Power distribution network energy storage locating and sizing method and device and electronic equipment

Abstract

The invention discloses a power distribution network energy storage locating and sizing method and device and electronic equipment. The method comprises the following steps: collecting data of a target power distribution network, including branch information of the target power distribution network and historical load data of the target power distribution network; determining a topological structure of the target power distribution network based on the branch information, and establishing a topological model of the target power distribution network; based on a bilevel planning model, the historical load data and the topology model, a target locating and sizing planning scheme of the energy storage equipment of the target power distribution network is obtained, the bilevel planning model comprises an upper-layer locating planning model and a lower-layer optimization operation model, the upper-layer locating planning model is used for determining an alternative locating and sizing planning scheme of the energy storage equipment, and the lower-layer optimization operation model is used for determining the alternative locating and sizing planning scheme of the energy storage equipment; the lower-layer optimization operation model is used for determining the lowest operation cost corresponding to the alternative locating and sizing planning scheme. The technical problems that the convergence speed is low and the planning scheme is not accurate enough when the energy storage equipment in the power distribution network is subjected to site selection and sizing are solved.

Description

Technical Field

[0001] The present invention relates to the field of electric power, and in particular to a method, device and electronic equipment for selecting a site and determining the capacity of energy storage in a distribution network. Background Art

[0002] In large-scale distribution networks, it is very important to rationally arrange energy storage equipment, which can not only improve the stability of power supply, but also promote the use of green energy and reduce the carbon emissions of the system. In related technologies, heuristic algorithms are used to plan the location and capacity of energy storage equipment based on typical daily load data, which lacks interpretability, has a slow convergence speed, and has unsatisfactory planning effects.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present invention provide a method, device and electronic device for site selection and capacity determination of energy storage in a distribution network, so as to at least solve the technical problems of slow convergence speed and inaccurate planning scheme when site selection and capacity determination of energy storage equipment in a distribution network.

[0005] According to one aspect of an embodiment of the present invention, a method for site selection and sizing of energy storage in a distribution network is provided, comprising: collecting data of a target distribution network, wherein the data comprises: branch information of the target distribution network, and historical load data of the target distribution network; based on the branch information, determining the topological structure of the target distribution network, and establishing a topological model of the target distribution network; based on a two-layer planning model, the historical load data and the topological model, obtaining a target site selection and sizing planning scheme for energy storage equipment of the target distribution network, wherein the two-layer planning model comprises: an upper-layer site selection planning model and a lower-layer optimization operation model, wherein the upper-layer site selection planning model is used to determine alternative site selection and sizing planning schemes for the energy storage equipment, and the lower-layer optimization operation model is used to determine the minimum operating cost corresponding to the alternative site selection and sizing planning schemes.

[0006] Optionally, determining the topological structure of the target distribution network based on the branch information and establishing the topological model of the target distribution network includes: connecting multiple branches based on the branch information to form the topological structure of the target distribution network; establishing the topological model of the target distribution network based on the topological structure, wherein the topological model is represented by two upstream and downstream 0-1 state variable matrices: matrix U and matrix D, element Uij=1 in matrix U, if and only if the edge numbered j supplies active power flow to the point numbered i under natural power flow, element Dij=1 in matrix D, if and only if the point numbered i supplies active power flow to the edge numbered j under natural power flow.

[0007] Optionally, before obtaining the target site selection and sizing planning scheme for the energy storage equipment of the target distribution network based on the two-layer planning model, the historical load data and the topological model, the method also includes: determining a first objective function and a first constraint based on the historical load data and the topological model, wherein the first objective function is used to minimize the operating cost of the alternative site selection and sizing scheme; constructing a lower-level optimization operation model based on the first objective function and the first constraint; determining a second objective function and a second constraint based on the installation cost of the energy storage equipment and the solution of the lower-level optimization operation model, wherein the second objective function is used to minimize the total cost of the alternative site selection and sizing scheme, and the total cost includes the operating cost and the installation cost; constructing an upper-level site selection planning model based on the second objective function and the second constraint; constructing the two-layer planning model based on the upper-level site selection planning model and the lower-level optimization operation model.

[0008] Optionally, the first constraint condition includes at least one of the following: energy storage charging and discharging constraint, energy storage capacity upper and lower limit constraints, node active power balance constraint, node reactive power balance constraint, network Ohm's law constraint, network power flow second-order cone constraint, node voltage constraint, generator power constraint, branch power constraint, and the second constraint condition is energy storage capacity upper and lower limit constraints, energy storage output upper and lower limit constraints.

[0009] Optionally, the method further includes: acquiring a total installation cost of the energy storage device; and obtaining the installation cost of the energy storage device based on the total installation cost and a corresponding discount rate.

[0010] Optionally, the target site selection and sizing planning scheme for the energy storage equipment of the target distribution network is obtained based on the two-layer planning model, the historical load data and the topological model, including: generating a plurality of alternative site selection and sizing planning schemes by the upper-layer site selection and planning model; solving the minimum operating cost corresponding to each of the plurality of alternative site selection and sizing planning schemes by the lower-layer optimization operation model based on the historical load data and the topological model; solving the optimal alternative site selection and sizing planning scheme among the plurality of alternative site selection and sizing planning schemes by the upper-layer site selection and planning model based on the minimum operating cost corresponding to each alternative site selection and sizing planning scheme and the corresponding energy storage equipment installation cost, to obtain the target site selection and sizing planning scheme.

[0011] Optionally, based on the minimum operating cost and the corresponding energy storage equipment installation cost corresponding to each alternative site selection and sizing planning scheme, the upper-level site selection and sizing planning model is used to solve the optimal alternative site selection and sizing planning scheme among the multiple alternative site selection and sizing planning schemes to obtain the target site selection and sizing planning scheme, including: based on the minimum operating cost and the energy storage equipment installation cost corresponding to each alternative site selection and sizing planning scheme, using an improved Bayesian algorithm to solve the upper-level site selection and sizing planning model to obtain the optimal alternative site selection and sizing planning scheme among the multiple alternative site selection and sizing planning schemes to obtain the target site selection and sizing planning scheme, wherein the improved Bayesian algorithm is a Bayesian algorithm combined with a Hyperband strategy.

[0012] Optionally, based on the minimum operating cost and energy storage equipment installation cost corresponding to each alternative site selection and sizing planning scheme, the improved Bayesian algorithm is used to solve the upper-level site selection and sizing planning model to obtain the optimal alternative site selection and sizing planning scheme among the multiple alternative site selection and sizing planning schemes, and the target site selection and sizing planning scheme is obtained, including: using the Hyperband strategy to generate the multiple alternative site selection and sizing planning schemes; based on the minimum operating cost and energy storage equipment installation cost corresponding to each alternative site selection and sizing planning scheme, obtaining the total cost corresponding to each alternative site selection and sizing planning scheme; when the total cost is less than a cost threshold, retaining the corresponding alternative site selection and sizing planning scheme, and when the total cost is greater than or equal to the cost threshold, eliminating the corresponding alternative site selection and sizing planning scheme; based on the retained alternative site selection and sizing planning schemes, multiple alternative site selection and sizing planning schemes for the next round of iterations are obtained, and after multiple rounds of iterations, the target site selection and sizing planning scheme is obtained.

