Pseudospectral method-based safe motion planning for quadrotors in a cluttered environment

Abstract

This paper presents an optimal-control-based motion planning framework for quadrotors in a dense obstacle field. On the backbone of a pseudospectral (PS) transcription of the optimal motion control problem (OCP), we compute and embed the safe flight corridor (SFC) to ensure feasibility of a safe path and to improve computational efficiency by employing linear inequality constraints from SFC. However, the solution path is imperfect in terms of keeping constraints since only state and control input at collocation points are enforced to stay within the constraints. To tackle these challenges, model predictive controller (MPC) tracks the computed path while keeping the constraints by locally solving the OCP in receding time horizon. A numerical example of minimum-time motion planning of a quadrotor in a three-dimensional cluttered environment validates the proposed framework. © 2022, American Institute of Aeronautics and Astronautics Inc. All rights reserved.