A method, system and device for double-layer optimization configuration of network-type energy storage power station considering source-load price uncertainty and a storage medium

Abstract

The application discloses a kind of network type energy storage power station double-layer optimization configuration method, system, equipment and storage medium considering source load price uncertainty, it is related to energy storage optimization configuration field, method includes: the data obtained is normalized and cluster generates typical operating scenario and probability;Based on scenario and probability, establish double-layer optimization model, upper layer determines energy storage capacity, lower layer optimization access position, adopt non-dominated sorting genetic algorithm to solve and iteratively update Pareto solution set;After convergence, Pareto solution set is comprehensively evaluated, and the optimal access node position, rated power and rated energy capacity are output;The application can effectively deal with multiple uncertainties of source load price, while reducing energy storage investment and operating cost, improve the uniformity of power grid short-circuit ratio and operating stability, realize economic and safety win-win.

Description

Technical Field

[0001] This invention relates to the field of energy storage optimization configuration, and in particular to a two-layer optimization configuration method, system, equipment and storage medium for grid-type energy storage power stations that takes into account the uncertainty of source-load price. Background Technology

[0002] With the advancement of the global energy transition, the penetration rate of new energy sources such as wind power and photovoltaics in the power grid is constantly increasing. However, the strong randomness and volatility of new energy power generation, along with load fluctuations caused by demand-side response and electric vehicles, as well as price fluctuations under the spot trading mechanism of the electricity market, pose significant challenges to the safe and stable operation of the power system and the economic viability of energy storage.

[0003] Energy storage systems employ grid-type converters to simulate the operating characteristics of synchronous generators, providing short-circuit capacity, voltage support, and virtual inertia to the grid, thus becoming a key means to improve the stability of high-proportion renewable energy power systems. However, current optimization configuration methods for energy storage mostly focus on single deterministic scenarios, failing to fully consider the impact of multiple uncertainties in the source-load-price relationship. Furthermore, in terms of optimization architecture, existing single-layer optimization or conventional configuration methods struggle to balance the complex relationship between the global economic capacity of the energy storage system and the local grid support location, resulting in the inability to maximize the economic value and grid support potential of energy storage. Simultaneously, with the increasing proportion of renewable energy, the phenomena of decreasing short-circuit ratios and uneven distribution in the system become more prominent. The virtual synchronous grid-type converter control method can be used to improve the uniformity of the system's short-circuit ratio and distribution. Grid-type energy storage power stations can effectively address the problems faced by current high-proportion renewable energy power systems. Several existing technologies also involve optimization configuration methods for grid-type energy storage power stations. For example, by establishing multiple power station nodes, constructing a multi-objective function based on cost, renewable energy utilization rate, and transient stability effect, and using genetic algorithms and grey relational projection methods to finely allocate the energy storage capacity of each power station, this method has made some progress in multi-objective coordination, but its applicability in a high-proportion renewable energy grid environment is insufficient. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention provides a two-layer optimization configuration method, system, equipment and storage medium for grid-type energy storage power stations that takes into account the uncertainty of source-load price.

[0005] Therefore, the technical problem solved by this invention is: how to simultaneously cope with the multiple uncertainties of source, load and price under the background of high proportion of new energy access, and balance the economic configuration of grid-type energy storage power station with grid support performance.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a two-layer optimization configuration method for grid-connected energy storage power stations that takes into account the uncertainty of source-load prices, including: Acquire regional power grid topology parameters, historical renewable energy output data, historical load data, and electricity spot market node price data, and set parameters for K-means clustering algorithm and non-dominated sorting genetic algorithm; The acquired data is normalized to construct a scenario vector containing source output, load and electricity price sequences. The K-means clustering algorithm is used to generate typical operating scenarios and calculate the probability of each scenario. Based on typical operating scenarios and corresponding scenario probabilities, a two-level optimization model is established and the decision variables of the upper-level optimization model and the lower-level optimization model are determined respectively. The initial candidate solution set is generated by using a non-dominated sorting genetic algorithm. Substitute the initial candidate solution set into the two-layer optimization model for iterative solution and update the Pareto solution set; If the preset convergence accuracy or maximum number of iterations is met, the current Pareto optimal solution set is output; otherwise, the feedback data from the lower layer is used to correct the capacity scheme of the upper layer, and selection, crossover and mutation operations are performed on the current population to update the energy storage capacity and access node position before iterating again to solve the problem. The Pareto optimal solution set is comprehensively evaluated, and the solution with the highest comprehensive score is selected as the final scheme. The optimal access node location, rated power capacity, and rated energy capacity of the grid-type energy storage power station are output.

[0007] As a preferred scheme of a two-layer optimal configuration method for grid-connected energy storage power stations that takes into account the uncertainty of source-load prices, wherein: The process involves normalizing the acquired data, constructing a scenario vector containing source output, load, and electricity price sequences, and using the K-means clustering algorithm to generate typical operating scenarios and calculate the probability of each scenario, including: Historical renewable energy output data, historical load data, and historical electricity price data are organized according to a unified time scale, and the organized data is normalized. The normalized source output, load and electricity price data at each moment are combined into a scene vector to form a high-dimensional feature space that represents multi-dimensional time series features. Clustering algorithms are used to iteratively cluster all scene vectors in the high-dimensional feature space, grouping original scenes with similar features into the same category, with each category corresponding to a typical running scene; Calculate the probability of occurrence of the corresponding typical scene based on the proportion of the original scene in each category to the total number of scenes; The final output is a probability-weighted set of typical scenarios consisting of each typical scenario and its corresponding probability.

