Self-organizing Fuzzy Neural Tracking Control for Surface Ships with Unmodelled Dynamics and Unknown Disturbances

In this paper, a novel self-organizing fuzzy neural control (SOFNC) scheme for tracking surface ships, whereby a self-organizing fuzzy neural network (SOFNN) is used to approximate unmodelled dynamics and unknown disturbances, is proposed. The salient features of the SOFNC are as follows: (1) Unlike previous fuzzy neural networks (FNN), the SOFNN is able to dynamically self-organize compact T-S fuzzy rules according to structure learning criteria. (2) The SOFNN-based SOFNC scheme is designed by combining the sliding-mode control (SMC) with the improved projection-based adaptive laws which avoid parameter drift. (3) A robust supervisory controller is presented to enhance the robustness to approximation errors. (4) The SOFNC achieves excellent tracking performance, whereby tracking errors and their first derivatives are globally asymptotical stable in addition that all signals are bounded. Simulation studies demonstrate remarkable performance the SOFNC in terms of tracking error and online approximation.