the active regulation capacity, the reactive regulation capacity, the load rate and the net load, the reactive power sensitivity-based electrical distance matrix can be fused to construct the partition similarity matrix. When the reactive power sensitivity-based electrical distance is fused to construct the partition similarity matrix, the fusion method is as follows:

[0150]

[0151] Wherein, s ij represents the element of the i-th row and j-th column of the similarity matrix, a m represents the weight of the mth index, x i,m ' represents the normalized mth index of photovoltaic node i, x j,m ' represents the normalized mth index of photovoltaic node j, d ij represents the electrical distance between node i and node j based on reactive power sensitivity, d max represents the maximum value in the reactive power sensitivity-based electrical distance matrix, and λ1 and λ2 represent the weights of the comprehensive index and the electrical distance, respectively.

[0152] The normalized active regulation capacity and the reactive regulation capacity are vector dot products to embody the similarity of the active regulation capacity and the reactive regulation capacity indexes of the two nodes corresponding to the element in the i-th row and the j-th column of the similarity matrix when the similarity is calculated; since the greater the electrical distance is, the smaller the similarity between the two nodes is, the electrical distance is taken as negative; the range of the similarity is generally between 0 and 1, so each index of the similarity is divided by the maximum value thereof.

[0153] In some optional embodiments, the Jaccard similarity matrix of the distributed photovoltaic nodes is constructed based on the adjacency relationship of the distributed photovoltaic nodes, including

[0154] The Jaccard similarity matrix of the distributed photovoltaic nodes is constructed based on the adjacency relationship of the distributed photovoltaic nodes by a third formula.

[0155] The third formula is represented as:

[0156]

[0157] wherein, Sim jaccard，ij represents the element in the i-th row and the j-th column of the Jaccard similarity matrix, Sim jaccard，ij represents the Jaccard similarity of the photovoltaic node i and the photovoltaic node j, Y i represents the neighbor node set of the distributed photovoltaic node i, Y j represents the neighbor node set of the distributed photovoltaic node j, Ψ(Y i , Y j ) represents the modulus of the intersection when the intersection of Y i and Y j is not empty, otherwise, the value is 1.

[0158] Specifically, the Jaccard similarity matrix of the distributed photovoltaic nodes is constructed based on the adjacency relationship of the distributed photovoltaic nodes, and the calculation method is as follows:

[0159]

[0160] wherein, Sim jaccard，ij represents the Jaccard similarity of the distributed photovoltaic node i and the distributed photovoltaic node j, Y i represents the neighbor node set of the distributed photovoltaic node i, Y j represents the neighbor node set of the distributed photovoltaic node j, Ψ(Y i , Y j ) represents the modulus of the intersection when the intersection of Y i and Y j is not empty, otherwise, the value is 1.

[0161] Photovoltaic nodes with similar geographical positions have similar electrical characteristics. The present application uses this to make a rough division and then a detailed division to improve the division efficiency and timeliness.

[0162] In some optional embodiments, the construction of the cluster network with the at least one cluster as at least one network node, and obtaining the second division similarity matrix of the at least one network node, comprises:

[0163] The cluster network is constructed with the at least one cluster as at least one network node;

[0164] The zero matrix is taken as the initial similarity matrix of the cluster network;

[0165] Based on the sum of the pairwise similarity of the nodes in the at least one cluster, the diagonal elements of the initial similarity matrix are updated;

[0166] Based on the fourth formula, the pairwise division similarity of the at least one cluster is calculated, the initial similarity matrix is updated, and the second division similarity matrix is obtained;

[0167] The fourth formula is represented as:

[0168]

[0169] Wherein, S (t, c) represents the division similarity between cluster t and cluster c in the at least one cluster, and distributed photovoltaic nodes i and j are nodes in cluster t and cluster c, respectively, S i,j S (i, j) represents the division similarity between distributed photovoltaic node i and node j.

[0170] Figure 3 The flowchart of the cluster network construction method provided by the present application is shown in FIG. 1, and the specific steps of reconstructing the distributed photovoltaic node network based on the first layer division result are as follows: Figure 3

[0171] 1) Obtain all the distributed photovoltaic clusters obtained after the first layer division, construct a new network (i.e. a cluster network) and the division similarity matrix of the network, i.e. a second division similarity matrix, each node of the cluster network is represented as a cluster of the first layer division result (i.e. abstract the divided cluster i as a new node j), and the second division similarity matrix is initialized as a zero matrix.

