| 英文摘要 | This thesis is deals with the adaptive control of a class of nonlinear systems. The Author gives a survey of development of adaptive control theory. According to this ruler, the contents of the thesis include: · Propose some new conceptions for adaptive control · Extend adaptive control theory of linear systems in order to be adapted to nonlinear systems · Look for new, good control schemes · Study the stability, convergence and robustness of systems · Reduce the assumptions in order to improve the applicability of control theory · Simulation results In chapter one, the problem of adaptive control of studying nonlinear systems is introduced and development outline of adaptive control is given. The main research problem are presented. In this chapter, the author proposes new conceptions about adaptive control and intelligent control, adaptive control belongs to the field of intelligent control. Intelligent control integrates many theories. In chapter two, several kinds of nonlinear models and mathematical representation are presented, which makes readers know nonlinear system models. In this chapter, background and ordinary, control schemes of nonlinear system models are also introduced. Up to now, the identification and adaptive control of HAMMERSTEIN models are difficult. Theory about the identification and adaptive control of HAMMERSTEIN models is still imperfect. Especially weighted control problem of the models is not yet solved. In the bellowing chapters, the problem of adaptive control about the models and extended models are described in detail. In chapter three, adaptive control of deterministic HAMMERSTEIN models is discussed. According to rule of minimal error, first, direct adaptive control scheme and indirect adaptive control scheme are described. Basic principle of designing adaptive controllers of the models extends adaptive control theories of linear systems. The difficult point of adaptive control systems is not the design of controller but analysis of the stability of closed-loop systems became of the nonlinear gain. The design of controller requires the inverse stability. How to reduce assumption is the main problem in author' research. If control horizon is weighted as usual, it results in constraint control law. Author proposes pseudo weighted control scheme which avoids assumption of inverse stability.In chapter four, stochastic adaptive control of the models is presented. In the chapter, the author proposes adaptive inverse dynamic of nonlinear gain of the models, so control scheme of combined pole placement is used for the models. In chapter fry'e, adaptive control of extended HAMMERSTEIN models-output nonlinearity, is discussed. First. the author discusses adaptive control of deterministic models. Second, the author deals with the stochastic adaptive control of the models. On the basis of chapters three, four and five, in chapter six, the author discusses adaptive control of d |
修改评论