A Two-Stage Optimal Allocation Method For Energy Storage In Active Distribution Networks Based On Improved Butterfly Optimization Algorithm

Under the "two-carbon" vision,integrating a significant amount of clean energy into the distribution network increases grid tide uncertainty,compromising voltage and frequency stability and threatening power supply security.Energy storage plays a crucial role in accommodating high levels of new energy in the power system.Accordingly,China actively develops energy storage power plants,improves the energy storage system’s participation in power market trading,and diversifies the operation mode of these plants.However,ensuring grid safety remains a fundamental requirement.Key factors for the effective support of energy storage include planning storage capacity and location,expanding profit models,and shortening investment recovery cycles.Consequently,this study focuses on the optimal configuration strategy of energy storage plants in active distribution networks,summarized as follows:(1)investigate the development status of energy storage at home and abroad from the levels of technology,construction and commercial policies respectively;elaborate the impact of different capacities of energy storage connected to different locations of the active distribution network on voltage and network loss,and analyze the importance of reasonable configuration of energy storage;introduce the vulnerability assessment method of the active distribution network,apply the vulnerability evaluation method to the feasibility of energy storage planning,and provide the theoretical basis for the configuration;use k means clustering to obtain typical scenarios to address the impact of source-load uncertainty on energy storage configuration in active distribution grids to improve model rationality and reduce computational complexity.(2)To address the shortcomings of the butterfly optimization algorithm for solving energy storage allocation problems that fall into local optimality too early,two improvement strategies are used to enhance the global optimization-seeking capability of the algorithm,improve the solution accuracy,and meet the requirements of solving nonlinear problems.The dynamic switching probability is established to coordinate the global search and local search ability of the algorithm in the early and late stages;the dynamic Gaussian variation strategy is used to improve the search range of the algorithm based on the number of iterations and reduce the probability of falling into the local optimum.From the perspective of the whole life cycle,we model the various expenditures and direct and indirect benefits of energy storage for active distribution network operators,and use the typical daily charging and discharging behavior of energy storage and the installation node as optimization variables to verify the effectiveness of the proposed model for optimal allocation of energy storage in active distribution networks with the improved algorithm.(3)In order to reduce the impact of the source-load characteristics of energy storage on the operational safety of the active distribution network and to enhance the economic returns of energy storage investors,a two-stage energy storage optimal allocation model is proposed.In the first stage,the node vulnerability assessment is used to reflect the importance of each node to the operational security of the grid,and the weak links of the grid are screened as nodes for energy storage placement;the global vulnerability index is used to quantify the impact of energy storage access on the overall security of the grid,and the planning scheme that does not meet the index improvement requirements is eliminated;in the second stage,a two-tier optimization method is used to plan the energy storage capacity,and the upper model uses the maximum profit of equal annual value for the configuration scheme of energy storage The lower model develops the operation strategy of energy storage for energy arbitrage and participation in peak and frequency regulation auxiliary market trading to improve the economics of energy storage applications,and returns the operation revenue to the upper layer,which is solved by the particle swarm algorithm and the improved butterfly algorithm according to the respective characteristics of the twolayer model.The two phases jointly determine the energy storage allocation strategy.The extended IEEE-33 node model simulation verifies that the improved butterfly optimization algorithm achieves faster convergence,better fitness values,and applicable energy storage planning results across various scenarios.The proposed two-stage model reduces grid vulnerability,shortens payback time,and fulfills safety and economic requirements for energy storage configuration,serving as a reference for constructing energy storage plants in active distribution networks.