Power distribution network voltage cooperative control method considering cluster division-optical storage all-in-one machine

Abstract

A power distribution network voltage cooperative control method considering a cluster division-optical storage all-in-one machine comprises the following steps: step 1, combining an electrical modularity index, a balance index and a membership index to construct a cluster comprehensive division index; 2, taking the comprehensive division index as a convergence condition for dividing a cluster region, and solving an optimal cluster division result of the power distribution network by utilizing an improved spectral clustering algorithm; 3, with the minimum power distribution network node voltage deviation punishment cost, the minimum network loss cost and the minimum power generation loss cost of the optical storage all-in-one machine as the target, a power distribution network voltage optimization model containing the optical storage all-in-one machine is constructed; and 4, based on the optimal cluster division result, solving the power distribution network voltage optimization model of the optical storage all-in-one machine through a cluster voltage optimization control strategy, optimizing active power, reactive power and voltage of the power distribution network, and improving power flow distribution of the whole network. The voltage regulation task sharing and power balance are realized, the voltage regulation effect is more reasonable, and the voltage out-of-limit problem under the random fluctuation of the source load power can be effectively solved.

Description

TECHNICAL FIELD

[0001] The application relates to a power distribution network voltage coordination control method considering cluster division-optical storage integrated machines and belongs to the technical field of power distribution networks. BACKGROUND

[0002] At present, the construction of a power distribution network voltage control index system based on cluster division mainly considers two aspects. On the one hand, the index system is constructed from the perspective of power supply and demand relationship by mainly constructing active and reactive power balance degree indexes to measure the voltage regulation capacity of a cluster. On the other hand, the index system is constructed from the perspective of cluster voltage regulation tasks by using cluster capacity to describe the relationship between distributed photovoltaic output and load and measuring the voltage regulation capacity of a cluster from the essence of causing node voltage out-of-limit. Based on this, the current power distribution network voltage coordination control strategies based on cluster division mainly include the following: dynamic cluster division and distributed coordination control, static cluster division based on the K-means algorithm, distributed power and energy storage coordination control, multi-level coordination control of main distribution networks, global optimization and cluster optimization control, so as to achieve the control of the voltage of the power distribution network. The existing power distribution network voltage coordination control strategies based on cluster division have made significant progress in improving control efficiency and robustness, but the existing cluster division is mostly based on electrical distance or fixed indexes, and the matching relationship between cluster capacity and regulation resources (such as energy storage and flexible load) is not fully considered. Moreover, the existing photovoltaic storage coordination strategies mostly focus on a single target (such as voltage out-of-limit management or economic optimization), lack the combination with cluster division, and do not fully exploit the adaptive voltage regulation capacity of photovoltaic storage integrated machines. Therefore, the existing power distribution network voltage coordination control strategies based on cluster division have problems such as insufficient scientificity of cluster division, limited adaptability of control strategies, communication and calculation burden, and lack of economic efficiency evaluation. These problems lead to the fact that in response to complex scenarios such as large-scale access of distributed photovoltaic and random fluctuations of load, the existing strategies are difficult to effectively solve the problems of voltage regulation task allocation, power balance and voltage out-of-limit. SUMMARY

[0003] In order to solve the problems existing in the prior art, the application provides a power distribution network voltage coordination control method considering cluster division-photovoltaic storage integrated machines, which overcomes the deficiencies of traditional methods in dynamic matching, randomness response and economic evaluation. By quantifying the voltage regulation capacity, designing a photovoltaic storage integrated machine coordination control strategy and establishing a comprehensive optimization model, the voltage regulation task allocation, power balance and adaptive voltage regulation are realized, which provides an innovative solution to the voltage control problem of large-scale access of distributed photovoltaic to a power distribution network.

[0004] This invention first establishes a comprehensive index for cluster division based on electrical modularity index, balance index, and membership index, and divides the clusters accordingly to balance the voltage regulation capabilities of each cluster. Secondly, based on the cluster division, a distribution network voltage optimization model considering cluster division and integrated photovoltaic-storage unit is constructed. Through cluster voltage optimization control methods, the active and reactive power of the integrated photovoltaic-storage unit is adjusted to optimize the active and reactive power and voltage of the distribution network, thereby improving the power flow distribution of the entire network.

[0005] The technical solution of the present invention is as follows: A distribution network voltage coordinated control method considering cluster partitioning and integrated photovoltaic-storage unit, comprising the following steps:

[0006] Step 1: Construct a comprehensive cluster partitioning index by combining electrical modularity index, balance index, and membership index;

[0007] Step 2: Using the comprehensive partitioning index as the convergence condition for dividing the cluster regions, the improved spectral clustering algorithm is used to obtain the optimal cluster partitioning result of the distribution network.

