Power distribution network energy storage optimization method, device, equipment, medium and program product

Abstract

The embodiment of the invention provides a power distribution network energy storage optimization method and device, equipment, a medium and a program product. The method comprises the following steps: acquiring historical data of distributed photovoltaic, constructing an output uncertainty model of the distributed photovoltaic according to the historical data and a connection function, establishing an objective function, and performing linear reconstruction on nonlinear characteristics of the objective function through a relaxation algorithm, and solving a target function based on a linear reconstruction result to optimize energy storage configuration of the power distribution network. By introducing the connection function, the potential dependency relationship of the distributed photovoltaic output characteristics can be captured, the objective function is established by using the connection function, and the non-linear characteristics are processed through the relaxation algorithm, so that the energy storage configuration of the power distribution network can be more accurate and reasonable, and the cost and the resource waste are reduced on the basis of ensuring the power supply reliability of the power distribution network.

Description

Technical Field

[0001] The present application relates to the field of power grid energy storage technology, and in particular to a distribution network energy storage optimization method, device, equipment, medium and program product. Background Art

[0002] Renewable energy technologies such as photovoltaics are developing rapidly and are increasingly becoming a significant part of power grids, posing significant challenges to grid planning, scheduling, and operation. Energy storage systems (ESS), as a flexible resource, can be widely used in distribution networks to mitigate fluctuations in power generation from photovoltaics and other energy sources. Consequently, planning issues within these networks have garnered significant attention.

[0003] At present, there are a large number of distributed photovoltaics in the distribution network. The output of these distributed photovoltaics is random, which makes it very difficult to set up energy storage systems for distributed photovoltaics during grid planning, and easily leads to waste of resources. Summary of the Invention

[0004] The embodiments of the present application provide a distribution network energy storage optimization method, device, equipment, medium and program product to reduce the difficulty of distribution network planning.

[0005] In a first aspect, an embodiment of the present application provides a distribution network energy storage optimization method, comprising:

[0006] Obtain historical data of distributed photovoltaics;

[0007] Constructing a distributed photovoltaic output uncertainty model based on the historical data and the connection function Copula;

[0008] Establishing an objective function according to the output uncertainty model, wherein the objective function includes nonlinear characteristics;

[0009] Using a relaxation algorithm to process the nonlinear characteristics of the objective function to obtain a linear reconstruction feature;

[0010] The objective function is solved according to the linear reconstruction characteristics, and the energy storage configuration of the distribution network is optimized according to the solution result.

[0011] In one possible implementation, constructing a distributed photovoltaic output uncertainty model based on the historical data and the connection function Copula includes:

[0012] determining a fuzzy set based on the historical data and the Wasserstein distance;

[0013] defining Copula parameters based on the historical data, wherein the Copula parameters represent the dependency relationship of distributed photovoltaic output characteristics;

[0014] The fuzzy set and the Copula parameters are fused, and a distributed photovoltaic output uncertainty model is constructed according to the fusion result.

[0015] In a possible implementation, determining a fuzzy set based on the historical data and the Wasserstein distance includes:

[0016] defining an empirical distribution characterizing the distributed photovoltaic output characteristics based on the historical data;

[0017] Calculating the Wasserstein distance between the empirical distribution and the target distribution;

[0018] A fuzzy set representing the probability distribution of distributed photovoltaic output is defined according to the Wasserstein distance.

[0019] In a possible implementation, establishing an objective function according to the output uncertainty model includes:

[0020] Obtaining decision variables for distribution network energy storage optimization, the decision variables including energy storage planning and scheduling decisions;

[0021] Determining the worst probability distribution of distributed photovoltaic output according to the output uncertainty model;

[0022] An objective function is established with the goal of minimizing the sum of the energy storage planning and the scheduling decision under the worst probability distribution.

[0023] In a possible implementation, a relaxation algorithm is used to process the nonlinear characteristics of the objective function to obtain a linear reconstruction feature, including:

[0024] Detecting features with strong duality in the objective function through convex optimization conditions and obtaining corresponding cone reconstruction results, wherein the cone reconstruction results include nonlinear constraints;

[0025] A relaxation algorithm is used to perform linear reconstruction on the nonlinear constraint, and a linear reconstruction result is obtained as a linear reconstruction feature.

[0026] In a possible implementation, the constraint conditions of the objective function include at least one of an energy storage planning constraint, an energy storage operation constraint, a photovoltaic output constraint, a power balance constraint, and a distributed robustness opportunity constraint.

