Observer based adaptive robust control with application to coordinated precision control of linear motor driven high speed electro-mechanical systems

Every physical system is subject to certain degrees of model uncertainties, which makes the design of high performance control algorithms a very challenging job. Recently, a state feedback adaptive robust control (ARC) approach was proposed, which effectively integrates the design methods of deterministic robust control (DRC) and adaptive control (AC) for high performance.;The ARC design assumes that all the states are measured and available for feedback. For many applications, however, only a part of the states or just the plant output is available for measurement. This work aims to remove this assumption and extend the ARC design to the output feedback or partial state feedback cases. The work focuses on the development of adaptive robust nonlinear observers and on how to incorporate the observers into the general ARC framework. In addition, trajectory tracking performance of the ARC schemes can be further improved by making full use of the available structural information and the prior knowledge of the bounds of the unmeasurable states, as is in the case of dynamic friction compensation.;The dissertation also extends ARC framework to the systems with periodic unknown nonlinearities or "repetitive control" cases. It is shown that, in continuous time domain, what the conventional repetitive learning algorithm does is equivalent to adapting an infinite number of parameters, which is sensitive to noise. The problem can be easily addressed in the proposed adaptive robust repetitive control (ARRC) strategy. In addition, ARRC has a much better ability in dealing with non-repeatable uncertainties and is applicable to nonlinear systems as well.;The other focus of the work is the coordinated precision motion control of linear motor drive systems. Linear motors offer several advantages over their rotary counterparts and show promise for widespread use in high-speed/high-accuracy positioning systems. However, these advantages are obtained at the expense of added difficulties in controlling such systems. In order for a linear motor system to be able to deliver its high performance potential, the proposed ARC algorithms which can handle various parametric uncertainties and uncertain nonlinear effects effectively are employed. Extensive experimental results are provided to verify the effectiveness and the achievable control performance of the ARC designs.