A power distribution system-oriented energy storage planning method, device, equipment and storage medium

Abstract

The application discloses a power distribution system-oriented energy storage planning method and device, equipment and a storage medium, applied to the technical field of power grid control, and the method comprises the following steps: acquiring source and load historical operation data and a plurality of power distribution nodes; sequentially processing the source and load historical operation data to obtain power curves; determining electrical distances between the plurality of power distribution nodes, source and load complementarity and node reliability similarity based on the power curves; processing the electrical distances, source and load complementarity and node reliability similarity to obtain node comprehensive similarity; constructing a modularity function for the node comprehensive similarity; performing modularity optimization processing on the modularity function to obtain maximum modularity, and determining a source and load cluster division scheme; and inputting the source and load cluster division scheme into an energy storage configuration optimization double-layer model for processing to obtain an energy storage planning strategy. The method provided by the application can solve the problem that a planning scheme is disengaged from actual operation conditions due to the neglect of source and load double uncertainty.

Description

Technical Field

[0001] This invention relates to the field of power grid control technology, and in particular to an energy storage planning method, device, equipment and storage medium for power distribution systems. Background Technology

[0002] The new power distribution system differs significantly from the traditional system in its three components: source, grid, and load. On the source side, the capacity of distributed energy sources is continuously increasing, making local consumption increasingly challenging. On the grid side, the integration of energy storage devices and new power electronic devices is making control strategies increasingly complex. On the load side, the increase in new loads and the development of electricity substitution technologies are making load characteristics more complex, involving source-load interactions and increasing uncertainties. Energy storage systems, with their flexible adjustment capabilities, can effectively mitigate the randomness and volatility problems caused by the dual uncertainties of source and load.

[0003] In existing technologies, energy storage planning methods generally allocate energy storage capacity at a fixed multiple based on the peak characteristics of historical load and renewable energy output, and determine the access point based on the static investment return cycle. However, such methods seriously ignore the spatiotemporal coupling relationship between the dual uncertainties of source and load, which often leads to overly rigid planning schemes. In actual operation, it is very easy to have alternating phenomena of insufficient energy storage or idle energy storage, which will in turn cause the distribution network to suffer from increased reverse power flow impact, increased risk of operating voltage exceeding limits, and reduced asset utilization efficiency of the entire system. Summary of the Invention

[0004] This invention provides an energy storage planning method, device, equipment, and storage medium for power distribution systems, in order to solve the technical problem that traditional energy storage planning leads to rigid planning schemes and disconnection from actual operating conditions due to neglecting the spatiotemporal coupling relationship between the dual uncertainties of source and load.

[0005] To address the aforementioned technical problems, this invention provides an energy storage planning method for power distribution systems, the method comprising: Acquire historical source-load operation data and several distribution nodes of the target power distribution system; The source-load historical operating data are sequentially processed by probabilistic power output modeling and three-point estimation to obtain the source-load probabilistic time-series power curve of the target power distribution system. Based on the source-load probability time-series power curves, the electrical distance, source-load complementarity, and node reliability similarity among several distribution nodes are determined; the entropy weight method is used to process the electrical distance, the source-load complementarity, and the node reliability similarity to obtain the comprehensive node similarity. A modularity function is constructed using the distribution node and the node comprehensive similarity; the modularity function is then optimized to obtain the maximum modularity of the target distribution system, and the maximum modularity is used to determine the source-load cluster partitioning scheme. The source-load cluster partitioning scheme is input into the constructed two-layer energy storage configuration optimization model for processing to obtain the energy storage planning strategy.

[0006] As one preferred embodiment, the process of sequentially performing probabilistic output modeling and three-point estimation on the historical operating data of the source loads to obtain the probabilistic time-series power curves of the target power distribution system includes: According to the power supply type and load type, the historical operating data of the power source and load are statistically analyzed and fitted to obtain the power supply probability distribution model and the load probability distribution model. Based on the power supply probability distribution model and the load probability distribution model, three estimation points for several distribution nodes are selected using the three-point estimation method, and the power value corresponding to each estimation point is calculated. By combining the power values ​​corresponding to all the estimated points, the source-load probability time-series power curve is obtained.

[0007] As one preferred embodiment, the construction of the modularity function using the distribution node and the node comprehensive similarity includes: A node similarity matrix is ​​constructed based on the node comprehensive similarity, and the elements of the node similarity matrix are the node comprehensive similarity between any two of the distribution nodes; The node degree of each distribution node is calculated using the node similarity matrix, where the node degree is the sum of the overall node similarity between the distribution node and all other distribution nodes. The modularity function is constructed based on the node similarity matrix and the node degree of all the distribution nodes.