[0008] Step 3: With the goal of minimizing the voltage deviation penalty cost of distribution network nodes, network loss cost, and power generation loss cost of integrated photovoltaic and energy storage units, construct a distribution network voltage optimization model including integrated photovoltaic and energy storage units;

[0009] Step 4: Based on the optimal cluster partitioning results, the distribution network voltage optimization model of the photovoltaic-storage integrated machine is solved through the cluster voltage optimization control strategy to optimize the active and reactive power and voltage of the distribution network and improve the power flow distribution of the entire network.

[0010] In step one,

[0011] Modularity index is used to determine the size and structural strength of partitioned networks, and electrical distance is used to measure the tightness of electrical coupling between nodes. Combining modularity index and electrical coupling relationship between nodes, an electrical modularity index is constructed. for:

[0012]

[0013]

[0014] In the formula: This represents a weighted network adjacency matrix defined by electrical distance. For nodes in the distribution network and nodes Electrical distance between them Represents nodes in a distribution network and nodes The maximum electrical distance between them; Given a network adjacency matrix, if nodes and nodes There is a connection relationship, If node And node There is no connection relationship, ; is the sum of all edge weights in the network adjacency matrix ; And respectively represent the sum of all edge weights of node And node Assuming that node And node are in the same cluster, then , otherwise ;

[0015] Based on the load balancing theory, the balance index is defined as follows:

[0016]

[0017] In the formula: is the number of clusters, is the sum of the number of nodes in cluster in cluster , represents the reactive resource call amount of a certain key node of cluster in cluster , is the total sum of all node reactive resource call amounts, is the balance coefficient, and the balance index The smaller the value is, the more similar the cluster capacity is, and the more balanced the internal allocation of reactive resource call amount is.

[0018] The membership index is:

[0019]

[0020] In the formula: is the node membership index, and the smaller the value is, the more balanced the size of each cluster is; is the membership degree of the node in cluster which is connected to node ; is the sum of the number of nodes between clusters other than cluster ; is the membership degree of the node in cluster which is not in the same cluster but is connected to node .

[0021] In step one, the cluster comprehensive division index is constructed by combining the electrical modularity index, the balance index and the membership index, which is specifically:

[0022] Based on the three clustering indicators—combining electrical modularity, balance, and membership—the following comprehensive invention classification index is constructed:

[0023]

[0024] In the formula: , , The weights of the electrical modularity index, net load balancing index, and membership index are respectively represented, and they satisfy the following conditions: .

[0025] Step two specifically involves:

[0026] Step 1: Input the initial number of clusters, k=2;

[0027] Step 2: Based on the electrical modularity index Calculate the electrical modularity index for a cluster with k members. Calculate the electrical modularity index for a cluster with k+1 nodes. ;

[0028] Step 3: If Q k+1 Q k If k = k + 1, return to Step 2; otherwise, go to Step 4.

[0029] Step 4: Obtain the optimal number of clusters And through the adjacency matrix Construct degree matrix :

[0030]

[0031] Using the degree matrix of each node You can get The degree matrix is:

[0032]

[0033] Step 5: For the node degree matrix Calculate the corresponding normalized Laplace matrix. for:

[0034]

[0035] In the formula: for identity matrix;

[0036] Step 6: Obtain the Laplace matrices The eigenvalues ​​and eigenvectors are used to obtain the frontier. The eigenvectors corresponding to the smallest eigenvalues ​​will The feature vectors are merged to obtain a low-dimensional... Feature matrix ;

[0037] Step 7: Divide the indicators comprehensively For the characteristic matrix Divided into columns Each region is divided into a cluster based on the corresponding columns, and the optimal cluster partitioning result is obtained.

[0038] The objective function of the distribution network voltage optimization model including the photovoltaic-storage integrated unit in step three is:

[0039]

[0040] In the formula: Costs related to distribution network losses; Penalty cost for voltage deviation at distribution network nodes; The cost of generating electricity for integrated photovoltaic and energy storage systems.

[0041] Minimize the cost of distribution network losses Represented as:

[0042]

[0043] In the formula: The on-grid tariff is based on active power. Number of clusters; This is the set of all lines in each cluster; for Timetable The transmission current value; For the line The resistance value; To optimize the calculation step size for rolling;

[0044] Distribution network node voltage deviation penalty cost Represented as:

[0045]

[0046] In the formula: The unit penalty cost for node voltage deviation in the distribution network. This represents the total number of nodes in each cluster. For nodes Voltage upper limit, For nodes Lower voltage limit for Time Node The voltage deviation value can be specifically expressed as:

[0047]

[0048] Loss cost of photovoltaic and energy storage integrated generator For:

[0049]

[0050] The revenue of photovoltaic and energy storage integrated generator per unit of electricity generation; The set of all photovoltaic and energy storage integrated generator grid-connected nodes; For Instantaneous node Distributed photovoltaic active power reduction of photovoltaic and energy storage integrated generator; For Instantaneous node Energy storage charging power of photovoltaic and energy storage integrated generator, For Instantaneous node Energy storage discharging power of photovoltaic and energy storage integrated generator; The charging and discharging efficiency of energy storage unit, The rolling optimization calculation step, The node belonging to the set of photovoltaic and energy storage integrated generator grid-connected nodes .

