In order to meet the needs of informationized warfare, overcoming the electromagnetic compatibility of combat platforms, as well as the integration of multifunctional electronic systems with greatly improved combat performance, has become the future direction of weapons and equipment development. Mission planning methods for electronic systems need to maximize system performance under limited resource conditions, but due to the complexity of multi-task planning, existing mission planning methods often lead to significant performance degradation in real-world operations as the number of tasks increases. In order to resolve the issues raised above, based on a deep reinforcement learning approach to multi-task planning, we propose a trajectory optimization problem for solving Unmanned Aerial Vehicles (UAV) trajectories when they fly to multiple designated regions to perform tasks. For multi-task planning, this paper takes the above trajectory optimization problem as an entry point and employs the Deep Deterministic Policy Gradient (DDPG) algorithm to compute the best planning results of this system under the constraints of limited resource conditions. Simulation results indicate that the planning algorithm used in the paper has significant advantages in improving comprehensive performance of this system compared to traditional task planning methods.