Model Reference Robust Control And Its Applications

This dissertation explores the model reference robust control, which is derived from both classical model reference adaptive control and backstepping adaptive control, for the purpose to solve some problems encountered in its research and applications.The main contributions of this dissertation are as follows.1. Based on the exiting model reference robust control, a new control scheme is proposed to cope with single input and single output linear time-invariant systems with nonlinear input disturbance and unmodelled dynamics, which significantly improves the tracking performance while reducing the amplitude of the control signal compared with the original model reference robust control. Moreover, our scheme clearly gives the relationship between the L∞performance of the tracking error and the design parameters even though the high frequency gain of the controlled plant being uncertain.2. Motivated by the dynamic surface control, a new model reference robust control scheme for plants with relative degree greater than one is proposed for the purpose to solve the"explosion of complexity"problem for both the existing model reference robust control and backstepping control. The key of this controller is that a first-order filter is used to link each subsystem, so that the derivative of some nonlinear signals can be avoided. Compared with the dynamic surface control, our new scheme is based on output feedback, i.e., only input and output signals of the controlled plant are needed.3. A new model reference robust control for plants with fast time-varying parameters is proposed to improve the transient performance and the amplitude of the control signal.4. Finally, in this dissertation, the model reference robust control for plants with unknown high frequency gain sign is investigated. Based on an appropriate monitoring function, a switching scheme is proposed so that after a finite number of switching, the tracking error can converge exponentially to zero if the plant relative degree is one, or a residual set whose radius is proportional to some design parameters if the plant relative degree is greater than one.