Transformer area distributed resource aggregation and cluster optimization control method for suppressing voltage fluctuation

Abstract

The invention belongs to the technical field of new energy, and particularly relates to a transformer area distributed resource aggregation and cluster optimization control method for suppressing voltage fluctuation, which comprises the following steps of: accessing various resource devices in a power distribution transformer area into a unified communication network by utilizing the Internet of Things and an ad hoc network technology, and constructing to obtain a distributed resource aggregation network; carrying out cluster division on the distributed resources in the distributed resource aggregation network by adopting an improved Louvain algorithm, generating an optimal cluster division scheme of the resources in the power distribution area, and obtaining an optimal cluster division result; the method comprises the following steps: establishing an optimization control model of a distributed resource cluster, solving the optimization control model to obtain active output and reactive output of each cluster, and carrying out output distribution again in the cluster according to the active output and the reactive output of each cluster based on the capacity of a photovoltaic power station so as to realize intelligent scheduling and optimization control of the distributed resource cluster. Therefore, coordination control is realized in the cluster, voltage fluctuation is suppressed, and electric energy quality is improved.

Description

Technical Field

[0001] The present invention belongs to the field of new energy technology, and in particular relates to a distributed resource aggregation and cluster optimization control method for suppressing voltage fluctuations in an area. Background Art

[0002] In recent years, with the transformation of the global energy structure and the in-depth implementation of the sustainable development strategy, the application of distributed energy resources (DER) in regional power grids has become increasingly widespread. In particular, renewable energy represented by photovoltaic power generation has been widely connected to the distribution network due to its clean and renewable advantages. However, the randomness and volatility of distributed energy have brought unprecedented technical challenges to the power grid.

[0003] The output of distributed power sources such as photovoltaic power generation has significant uncertainty and intermittency, which leads to frequent fluctuations in grid voltage, especially during periods of good lighting conditions and high photovoltaic output, when local areas are prone to overvoltage problems. Traditional voltage control methods, such as those that rely on reactive power compensation equipment in substations and voltage regulation measures on trunk lines, are unable to cope with the new scenario of large-scale access to distributed power sources. These centralized control methods lack real-time performance and flexibility, and cannot effectively reduce voltage fluctuations in distribution areas. Summary of the invention

[0004] In view of this, the present invention provides a method for aggregating distributed resources in a substation and optimizing cluster control for suppressing voltage fluctuations, so as to solve the technical problem that traditional voltage regulation means in the prior art cannot effectively reduce voltage fluctuations in distribution substations when facing a new scenario with a large number of distributed power sources connected.

[0005] To achieve the above objectives, this application adopts the following scheme:

[0006] A distributed resource aggregation and cluster optimization control method for suppressing voltage fluctuations in a substation includes the following steps:

[0007] S10. Using the Internet of Things and ad hoc network technologies, various resource devices in the distribution area are connected to a unified communication network to build a distributed resource aggregation network;

[0008] S20. Using the Louvain algorithm to cluster the distributed resources in the distributed resource aggregation network, generate the best clustering scheme for the distributed resources in the distribution station area, and obtain the best clustering result according to the best clustering scheme;

[0009] S30. Based on the best result of the cluster division, an optimization control model of the distributed resource cluster is established, and the optimization control model is solved to obtain the overall active output of each divided cluster and the overall reactive output of each divided cluster. According to the overall active output of each divided cluster and the overall reactive output of each divided cluster and the capacity of the photovoltaic power station, the output is redistributed within each cluster to realize intelligent scheduling and optimization control of the distributed resource cluster, thereby realizing coordinated control within the cluster, suppressing voltage fluctuations and improving power quality.

[0010] Preferably, in step S10, constructing the distributed resource aggregation network includes the following steps:

[0011] S11. Resource equipment access and information collection: All kinds of distributed resource equipment in the distribution area are connected to the Internet of Things communication network. All kinds of distributed resource equipment are equipped with intelligent collection terminals to collect the corresponding operating status information of various distributed resource equipment in real time, form operating status information data, and transmit the operating status information to the centralized management node;

[0012] S12. Building a self-organizing network communication architecture: The self-organizing network is used to establish communication links between the various types of resource devices to form a self-organizing network of distributed resources;

[0013] S13. Edge computing of operation status information data: Through edge computing technology, real-time operation status information data processing and preliminary analysis are performed at the edge nodes of the distribution station area to realize local processing of operation status information data;

[0014] S14. Information sharing and status synchronization: In a distributed resource aggregation network, the status information data of all resource devices is synchronously updated on the centralized management node, and the updated status information data of the resource devices can be shared;

[0015] S15. Unified management and scheduling preparation: After the collected operating status information data is aggregated, edge computed and shared, the distributed resources within the distribution station area are integrated into a resource pool that can be uniformly managed and scheduled, and the construction of the distributed resource aggregation network is completed.

