In recent years, the frequent occurrence of extreme disaster weather has greatly influenced the safe operation of distribution network. Therefore, it is an effective way to improve the resilience of distribution network to formulate reasonable planning strategy before the disasters. With the rapid growth of distributed power generation and electric vehicle users, a multi-flexible resource planning strategy is proposed to improve the resilience of distribution network in typhoon weather. Firstly, the typical typhoon disaster scenarios are screened by constructing the correlation model of typhoon wind speed and failure rate, combining the conditional generation adversarial network and information entropy. Secondly, the upper planning model aiming at the minimum annual comprehensive cost and the lower multi-objective optimization model taking into account the operating cost and performance loss area are established. The electric vehicle charging pile planning is innovatively incorporated into the model, and the distributed and mobile characteristics of electric vehicles are utilized to enhance the flexibility and elasticity of the power grid. Then, the optimization algorithm combining non-dominated sorting genetic algorithm and interior point method is used to solve the model. Finally, through the improved IEEE33-node distribution system, the results show that the proposed method can effectively reduce the network loss and load loss, and improve the overall resilience of the distribution network.