Robust Motion Control Of Unmanned Underwater Vehicles Based On T-S Fuzzy Model

Motion control performance of an unmanned underwater vehicle(UUV)determines its stability,maneuverability and capability of disturbance rejection,and in turn influences the quality of survey data of onboard inspection equipment and/or operation efficiency of onboard interventional tools.Taking into account the strong nonlinear dynamics,model uncertainty and environmental disturbances of UUV,Takagi-Sugeno(T-S)fuzzy model is adopted to model UUV dynamics accurately.Linear state feedback controller is design based on the T-S fuzzy model.Through the Lyapunov stability analysis of the control system,linear matrix inequalities(LMIs)that controller parameters should satisfy are derived and discrete and continuous design methods suitable for robust motion control of UUVs are obtained.The discrete method is designed based on the continuous T-S fuzzy model,where the control period is introduced as a parameter.It could overcome the serious degradation of control performance when the control period is relatively long.This approach has been successfully verified in the depth control of a hovering autonomous underwater vehicle(AUV).The continuous design method based on the continuous T-S fuzzy model is effective to computer control systems with short control periods.It is unnecessary to consider the influence of control periods in the design,so this method is relatively simple.However,the method is unable to suppress external disturbances.Incorporating the feedforward control of the Nonlinear Disturbance Observer(NDOB)is needed to eliminate steadystate errors when the method is directly applied.In order to simplify the controller structure and improve performance,this method is further modified in this paper.The integral of tracking error is augmented to state variables of the plant model at first.Parametric uncertainties involved in hydrodynamic coefficients are described explicitly and the T-S fuzzy model is then derived.A state feedback controller with a single gain which guarantees the global stability is then found based on the T-S fuzzy model.This approach has an obvious advantage over the conventional T-S fuzzy model-based control methods in that the controller has significantly reduced structure.Finally,the approach has been validated via tank tests of depth and heading controls of the remotely operated vehicle(ROV).