[0013] According to another aspect of the present invention, there is provided a distribution network energy storage site selection and sizing device, comprising: a collection module, for collecting data of a target distribution network, wherein the data comprises: branch information of the target distribution network, historical load data of the target distribution network; an establishment module, for determining the topological structure of the target distribution network based on the branch information, and establishing a topological model of the target distribution network; an acquisition module, for obtaining a target site selection and sizing planning scheme for energy storage equipment of the target distribution network based on a two-layer planning model, the historical load data and the topological model, wherein the two-layer planning model comprises: an upper-layer site selection planning model and a lower-layer optimization operation model, wherein the upper-layer site selection planning model is used to determine alternative site selection and sizing planning schemes for the energy storage equipment, and the lower-layer optimization operation model is used to determine the minimum operating cost corresponding to the alternative site selection and sizing planning schemes.

[0014] According to another aspect of the present invention, there is provided a computer-readable storage medium, the computer-readable storage medium comprising a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned methods for site selection and sizing of energy storage in a distribution network.

[0015] According to another aspect of the present invention, there is provided an electronic device, comprising: a memory storing an executable program; and a processor for running the program, wherein when the program is run, any one of the above-mentioned methods for site selection and sizing of energy storage in a distribution network is executed.

[0016] According to yet another aspect of the present invention, there is provided a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the steps of any one of the methods for selecting a site and sizing energy storage in a distribution network are implemented.

[0017] In the embodiment of the present invention, a two-layer model is solved, and by using complete historical data and an improved Bayesian optimization algorithm combined with a Hyperband strategy, the purpose of selecting a site and sizing the energy storage equipment in the distribution network is achieved, thereby achieving the technical effect of improving the planning robustness and model interpretability of the site selection and sizing of the energy storage in the distribution network, and further solving the technical problems of slow convergence speed and inaccurate planning schemes when selecting a site and sizing the energy storage equipment in the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0019] Figure 1 is a flow chart of a method for selecting a site and determining the capacity of energy storage in a distribution network according to an embodiment of the present invention;

[0020] Figure 2 It is a schematic diagram of a large-scale distribution network energy storage site selection and capacity planning method based on an improved Bayesian algorithm according to an optional implementation mode of the present invention;

[0021] Figure 3 is a flow chart of a BOHB algorithm according to an optional embodiment of the present invention;

[0022] Figure 4 The present invention is a structural block diagram of a distribution network energy storage site selection and capacity determination device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0024] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0025] First, some nouns or terms that appear in the description of the embodiments of the present application are subject to the following explanations:

[0026] Bayesian Optimization (BO) is a global optimization method that is particularly suitable for optimizing black-box functions with high computational costs and limited call budgets. The core idea of ​​Bayesian optimization is to build a probability model (usually a Gaussian process model) to predict the value of the function at unexplored points, and use this model to guide the next search direction to find the global optimal value of the function.

[0027] Hyperband strategy: is an algorithm for fast hyperparameter optimization, which is particularly suitable for machine learning models that require a lot of computing resources (such as time or computing power). Hyperband was proposed by Google scientist Liane Li and others in a 2016 paper. It is a method based on the principle of multi-armed bandit, which aims to find the optimal hyperparameter configuration of the model at a lower cost. The core idea of ​​Hyperband is to dynamically allocate resources to hyperparameter configurations that perform well, while quickly eliminating configurations that perform poorly. It accelerates the process of finding the best configuration by running multiple configuration subsets at different resource levels and screening them based on their initial performance.

[0028] Cplex solver: It is a high-performance commercial mathematical optimization solver developed by IBM, mainly used to solve optimization problems such as linear programming, mixed integer programming, quadratic programming, quadratic constraint programming and quadratic assignment problems. Cplex solver has the ability to solve efficiently, and can also achieve effective preprocessing, multi-threading support, programmable interface, etc.

[0029] Optimal Power Flow (OPF) is a core optimization problem in the power system. It involves finding the power generation scheduling and network control strategy that makes the system operate most economically or meets other specific goals (such as minimizing network losses, maximizing system stability, etc.) under the conditions of satisfying system safety and operational constraints.

[0030] With the continuous expansion of the distribution network and the increasing importance of renewable energy, energy storage technology plays an increasingly important role in modern power systems. Effective energy storage site selection and capacity planning can optimize the allocation of power resources and improve the reliability and flexibility of the power grid. Especially in large-scale distribution networks, the reasonable layout of energy storage equipment can not only improve the stability of power supply, but also promote the use of green energy and reduce the carbon emissions of the system. Therefore, the development of efficient energy storage site selection and capacity planning schemes is of great significance to the realization of a sustainable power system.

[0031] In the related technologies, the research on the site selection and capacity determination of energy storage in distribution networks mostly adopts heuristic algorithms based on load clustering. The complete load data for the whole year is clustered, several typical days are extracted, and then the heuristic algorithm is used for planning optimization and solution. This method can reduce the computational complexity to a certain extent and provide a preliminary reference for the site selection of energy storage. However, this technology also has obvious limitations. For example, the extracted typical days often lack interpretability and are difficult to reflect the full picture of load fluctuations. In addition, when facing high-dimensional problems, the heuristic algorithm converges slowly and is prone to fall into local optimal solutions, resulting in unsatisfactory planning results. More importantly, the typical day selection method in the related technology fails to fully consider extreme load conditions, and the site selection and capacity determination of energy storage equipment need to cope with a variety of complex operating scenarios. Therefore, if the extracted typical days do not cover extreme conditions, the reliability and robustness of the power supply will be insufficient. This means that under sudden load demands or extreme weather conditions, the power grid may not be able to respond effectively, thereby affecting the overall operational stability of the power system.

[0032] Therefore, in order to improve the energy storage site selection and sizing planning scheme and improve its adaptability and explainability in extreme situations, a distribution network energy storage site selection and sizing method is provided. By using comprehensive and complete load data, considering multiple load scenarios and random factors, and considering accidental and emergency situations that have occurred, it helps to improve the reliability of the planning scheme. In addition, by developing efficient computing methods to deal with high-dimensional problems, reduce algorithm complexity, and achieve better energy storage configuration. This method not only helps to improve the operating efficiency of the power grid, but also lays a solid foundation for the future development of smart grids.

[0033] According to an embodiment of the present invention, an embodiment of a method for site selection and sizing of energy storage in a distribution network is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in an order different from that shown here.

[0034] Figure 1 is a flow chart of a method for selecting a site and determining the capacity of energy storage in a distribution network according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0035] Step S102, collecting data of the target distribution network, wherein the data includes: branch information of the target distribution network, and historical load data of the target distribution network.

