Distribution network distributed energy storage configuration method and device considering comprehensive elasticity

Abstract

The invention relates to a distribution network distributed energy storage configuration method and equipment considering comprehensive elasticity, and the method comprises the steps: constructing a comprehensive elasticity index system which comprises an elasticity index, a new energy matching index, a modularity index and a voltage qualification index; performing cluster division on the power distribution network by taking the comprehensive elastic index system as a reference basis; and a multi-target particle swarm algorithm is adopted to solve the energy storage locating and sizing model in the cluster, a distributed energy storage configuration scheme of each cluster of the power distribution network is obtained, and multiple targets of the locating and sizing model comprise the minimum cost target, the maximum voltage stability and the optimal self-healing rate index. Compared with the prior art, the method has the advantages that the overall self-healing rate and economical efficiency of the power distribution network are improved, the power distribution network runs more stably, and the resource utilization rate is higher.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network planning, and in particular to a method and device for configuring distributed energy storage in a distribution network taking comprehensive elasticity into consideration. Background Art

[0002] With the large-scale integration of renewable energy into the distribution network, the unpredictability and time-varying characteristics of various distributed energy sources and loads have posed challenges to the planning, operation and dispatching of the distribution network. In terms of grid operation, the power supply mode of the power system has shifted from "generation on demand" to "generation on demand interaction". Traditional distribution network planning usually focuses on the radial power supply network of a single power source. However, in the context of a high proportion of new energy access, the uncertainty of multi-source supply and renewable energy output makes distribution network planning more complicated. In terms of grid dispatching, the individual capacity of distributed energy is small and scattered, and it is constrained by access rules, which limits its flexibility to actively participate in grid dispatching. This limitation may lead to waste of resources due to insufficient system accommodation capacity.

[0003] Considering the integration of distributed energy and loads in urban power grids, the traditional centralized control strategy needs to shift to decentralized control. This change has led to the need for demand-side resource clustering. Demand-side resource clustering mainly involves the combination of distributed energy devices to promote self-organization and self-management within the cluster, while achieving inter-cluster coordination through grid dispatching and management. At present, there are some studies on demand-side resource allocation strategies in academia. Current research on demand-side resource partitioning mainly revolves around two key issues: partitioning criteria and partitioning algorithms. This process usually starts with establishing partitioning criteria based on specific goals. Then, a suitable clustering algorithm is selected, and the iterative index is used as the clustering criterion to finally obtain the best possible clustering solution. The clustering criteria usually revolve around evaluating the degree of interdependence between clusters. Specifically, they emphasize weaker interdependence between clusters to promote specialization; and emphasize stronger connections within clusters to encourage collaboration. Clustering algorithms are usually determined by the characteristics of clustering criterion indicators, focusing mainly on cluster analysis and optimization algorithms. Luo X et al. Optimization of a solar-based integrated energy system considering interaction between generation, network, and demand side[J]. Applied Energy, 2021, 294(8): 116931. The electrical distance between load nodes is defined as the weight of the graph edge, and the spectral clustering algorithm is used for clustering. Ma Qipeng et al. Network cluster and voltage coordination control of high-penetration distributed photovoltaic distribution network[J]. Power System and Clean Energy, 2023, 39(03): 93-102+108. The quality of the power grid cluster was evaluated using electrical modularization and reactive power reserve verification as two evaluation criteria. In addition, an improved k-means clustering algorithm was proposed to perform clustering analysis on the feature sequence after dimensionality reduction. Liu N et al. Distributed energy management for interconnected operation of combined heat and power-based microgrids with demand response[J]. Journal of Modern Power System and Clean Energy, 2017(3): 11. The concept of complex network modularization function was introduced, and the photovoltaic power generation cluster was divided using the fast expansion clustering algorithm. Applying hierarchical clustering algorithms to demand-side resource partitioning has the advantages of simple calculation and high efficiency. However, it must be recognized that this method often requires a more robust global search function and is prone to falling into local optimal states.

[0004] Mi Yang, Zhou Jie, Lu Changkun, et al. Multi-objective bi-level planning of distribution network based on improved generative adversarial network and carbon footprint [J / OL]. Proceedings of the CSEE: 1-15 [2024-05-01]. When the output of wind and solar power is unstable, the output of new energy units will drop significantly. Considering that there is a certain slope limit for the distribution network to purchase electricity upward, it may cause large-scale power outages, causing serious losses to the economy. Large-scale power outages usually lead to production interruptions and service interruptions in multiple industries, causing serious impacts on the economic activities of enterprises and individuals. Power outages may cause factories to stop production, stores to be unable to operate normally, traffic to be blocked, and medical facilities to be unable to operate normally. These impacts will not only lead to direct economic losses, but may also have a negative impact on social stability and people's lives. At the same time, in order to cope with the dual crises of energy depletion and environmental pollution, the large-scale introduction of distributed power sources (DG) and their connection with the distribution network to realize power substitution load access provide a feasible solution for load recovery. Therefore, making full use of various distributed resources to reduce power outage losses in the distribution network planning stage is of great significance to improving the resilience of the distribution network.

