Indirect Adaptive Robust Control of Nonlinear Systems with Application to Electro-mechanical/hydraulic Systems

Palmer, Andrew P., M.S., Purdue University, August, 2000. Indirect Adaptive Robust Control of Nonlinear Systems with Application to Electro-Mechanical/Hydraulic Systems. Major Professor: Dr. Bin Yao, School of Mechanical Engineering. The focus of the thesis is on the nonlinear model based control of systems with unknown parameters and uncertain nonlinearities. The objective is to maximize the achievable performance of a controlled system and to obtain accurate parameter estimates for other purposes. This is achieved by integrating the excellent output tracking performance achieved by the recently proposed direct adaptive robust control with the good parameter estimation process of indirect adaptive designs. Two types of indirect adaptive robust controllers (IARC) are rst constructed. The methods are based on the recently proposed modularized adaptive backstepping designs with zswapping and x-swapping estimation algorithms. However, there are some signi cant di erences in terms of the fundamental view of the problem and the use of parameter adaptation. The modularized adaptive backstepping designs assume the system free of disturbances and uncertain nonlinearities. This may cause problems in application due to unavoidable disturbances and uncertain nonlinearities. The proposed methods, in contrast, use available knowledge of the physical bounds of unknown parameters, along with preset adaptation rate limits, to construct projection type parameter estimation algorithms with rate limits for a controlled estimation process. By doing so, a guaranteed transient performance and nal tracking accuracy is achieved even in general, regardless of the estimation algorithm used. In addition, the theoretical performance of the adaptive designs is preserved. An integrated direct/indirect adaptive robust controller (DIARC) is then developed that combines the excellent transient performance of the direct ARC designs with the good parameter estimation process of the IARC designs. The three controllers are applied to a linear motor system to