To guarantee the reliable and economical operation of standalone microgrids,the optimal configuration of power capacity must be determined during the planning phase.The selection of power source types and power capacities of standalone microgrids is affected by internal load levels and the unique natural resource conditions such as wind,photovoltaic (PV),water and storage potential of the respective regions.This paper investigates the comprehensive natural resource conditions of "wind/PV/water/storage" across different regions and establishes a multi-objective optimization model aimed at minimizing the annual generation cost,while considering the reliable power supply and environmentally friendly power generation.Constraints related to electric power and energy balance are incorporated into the optimization model,which is solved via a linearization algorithm.To address the uncertainty in the output of wind,PV and water power,we employ Generative Adversarial Networks (GANs) to simulate multiple scenarios,which are then reduced via an improved K-Medoids clustering algorithm,thereby enhancing the computational efficiency.Furthermore,an index evaluation system is constructed to analyze the characteristics of wind,PV and water resources,and the natural resource levels of 31 provincial-level administrative regions in China's mainland are obtained using a fuzzy evaluation method.By comparing the power capacity configurations of standalone microgrids in 5 representative regions with varying natural resources,this study validates the feasibility and rationality of the proposed approach,and provides valuable insights for the planning of standalone microgrid system.