A Neural Network Perspective to Extended Luenberger Observers

In this paper, we investigate the use of adaptive extended Luenberger state estimators for general nonlinear and possibly time-varying systems. We identify the connection between the extended Luenberger observer and Grossberg’s additive model for dynamic neural networks. The association between dynamic neural networks and the Luenberger observer leads to an obvious modification on the proposed observer scheme that would allow handling state estimation for those systems whose dynamic equations are partially known or not known at all. The performance of the adaptive observer is demonstrated on a number of systems including an LTI system, the Van der Pol oscillator, the Lorenz attractor and a realistic partial gasoline engine model.