Design of sliding mode fuzzy controllers for an autonomous underwater vehicle without system model

For 2 dimensional systems, the time optimal control law with physically limited control command follows Pontryagin´s maximum principle. The control switches its sign on both sides of a nonlinear switching curve. The key factor in robust time optimal controller design is to approximate that nonlinear curve while maintaining robustness. Time optimality and robustness are contradictory to each other. This paper describes the development of a sliding mode fuzzy pitch controller and a sliding mode fuzzy heading controller for Ocean Explorer (OEX) series autonomous underwater vehicles (AUVs) developed in Florida Atlantic University. A sliding mode fuzzy controller (SMFC) inherits the robustness property of sliding mode control and interpolation property of fuzzy logic control such that the nonlinear switching curve can be approximated and the robustness can be maintained. Moreover, since the physical meaning of the parameters in a SMFC is straightforward, the on-line tuning with an SMFC is much easier than with the old fuzzy logic controllers used on the OEX series AUVs. The at-sea experimental results show the effectiveness of the design philosophy