Nonlinear Adaptive Observer Design Based on B-spline Neural Network

H.Y. Zhang,C.W. Chan,K.C. Cheung,Hong Jin
DOI: https://doi.org/10.1016/s1474-6670(17)56397-1
1999-01-01
IFAC Proceedings Volumes
Abstract:A design approach of adaptive observer for a class of nonlinear systems with unknown time-varying nonlinearity is proposed. The nonlinear adaptive observer is constructed using a linear observer and a nonlinear compensation term. The latter term is determined recursively using a proposed “decon volution procedure”. The determined term is then approximated using a B-spline neural network with state estimate as its input. The convergence of state estimation error and weight error of the network is proved, and a weight updating rule is derived. Other weight updating rules, such as Instantaneous Gradient Descent method and Recursive Least-Squares method, are also discussed. Simulation results show that the proposed approach is effective
What problem does this paper attempt to address?