[0051] The constraint conditions of the photovoltaic and energy storage integrated generator-containing distribution network voltage optimization model in step three include:

[0052] The branch power flow equation constraint is:

[0053]

[0054]

[0055]

[0056]

[0057] In the formula: The active power from node to node at the moment, The reactive power from node to node at the moment, The active power from node to node at the moment, The active power from node to node at the moment, The active power from node ​instantaneous from node instantaneous to node reactive power; is instantaneous line transmission current value, instantaneous node active power of the net load, is instantaneous node reactive power of the net load; is reactance value of the line; is resistance value of the line; is instantaneous node voltage amplitude, is instantaneous node voltage amplitude;

[0058] The net load power of each node is:

[0059]

[0060]

[0061] In the formula: is instantaneous node net active power of the load, is instantaneous node net reactive power of the load; instantaneous node active power of the load, is instantaneous node reactive power of the load; is instantaneous node active power of the maximum power point of the distributed photovoltaic unit; is instantaneous node reactive output power of the light storage;

[0062] The safe operation constraint of the power distribution network is:

[0063]

[0064]

[0065] , Load nodes Upper and lower voltage limits Line Maximum transmission current

[0066] Operation constraints of the optical storage integrated machine:

[0067] The operation constraints of the energy storage unit mainly include the charge and discharge power constraints and the upper and lower limits of the remaining capacity, i.e.

[0068]

[0069] In the formula: is Time node The energy storage charging power of the optical storage integrated machine, is Time node The energy storage discharging power of the optical storage integrated machine; is a binary variable, indicating Time node The charge and discharge state of the energy storage, taking the value of 1 when the energy storage is discharging and 0 when the energy storage is charging; is the charge and discharge efficiency of the energy storage unit, is the rolling optimization calculation step size; is the node The maximum charge and discharge power of the energy storage; is the node The remaining charge capacity of the energy storage at the node at the time, is the node The remaining charge capacity of the energy storage at the node at the time; is the node The minimum remaining capacity allowed in the operation process of the energy storage, is the node The maximum remaining capacity allowed in the operation process of the energy storage;

[0070] Active power reduction amount of the distributed photovoltaic unit Subject to the maximum power point constraint:

[0071]

[0072] The reactive power output of the optical storage integrated machine is mainly subject to the capacity and power factor constraints of the inverter, and is:

[0073]

[0074] In the formula: is the node The capacity of the optical storage inverter; The maximum power factor angle of the light storage inverter output power.

[0075] The step four cluster voltage optimization control strategy adopts the improved particle swarm algorithm for optimization control.

[0076] Step four is specifically:

[0077] Step 41: population parameter initialization, set the parameters of the improved particle swarm algorithm combined with the optimal cluster division result; The parameters set include the number of population , the space dimension is , the maximum number of iterations is , the initial position and speed of each cluster particle;

[0078] Step 42: set the search range and constraint conditions of the particles in each cluster of the power distribution network, and set the constraint conditions as the upper and lower limits of the node voltage and the upper and lower limits of the remaining capacity of each node of the light storage integrated machine;

[0079] Step 43: calculate the fitness function , The objective function of the voltage optimization model of the power distribution network containing the light storage integrated machine, judge whether the fitness of the adjacent iteration changes, if yes, jump to step 44; If not, jump to step 47;

[0080] Step 44: find the individual fitness and group fitness of the particles in each cluster of the power distribution network according to the constraint conditions and fitness set in step 42;

[0081] Step 45: update the particle position and speed of each cluster in the power distribution network, update the individual fitness and group fitness of each cluster in the power distribution network;

[0082] Step 46: judge the number of iterations, if the maximum number of iterations is not reached, return to step 43;

[0083] Step 47: get the optimal solution, output the active and reactive power of the light storage integrated machine in each cluster of the power distribution network, and thus improve the power flow distribution of the whole network.

[0084] The improved particle swarm algorithm is:

[0085] On the basis of the linear decreasing strategy, the inertia weight combination update formula is constructed:

[0086]

[0087] Among them, is the current iteration number, is the maximum iteration number; indicates the weight of the current iteration number, and These represent the maximum weight and the minimum weight, respectively. Indicates the variance of group fitness. This represents the threshold of population fitness variance.

