To enhance the resilience of distribution systems and fight against extreme disasters, a novel planning-attack-reconfiguration optimization method is proposed in this paper. Firstly, according to the processes of prevention, defence, and restoration for a resilient distribution system through disruption, the novel resilience evaluation indicators are presented, which include the node degree of distributed generation (DG) bus, survival rate, and recovery ability. Secondly, a novel planning-attack-reconfiguration optimization model is developed to improve the resilience of distribution systems. In DG planning stage, the multi-objective planning model is formulated, which includes the minimization of the total cost of investment and operation, and the maximization of the node degree of DG buses for critical loads. In the attack stage, a clear worst case of N-k contingencies on the basis of generalized nodes is presented to reduce the computational complexity. Then, the post-disaster network reconfiguration model is formulated to maximize the restoration rate of critical loads (RRCL). Finally, the proposed method is illustrated by the case study on PG&E 69-bus distribution system. The simulation results indicate that all the RRCL can reach about 90% in the four multipoint fault scenarios. Meanwhile, other evaluation indicators are greatly improved. It is shown that the resilience of distribution systems can be dramatically enhanced by the proposed method.