[0005] Energy storage systems (ESS) have the ability of rapid response and bidirectional power regulation, can transfer electric energy in time and space, and effectively balance system fluctuations to maintain system resilience. Therefore, proposing an optimization scheme to improve the resilience of the distribution network is of practical significance for efficient resource utilization and ensuring the safety and stability of the power grid.

[0006] At present, domestic and foreign scholars have conducted extensive research on the reasonable site selection and capacity configuration of energy storage for improving the elasticity of distribution networks. Wang Fashun, Research on Multi-objective Optimal Configuration of Energy Storage and Controllable DG in Smart Distribution Network Considering Elasticity Improvement [C] / / Zhejiang Electric Power Society. Zhejiang Electric Power Society 2022 Annual Excellent Paper Collection. China Electric Power Press, 2023: 10, considering both the elasticity and economy of the distribution network, proposed a joint optimization configuration method of energy storage and controllable DG, aiming to maximize the elasticity of the distribution network while taking into account the economy. Lou Ping et al., Research on Optimal Configuration of Energy Storage Considering the Comprehensive Elasticity and Economy of Distribution Networks [J]. Journal of Wuhan University (Engineering Edition), 2022, 55(09): 901-909, proposed a comprehensive elasticity evaluation method for distribution networks based on subjective and objective weights, and constructed an optimal configuration model for energy storage with the optimal economy and comprehensive elasticity as the objective function. Ma Liye et al. Enhanced distribution network resilience and flexibility resource planning under typhoon disasters taking into account the influence of correlation [J]. Automation of Electric Power Systems, 2022, 46(07): 60-68. The PageRank algorithm was used to construct a distribution network failure rate model under typhoon disasters taking into account the influence of correlation, and a resilience quantitative index reflecting the influence of single load and flexibility resources on grid resilience was proposed. Through coordinated planning, the distribution network achieved optimal economic performance while improving resilience. Li Jitong et al. Multi-layer planning method for energy storage power stations in elastic distribution networks taking into account the spatiotemporal correlation of severe weather [J]. Power System Protection and Control, 2023, 51(09): 128-137. A joint disaster intensity probability distribution model considering the correlation between rainfall intensity and wind speed under severe weather with heavy rain and strong wind was established, and a multi-layer planning strategy for energy storage power stations to enhance the operational flexibility of distribution networks under compound disaster weather was proposed.

[0007] The above studies did not consider the division of demand-side resources in the distribution network based on elastic demand. If the demand-side resources in the distribution network are divided based on elasticity and cluster division, and the site selection and capacity determination of distributed energy storage are carried out on this basis, it will be of great significance to promote the overall regulation capacity of the distribution network and improve the network loss of the distribution network.

[0008] How to achieve distributed energy storage configuration based on elasticity and cluster division has become a technical problem that needs to be solved. Summary of the invention

[0009] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a distributed energy storage configuration method and device for a distribution network taking comprehensive elasticity into consideration.

[0010] The purpose of the present invention can be achieved by the following technical solutions:

[0011] According to one aspect of the present invention, a method for configuring distributed energy storage in a distribution network considering comprehensive elasticity is provided, the method comprising the following steps:

[0012] Step S1, constructing a comprehensive elasticity index system, including elasticity index, new energy matching index, modularity index and voltage qualification index;

[0013] Step S2, clustering the distribution network based on the comprehensive elasticity index system;

[0014] In step S3, a multi-objective particle swarm algorithm is used to solve the location and capacity determination model of energy storage within the cluster to obtain the distributed energy storage configuration plan for each cluster of the distribution network, wherein the multi-objectives of the location and capacity determination model include minimum cost target, maximum voltage stability and optimal self-healing rate index.

[0015] Preferably, the elasticity index is a self-healing rate index, specifically

[0016]

[0017] In the formula, τ 1 , τ 2 and τ 3 are the weighting coefficients of primary load, secondary load and tertiary load respectively; T f Indicates the fault duration, P 1,t , P 2,t and P 3,t They represent the actual restored power of the primary load, secondary load and tertiary load after the fault is restored in the distribution network. and They represent the primary load, secondary load and tertiary load before the fault occurs, and t represents the time when the fault occurs.

