In order to improve the autonomous ability of unmanned aerial vehicles (UAVs) and achieve the integral framework of task assignment and path planning, a mission planning system based on path prediction algorithm for multiple UAVs is presented. To model obstacles more accurately, the forbidden areas are defined as polygons. Then, the optimal path segment avoiding all obstacles between two waypoints is computed by using improved A^* path prediction algorithm. According to this path segment, the task assignment is determined by improved particle swarm optimization (PSO) algorithm. Finally, the B-spline method is adopted to smooth the flight path, which consists of the sequential path segments. Numerical results demonstrate that the proposed method can achieve the near-optimal task assignment and best flight routes with effectiveness of computation speed and precision.