[0008] As a preferred scheme of a two-layer optimal configuration method for grid-connected energy storage power stations that takes into account the uncertainty of source-load prices, wherein: Based on typical operating scenarios and corresponding scenario probabilities, a two-layer optimization model is established, and the decision variables of the upper and lower optimization models are determined respectively. An initial candidate solution set is generated using a non-dominated sorting genetic algorithm, including: Based on typical operating scenarios and their corresponding probabilities of occurrence, a two-layer optimization model is constructed, which includes an upper-layer capacity planning model and a lower-layer location selection and operation model. The decision variables for the upper-level model are determined to be the rated power capacity and rated energy capacity of the grid-type energy storage power station, and the decision variables for the lower-level model are determined to be the location of the access node of the grid-type energy storage power station.

[0009] As a preferred scheme of a two-layer optimal configuration method for grid-connected energy storage power stations that takes into account the uncertainty of source-load prices, wherein: The process of establishing a two-layer optimization model based on typical operating scenarios and corresponding scenario probabilities, determining the decision variables for the upper and lower optimization models respectively, and generating an initial candidate solution set using a non-dominated sorting genetic algorithm, further includes: The information transmission relationship between the upper-level model and the lower-level model is defined. The upper-level model transmits the capacity scheme determined by the decision variables to the lower-level model as a constraint. The lower-level model transmits the optimized operating benefit index and system strength index to the upper-level model as feedback information to guide the upper-level model to revise the capacity scheme and establish a feedback iteration mechanism between the two layers. An initial population is generated using a non-dominated sorting genetic algorithm. Each individual in the initial population contains a combination of energy storage capacity and access node location. A candidate solution set consisting of multiple such combinations is constructed as the initial solution space for two-level iterative optimization.

[0010] As a preferred scheme of a two-layer optimal configuration method for grid-connected energy storage power stations that takes into account the uncertainty of source-load prices, wherein: The step of substituting the initial candidate solution set into the two-layer optimization model for iterative solution and updating the Pareto solution set includes: A non-dominated sorting genetic algorithm was used to solve the upper and lower models in the bi-level optimization model, respectively. The upper-level model aims to minimize the annualized comprehensive cost of the energy storage system and the power penalty of energy storage operation. It optimizes the rated power capacity and rated energy capacity of energy storage under the constraints of grid power balance conditions and the operating characteristics of energy storage itself, and outputs a capacity configuration scheme. The capacity configuration scheme is then passed to the lower-level model. The lower-level model takes capacity configuration scheme as constraint, maximizes the uniformity of the equivalent short-circuit ratio distribution of each node in the regional power grid and maximizes the operating benefits of the energy storage system as optimization objectives, and takes power system operation safety and the dynamic support characteristics of grid-type energy storage as constraints to optimize the access node location of energy storage and output the site selection scheme.

[0011] As a preferred scheme of a two-layer optimal configuration method for grid-connected energy storage power stations that takes into account the uncertainty of source-load prices, wherein: The step of substituting the initial candidate solution set into the two-layer optimization model for iterative solution and updating the Pareto solution set further includes: Multi-scenario power flow calculations are performed on the location schemes obtained by optimizing the lower-level model. Based on the calculation results, the equivalent short-circuit ratio distribution uniformity index and the operation benefit index are updated and fed back to the upper-level model. The upper-level model adjusts the current capacity configuration scheme based on the feedback indicators.

[0012] As a preferred scheme of a two-layer optimal configuration method for grid-connected energy storage power stations that takes into account the uncertainty of source-load prices, wherein: The response is to output the current Pareto optimal solution set if the preset convergence accuracy or maximum number of iterations is met; otherwise, the lower-level feedback data is used to correct the upper-level capacity scheme, and selection, crossover, and mutation operations are performed on the current population to update the energy storage capacity and access node positions before iterating again, including: If the Pareto front solution set of the current population meets the preset convergence accuracy requirement or reaches the preset maximum number of iterations, stop iterating and output the current Pareto optimal solution set; Otherwise, the operational benefit indicators and system strength indicators obtained from the optimization of the lower-level model are fed back to the upper-level model, and the upper-level model corrects the current capacity configuration scheme based on the received feedback indicators; The selection, crossover and mutation operations of the non-dominated sorting genetic algorithm are performed on the current population to retain superior individuals and generate new candidate individuals, and update the energy storage capacity and access node position in each candidate solution. The updated population is then substituted back into the bi-level optimization model to initiate a new round of iterative solutions.

