Study on Siting and Capacity Determination of Energy Storage in Distribution Network Based on Improved MOPSO Algorithm

With the large-scale access of distributed new energy to the distribution network, the intermittency and stochasticity of its output seriously affect the stability and economy of system operation. Aiming at this problem, this paper constructs a multi-objective optimization model containing load fluctuation, active network loss and storage capacity minimisation, and proposes an energy storage siting and capacity determination method based on the improved multi-objective particle swarm optimization algorithm. The method significantly improves the uniformity of population distribution by introducing the initialisation strategy combining Tent chaotic mapping and Latin hypercubic sampling; adopts adaptive inertia weight and learning factor dynamic adjustment mechanism to effectively balance the algorithm's ability of global exploration and local development; and integrates entropy weight and TOPSIS methods to construct a multi-objective decision-making framework, realising the objective screening of the optimal Pareto solution. Based on the IEEE-33 node system simulation, compared with the standard MOPSO algorithm, this method effectively reduces the load fluctuation and network loss, optimises the energy storage allocation, and balances the grid operation characteristics and the investment cost of energy storage.