[0088] The present invention has the following beneficial effects:

[0089] By dividing clusters based on comprehensive indicators, the resulting cluster sizes are more reasonable, the net load distribution among clusters tends to be more balanced, and the degree of aggregation of nodes within the clusters is high.

[0090] This invention employs a cluster-based approach for distributed voltage control of distribution network nodes. Specifically, network nodes are divided into multiple clusters according to the comprehensive indicators described in this application. By optimizing the reactive and active power outputs of key nodes within each cluster (such as the integrated photovoltaic and energy storage unit), the power flow distribution is improved to achieve global voltage control of the distribution network. This results in a more balanced distribution of voltage regulation tasks and power equilibrium, leading to a more reasonable voltage regulation effect and effectively solving the problem of voltage exceeding limits under random fluctuations in source and load power.

[0091] This invention takes into account both cluster capacity and regional regulation resources, quantifies the voltage regulation capacity within the region, and divides the cluster accordingly. This achieves even distribution of voltage regulation tasks and power balance, resulting in a more reasonable voltage regulation effect. Furthermore, considering the potential voltage fluctuations and limit exceedances caused by random fluctuations in distributed photovoltaic power output and load power consumption, this invention integrates distributed photovoltaics with energy storage to address voltage limit exceedances within the cluster and distributed photovoltaic absorption issues. By adjusting the active and reactive power of the integrated photovoltaic-energy storage unit, the voltage limit exceedance problem can be effectively solved, and it possesses adaptive voltage regulation capabilities under random fluctuations in source and load power. Attached Figure Description

[0092] Figure 1 This is a flowchart of the method of the present invention;

[0093] Figure 2 This is a flowchart for obtaining the optimal cluster partitioning result based on the improved spectral clustering algorithm. Detailed Implementation

[0094] like Figure 1 This invention provides a distribution network voltage coordinated control method considering cluster partitioning and integrated photovoltaic-storage units, comprising the following steps:

[0095] Step 1: Construct a comprehensive cluster partitioning index by combining three indicators: electrical modularity index, balance index, and membership index;

[0096] Step 2: Using the comprehensive partitioning index as the convergence condition for dividing the cluster regions, the improved spectral clustering algorithm is used to obtain the optimal cluster partitioning result of the distribution network.

[0097] Step 3: Based on the cluster partitioning, construct a distribution network voltage optimization model that includes the photovoltaic-storage integrated unit, with the goal of minimizing the distribution network voltage deviation penalty cost, network loss cost, and photovoltaic-storage integrated unit power generation loss cost;

[0098] Step 4: Based on the optimal cluster partitioning results, the voltage optimization model of the distribution network considering the cluster partitioning of the photovoltaic-storage integrated unit is solved through the cluster voltage optimization control strategy to optimize the active and reactive power and voltage of the distribution network and improve the power flow distribution of the entire network.

[0099] In step one, the cluster comprehensive division index includes three indicators: electrical modularity index, balance index, and membership index.

[0100] (1) Electrical modularity index

[0101] In selecting cluster partitioning metrics, the modularity metric is used to determine the size and structural strength of the partitioned network, and electrical distance is used to measure the tightness of electrical coupling between nodes. Combining the modularity metric and the electrical coupling relationship between nodes, an electrical modularity metric is constructed. for:

[0102]

[0103]

[0104] In the formula: This represents a weighted network adjacency matrix defined by electrical distance. For nodes in the distribution network and nodes Electrical distance between them Represents nodes in a distribution network and nodes The maximum electrical distance between them; Given a network adjacency matrix, if nodes and nodes There is a connection. If node and nodes There is no connection relationship. If the node and nodes Within the same cluster ,otherwise . It is a network adjacency matrix The sum of all edge weights in the middle; and Representing the nodes respectively and nodes The sum of all edge weights;

[0105] (2) Balance index

[0106] When the node voltage of the power distribution network is out of limit, in order to distribute the total voltage regulation task as much as possible in each cluster, the nodes in the cluster call the reactive power resources of the optical storage nodes to undertake the same or similar voltage regulation task, and the task digestion in the cluster is maximized. Based on the load balancing theory, the balance index is defined as follows:

[0107]

[0108] In the formula: is the number of clusters, is the cluster in the cluster , the sum of the number of nodes in the cluster , the reactive power resource calling amount of a certain key node in the cluster , the reactive power resource calling amount of all nodes, is the balance coefficient, which is set to 1. The obtained balance index , the smaller the value, the more similar the cluster capacity, and the more balanced the internal distribution of reactive power resources. (3) Membership index

[0109] After limiting the size and capacity of the cluster division, in order to improve the balance degree of the cluster size and reduce the complexity of the cluster optimization model, the membership index is defined as:

[0110]

[0111]

[0112] In the formula: is the node membership index, the smaller the value, the more balanced the size of each cluster; is the membership degree of the node in the same cluster ; is the sum of the number of edges between nodes in the remaining clusters except the cluster ; is the membership degree of the node which is not in the same cluster but connected to the node in the cluster .