[0013] Secondly, the present invention provides a two-layer optimized configuration system for grid-connected energy storage power stations that takes into account the uncertainty of source-load prices, including: The basic data acquisition and algorithm parameter initialization module is used to acquire regional power grid topology parameters, historical renewable energy output data, historical load data, and electricity spot market node price data, and to set the parameters of the K-means clustering algorithm and the non-dominated sorting genetic algorithm. The typical scenario generation module for source-load price is used to normalize the acquired data, construct a scenario vector containing source output, load and electricity price sequences, and use the K-means clustering algorithm to generate typical operating scenarios and calculate the probability of each scenario. The two-level optimization modeling and initial solution generation module is used to establish a two-level optimization model based on typical operating scenarios and corresponding scenario probabilities, and to determine the decision variables of the upper-level optimization model and the lower-level optimization model respectively, and to generate an initial candidate solution set using a non-dominated sorting genetic algorithm. The bi-layer iterative solution and Pareto solution update module is used to substitute the initial candidate solution set into the bi-layer optimization model for iterative solution and update the Pareto solution set. The convergence determination and population update module is used to output the current Pareto optimal solution set in response to meeting the preset convergence accuracy or maximum number of iterations; otherwise, it uses the feedback data from the lower layer to correct the upper layer capacity scheme, and performs selection, crossover and mutation operations on the current population to update the energy storage capacity and access node positions before iterating again to solve the problem. The optimal solution comprehensive evaluation and output module is used to comprehensively evaluate the Pareto optimal solution set, select the solution with the highest comprehensive score as the final solution, and output the optimal access node location, rated power capacity and rated energy capacity of the grid-type energy storage power station.

[0014] Thirdly, the present invention provides a computer device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the two-layer optimization configuration method for grid-type energy storage power stations that takes into account the uncertainty of source-load price.

[0015] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of a two-layer optimal configuration method for a grid-type energy storage power station that takes into account the uncertainty of source-load price.

[0016] The beneficial effects of this invention are as follows: Addressing the multiple uncertainties arising from the random power output, frequent load fluctuations, and drastic changes in spot electricity prices under high-proportion renewable energy integration, this invention achieves a balance between the economic efficiency and grid support performance of grid-connected energy storage power stations in actual power systems. By processing historical wind and solar power, load, and electricity price data through deviation standardization and clustering algorithms, representative typical operating scenarios are generated, enabling energy storage configurations to accurately reflect the combined impact of multiple uncertainties and avoiding significant deviations between configuration results and actual operation in a single scenario. A two-layer optimization architecture is adopted. The upper layer aims to minimize the annualized cost and operating power penalty of the energy storage system throughout its entire lifecycle, effectively controlling energy storage investment and maintenance costs while suppressing the adverse effects of frequent high-power charging and discharging on equipment lifespan. The lower layer aims to improve the uniformity of the equivalent short-circuit ratio distribution at various nodes in the regional power grid and maximize the operating benefits of energy storage, optimizing the energy storage access location and effectively improving the weak link of low and unevenly distributed short-circuit ratios in high-proportion renewable energy power grids. Through a two-layer feedback iteration, the resulting capacity and location scheme significantly improves the equivalent short-circuit ratio of each node and makes its distribution more uniform in a real power grid, thereby enhancing the system's voltage support capability and operational stability. This invention can be directly applied to energy storage planning in real-world power grids with high proportions of renewable energy, balancing both economic investment and safety and stability requirements. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is an overall flowchart of a two-layer optimization configuration method for a grid-type energy storage power station that takes into account the uncertainty of source-load price, provided by the present invention.

[0019] Figure 2 This is a simulation example of a two-layer optimization configuration method for a grid-type energy storage power station that takes into account the uncertainty of source-load price, based on the present invention, showing typical scenarios of new energy output, load power, and electricity price.

[0020] Figure 3 This is a heat map of the short-circuit ratio distribution uniformity index of a grid-type energy storage power station connected to each node in a simulation example of a two-layer optimization configuration method for grid-type energy storage power stations that takes into account the uncertainty of source-load price, provided by the present invention.

[0021] Figure 4This is a simulation example of a two-layer optimization configuration method for a grid-type energy storage power station that takes into account the uncertainty of source-load price, showing a comparison of the equivalent short-circuit ratios of nodes in the grid-type energy storage power station connected to the optimal scheme and nodes without energy storage in the simulation example. Detailed Implementation

[0022] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0023] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a two-layer optimal configuration method for grid-connected energy storage power stations that takes into account the uncertainty of source-load prices, including: S1: Obtain regional power grid topology parameters, historical renewable energy output data, historical load data, and electricity spot market node price data, and set the parameters for the K-means clustering algorithm and the non-dominated sorting genetic algorithm; S2: Normalize the acquired data, construct a scenario vector containing source output, load and electricity price sequences, use K-means clustering algorithm to generate typical operating scenarios and calculate the probability of each scenario; S3: Based on typical operating scenarios and corresponding scenario probabilities, establish a two-level optimization model and determine the decision variables of the upper-level optimization model and the lower-level optimization model respectively. Use a non-dominated sorting genetic algorithm to generate an initial candidate solution set. S4: Substitute the initial candidate solution set into the bi-layer optimization model for iterative solution and update the Pareto solution set; S5: In response to meeting the preset convergence accuracy or maximum number of iterations, output the current Pareto optimal solution set; otherwise, use the feedback data from the lower layer to correct the upper layer capacity scheme, and at the same time perform selection, crossover and mutation operations on the current population, update the energy storage capacity and access node position, and then iterate to solve again. S6: Perform a comprehensive evaluation of the Pareto optimal solution set, select the solution with the highest comprehensive score as the final scheme, and output the optimal access node location, rated power capacity, and rated energy capacity of the grid-type energy storage power station.

