Unmanned aerial vehicle trajectory planning by an integrated algorithm in a complex obstacle environment

Both the artificial potential field method and direct method for the optimal control problem have shortcomings in terms of effectiveness and computational complexity for the trajectory-planning problem. This paper proposes an integrated algorithm combining the artificial potential field method and direct method for planning in a complex obstacle-rich environment. More realistic unmanned aerial vehicle dynamics equations, which are rarely applied in the traditional artificial potential field method, are considered in this paper. Furthermore, an additional control force is introduced to transcribe the artificial potential field model into an optimal control problem, and the equality/inequality constraints on the description of the shape of the obstacles are substituted by the repulsive force originating from all the obstacles. The Legendre pseudospectral method and virtual motion camouflage are both utilized to solve the modified optimal control problem for comparison purposes. The algorithm presented in this paper improves the performance of solving the trajectory-planning problem in an obstacle-rich environment. In particular, the algorithm is suitable for addressing some conditions that cannot be considered by the traditional artificial potential field method or direct method individually, such as local extreme value points and a large numbers of constraints. Two simulation examples, a single cube-shaped obstacle and a different-shaped obstacle-rich environment, are solved to demonstrate the capabilities and performance of the proposed algorithm.