[0027] In a second aspect, an embodiment of the present application provides a distribution network energy storage optimization device, comprising:

[0028] Acquisition module, used to obtain historical data of distributed photovoltaic;

[0029] A model building module, configured to build a distributed photovoltaic output uncertainty model based on the historical data and the connection function Copula;

[0030] A function generation module, configured to establish an objective function according to the output uncertainty model;

[0031] A linear processing module, configured to process the nonlinear characteristics of the objective function using a relaxation algorithm to obtain linear reconstruction characteristics;

[0032] An optimization module is used to solve the objective function according to the linear reconstruction characteristics and optimize the energy storage configuration of the distribution network according to the solution result.

[0033] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a memory, a processor;

[0034] The memory stores computer-executable instructions;

[0035] The processor executes the computer-executable instructions stored in the memory, so that the processor executes the above first aspect and / or various possible implementations of the first aspect.

[0036] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the first aspect above and / or various possible implementation methods of the first aspect.

[0037] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the above first aspect and / or various possible implementation methods of the first aspect.

[0038] The distribution network energy storage optimization method, device, equipment, medium and program product provided in the embodiments of the present application can obtain historical data of distributed photovoltaics, construct a distributed photovoltaic output uncertainty model based on historical data and a connection function and establish an objective function, then linearly reconstruct the nonlinear characteristics of the objective function through a relaxation algorithm, and solve the objective function based on the linear reconstruction result to optimize the distribution network energy storage configuration. By introducing a connection function, the potential dependency of distributed photovoltaic output characteristics can be captured, and the objective function can be established by using it. By processing the nonlinear characteristics through a relaxation algorithm, the energy storage configuration of the distribution network can be made more accurate and reasonable, reducing costs and reducing resource waste while ensuring the reliability of the distribution network power supply. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0040] Figure 1 A schematic diagram of the process of a distribution network energy storage optimization method provided in this application Figure 1;

[0041] Figure 2 A schematic diagram of the process of a distribution network energy storage optimization method provided in this application Figure 2 ;

[0042] Figure 3 A schematic diagram of the process of a distribution network energy storage optimization method provided in this application Figure 3 ;

[0043] Figure 4 A schematic diagram of an exemplary energy storage configuration provided for this application;

[0044] Figure 5 A schematic diagram of the structure of a distribution network energy storage optimization device provided in this application;

[0045] Figure 6 This is a schematic diagram of the structure of an electronic device provided in this application.

[0046] The above drawings illustrate specific embodiments of the present application, which will be described in more detail below. These drawings and the textual description are not intended to limit the scope of the present application in any way, but rather to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0047] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0048] The rapid development of renewable energy such as photovoltaics has brought problems to the distribution system, such as the probabilistic balance of power and electricity, the difficulty in absorbing renewable energy, and the increase in the operating costs of the power system, which has brought severe challenges to the planning, scheduling and operation of the power grid.

[0049] As a flexible resource, energy storage system (ESS) can be widely used in distribution networks due to its energy time-shifting characteristics and rapid response capabilities to smooth out fluctuations in renewable energy, promote local consumption of renewable energy, achieve friendly grid connection of distributed power sources, and delay grid upgrades.

[0050] Distribution networks often include a large number of distributed photovoltaic systems. Their output is random, meaning it's difficult to determine the amount of energy they can provide to the grid at any given moment in the future. This is a key factor in grid planning. Due to this output uncertainty, ensuring reliable power supply requires significant margin for distributed photovoltaic energy storage during grid planning, which can lead to high costs and a waste of resources.

[0051] After extensive research and analysis of historical distributed photovoltaic power generation data, the inventors discovered that while distributed photovoltaic power generation exhibits output uncertainty, when historical output is correlated with factors such as geographic proximity and meteorological conditions, this data exhibits potential dependencies. The inventors developed the idea of ​​capturing these dependencies and leveraging them to optimize the energy storage configuration of distribution networks. This approach, in turn, allows for more accurate and rational energy storage configuration, reducing costs and minimizing resource waste while ensuring reliable power supply.

[0052] Based on this, the present application proposes a distribution network energy storage optimization method, device, equipment, medium and program product to solve the above technical problems.

[0053] The following specific embodiments describe in detail the technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0054] Figure 1 A schematic diagram of the process of a distribution network energy storage optimization method provided in this application Figure 1 ,like Figure 1 As shown, the method includes:

[0055] Step S101: Obtain historical data of distributed photovoltaics.

[0056] In the embodiment of the present application, the historical data of distributed photovoltaics may include data such as the power generation and energy storage cost of distributed photovoltaics in a certain period or at a certain moment in the past.