[0024] It should be noted that, through steps S1-S6, this invention achieves coordinated optimization of capacity determination and access location addressing for grid-type energy storage power stations under the conditions of source output randomness, load fluctuation and spot electricity price uncertainty, and finally outputs the optimal configuration scheme that takes into account both the economic efficiency of the entire life cycle of energy storage and the regional power grid strength support performance.

[0025] Example 2, refer to Figure 1 As an embodiment of the present invention, based on the previous embodiment, a two-layer optimization configuration method for grid-connected energy storage power stations considering source-load price uncertainty is provided, including: In this embodiment, the steps S1 described above, including obtaining regional power grid topology parameters, historical renewable energy output data, historical load data, and electricity spot market node price data, and setting parameters for the K-means clustering algorithm and the non-dominated sorting genetic algorithm, include: Obtain the basic data and algorithm parameters required for the two-layer optimal configuration of a grid-type energy storage power station.

[0026] The basic data includes regional power grid topology parameters, historical renewable energy output data, historical load data, and electricity spot market node price data; the algorithm parameters include the number of clusters and maximum number of iterations for the K-means clustering algorithm, and the population size, crossover probability, and mutation probability for the non-dominated sorting genetic algorithm.

[0027] It should be noted that this step completes the unified initialization of multi-source data and optimization algorithms, providing data support and parameter basis for subsequent generation of typical scenarios and two-layer optimization calculations.

[0028] In this embodiment, the data obtained in step S2 above is normalized to construct a scenario vector containing source output, load, and electricity price sequences. The K-means clustering algorithm is used to generate typical operating scenarios and calculate the probability of each scenario, including: Historical renewable energy output data, load data, and electricity price data are organized according to a unified time scale.

[0029] Specifically, due to the significant differences in the dimensions and numerical magnitudes of new energy output, load, and electricity price, direct clustering would lead to the large numerical indicators completely dominating distance calculations. Therefore, a deviation standardization method is first used to map the source-load-price data to the [0, 1] per-unit value range to eliminate the influence of dimensions and prevent any one type of data from dominating the results during clustering. Each scenario vector contains the source-end output sequence, load power sequence, and electricity price sequence, thus forming a high-dimensional feature space for subsequent clustering analysis.

[0030] We obtain N original scenarios, each of which is a long vector containing source-load-price data information: In the formula, S i P is the vector of the i-th scene; res,1…T For the source-end output sequence; P load,1…T For load power sequence; q 1…T This is the electricity price series.

[0031] The K-means clustering algorithm is used to cluster the standardized scene vectors. By continuously updating the cluster centers, the original scene is divided into several categories, each corresponding to a typical operating scene.

[0032] Based on the proportion of samples in each category to the total number of samples, the probability of occurrence of the corresponding typical scenario is calculated, thereby constructing a probability-weighted set of typical scenarios. The final generated set of typical scenarios and its corresponding probability distribution are output, providing input data for the two-layer optimization model and achieving an effective characterization of multiple uncertainties in the source-load-price relationship.

[0033] In this embodiment, step S3 above, based on typical operating scenarios and corresponding scenario probabilities, establishes a two-layer optimization model and determines the decision variables for the upper and lower optimization models respectively, and uses a non-dominated sorting genetic algorithm to generate an initial candidate solution set, including: Based on typical operating scenarios and their corresponding probabilities, a two-level optimization model is established, and the decision variables in the two-level optimization model are initialized.

[0034] The two-layer optimization model consists of an upper-layer capacity planning model and a lower-layer location and operation model. The upper-layer model passes the calculated power and energy capacity as constraints to the lower-layer model. The lower-layer model feeds back the operating benefits and short-circuit ratio standard deviation to the upper-layer model to correct the capacity scheme, ultimately achieving a globally optimal solution through two-layer collaboration. The two layers iterate through a feedback mechanism.

[0035] Among them, the decision variable of the upper-level optimization model is the rated power capacity P of the grid-type energy storage power station. es and energy capacity E es The decision variable for the lower-level optimization model is the location of the access node of the energy storage power station.

[0036] Meanwhile, an initial population is generated using a non-dominated sorting genetic algorithm, and a candidate solution set consisting of multiple energy storage capacity and location combinations is constructed to provide an initial solution space for subsequent two-layer optimization iterations.

[0037] In this embodiment, step S4 above, which substitutes the initial candidate solution set into the bi-layer optimization model for iterative solution and updates the Pareto solution set, includes: For the multi-objective optimization of the upper and lower layers in the bi-level optimization model, the non-dominated sorting genetic algorithm is used to solve the problem. The optimization of the decision variables at each layer is completed through the iterative mechanism of the non-dominated sorting genetic algorithm.

[0038] Upper-level capacity planning represents the macro-decision-making layer in the two-layer optimization, and is responsible for determining the capacity of energy storage power stations from the perspective of the overall economic efficiency of the system.