[0016] Preferably, in step S20, the Louvain algorithm includes:

[0017] S211. Get the modularity function expression of the Louvain algorithm:

[0018]

[0019] Where: A ijis the weight of the edge between node i and node j, indicating the connectivity between nodes. It is 1 when nodes i and j are directly connected, otherwise it is 0; k i is the sum of the weights of all edges connected to node i; k j is the sum of the weights of all edges connected to node j; m = ∑ i ∑ j A ij / 2 is the sum of the weights of all edges in the network; if nodes i and j are in the same cluster, then δ(i,j) is 1, otherwise it is 0;

[0020] S212. The weight of the modularity function is improved, and the improved weight is expressed as:

[0021]

[0022] Where: S OUij and S OUji They represent the reactive voltage sensitivity factors of node i to node j and node j to node i respectively;

[0023] S213. Adjust the reactive power of node i to the reactive support capacity of node j, the weight of the reactive support capacity is expressed as:

[0024]

[0025] Where: Q PV,i , Q VSC,i and Q SVG,i are the adjustable reactive capacities of the PV inverter, VSC and SVG at node i respectively;

[0026] S214. The improved weight matrix is ​​expressed as:

[0027] A OUij =α OUij +β OUij

[0028] S215. Obtain an improved modularity function of the Louvain algorithm, wherein the improved modularity function is:

[0029]

[0030] S216. The cluster coupling index can characterize the degree of coupling between nodes in a cluster. The average value of the coupling within each cluster area is taken as the total intra-area coupling index, which is expressed as:

[0031]

[0032] Where: k is the number of the cluster; N is the total number of clusters; avg is the mean function;

[0033] S217. The comprehensive modularity function index of the network topology, the reactive power support capability of the cluster and the node coupling degree of the cluster can be expressed as:

[0034] ρ=ρ a +ρ b

[0035] Where: a is the improved modularity function; ρ b It is an improved comprehensive modularity function indicator.

[0036] Preferably, in the step S20, clustering the distributed resources in the distributed resource aggregation network using the Louvain algorithm includes the following steps:

[0037] S221. Obtain the operating status information of various resource devices in the distributed resource aggregation network, regard each node in the distributed resource aggregation network as a new cluster, and calculate the improved modularity value ρ;

[0038] S222. randomly select node j and combine it with the node i and calculate to obtain the modularity increment, select the node with the largest modularity increment and the node i to form a new cluster, and update the improved modularity value;

[0039] S223. The current cluster is continued to be combined with other clusters as a new cluster, and the step S222 is repeated to form a new cluster division result until all nodes in the distributed resource aggregation network are traversed;

[0040] S224. When the modularity value in the distributed resource aggregation network reaches the maximum and no node can be merged, cluster division is stopped, and the cluster division result at this time is the optimal cluster division result.

[0041] Preferably, in the step S30, establishing an optimization control model of a distributed resource cluster and solving the optimization control model include the following steps:

[0042] S31. Selecting each cluster leading node based on cluster division, the leading node is a node that can sensitively respond to active changes of other nodes in the cluster and can sensitively respond to reactive changes of other nodes in the cluster;

[0043] S32. Taking the minimum voltage deviation and the minimum system network loss at each of the cluster leading nodes as the control target, and taking the suppression of voltage fluctuation in the distribution station area as the strategy, establish objective function 1 and objective function 2, and perform weighted combination on objective function 1 and objective function 2, and meet the constraints of power flow constraint, cluster power constraint and safe operation constraint. The objective function 1 and objective function 2 are respectively expressed as:

[0044]

[0045] In the formula, j represents the number of dominant nodes; U j represents the voltage amplitude of the dominant node j, U j It is expressed in per-unit value; U0 represents the node voltage reference value, U0 = 1p.u.; M represents the number of clusters involved in the regulation; P loss,m represents the network loss of cluster m;

[0046] The objective function 1 and the objective function 2 are weighted and combined, which is expressed as:

[0047] minf=ω1minf1+ω2minf2

[0048] Where: ω1 and ω2 represent the weight coefficients of objective function 1 and objective function 2 respectively;

[0049] S34. Based on the established optimization control model of the distributed resource cluster, the overall active output of each cluster and the overall reactive output of each cluster are obtained by solving the optimization control model of the distributed resource cluster. According to the overall active output of each cluster and the overall reactive output of each cluster, as well as the capacity of the photovoltaic power station, the output of each cluster is redistributed, as shown in the following formula:

[0050]

[0051] Where: P PV,m , Q PV,m They represent the distributed photovoltaic active output and reactive output of cluster m respectively; P PV,n With Q PV,n They represent the active output and reactive output of the nth PV power station in cluster m respectively; S PV,n represents the capacity of the nth PV power station; N represents the number of PV power stations in cluster m.