[0057] Step S102: constructing a distributed photovoltaic output uncertainty model based on historical data and the connection function Copula.

[0058] Among them, the connection function Copula, also known as the Copula function, is a distribution function with uniform marginal distribution. It can be used to capture the dependency structure information between uncertain parameters and accurately describe the dependency relationship between random variables.

[0059] Specifically, the probability distribution of distributed photovoltaic output can be determined based on its historical data, and a fuzzy set that characterizes the uncertainty of distributed photovoltaic output can be defined based on the probability distribution. Then, a copula is defined based on the cumulative distribution function of the historical distributed photovoltaic output data, and the copula is embedded in the fuzzy set to construct an output uncertainty model.

[0060] Step S103: establishing an objective function according to the output uncertainty model.

[0061] Specifically, after obtaining the output uncertainty model of distributed photovoltaics, the distributed robust chance constraint can be used to describe the output uncertainty of distributed photovoltaics, and a two-stage distributed robust optimization model (DRO model) of distribution network energy storage can be constructed.

[0062] Among them, the first stage can be the planning stage, which is used to plan the site selection and capacity determination of the distribution network energy storage system; the second stage can be the scheduling stage, which is used to make scheduling decisions based on the actual output of distributed photovoltaics.

[0063] The decision variables of the two-stage distributed robust optimization model for distribution network energy storage can include energy storage planning schemes and scheduling decisions. The objective function is to minimize the sum of the distribution network energy storage investment and scheduling operation costs under the worst probability distribution of the fuzzy set taking into account the uncertainty of distributed photovoltaic output.

[0064] Optionally, the constraint conditions of the objective function may include at least one of an energy storage planning constraint, an energy storage operation constraint, a photovoltaic output constraint, a power balance constraint, and a distributed robustness opportunity constraint.

[0065] Step S104: Using a relaxation algorithm to process the nonlinear characteristics of the objective function to obtain linear reconstruction characteristics.

[0066] Step S105 , solving the objective function according to the linear reconstruction characteristics, and optimizing the energy storage configuration of the distribution network according to the solution result.

[0067] In an embodiment of the present application, the constructed two-stage distributed robust optimization model can be linearized, and a relaxation algorithm can be used to process the nonlinear characteristics involved in the objective function to obtain a linear reconstruction form of the optimization problem, so that the model can be solved using a commercial solver, and the energy storage system and other configurations of the distribution network can be set according to the solution results.

[0068] In the above embodiment, historical data for distributed photovoltaic systems is obtained, and a distributed photovoltaic output uncertainty model and objective function are constructed based on the historical data and the connection function. The nonlinear characteristics of the objective function are then linearly reconstructed using a relaxation algorithm. The objective function is then solved based on the linear reconstruction results to optimize the energy storage configuration of the distribution network. By introducing the connection function, the potential dependencies of distributed photovoltaic output characteristics can be captured, utilized to establish the objective function, and the nonlinear characteristics processed using the relaxation algorithm. This allows for more accurate and reasonable energy storage configuration in the distribution network, reducing costs and resource waste while ensuring the reliability of the distribution network.

[0069] Figure 2 A schematic diagram of the process of a distribution network energy storage optimization method provided in this application Figure 2 ,like Figure 2 As shown in the figure, the output uncertainty model of distributed photovoltaics is constructed based on historical data and the connection function Copula, which can include:

[0070] Step S201 : determining a fuzzy set based on historical data and Wasserstein distance.

[0071] Among them, Wasserstein distance is a measurement method that can be used to measure the difference between two probability distributions.

[0072] In some possible implementations, determining a fuzzy set based on historical data and the Wasserstein distance may include: defining an empirical distribution characterizing distributed photovoltaic output characteristics based on the historical data; calculating the Wasserstein distance between the empirical distribution and a target distribution; and defining a fuzzy set characterizing the probability distribution of distributed photovoltaic output based on the Wasserstein distance.

[0073] Step S202: define Copula parameters based on historical data.

[0074] Among them, the Copula parameter represents the dependence of distributed photovoltaic output characteristics.

[0075] Step S203: Fuzzy sets and Copula parameters are fused, and a distributed photovoltaic output uncertainty model is constructed based on the fusion result.

[0076] The following describes the process of constructing a distributed photovoltaic output uncertainty model in an embodiment of the present application with reference to an example.

[0077] Assume that the output error of distributed photovoltaic at time t is Consider N historical observation data sets {δ1, δ2, ..., δ N}, define the empirical distribution Q N .