[0039] The specific decision variables at the upper level are energy storage power capacity and energy capacity.

[0040] The upper-level optimization objectives are to minimize the annualized total cost of the energy storage system and the energy storage operating power penalty, so as to ensure the economic feasibility of the planning scheme.

[0041] Given the known capacity, the next-level site selection optimizes the specific access location and operation strategy of energy storage in the power grid.

[0042] The lower-level decision variable is the location of grid-type energy storage access.

[0043] The lower-level optimization objectives are to maximize the uniformity of the regional power grid short-circuit ratio distribution and to maximize the operating benefits of the energy storage system.

[0044] In the upper-level optimization model process, the operating benefits and short-circuit ratio standard deviation data of the optimal access location under the current capacity configuration are obtained from the feedback of the lower-level model. The objective function is constructed to minimize the annualized total cost of the energy storage system. Where F1 is the objective function for the annualized total cost of the energy storage system; B in B represents the annual value of the initial investment cost for energy storage. om is the annual operating and maintenance cost; w is the discount rate; n e For life cycle; a p and a e Cost per unit power and cost per unit capacity, respectively; K M K represents the annual fixed maintenance cost coefficient per unit power of energy storage; O This represents the variable operating cost coefficient per unit of energy storage; g s Let s be the probability of scenario s occurring. Let be the output power of energy storage at time t under scenario s; T is the total number of operating periods throughout the year; Δt represents the calculation time step for energy storage operation optimization.

[0045] Construct a second objective function by minimizing the power penalty of energy storage operation: Here, F2 is the objective function for energy storage operation power penalty, used to suppress frequent high-power charging and discharging behavior of energy storage. Based on this, constraints are imposed on the upper-level optimization, including grid power balance constraints, upper and lower limits of energy storage state of charge constraints, and their dynamic constraints. Then, the annualized cost objective and the operation power penalty objective are jointly evaluated, the fitness value of each individual is calculated, and a Pareto solution set is constructed based on a non-dominated sorting algorithm. Finally, the energy storage capacity configuration scheme obtained from the current optimization is output as the input parameters for the lower-level model.

[0046] In the lower-level optimization model process, the rated power and energy capacity parameters of the energy storage system output from the upper-level optimization model are first received and used as constraints. Then, with the objective of minimizing the standard deviation of the equivalent short-circuit ratio of each node in the regional power grid, the connection location of the grid-connected energy storage power station is optimized. Where G1 is the objective function of the standard deviation of the equivalent short-circuit ratio of regional power grid nodes; N b The total number of nodes; The equivalent short-circuit ratio of node i after configuring a grid-type energy storage power station; U Ni Let be the rated voltage of node i; Before grid-connected energy storage power stations are connected to the grid, the system impedance matrix is... Z EQ The element in row i and column j, , For the grid-connected energy storage power station access node t, in Z EQ The element in row t, column i, and column t; z 0 represents the equivalent impedance of the virtual impedance plus branch impedance generated by the network-type virtual synchronous control; P j This represents the active power of node j.

[0047] The objective function is constructed with the goal of maximizing the operational benefits of the energy storage system as follows: Wherein, G2 is the objective function for the operating revenue of the energy storage system; Let be the output power of energy stored at node i in scenario s at time t; Let be the electricity price of node i at time t in scenario s.

[0048] Based on this, multi-scenario power flow calculations are performed on each candidate node using typical scenario data to obtain the voltage, current, and power distribution of each node. Constraints related to power system operation and grid-based energy storage characteristics are then applied, including short-circuit ratio safety constraints, dynamic support constraints for the grid-based energy storage virtual synchronous machine, and power flow balance constraints. Next, the short-circuit ratio distribution uniformity objective and the operational benefit objective are jointly evaluated, the fitness value of each individual is calculated, and a Pareto solution set is constructed based on a non-dominated sorting algorithm. Finally, the site selection optimization results are fed back to the upper-level model to update the objective function evaluation of the two-layer optimization, achieving information interaction and closed-loop iterative optimization between the upper and lower layers.

[0049] In this embodiment, in step S5 above, in response to meeting the preset convergence accuracy or maximum number of iterations, the current Pareto optimal solution set is output; otherwise, the lower-level feedback data is used to correct the upper-level capacity scheme, and at the same time, selection, crossover, and mutation operations are performed on the current population to update the energy storage capacity and access node positions before iterating again to solve the problem. If the Pareto front solution set of the current population meets the preset accuracy requirement, or reaches the maximum number of iterations, output the current Pareto optimal solution set; If the convergence condition is not met, the operating benefits and short-circuit ratio standard deviation calculated by the lower-level model are fed back to the upper-level model, which then corrects the capacity scheme accordingly. At the same time, a non-dominated sorting genetic algorithm is used to perform selection, crossover, and mutation operations on the current population. By retaining superior individuals and generating new candidate solutions, the energy storage capacity and access node positions are updated, and the updated population is re-substituted into the two-level optimization model for iterative solution.

[0050] In this embodiment, step S6 above involves a comprehensive evaluation of the Pareto optimal solution set, selecting the solution with the highest comprehensive score as the final scheme. The optimal access node location, rated power capacity, and rated energy capacity of the grid-type energy storage power station are output, including: A comprehensive evaluation is performed on the final Pareto optimal solution set.

