A Distributed Synchronisation-based Predictor for Lur'e-type Nonlinear Systems

Abstract

This paper proposes a distributed state predictor for nonlinear systems using sampled measured data. In particular, we attempt to develop a distributed state predictor for Lur'e systems based on anticipating synchronisation. Anticipating synchronisation is a leader-follower synchronisation such that the follower system synchronises with the future values of the leader system by using delayed feedback. Since the prediction scheme has a closed-loop structure, it is expected to have robustness for disturbance and model mismatch. In this paper, we develop a distributed predictor that consists of networks of multiple predictors based on anticipatory synchronisation to achieve the state prediction for sampled-data nonlinear systems. First, we consider a distributed state predictor in the case where all measurements are sampled synchronously. After that, we develop a distributed state predictor based on asynchronously acquired measurements. The validity of the obtained results is illustrated in a numerical example.