Nonlinear observers for a class of neuronal oscillators in the presence of strong measurement noise

Abstract

The objective of this paper is to design observers for a class of neuronal oscillators on the one hand, and to give a comparative study of the observer performance as the number of synchronized observer increases, on the other hand. More specifically, we apply the methodology of observer design in [4] for a class of neural oscillators. Contraction tool [7] is applied to obtain an exponentially convergent reduced-order observer, which serves as a building-block to construct a complete-order observer when the output is corrupted by moderate level of noise. In presence of strong measurement noise, several identical complete-order observers are coupled to synchronize.