As the penetration rate of renewable energy sources and the proportion of electric vehicles continue to increase

the uncertainties in the operation of microgrids have increased significantly. Accurate assessment of the influence of uncertainties on the status of microgrids will help improve the safe and stable operation of the system. Considering the uncertainties of source and loads

the probabilistic power flow model of islanded microgrids is established and the global sensitivity analysis indices are introduced. Then

the global sensitivity analysis of islanded microgrids based on sparse polynomial chaos expansion is proposed. The proposed method is used to accurately and quickly identify critical input random variables that affect the operation status of the system. Through the simulation of a 40-bus islanded microgrid with distributed generators

the accuracy and efficiency of the proposed method are verified. The effect of the source and load uncertainty on the power flow of islanded microgrid is analyzed and the importance ranking of input random variables is given.