[0172] 2) Sum of the pairwise similarity of the nodes in the cluster, assign the sum of the pairwise similarity of the nodes in the cluster to the diagonal line of the new division similarity matrix; that is, for each divided cluster i (node j of the cluster network), sum the similarity between the nodes in the cluster i to obtain the self-loop weight edge of the new node j. ​

[0173] 3) Traverse each cluster, and calculate the partition similarity between the cluster and other clusters according to the following formula:

[0174]

[0175] wherein, represents the new partition similarity between cluster t and cluster c, and the distributed photovoltaic nodes i and j are nodes in cluster t and cluster c respectively, S i,j represents the partition similarity between the distributed photovoltaic nodes i and j.

[0176] 4) Correspond the partition similarity between two clusters to the partition similarity between nodes of the new network (i.e. the cluster network).

[0177] In some optional embodiments, the distributed photovoltaic nodes are distributedly clustered based on the second partition similarity matrix and the improved modularity gain of reactive power balance, to obtain a clustering result of the distributed photovoltaic nodes, by using the Louvain algorithm, including:

[0178] The first process is performed at least once until the clustering results of the distributed photovoltaic nodes obtained in the last two first processes are both unchanged, and the clustering result of the distributed photovoltaic nodes obtained in the last first process is taken as the clustering result of the distributed photovoltaic nodes;

[0179] wherein, the first process includes:

[0180] The plurality of distributed photovoltaic nodes are traversed until the second process is performed once for each distributed photovoltaic node;

[0181] The second process includes:

[0182] For the photovoltaic node i, the improved modularity gain of reactive power balance of the photovoltaic node i to each photovoltaic node adjacent to the photovoltaic node i is calculated;

[0183] The photovoltaic node j corresponding to the maximum value in the improved modularity gain of reactive power balance of the photovoltaic node i to each photovoltaic node adjacent to the photovoltaic node i is selected, and the photovoltaic node i is added to the cluster in which the photovoltaic node j is located;

[0184] wherein, the improved modularity gain of reactive power balance is represented as: ΔM new = β1ΔM + β2Δω; ΔM represents the improved modularity gain of the moving node, Δω represents the reactive power balance gain of the moving node, and β1 and β2 represent the weights of the improved modularity gain and the reactive power balance gain, respectively.

[0185] Specifically, the improved modularity gain combined with the reactive power balance degree is composed of the reactive power balance gain and the improved modularity gain; wherein, the improved modularity gain calculation formula is as follows:

[0186]

[0187] Wherein, ΔM represents the improved modularity gain, cluster t and cluster c are the original cluster and the cluster to be moved in of the i distributed photovoltaic node respectively, s t represents the sum of the partition similarity of all distributed photovoltaic nodes in cluster t, s i represents the sum of the partition similarity of i node and other distributed photovoltaic nodes, s i，t represents the sum of the partition similarity of i node to the distributed photovoltaic nodes in cluster t.

[0188] The reactive power balance calculation formula is as follows:

[0189]

[0190]

[0191] Wherein, ω sup,i represents the maximum value of the reactive power output of cluster i, ω need,i represents the reactive power demand value of cluster i, ω i represents the reactive power balance degree of cluster i, c represents the total number of clusters, and ω represents the overall reactive power balance degree of the photovoltaic node network.

[0192] The improved modularity gain combined with the reactive power balance degree calculation formula is as follows:

[0193] ΔM new = β1ΔM + β2Δω (14)

[0194] Wherein, ΔM new represents the improved modularity gain combined with the reactive power balance degree, Δω represents the reactive power balance gain of the moving node, and β1 and β2 represent the weights of the improved modularity gain and the reactive power balance gain respectively.

[0195] The traditional modularity gain of Louvain algorithm does not calculate the modularity gain brought by the node moving when the node moves, which will lead to poor subsequent partition effect and instability. The improved modularity gain increases the stability and efficiency of subsequent partition.

[0196] The reactive power balance degree plays an important role and significance in evaluating system stability, predicting potential problems, optimizing reactive power distribution, guiding power system operation strategy and energy efficiency evaluation, etc. Adding the reactive power balance degree to the optimization index for optimization helps to improve the reliability, stability and operation quality of the power system.

