Trajectory tracking control for autonomous underwater vehicle based on rotation matrix attitude representation

This paper provides a rotation-matrix-based control scheme for fully actuated autonomous underwater vehicles (AUVs) considering model uncertainties and external disturbances. By resorting to the rotation matrix, AUV's attitude dynamics will be globally and uniquely represented without unwinding. However, this kind of attitude representation approach complicates the controller design process in the sense that the attitude tracking error is defined by a rotation matrix instead of a three-dimension vector. Regarding this, an alternative error vector is introduced for the simplicity of calculation and implementation. Then, a novel sliding mode surface that consists of two components is devised to provide the characters of finite-time convergence and singularity-free. Upon the application of this variable, an artful adaptive law is constructed such that the estimated information always appears as limited values in the controller. Eventually, theoretical analysis and comparative simulations can illustrate the effectiveness of the proposed control scheme.