[0052] Preferably, in the step S32, the network loss of the cluster m is calculated by the following formula:

[0053]

[0054] Where: N represents the total number of nodes in the cluster; l represents the set of nodes connected to node n; inl represents the current in the branch connecting node n and node l; r nl Indicates the resistance of this branch.

[0055] Preferably, in the step S32, the power flow constraint can be expressed as:

[0056]

[0057] Where: u i with u j Represent the node voltages of node i and node j respectively; i ij represents the current of branch ij starting from node i and ending at node j; P ij With Q ij Respectively represent the active and reactive power flowing through branch ij; r ij With x ij Respectively represent the resistance and reactance of branch ij; P j With Q j They represent the net active and reactive loads injected into node j respectively; jl represents the set of branches connected to node j with node j as the starting point and node l as the end point, P jl With Q jl They represent the active power and reactive power flowing through branch jl respectively.

[0058] Preferably, in the step S32, the cluster power constraint is expressed as:

[0059]

[0060]

[0061] Where: P PV,m , Q PV,m They represent the distributed photovoltaic active output and reactive output of cluster m respectively; represents the maximum value of distributed photovoltaic active output of cluster m; and They respectively represent the upper and lower limits of the distributed photovoltaic reactive output of cluster m.

[0062] Preferably, in the step S32, the operation safety constraint is expressed as:

[0063]

[0064] Where: and They represent the upper and lower limits of the dominant node voltage respectively, and the node voltage operating range is taken as 0.95pu~1.07pu.

[0065] The technical solution adopted in this application can achieve the following beneficial effects:

[0066] In a distributed resource aggregation and cluster optimization control method for suppressing voltage fluctuations provided by the present invention, various resource devices in the distribution station area are connected to a unified communication network by using the Internet of Things and self-organizing network technology to construct a distributed resource aggregation network. The distributed resource aggregation network can monitor the operating status of various resource devices in real time, and synchronize the status and share information through edge computing and fast communication, thereby realizing intelligent management of distributed resources. This aggregation technology provides technical support for the coordinated scheduling of various resource devices in the distribution station area, and effectively improves the utilization rate of distributed resources. The Louvain algorithm is used to cluster the distributed resources in the distributed resource aggregation network, and the optimal cluster division scheme for the resources in the distribution station area is generated. The optimal cluster division result is obtained according to the optimal cluster division scheme. Through cluster division, the distributed resources in the distribution station area are successfully clustered according to their Geographical location, load characteristics and grid demand are reasonably divided into multiple functional clusters, and the resources within each cluster are highly similar and strongly correlated, and the coupling degree between different clusters is low, which ensures the coordination control efficiency and response speed within the cluster. The divided clusters can accurately reflect the voltage demand and power fluctuation of the substation, and further improve the flexibility and accuracy of resource scheduling in the optimization control; an optimization control model of distributed resource clusters is established, and the optimization control model is solved to obtain the overall active output of each divided cluster and the overall reactive output of each divided cluster. According to the overall active output of each divided cluster and the overall reactive output of each divided cluster and the capacity of the photovoltaic power station, the output is redistributed within each cluster, so that the maximum deviation value of the node voltage is reduced from 0.1049pu to 0.0579pu, which can effectively suppress the fluctuation of the node voltage and improve the power quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 It is the overall flow chart of the present invention.

[0068] Figure 2 This is a flow chart of cluster division in the present invention.

[0069] Figure 3 The present invention establishes an optimization control model for a distributed resource cluster and shows a flow chart for solving the model.

[0070] Figure 4 It is a schematic diagram of the cluster division results in the present invention. DETAILED DESCRIPTION

[0071] In order to facilitate the understanding of the present application, the present application will be described more comprehensively below in conjunction with the accompanying drawings. And the preferred implementation of the present application is given. However, the present application can be implemented in many different forms and is not limited to the implementation described herein. On the contrary, the purpose of providing these implementations is to make the disclosure of the present application more thoroughly and comprehensively understood.

[0072] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by those skilled in the art to which this application belongs. The terms used herein in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0073] Please see Figure 1 This embodiment provides a method for controlling distributed resource aggregation and cluster optimization in a substation area for suppressing voltage fluctuations, including the following steps:

[0074] S10. Using the Internet of Things and ad hoc network technologies, various resource devices in the distribution area are connected to a unified communication network to build a distributed resource aggregation network;

[0075] In the S10 step, this embodiment utilizes the Internet of Things (IoT) and self-organizing network technology to achieve effective aggregation of various distributed resources in the substation area. Specifically, by connecting various resource devices such as photovoltaic systems, energy storage equipment, and reactive compensation devices in the distribution substation area to a unified communication network, a distributed resource aggregation network is constructed, and the capacity of the photovoltaic power station in the photovoltaic system is obtained. The distributed resource aggregation network can monitor the operating status of various resource devices in real time, and synchronize the status and share information through edge computing and fast communication, thereby realizing intelligent management of distributed resources. This aggregation technology provides technical support for the coordinated scheduling of various resource devices in the distribution substation area, effectively improving the utilization rate of distributed resources.