[0078]

[0079] In the above formula (1), Represented by δ i The Dirac distribution centered on Next, using the Wasserstein distance d W (Q N ,Q) calculate the distance between two distribution functions, d W (Q N ,Q) is expressed as follows:

[0080]

[0081] In the above formula (2), A is the support space that defines all distribution functions; δ and δ are the marginal distributions Q N , Q’s random variables; ||δ-δ|| is the distance between random variables, which is processed using the 1-norm; ∏(ζ δ ,ζ δ ) means defined in Q N , the joint probability distribution on Q.

[0082] The fuzzy set Ω1 of distributed photovoltaic output probability distribution based on Wasserstein distance is defined as follows:

[0083] Ω1={Q∈A|d W (Q N , Q)≤R W} (3)

[0084] In the above formula (3), R W is the Wasserstein radius, and its value reflects the range of probability distribution covered under a given confidence level β∈(0,1). The larger the β value, the more distributions are included in the fuzzy set, and the system uncertainty increases accordingly.

[0085] Among them, R W It can be expressed as the following formula (4):

[0086]

[0087] In the above formula (5), β is the confidence level; Represents the sample mean; λ is an auxiliary variable, which can be completed by segmented search method; after λ is given, the auxiliary variable C W Can be determined by dichotomy.

[0088] The copula of the distribution Q is defined as the cumulative distribution function of the uncertainty parameter U:

[0089]

[0090] In the above formula (6), N DPV Indicates the number of distributed photovoltaic units; It represents the output deviation uncertainty of distributed photovoltaic units and is associated with the distribution Q; Associated with the distribution C; function Represents the dimension N DPV The uncertainty vector Chinese elements The cumulative distribution function, probability operator Q i (.) represents a random variable The marginal distribution of F i (.) defines a random variable uniformly distributed on [0,1].

[0091] For the marginal cumulative distribution function F i (.), μ∈R is defined as the function variable. For a given N equally probable historical observations, the empirical marginal cumulative distribution of distributed photovoltaic unit i can be written as:

[0092]

[0093] In the above formula (7), st is (subject to), which represents the constraint condition. ik is an auxiliary variable. Copula is embedded in Ω1 to construct the improved Wasserstein fuzzy set Ω2 as shown in the following formula (8):

[0094]

[0095] In the above formula (8), r W It is a parameter determined by the decision maker, reflecting the decision maker's tolerance for the uncertainty of the dependencies between distributed photovoltaic units.

[0096] Figure 3 A schematic diagram of the process of a distribution network energy storage optimization method provided in this application Figure 3 ,like Figure 3 As shown in Figure 2, the objective function established based on the output uncertainty model may include:

[0097] Step S301: Obtain decision variables for distribution network energy storage optimization.

[0098] Among them, decision variables include energy storage planning and scheduling decisions.

[0099] Step S302: determining the worst probability distribution of distributed photovoltaic output according to the output uncertainty model.

[0100] Step S303 , establishing an objective function with the goal of minimizing the sum of energy storage planning and scheduling decisions under the worst probability distribution.

[0101] The process of establishing the objective function in the embodiment of the present application is described below with reference to an example.

[0102] The decision variables of the two-stage DRO model for energy storage planning and operation in the distribution network include the energy storage planning scheme x and the scheduling decision y. The objective function is to minimize the sum of the distribution network energy storage investment and operation costs under the worst probability distribution of the fuzzy set taking into account the uncertainty of distributed photovoltaic output, namely:

[0103]

[0104] In equation (9), C1(x) is the cost of configuring alternative energy storage for the distribution network during the planning phase; C2(y,δ) is the dispatching and operating cost of the distribution network during the operation phase; X and Y are feasible sets; the probability distribution of the uncertainty of distributed PV output is Q; and Ω2 is a Wasserstein fuzzy set based on the improved Copula. The cost calculation methods for the planning and operation phases are as follows:

[0105] First, the cost of energy storage configuration in the distribution network during the planning stage:

[0106]

[0107] In the above formula (10), N node is the number of distribution network nodes; X i is the decision variable for whether node i should invest in energy storage, which is a variable in the interval 0 to 1, where 0 represents not investing and 1 represents investing; is the unit capacity investment cost of energy storage, Invest in energy storage capacity for node i; is the unit power investment cost of energy storage, The rated power of the energy storage system at node i.

[0108] Second, the distribution network dispatching and operating costs during the operation phase:

[0109] The operation cost of the distribution network dispatching in the operation phase includes the operation and maintenance cost C caused by calling energy storage. main , power purchase cost of distribution network C pur , the cost of curtailing wind and solar power C pun and the system load loss cost C loss .