[0036] As an optional embodiment, the execution subject of the method of this embodiment can be a terminal or server used for distribution network energy storage site selection and capacity determination. When applied to the terminal, it can easily realize the distribution network energy storage site selection and capacity determination. When applied to the server, it can call on the server's rich computing resources, and can more accurately perform distribution network energy storage site selection and capacity determination. Among them, the above-mentioned terminal can be of various types, for example, it can be a mobile terminal with certain computing capabilities, or it can be a fixed computer device with recognition capabilities, etc. The above-mentioned server can also be of various types, for example, it can be a local server or a virtual cloud server. According to the computing power, it can be a single computer device or a computer cluster integrated with multiple computer devices.

[0037] As an optional embodiment, when collecting data of the target distribution network, multiple data can be collected, for example, the branch information of the above-mentioned target distribution network, and the historical load data of the above-mentioned target distribution network. Among them, the above-mentioned branch information is used to describe the structure and electrical characteristics of the distribution network, and may include branch number, branch endpoint, branch resistance and reactance, branch conductance and susceptance, maximum transmission capacity, minimum and maximum voltage, branch type, branch status and other information. The above-mentioned historical load data is the historical load data of the target distribution network over a continuous and complete period of time. After collecting the data of the target distribution network, the above-mentioned data can also be cleaned and preprocessed to ensure that it does not contain missing values, outliers and the like, improve the data quality, and avoid affecting the site selection and capacity determination results due to data quality.

[0038] Step S104: based on the branch information, determine the topology structure of the target distribution network and establish a topology model of the target distribution network.

[0039] As an optional embodiment, when determining the topological structure of the target distribution network based on the branch information and establishing the topological model of the target distribution network, multiple models can be established, for example, a node-branch model, an adjacency matrix model, a state variable matrix model, a tree and ring model, and other models can be established. When establishing the topological model of the target distribution network, multiple methods can be used. For example, based on the branch information, multiple branches can be connected to form the topological structure of the target distribution network; based on the topological structure, the topological model of the target distribution network is established, wherein the topological model is represented by two upstream and downstream 0-1 state variable matrices: matrix U and matrix D, the element Uij in matrix U is 1, if and only if the edge numbered j supplies active power flow to the point numbered i under the natural power flow, and the element Dij in matrix D is 1, if and only if the point numbered i supplies active power flow to the edge numbered j under the natural power flow. By establishing the topological model of the above-mentioned target distribution network in the form of a state variable matrix, the model can more directly reflect the directionality of power flow and the connectivity of the network, making it easier to solve related optimization problems.

[0040] Step S106, based on the two-layer planning model, the above-mentioned historical load data and the above-mentioned topological model, obtain the target site selection and sizing planning scheme for the energy storage equipment of the above-mentioned target distribution network, wherein the above-mentioned two-layer planning model includes: an upper-layer site selection planning model and a lower-layer optimization operation model, wherein the above-mentioned upper-layer site selection planning model is used to determine the alternative site selection and sizing planning schemes for the above-mentioned energy storage equipment, and the above-mentioned lower-layer optimization operation model is used to determine the minimum operating cost corresponding to the above-mentioned alternative site selection and sizing planning schemes.

[0041] As an optional embodiment, before obtaining the target site selection and capacity planning scheme for the energy storage equipment of the target distribution network based on the two-layer planning model, the historical load data and the topology model, the two-layer planning model can be constructed first. Specifically, the upper and lower optimization operation models can be constructed separately, wherein the upper site selection planning model is the leading layer and the lower optimization operation model is the following layer. A planning model can be constructed in a variety of ways. For example, based on the above-mentioned historical load data and the above-mentioned topological model, a first objective function and a first constraint condition are determined, wherein the above-mentioned first objective function is used to minimize the operating cost of the alternative site selection and sizing scheme; based on the above-mentioned first objective function and the above-mentioned first constraint condition, a lower-level optimization operation model is constructed; based on the installation cost of the above-mentioned energy storage equipment and the solution of the above-mentioned lower-level optimization operation model, a second objective function and a second constraint condition are determined, wherein the above-mentioned second objective function is used to minimize the total cost of the above-mentioned alternative site selection and sizing scheme, and the above-mentioned total cost includes the above-mentioned operating cost and the above-mentioned installation cost; based on the above-mentioned second objective function and the above-mentioned second constraint condition, an upper-level site selection planning model is constructed; based on the above-mentioned upper-level site selection planning model and the above-mentioned lower-level optimization operation model, the above-mentioned two-level planning model is constructed. The above two-level programming model can ensure that the operation and installation costs of energy storage equipment are minimized while the power grid is operating safely by optimizing the first objective function and the second objective function. By allowing decisions on the location and sizing of energy storage equipment to be made at the upper level and then calculating the detailed operating effects of this decision at the lower level, a more comprehensive consideration can be made, reducing the complexity of the global optimization problem and making the overall solution process more efficient.

[0042] As an optional embodiment, the first constraint condition includes at least one of the following: energy storage charging and discharging constraint, energy storage capacity upper and lower limit constraint, node active power balance constraint, node reactive power balance constraint, network Ohm's law constraint, network power flow second-order cone constraint, node voltage constraint, generator power constraint, branch power constraint, and the second constraint condition is the upper and lower limit constraint of energy storage capacity. Among them, the energy storage charging and discharging constraint indicates that the energy storage device cannot be charged and discharged at the same time when working, and needs to maintain the continuity of the charging and discharging state within a certain period of time. The upper and lower limit constraints of energy storage capacity indicate that the capacity of the energy storage device must be within a given range to avoid exceeding the physical or economic limitations of the device. The node active power balance constraint indicates that the active power input and output of each node must be balanced, that is, the power generation plus the energy storage discharge must be equal to the load consumption plus the energy storage charge. The node reactive power balance constraint can indicate that the reactive power input and output of each node must also be balanced, involving reactive exchange of generators, loads and energy storage devices. The network Ohm's law constraint ensures that Ohm's law is followed when power flows through the line. The network power flow second-order cone constraint uses the second-order cone programming method to approximate the power flow equation of the power system, making the optimization problem a convex optimization problem. The node voltage constraint limits the voltage of each node to a certain range. The generator power constraint limits the power output of the generator between the minimum and maximum to avoid overloading or underutilization of the generator. The branch power constraint limits the power transmission of the line or transformer to a safe range to avoid overloading or inefficient use. Through the above constraints, the safety, stability and economy of the power system can be fully considered when finding the optimal energy storage site selection and sizing solution.

[0043] As an optional embodiment, when obtaining the target site selection and sizing planning scheme for the energy storage equipment of the target distribution network based on the above-mentioned two-layer planning model, the above-mentioned historical load data and the above-mentioned topological model, the total installation cost of the above-mentioned energy storage equipment can also be obtained; based on the above-mentioned total installation cost and the corresponding conversion rate, the installation cost of the above-mentioned energy storage equipment can be obtained. The above-mentioned installation cost can be obtained in a variety of ways. For example, the unit price of the energy storage equipment can be directly obtained from the market; all costs of the energy storage equipment throughout its life cycle can be considered to conduct a life cycle cost analysis; and cost-benefit analysis can also be used to evaluate the economic benefits brought by the energy storage equipment while calculating the total cost. By obtaining the total installation cost of the above-mentioned energy storage equipment, and obtaining the installation cost of the above-mentioned energy storage equipment based on the total installation cost and the corresponding conversion rate, the future cost is converted into present value using the conversion rate, which facilitates comparison of costs at different time points in decision-making. After conversion, it can also be better integrated into long-term planning, and can be used to optimize decisions to find the lowest cost solution.

