Robust trajectory tracking and control allocation of X-rudder AUV with actuator uncertainty

This paper proposes a robust trajectory tracking control method for an X-rudder autonomous underwater vehicle subject to imprecise model parameters, unknown disturbances, and actuator uncertainty. The control scheme consists of a kinematics controller, dynamics controller, and rudder allocator constructed using backstepping techniques. The kinematics controller employs a line-of-sight guidance law, while the dynamics controller uses integral sliding mode and a Radial Basis Function Neural Network to handle modeling errors and unknown disturbances. Finally, a Recursive Least Square Control Allocation method is developed to minimize the allocation error considering the uncertainty in the control effectiveness matrix by solving a dual second-order cone programming problem. Simulation results demonstrate that the RBFNN-based trajectory tracking method achieves satisfactory performance in the presence of time-varying environmental disturbances, and the Recursive Least Square Control Allocation method effectively deals with actuator uncertainty.