[0110] C2(y,δ)=C main +C pur +C pun +C loss (11)

[0111]

[0112] R j=(1+r) -j (16)

[0113] In the above formula, N year is the total number of years included in the planning period; T is the number of scheduling periods; R j is the present value coefficient for the jth year, j is the number of years from the time the cost is incurred to the beginning of the planning period, and r is the discount rate; is the unit power operation and maintenance cost of energy storage; is the charge and discharge power of energy storage device i in time period t; is the charge and discharge power of energy storage device i in time period t to balance the predicted deviation of distributed photovoltaic output; is the real-time electricity price of the power purchased by the distribution network from the upper power grid during period t. This paper adopts the time-of-use electricity price. The power purchased by the distribution network from the upper power grid during period t; is the unit power curtailment cost of photovoltaic power; is the amount of abandoned solar power in the distribution network during period t; p loss is the load loss cost per unit power; is the load loss power of node i in time period t.

[0114] In the above example, the constraints of the objective function may include energy storage planning constraints, energy storage operation constraints, photovoltaic output constraints, power balance constraints, network flow constraints, power purchase cap constraints, and distributed flooding opportunity constraints.

[0115] The energy storage planning constraints can be expressed as:

[0116]

[0117] In the above formula, is the maximum energy storage capacity that can be configured for node i, is the maximum energy storage power that can be configured for node i, η min ,η max are the minimum and maximum power coefficients of energy storage, The maximum number of energy storage devices that can be configured in the distribution network.

[0118] The energy storage operation constraints can be expressed as:

[0119]

[0120]

[0121] In the above formula, are variables between 0 and 1 representing the charging and discharging states of the energy storage at node i, respectively; is the state of charge (SOC) of the energy storage at node i at time t, is the upper and lower limits of SOC; η loss is the energy loss coefficient of energy storage, η cha ,η dis are the charging efficiency and discharging efficiency of energy storage respectively; Δt is the time interval between adjacent scheduling times, which is taken as 1 hour in this model.

[0122] The photovoltaic output constraint can be expressed as:

[0123]

[0124] In the above formula, Ω PV is the set of nodes equipped with distributed photovoltaic units, N DPV Ω PV The total number of nodes connected to distributed photovoltaics; is the total output power of distributed photovoltaic power in the distribution network at time t, is the actual photovoltaic output at node l at time t; is the amount of abandoned light at node l at time t, is the total amount of abandoned solar power in the distribution network.

[0125] The power balance constraint can be expressed as:

[0126]

[0127]

[0128] In the above formula, is the total load of the distribution network during period t, is the load of node i in period t; is the total load loss power of the distribution network during period t, is the load loss power at node i during period t; The net discharge power of alternative energy storage for distribution networks; is the net discharge power of the distribution network to balance the distributed photovoltaic prediction error.

[0129] The network power flow constraint can be expressed as:

[0130]

[0131] In the above formula, are distribution network lines and node sets respectively; are the sets of lines with node n as the starting and ending points respectively; and are the voltage at node n, the current of line l, and the power of line l at time t respectively; is the outflow power of node n at time t; R line，l is the resistance value of line l; Pline，max 、 and They are and The upper limit value of .

[0132] The upper limit constraint of electricity purchase can be expressed as:

[0133]

[0134] In the above formula, P pur，max The upper limit of the amount of electricity that the distribution network can purchase from the upper-level power grid.

[0135] The distributional robustness chance constraint can be expressed as:

[0136]

[0137] In the above formula, Ω node is the node set; ES is the confidence level of the distribution constraint; α g is the response factor. The distribution network energy storage installed at node g charges and discharges according to the response factor to balance the imbalance of distributed photovoltaics during the daily stage.

[0138] In one embodiment, a relaxation algorithm is used to process the nonlinear characteristics of the objective function to obtain linear reconstruction characteristics, which may include:

[0139] In the first step, convex optimization conditions are used to detect features with strong duality in the objective function and obtain the corresponding cone reconstruction results.

[0140] In the second step, a relaxation algorithm is used to linearly reconstruct the nonlinear constraints and obtain the linear reconstruction results as linear reconstruction features.

[0141] The cone reconstruction result includes nonlinear constraints.

[0142] The following is an example to illustrate the process of processing the objective function through the relaxation algorithm in the embodiment of the present application.