[0076] S20. Using the Louvain algorithm to cluster the distributed resources in the distributed resource aggregation network, generate the best clustering scheme for the distributed resources in the distribution station area, and obtain the best clustering result according to the best clustering scheme;

[0077] On the basis of the distributed resource aggregation network constructed by the above-mentioned step S10, the step S20 in this embodiment further uses the Louvain algorithm to cluster the distributed resources in the substation area. The Louvain algorithm is a community discovery algorithm based on modularity optimization, which is suitable for the partitioning task of large-scale networks. The Louvain algorithm described in this embodiment is an improved Louvain algorithm. The algorithm is used to cluster the distributed resources in the distributed resource aggregation network, and the optimal clustering scheme for distributed resources in the distribution substation area can be generated more quickly and accurately. The obtained clustering results are helpful to identify and group resources in the same area, which is convenient for the subsequent (step S30) collaborative control. In this way, under different grid loads and photovoltaic output conditions, the control strategy can be quickly adapted and adjusted to improve the overall stability of the power grid.

[0078] S30. Based on the optimal result of the cluster division, an optimization control model of the distributed resource cluster is established, and the optimization control model is solved to obtain the overall active output of each divided cluster and the overall reactive output of each divided cluster. According to the overall active output of each divided cluster and the overall reactive output of each divided cluster and the capacity of the photovoltaic power station, the output is redistributed within each cluster to realize intelligent scheduling and optimization control of the distributed resource cluster, thereby realizing coordinated control within the cluster, suppressing voltage fluctuations and improving power quality;

[0079] After completing the resource cluster division, the next step (step S30) is to establish an optimization control model for the distributed resource cluster, which aims to reduce the voltage fluctuation in the distribution station area and improve the voltage quality, and combines the operating characteristics of the resources inside the cluster and the voltage demand of the distributed resource aggregation network to perform multi-objective optimization. By solving the optimization control model of the distributed resource cluster, the active output and reactive output of each cluster as a whole are obtained. According to the active output of each cluster as a whole and the reactive output of each cluster as a whole and the capacity of the photovoltaic power station, the output is redistributed within each cluster, which can realize intelligent scheduling and optimization control of the distributed resource cluster, thereby realizing coordinated control within the cluster, suppressing voltage fluctuations and improving power quality, and the distributed resource aggregation and cluster optimization control method for suppressing voltage fluctuations provided in this embodiment can realize fine regulation of distributed resources at the distribution station area level, meet voltage quality requirements, and improve the operation efficiency of the power grid.

[0080] The technical solutions and technical effects of the present invention are further illustrated below through specific examples. It should be noted that the following examples are only for further explanation of the present invention and do not limit the technical solutions of the present invention.

[0081] Example 1

[0082] Constructing the distributed resource aggregation network specifically includes the following steps:

[0083] S11. Resource equipment access and information collection: connect various types of distributed resource equipment in the distribution area to the Internet of Things communication network, such as photovoltaic systems, energy storage equipment, reactive power compensation devices, etc. All of the distributed resource equipment are equipped with intelligent collection terminals to collect the corresponding operating status information of the various distributed resource equipment in real time, form operating status information data, and transmit the operating status information to the centralized management node. The operating status information includes voltage, current, power, etc. The centralized management node can be a cloud platform or other server;

[0084] S12. Construction of self-organizing network communication architecture: The self-organizing network is used to establish communication links between the various types of resource devices to form a self-organizing network of distributed resources. The advantage of the self-organizing network is that it can flexibly adapt to the distribution of equipment in the distribution area, support dynamic topology adjustment, and ensure the stability and reliability of the network;

[0085] S13. Edge computing of operation status information data: Through edge computing technology, real-time operation status information data processing and preliminary analysis are performed at the edge nodes of the distribution station area to realize local processing of operation status information data. The edge nodes are, for example, data gateways or controllers. Edge computing can realize local processing of data, such as anomaly detection, equipment status monitoring, etc., reducing the delay and bandwidth pressure of uploading data to the cloud;

[0086] S14. Information sharing and status synchronization: In a distributed resource aggregation network, the status information data of all resource devices are synchronously updated on the centralized management node, and the updated status information data of the resource devices can be shared. The information data sharing of the resource device status enables the operating status of various devices in the distribution station area to be monitored in a timely manner, providing data support for subsequent collaborative control;

[0087] S15. Unified management and scheduling preparation: After the collected operating status information data is aggregated, edge computed and shared, the distributed resources within the distribution station area are integrated into a resource pool that can be uniformly managed and scheduled, and the construction of the distributed resource aggregation network is completed, laying the foundation for subsequent cluster division and optimized control.

