Adaptive output feedback control of uncertain multi-input multi-output systems using single hidden layer neural networks

N. Hovakimyan, A.J. Calise
2002 Proceedings of the 2002 American Control Conference (IEEE Cat. No.CH37301)  
We consider adaptive output feedback control of uncertain multi-input multi-output nonlinear systems, in which both the dynamics and the dimension of the regulated plant may be unknown, but knowledge of vector relative degree is required. Given smooth reference trajectories, the problem is to design controllers that force the system measurements to track them with bounded errors. The classical approach necessitates building a state observer. However, finding a good observer for an uncertain
more » ... inear multi-input multi-output system is not an obvious task. We argue that for an observable and stabilizable minimum phase system it should be sufficient to build a linear observer for the output tracking error vector. A Single Hidden Layer Neural Network is introduced to cancel the modelling errors. Ultimate boundedness of the error signals is shown through Lyapunov's direct method. Simulations of a fourth order two-input two-output nonlinear system illustrate the theoretical results. * Research Scientist II, Member IEEE † Professor, Member IEEE
doi:10.1109/acc.2002.1023243 fatcat:bzb42hmlf5bfznnlnceoqqjxny