[0018] Preferably, the new energy matching index is specifically:

[0019]

[0020] In the formula, N represents the number of clusters, T represents the scheduling period, represents the load in the ith population at time t, Represents the new energy output within the i-th population at time t.

[0021] Preferably, the modularity index is specifically:

[0022]

[0023] In the formula, A represents the edge weight matrix of the network, A i,j is the edge weight between node i and node j; k i is the sum of all edge weights of node i; m is the sum of all edge weights in the entire distribution network; if node i and node j are in the same cluster, then δ(i,j) = 1, and if they are not in the same cluster, then 0.

[0024] Preferably, the voltage qualification index is specifically:

[0025]

[0026] In the formula, V , V are the upper and lower operating voltage bus limits and operating voltage value of the bus under assessment respectively, and T is the dispatching period.

[0027] Preferably, the process of dividing the distribution network into clusters includes:

[0028] Step S2-1, inputting the relevant parameters of the distribution network and the distributed power source, calculating the sensitivity matrix of voltage to reactive power and the electrical distance between nodes;

[0029] Step S2-2, based on the electrical distance between nodes, the distribution network is initially clustered using the kmeans algorithm;

[0030] Step S2-3, based on the comprehensive elasticity index, use the particle swarm algorithm to further cluster the distribution network and output the optimal clustering result.

[0031] More preferably, the electrical distance between nodes is calculated as follows:

[0032] When there are n nodes in the distribution network, the electrical distance between node i and node j is expressed as:

[0033]

[0034] Where D ij is the electrical distance influencing factor;

[0035] The calculation of the voltage to reactive power sensitivity matrix is ​​specifically as follows:

[0036] ΔV i =ΔQ j S VQ

[0037] In the formula, S VQ is the sensitivity matrix of voltage to reactive power, ΔQ j and ΔV represent the reactive power change at node j and the voltage amplitude at node i, respectively.

[0038] More preferably, the use of the particle swarm algorithm to further cluster the distribution network includes: using the initial cluster division result as the initial particle position and initializing the particle velocity, calculating the objective function of each particle according to the comprehensive elasticity index, iteratively calculating, updating the individual optimal value and the group maximum value until the convergence criteria are met, and outputting the optimal cluster division result.

[0039] Preferably, step S3 specifically comprises: for each cluster divided into clusters, a multi-objective particle swarm algorithm is used to perform multi-objective solution on the site selection and sizing model, and a fuzzy multi-attribute decision method is used to select the best compromise solution as the distributed energy storage configuration scheme for each cluster of the distribution network.

[0040] According to another aspect of the present invention, an electronic device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the method described above is implemented when the processor executes the program.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] 1) The present invention aims to optimize the balanced comprehensive elasticity index system including elasticity, clusters the distribution network to obtain the optimal cluster, adopts a multi-objective particle swarm algorithm to solve the multi-objective site selection and sizing model, and configures distributed energy storage clusters based on the load and new energy output characteristics of each cluster, with the goal of maximizing the conventional energy storage income and minimizing the cost recovery period. Compared with non-clustered energy storage configuration, the overall self-healing rate and economy of the distribution network are significantly improved.

[0043] 2) The present invention uses the kmeans algorithm to classify the parts with denser comprehensive elasticity coefficients into the same cluster, so as to efficiently formulate energy storage control strategies. The particle swarm algorithm is combined with the comprehensive elasticity index to iteratively calculate and output the optimal cluster division result. The indicators between the clusters are relatively balanced, and the network structure and electrical coupling perform well. The power quality is high, which is conducive to the stable operation of the distribution network.

[0044] 3) Energy storage is optimized and configured based on cluster division, with low cost, significantly higher load self-healing rate than the non-clustered solution, and significantly improved resource utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A schematic diagram of a cluster division process of a distribution network in the present invention;

[0046] Figure 2 It is a schematic diagram of wind and solar power output and daily load curve in the present invention;

[0047] Figure 3 This is a schematic diagram of the initial cluster structure of the distribution network in the present invention;

[0048] Figure 4 This is a schematic diagram of the final cluster structure of the distribution network in the present invention;

[0049] Figure 5 It is a schematic diagram of the fitness curve of the cluster algorithm in the present invention;

[0050] Figure 6 This is a schematic diagram of distributed energy and load data in the present invention;

[0051] Figure 7 It is a schematic diagram of the power purchase curve and time-of-use electricity price of the distribution network in the present invention;

[0052] Figure 8 It is a schematic diagram of the flow of the distributed energy storage configuration method of the distribution network in the present invention. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0054] This embodiment relates to a method for configuring distributed energy storage in a distribution network taking into account comprehensive elasticity. Figure 8 , including the following:

[0055] Step S1, establish a comprehensive elasticity index system, combine elasticity index, new energy matching index, modularity index, and voltage qualification index to establish a comprehensive elasticity index. Based on the established index as a reference, consider the demand side resources to cluster the entire distribution network.