[0044] As an optional embodiment, when obtaining the target site selection and sizing planning scheme for the energy storage equipment of the target distribution network based on the above-mentioned two-layer planning model, the above-mentioned historical load data and the above-mentioned topological model, a variety of methods can be used. For example, the upper-level site selection planning model and the lower-level optimization operation model can be nested for solution, and the upper-level site selection planning model generates a preliminary site selection and sizing plan, which is then handed over to the lower-level optimization operation model to solve the operating cost, and iterates multiple times until the optimal solution is found. For another example, the upper-level site selection planning model and the lower-level optimization operation model can be collaboratively optimized, and through interactive iterative solutions, the site selection and sizing plan and the operation strategy can be optimized simultaneously until the convergence conditions are met. The upper-level site selection and planning model can also generate multiple alternative site selection and sizing planning schemes; based on historical load data and topological models, the lower-level optimization operation model solves the minimum operating cost corresponding to each of the multiple alternative site selection and sizing planning schemes; based on the minimum operating cost corresponding to each alternative site selection and sizing planning scheme and the corresponding energy storage equipment installation cost, the upper-level site selection and planning model solves the optimal alternative site selection and sizing planning scheme among the multiple alternative site selection and sizing planning schemes to obtain the target site selection and sizing planning scheme. The above two-level planning model can provide decision support based on data and models, so that the site selection and sizing planning scheme for energy storage equipment is optimal at the technical, economic and strategic levels, that is, it can optimize the use of energy storage resources, ensure that resources are used most effectively while meeting system requirements, and at the same time enable the obtained target site selection and sizing planning scheme to cope with complex and uncertain conditions.

[0045] As an optional embodiment, based on the minimum operating cost corresponding to each of the above-mentioned alternative site selection and capacity planning schemes and the corresponding energy storage equipment installation cost, the above-mentioned upper-level site selection and capacity planning model solves the optimal alternative site selection and capacity planning scheme among the above-mentioned multiple alternative site selection and capacity planning schemes to obtain the above-mentioned target site selection and capacity planning scheme. A variety of methods can be used to solve the upper-level site selection and capacity planning model. For example, a heuristic algorithm can be used for solving, including: genetic algorithm, particle swarm optimization, simulated annealing, ant colony algorithm, etc., and a meta-heuristic algorithm can also be used, including: differential evolution, Bayesian optimization, etc. Among them, Bayesian optimization uses a probability model to predict the performance of unknown parameters, which is suitable for parameter tuning and black box function optimization. It is more efficient when dealing with large-scale optimization problems, but it may not be possible to guarantee that the global optimal solution is found. Therefore, the Bayesian algorithm can be improved to obtain an improved Bayesian algorithm. Specifically, based on the minimum operating cost and energy storage equipment installation cost corresponding to each of the above-mentioned alternative site selection and sizing planning schemes, the improved Bayesian algorithm is used to solve the above-mentioned upper-level site selection and sizing planning model, and the optimal alternative site selection and sizing planning scheme among the above-mentioned multiple alternative site selection and sizing planning schemes is obtained, and the above-mentioned target site selection and sizing planning scheme is obtained, wherein the above-mentioned improved Bayesian algorithm is a Bayesian algorithm combined with the Hyperband strategy. The Hyperband strategy is a resource allocation strategy. By combining the Bayesian algorithm with the Hyperband strategy, the convergence of the Bayesian algorithm can be accelerated, and the optimal solution can be found by dynamically adjusting the resource allocation of the candidate configuration.

[0046] As an optional embodiment, based on the minimum operating cost and energy storage equipment installation cost corresponding to each of the above-mentioned alternative site selection and sizing planning schemes, an improved Bayesian algorithm is used to solve the above-mentioned upper-level site selection and sizing planning model to obtain the optimal alternative site selection and sizing planning scheme among the above-mentioned multiple alternative site selection and sizing planning schemes, and when the above-mentioned target site selection and sizing planning scheme is obtained, a Hyperband strategy can be used to generate the above-mentioned multiple alternative site selection and sizing planning schemes; based on the minimum operating cost and energy storage equipment installation cost corresponding to each of the above-mentioned alternative site selection and sizing planning schemes, the total cost corresponding to each of the above-mentioned alternative site selection and sizing planning schemes is obtained; when the above-mentioned total cost is less than the cost threshold, the corresponding alternative site selection and sizing planning scheme is retained, and when the above-mentioned total cost is greater than or equal to the above-mentioned cost threshold, the corresponding alternative site selection and sizing planning scheme is eliminated; based on the retained alternative site selection and sizing planning schemes, multiple alternative site selection and sizing planning schemes for the next round of iterations are obtained, and after multiple rounds of iterations, the above-mentioned target site selection and sizing planning scheme is obtained. By combining the Hyperband strategy, a large number of alternative solutions can be quickly evaluated, and more computing resources can be concentrated on those solutions that perform well and are expected to achieve the best, thus achieving efficient resource allocation. Through multiple rounds of iterations, the solution with lower cost is retained in each iteration, which can significantly accelerate the solution process and quickly approach the optimal solution. By using the cost threshold as a judgment indicator, better options can be screened out from multiple alternative solutions, that is, those with a total cost lower than the preset threshold, thereby ensuring the economic feasibility of the solution and improving the quality of the final solution. Through the detailed evaluation of the alternative solutions by the lower-level model, the energy storage layout effect under different operating conditions (such as changes in historical load data) can be better reflected, and the robustness of the final solution can be enhanced. Therefore, the Bayesian algorithm combined with the Hyperband strategy is used to solve the upper-level site selection planning model, which can provide an efficient, accurate and economical solution for the site selection and sizing of energy storage equipment in large-scale distribution networks.

[0047] Through the above steps, a better distribution network energy storage site selection and sizing planning scheme can be obtained to ensure that the installation location and energy storage capacity of the energy storage equipment can simultaneously meet the requirements of safe operation and economic benefits of the distribution network. By using complete and continuous historical load data as one of the bases for site selection and sizing, different application conditions and scenarios can be taken into account when determining the plan, avoiding omissions caused by only using typical daily historical data. By using the Bayesian algorithm combined with the Hyperband strategy to solve the distribution network energy storage site selection and sizing planning scheme, the solution speed can be accelerated while ensuring the accuracy of the solution, so that the proposed distribution network energy storage site selection and sizing method can still quickly obtain the target site selection and sizing planning scheme even if complete and continuous historical load data is used.

