Abstract: This paper presents a multi-stage multi-objective power expansion planning model with fault uncertainty,aiming at the situation that the generator set suffers from power interruption and other heavy losses due to random faults during operation. The outer layer optimizes the type and quantity of the added units,and the inner layer optimizes the dual objective function of operating cost and power loss cost for all units. By means of fault probability matrix,LOLP,EENS and other indicators,the possible fault types of units are materialized,and the difficulty of solving problems is reduced by load clustering and ignoring higher-order fault terms. Finally,the simulation verification is carried out with IEEE-RTS system,state enumeration method and the NSGA-Ⅱ algorithm is used for optimizing the dual objective function with multi-dimensional fault state. The results show that the model and method used in this paper can effectively take into account the economic cost of each unit while reducing power loss,and can provide theoretical basis for power expansion planning in practical projects in terms of spare capacity and planned output.