[0051] Specifically, firstly, the objective functions are made dimensionless, and weight coefficients are assigned to each indicator. Then, the Euclidean distance between each candidate scheme and the positive and negative ideal solutions is calculated to obtain the relative closeness of each configuration scheme. Finally, the solution with the highest comprehensive score is selected as the final optimal configuration scheme for grid-type energy storage, and its optimal access node location, rated power, and rated energy capacity are output.

[0052] Example 3 illustrates a schematic scheme of a two-layer optimal configuration method for a grid-connected energy storage power station considering source-load price uncertainty. It should be noted that the technical solution of a two-layer optimal configuration system for a grid-connected energy storage power station considering source-load price uncertainty is based on the same concept as the aforementioned two-layer optimal configuration method for a grid-connected energy storage power station considering source-load price uncertainty. Details not described in detail in the technical solution of the two-layer optimal configuration system for a grid-connected energy storage power station considering source-load price uncertainty in this embodiment can be found in the description of the aforementioned two-layer optimal configuration method for a grid-connected energy storage power station considering source-load price uncertainty.

[0053] This embodiment also provides a two-layer optimized configuration system for grid-connected energy storage power stations that takes into account the uncertainty of source-load prices, including: The basic data acquisition and algorithm parameter initialization module is used to acquire regional power grid topology parameters, historical renewable energy output data, historical load data, and electricity spot market node price data, and to set the parameters of the K-means clustering algorithm and the non-dominated sorting genetic algorithm. The typical scenario generation module for source-load price is used to normalize the acquired data, construct a scenario vector containing source output, load and electricity price sequences, and use the K-means clustering algorithm to generate typical operating scenarios and calculate the probability of each scenario. The two-level optimization modeling and initial solution generation module is used to establish a two-level optimization model based on typical operating scenarios and corresponding scenario probabilities, and to determine the decision variables of the upper-level optimization model and the lower-level optimization model respectively, and to generate an initial candidate solution set using a non-dominated sorting genetic algorithm. The bi-layer iterative solution and Pareto solution update module is used to substitute the initial candidate solution set into the bi-layer optimization model for iterative solution and update the Pareto solution set. The convergence determination and population update module is used to output the current Pareto optimal solution set in response to meeting the preset convergence accuracy or maximum number of iterations; otherwise, it uses the feedback data from the lower layer to correct the upper layer capacity scheme, and performs selection, crossover and mutation operations on the current population to update the energy storage capacity and access node positions before iterating again to solve the problem. The optimal solution comprehensive evaluation and output module is used to comprehensively evaluate the Pareto optimal solution set, select the solution with the highest comprehensive score as the final solution, and output the optimal access node location, rated power capacity and rated energy capacity of the grid-type energy storage power station.

[0054] This embodiment also provides a computer device applicable to a two-layer optimization configuration method for grid-connected energy storage power stations that takes into account the uncertainty of source-load prices, including: The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement a two-layer optimization configuration method for grid-type energy storage power stations that takes into account the uncertainty of source-load prices, as proposed in the above embodiments.

[0055] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a two-layer optimization configuration method for grid-type energy storage power stations that takes into account the uncertainty of source-load prices, as proposed in the above embodiment.

[0056] The storage medium proposed in this embodiment belongs to the same inventive concept as the two-layer optimization configuration method for grid-type energy storage power stations that takes into account the uncertainty of source-load price proposed in the above embodiment. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0057] Example 4, refer to Figures 2-4 As an embodiment of the present invention, a two-layer optimization configuration method for grid-type energy storage power stations that takes into account the uncertainty of source-load price is provided. In order to verify the beneficial effects of the present invention, a simulation experiment is conducted for scientific demonstration.

[0058] A simulation analysis was conducted on a power grid in a region with a high proportion of renewable energy. Based on the actual historical data of the power grid in this region, the simulation acquired grid topology parameters, historical wind and solar power output data, historical load data, and electricity spot market nodal price data, and organized them according to a unified time scale. The data was normalized using deviation standardization, and a high-dimensional scenario vector containing source-end output sequences, load power sequences, and price sequences was constructed. K-means clustering was used for cluster analysis to generate several typical operating scenarios, and the probability of each scenario was calculated. The results of generating typical scenarios are shown below. Figure 2 As shown in (a), (b), and (c), the time-series distribution characteristics of new energy output, load power, and electricity price under various typical scenarios are presented.

[0059] The generated typical scenarios and their probabilities are input into the constructed two-layer optimization model. The upper-layer model aims to minimize the annualized total cost of the energy storage system and the operating power penalty, and uses a non-dominated sorting genetic algorithm to optimize the rated power capacity and rated energy capacity of the energy storage system. The lower-layer model receives the capacity scheme output by the upper layer and optimizes the location of the energy storage access nodes with the aim of minimizing the standard deviation of the equivalent short-circuit ratio of each node in the regional power grid and maximizing the operating benefits of energy storage. The upper and lower layers iteratively solve the problem through a feedback mechanism, updating the Pareto solution set round by round.