[0088] The Louvain algorithm includes:

[0089] S211. Get the modularity function expression of the Louvain algorithm:

[0090]

[0091] Where: A ijis the weight of the edge between node i and node j, indicating the connectivity between nodes. It is 1 when nodes i and j are directly connected, otherwise it is 0; k i is the sum of the weights of all edges connected to node i; k j is the sum of the weights of all edges connected to node j; m = ∑ i ∑ j A ij / 2 is the sum of the weights of all edges in the network; if nodes i and j are in the same cluster, then δ(i,j) is 1, otherwise it is 0;

[0092] S212. The weight of the modularity function is improved, and the improved weight is expressed as:

[0093]

[0094] Where: S OUij and S OUji They represent the reactive voltage sensitivity factors of node i to node j and node j to node i respectively;

[0095] S213. Reactive power support capacity will affect the reactive power balance within the cluster. It mainly considers the equipment that can quickly adjust reactive power. DPV can adjust the inverter to change its output reactive power. SVG, VSC and other equipment can also quickly adjust reactive power output to provide reactive support for the system. The reactive power of node i is adjusted to support the reactive power of node j. The weight of the reactive power support capacity is expressed as:

[0096]

[0097] Where: Q PV,i , Q VSC,i and Q SVG,i are the adjustable reactive capacities of the PV inverter, VSC and SVG at node i respectively;

[0098] S214. The improved weight matrix is ​​expressed as:

[0099] A OUij =α OUij +β OUij

[0100] S215. Obtain an improved modularity function of the Louvain algorithm, wherein the improved modularity function is:

[0101]

[0102] S216. The cluster coupling index can characterize the degree of coupling between nodes in a cluster. The average value of the coupling within each cluster area is taken as the total intra-area coupling index, which is expressed as:

[0103]

[0104] Where: k is the number of the cluster; N is the total number of clusters; avg is the mean function;

[0105] S217. The comprehensive modularity function index of the network topology, the reactive power support capability of the cluster and the node coupling degree of the cluster can be expressed as:

[0106] ρ=ρ a +ρ b

[0107] Where: a is the improved modularity function; ρ b It is an improved comprehensive modularity function indicator.

[0108] Please see Figure 2 , using the above Louvain algorithm to cluster the distributed resources in the distributed resource aggregation network, specifically including the following steps:

[0109] S221. Obtain the running status information of the distributed resource device when performing cluster division, regard each node in the distributed resource aggregation network as a new cluster, and calculate the improved modularity value ρ according to the calculation method in the above step S217;

[0110] S222. randomly select node j and combine it with the node i and calculate to obtain the modularity increment, select the node with the largest modularity increment and the node i to form a new cluster, and update the improved modularity value;

[0111] S223. The current cluster is continued to be combined with other clusters as a new cluster, and the step S222 is repeated to form a new cluster division result until all nodes in the distributed resource aggregation network are traversed;

[0112] S224. When the modularity value in the distributed resource aggregation network reaches the maximum and no node can be merged, the cluster division is stopped. The cluster division result at this time is the optimal cluster division result, such as Figure 4 shown.

[0113] Depend on Figure 4It can be seen that the cluster division method provided in this embodiment 1 achieves a significant cluster division effect, and successfully divides the distributed resources in the distribution station area into multiple functional clusters according to their geographical location, load characteristics and grid demand, that is, into multiple regions, and the resources in each cluster are highly similar and highly correlated, and the coupling degree between different clusters is low, ensuring the coordination control efficiency and response speed within the cluster. The divided clusters can accurately reflect the voltage demand and power fluctuations in the substation area, and further improve the flexibility and accuracy of resource scheduling in the optimization control. The optimized clusters can not only independently cope with the voltage fluctuations in their respective areas, but also respond quickly to sudden load changes through coordinated linkage, thereby effectively reducing the overall voltage fluctuations in the substation area and improving the power quality.

[0114] Among them, please see Figure 3 , establishing an optimization control model of a distributed resource cluster and solving the optimization control model, including the following steps:

[0115] S31. Selecting each cluster leading node based on cluster division, the leading node is a node that can sensitively respond to active changes of other nodes in the cluster and can sensitively respond to reactive changes of other nodes in the cluster;

[0116] The selection of the cluster leading node is based on the principle of eliminating the voltage deviation of the node and minimizing the voltage deviation of other nodes in the cluster.

[0117] The observability index of node n is denoted by S Gn express:

[0118]

[0119]

[0120] Where: N represents the number of nodes in the cluster; α nk represents the voltage sensitivity of node n to node k; ΔU n With ΔU k Represents the voltage change of node n and node k respectively

[0121] The controllability index of node n is denoted by S jn express:

[0122]

[0123] In the formula, β nk represents the reactive voltage sensitivity of node n to node k; ΔQ k Represents the reactive power change of node j and defines the comprehensive index S of node n n for:

[0124] Sn =δ1S Gn +δ2S Kn

[0125] Where: δ1 and δ2 represent the weight coefficients of observability index and controllability index respectively, the selection of dominant node and voltage sensitivity α nk and reactive voltage sensitivity β nk It is mainly affected by the system topology and structural parameters.