[0143] The maximum function (max) and minimum function (min) in the objective function are intended to determine the worst probability distribution within a given fuzzy set and make a decision. The objective function can be tested by the convex optimization Slater condition. After testing, it is determined to have strong duality. After cone reconstruction, it can be expressed as:

[0144]

[0145]

[0146] In the above formula, ω,v∈R +,θ k ∈R and All are auxiliary variables introduced; ||.|| * It can be expressed as the dual norm of the calculation vector. According to the above formula and argument, the cumulative distribution function It can be expressed as:

[0147]

[0148] On this basis, the duality theory can be used to transform the maximization operator in F(δ) into an external constraint:

[0149]

[0150] Furthermore, substituting the above formulas (53) and (54) into formula (51) yields:

[0151]

[0152] In the above formula (55), there is a nonlinear term z ilk δ l and τ lik δ l , McCormick relaxation can be used to deal with nonlinear terms.

[0153] Among them, McCormick relaxation is a mathematical technique used to deal with bilinear terms (such as z = x·y) in optimization problems. Its core idea is to approximate nonlinear product terms through linear constraints, thereby converting non-convex problems into convex or linear forms that are easier to solve. This method is based on the known upper and lower bounds of variables x and y (x∈[x L ,x U ], y∈[y L ,y U ]), construct the convex envelope and concave envelope of the bilinear term, introduce the combination of extreme points (such as x L y L and x U y U ) so that the introduced constraints cover the possible range of bilinear terms and the original nonlinear terms are relaxed into a set of linear inequalities, thereby simplifying the optimization model.

[0154] In some possible implementations, nonlinear terms may be processed using piecewise linear approximation algorithms, reconstruction linearization algorithms, and outer approximation methods.

[0155] Piecewise linear approximation is a method that approximates a nonlinear function using multiple linear segments. The basic idea is to divide the domain of the nonlinear function into several small intervals and use a linear function to approximate the original function in each interval.

[0156] The exterior approximation method is a method for solving nonlinear problems by constructing a series of exterior approximations. Its basic principle is to use a set of linear inequalities to approximate the exterior boundary of the nonlinear feasible region.

[0157] In this example, the McCormick relaxation is used to handle the nonlinear term z ilk δ l and τ lik δ l , formula (55) can be further transformed into:

[0158]

[0159] In the above formula, τ lik ,π lik is an auxiliary variable. Optionally, the Lagrangian method can be used to dualize the above equation to get rid of the maximization operator, and finally a linear reconstruction form of the optimization problem can be obtained, so that the model can be solved using a commercial solver. The solution process can refer to the following formula:

[0160]

[0161]

[0162] In the above formula, μ l are auxiliary variables, are the upper and lower bounds of the uncertainty of distributed photovoltaic output, is the upper bound of the McCormick relaxation factor.

[0163] In order to verify the effectiveness of the distribution network energy storage optimization method of this application, a certain 10kV distribution system is taken as the research object as an example.

[0164] First, the economic benefits of building shared energy storage on the distribution network side are analyzed, and two optimization scenarios are set: 1) no energy storage is configured on the distribution network side; 2) shared energy storage is configured on the distribution network side. The energy storage configuration results of scenario 2 can be referred to the following table and Figure 4 . Figure 4 In the diagram, black circular nodes represent configuration nodes.

[0165] Configure Node Energy storage capacity / MWh Energy storage power / MW 5 4.94 1.79 10 4.86 1.77 14 5.06 2.01 15 4.29 1.65 17 4.41 1.88

[0166] The comparison of economic benefits of distribution network is shown in the following table:

[0167] Cost Category With energy storage configuration No energy storage configuration Total cost 9020 19561 Energy storage investment and construction costs 1793.2 / Energy storage operation and maintenance costs 117.8 / Power purchase cost 7109 10039 Loss of load 0 2762 Cost of curtailing wind and solar power 0 6760

[0168] According to the above table, when energy storage is configured according to the method proposed in this application, the total cost of the distribution network is 90.2 million yuan, while the total cost is 195.61 million yuan when no energy storage is configured. This solution can save 53.89% of the distribution network cost. When energy storage is configured, the load loss and wind and solar power curtailment are both 0. However, when energy storage is not configured, even if the system is allowed to partially feed power back to the upper grid, a large amount of wind and solar power curtailment and load loss still occurs.

