Adaptive Neural Network Control Of Non-strictly Feedback Interconnected Systems

In practical control systems,the objects often present contain to be nonlinearity,uncertainty,and may be input saturation and unobservable.The classical linear system theory is hard to explain and analyze such control systems.On the other hand,there are a large number of interconnected large-scale systems in industrial productions.So,it is of great value to study the control problem of the interconnected large-scale systems.Based on the current research results on the control of nonlinear interconnected large-scale systems,this thesis will utilize adaptive neural network control theory,backstepping technology,and convex combination approach to investigate the control problem of nonlinear interconnected large-scale systems and develop new adaptive neural control methods via state feedback or output feedback.The main results of this thesis are shown as follows:1.Firstly,introduce the application of Radical Basis Function(RBF)neural networks to approximate nonlinear functions,some necessary definition of stability based on Lyapunov theorem,and the main lemmas and hypotheses in this thesis.2.Secondly,the problem of state feedback decentralized control is considered for a class of interconnected large-scale systems with non-strict-feedback form.In the control design,radial basis function neural network is used to approximate the unknown nonlinear functions,then a property of the vector norm of the basis vector functions is developed to overcome the difficulty caused by the non-strict-feedback structures in interconnected large-scale systems.Furthermore,backstepping technique is applied to design adaptive neural controller.At last Lyapunov theorem is employed to analyze the stability of the closed-loop systems.The proposed scheme guarantees the boundedness of all the signals in the resulting closed-loop systems,while the system output tracking the desired reference signal and the tracking error converges to a small enough neighborhood of origin.The character of the proposed scheme is that the designed controller has a simple structure and fewer adaptive parameters.3.Secondly,Adaptive neural networks output feedback tracking control is presented for a class of large-scale nonlinear systems with non-strict feedback form and unmeasurable system state.A state observer is first constructed to estimate the unmeasurable state variables,Then by combing the adaptive neural networks method and backstepping technique,adaptive laws and controller parameters are designed.At last the stability of the resulting closedloop system has been proved by the Lyapunov stability theory.It is shown that all the signals in the closed-loop systems are semi-globally uniformly ultimately bounded,while the system outputs can track the reference signals as closely as possible.A simulation example is used to demonstrate the effectiveness of the proposed scheme.4.Finally,summarize the results on adaptive neural networks control for a class of interconnected large-scale systems,and give the expectation of future research.