[0197] Figure 4 The flowchart of using the improved modularity gain combined with the reactive balance degree on the new network by the Louvain algorithm in the application; specifically, the new distributed photovoltaic node network is calculated, and the distributed photovoltaic cluster division is performed on the new network by using the Louvain algorithm based on the partition similarity matrix using the improved modularity gain combined with the reactive balance degree, and the specific steps are as follows:

[0198] 1) Traverse to select a distributed photovoltaic node i, and calculate the improved modularity gain combined with the reactive balance degree for all adjacent distributed photovoltaic nodes of the node;

[0199] 2) For the node i, select the maximum improved modularity gain combined with the reactive balance degree node j, and add the node i to the distributed photovoltaic cluster in which the node j is located, that is, complete the second process of the node i, and repeat step 1) until all distributed photovoltaic nodes are selected once, that is, complete one first process;

[0200] 3) If all distributed photovoltaic nodes are traversed once and no node movement occurs, the cluster division algorithm ends, otherwise, the distributed photovoltaic node network is reconstructed and step 1) is entered;

[0201] 4) The improved modularity gain based Louvain algorithm cluster division ends, and the final distributed photovoltaic cluster division result is obtained.

[0202] In one embodiment, Figure 5 The schematic diagram of the data set network structure provided by the application is shown in FIG. 1, in order to verify the effectiveness and timeliness of the improved hierarchical Louvain algorithm and simplify the experimental steps, the application tests the Louvain algorithm and the improved hierarchical Louvain algorithm on the Matlab based on the adjacency matrix distribution of the network structure shown in FIG. 1, wherein the edge weight is all set to 1, and the program outputs the running time, the cluster division list and the modularity. Figure 5

[0203] The Louvain algorithm and the improved hierarchical Louvain algorithm are tested for 10 times, the cluster modularity and the running time of each test result are counted, and the test results are shown in Table 1:

[0204] Table 1 test results

[0205] Partitioning algorithm Average modularity Average running time Louvain algorithm 0.40206 0.92651 Improved hierarchical Louvain algorithm 0.40549 0.45216

[0206] ​The improved hierarchical Louvain algorithm increases 0.243% on average modularity, makes the clusters after division more compact, and reduces the average running time by nearly one time, which proves that the improved hierarchical Louvain algorithm guarantees the efficiency of cluster division and greatly increases the timeliness.

[0207] Figure 6 A schematic diagram of the network structure divided by the Louvain algorithm provided by the application; Figure 7 A schematic diagram of the network structure divided by the improved hierarchical Louvain algorithm provided by the application, and the optimal division results of the Louvain algorithm and the improved hierarchical Louvain algorithm for 10 test division results are analyzed, as shown in Figure 6 、 Figure 7 and Table 2:

[0208] Table 2: Experimental division results

[0209]

[0210] The number of clusters divided by the two algorithms is 4 clusters, and the difference between the division results is that node 10 is in cluster 1 by the Louvain algorithm and is in cluster 4 by the improved hierarchical Louvain algorithm. Due to the movement of node 10, the improved hierarchical Louvain algorithm improves 0.091% on modularity compared with the Louvain algorithm, making the cluster structure more coupled, and the running time is reduced by more than one time. Compared with the average modularity, the optimal result improves less on modularity and increases more on running time, which proves that the improved hierarchical Louvain algorithm has stability in division results.

[0211] Therefore, the algorithm of the application greatly reduces the running time of the algorithm for dynamic division while guaranteeing that the cluster division result has strong internal coupling and weak external coupling, thereby improving the potential of short-time-scale control and scheduling of the distributed photovoltaic cluster. The short-time-scale control and scheduling is realized, the distributed photovoltaic cluster is quickly divided in real time, and the control and scheduling efficiency is improved. The running time of the distributed photovoltaic cluster division algorithm is reduced, the strong internal coupling and weak external coupling of the distributed photovoltaic cluster are improved, and the working effect of the cluster technology is further improved.

[0212] The distributed photovoltaic cluster dynamic division device provided by the application is described below, and the distributed photovoltaic cluster dynamic division device described below can be correspondingly referred to the distributed photovoltaic cluster dynamic division method described above.

[0213] Figure 8is a structural schematic diagram of a distributed photovoltaic cluster dynamic division device provided by the application, as shown in the figure, the device 800 comprises: Figure 8

[0214] A first construction module 810 is configured to construct a first division similarity matrix of a plurality of distributed photovoltaic nodes based on an integrated index of electrical characteristics, wherein the integrated index of electrical characteristics comprises at least one of the following: an electrical distance matrix based on reactive sensitivity, active regulation capacity, reactive regulation capacity, load rate, and net load.

