On Decentralized Adaptive Control Of Interconnected Large-scale Systems With Unmodeled Dynamics

Along with the development of technology, more and more problems about large-scale system appear, the decentralized adaptive control problems of the large-scale system with unmodeled dynamics have received much attension of the control scholar. The unmodeled dynamics widely exists in the practical complicated systems and it is easy to cause system oscillation and divergence. Combining variable structure control with backstepping design method, dynamic surface control and neural networks technology, several decentralized adaptive control schemes are proposed in this paper. The main contributions of the thesis are summarized as follows:Firstly, based on the approximation capability of the neural networks, a decentralized adaptive neural network control scheme is presented for a class of large-scale nonlinear systems in pure feedback form with unmodeled dynamics. A dynamic signal is introduced to dominate the unmodeled dynamics in the control process. The unknown nonaffine functions are separated by the mean value theorem, while the restrictions of disturbance and interconnections are relaxed by utilizing the separation technique. Decentralized control law and parameter adaptive law are designed by using variable structure technology. By theoretical analysis, the closed-loop control system is shown to be semi-globally uniformly ultimately bounded. The effectiveness of the proposed approach is illustrated by using simulation results in the end.Secondly, based on the approximation capability of the neural networks, a decentralized adaptive dynamic surface control scheme is presented for a class of large-scale nonlinear interconnected systems with unmodeled dynamics. A dynamic signal is introduced to dominate the unmodeled dynamics in the scheme. The restrictions of disturbance and interconnections are relaxed by using the separation technique. By combining dynamic surface control with the backstepping design method, such as the "explosion of complexity" problem caused by the repeated differentiations on virtual control in traditional backstepping design is effectively solved, and finally the decentralized adaptive control of interconnected large-scale systems is realized. By theoretical analysis, the closed-loop control system is shown to be semi-globally uniformly ultimately bounded, and the tracking errors converge to a small neighborhood of the origin. The effectiveness of the proposed approach is illustrated by using simulation results.Thirdly, a decentralized adaptive neural network control scheme is proposed for a class of time-delay interconnected large-scale systems with unknown dead-zone and unmodeled dynamic. Neural networks are used to approximate the unknown continuous functions in the procedure. By constructing proper Lyapunov-Krasovskii functions, the unknown time-varying delay uncertainties are compensated for. A dynamic signal is introduced to dominate the dynamic disturbances. Separation technology is used to relax the limitation on the interconnection terms. Finally the decentralized adaptive control of interconnected large-scale systems is implemented. By theoretical analysis, the closed-loop control system is shown to be semi-globally uniformly ultimately bounded, and the tracking errors converge to a small neighborhood of the origin. Simulation results are given to illustrate the effectiveness of the proposed control scheme.