Real-Time Nonlinear Model Predictive Control of Unmanned Surface Vehicles for Trajectory Tracking and Collision Avoidance

摘要

As the demand increases for marine activities such as environmental monitoring, search and rescue. Research on unmanned surface vehicles (USVs) has attracted considerable attention. The control technologies in the motion stabilization and tracking are the fundamental issues that need to be solved for USV. Moreover, the collision avoidance should be considered in the control problem to guarantee the safety. Here, we present a real-time nonlinear model predictive control (NMPC) scheme for constrained 3-DOF dynamical USVs. The obstacle avoidance problem in navigation is transformed into a variable constraint problem in NMPC, and the ship avoidance rule in International Regulations for Preventing Collisions at Sea (COLREGS) is implemented by the value of the weight in the cost function. The optimization problem with constraints is solved by an open-source powerful optimization software-CasADi in a finite horizon and finite risk minimization. A wave adaptive modular USV with two electric propulsion modules is used as an object for validation our method, three sailing scenarios including point shooting, trajectory tracking, and head-on situation following Rule 14 in COLREGS are simulated in MATLAB. The results demonstrate that the proposed method has good controller performance and real-time requirements.