[0215] An adjacency relationship acquisition module 820 is configured to obtain an adjacency relationship of the distributed photovoltaic nodes based on the first division similarity matrix.

[0216] A second construction module 830 is configured to construct a Jaccard similarity matrix of the distributed photovoltaic nodes based on the adjacency relationship of the distributed photovoltaic nodes.

[0217] A first division module 840 is configured to perform first layer division on the distributed photovoltaic nodes based on the Jaccard similarity matrix to obtain at least one cluster, wherein different clusters comprise different distributed photovoltaic nodes, each cluster comprises two distributed photovoltaic nodes, one of the two distributed photovoltaic nodes is the distributed photovoltaic node with the largest Jaccard similarity of the other distributed photovoltaic node, and the two distributed photovoltaic nodes have a number of common neighbor nodes greater than or equal to 2.

[0218] A third construction module 850 is configured to construct a cluster network taking the at least one cluster as at least one network node and obtain a second division similarity matrix of the at least one network node.

[0219] A second division module 860 is configured to perform distributed cluster division on the plurality of distributed photovoltaic nodes using a Louvain algorithm based on the second division similarity matrix and combined with an improved module degree gain of reactive balance, to obtain a division result of the plurality of distributed photovoltaic nodes.

[0220] It should be noted that the distributed photovoltaic cluster dynamic division device of the application can realize each embodiment of the distributed photovoltaic cluster dynamic division method as described above and achieve the same technical effect, which will not be described here.

[0221] ​The application provides a distributed photovoltaic cluster dynamic division device, which is divided in layers by using an electrical characteristic-based comprehensive index system and constructing two similarity matrices based on position characteristics; the timeliness of the Louvain algorithm is improved by improving the Louvain algorithm in layers, wherein the first layer division is based on position characteristics, and the division is allowed to have only two nodes in a cluster and each node is allowed to be divided only once, so that the timeliness is improved and the effectiveness of the division is ensured, and the influence of the first layer division result on subsequent division is reduced; the second layer division is an improved Louvain algorithm division based on electrical characteristics, the improved module gain is combined with the reactive power balance degree, and the reactive power balance degree is introduced for optimization, so that the division effect is improved, the effectiveness and rationality of the distributed photovoltaic cluster dynamic division are ensured, and the timeliness of the distributed photovoltaic cluster dynamic division algorithm is improved.

[0222] Figure 9 An example of a schematic diagram of a physical structure of an electronic device is shown in Figure 9 The electronic device can include a processor 910, a communications interface 920, a memory 930, and a communications bus 940, wherein the processor 910, the communications interface 920, and the memory 930 communicate with each other through the communications bus 940. The processor 910 can invoke a logical instruction in the memory 930 to execute a distributed photovoltaic cluster dynamic division method, including:

[0223] Construct a first division similarity matrix of a plurality of distributed photovoltaic nodes based on an electrical characteristic-based comprehensive index, wherein the electrical characteristic-based comprehensive index includes at least one of the following: an electrical distance matrix based on reactive sensitivity, active regulation capacity, reactive regulation capacity, load rate, and net load;

[0224] Obtain an adjacency relationship of the distributed photovoltaic nodes based on the first division similarity matrix;

[0225] Construct a Jaccard similarity matrix of the distributed photovoltaic nodes based on the adjacency relationship of the distributed photovoltaic nodes;

[0226] Perform first layer division on the distributed photovoltaic nodes based on the Jaccard similarity matrix to obtain at least one cluster; wherein different clusters include different distributed photovoltaic nodes, each cluster includes two distributed photovoltaic nodes, one of the two distributed photovoltaic nodes is the distributed photovoltaic node with the largest Jaccard similarity with the other distributed photovoltaic node, and the two distributed photovoltaic nodes have a number of common neighbor nodes greater than or equal to 2;

[0227] construct a cluster network with the at least one cluster as at least one network node, and obtain a second partition similarity matrix of the at least one network node;

[0228] Based on the second partition similarity matrix, combined with the improved module degree gain of reactive power balance, the Louvain algorithm is used for distributed cluster partition of the plurality of distributed photovoltaic nodes to obtain a partition result of the plurality of distributed photovoltaic nodes.

