The Design Of Neural Network Adaptive Controller Based On Catastrophe Theory

The found of Catastrophe theory is one of important achievements in nonlinear science.With the complication of controlled object on aeronautic and marine engineering,heavy electric power system and complex productive progress, together with the desire of performance index on controlled system becomes higher and higher.The controlling problem of complex system,especially the complex system with catastrophic characteristic,is an austere challenge the traditional control theory has to be faced with.So that new control methods are urgently required for the control problem of nonlinear catastrophe system.To adapt this actual demand,Catastrophe control theory emerges as the times require.The research of catastrophe control theory is just in its first stages,this dissertation starts with the investigation of catastrophe phenomena,catastrophe theory is introduced to system controller design,to research catastrophe control mechanism and it's application.There are two aspects in the combination of catastrophe theory and control technologyrone is to create new beneficial catastrophe while preserve all initial equilibria,named catastrophe control; another is to eliminate or restrain harmful catastrophe,called averse-catastrophe, it is not only preserve the original equilibrium even under model uncertainty,but also can simplify the controller,to expand the range of control parameter.By combining model reference adaptive control and neural netword control,a novel model reference adaptive neural network control scheme is proposed for the ship's course nonlinear model,design an averse-catastrophe controller,In the procedure,the RBF neural networks are used to approximate nonlinear systems;The weights can be achieved by use of Lyapunove approach.The tracking errors convergence to a neighborhood of zero.At last,the scheme is appilied to ship's course steering control and the simulation curves show that the desired results are attained.