Output feedback neural network adaptive robust control of a class of SISO nonlinear systems

Through the use of high-gain observer to estimate the unmeasurable system states, neural networks (NNs) and adaptive robust control (ARC) method are integrated to design performance oriented output feedback control laws for a class of single-input-single-output (SISO) nth order nonlinear systems in normal form. Multi-layer neural networks (MLNNs) with the estimated states as inputs are used to approximate all unknown but repeatable nonlinear functions in the system. A controlled learning is achieved through the use of discontinuous projections with fictitious bounds in the tuning laws for NN weights. Certain robust control terms are constructed to effectively attenuate various model uncertainties and estimation errors for a guaranteed output tracking transient performance and a guaranteed final tracking accuracy in general. In addition, asymptotic output tracking is achieved in the ideal case. Experimental results on the output feedback control of a linear motor drive system are obtained to illustrate the proposed algorithm.