[0126] In this embodiment 1, since the number of dominant nodes needs to be reasonably selected, too few nodes cannot fully represent the voltage levels of nodes in the entire area, and too many nodes will increase the computing pressure of the monitoring center. Therefore, a cluster selects one dominant node.

[0127] S32. Taking the minimum voltage deviation and the minimum system network loss at each of the cluster leading nodes as the control target, and taking the suppression of voltage fluctuation in the distribution station area as the strategy, establish objective function 1 and objective function 2, and perform weighted combination on objective function 1 and objective function 2, and meet the constraints of power flow constraint, cluster power constraint and safe operation constraint. The objective function 1 and objective function 2 are respectively expressed as:

[0128]

[0129] In the formula, j represents the number of dominant nodes; U j represents the voltage amplitude of the dominant node j, U j It is expressed in per-unit value; U0 represents the node voltage reference value, U0 = 1p.u.; M represents the number of clusters involved in the regulation; P loss,m Indicates the network loss of cluster m. The cluster network loss is calculated as follows:

[0130]

[0131] Where: N represents the total number of nodes in the cluster; l represents the set of nodes connected to node n; i nl represents the current in the branch connecting node n and node l, r nl Indicates the branch resistance

[0132] The objective function 1 and the objective function 2 are weighted and combined, which is expressed as:

[0133] minf=ω1minf1+ω2minf2

[0134] Where: ω1 and ω2 represent the weight coefficients of objective function 1 and objective function 2 respectively;

[0135] Among them, the constraints include power flow constraints, cluster power constraints and safe operation constraints. The power flow constraints can be expressed as:

[0136]

[0137] Where: u i with u j Represent the node voltages of node i and node j respectively; i ij represents the current of branch ij starting from node i and ending at node j; P ij With Q ij Respectively represent the active and reactive power flowing through branch ij; r ij With x ij Respectively represent the resistance and reactance of branch ij; P j With Q j They represent the net active and reactive loads injected into node j respectively; jl represents the set of branches connected to node j with node j as the starting point and node l as the end point, P jl With Q jl They represent the active and reactive power flowing through branch jl respectively.

[0138] The cluster power constraint can be expressed as:

[0139]

[0140] Where: P PV,m , Q PV,m They represent the distributed photovoltaic active and reactive output of cluster m respectively; represents the maximum value of distributed photovoltaic active output of cluster m; and They respectively represent the upper and lower limits of the distributed photovoltaic reactive output of cluster m.

[0141] The operational safety constraints can be expressed as:

[0142]

[0143] Where: and They represent the upper and lower limits of the dominant node voltage respectively, and the node voltage operating range is 0.95pu~1.07pu.

[0144] S34. Based on the established optimization control model of the distributed resource cluster, the active output and reactive output of each cluster as a whole are obtained by solving the optimization control model of the distributed resource cluster. According to the active output of each cluster as a whole and the reactive output of each cluster as a whole and the capacity of the photovoltaic power station, the output is redistributed within each cluster, as shown in the following formula:

[0145]

[0146] Where: P PV,m , Q PV,m They represent the distributed photovoltaic active output and reactive output of cluster m respectively; P PV,n With Q PV,n They represent the active output and reactive output of the nth PV power station in cluster m respectively; S PV,n represents the capacity of the nth PV power station; N represents the number of PV power stations in cluster m.

[0147] The maximum node voltage deviation value, the sum of all node voltage deviation values, and the network loss of the distribution network in the above-mentioned embodiment 1 are counted, and compared with the maximum node voltage deviation value, the sum of all node voltage deviation values, and the network loss of the distribution network before the distributed resource aggregation and cluster optimization control method for suppressing voltage fluctuations provided by the present invention is adopted. The results are shown in Table 1:

[0148] Table 1 Comparison of distribution network data between Example 1 and before the method in Example 1 is adopted

[0149] project Maximum node voltage deviation The sum of all node voltage deviations Network loss Before using Example 1 0.1049pu 0.5997pu 98.49kW Example 1 0.0579pu 0.3503pu 48.98kW

[0150] It can be seen from the data in Table 1 above that before using Example 1, that is, before control (the optimization control model of the distributed resource cluster is not adopted), the maximum deviation value of the node voltage is 0.1049pu, and in Example 1, that is, after control (the optimization control model of the distributed resource cluster), the maximum deviation value of the node voltage drops to 0.0579pu, indicating that: after adopting the method provided in this Example 1, the voltage deviation can be significantly reduced, and the fluctuation of the node voltage can be effectively suppressed; at the same time, the sum of the voltage deviation values ​​of all nodes is reduced from 0.5997pu before control to 0.3503pu after control, indicating that the overall voltage deviation is significantly optimized; in addition, the network loss is reduced from 98.49kW before control to 48.98kW after control, which significantly improves the energy efficiency of the system.