[0229] In addition, the logical instructions in the memory 930 described above can be implemented in the form of a software function unit and sold or used as an independent product, and can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the parts that contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0230] On the other hand, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program is executed by a processor, so that the computer can execute the distributed photovoltaic cluster dynamic partitioning method provided by the above-mentioned method, comprising:

[0231] Based on the comprehensive index of electrical characteristics, a first partition similarity matrix of the plurality of distributed photovoltaic nodes is constructed, and the comprehensive index of electrical characteristics includes at least one of the following: electrical distance matrix based on reactive sensitivity, active regulation capacity, reactive regulation capacity, load rate and net load;

[0232] Based on the first partition similarity matrix, the adjacency relationship of the distributed photovoltaic nodes is obtained;

[0233] Based on the adjacency relationship of the distributed photovoltaic nodes, a Jaccard similarity matrix of the distributed photovoltaic nodes is constructed;

[0234] performing first-layer partitioning on the distributed photovoltaic nodes based on the Jaccard similarity matrix to obtain at least one cluster; different clusters include different distributed photovoltaic nodes, each cluster includes two distributed photovoltaic nodes, one of the two distributed photovoltaic nodes is the distributed photovoltaic node with the largest Jaccard similarity of the other distributed photovoltaic node, and the two distributed photovoltaic nodes have a number of common neighbor nodes greater than or equal to 2;

[0235] constructing a cluster network with the at least one cluster as at least one network node, and obtaining a second partition similarity matrix of the at least one network node;

[0236] performing distributed cluster partitioning on the plurality of distributed photovoltaic nodes using the Louvain algorithm based on the second partition similarity matrix and in combination with improved module degree gain of reactive power balance degree.

[0237] In another aspect, the present application also provides a non-transitory computer readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the distributed photovoltaic cluster dynamic partitioning method provided by the above-mentioned methods, comprising:

[0238] constructing a first partition similarity matrix of a plurality of distributed photovoltaic nodes based on a comprehensive index of electrical characteristics, the comprehensive index of electrical characteristics including at least one of the following: electrical distance matrix based on reactive sensitivity, active regulation capacity, reactive regulation capacity, load rate, and net load;

[0239] obtaining an adjacency relationship of the distributed photovoltaic nodes based on the first partition similarity matrix;

[0240] constructing a Jaccard similarity matrix of the distributed photovoltaic nodes based on the adjacency relationship of the distributed photovoltaic nodes;

[0241] performing first-layer partitioning on the distributed photovoltaic nodes based on the Jaccard similarity matrix to obtain at least one cluster; different clusters include different distributed photovoltaic nodes, each cluster includes two distributed photovoltaic nodes, one of the two distributed photovoltaic nodes is the distributed photovoltaic node with the largest Jaccard similarity of the other distributed photovoltaic node, and the two distributed photovoltaic nodes have a number of common neighbor nodes greater than or equal to 2;

[0242] constructing a cluster network with the at least one cluster as at least one network node, and obtaining a second partition similarity matrix of the at least one network node;

[0243] Based on the second partition similarity matrix, combined with the improved module degree gain of reactive power balance, the Louvain algorithm is used for distributed cluster partition of the plurality of distributed photovoltaic nodes, and a partition result of the plurality of distributed photovoltaic nodes is obtained.

[0244] The device embodiments described above are only illustrative, wherein the units illustrated as separate components can or can not be physically separated, and the components illustrated as units can or can not be physical units, i.e., can be located in one place or distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0245] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software and the necessary general hardware platform, and of course, can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of software products, which can be stored in a computer readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and include a plurality of instructions for making a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0246] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for distributed photovoltaic cluster dynamic partitioning, characterized in that, The method comprises the following steps: Based on the comprehensive index of electrical characteristics, a first partition similarity matrix of a plurality of distributed photovoltaic nodes is constructed, and the comprehensive index of electrical characteristics comprises at least one of the following: reactive sensitivity-based electrical distance matrix, active regulation capacity, reactive regulation capacity, load rate and net load; Based on the first partition similarity matrix, the adjacency relationship of the distributed photovoltaic nodes is obtained; Based on the adjacency relationship of the distributed photovoltaic nodes, a Jaccard similarity matrix of the distributed photovoltaic nodes is constructed; Based on the Jaccard similarity matrix, the first layer partition of the distributed photovoltaic nodes is performed to obtain at least one cluster; wherein different clusters comprise different distributed photovoltaic nodes, each cluster comprises two distributed photovoltaic nodes, one of the two distributed photovoltaic nodes is the distributed photovoltaic node with the maximum Jaccard similarity of the other distributed photovoltaic node, and the number of common neighbor nodes possessed by the two distributed photovoltaic nodes is greater than or equal to 2; A cluster network taking the at least one cluster as at least one network node is constructed, and a second partition similarity matrix of the at least one network node is obtained; Based on the second partition similarity matrix, combined with the improved module degree gain of reactive balance degree, the Louvain algorithm is used to perform distributed cluster partition on the plurality of distributed photovoltaic nodes to obtain the partition result of the plurality of distributed photovoltaic nodes.