[0008] As one preferred embodiment, the step of performing modularity optimization on the modularity function to obtain the maximum modularity of the target power distribution system includes: Each of the power distribution nodes is used as the initial cluster, and the current modularity value is calculated based on the modularity function of the power distribution node; Traverse all node merging operations. Each merging operation merges two of the aforementioned distribution nodes or existing clusters into a new cluster, and calculates the modularity value of the merged cluster based on the modularity function. If the merged modularity value is greater than the current modularity value, then the merge operation is executed and the merged new cluster is used as a new initial cluster to participate in subsequent merges; otherwise, the merge operation is abandoned. Repeat the above traversal and merging operations until no merging operation can be found that increases the modularity value. At this point, the corresponding modularity value is the maximum modularity value.

[0009] As one preferred embodiment, the process of determining the source-load cluster partitioning scheme includes: Obtain the cluster label of the power distribution node corresponding to the maximized modularity; Distribution nodes with the same cluster label are grouped into the same source-load cluster to generate the source-load cluster partitioning scheme.

[0010] As one preferred embodiment, the construction process of the energy storage configuration optimization two-layer model includes: With the minimum comprehensive annual cost as the objective function, and with the power self-balancing, power fluctuation and unit construction capacity multiple of each source-load cluster as constraints, an upper-level optimization model is constructed. With the objective function of minimizing annual network loss, and with Distflow power flow constraints, energy storage operation constraints, and system security constraints as constraints, a lower-level optimization model is constructed. The upper-level optimization model and the lower-level optimization model are combined to obtain the energy storage configuration optimization two-layer model.

[0011] As one preferred embodiment, the process of inputting the source-load cluster partitioning scheme into the constructed two-layer energy storage configuration optimization model for processing to obtain an energy storage planning strategy includes: The source-load cluster partitioning scheme and the source-load probabilistic time-series power curve are input into the upper optimization model of the energy storage configuration optimization two-layer model to solve for the energy storage candidate installation scheme, which includes the energy storage installation location, rated capacity and rated power. The energy storage candidate installation scheme is passed to the lower optimization model of the energy storage configuration optimization two-layer model, and the source load probability time series power curve is simulated to obtain the energy storage operation scheduling strategy. The energy storage operation scheduling strategy includes the energy storage charging and discharging power at several times. Determine whether the energy storage operation and scheduling strategy satisfies all the constraints of the lower-level optimization model. If yes, combine the energy storage candidate installation scheme and the energy storage operation and scheduling strategy to output the energy storage planning strategy. If no, adjust the energy storage candidate installation scheme and return to the lower-level optimization model to solve it again until all constraints are satisfied.

[0012] The present invention also provides an energy storage planning device for power distribution systems, comprising: The acquisition module is used to acquire historical operating data of the source loads and several distribution nodes of the target power distribution system; The processing module is used to sequentially perform probabilistic output modeling and three-point estimation processing on the historical operating data of the source load to obtain the source load probabilistic time-series power curve of the target power distribution system. The determination module is used to determine the electrical distance, source-load complementarity, and node reliability similarity among several distribution nodes based on the source-load probability time-series power curve; and to process the electrical distance, source-load complementarity, and node reliability similarity using the entropy weight method to obtain the comprehensive node similarity. A module is constructed to build a modularity function using the distribution node and the node comprehensive similarity; the modularity function is optimized to obtain the maximum modularity of the target distribution system, and the maximum modularity is used to determine the source-load cluster partitioning scheme. The planning module is used to input the source-load cluster partitioning scheme into the constructed two-layer energy storage configuration optimization model for processing, so as to obtain the energy storage planning strategy.

[0013] The present invention also provides an energy storage planning device for power distribution systems, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the energy storage planning method for power distribution systems as described above.

[0014] The present invention further provides a computer-readable storage medium storing a computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements the energy storage planning method for power distribution systems as described above.

[0015] Compared with the prior art, the beneficial effects of the present invention are at least one of the following: This invention acquires historical source-load operation data and several distribution nodes of a target power distribution system; performs probabilistic output modeling and three-point estimation on the historical source-load operation data to obtain the source-load probabilistic time-series power curve of the target power distribution system; based on the source-load probabilistic time-series power curve, determines the electrical distance, source-load complementarity, and node reliability similarity among the distribution nodes; processes the electrical distance, source-load complementarity, and node reliability similarity using the entropy weight method to obtain the node comprehensive similarity; constructs a modularity function using the distribution nodes and the node comprehensive similarity; optimizes the modularity function to obtain the maximum modularity of the target power distribution system, which is used to determine the source-load cluster partitioning scheme; inputs the source-load cluster partitioning scheme into a constructed two-layer energy storage configuration optimization model for processing to obtain an energy storage planning strategy.