[0113] (4) Comprehensive division index

[0114] The above three cluster division indexes are comprehensively selected to construct the comprehensive division index of the application as follows:

[0115]

[0116] In the formula:​ 、 、 respectively represent the weight of electrical modularity index, net load balance index and membership index, and satisfy .

[0117] In the step two, the comprehensive division index is used as the convergence condition, and the improved spectral clustering algorithm is used to obtain the optimal cluster division result, and the process is as Figure 2 , including the following steps:

[0118] Step 1: input the initial cluster number k=2;

[0119] Step 2: according to the electrical modularity index , the electrical modularity index of the cluster number k is calculated , the electrical modularity index of the cluster number k+1 is calculated ;

[0120] Step 3: when Q k+1 >Q k , k=k+1, return to Step 2; otherwise, go to Step 4;

[0121] Step 4: obtain the optimal cluster number , and construct the degree matrix through the adjacency matrix :

[0122]

[0123] Using the degree matrix of each node , the degree matrix can be obtained :

[0124]

[0125] Step 5: for each node degree matrix , the corresponding normalized Laplacian matrix is calculated as:

[0126]

[0127] In the formula: is the unit matrix.

[0128] Step 6: the characteristic value and the characteristic vector of each are calculated, the characteristic vector corresponding to the first minimum characteristic value is obtained, and the low-dimensional characteristic matrix is obtained after arrangement and combination. . ​

[0129] Step 7: Comprehensive index division Feature matrix Split by column into regions, and the nodes corresponding to the columns in the region are divided into a cluster to obtain the optimal cluster division result.

[0130] Step 3, based on the cluster division result, a power distribution network voltage optimization model considering cluster division and light storage integrated machine is constructed with the power distribution network voltage deviation penalty cost, network loss cost, and light storage integrated machine power generation loss cost as targets; the power distribution network voltage optimization model considering cluster division and light storage integrated machine includes an objective function and constraint conditions;

[0131] The objective function is established with the power distribution network voltage deviation penalty cost, network loss cost, and light storage integrated machine power generation loss cost; the objective function is as follows:

[0132]

[0133] In the formula: is the power distribution network network loss cost; is the power distribution network node voltage deviation penalty cost; is the light storage integrated machine power generation loss cost.

[0134] (1) The minimum power distribution network network loss cost is taken as the objective function, which can be expressed as:

[0135]

[0136] In the formula: is the active power online price; is the number of clusters; is the set of all lines in each cluster; is the transmission current value of line at time t; is the resistance value of line ; and is the rolling optimization calculation step, which is 1 hour. (2) The power distribution network node voltage deviation penalty cost

[0137] is taken as the objective function, which can be expressed as:

[0138]

[0139] In the formula: is the unit penalty cost of node voltage deviation in the power distribution network, is the total number of nodes in each cluster,​ , They are nodes Upper voltage limit and lower voltage limit, for Time Node The voltage deviation value can be specifically expressed as:

[0140]

[0141] (3) Optimize the active and reactive power output of the photovoltaic-storage integrated machine from the current time to 24:00, and calculate the power generation loss cost of the photovoltaic-storage integrated machine. for:

[0142]

[0143] The revenue per unit of electricity generated by the photovoltaic-storage integrated unit; This refers to the collection of all grid-connected nodes of the integrated photovoltaic and energy storage system. for Time Node The reduction in active power of distributed photovoltaic power generation in integrated photovoltaic and energy storage systems; and They are respectively Time Node The energy storage charging and discharging power of the photovoltaic-energy storage integrated machine; The charging and discharging efficiency of the energy storage unit. To optimize the calculation step size for rolling, a value of 1 hour is used.

[0144] In step three, the power grid voltage optimization model for the integrated photovoltaic and energy storage unit that takes into account cluster partitioning includes constraints.

[0145] 9) The branch power flow equation constraint is:

[0146]

[0147]

[0148]

[0149]

[0150] In the formula: and They are respectively From the node Flow to Node Active and reactive power, for From the node Flow to Node active power, for From the node Flow to Node reactive power; for Timetable The transmission current value, and for Time Node The active and reactive power of the net load; For the line The reactance value; For the line The resistance value; and They are respectively Time Node and nodes The voltage amplitude.

[0151] 10) The net load power of each node is:

[0152]

[0153]

[0154] In the formula: and for Time Node Net active and net reactive power of the load; and for Time Node The active and reactive power of the load; for Time Node The active power generation at the maximum power point of a distributed photovoltaic unit; for Time Node The reactive power output of photovoltaic energy storage.