[0060] To evaluate the impact of grid-connected energy storage power stations on the system intensity distribution by connecting them to different nodes, energy storage power stations were sequentially configured at each candidate node of the regional power grid. The equivalent short-circuit ratio of each node under each connection scheme was recorded, and its standard deviation was calculated. The heat index of the uniformity of the equivalent short-circuit ratio distribution after the grid-connected energy storage power stations are connected to each node is as follows: Figure 3 As shown in the figure, the color intensity reflects the degree to which the equivalent short-circuit ratio of each node deviates from the mean. The darker the color, the more the system strength deviates from the average level at that node; the lighter the color, the more uniform the distribution of the short-circuit ratio among the nodes. Figure 3 It can be seen that when the energy storage is connected to node 10, the short-circuit ratio distribution is the most uniform and the overall deviation is the smallest.

[0061] After convergence through a two-layer feedback iteration, the obtained Pareto optimal solution set is comprehensively evaluated using the superior-inferior solution distance method. Each objective function is dimensionless and assigned weight coefficients. The Euclidean distance between each candidate solution and the positive and negative ideal solutions is calculated to obtain the relative proximity, and the solutions are ranked accordingly. The solution with the highest score is selected as the optimal configuration. The final optimal configuration scheme for the grid-connected energy storage power station is: energy capacity 514MWh, power capacity 432MW, and 10 access nodes.

[0062] To visually reflect the optimization effect, the equivalent short-circuit ratio of each node under the optimal configuration scheme is compared with that of each node without energy storage. The results are as follows: Figure 4 As shown. By Figure 4 It is evident that without energy storage, the short-circuit ratios of some nodes are low, and the distribution among nodes varies significantly, resulting in a prominent problem of uneven system strength distribution. After connecting to an optimally configured grid-type energy storage power station, the overall equivalent short-circuit ratio of each node is significantly improved, and the distribution among nodes becomes more uniform. This result demonstrates that the method of this invention can effectively improve the voltage support capability and operational stability of high-proportion renewable energy power systems.

[0063] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A two-layer optimal configuration method for grid-connected energy storage power stations considering source-load price uncertainty, characterized in that, include: Acquire regional power grid topology parameters, historical renewable energy output data, historical load data, and electricity spot market node price data, and set parameters for K-means clustering algorithm and non-dominated sorting genetic algorithm; The acquired data is normalized to construct a scenario vector containing source output, load and electricity price sequences. The K-means clustering algorithm is used to generate typical operating scenarios and calculate the probability of each scenario. Based on typical operating scenarios and corresponding scenario probabilities, a two-level optimization model is established and the decision variables of the upper-level optimization model and the lower-level optimization model are determined respectively. The initial candidate solution set is generated by using a non-dominated sorting genetic algorithm. Substitute the initial candidate solution set into the two-layer optimization model for iterative solution and update the Pareto solution set; If the preset convergence accuracy or maximum number of iterations is met, the current Pareto optimal solution set is output; otherwise, the feedback data from the lower layer is used to correct the capacity scheme of the upper layer, and selection, crossover and mutation operations are performed on the current population to update the energy storage capacity and access node position before iterating again to solve the problem. The Pareto optimal solution set is comprehensively evaluated, and the solution with the highest comprehensive score is selected as the final scheme. The optimal access node location, rated power capacity, and rated energy capacity of the grid-type energy storage power station are output.

2. The two-layer optimal configuration method for grid-connected energy storage power stations considering source-load price uncertainty as described in claim 1, characterized in that, The process involves normalizing the acquired data, constructing a scenario vector containing source output, load, and electricity price sequences, and using the K-means clustering algorithm to generate typical operating scenarios and calculate the probability of each scenario, including: Historical renewable energy output data, historical load data, and historical electricity price data are organized according to a unified time scale, and the organized data is normalized. The normalized source output, load and electricity price data at each moment are combined into a scene vector to form a high-dimensional feature space that represents multi-dimensional time series features. Clustering algorithms are used to iteratively cluster all scene vectors in the high-dimensional feature space, grouping original scenes with similar features into the same category, with each category corresponding to a typical running scene; Calculate the probability of occurrence of the corresponding typical scene based on the proportion of the original scene in each category to the total number of scenes; The final output is a probability-weighted set of typical scenarios consisting of each typical scenario and its corresponding probability.

3. The two-layer optimal configuration method for grid-type energy storage power stations considering source-load price uncertainty as described in claim 2, characterized in that, Based on typical operating scenarios and corresponding scenario probabilities, a two-layer optimization model is established, and the decision variables of the upper and lower optimization models are determined respectively. An initial candidate solution set is generated using a non-dominated sorting genetic algorithm, including: Based on typical operating scenarios and their corresponding probabilities of occurrence, a two-layer optimization model is constructed, which includes an upper-layer capacity planning model and a lower-layer location selection and operation model. The decision variables for the upper-level model are determined to be the rated power capacity and rated energy capacity of the grid-type energy storage power station, and the decision variables for the lower-level model are determined to be the location of the access node of the grid-type energy storage power station.