[0169] Furthermore, to verify the effectiveness of the distribution network energy storage optimization method (hereinafter referred to as "C-WDRO") proposed in this application, the site selection and sizing problem of distribution network energy storage devices was solved using the stochastic optimization (SP) algorithm, the traditional Wasserstein distributed robust optimization algorithm (WDRO), and the C-WDRO algorithm of this application. The calculated distribution network operating costs were compared, and the results are shown in the following table:

[0170] Optimization methods SP C-WDRO WDRO Total cost 7567.8 9020 11216.5 Energy storage investment and construction costs 1476.8 1793.2 2256.32 Energy storage investment and construction 3 5 5

[0171] The total cost of the proposed C-WDRO algorithm is 90.2 million yuan, which is 14.522 million yuan higher than SP and 21.965 million yuan less than WDRO. This is because SP only considers scenarios with fixed probabilities, and its planning strategy is overly optimistic. Traditional WDRO, on the other hand, contains many distributions that do not match the actual situation, and its planning strategy is overly conservative, resulting in a waste of resources. The cost of the C-WDRO algorithm is between SP and WDRO, and it can take into account both the robustness and economy of the strategy.

[0172] It can be seen from the above embodiments that the traditional SP algorithm cannot take into account the output of distributed photovoltaics under the worst probability distribution, which will result in insufficient reliability and security of the distribution network power supply; although the traditional WDRO algorithm can improve the reliability of the distribution network power supply, it increases the cost of energy storage configuration in order to take into account the uncertainty of distributed photovoltaic output; and the distribution network energy storage optimization method (C-WDRO) of the present application uses a Copula function to capture and utilize the potential connections of the uncertainty of distributed photovoltaic output. By using this potential connection, a more accurate probability distribution and fuzzy set that can better reflect the actual output of distributed photovoltaics can be defined, thereby reducing the cost of energy storage configuration while ensuring the reliability of the distribution network power supply, making the final solution cost lower than the traditional WDRO algorithm and reducing resource waste.

[0173] Figure 5 A schematic diagram of the structure of a distribution network energy storage optimization device provided in this application is shown as follows: Figure 5 As shown, the distribution network energy storage optimization device 500 provided in this embodiment includes:

[0174] Acquisition module 501, used to acquire historical data of distributed photovoltaic;

[0175] A model building module 502 is used to build a distributed photovoltaic output uncertainty model based on the historical data and the connection function Copula;

[0176] Function generation module 503, used to establish an objective function according to the output uncertainty model;

[0177] A linear processing module 504 is configured to process the nonlinear characteristics of the objective function using a relaxation algorithm to obtain a linear reconstruction characteristic;

[0178] The optimization module 505 is configured to solve the objective function according to the linear reconstruction feature and optimize the energy storage configuration of the distribution network according to the solution result.

[0179] In one possible implementation, the model building module 502 is further configured to: determine a fuzzy set based on the historical data and the Wasserstein distance; define Copula parameters based on the historical data, wherein the Copula parameters characterize the dependency of distributed photovoltaic output characteristics; fuse the fuzzy set and the Copula parameters, and construct a distributed photovoltaic output uncertainty model based on the fusion result.

[0180] In one possible implementation, the model building module 502 is further configured to: define an empirical distribution characterizing the distributed photovoltaic output characteristics based on the historical data; calculate a Wasserstein distance between the empirical distribution and a target distribution; and define a fuzzy set characterizing the probability distribution of the distributed photovoltaic output based on the Wasserstein distance.

[0181] In one possible implementation, the function generation module 503 is further used to: obtain decision variables for distribution network energy storage optimization, the decision variables including energy storage planning and scheduling decisions; determine the worst probability distribution of distributed photovoltaic output based on the output uncertainty model; and establish an objective function with the goal of minimizing the sum of the energy storage planning and the scheduling decision under the worst probability distribution.

[0182] In one possible embodiment, the linear processing module 504 is also used to: detect features with strong duality in the objective function through convex optimization conditions and obtain corresponding conical reconstruction results, wherein the conical reconstruction results include nonlinear constraints; use a relaxation algorithm to linearly reconstruct the nonlinear constraints to obtain linear reconstruction results as linear reconstruction features.

[0183] In a possible implementation, the constraint conditions of the objective function include at least one of an energy storage planning constraint, an energy storage operation constraint, a photovoltaic output constraint, a power balance constraint, and a distributed robustness opportunity constraint.

[0184] The distribution network energy storage optimization device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effects are similar and will not be described in detail in this embodiment.

[0185] Figure 6 This is a schematic diagram of the structure of an electronic device provided by this application. Figure 6 As shown, the electronic device 60 provided in this embodiment includes: at least one processor 601 and a memory 602. Optionally, the device 60 further includes a communication component 603. The processor 601, the memory 602 and the communication component 603 are connected via a bus 604.

[0186] During the specific implementation process, at least one processor 601 executes the computer-executable instructions stored in the memory 602, so that the at least one processor 601 performs the above method.

[0187] The specific implementation process of the processor 601 can be found in the above method embodiment. Its implementation principle and technical effects are similar and will not be repeated here in this embodiment.