[0056] Step S2, cluster the distribution network. The present invention sets the initial particles taking into account the complex network structure of the distribution network and the fact that the change of node combination in the newly generated partitioning result in each iteration will cause the change of the whole cluster performance. First, kmeans is used to perform preliminary clustering on the distribution network based on the modularity index. On this basis, the particle swarm algorithm is used to further partition the whole distribution network according to the comprehensive elasticity index, and the optimal clustering result is output. The whole process is as follows: Figure 1 .

[0057] Step S3, solving the site selection and sizing model of cluster energy storage: according to the cluster results, the site selection and sizing of energy storage is carried out for each cluster, and the site selection and sizing model of distributed cluster energy storage is solved by using a multi-objective particle swarm algorithm according to the cost target, maximum voltage stability and self-healing rate indicators. For the Pareto solution set obtained by solving the multi-objective problem, the fuzzy multi-attribute decision-making method is used to select the best compromise solution.

[0058] The distribution network cluster division and distributed energy storage configuration method designed by the present invention proposes a comprehensive elastic cluster index system that considers elasticity, structurality, and processing characteristics based on the traditional distribution network division based on the "modularity" index. Then, based on the load of each cluster and the output characteristics of new energy sources, the energy storage cluster is configured with the goal of maximizing the conventional benefits of energy storage and minimizing the cost recovery period. Finally, the superiority of the comprehensive elastic division method and distributed energy storage cluster configuration method proposed in this paper is demonstrated through simulation examples.

[0059] The present invention is further explained below from several aspects such as model establishment, design principle, design method, and effectiveness verification.

[0060] (1) Comprehensive elasticity index

[0061] 1) Self-healing rate index

[0062]

[0063] In the above formula, τ 1 , τ 2 and τ 3 are the weighting coefficients of primary load, secondary load and tertiary load respectively; T f Indicates the fault duration, P 1,t , P 2,t , P 3,t It indicates the actual restored power of the primary load, secondary load and tertiary load after the fault is restored in the distribution network. and They represent the primary load, secondary load and tertiary load before the fault occurs, and t represents the time when the fault occurs.

[0064] 2) New energy matching indicators

[0065]

[0066] In the above formula, N represents the number of clusters, T represents the scheduling period, represents the load in the ith population at time t, Represents the new energy output within the i-th population at time t.

[0067] 3) Modularity index

[0068]

[0069] In the above formula, A represents the edge weight matrix of the network. i,j is the edge weight between node i and node j, k iis the sum of all edge weights of node i, m is the sum of all edge weights in the entire distribution network, if node i and node j are in the same cluster, then δ(i,j) = 1, when they are not in the same cluster, it is 0.

[0070] ΔV i =ΔQ j S VQ

[0071] In the formula, S VQ is the sensitivity matrix of voltage to reactive power, generally a constant matrix, ΔQ j and ΔV represent the reactive power change at node j and the voltage amplitude at node i, respectively.

[0072] The node electrical distance impact factor is defined as:

[0073]

[0074] Where D ij is the electrical distance impact factor. The main significance of the electrical distance impact factor is that it indicates how external nodes affect a given node. The higher the value, the smaller the impact between nodes, and the greater the electrical distance between nodes. When there are n nodes in the network, the electrical distance between node i and node j is expressed as:

[0075]

[0076] Edge weight A i,j as follows:

[0077]

[0078] Where max(A) represents the maximum edge weight in the entire cluster.

[0079] 4) Voltage qualification index

[0080]

[0081] Where: V and V are the upper and lower operating voltage bus limits and operating voltage value of the bus under assessment respectively.

[0082] Combining the above four indicators, the comprehensive elasticity index is expressed as:

[0083] f a =ω 1 f 11 +ω 2 f 21 +ω 3 f 31 +ω 4 f41

[0084] In the above formula, ω 1 ,ω 2 ,ω 3 and ω 4 is a weighted coefficient, and the sum is 1. In the process of clustering, the parts with denser comprehensive elasticity coefficients are classified into the same cluster. This cluster can better detect the operation status according to the comprehensive elasticity coefficient and arrange the energy storage control strategy more effectively.