[0048] In combination with the above embodiments and optional embodiments, an optional implementation is provided. In this optional implementation, a large-scale distribution network energy storage site selection and sizing two-layer planning model based on the Bayesian Optimization with Hyperband (BOHB) algorithm improved by the Hyperband strategy is provided. The core technology is to use BOHB to optimize the upper model, perform site selection and sizing planning for energy storage, and use the Cplex solver to optimize the lower model, and solve the optimal power flow for 365 days of data day by day. This BOHB-based optimization method can obtain a relatively optimal solution by minimizing the number of calculations of the lower model, and does not waste any day's data. It is suitable for the field of pursuing power supply reliability of the power system. Figure 2 is a schematic diagram of a large-scale distribution network energy storage site selection and capacity planning method based on an improved Bayesian algorithm according to an optional implementation mode of the present invention, such as Figure 2 As shown, the method includes the following processing.

[0049] S1, data collection and cleaning.

[0050] Before planning the site selection and capacity of energy storage equipment in the distribution network, we first obtain the information of the distribution network nodes and branches to be planned, as well as the corresponding 8760 hours (i.e., the whole year) of historical load information for planning purposes. We also clean the missing values ​​and outliers to ensure that they do not contain missing values, outliers, etc., to provide high-quality data for the subsequent planning model and prevent planning failures due to data quality issues.

[0051] S2, establish a topological model.

[0052] Convert the distribution network nodes and branch information of the above-mentioned planned area into the format required by the planning model, obtain the topological structure of the distribution network to be planned, and establish a topological model that can be used by the planning model. Construct two upstream and downstream 0-1 state variable matrices U and D, and use the line storage method in the form of an adjacency list to provide an easy-to-use topological model for the optimal power flow model. Specifically, Uij = 1, if and only if the edge numbered j supplies active power flow to the point numbered i under the natural power flow; Dij = 1, if and only if the point numbered i supplies active power flow to the edge numbered j under the natural power flow.

[0053] S3, upper-level model establishment and solution.

[0054] After the topology model is established, the upper-level model is constructed. The upper-level model is used to determine the planning dimensions, upper and lower limits of installed capacity, etc. of the energy storage equipment. In order to improve planning flexibility, a decision vector can be used instead of the site selection process. Assuming that there are n points to be installed, the decision vector has 2n dimensions. The first n dimensions represent the maximum capacity of the energy storage equipment installed at each point to be installed, and the last n dimensions represent the maximum output of the energy storage equipment installed at each installation point. The corresponding value of 0 represents that no energy storage is installed at this point. By setting the above decision variables, the flexibility of decision-making can be improved by expanding the dimensions. Among them, the above decision variables can be generated in the form of vectors, as shown in the following formula:

[0055] X=[s1,s2,...,s i ,...,s N ,p1,p2,...,p i ,...,p N ]

[0056] Where X is the decision vector, s i represents the maximum capacity configuration of energy storage equipment installed at the i-th point, p i is the maximum output power configuration of the energy storage device installed at the i-th point, and N is the number of points to be installed.

[0057] The objective function of the upper model is the annual total cost consisting of the installation discount cost and the annual operation cost. The goal of the upper model is to minimize the above objective function. Among them, the above annual operation cost is obtained by the lower optimization operation model, and the above installation discount cost can be calculated by the following formula:

[0058]

[0059] Where W is the total installation cost; σ% is the discount rate, generally 5%; N is the life span.

[0060] The constraints of the upper model are represented by the upper and lower limit constraints of the elements in the above decision variables, and their mathematical expressions are as follows:

[0061] s min ≤s i ≤s max

[0062] 0≤p i ≤s i

[0063] In the formula, s min 、s max are the lower and upper limits of the maximum capacity of energy storage, respectively. Considering that the sampler will be in 0-s max The sampling time is smaller than s min The installation point is set to 0, which does not play the role of energy storage and does not take into account the cost. iSimilarly, as shown in the following formula:

[0064]

[0065] In order to optimize and solve the above-mentioned upper-level model, an improved Bayesian algorithm is proposed. Bayesian optimization is an optimization technique based on a probability model, which is used to find the optimal solution in the possible parameter space. Its core is to establish a probability model to predict the performance of unknown parameters, and then select new parameters for testing based on this prediction. Specifically, the above-mentioned improved Bayesian algorithm adopts a tree-structured Parzen (Tree-structured Parzen Estimator, referred to as TPE) sampler and a pruner based on the Hyperband strategy. Among them, TPE is a probability model in the Bayesian optimization framework, and kernel density estimation is used to construct two key probability density functions: l(x) and g(x), which represent the probability density functions of the parameter site selection and capacity configuration that performs better than a certain threshold and performs worse than the threshold, respectively. The formula for kernel density estimation is:

[0066]

[0067] Where n is the total number of sample points for site selection and capacity configuration that have been measured; x i is the i-th configuration sample point; h is the bandwidth, which determines the width of the kernel; K h is the kernel function, generally using the Gaussian kernel function, and its mathematical expression is as follows:

[0068]

[0069] The following two probability density functions can be calculated by kernel density estimation:

[0070]

[0071] Where y is the objective function value, which generally refers to the cost for planning problems; γ is the threshold parameter.

[0072] After obtaining the two probability densities, the expected lift is calculated using the ratio of l(x) and g(x), and the mathematical expression is as follows:

[0073]

[0074] The expected improvement helps the algorithm determine the parameters that are most likely to improve the performance of the objective function in the next step. If the ratio is large, it means that there is a higher probability of obtaining a performance better than the threshold γ at position x, so the parameter configuration at this position is worth trying. Through this mechanism, the TPE sampler can effectively guide parameter configurations that may bring better performance while avoiding oversampling in poor quality areas.

[0075] The Hyperband strategy is an optimization strategy based on the multi-armed bandit algorithm. Its core idea is to dynamically allocate resources to more promising configurations. Hyperband accelerates the search for the best configuration by running multiple subsets at different resource levels and quickly discarding poorly performing configurations. Specifically, Hyperband first generates a large number of random site selection and capacity configurations, and initially tests these configurations under limited resources, which is considered their "first round" performance; then based on their performance, some of the configurations with poor performance are eliminated and no longer undergo subsequent testing. The remaining configurations can obtain more resources for the next test, which means that these configurations can more fully demonstrate their potential; repeat the evaluation and elimination process, reducing the number of configurations in each round and increasing the resources allocated to each configuration; after several rounds of iterations, the best site selection and capacity planning scheme remains, which is the best performing configuration under given resources. According to the above process, Hyperband is also regarded as an optimization of the continuous halving strategy.

[0076] Combining the above TPE sampling and Hyperband pruning, the improved Bayesian BOHB algorithm is Figure 3 is a flow chart of a BOHB algorithm according to an optional embodiment of the present invention. Figure 3 shown.

[0077] S4, data grouping.

[0078] In order to be able to call the solver to solve the optimal power flow problem, the annual load data should be processed in groups to reduce memory overhead and computing time, which is convenient for solver calculation. If the distribution network is too large, the number of groups should be increased to 365; if the distribution network is small, the number of groups can be appropriately reduced to reduce the number of solver calls and modeling times, and improve the solution speed.

[0079] S5, lower-level model establishment and total cost calculation.

