Smooth path planning for unmanned aerial vehicles in unknown complex environments

A smooth real-time path planning approach is proposed based on the bilevel programming (BLP) for unmanned aerial vehicles (UAVs) in complex environments, to improve the flight path smoothness which has not been achieved by most existing methods. Firstly we define the measure for the path smoothness, and then we build the model for the bi-level decision objectives, the model of obstacle avoidances and the model of performances of the UAV, and introduce a variable planning time interval. On this basis, we proceed to develop the path planning model based on the bi-level decision. In the process of the planning, we introduce heuristic optimal strategies to further improve the smoothness for the local path and the global path, and to raise the efficiency in path searching. Results from simulations of the proposed approach in complex scenarios are compared with those obtained from classical methods; the conclusions indicate that the proposed approach can successfully plan a shorter and smoother flight path in real-time when passing around a wide dangerous region.