[0085] (2) Cluster content location selection and capacity determination model

[0086] 1) Cost target

[0087] The optimization objectives are cluster energy storage costs, distribution network operation and maintenance costs, network loss costs, and upstream power purchase costs within the cycle. The formula is as follows:

[0088] F=F 1 +F 2 +F 3 +F 4

[0089]

[0090] Among them, F 1 is the cluster energy storage cost during the period, F 2 is the distribution network operation and maintenance cost within the period, F 3 is the distribution network loss cost within the period, F 4 is the cost of electricity purchased from the distribution network during the period. m p is the unit price of energy storage power; m B is the unit price of energy storage capacity; m y P is the maintenance and operation cost of unit energy storage in the whole life cycle; n,j is the rated power of the energy storage of the i-th cluster; S n,i is the rated capacity of the energy storage of the i-th cluster; 1 and λ 2 are the annual maintenance coefficients of distributed power generation and energy storage system respectively; C DG is the initial investment cost of distributed power generation; C BSS is the initial investment cost of the energy storage system; E X represents the equipment replacement cost; Ω is the set of all branches of the distribution network; λ 3 is the unit network loss cost; I ij is the current of the branch at time; r ij is the resistance of branch ij; P t is the amount of electricity purchased by the distribution network from the upper main grid at time t, 4 is the upward electricity purchase cost coefficient.

[0091] 2) Maximum voltage stability

[0092]

[0093] Where: N L is the number of nodes where the load is located; U i,s is the actual voltage of node i in the sth scenario; U i* Assume that U i* =1; ΔU i,max is the maximum allowable deviation voltage value of node i, that is, the difference between the maximum voltage and the minimum voltage within the allowable range.

[0094] ΔU i,max =U i,max -U i,min

[0095] 3) Elasticity index

[0096] The present invention uses the self-healing rate to represent the elasticity index, which is expressed as ensuring that the maximum amount of important loads remain in a working state after the distribution network is disturbed by the uncertainty of wind and solar power during the operation cycle. The formula is as follows:

[0097] minE=1-f 11

[0098] (3) Constraints

[0099] The energy storage configuration of distribution network clusters needs to consider the network constraints of the distribution network (i.e., node flow power constraints, node voltage constraints, and the physical characteristics constraints of the energy storage itself, i.e., the charging and discharging power constraints, operating status constraints, and capacity constraints of the energy storage).

[0100] 1) Node voltage constraints:

[0101] V i,min ≤V i,t ≤V i,max

[0102] Where: V i,max and V i,min are the upper and lower limits of the voltage amplitude at node i respectively.

[0103] 2) Node power constraints:

[0104]

[0105] θ ij =θ i -θ j

[0106] Where: P i and Q i are the active power and reactive power injected into node i, Vi 、V j are the voltage amplitudes of nodes i and j, G ij , B ij are the conductance and susceptance of the path ij respectively; θ i and θ j are the voltage phase angles at nodes i and j respectively.

[0107] 3) Energy storage operation constraints

[0108]

[0109] Where: are the maximum values ​​of the charging and discharging power of the ESS in the i-th cluster, P b,i is the exchange power between ESS and distribution network at node i at time t; SOC i,min and SOC i,max are the upper and lower limits of the state of charge of the ESS at node i, SOC i (t) is the charge state of ESS at time t; Q i,max is the maximum value of the energy storage power exchange of the i-th cluster in one day; η is the performance factor, P b,in is the rated power of the energy storage of the i-th cluster; S n,i is the rated capacity of the energy storage of the i-th cluster; e i,t+Δt 、e i,t are the capacity of ESS at node i at time t and t+Δt respectively; SOC i,0 , SOC i,T are the charge states of ESS at node i at time 0 and T respectively.

[0110] 4) Line power constraints

[0111]

[0112] Where: I ij,t ,I ij,max are the current flowing through branch ij at time t and the maximum current allowed in branch ij.

[0113] 5) Power reverse constraints

[0114] Considering that excessive power transmission between clusters will increase network losses, energy storage within the cluster is required to help consume new energy locally, avoid the distribution network sending power back to the superior power grid and exchange power between clusters.

[0115] 0≤P l n (t)

[0116] Where: P l n(t) is the net load of the nth cluster of the distribution network at time t, which is equal to the load power minus the sum of the output power of new energy and energy storage.

[0117] 5) Branch capacity constraints

[0118] S ij ≤S ij,max

[0119] Where: S ij represents the transmission power of the branch between node i and node j, S ij,max It represents the maximum transmission power allowed by the branch between node i and node j.

