AUV docking control based on stochastic model predictive control

The AUV docking control based on stochastic model predictive control is studied with the fork-carrying-pole recovery method as the background. Firstly, based on the traditional AUV model, the unmodeled dynamics that may occur in the model simplification process are fully considered. And the five-degree-of-freedom(5-DOF) AUV model considering the unmodeled dynamics is established. Secondly, the docking control problem is transformed into a standard convex quadratic programming problem according to the stochastic model predictive control algorithm. In the design phase of control strategy, the actual constraints of system input and state are effectively considered, especially the speed constraints in docking phase. According to different reference variables, safe longitudinal and vertical expected speeds are designed to enable AUV to enter the docking station at a safe speed as soon as possible. The collision between the AUV and the mothership due to too fast speed is avoided. Finally, some simulation experiments are carried out. The results show that the control accuracy can be improved by compensating the nonlinearity of AUV.