[0016] Compared with existing technologies, the present invention transforms historical source-load data into quantified time-series curves through probabilistic output modeling and three-point estimation to accurately capture uncertainties. It then integrates electrical distance, source-load complementarity, and reliability similarity to construct a comprehensive evaluation system. Furthermore, it utilizes entropy weighting and modularity optimization algorithms to overcome the limitations of traditional single-point planning, achieving a scientific division of source-load clusters that considers spatiotemporal coupling. Finally, through iterative feedback of a two-layer optimization model, it ensures that energy storage configuration is deeply aligned with operational conditions at the planning level. This solves the problem of rigid schemes and operational disconnect caused by neglecting the spatiotemporal coupling of dual uncertainties between source and load in traditional methods, ensuring that the planning strategy is both robust and economical. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating an energy storage planning method for power distribution systems in one embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of an energy storage planning device for power distribution systems in one embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an energy storage planning device for power distribution systems in one embodiment of the present invention; Figure label: Among them, 11. Acquisition module; 12. Processing module; 13. Determination module; 14. Construction module; 15. Planning module; 21. Processor; 22. Memory. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] In the description of this invention, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0020] One embodiment of the present invention provides an energy storage planning method for power distribution systems. For details, please refer to [link to relevant documentation]. Figure 1 , Figure 1The diagram shown is a flowchart illustrating an energy storage planning method for power distribution systems according to one embodiment of the present invention. The method includes: S1: Obtain historical operating data of the source and load of the target power distribution system and several power distribution nodes; S2: Perform probabilistic power output modeling and three-point estimation processing on the historical operating data of the source load in sequence to obtain the probabilistic time-series power curve of the source load of the target power distribution system. S3: Based on the source-load probability time-series power curve, determine the electrical distance, source-load complementarity, and node reliability similarity among several distribution nodes; use the entropy weight method to process the electrical distance, source-load complementarity, and node reliability similarity to obtain the comprehensive node similarity. S4: Construct a modularity function using the distribution node and the node comprehensive similarity; perform modularity optimization processing on the modularity function to obtain the maximum modularity of the target distribution system, and the maximum modularity is used to determine the source-load cluster partitioning scheme; S5: Input the source-load cluster partitioning scheme into the constructed two-layer energy storage configuration optimization model for processing to obtain the energy storage planning strategy.

[0021] Historical operating data of power sources and loads includes two categories: output data of distributed photovoltaic power generation and energy consumption data of conventional loads. These are generally extracted from the SCADA system, electricity consumption information collection system, distributed energy management system, or historical database of the target power distribution system. Distribution nodes, on the other hand, are extracted from the geographic information system or power grid topology database of the target power distribution system and generally include information such as the distribution node number, type identifier, geographical coordinates, and voltage level.

[0022] In step S2, the historical operating data of the source and load are sequentially processed by probabilistic output modeling and three-point estimation to obtain the probabilistic time-series power curve of the target power distribution system. This includes: statistically analyzing the historical operating data of the source and load according to the power source type and load type, and fitting the power source probability distribution model and the load probability distribution model; based on the power source probability distribution model and the load probability distribution model, selecting three estimation points of several distribution nodes using the three-point estimation method, and calculating the power value corresponding to each estimation point; and combining the power values ​​corresponding to all the estimation points to obtain the probabilistic time-series power curve of the source and load.

[0023] For distributed photovoltaic (PV) output data, the Beta distribution is usually used for probability modeling. The Beta distribution can well describe the distribution pattern of PV output between 0 and rated capacity.

[0024] Based on the historical power output sequence of photovoltaic power at each distribution node, the two shape parameters α and β of the Beta distribution are calculated using the maximum likelihood estimation method, thereby obtaining the photovoltaic power output probability density function of the node at each time point.

[0025] For load data, considering that load fluctuations statistically approximate a normal distribution, a normal distribution is used for fitting. Similarly, based on the historical load data of each distribution node, the mean μ and standard deviation σ at each moment are calculated to obtain the normal distribution model of the load.

[0026] After completing the probability modeling, the continuous probability distribution is discretized into three typical estimation points using the three-point estimation method.

[0027] Specifically, for the probability distribution of each distribution node at each time point, three estimation points are selected: the first estimation point is the mean μ of the probability distribution, the second estimation point is the mean minus one standard deviation (μ - σ), and the third estimation point is the mean plus one standard deviation (μ + σ).

[0028] Based on the cumulative probability values ​​corresponding to these three estimated points, the probability distribution function is looked up to calculate the actual power value corresponding to each estimated point.

[0029] Finally, by arranging and combining the power values ​​corresponding to the three estimated points of all distribution nodes at all times in chronological and node order, a three-dimensional data structure is formed, namely the source-load probability time-series power curve.