2. The method of claim 1, wherein, The method for constructing the partition similarity matrix of the distributed photovoltaic nodes based on the comprehensive index of electrical characteristics comprises: Normalizing the active regulation capacity, reactive regulation capacity, load rate and net load of each photovoltaic node to obtain a regulation capacity normalized index; Based on the regulation capacity normalized index, the reactive sensitivity-based electrical distance matrix is fused to construct the partition similarity matrix; Wherein, the electrical distance matrix based on the reactive sensitivity is expressed as: d ij represents the electrical distance between node i and node j, the element S ij represents the change value of the voltage of node i corresponding to the injection of unit reactive power by node j, the element S jj represents the change value of the voltage of node j corresponding to the injection of unit reactive power by node j, the relevant expression of the reactive sensitivity matrix S is ΔV=SΔQ, ΔV represents the change amount of the voltage amplitude quantity, and ΔQ represents the reactive change amount. Active regulation capacity P of photovoltaic node i i is expressed as: P i ∈ [0, P max ], P max represents the active output of the photovoltaic power generation system under the control of the maximum power point tracking system; Reactive regulation capacity Q of photovoltaic node i j is represented as: S max P is the grid-connected capacity of the photovoltaic inverter corresponding to the photovoltaic cluster max λ represents the active power output of the photovoltaic power generation system where the photovoltaic cluster is located under the control of the maximum power point tracking system max λ represents the upper limit value of the power factor min λ represents the lower limit value of the power factor 3. The method of claim 2, wherein, The method for normalizing the active regulation capacity, reactive regulation capacity, load rate and net load of each photovoltaic node to obtain a regulation capacity normalized index comprises: The active regulation capacity, reactive regulation capacity, load rate and net load of each photovoltaic node are normalized by a first formula to obtain the regulation capacity normalized index of each photovoltaic node; The first formula is represented as: where x i,m represents the mth regulation capacity normalized indicator of the photovoltaic node i, x i,m represents the mth indicator of the photovoltaic node i, said indicator comprising active regulation capacity, reactive regulation capacity, load rate and net load, x 1 / 2,m represents the median value of the mth indicator, x 1 / 4,m represents the quartile value of the mth indicator.

4. The method of claim 3, wherein, The method for fusing the reactive sensitivity-based electrical distance matrix based on the regulation capacity normalized index to construct the partition similarity matrix comprises: Based on the regulation capacity normalized index, the reactive sensitivity-based electrical distance matrix is fused by a second formula to obtain the partition similarity matrix; The second formula is represented as: wherein s ij represents the element of the similarity matrix in the i-th row and j-th column, α m represents the weight of the m-th index, x i,m ' represents the m-th normalized index of the regulating capacity of the photovoltaic node i, x j,m ' represents the m-th normalized index of the regulating capacity of the photovoltaic node j, d ij represents the electrical distance between the node i and the node j based on the reactive sensitivity, d max represents the maximum value in the electrical distance matrix based on the reactive sensitivity, λ1 and λ2 represent the weights of the comprehensive index and the electrical distance, respectively.

5. The method of claim 1, wherein, The method for constructing the Jaccard similarity matrix of the distributed photovoltaic nodes based on the adjacency relationship of the distributed photovoltaic nodes comprises Based on the adjacency relationship of the distributed photovoltaic nodes, the Jaccard similarity matrix of the distributed photovoltaic nodes is constructed by a third formula; The third formula is represented as: Among them, Sim jaccard，ij Sim represents the element in the i-th row and j-th column of the Jaccard similarity matrix. jaccard，ij Y represents the Jaccard similarity between photovoltaic node i and photovoltaic node j. i Y represents the set of neighboring nodes of distributed photovoltaic node i. j Let Ψ(Y) represent the set of neighboring nodes of distributed photovoltaic node j. i Y j ) indicates when Y i With Y j If the intersection is not empty, take the modulus of the intersection; otherwise, take the value 1.

