Abstract: Resilience is an ability index of a system to absorb and withstand disturbances until returning to a normal state under extreme disasters. Based on the characteristics of the tight coupling between the Power-Information-Transportation Networks, the strategy for a multi-stage distribution network to improve its resilience considering the coupling characteristics of multiple networks is proposed. First of all, in the prevention stage before a disarster, the optimal line reinforcement strategy considering the network survivability and the risk value of the power information is proposed, and the emergency centers are set optimally. Secondly, based on the topology changes caused by the network damages, the distributed power generation and the tie switch are dynamically and jointly scheduled to minimize the load loss value in the disaster resistance stage. Then, combined with the associated constraints between the coupled networks, a multi-network collaborative recovery strategy is proposed in the post-disaster phase. Finally, a combined evaluation framework is established to conduct a multi-period refined analysis of the system resilience. Based on the IEEE33 node system to build a three-network coupling system example, the results show that the proposed strategy is able to effectively improve the resilience of the system under disasters and improve the recovery efficiency.