[0030] This curve contains time-series power sets for multiple scenarios, with each scenario corresponding to a combination of three estimation points. It can comprehensively reflect the uncertainty of source load and is more efficient than directly using a large number of Monte Carlo simulation samples, making it easier to use for subsequent cluster partitioning and energy storage configuration optimization.

[0031] Based on the source-load probability time-series power curves, the electrical distance, source-load complementarity, and node reliability similarity among several distribution nodes are determined; the entropy weight method is used to process the electrical distance, the source-load complementarity, and the node reliability similarity to obtain the comprehensive node similarity.

[0032] Specifically, electrical distance is used to measure the tightness of electrical connection between two distribution nodes. Based on the node admittance matrix of the distribution system, the sensitivity coefficient of each node to voltage changes is calculated to obtain a sensitivity matrix. For any two nodes i and j, the electrical distance is calculated based on their Euclidean distance in the corresponding column of the sensitivity matrix.

[0033] Source-load complementarity is used to measure the complementary characteristics of two distribution nodes in terms of their power curves. Based on the source-load probabilistic time-series power curves, the net load curve of each distribution node is calculated, which is the load power of the node minus the photovoltaic output. For any two nodes i and j, the Pearson correlation coefficient of their net load curves is calculated, and then 1 is subtracted from this correlation coefficient to obtain the complementarity index.

[0034] Node reliability similarity is measured using the average service availability index (ASAI), which is the ratio of the total uninterrupted power supply time received by users in a year to the total power supply time requested by users. Alternatively, the average service availability index can be calculated based on the historical number of faults, outage time, and number of users for each distribution node.

[0035] Entropy weighting is an objective weighting method that does not require manually setting weights, but determines weights based on the information content of the data itself.

[0036] First, an evaluation matrix is ​​constructed based on the electrical distance, source-load complementarity, and node reliability similarity of all node pairs. Then, this matrix is ​​standardized to eliminate the influence of different dimensions. The entropy value of each indicator is calculated. The smaller the entropy value, the greater the difference between different node pairs, the more information it carries, and the greater the weight it should be assigned. The weight of each indicator is calculated based on the entropy value. Then, the three indicators are weighted and summed to obtain the comprehensive node similarity between each node pair.

[0037] Constructing a modularity function using the distribution nodes and their comprehensive similarity includes: constructing a node similarity matrix based on the comprehensive similarity, where each element of the node similarity matrix is ​​the comprehensive similarity between any two distribution nodes; calculating the node degree of each distribution node using the node similarity matrix, where the node degree is the sum of the comprehensive similarities between that distribution node and all other distribution nodes; and constructing the modularity function based on the node similarity matrix and the node degrees of all distribution nodes.

[0038] Specifically, each distribution node is regarded as a vertex in the network. The comprehensive similarity obtained by fusing any two different nodes through the entropy weight method is used as the weight of the edge between them. A symmetric similarity matrix with diagonal elements set to zero is constructed. The element in the i-th row and j-th column of the matrix is ​​equal to the comprehensive similarity value between node i and node j.

[0039] For each distribution node, sum all matrix elements in its row except itself. The sum is the node degree of that node. This node degree actually represents the overall affinity between that node and all other nodes in the network in terms of source-load characteristics.

[0040] The formula for constructing the modularity function is: in, It is a node and nodes Similarity between them It is a node The sum of similarities between a node and other nodes, i.e., the node's similarity to other nodes. degree of nodes, It is a node degree of nodes, It is the sum of the similarities between nodes in the network. It is a modularity function.

[0041] The modularity optimization process is performed on the modularity function to obtain the maximum modularity of the target power distribution system. This includes: treating each power distribution node as an initial cluster and calculating the current modularity value based on the modularity function of the power distribution node; traversing all node merging operations, where each merging operation merges two power distribution nodes or existing clusters into a new cluster, and calculating the merged modularity value based on the modularity function; if the merged modularity value is greater than the current modularity value, then the merging operation is executed, and the merged new cluster is used as a new initial cluster for subsequent merging; otherwise, the merging operation is abandoned; the above traversal and merging operations are repeated until no merging operation that increases the modularity value can be found, at which point the corresponding modularity value is the maximum modularity.

[0042] Specifically, in optimizing the modularity function to maximize modularity, a bottom-up agglomerative greedy search strategy is adopted.

[0043] At the start of the process, each distribution node in the target power distribution system is treated as an initial cluster, and the modularity value of the entire network under this initial partitioning state is calculated using the previously constructed modularity function, which serves as the starting benchmark for iterative optimization.