[0080] After obtaining the distribution network side energy storage site selection and capacity planning scheme transmitted by the upper layer, the lower model is established, and a mixed integer second-order cone programming model can be built to solve the annual minimum operating cost under the planning scheme. The objective function of the lower model is the annual operating cost composed of network loss cost and energy storage charging and discharging depreciation cost. The constraints include network topology constraints, power flow equation constraints, node voltage upper and lower limit constraints, line transmission power constraints, energy storage operation constraints, etc. After the lower model obtains the annual operating cost, it should be combined with the annual discounted cost of the installation cost of the upper model as the total cost and returned to the upper model for optimization.

[0081] Among them, the mathematical expression of the objective function of the lower model is as follows:

[0082]

[0083] Where C is the operating cost, including network loss cost and dispatching cost; is the square of the current in branch l; r l is the resistance value of branch l; ρ t is the network loss cost in time period t; m is the number of branches; T is the dispatching period, if it is the annual load, then T = 8760h; n s is the amount of energy storage installed; P ch,t,i is the charging capacity of the i-th energy storage period t; P dc,t,i is the discharge amount of the i-th energy storage period t; ρ s It is the depreciation cost of charging and discharging unit electricity of energy storage, which is calculated comprehensively based on the installation cost and maintenance cost of the energy storage.

[0084] The constraints of the underlying model include:

[0085] Energy storage charging and discharging constraints are expressed as follows:

[0086] μ dc,i +μ ch,i ≤1

[0087] 0≤p dc,i ≤μ dc,i ×p i

[0088] 0≤p ch,i ≤μ ch ×p i

[0089] In the formula, μ dc,i is the discharge state of the i-th energy storage, if the energy storage is discharged, it is 1, otherwise it is 0, μ ch,i is the charging state of the i-th energy storage, which is 1 if the energy storage is discharged, otherwise it is 0, both of which are 0-1 variables; p dc,i is the discharge power of the i-th energy storage; p ch,i is the charging power of the i-th energy storage.

[0090] The upper and lower limits of energy storage capacity are as follows:

[0091] s i,min ≤S i ≤s i,max

[0092] In the formula, s i,min 、s i,max are the minimum and maximum capacities of the i-th energy storage, respectively; S i is the remaining capacity of the i-th energy storage. In order to ensure safety and battery life, the storage cannot be overcharged or over-discharged. The mathematical expression is as follows:

[0093]

[0094] The capacity of energy storage is updated, and the mathematical expression is as follows:

[0095] S i,t+1 =S i,t +ηP ch,t,i -ηP dc,t,i

[0096] In the formula, S i,t is the capacity of energy storage i in period t, and η is the energy storage charging and discharging efficiency.

[0097] For the convenience of scheduling, the energy storage capacity can be set to be the same at the beginning of each day. The mathematical expression is as follows:

[0098]

[0099] Where T' is the time span of each set of data (h).

[0100] The node active power balance constraint is expressed as follows:

[0101] P g +P dc -P ch +P in =P load

[0102] Where P g P is the active output of the generator at this node; dc is the discharge power of energy storage; P ch is the charging power of energy storage; P load is the active load of the node; P in The active power injected into the node is expressed as follows:

[0103]

[0104] Where P in,i is the injected power of node i; P ij is the active power of branch ij; is the square of the current in branch ij; j:j->i is all branches ij with starting point j and end point i; r ij is the resistance value of branch ij; P ik is the active power of branch ik; k:i->k is all branches ik with starting point i and end point k.

[0105] The node reactive power balance constraint is expressed as follows:

[0106] Q g +Q in =Qload

[0107] In the formula, Q g is the reactive power output of the generator at this node; Q load is the reactive load of the node; Q in Inject reactive power into the node, the expression is as follows:

[0108]

[0109] In the formula, Q in,i Inject reactive power into node i; Q ij is the reactive power of branch ij; is the square of branch ij; x ij is the reactance of branch ij.

[0110] The network is constrained by Ohm's law, and the expression is as follows:

[0111]

[0112] In the formula, is the square of the voltage at node i, and the power flow is j->i; is the square of the voltage at node j; r ij is the resistance value of branch ij; P ij is the active power of branch ij; x ij is the reactance of branch ij; Q ij is the reactive power of branch ij; is the square of the current in branch ij.

[0113] The second-order cone constraint of the network power flow is expressed as follows:

[0114]

[0115] The node voltage constraint is expressed as follows:

[0116]

[0117] In the formula, is the lower limit of the square voltage of node i; is the upper limit of the voltage squared at node i.

[0118] Generator power constraint, the expression is as follows:

[0119] P g,min ≤P g ≤P g,max ,Q g,min ≤Q g ≤Q g,max

[0120] Where Pg,max P is the upper limit of the active power of the generator; g,min is the lower limit of the generator active power; Q g,max is the upper limit of the reactive power output of the generator; Q g,min It is the lower limit of the reactive power output of the generator.

[0121] The branch power constraint is expressed as follows:

[0122] P l,min ≤P l ≤P l,max ,Q l,min ≤Q l ≤Q l,max

[0123] Where P l,max is the maximum active power that can flow through branch l; power; P l,min is the minimum active power that can flow through branch l; Q l,max is the maximum reactive power that can flow through branch l; Q l,min is the minimum reactive power that can flow through branch l.

[0124] A solver can be used to solve the mixed integer second-order cone programming model for each group of load data, and the resulting costs are accumulated to obtain the lowest operating cost for the whole year. The solver can be adjusted according to the actual scenario. In this optional implementation, a Cplex solver is used; the lowest operating cost for the whole year can be superimposed with the annual cost converted from the equipment installation cost, and returned to the upper layer as the total cost of the configuration as the objective function of the upper model.

[0125] S6, solution of the bi-level programming model.

[0126] The two-level planning model established above is solved using the BOHB nested Cplex solver to obtain the minimum total cost after adding energy storage to the annual load, as well as the energy storage site selection and sizing results under the corresponding scheme. That is, in order to obtain the optimal configuration plan, the model established by S3 and S5 is solved, and it is continuously iterated and optimized to obtain the optimal energy storage site selection and sizing planning scheme. This planning scheme fully takes into account the data of every day of the year, ensuring that the planning results are sufficiently robust and can show good reliability under extreme conditions.

[0127] The advantages of this optional implementation method in model construction and model solving are very significant, which improves the planning robustness and model interpretability of power system energy storage site selection and sizing.

[0128] In terms of model construction, the complete annual load data is used instead of the cluster typical daily data, which improves the data utilization rate and avoids the problem that extreme weather and conditions are difficult to show in the cluster, which leads to the situation being ignored by the planning model, thus improving the power supply stability of the planning scheme under extreme conditions. At the same time, the problem of low computational efficiency is solved by the BOHB algorithm.

[0129] In terms of model solving, the BOHB algorithm is used to optimize the energy storage site selection and sizing planning scheme. As an algorithm commonly used in machine learning hyperparameter tuning, the Bayesian algorithm can converge to a relatively good solution with a relatively small number of runs, and can also handle ordinary optimization problems with large objective function calculation time overhead. The Hyperband strategy is an improved strategy based on the multi-armed bandit machine. Its core idea is to increase the budget of configurations with better performance and reduce the time overhead of configurations with poorer performance, thereby expanding the search range while improving the search accuracy and the excellence of the results. Using Hyperband to improve the Bayesian algorithm can improve its search capabilities in high-dimensional space, so the BOHB algorithm is suitable for the rapid solution of the two-level programming model in this optional implementation.