[0155] 11) The constraints for the safe operation of the distribution network are:

[0156]

[0157]

[0158] , They are respectively load nodes Upper voltage limit and lower voltage limit; For the line The maximum transmission current.

[0159] 12) Operation constraints of the light storage integrated machine:

[0160] The operation constraints of the energy storage unit mainly include the charge and discharge power constraints and the upper and lower limits of the remaining capacity constraints, that is,

[0161]

[0162] In the formula: is a binary variable, indicating the node at time the charge and discharge state of the energy storage, taking 1 to represent discharging of the energy storage and 0 to represent charging of the energy storage; is the node the maximum charge and discharge power of the energy storage; , are the remaining charge capacities of the node the energy storage at the time and the time ; and are the minimum and maximum remaining capacities of the node the energy storage during the operation process.

[0163] The active power reduction amount of the distributed photovoltaic unit is subject to the maximum power point constraint:

[0164]

[0165] The reactive power output of the light storage integrated machine is mainly subject to the capacity and power factor of the inverter, and is:

[0166]

[0167] In the formula: is the capacity of the node the light storage inverter; is the maximum power factor angle of the output power of the light storage inverter.

[0168] The fourth step is to adjust the active and reactive power of the light storage integrated machine through the cluster voltage optimization control strategy, optimize the voltage of the power distribution network, and improve the distribution of the whole network power flow. The cluster voltage optimization control strategy adopts an improved particle swarm algorithm for optimization control, and the steps are as follows:

[0169] Step 41: population parameter initialization, set the parameters in the improved particle swarm algorithm in combination with the optimal cluster division result; the parameters set include the number of population , the space dimension is , and the maximum number of iterations is ; set the initial position and speed of each cluster particle;

[0170] Step 42: set the search range of particles in each cluster of the power distribution network and the constraint condition, and set the constraint condition as the upper and lower limits of node voltage and the upper and lower limits of the capacity of each node optical storage integrated machine;

[0171] Step 43: calculate the fitness function , judge whether the fitness of the next iteration changes, if yes, jump to step 44; if not, jump to step 47;

[0172] Step 44: find the individual fitness value and group fitness value of the particles in each cluster of the power distribution network according to the constraint condition and the fitness;

[0173] Step 45: update the particle position and speed of each cluster in the power distribution network, and update the individual fitness value and group fitness value of each cluster in the power distribution network;

[0174] Step 46: judge the iteration number, if the maximum iteration number is not reached, return to step 43;

[0175] Step 47: obtain the optimal solution, output the active power and reactive power of the optical storage integrated machine in each cluster of the power distribution network, and thus improve the power flow distribution of the whole network.

[0176] In order to improve the search performance of the algorithm and enhance the ability of the algorithm to jump out of the local optimum, the particle swarm algorithm is improved, and on the basis of the linear decreasing strategy, a new inertia weight combination update formula is constructed:

[0177]

[0178] Wherein, is the current iteration number, is the maximum iteration number; indicates the weight of the current iteration number, and respectively indicate the maximum weight and the minimum weight, indicates the group fitness variance, indicates the group fitness variance threshold.

[0179] In the calculation process, the inertia weight will be adaptively adjusted according to the particle fitness value, for the particles with poor fitness value, the inertia weight is matched to a larger value, which can drive the particles to fly to a farther solution space to search for the optimal solution, and the ability to jump out of the local optimum is improved; and for the particles with better fitness value, the inertia weight is matched to a smaller value, so that the particles can search for the optimal solution in a local range, and the local search ability of the particles is improved.​

Claims

1. A distribution network voltage coordinated control method considering cluster partitioning and integrated photovoltaic-storage unit, characterized in that, Includes the following steps: Step 1: Construct a comprehensive cluster partitioning index by combining electrical modularity index, balance index, and membership index; Step 2: Using the comprehensive partitioning index as the convergence condition for dividing the cluster regions, the improved spectral clustering algorithm is used to obtain the optimal cluster partitioning result of the distribution network. Step 3: With the goal of minimizing the voltage deviation penalty cost of distribution network nodes, network loss cost, and power generation loss cost of integrated photovoltaic and energy storage units, construct a distribution network voltage optimization model including integrated photovoltaic and energy storage units; Step 4: Based on the optimal cluster partitioning results, the distribution network voltage optimization model of the photovoltaic-storage integrated machine is solved through the cluster voltage optimization control strategy to optimize the active and reactive power and voltage of the distribution network and improve the power flow distribution of the entire network.