4. The two-layer optimal configuration method for grid-connected energy storage power stations considering source-load price uncertainty as described in claim 3, characterized in that, The process of establishing a two-layer optimization model based on typical operating scenarios and corresponding scenario probabilities, determining the decision variables for the upper and lower optimization models respectively, and generating an initial candidate solution set using a non-dominated sorting genetic algorithm, further includes: The information transmission relationship between the upper-level model and the lower-level model is defined. The upper-level model transmits the capacity scheme determined by the decision variables to the lower-level model as a constraint. The lower-level model transmits the optimized operating benefit index and system strength index to the upper-level model as feedback information to guide the upper-level model to revise the capacity scheme and establish a feedback iteration mechanism between the two layers. An initial population is generated using a non-dominated sorting genetic algorithm. Each individual in the initial population contains a combination of energy storage capacity and access node location. A candidate solution set consisting of multiple such combinations is constructed as the initial solution space for two-level iterative optimization.

5. The two-layer optimal configuration method for grid-connected energy storage power stations considering source-load price uncertainty as described in claim 4, characterized in that, The step of substituting the initial candidate solution set into the two-layer optimization model for iterative solution and updating the Pareto solution set includes: A non-dominated sorting genetic algorithm was used to solve the upper and lower models in the bi-level optimization model, respectively. The upper-level model aims to minimize the annualized comprehensive cost of the energy storage system and the power penalty of energy storage operation. It optimizes the rated power capacity and rated energy capacity of energy storage under the constraints of grid power balance conditions and the operating characteristics of energy storage itself, and outputs a capacity configuration scheme. The capacity configuration scheme is then passed to the lower-level model. The lower-level model takes capacity configuration scheme as constraint, maximizes the uniformity of the equivalent short-circuit ratio distribution of each node in the regional power grid and maximizes the operating benefits of the energy storage system as optimization objectives, and takes power system operation safety and the dynamic support characteristics of grid-type energy storage as constraints to optimize the access node location of energy storage and output the site selection scheme.

6. The two-layer optimal configuration method for grid-connected energy storage power stations considering source-load price uncertainty as described in claim 5, characterized in that, The step of substituting the initial candidate solution set into the two-layer optimization model for iterative solution and updating the Pareto solution set further includes: Multi-scenario power flow calculations are performed on the location schemes obtained by optimizing the lower-level model. Based on the calculation results, the equivalent short-circuit ratio distribution uniformity index and the operation benefit index are updated and fed back to the upper-level model. The upper-level model adjusts the current capacity configuration scheme based on the feedback indicators.

7. The two-layer optimal configuration method for grid-connected energy storage power stations considering source-load price uncertainty as described in claim 6, characterized in that, The response is to output the current Pareto optimal solution set if the preset convergence accuracy or maximum number of iterations is met; otherwise, the lower-level feedback data is used to correct the upper-level capacity scheme, and selection, crossover, and mutation operations are performed on the current population to update the energy storage capacity and access node positions before iterating again, including: If the Pareto front solution set of the current population meets the preset convergence accuracy requirement or reaches the preset maximum number of iterations, stop iterating and output the current Pareto optimal solution set; Otherwise, the operational benefit indicators and system strength indicators obtained from the optimization of the lower-level model are fed back to the upper-level model, and the upper-level model corrects the current capacity configuration scheme based on the received feedback indicators; The selection, crossover and mutation operations of the non-dominated sorting genetic algorithm are performed on the current population to retain superior individuals and generate new candidate individuals, and update the energy storage capacity and access node position in each candidate solution. The updated population is then substituted back into the bi-level optimization model to initiate a new round of iterative solutions.

8. A two-layer optimal configuration system for a grid-connected energy storage power station considering the uncertainty of source-load prices, using the method described in any one of claims 1 to 7, characterized in that, include: The basic data acquisition and algorithm parameter initialization module is used to acquire regional power grid topology parameters, historical renewable energy output data, historical load data, and electricity spot market node price data, and to set the parameters of the K-means clustering algorithm and the non-dominated sorting genetic algorithm. The typical scenario generation module for source-load price is used to normalize the acquired data, construct a scenario vector containing source output, load and electricity price sequences, and use the K-means clustering algorithm to generate typical operating scenarios and calculate the probability of each scenario. The two-level optimization modeling and initial solution generation module is used to establish a two-level optimization model based on typical operating scenarios and corresponding scenario probabilities, and to determine the decision variables of the upper-level optimization model and the lower-level optimization model respectively, and to generate an initial candidate solution set using a non-dominated sorting genetic algorithm. The bi-layer iterative solution and Pareto solution update module is used to substitute the initial candidate solution set into the bi-layer optimization model for iterative solution and update the Pareto solution set. The convergence determination and population update module is used to output the current Pareto optimal solution set in response to meeting the preset convergence accuracy or maximum number of iterations; otherwise, it uses the feedback data from the lower layer to correct the upper layer capacity scheme, and performs selection, crossover and mutation operations on the current population to update the energy storage capacity and access node positions before iterating again to solve the problem. The optimal solution comprehensive evaluation and output module is used to comprehensively evaluate the Pareto optimal solution set, select the solution with the highest comprehensive score as the final solution, and output the optimal access node location, rated power capacity and rated energy capacity of the grid-type energy storage power station.

9. A computer device, characterized in that, include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores computer-executable instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.

Images