[0130] In summary, this optional implementation method solves the energy storage site selection and sizing planning problem by replacing clustered typical days with complete load data throughout the year, and uses the BOHB algorithm to solve the resulting large-scale distribution network dimensional disasters and large computing time overhead. It is helpful to improve the planning robustness and model interpretability of energy storage site selection and sizing.

[0131] According to an embodiment of the present invention, a device for selecting a location and determining the capacity of energy storage in a distribution network is provided. Figure 4 is a structural block diagram of a distribution network energy storage site selection and capacity determination device according to an embodiment of the present invention, such as Figure 4 As shown, the device includes: a collection module 402, a creation module 404, and an acquisition module 406. The device is described below.

[0132] An acquisition module 402 is used to acquire data of a target distribution network, wherein the data include: branch information of the target distribution network and historical load data of the target distribution network; an establishment module 404 is connected to the acquisition module 402 and is used to determine the topological structure of the target distribution network based on the branch information and to establish a topological model of the target distribution network; an acquisition module 406 is connected to the establishment module 404 and is used to obtain a target site selection and sizing planning scheme for energy storage equipment of the target distribution network based on a two-layer planning model, the historical load data and the topological model, wherein the two-layer planning model includes: an upper-layer site selection planning model and a lower-layer optimization operation model, wherein the upper-layer site selection planning model is used to determine alternative site selection and sizing planning schemes for the energy storage equipment, and the lower-layer optimization operation model is used to determine the minimum operating cost corresponding to the alternative site selection and sizing planning schemes.

[0133] It should be noted here that the above-mentioned acquisition module 402, establishment module 404, and acquisition module 406 correspond to steps S102 to S106 in the embodiment, and the instances and application scenarios implemented by multiple modules and corresponding steps are the same, but are not limited to the contents disclosed in the above-mentioned embodiment.

[0134] As an optional embodiment, the establishment module 404 includes: a formation unit and an establishment unit. The formation unit is used to connect multiple branches based on the branch information to form the topological structure of the target distribution network; the establishment unit is connected to the formation unit and is used to establish a topological model of the target distribution network based on the topological structure, wherein the topological model is represented by two upstream and downstream 0-1 state variable matrices: matrix U and matrix D, the element Uij in matrix U is 1 if and only if the edge numbered j supplies active power flow to the point numbered i under natural power flow, and the element Dij in matrix D is 1 if and only if the point numbered i supplies active power flow to the edge numbered j under natural power flow.

[0135] As an optional embodiment, the device further includes: a first determination module, a first construction module, a second determination module, a second construction module, and a third construction module. The first determination module is used to determine the first objective function and the first constraint based on the historical load data and the topological model, wherein the first objective function is used to minimize the operating cost of the alternative site selection and sizing scheme; the first construction module is connected to the first determination module and is used to construct a lower-level optimization operation model based on the first objective function and the first constraint; the second determination module is connected to the first construction module and is used to determine the second objective function and the second constraint based on the installation cost of the energy storage device and the solution of the lower-level optimization operation model, wherein the second objective function is used to minimize the total cost of the alternative site selection and sizing scheme, and the total cost includes the operating cost and the installation cost; the second construction module is connected to the second determination module and is used to construct an upper-level site selection planning model based on the second objective function and the second constraint; the third construction module is connected to the second construction module and is used to construct the double-level planning model based on the upper-level site selection planning model and the lower-level optimization operation model.

[0136] As an optional embodiment, the above-mentioned first constraint condition includes at least one of the following: energy storage charging and discharging constraint, energy storage capacity upper and lower limit constraints, node active power balance constraint, node reactive power balance constraint, network Ohm's law constraint, network flow second-order cone constraint, node voltage constraint, generator power constraint, branch power constraint, and the above-mentioned second constraint condition is energy storage capacity upper and lower limit constraints, and energy storage output upper and lower limit constraints.

[0137] As an optional embodiment, the device further includes: an acquisition module and a conversion module. The acquisition module is used to acquire the total installation cost of the energy storage device; the conversion module is connected to the acquisition module and is used to obtain the installation cost of the energy storage device based on the total installation cost and the corresponding conversion rate.

[0138] As an optional embodiment, the above-mentioned obtaining module 406 includes: a generation unit, a first solving unit and a second solving unit. The generation unit is used to generate multiple alternative site selection and capacity planning schemes from the above-mentioned upper-level site selection and planning model; the first solving unit is connected to the above-mentioned generation unit, and is used to solve the minimum operating cost corresponding to each alternative site selection and capacity planning scheme in the above-mentioned multiple alternative site selection and capacity planning schemes by the above-mentioned lower-level optimization operation model based on the above-mentioned historical load data and the above-mentioned topological model; the second solving unit is connected to the above-mentioned first solving unit, and is used to solve the optimal alternative site selection and capacity planning scheme in the above-mentioned multiple alternative site selection and capacity planning schemes by the above-mentioned upper-level site selection and planning model based on the minimum operating cost corresponding to each of the above-mentioned alternative site selection and capacity planning schemes and the corresponding energy storage equipment installation cost, so as to obtain the above-mentioned target site selection and capacity planning scheme.

[0139] As an optional embodiment, the second solution unit includes: a solution subunit. The solution subunit is used to solve the upper-level site selection and capacity planning model using an improved Bayesian algorithm based on the minimum operating cost and energy storage equipment installation cost corresponding to each of the above-mentioned alternative site selection and capacity planning schemes, obtain the optimal alternative site selection and capacity planning scheme among the above-mentioned multiple alternative site selection and capacity planning schemes, and obtain the above-mentioned target site selection and capacity planning scheme, wherein the above-mentioned improved Bayesian algorithm is a Bayesian algorithm combined with a Hyperband strategy.

[0140] As an optional embodiment, the above-mentioned solving sub-unit is also used to adopt the Hyperband strategy to generate the above-mentioned multiple alternative site selection and sizing planning schemes; based on the minimum operating cost and energy storage equipment installation cost corresponding to each of the above-mentioned alternative site selection and sizing planning schemes, the total cost corresponding to each of the above-mentioned alternative site selection and sizing planning schemes is obtained; when the above-mentioned total cost is less than the cost threshold, the corresponding alternative site selection and sizing planning scheme is retained, and when the above-mentioned total cost is greater than or equal to the above-mentioned cost threshold, the corresponding alternative site selection and sizing planning scheme is eliminated; based on the retained alternative site selection and sizing planning schemes, multiple alternative site selection and sizing planning schemes for the next round of iteration are obtained, and after multiple rounds of iterations, the above-mentioned target site selection and sizing planning scheme is obtained.

[0141] According to an embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned distribution network energy storage site selection and sizing methods.