2. The distribution network voltage coordinated control method considering cluster partitioning and integrated photovoltaic-storage unit as described in claim 1, characterized in that, In step one, Modularity index is used to determine the size and structural strength of partitioned networks, and electrical distance is used to measure the tightness of electrical coupling between nodes. Combining modularity index and electrical coupling relationship between nodes, an electrical modularity index is constructed. for: ； ； In the formula: This represents a weighted network adjacency matrix defined by electrical distance. For nodes in the distribution network and nodes Electrical distance between them Represents nodes in a distribution network and nodes The maximum electrical distance between them; Given a network adjacency matrix, if the nodes and nodes There is a connection. If node and nodes There is no connection relationship. ; It is a network adjacency matrix The sum of all edge weights in the middle; and Representing the nodes respectively and nodes The sum of all edge weights; assuming a node and nodes In the same cluster, ,otherwise ; Based on load balancing theory, the following load balancing metrics are defined: ； In the formula: Number of clusters For cluster Clusters in Total number of internal nodes Represents a cluster Clusters in The amount of reactive power used at a certain critical node This represents the total amount of reactive power resource usage across all nodes. Equilibrium coefficient, equilibrium index The smaller the value, the more similar the cluster capacity, and the more balanced the internal allocation of reactive resource usage. The membership index is: ； In the formula: This is a node membership index; the smaller the value, the more balanced the size of the clusters. For nodes Nodes that are connected within the same cluster are in the cluster The degree of subordination within; To remove cluster The sum of the number of edges between nodes in the remaining clusters; For nodes Nodes that are not in the same cluster but are connected are in the cluster. The degree of subordination.

3. A distribution network voltage coordinated control method considering cluster partitioning and integrated photovoltaic-storage unit as described in claim 1 or 2, characterized in that, In step one, a comprehensive cluster partitioning index is constructed by combining the electrical modularity index, the balance index, and the membership index, specifically as follows: Based on the three clustering indicators—combining electrical modularity, balance, and membership—the following comprehensive invention classification index is constructed: ； In the formula: , , The weights of the electrical modularity index, net load balancing index, and membership index are respectively represented, and they satisfy the following conditions: .

4. The distribution network voltage coordinated control method considering cluster partitioning and integrated photovoltaic-storage unit as described in claim 1, characterized in that, Step two specifically involves: Step 1: Input the initial number of clusters, k=2; Step 2: Based on the electrical modularity index Calculate the electrical modularity index for a cluster with k members. Calculate the electrical modularity index for a cluster with k+1 nodes. ; Step 3: If Q k+1 Q k If k = k + 1, return to Step 2; otherwise, go to Step 4. Step 4: Obtain the optimal number of clusters And through the adjacency matrix Construct degree matrix : ； Using the degree matrix of each node You can get The degree matrix is: ； Step 5: For the node degree matrix Calculate the corresponding normalized Laplace matrix. for: ； In the formula: for identity matrix; Step 6: Obtain the Laplace matrices The eigenvalues ​​and eigenvectors are used to obtain the frontier. The eigenvectors corresponding to the smallest eigenvalues ​​will The feature vectors are merged to obtain a low-dimensional... Feature matrix ; Step 7: Divide the indicators comprehensively For the characteristic matrix Divided into columns Each region is divided into a cluster based on the corresponding columns, and the optimal cluster partitioning result is obtained.

5. The distribution network voltage coordinated control method considering cluster partitioning and integrated photovoltaic-storage unit as described in claim 1, characterized in that, The objective function of the distribution network voltage optimization model including the photovoltaic-storage integrated unit in step three is: ； In the formula: Costs related to distribution network losses; Penalty cost for voltage deviation at distribution network nodes; The cost of generating electricity for integrated photovoltaic and energy storage systems.

6. The distribution network voltage coordinated control method considering cluster partitioning-photovoltaic-storage integrated unit according to claim 5, characterized in that, Minimize the cost of distribution network losses Represented as: ； In the formula: The on-grid tariff is based on active power. Number of clusters; This is the set of all lines in each cluster; for Timetable The transmission current value; For the line The resistance value; To optimize the calculation step size for rolling; Distribution network node voltage deviation penalty cost Represented as: ； In the formula: The unit penalty cost for node voltage deviation in the distribution network. This represents the total number of nodes in each cluster. For nodes Voltage upper limit, For nodes Lower voltage limit for Time Node The voltage deviation value can be specifically expressed as: ； Photovoltaic-storage integrated power generation loss cost for: ； The revenue per unit of electricity generated by the photovoltaic-storage integrated unit; This refers to the collection of all grid-connected nodes of the integrated photovoltaic and energy storage system. for Time Node The reduction in active power of distributed photovoltaic power generation in integrated photovoltaic and energy storage systems; for Time Node The energy storage and charging power of the integrated photovoltaic and energy storage unit for Time Node The energy storage discharge power of the integrated photovoltaic and energy storage unit; The charging and discharging efficiency of the energy storage unit. To optimize the calculation step size for rolling, This represents the set of grid-connected nodes belonging to the integrated photovoltaic and energy storage system. nodes .