[0151] To sum up, the distributed resource aggregation and cluster optimization control method provided by the present invention can successfully reduce the voltage fluctuation of the distribution network and ensure that the system operates in an economical and safe state. The voltage of each node can be kept within a safe range, the stability of the system when affected by load fluctuations is improved, and the power quality is improved. Under the intelligent regulation of distributed resources in the substation, the network loss of the system can be effectively reduced, and the safe, stable and economical operation of the substation power supply system can be achieved.

[0152] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or substitutions within the technical scope disclosed by the present invention, and these modifications or substitutions should be included in the protection scope of the present invention.

Claims

1. A method for controlling distributed resources aggregation and cluster optimization in a substation area for suppressing voltage fluctuations, characterized in that: The following steps are involved: S10. Using the Internet of Things and ad hoc network technologies, various resource devices in the distribution area are connected to a unified communication network to build a distributed resource aggregation network; S20. Using the Louvain algorithm to cluster the distributed resources in the distributed resource aggregation network, generate the best clustering scheme for the distributed resources in the distribution station area, and obtain the best clustering result according to the best clustering scheme; S30. Based on the best result of the cluster division, an optimization control model of the distributed resource cluster is established, and the optimization control model is solved to obtain the overall active output of each divided cluster and the overall reactive output of each divided cluster. According to the overall active output of each divided cluster and the overall reactive output of each divided cluster and the capacity of the photovoltaic power station, the output is redistributed within each cluster to realize intelligent scheduling and optimization control of the distributed resource cluster, thereby realizing coordinated control within the cluster, suppressing voltage fluctuations and improving power quality.

2. The method for controlling distributed resources aggregation and cluster optimization in a substation area for suppressing voltage fluctuations according to claim 1 is characterized in that: In step S10, constructing the distributed resource aggregation network includes the following steps: S11. Resource equipment access and information collection: All kinds of distributed resource equipment in the distribution area are connected to the Internet of Things communication network. All kinds of distributed resource equipment are equipped with intelligent collection terminals to collect the corresponding operating status information of various distributed resource equipment in real time, form operating status information data, and transmit the operating status information to the centralized management node; S12. Building a self-organizing network communication architecture: The self-organizing network is used to establish communication links between the various types of resource devices to form a self-organizing network of distributed resources; S13. Edge computing of operation status information data: Through edge computing technology, real-time operation status information data processing and preliminary analysis are performed at the edge nodes of the distribution station area to realize local processing of operation status information data; S14. Information sharing and status synchronization: In a distributed resource aggregation network, the status information data of all resource devices is synchronously updated on the centralized management node, and the updated status information data of the resource devices can be shared; S15. Unified management and scheduling preparation: After the collected operating status information data is aggregated, edge computed and shared, the distributed resources within the distribution station area are integrated into a resource pool that can be uniformly managed and scheduled, and the construction of the distributed resource aggregation network is completed.

3. The method for controlling distributed resources aggregation and cluster optimization in a substation area for suppressing voltage fluctuations according to claim 2 is characterized in that: In the step S20, the Louvain algorithm includes: S211. Get the modularity function expression of the Louvain algorithm: Where: A ij is the weight of the edge between node i and node j, indicating the connectivity between nodes. It is 1 when nodes i and j are directly connected, otherwise it is 0; k i is the sum of the weights of all edges connected to node i; k j is the sum of the weights of all edges connected to node j; m = ∑ i ∑ j A ij / 2 is the sum of the weights of all edges in the network; if nodes i and j are in the same cluster, then δ(i,j) is 1, otherwise it is 0; S212. The weight of the modularity function is improved, and the improved weight is expressed as: Where: S OUij and S OUji They represent the reactive voltage sensitivity factors of node i to node j and node j to node i respectively; S213. Adjust the reactive power of node i to the reactive support capacity of node j, the weight of the reactive support capacity is expressed as: Where: Q PV,i , Q VSC,i and Q SVG,i are the adjustable reactive capacities of the PV inverter, VSC and SVG at node i respectively; S214. The improved weight matrix is ​​expressed as: A OUij ＝α OUij +β OUij S215. Obtain an improved modularity function of the Louvain algorithm, wherein the improved modularity function is: S216. The cluster coupling index can characterize the degree of coupling between nodes in a cluster. The average value of the coupling within each cluster area is taken as the total intra-area coupling index, which is expressed as: Where: k is the number of the cluster; N is the total number of clusters; avg is the mean function; S217. The comprehensive modularity function index of the network topology, the reactive power support capability of the cluster and the node coupling degree of the cluster can be expressed as: p=p a +r b Where: a is the improved modularity function; ρ b It is an improved comprehensive modularity function indicator.