[0188] In the above embodiments, it should be understood that the processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the present invention may be directly implemented by a hardware processor or implemented by a combination of hardware and software modules in the processor.

[0189] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include a non-volatile memory (NVM), such as at least one disk memory.

[0190] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be classified into address buses, data buses, and control buses. For ease of illustration, the buses in the drawings of this application are not limited to just one bus or just one type of bus.

[0191] The present application also provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0192] The present application also provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the above method is implemented.

[0193] The above-mentioned readable storage medium can be implemented by any type of volatile or non-volatile memory device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The readable storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0194] An exemplary readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be an integral part of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (ASIC). Of course, the processor and the readable storage medium can also exist in the device as discrete components.

[0195] The division of units is merely a logical functional division; actual implementations may employ alternative divisions, such as combining or integrating multiple units or components into another system, or omitting or disabling certain features. Furthermore, any direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between devices or units, either through an interface, electrical, mechanical, or other means.

[0196] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0197] 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.

[0198] If the function 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, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0199] Those skilled in the art will appreciate that all or part of the steps in the above-described method embodiments can be implemented using hardware associated with program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0200] Finally, it should be noted that those skilled in the art will readily identify other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art not disclosed herein. The present invention is not limited to the precise structure described above and illustrated in the accompanying drawings, and various modifications and variations may be made without departing from the scope thereof. The scope of the present invention is limited solely by the appended claims.

Claims

1. A distribution network energy storage optimization method, characterized in that: include: Obtain historical data of distributed photovoltaics; Constructing a distributed photovoltaic output uncertainty model based on the historical data and the connection function Copula; Establishing an objective function according to the output uncertainty model; Using a relaxation algorithm to process the nonlinear characteristics of the objective function to obtain a linear reconstruction feature; The objective function is solved according to the linear reconstruction characteristics, and the energy storage configuration of the distribution network is optimized according to the solution result.

2. The method according to claim 1, characterized in that The method of constructing a distributed photovoltaic output uncertainty model based on the historical data and the connection function Copula includes: determining a fuzzy set based on the historical data and the Wasserstein distance; defining Copula parameters based on the historical data, wherein the Copula parameters represent the dependency relationship of distributed photovoltaic output characteristics; The fuzzy set and the Copula parameters are fused, and a distributed photovoltaic output uncertainty model is constructed according to the fusion result.

3. The method according to claim 2, characterized in that Determining a fuzzy set based on the historical data and the Wasserstein distance includes: defining an empirical distribution characterizing the distributed photovoltaic output characteristics based on the historical data; Calculating the Wasserstein distance between the empirical distribution and the target distribution; A fuzzy set representing the probability distribution of distributed photovoltaic output is defined according to the Wasserstein distance.

4. The method according to any one of claims 1 to 3, characterized in that The establishing of the objective function according to the output uncertainty model includes: Obtaining decision variables for distribution network energy storage optimization, the decision variables including energy storage planning and scheduling decisions; Determining the worst probability distribution of distributed photovoltaic output according to the output uncertainty model; An objective function is established with the goal of minimizing the sum of the energy storage planning and the scheduling decision under the worst probability distribution.

5. The method according to claim 4, characterized in that The nonlinear characteristics of the objective function are processed using a relaxation algorithm to obtain linear reconstruction characteristics, including: Detecting features with strong duality in the objective function through convex optimization conditions and obtaining corresponding cone reconstruction results, wherein the cone reconstruction results include nonlinear constraints; A relaxation algorithm is used to perform linear reconstruction on the nonlinear constraint, and a linear reconstruction result is obtained as a linear reconstruction feature.

6. The method according to any one of claims 1 to 3, characterized in that The constraint conditions of the objective function include at least one of energy storage planning constraint, energy storage operation constraint, photovoltaic output constraint, power balance constraint and distributed robustness opportunity constraint.

7. A distribution network energy storage optimization device, characterized in that: include: Acquisition module, used to obtain historical data of distributed photovoltaic; A model building module, configured to build a distributed photovoltaic output uncertainty model based on the historical data and the connection function Copula; A function generation module, configured to establish an objective function according to the output uncertainty model; A linear processing module, configured to process the nonlinear characteristics of the objective function using a relaxation algorithm to obtain linear reconstruction characteristics; An optimization module is used to solve the objective function according to the linear reconstruction characteristics and optimize the energy storage configuration of the distribution network according to the solution result.

8. An electronic device, characterized in that: include: a processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 6 when executed by a processor.

10. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 6 when being executed by a processor.

Images