The integration of Unmanned Aerial Vehicles (UAVs) and Mobile Edge Computing (MEC) enhances the coverage and performance of communication networks. However, achieving full communication coverage and efficient task offloading with a minimal number of UAVs remains challenging due to their limited range and energy. To address this issue, we propose a UAV-assisted two-stage task scheduling model to optimize the costs of the MEC system, focusing on both UAV positioning and task scheduling. Accordingly, we design a UAV-assisted two-stage intelligent collaborative method, which includes two algorithms: the enhanced particle swarm optimization algorithm and the deep reinforcement learning algorithm, to find the optimal solution. Simulation results show that the proposed method converges well and outperforms three classical reinforcement learning algorithms in terms of reducing latency and energy consumption.