NMPC-Based Obstacle Avoidance Trajectory Planning for Bionic Ray-Inspired Amphibious Robots with Dead Zone Compensation

This paper focuses on a bionic ray-inspired amphibious robot (BRAR), which is modeled through the differential steering approach. A nonlinear model predictive control (NMPC) obstacle avoidance method is proposed for trajectory planning. First, a BRAR’s differential drive system is designed, followed by kinematic and dynamic modeling. Subsequently, an NMPC-based obstacle avoidance trajectory planning method is developed to constitute safe trajectories in complex workspaces. Further, a dead zone compensation method is proposed to improve control precision. Finally, the effectiveness of the proposed method is validated through both simulations and experiments. Simulation and experimental results demonstrate the feasibility and effectiveness of the proposed methods.