4. The method for controlling distributed resources aggregation and cluster optimization in a substation area for suppressing voltage fluctuations according to claim 3 is characterized in that: In the step S20, clustering the distributed resources in the distributed resource aggregation network using the Louvain algorithm includes the following steps: S221. Obtain the operating status information of various resource devices in the distributed resource aggregation network, regard each node in the distributed resource aggregation network as a new cluster, and calculate the improved modularity value ρ; S222. randomly select node j and combine it with the node i and calculate to obtain the modularity increment, select the node with the largest modularity increment and the node i to form a new cluster, and update the improved modularity value; S223. The current cluster is continued to be combined with other clusters as a new cluster, and the step S222 is repeated to form a new cluster division result until all nodes in the distributed resource aggregation network are traversed; S224. When the modularity value in the distributed resource aggregation network reaches the maximum and no node can be merged, cluster division is stopped, and the cluster division result at this time is the optimal cluster division result.

5. The method for controlling distributed resources aggregation and cluster optimization in a substation area for suppressing voltage fluctuations according to claim 4 is characterized in that: In the step S30, the step of establishing the optimization control model of the distributed resource cluster and solving the optimization control model includes the following steps: S31. Selecting each cluster leading node based on cluster division, the leading node is a node that can sensitively respond to active changes of other nodes in the cluster and can sensitively respond to reactive changes of other nodes in the cluster; S32. Taking the minimum voltage deviation and the minimum system network loss at each of the cluster leading nodes as the control target, and taking the suppression of voltage fluctuation in the distribution station area as the strategy, establish objective function 1 and objective function 2, and perform weighted combination on objective function 1 and objective function 2, and meet the constraints of power flow constraint, cluster power constraint and safe operation constraint. The objective function 1 and objective function 2 are respectively expressed as: In the formula, j represents the number of dominant nodes; U j represents the voltage amplitude of the dominant node j, U j It is expressed in per-unit value; U0 represents the node voltage reference value, U0 = 1p.u.; M represents the number of clusters involved in the regulation; P loss,m represents the network loss of cluster m; The objective function 1 and the objective function 2 are weighted and combined, which is expressed as: minf=ω1minf1+ω2minf2 Where: ω1 and ω2 represent the weight coefficients of objective function 1 and objective function 2 respectively; S34. Based on the established optimization control model of the distributed resource cluster, the overall active output of each cluster and the overall reactive output of each cluster are obtained by solving the optimization control model of the distributed resource cluster. According to the overall active output of each cluster and the overall reactive output of each cluster, as well as the capacity of the photovoltaic power station, the output of each cluster is redistributed, as shown in the following formula: Where: P PV,m , Q PV,m They represent the distributed photovoltaic active output and reactive output of cluster m respectively; P PV,n With Q PV,n They represent the active output and reactive output of the nth PV power station in cluster m respectively; S PV,n represents the capacity of the nth PV power station; N represents the number of PV power stations in cluster m.

6. The method for controlling distributed resources aggregation and cluster optimization in a substation area for suppressing voltage fluctuations according to claim 5, characterized in that: In the step S32, the network loss of the cluster m is calculated by the following formula: Where: N represents the total number of nodes in the cluster; l represents the set of nodes connected to node n; i nl represents the current in the branch connecting node n and node l; r nl Indicates the resistance of this branch.

7. The method for controlling distributed resources aggregation and cluster optimization in a substation area for suppressing voltage fluctuations according to claim 5, characterized in that: In the step S32, the power flow constraint can be expressed as: Where: u i with u j Represent the node voltages of node i and node j respectively; i ij represents the current of branch ij starting from node i and ending at node j; P ij With Q ij Respectively represent the active and reactive power flowing through branch ij; r ij With x ij Respectively represent the resistance and reactance of branch ij; P j With Q j They represent the net active and reactive loads injected into node j respectively; jl represents the set of branches connected to node j with node j as the starting point and node l as the end point, P jl With Q jl They represent the active power and reactive power flowing through branch jl respectively.

8. The method for controlling distributed resources aggregation and cluster optimization in a substation area for suppressing voltage fluctuations according to claim 5, characterized in that: In the step S32, the cluster power constraint is expressed as: Where: P PV,m , Q PV,m They represent the distributed photovoltaic active output and reactive output of cluster m respectively; represents the maximum value of distributed photovoltaic active output of cluster m; and They respectively represent the upper and lower limits of the distributed photovoltaic reactive output of cluster m.

9. The method for controlling distributed resources aggregation and cluster optimization in a substation area for suppressing voltage fluctuations according to claim 5, characterized in that: In the step S32, the operation safety constraint is expressed as: Where: and They represent the upper and lower limits of the dominant node voltage respectively, and the node voltage operating range is taken as 0.95pu~1.07pu.

Images