6. The method of claim 1, wherein, The construction takes the at least one cluster as at least one network node to construct the cluster network, and obtains a second partition similarity matrix of the at least one network node, comprising: constructing the cluster network by taking the at least one cluster as at least one network node; taking a zero matrix as an initial similarity matrix of the cluster network; updating diagonal elements of the initial similarity matrix based on a sum of pairwise similarity of nodes in the at least one cluster; updating the initial similarity matrix based on a fourth formula to obtain a second partition similarity matrix; the fourth formula is expressed as: wherein, represents the partition similarity between cluster t and cluster c in at least one cluster, distributed photovoltaic node i and distributed photovoltaic node j are nodes within cluster t and cluster c respectively, S i,j represents the partition similarity between distributed photovoltaic node i and node j.

7. The method of claim 1, wherein, based on the second partition similarity matrix, combining the improved modularity gain of the reactive power balance degree, using the Louvain algorithm to perform distributed cluster partition on the plurality of distributed photovoltaic nodes to obtain a partition result of the plurality of distributed photovoltaic nodes, comprising: performing at least one first process until the partition results of the plurality of distributed photovoltaic nodes obtained in the last two first processes are all unchanged, and taking the partition result of the plurality of distributed photovoltaic nodes obtained in the last first process as the partition result of the plurality of distributed photovoltaic nodes; wherein, once the first process comprises: traversing the plurality of distributed photovoltaic nodes until performing a second process once for each distributed photovoltaic node; the second process comprises: for photovoltaic node i, calculating the improved modularity gain of the reactive power balance degree of the photovoltaic node i to each photovoltaic node adjacent to the photovoltaic node i; selecting a photovoltaic node j corresponding to the maximum value in the improved modularity gain of the reactive power balance degree of the photovoltaic node i to each photovoltaic node adjacent to the photovoltaic node i, and adding the photovoltaic node i to the cluster in which the photovoltaic node j is located; wherein the improvement module degree gain of the reactive power balance degree is expressed as: ΔM new = β1ΔM + β2Δω; ΔM represents the improvement module degree gain of the mobile node, Δω represents the improvement module degree gain of the mobile node, and β1 and β2 represent the weights of the improvement module degree gain and the reactive power balance degree gain, respectively.

8. A distributed photovoltaic cluster dynamic partitioning apparatus, characterized in that, comprising: a first construction module for constructing a first partition similarity matrix of a plurality of distributed photovoltaic nodes based on a comprehensive index of electrical characteristics, the comprehensive index of electrical characteristics comprising at least one of: an electrical distance matrix based on reactive power sensitivity, active regulation capacity, reactive regulation capacity, load rate and net load; an adjacency relationship acquisition module for obtaining an adjacency relationship of the distributed photovoltaic nodes based on the first partition similarity matrix; a second construction module for constructing a Jaccard similarity matrix of the distributed photovoltaic nodes based on the adjacency relationship of the distributed photovoltaic nodes; a first partition module for performing first layer partition on the distributed photovoltaic nodes based on the Jaccard similarity matrix to obtain at least one cluster; wherein different clusters include different distributed photovoltaic nodes, each cluster includes two distributed photovoltaic nodes, one of the two distributed photovoltaic nodes is the distributed photovoltaic node with the largest Jaccard similarity of the other distributed photovoltaic node, and the two distributed photovoltaic nodes have a number of common neighbor nodes greater than or equal to 2; a third construction module for constructing a cluster network taking the at least one cluster as at least one network node, and obtaining a second partition similarity matrix of the at least one network node; A second division module is configured to perform distributed cluster division on the plurality of distributed photovoltaic nodes based on the second division similarity matrix and the improved module gain of the reactive power balance degree using a Louvain algorithm to obtain a division result of the plurality of distributed photovoltaic nodes.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor implements the steps of the distributed photovoltaic cluster dynamic division method according to any one of claims 1 to 7 when executing the program. 10.A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by the processor, implements the steps of the distributed photovoltaic cluster dynamic division method according to any one of claims 1 to 7.

Images