Trajectory Online Planning for UCAV Air to Ground Attack

This paper presents an efficient trajectory online planning method for unmanned combat aerial vehicle (UCAV) performing autonomous air to ground attack in an unexpected threat condition. It combines the benefits of artificial bee colony algorithm (ABC) and hp adaptive pseudospectral method. First of all, the trajectory online planning model with kinematic and dynamic constraints is built. Then, the entire planning procedure is divided into two phases: waypoints optimization phase and trajectory re-planning phase. In waypoints optimization phase, the ABC algorithm is used to heuristically search optimal waypoints in the neighborhood of unexpected threat space. For typical ABC algorithm converges too fast and is easy to fall into local optimum solution, an improved ABC algorithm based on chaotic mapping is proposed. The chaotic mapping based ABC algorithm (CMABC) improves swarm initialization and the updating way of food sources, also maintains the speed and swarm diversity in the optimization process, and ultimately reaches the global optimal solution. In trajectory re-planning phase, in order to meet kinematic and dynamic constraints of UCAV in the flight, a local trajectory planner is built based on hp adaptive pseudospectral method to generate feasible trajectory. Numerical experiment results demonstrate that the proposed method can generate both feasible and optimal trajectory for online purpose, the trajectory can be directly applied to flight, as the planning results contain vehicle states and controls information.