Path Following Method Research For Unmanned Surface Vessels Based On Predictive Control

Unmanned surface vessels (USV) refer to the intelligent platforms relying on autonomous navigation or remote control on the water. It is fast, flexible and can finish a multitude of extremely dangerous missions that human can't complete. It plays an irreplaceable role in contemporary ocean war and marine resources exploration. Because the environment of the surface is very complex, USV will be influenced by the real-time sea conditions such as static obstacles or other sailing ships in order to complete the specified cruising tasks at the same time. Therefore, the research on trajectory tracking of USV under the complicated environment has become a hot topic in ship control field.In the existing literature, many scholars have researched on several methods for the path tracking of unmanned surface vessel such as PID control, sliding mode control and predictive control and made significant achievements. The majority of these literature, however, have verified the validity of relevant algorithm through computer, and the research on the application of predictive control on the actual USV is rare. Therefore, the aim of this paper is to broaden the model predictive control for trajectory tracking of USV based on the existing predictive control algorithms.Firstly, this article has established trajectory tracking control system model of USV under the Serret-Frenet coordinate control containing external disturbances and input constraints. Using the Lagrange multiplier method to deal with amplitude input constraints and obtained convergent control law with constraint conditions by model predictive control method.Moreover, Using a disturbance observer to estimate the nonlinear disturbance of the outside on line, and the model predictive controller with disturbance observer is designed,which can compensate the disturbance in real-time. In order to describe actual static obstacles oversea, this paper has introduced the concept of risk of field through artificial potential field.And the model predictive controller with the augmented risk state is designed, which realize that USV can avoid the obstacles and track the trajectory accurately simultaneously.At last, the proposed algorithms are verified through computer simulation and applied to the modeling of unmanned surface vehicle experiment successfully reflecting the effectiveness of the proposed control algorithms and engineering application value.