[0120] 6) Active power constraints injected into the upper grid

[0121]

[0122] Where: are the active and reactive power injected into the distribution network by the upper power grid at time t; ΔP grid,max It is the maximum climbing power of the upper power grid.

[0123] Example 2

[0124] This embodiment also relates to a cluster division method of a distribution network, such as Figure 1 , including the following steps:

[0125] Step S2-1, inputting the relevant parameters of the distribution network and the distributed generation, and calculating the voltage sensitivity matrix and the electrical distance between nodes;

[0126] Step S2-2, using the kmeans algorithm to perform initial clustering on the distribution network based on electrical distance;

[0127] Step S2-3, use the particle swarm algorithm to further cluster the distribution network: the initial cluster division result is used as the initial particle position, and the particle velocity is initialized. The objective function of each particle is calculated according to the comprehensive elasticity index, and the individual optimal value and the group maximum value are updated iteratively until the convergence criteria are met, and the optimal cluster division result is output.

[0128] Example 3

[0129] This embodiment also relates to the verification of a distributed energy storage configuration method for a distribution network. In order to verify the effectiveness of the distributed energy storage planning and configuration model proposed in the present invention, two different schemes are set for comparison.

[0130] The applicability and effectiveness of the cluster division method were verified through the IEEE33-node test system. The system consists of 33 nodes, and the maximum system load is 9.285MW. The main load is divided into nodes 6, 23, 24, 29, 30, and 31. The second-level load nodes are nodes 1, 2, 3, 13, 17, 18, 19, 21, 22, and 28. The remaining nodes are classified as third-level loads. The weight coefficients of the main load, secondary load, and secondary load are 0.6, 0.3, and 0.1, respectively. The rated capacity of wind turbines at nodes 5, 12, 16, and 31 is 300kw, and the rated capacity of photovoltaics at nodes 7, 19, 23, and 27 is 600kW. The wind and solar output and daily load data use the historical data of a local distribution network. After correlation processing, the data of the 24h sampling point are obtained. The distributed energy output and daily load data in the distribution network are as follows: Figure 2 shown.

[0131] Solution 1: Instead of using the cluster partitioning model proposed in the present invention, the distributed energy storage is configured using the model proposed in this article.

[0132] Solution 2: Using the cluster partitioning model based on comprehensive elasticity index proposed in the present invention, the distributed energy storage is configured using the model proposed in this article without considering the elasticity index.

[0133] Solution 3: Using the cluster partition model based on comprehensive elasticity indicators proposed in the present invention, the distributed energy storage is configured using the model proposed in the present invention.

[0134] According to the typical daily load of the distribution network and the sequential output of each unit, the initial clustering is first performed according to the expected electrical distance. After the initial clustering is obtained, the distribution network is divided into 5 clusters by considering the comprehensive elasticity index. The particle swarm algorithm is used for solution, with the population set to 50, the number of iterations to 100, the initial value of the inertia weight to 0.9, the end value of the inertia weight to 0.4, the mutation rate to 0.5, and the crossover rate to 0.1.

[0135] The results of the initial cluster and the final cluster are as follows: Figure 3 and Figure 4 As shown, the comprehensive elasticity index parameters of the two cluster results are shown in Table 1 and Table 2 respectively.

[0136] Table 1

[0137]

[0138] Table 2

[0139]

[0140] According to the analysis of the results, we can see that kmeans is first used to perform preliminary clustering of the distribution network based on the modularity index, and then the particle swarm algorithm is used to perform optimal clustering. There are no isolated nodes in each cluster, and there are no overlapping nodes between clusters, which conforms to the principle of clustering. Among the several clusters divided by kmeans based on the modularity index, only the modularity index f 31 Maintaining a high level, the self-healing rate index f of cluster 2 and cluster 4 11 Obviously lower, voltage qualified index f 41 Clusters 1 to 4 all have relatively low indicators. 21 Cluster 1 and cluster 2 also show low performance. From the above results, it can be seen that using only kmeans based on modularity index division cannot balance various indicators, and some clusters have low indicators. The complementarity is poor, and the advantages of clusters cannot be fully utilized. Distributed energy storage configuration on this basis is not conducive to the stable operation of the distribution network.