[0044] The program then enters a loop iteration phase: it attempts to merge all existing clusters pairwise. Each merge attempt recalculates the modularity function after updating the cluster affiliation and records the modularity increment from all possible merges. If, during a iteration, a merge operation is found to have increased the modularity compared to before the merge, the merge operation that yields the largest increase in modularity is executed. This merges the corresponding two nodes or clusters into a new cluster, replacing the two existing clusters and updating the entire network's cluster structure as the initial state for the next iteration. If a merge attempt results in a decrease in modularity instead of an increase, the merge attempt is abandoned, and the cluster structure remains unchanged.

[0045] The algorithm will continuously repeat the above steps. After each iteration, it will recalculate the modularity changes of all possible merges using the updated cluster structure until no merge operation can be found that increases the modularity value on the current basis.

[0046] At this point, the algorithm converges, and the network reaches the cluster partitioning form with the maximum modularity function value. This final value is the maximum modularity we are looking for, which corresponds to the clustering scheme that divides the source load nodes into the most cohesive groups.

[0047] In another embodiment, this process further includes: Step 1: Calculate the modularity Q1 of the current cluster partitioning.

[0048] Step 2: Randomly select two nodes in the network and assume they are merged. Update the complementarity and reliability similarity indices of the merged nodes, and calculate the modularity Q2 of the merged nodes.

[0049] Step 3: If so, then merge the two nodes, and the merged node will be considered as a new node.

[0050] Step 4: Repeat steps 2 and 3 until Q no longer increases, which is the optimal result.

[0051] The process of determining the source-load cluster partitioning scheme includes: obtaining the cluster label of the distribution node corresponding to the maximum modularity; grouping distribution nodes with the same cluster label into the same source-load cluster, and generating the source-load cluster partitioning scheme.

[0052] After maximizing the modularity value through agglomerative greedy search, the final cluster affiliation of each distribution node in the network is determined.

[0053] When the algorithm converges, each distribution node is explicitly labeled with a cluster tag, which directly identifies the cluster number to which it belongs in the optimal partitioning structure. This process is actually decoding the cluster partitioning state corresponding to maximizing modularity: grouping all distribution nodes carrying the same cluster tag together to form an independent source-load cluster.

[0054] After traversing all nodes and classifying them according to their labels, the originally scattered power distribution nodes are organized into several clusters with highly similar internal source-load characteristics and significant differences in characteristics among themselves. This combination is the source-load cluster partitioning scheme output in this solution.

[0055] The entire solution does not require manual pre-specification of the number of clusters; it is adaptively generated entirely by a data-driven modularity optimization process.

[0056] Next, the source-load cluster partitioning scheme is input into the constructed two-layer energy storage configuration optimization model for processing to obtain the energy storage planning strategy.

[0057] The construction process of the two-layer energy storage configuration optimization model includes: constructing an upper-layer optimization model with the minimum comprehensive annual cost as the objective function and the power self-balancing, power fluctuation, and unit construction capacity multiple of each source-load cluster as constraints; constructing a lower-layer optimization model with the minimum annual network loss as the objective function and the Distflow power flow constraints, energy storage operation constraints, and system security constraints as constraints; and combining the upper-layer optimization model and the lower-layer optimization model to obtain the two-layer energy storage configuration optimization model.

[0058] When constructing a two-layer model for optimizing energy storage configuration, planning decisions and operational simulations are decoupled and layered.

[0059] The upper-level model focuses on the capacity and power configuration of energy storage. The objective function is set to minimize the comprehensive equivalent annual cost. Here, the comprehensive equivalent annual cost is the total life cycle cost of energy storage equipment, including initial investment, replacement cost, operation and maintenance expenses, which is uniformly discounted into an equivalent annual cost value according to a preset discount rate and service life. This is used as the sole criterion for economic evaluation.

[0060] The constraints closely adhere to the characteristics of the pre-defined source-load clusters, requiring that the energy storage power and capacity configured in each cluster must ensure that the self-balancing rate of the time-series net load within the cluster reaches a specified limit, meaning that the degree of matching between renewable energy output and local load consumption cannot be too low. Simultaneously, the rate of change of the exchange power at the access point is limited, characterizing power fluctuation constraints as the absolute value of the difference in grid-connected power between adjacent time periods not exceeding a set threshold, thereby suppressing the impact on the upstream power grid. Furthermore, considering that energy storage in actual engineering often uses modular prefabricated modules, and the capacity must be an integer multiple of the capacity of a single device or standard module, a unit construction capacity multiple constraint is introduced, limiting the decision variable to discrete modular increments.

[0061] The lower-level model takes minimizing the annual grid loss as the operation optimization objective. After knowing the energy storage configuration scheme passed down from the upper level, it performs a detailed simulation of the power flow distribution in various typical scenarios throughout the year.

