Proactive preparation is a crucial necessary condition for energy systems to handle disaster events. This paper proposes a proactive method to enhance the resilience of integrated energy systems by obtaining pre-disaster optimal scheduling strategies to improve system resilience. First, the paper analyzes the wind force model of typhoon disasters, calculates the fault model of overhead lines within the system, and performs targeted reinforcement based on line fault rates. Then, the total amount of electrical, thermal, and natural gas power storage within the worst-case scheduling range of the disaster is defined as the preparedness index. Increasing the preparedness index comes at the expense of load reduction within the pre-disaster scheduling range. To address this, the paper establishes a Markov model for the integrated energy system and uses the Actor-Critic algorithm to obtain the optimal scheduling strategy during a disaster, solving the multi-objective optimization problem of conflicting goals (preparedness index and load supply). Finally, through simulations, it is confirmed that the optimized scheduling strategy can significantly improve the system""s preparedness index and resilience indicators, greatly reducing load losses.