Optimal Sizing Considering Power Uncertainty and Power Supply Reliability Based on LSTM and MOPSO for SWPBMs

Abstract

For independent islands with poor grid and traffic facilities, the stand-alone photovoltaic/wind/battery microgrid (SWPBM) is an effective way to solve its power supply poser via integrating its local renewable energy sources. This article proposes a size planning scheme based on power prediction to ensure the economy, reliability, and environmental-friendly of SWPBMs. A new objective function covering the dynamic investment payback period and excess energy is designed to achieve the tradeoff between the economy and the utilization rate of RE in SWPBMs. Strict uninterrupted power supply and state of charge range of battery are taken as constraints of optimal problem to improve the reliability of power supply of the system. Moreover, a long short-term memory based power prediction method with prediction error correction is proposed to capture the uncertainty of PV, wind turbine, and load power, which is conducive to making a reliable size planning decision. A nested multiobjective particle swarm optimization algorithm, whose computational complexity is greatly reduced by grouping the net load, is implemented to solve the multiobjective optimal sizing problem. Finally, the feasibility of the proposed size planning method for SWPBM is verified by the simulation results.