[0062] The constraints adopt the Distflow branch power flow equation applicable to radial distribution networks to ensure that the active and reactive power balance of each node and the voltage drop relationship along the feeder are strictly satisfied; the energy storage operation constraints cover requirements such as charging and discharging power not exceeding the limit, dynamic and continuous change of state of charge within the allowable range, and energy balance at the beginning and end of the dispatch cycle; the system safety constraints force the voltage amplitude of each node and the current of each branch to always be maintained within the qualified range specified in the standard.

[0063] The two-layer model forms a closed loop through the issuance of energy storage configuration schemes and feedback of network loss during operation. The upper layer generates alternative configurations and passes them to the lower layer. The lower layer solves the optimal operation strategy based on the configuration and returns the minimum network loss value and the constraint overrun situation. The upper layer adjusts the configuration accordingly until it finds a solution that minimizes the comprehensive annual cost and makes the lower layer's operation feasible. This iterative cooperation between the upper and lower layers forms a two-layer model for energy storage configuration optimization oriented towards source-load clusters.

[0064] The source-load cluster partitioning scheme is input into the constructed two-layer energy storage configuration optimization model for processing to obtain an energy storage planning strategy. This includes: inputting the source-load cluster partitioning scheme and the source-load probabilistic time-series power curve into the upper-layer optimization model of the two-layer energy storage configuration optimization model to solve for candidate energy storage installation schemes, which include energy storage installation location, rated capacity, and rated power; passing the candidate energy storage installation scheme to the lower-layer optimization model of the two-layer energy storage configuration optimization model and performing time-series simulation on the source-load probabilistic time-series power curve to obtain an energy storage operation scheduling strategy, which includes energy storage charging and discharging power at several times; determining whether the energy storage operation scheduling strategy satisfies all constraints of the lower-layer optimization model. If yes, the candidate energy storage installation scheme and the energy storage operation scheduling strategy are combined and output as the energy storage planning strategy; if not, the candidate energy storage installation scheme is adjusted and returned to the lower-layer optimization model for re-solution until all constraints are satisfied.

[0065] Specifically, the source-load cluster partitioning scheme and the source-load probabilistic time-series power curve are fed into the upper-level optimization model. The upper-level model uses the installation nodes, rated capacity, and rated power of energy storage in each cluster as decision variables. Under constraints such as cluster power self-balancing rate, power fluctuation limit, and modular capacity multiple, the model seeks to minimize the comprehensive equivalent annual cost and solves a set of specific energy storage candidate installation schemes.

[0066] This set of schemes was then fixed and passed to the lower-level optimization model. The lower level used the candidate scheme as the boundary condition to perform time-series simulation of the source-load probability time-series power curve at the operation level for each time period, and calculated the charging and discharging power command of the energy storage system at each moment in the entire scheduling cycle, forming a set of energy storage operation scheduling strategies that are compatible with the candidate installation schemes.

[0067] Next, the lower-level model will rigorously verify whether the node voltage, branch current, energy storage state of charge, and charging and discharging power under this operation and scheduling strategy fully meet the Distflow power flow constraints, energy storage operation constraints, and system safety constraints.

[0068] If all verification results are satisfactory, it means that the candidate installation scheme is feasible for operation throughout the year. At this point, the energy storage candidate installation scheme output from the upper layer is combined with the energy storage operation and scheduling strategy output from the lower layer and output together as the final energy storage planning strategy.

[0069] If the verification finds that constraints are exceeded in certain periods, the feedback information of the infeasibility is returned to the upper layer model. Based on this, the upper layer adjusts the energy storage configuration scheme during the optimization process, such as increasing or decreasing the energy storage capacity or power of certain clusters, or changing the installation nodes. After generating a new candidate scheme, it is sent down to the lower layer again for time-series simulation and constraint verification. This process is repeated until a configuration and scheduling scheme that satisfies both the economic indicators of the upper layer and ensures that all operational constraints of the lower layer are met is found. Only then does the iteration stop and the feasible scheme is determined as the final delivered energy storage planning strategy.

[0070] And control the power distribution system to execute the energy storage planning strategy.

[0071] Another embodiment of the present invention provides an energy storage planning device for power distribution systems. For details, please refer to [link to relevant documentation]. Figure 2 , Figure 2 The diagram shown is a structural schematic of an energy storage planning device for power distribution systems according to one embodiment of the present invention. The device includes: Module 11 is used to acquire historical source-load operation data and several distribution nodes of the target power distribution system; Processing module 12 is used to sequentially perform probabilistic output modeling and three-point estimation processing on the historical operating data of the source load to obtain the source load probabilistic time-series power curve of the target power distribution system. The determination module 13 is used to determine the electrical distance, source-load complementarity, and node reliability similarity among several distribution nodes based on the source-load probability time-series power curve; and to process the electrical distance, source-load complementarity, and node reliability similarity using the entropy weight method to obtain the node comprehensive similarity. Module 14 is used to construct a modularity function using the distribution node and the node comprehensive similarity; perform modularity optimization processing on the modularity function to obtain the maximum modularity of the target distribution system, and the maximum modularity is used to determine the source-load cluster partitioning scheme; The planning module 15 is used to input the source-load cluster partitioning scheme into the constructed energy storage configuration optimization two-layer model for processing to obtain the energy storage planning strategy.