[0141] The fitness curve of the particle swarm algorithm is as follows: Figure 5 As shown in Figure 2, it can be seen that using the particle swarm algorithm to divide the initial clusters obtained by kmeans can improve the convergence speed and cluster efficiency of the algorithm, and the fitness value is also high. The optimal cluster obtained does not have isolated nodes in each cluster, and there are no overlapping nodes between clusters, which conforms to the logical principle of clustering. As shown in Table 2, the indicators between different clusters are relatively balanced, and the modularity index f 31 And voltage qualified index f 41 Maintaining a high level, it shows that the division results perform well in terms of network structure and electrical coupling, and the power quality is high. The comprehensive division index is 23.1% higher than that of the kmeans method alone, and there are significant improvements in the comprehensive elasticity evaluation index, new energy matching index and voltage qualification index.

[0142] In summary, the method proposed in the present invention has stronger global search capability, effectively avoids local optimum, has faster iteration speed and higher clustering efficiency.

[0143] Energy storage optimization configuration is performed based on the obtained cluster, and the results of energy storage optimization configuration are shown in Table 3.

[0144] Table 3

[0145]

[0146] As shown in Table 3, the cost of Scheme 1, which does not consider the elasticity index for the energy storage site selection and sizing of the distribution network, is lower than that of Scheme 2 and Scheme 3, and the cost of Scheme 2 is slightly higher than that of Scheme 3. Because Scheme 2 does not select and size the self-healing rate of the overall load, its self-healing rate index is lower than that of the distribution network after clustering in Scheme 3, but it is significantly higher than the load self-healing rate after energy storage site selection and sizing in Scheme 1. In addition, in the case of unstable wind and solar units and insufficient output of the upper power grid, the normal working rate of the load (corresponding to the self-healing rate) of Scheme 3 is significantly higher than that of Scheme 1 without clustering, because in the case of unstable wind and solar output, the storage energy at the randomly assigned location cannot supply the specific load, and the energy storage site selection and sizing is carried out according to the cluster results. The configured nodes are basically located close to the distributed energy in the cluster, which can further improve the network loss of the distribution network compared with Scheme 1, and the utilization rate of flexible resources can also be improved.

[0147] In the case of unstable wind and solar power output, the wind and solar power output may drop significantly before and after. At the same time, the insufficient regulation capacity of the distribution network will result in the inability to supply all node loads in the distribution network at this moment, resulting in a decrease in the normal working rate of the load. The impact of energy storage units on the self-healing rate of the distribution network is further analyzed by studying the hourly output of the energy storage unit cycle and the amount of electricity purchased by the distribution network.

[0148] Scheme 3: The distributed energy and load data of the distribution network, the power purchase curve of the distribution network and the time-of-use electricity price are as follows: Figure 6 and Figure 7 As shown in the figure, without considering the output of the energy storage unit, it is obvious that the wind and solar power cannot be fully absorbed at 5:00-6:00 and 10:00-17:00. It can be seen that after adding the energy storage unit, the overall wind and solar power absorption is helped in these time periods through charging and discharging. At 2:00-4:00 and 23:00-24:00, the photovoltaic unit does not work and the load demand and wind turbine output vary greatly, and the distribution network has limited regulation capacity. At this time, the energy storage unit adjusts the output to supply part of the load to work normally. At 8:00-10:00, the output of the photovoltaic unit increases significantly, the output of the wind turbine is low, and the distribution network purchases electricity upward due to slope restrictions, resulting in the inability to supply the load in the distribution network. The energy storage sends electricity at this time to help restore the normal operation of the load. By analyzing the output of energy storage units from 10:00 to 16:00, it is found that when meeting the elastic demand, a large amount of electricity is purchased for storage during the period of low electricity prices, and the purchase of electricity is reduced accordingly during the period of relatively high electricity prices, and the stored electricity is released at the same time, thereby reducing the cost of purchasing electricity. The energy storage site selection and capacity determination strategy proposed in this paper can improve the overall self-healing rate of the distribution network while maintaining a certain degree of economy.

[0149] Example 4

[0150] The electronic device of the present invention includes a central processing unit (CPU), which can perform various appropriate actions and processes according to computer program instructions stored in a read-only memory (ROM) or loaded from a storage unit into a random access memory (RAM). In the RAM, various programs and data required for device operation can also be stored. The CPU, ROM and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0151] Multiple components in the device are connected to the I / O interface, including: input units, such as keyboards, mice, etc.; output units, such as various types of displays, speakers, etc.; storage units, such as disks, optical disks, etc.; and communication units, such as network cards, modems, wireless communication transceivers, etc. The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunication networks.