[0142] According to an embodiment of the present invention, there is provided an electronic device, comprising: a memory storing an executable program; and a processor for running the program, wherein when the program is run, any one of the above-mentioned methods for selecting a location and determining the capacity of energy storage in a distribution network is executed.

[0143] According to an embodiment of the present invention, a computer program product is provided, including a computer program, wherein when the computer program is executed by a processor, the steps of any one of the above methods are implemented.

[0144] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0145] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0146] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0147] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0148] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0149] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0150] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for selecting a site and determining the capacity of energy storage in a distribution network, characterized in that: include: Collecting data of the target distribution network, wherein the data includes: branch information of the target distribution network, and historical load data of the target distribution network; Based on the branch information, determining the topological structure of the target distribution network, and establishing a topological model of the target distribution network; Based on the two-layer planning model, the historical load data and the topological model, a target site selection and sizing planning scheme for the energy storage equipment of the target distribution network is obtained, wherein the two-layer planning model includes: an upper-layer site selection planning model and a lower-layer optimization operation model, wherein the upper-layer site selection planning model is used to determine the alternative site selection and sizing planning schemes for the energy storage equipment, and the lower-layer optimization operation model is used to determine the minimum operating cost corresponding to the alternative site selection and sizing planning schemes.

2. The method according to claim 1, characterized in that The determining the topology structure of the target distribution network based on the branch information and establishing the topology model of the target distribution network includes: Based on the branch information, multiple branches are connected to form a topological structure of the target distribution network; Based on the topological structure, a topological model of the target distribution network is established, wherein the topological model is represented by two upstream and downstream 0-1 state variable matrices: matrix U and matrix D, the element Uij in matrix U is 1, if and only if the edge numbered j supplies active power flow to the point numbered i under natural power flow, and the element Dij in matrix D is 1, if and only if the point numbered i supplies active power flow to the edge numbered j under natural power flow.

3. The method according to claim 1, characterized in that Before obtaining a target site selection and capacity planning scheme for energy storage equipment of the target distribution network based on the two-layer planning model, the historical load data and the topology model, the method further includes: Based on the historical load data and the topological model, determining a first objective function and a first constraint condition, wherein the first objective function is used to minimize the operating cost of the alternative site selection and capacity determination scheme; Based on the first objective function and the first constraint condition, construct a lower-level optimization operation model; Based on the installation cost of the energy storage device and the solution of the lower-level optimization operation model, determine a second objective function and a second constraint condition, wherein the second objective function is used to minimize the total cost of the alternative site selection and sizing solution, and the total cost includes the operation cost and the installation cost; Based on the second objective function and the second constraint condition, construct an upper-level site selection planning model; The two-layer planning model is constructed based on the upper-layer site selection planning model and the lower-layer optimization operation model.

4. The method according to claim 3, characterized in that The first constraint condition includes at least one of the following: energy storage charging and discharging constraint, energy storage capacity upper and lower limit constraints, node active power balance constraint, node reactive power balance constraint, network Ohm's law constraint, network flow second-order cone constraint, node voltage constraint, generator power constraint, branch power constraint; the second constraint condition is energy storage capacity upper and lower limit constraints, energy storage output upper and lower limit constraints.

5. The method according to claim 3, characterized in that: The method further comprises: Obtaining a total installation cost of the energy storage device; Based on the total installation cost and the corresponding discount rate, the installation cost of the energy storage device is obtained.

6. The method according to claim 1, characterized in that The target location and capacity planning scheme for the energy storage equipment of the target distribution network is obtained based on the two-layer planning model, the historical load data and the topology model, including: Generate multiple alternative site selection and capacity planning schemes from the upper-level site selection and planning model; Based on the historical load data and the topology model, the lower-layer optimization operation model solves the minimum operation cost corresponding to each of the multiple alternative site selection and capacity planning schemes; Based on the minimum operating cost corresponding to each alternative site selection and capacity planning scheme and the corresponding energy storage equipment installation cost, the upper-level site selection and capacity planning model solves the optimal alternative site selection and capacity planning scheme among the multiple alternative site selection and capacity planning schemes to obtain the target site selection and capacity planning scheme.

7. The method according to claim 6, characterized in that The method of solving the optimal alternative site selection and capacity planning scheme among the multiple alternative site selection and capacity planning schemes based on the minimum operating cost corresponding to each alternative site selection and capacity planning scheme and the corresponding energy storage equipment installation cost by the upper-level site selection and planning model to obtain the target site selection and capacity planning scheme includes: Based on the minimum operating cost and energy storage equipment installation cost corresponding to each alternative site selection and sizing planning scheme, an improved Bayesian algorithm is used to solve the upper-level site selection and sizing planning model to obtain the optimal alternative site selection and sizing planning scheme among the multiple alternative site selection and sizing planning schemes, and to obtain the target site selection and sizing planning scheme, wherein the improved Bayesian algorithm is a Bayesian algorithm combined with the Hyperband strategy.

8. The method according to claim 7, characterized in that The method of solving the upper-layer site selection and capacity planning model based on the minimum operating cost and energy storage equipment installation cost corresponding to each of the alternative site selection and capacity planning schemes using an improved Bayesian algorithm to obtain the optimal alternative site selection and capacity planning scheme among the multiple alternative site selection and capacity planning schemes, and obtaining the target site selection and capacity planning scheme, includes: Using a Hyperband strategy, generating the plurality of alternative site selection and capacity planning schemes; Based on the minimum operating cost and energy storage equipment installation cost corresponding to each of the alternative site selection and capacity planning schemes, the total cost corresponding to each of the alternative site selection and capacity planning schemes is obtained; When the total cost is less than the cost threshold, the corresponding alternative site selection and capacity planning scheme is retained; when the total cost is greater than or equal to the cost threshold, the corresponding alternative site selection and capacity planning scheme is eliminated; Based on the retained alternative site selection and capacity planning schemes, multiple alternative site selection and capacity planning schemes for the next round of iterations are obtained, and after multiple rounds of iterations, the target site selection and capacity planning scheme is obtained.

9. A distribution network energy storage site selection and capacity determination device, characterized in that: include: A collection module, used for collecting data of a target distribution network, wherein the data includes: branch information of the target distribution network, and historical load data of the target distribution network; An establishing module, used to determine the topological structure of the target distribution network based on the branch information, and establish a topological model of the target distribution network; A module is obtained, which is used to obtain a target site selection and sizing planning scheme for the energy storage equipment of the target distribution network based on a two-layer planning model, the historical load data and the topology model, wherein the two-layer planning model includes: an upper-layer site selection planning model and a lower-layer optimization operation model, wherein the upper-layer site selection planning model is used to determine alternative site selection and sizing planning schemes for the energy storage equipment, and the lower-layer optimization operation model is used to determine the minimum operating cost corresponding to the alternative site selection and sizing planning schemes.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute the distribution network energy storage site selection and capacity determination method according to any one of claims 1 to 8.

11. An electronic device, characterized in that: include: A memory storing an executable program; A processor is used to run the program, wherein when the program is run, the distribution network energy storage site selection and sizing method described in any one of claims 1 to 8 is executed.

12. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images