[0072] See Figure 3 This is a schematic diagram of the structure of an energy storage planning device for power distribution systems provided in an embodiment of the present invention. The energy storage planning device for power distribution systems provided in this embodiment includes a processor 21, a memory 22, and a computer program stored in the memory 22 and configured to be executed by the processor 21. When the processor 21 executes the computer program, it implements the steps described in the above embodiment of the energy storage planning method for power distribution systems, for example... Figure 1The steps S1 to S5 described above; or, when the processor 21 executes the computer program, it implements the functions of each module in the above-described device embodiments, such as the acquisition module 11.

[0073] For example, the computer program can be divided into one or more modules, which are stored in the memory 22 and executed by the processor 21 to complete the present invention. The one or more modules can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the energy storage planning device for the power distribution system. For example, the computer program can be divided into an acquisition module 11, a processing module 12, a determination module 13, etc., with the specific functions of each module as follows: Module 11 is used to acquire historical source-load operation data and several distribution nodes of the target power distribution system; Processing module 12 is used to sequentially perform probabilistic output modeling and three-point estimation processing on the historical operating data of the source load to obtain the source load probabilistic time-series power curve of the target power distribution system. The determination module 13 is used to determine the electrical distance, source-load complementarity, and node reliability similarity among several distribution nodes based on the source-load probability time-series power curve; and to process the electrical distance, source-load complementarity, and node reliability similarity using the entropy weight method to obtain the node comprehensive similarity. Module 14 is used to construct a modularity function using the distribution node and the node comprehensive similarity; perform modularity optimization processing on the modularity function to obtain the maximum modularity of the target distribution system, and the maximum modularity is used to determine the source-load cluster partitioning scheme; The planning module 15 is used to input the source-load cluster partitioning scheme into the constructed energy storage configuration optimization two-layer model for processing to obtain the energy storage planning strategy.

[0074] The energy storage planning device for power distribution systems may include, but is not limited to, a processor 21 and a memory 22. Those skilled in the art will understand that the schematic diagram is merely an example of an energy storage planning device for power distribution systems and does not constitute a limitation on such a device. It may include more or fewer components than illustrated, or combine certain components, or use different components. For example, the energy storage planning device for power distribution systems may also include input / output devices, network access devices, buses, etc.

[0075] The processor 21 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor 21 is the control center of the energy storage planning equipment for the power distribution system, connecting various parts of the entire energy storage planning equipment for the power distribution system through various interfaces and lines.

[0076] The memory 22 can be used to store the computer programs and / or modules. The processor 21 implements various functions of the energy storage planning device for the power distribution system by running or executing the computer programs and / or modules stored in the memory 22 and calling the data stored in the memory 22. The memory 22 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 22 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0077] Wherein, if the modules integrated into the energy storage planning equipment for the power distribution system are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0078] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0079] Accordingly, embodiments of the present invention provide a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform steps in the energy storage planning method for power distribution systems as described in the above embodiments, for example... Figure 1 Steps S1 to S5 as described above.

[0080] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for energy storage planning in power distribution systems, characterized in that, include: Acquire historical source-load operation data and several distribution nodes of the target power distribution system; The source-load historical operating data are sequentially processed by probabilistic power output modeling and three-point estimation to obtain the source-load probabilistic time-series power curve of the target power distribution system. Based on the source-load probability time-series power curves, the electrical distance, source-load complementarity, and node reliability similarity among several distribution nodes are determined; the entropy weight method is used to process the electrical distance, the source-load complementarity, and the node reliability similarity to obtain the comprehensive node similarity. A modularity function is constructed using the power distribution node and the node comprehensive similarity. The modularity function is optimized to obtain the maximum modularity of the target power distribution system. The maximum modularity is used to determine the source-load cluster partitioning scheme. The source-load cluster partitioning scheme is input into the constructed two-layer energy storage configuration optimization model for processing to obtain the energy storage planning strategy.

2. The energy storage planning method for power distribution systems as described in claim 1, characterized in that, The process of sequentially performing probabilistic output modeling and three-point estimation on the historical operating data of the source loads yields the probabilistic time-series power curves of the target power distribution system, including: According to the power supply type and load type, the historical operating data of the power source and load are statistically analyzed and fitted to obtain the power supply probability distribution model and the load probability distribution model. Based on the power supply probability distribution model and the load probability distribution model, three estimation points for several distribution nodes are selected using the three-point estimation method, and the power value corresponding to each estimation point is calculated. By combining the power values ​​corresponding to all the estimated points, the source-load probability time-series power curve is obtained.