[0152] The processing unit performs the various methods and processes described above, such as methods S1 to S3. For example, in some embodiments, methods S1 to S3 may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and / or installed on the device via a ROM and / or a communication unit. When the computer program is loaded into the RAM and executed by the CPU, one or more steps of methods S1 to S3 described above may be performed. Alternatively, in other embodiments, the CPU may be configured to execute methods S1 to S3 in any other appropriate manner (e.g., by means of firmware).

[0153] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0154] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0155] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0156] The above is only a specific embodiment 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 replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for configuring distributed energy storage in a distribution network considering comprehensive elasticity, characterized in that: The method comprises the following steps: Step S1, constructing a comprehensive elasticity index system, including elasticity index, new energy matching index, modularity index and voltage qualification index; Step S2, clustering the distribution network based on the comprehensive elasticity index system; In step S3, a multi-objective particle swarm algorithm is used to solve the location and capacity determination model of energy storage within the cluster to obtain the distributed energy storage configuration plan for each cluster of the distribution network, wherein the multi-objectives of the location and capacity determination model include minimum cost target, maximum voltage stability and optimal self-healing rate index.

2. A method for configuring distributed energy storage in a distribution network considering comprehensive elasticity according to claim 1, characterized in that: The elasticity index is a self-healing rate index, specifically: Where τ1, τ2 and τ3 are the weighting coefficients of primary load, secondary load and tertiary load respectively; T f Indicates the fault duration, P 1,t , P 2,t and P 3,t They represent the actual restored power of the primary load, secondary load and tertiary load after the fault is restored in the distribution network. and They represent the primary load, secondary load and tertiary load before the fault occurs, and t represents the time when the fault occurs.

3. A method for configuring distributed energy storage in a distribution network considering comprehensive elasticity according to claim 1, characterized in that: The new energy matching index is specifically: In the formula, N represents the number of clusters, T represents the scheduling period, represents the load in the ith population at time t, Represents the new energy output within the i-th population at time t.

4. A method for configuring distributed energy storage in a distribution network considering comprehensive elasticity according to claim 1, characterized in that: The modularity index is specifically: In the formula, A represents the edge weight matrix of the network, A i,j is the edge weight between node i and node j; k i is the sum of all edge weights of node i; m is the sum of all edge weights in the entire distribution network; If node i and node j are in the same cluster, then δ(i,j)=1, and if they are not in the same cluster, then δ(i,j)=0.

5. A method for configuring distributed energy storage in a distribution network considering comprehensive elasticity according to claim 1, characterized in that: The voltage qualification index is specifically: In the formula, V , V are the upper and lower operating voltage bus limits and operating voltage value of the bus under assessment respectively, and T is the dispatching period.

6. A method for configuring distributed energy storage in a distribution network considering comprehensive elasticity according to claim 1, characterized in that: The process of clustering the distribution network includes: Step S2-1, inputting the relevant parameters of the distribution network and the distributed power source, calculating the sensitivity matrix of voltage to reactive power and the electrical distance between nodes; Step S2-2, based on the electrical distance between nodes, the distribution network is initially clustered using the kmeans algorithm; Step S2-3, based on the comprehensive elasticity index, use the particle swarm algorithm to further cluster the distribution network and output the optimal clustering result.

7. A method for configuring distributed energy storage in a distribution network considering comprehensive elasticity according to claim 6, characterized in that: The calculation of the electrical distance between the nodes is specifically as follows: When there are n nodes in the distribution network, the electrical distance between node i and node j is expressed as: Where D ij is the electrical distance influencing factor; The calculation of the voltage to reactive power sensitivity matrix is ​​specifically as follows: ΔV i =ΔQ j S VQ In the formula, S VQ is the sensitivity matrix of voltage to reactive power, ΔQ j and ΔV represent the reactive power change at node j and the voltage amplitude at node i, respectively.

8. A method for configuring distributed energy storage in a distribution network considering comprehensive elasticity according to claim 6, characterized in that: The method of further clustering the distribution network using the particle swarm algorithm includes: taking the initial clustering result as the initial particle position and initializing the particle velocity, calculating the objective function of each particle according to the comprehensive elasticity index, iteratively calculating, updating the individual optimal value and the group optimal value until the convergence criteria are met, and outputting the optimal clustering result.

9. A method for configuring distributed energy storage in a distribution network considering comprehensive elasticity according to claim 1, characterized in that: The step S3 is specifically as follows: for each cluster divided into clusters, a multi-objective particle swarm algorithm is used to perform multi-objective solutions to the site selection and sizing model, and a fuzzy multi-attribute decision method is used to select the best compromise solution as the distributed energy storage configuration plan for each cluster of the distribution network.

10. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the program, the method according to any one of claims 1 to 9 is implemented.

Images