7. A distribution network voltage coordinated control method considering cluster partitioning-photovoltaic-storage integrated unit as described in claim 1, 5, or 6, characterized in that, The constraints of the distribution network voltage optimization model containing the photovoltaic-storage integrated unit in step three include: 1) The branch power flow equation constraint is: ； ； ； ； In the formula: for From the node Flow to Node active power, for From the node Flow to Node reactive power, for From the node Flow to Node active power, for From the node Flow to Node reactive power; for Timetable The transmission current value, Time Node The active power of the net load, for Time Node Reactive power of net load; For the line The reactance value; For the line The resistance value; for Time Node voltage amplitude, for Time Node The voltage amplitude; 2) The net load power of each node is: ； ； In the formula: for Time Node The net active power of the load, for Time Node The net reactive power of the load; Time Node The active power of the load, for Time Node The reactive power of the load; for Time Node The active power generation at the maximum power point of a distributed photovoltaic unit; for Time Node The reactive power output of photovoltaic energy storage; 3) The constraints for the safe operation of the distribution network are: ； ； , They are respectively load nodes Upper voltage limit and lower voltage limit; For the line Maximum transmission current; 4) Operational constraints of integrated photovoltaic and energy storage systems: The operational constraints of energy storage units mainly include charging and discharging power constraints and upper and lower limits of remaining capacity constraints, namely... ； In the formula: for Time Node The energy storage and charging power of the integrated photovoltaic and energy storage unit for Time Node The energy storage discharge power of the integrated photovoltaic and energy storage unit; A binary variable, representing Time Node The charging and discharging state of energy storage is represented by a value of 1, which indicates energy storage is discharging, and a value of 0, which indicates energy storage is charging. The charging and discharging efficiency of the energy storage unit. To optimize the calculation step size for rolling; For nodes Maximum charge and discharge power of energy storage; For nodes Energy storage at nodes Remaining charge capacity at time t. For nodes Energy storage at nodes The remaining charge capacity at any given time; For nodes The minimum remaining capacity of energy storage during operation. For nodes Energy storage is allowed a maximum remaining capacity during operation; Active power reduction of distributed photovoltaic units The maximum power point constraint is: ； The reactive power output of the photovoltaic-storage integrated unit is mainly constrained by the inverter's capacity and power factor. ； In the formula: For nodes The capacity of the photovoltaic-storage inverter; The maximum power factor angle of the photovoltaic-storage inverter output power.

8. The distribution network voltage coordinated control method considering cluster partitioning-photovoltaic-storage integrated unit according to claim 1, characterized in that, The fourth step, cluster voltage optimization control strategy, employs an improved particle swarm optimization algorithm for optimization control.

9. A distribution network voltage coordinated control method considering cluster partitioning-photovoltaic-storage integrated unit as described in claim 1 or 8, characterized in that, Step four is as follows: Step 41: Population parameter initialization. Based on the optimal cluster partitioning results, set various parameters in the improved particle swarm optimization algorithm; the parameters set include the number of population members. Spatial dimension is The maximum number of iterations is The initial position and velocity of each cluster of particles; Step 42: Set the search range and constraints for particles in each cluster of the distribution network. Set the constraints as the upper and lower limits of node voltage and the upper and lower limits of the remaining capacity of each node's photovoltaic-storage integrated unit. Step 43: Calculate the fitness function , The objective function of the distribution network voltage optimization model including photovoltaic and energy storage units is to determine adjacent... If the fitness changes in the next iteration, then proceed to step 44. If not, skip to step 47; Step 44: Find the individual fitness value and the population fitness value of particles in each cluster of the distribution network according to the constraints and fitness set in Step 42; Step 45: Update the particle positions and velocities of each cluster in the distribution network, and update the individual fitness and population fitness of each cluster in the distribution network; Step 46; Check the number of iterations. If the maximum number of iterations has not been reached, return to step 43. Step 47: Obtain the optimal solution and output the active and reactive power of the photovoltaic-storage integrated units in each cluster of the distribution network, thereby improving the power flow distribution of the entire network.

10. A distribution network voltage coordinated control method considering cluster partitioning-photovoltaic-storage integrated unit as described in claim 8, characterized in that, The improved particle swarm optimization algorithm is as follows: Based on the linear decreasing strategy, a formula for updating the inertia weight combination is constructed: ； in, This represents the current iteration number. This represents the maximum number of iterations. The weight representing the current iteration number. and These represent the maximum weight and the minimum weight, respectively. Indicates the variance of group fitness. This represents the threshold of population fitness variance.

Images