Gauss-Newton moving horizon observer for state and parameter estimation of nonlinear networked control systems

In this paper a method to design an observer for state and parameter estimation of sampled nonlinear systems where the sensor possesses an unsynchronized time is presented. The sensor measurements are transmitted together with the relative time stamps as a packet to the observer via a communication network. These packets are subject to random delay or might even be completely lost. Using the information of a fixed number of past packets and the model equations, state and parameter estimation is expressed as a moving horizon optimization problem which is solved by using the Gauss-Newton algorithm. The performance of this method is illustrated using a simulated continuous stirred tank reactor.