Underwater Obstacle Avoiding Trajectory Tracking Approach for Amphibious Spherical Robots

As real environment is complex and unstable, obstacles need to be considered when a robot tracks a trajectory. In view of this, the problem of tracking a path with obstacle avoiding is addressed. A two-layer control system is designed, which contains a re-planning controller and a trajectory tracking controller. In the re-plannng controller, nonlinear Model Predictive Control (NMPC) and avoiding obstacle function are combined to realize trajectory re-planning and obstacle reliable avoidance. In addition, a reference point of re-planning path next to the final point on a global reference path is chosen, which prevents the robot from going back to realize obstacle avoidance. 5 times multiform fitting is adopted for transferring the local trajectory information to the trajectory tracking controller. In the trajectory tracking controller, Model Predictive Control (MPC) based on kinematics and dynamics of the amphibious spherical robot (ASR) is designed. Simulation results illustrate that the designed two-layer control system is able to track a desired trajectory and avoid obstacles reliably at the same time.