3. The energy storage planning method for power distribution systems as described in claim 1, characterized in that, The method of constructing a modularity function using the distribution node and the node comprehensive similarity includes: A node similarity matrix is ​​constructed based on the node comprehensive similarity, and the elements of the node similarity matrix are the node comprehensive similarity between any two of the distribution nodes; The node degree of each distribution node is calculated using the node similarity matrix, where the node degree is the sum of the overall node similarity between the distribution node and all other distribution nodes. The modularity function is constructed based on the node similarity matrix and the node degree of all the distribution nodes.

4. The energy storage planning method for power distribution systems as described in claim 1, characterized in that, The step of performing modularity optimization on the modularity function to obtain the maximum modularity of the target power distribution system includes: Each of the power distribution nodes is used as the initial cluster, and the current modularity value is calculated based on the modularity function of the power distribution node; Traverse all node merging operations. Each merging operation merges two of the aforementioned distribution nodes or existing clusters into a new cluster, and calculates the modularity value of the merged cluster based on the modularity function. If the merged modularity value is greater than the current modularity value, then the merge operation is executed and the merged new cluster is used as a new initial cluster to participate in subsequent merges; otherwise, the merge operation is abandoned. Repeat the above traversal and merging operations until no merging operation can be found that increases the modularity value. At this point, the corresponding modularity value is the maximum modularity value.

5. The energy storage planning method for power distribution systems as described in claim 1, characterized in that, The process of determining the source-load cluster partitioning scheme includes: Obtain the cluster label of the power distribution node corresponding to the maximized modularity; Distribution nodes with the same cluster label are grouped into the same source-load cluster to generate the source-load cluster partitioning scheme.

6. The energy storage planning method for power distribution systems as described in claim 5, characterized in that, The construction process of the two-layer model for optimizing energy storage configuration includes: With the minimum comprehensive annual cost as the objective function, and with the power self-balancing, power fluctuation and unit construction capacity multiple of each source-load cluster as constraints, an upper-level optimization model is constructed. With the objective function of minimizing annual network loss, and with Distflow power flow constraints, energy storage operation constraints, and system security constraints as constraints, a lower-level optimization model is constructed. The upper-level optimization model and the lower-level optimization model are combined to obtain the energy storage configuration optimization two-layer model.

7. The energy storage planning method for power distribution systems as described in claim 1, characterized in that, The process of inputting the source-load cluster partitioning scheme into the constructed two-layer energy storage configuration optimization model for processing to obtain the energy storage planning strategy includes: The source-load cluster partitioning scheme and the source-load probabilistic time-series power curve are input into the upper optimization model of the energy storage configuration optimization two-layer model to solve for the energy storage candidate installation scheme, which includes the energy storage installation location, rated capacity and rated power. The energy storage candidate installation scheme is passed to the lower optimization model of the energy storage configuration optimization two-layer model, and the source load probability time series power curve is simulated to obtain the energy storage operation scheduling strategy. The energy storage operation scheduling strategy includes the energy storage charging and discharging power at several times. Determine whether the energy storage operation and scheduling strategy satisfies all the constraints of the lower-level optimization model. If yes, combine the energy storage candidate installation scheme and the energy storage operation and scheduling strategy to output the energy storage planning strategy. If no, adjust the energy storage candidate installation scheme and return to the lower-level optimization model to solve it again until all constraints are satisfied.

8. An energy storage planning device for power distribution systems, characterized in that, include: The acquisition module is used to acquire historical operating data of the source loads and several distribution nodes of the target power distribution system; The processing module is used to sequentially perform probabilistic output modeling and three-point estimation processing on the historical operating data of the source load to obtain the source load probabilistic time-series power curve of the target power distribution system. The determination module is used to determine the electrical distance, source-load complementarity, and node reliability similarity among several distribution nodes based on the source-load probability time-series power curve; and to process the electrical distance, source-load complementarity, and node reliability similarity using the entropy weight method to obtain the comprehensive node similarity. A module is constructed to build a module degree function using the power distribution node and the node comprehensive similarity. The modularity function is optimized to obtain the maximum modularity of the target power distribution system. The maximum modularity is used to determine the source-load cluster partitioning scheme. The planning module is used to input the source-load cluster partitioning scheme into the constructed two-layer energy storage configuration optimization model for processing, so as to obtain the energy storage planning strategy.

9. An energy storage planning device for power distribution systems, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the energy storage planning method for power distribution systems as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements the energy storage planning method for power distribution systems as described in any one of claims 1 to 7.

Images