A linearization free recursive prediction error method for combined state and parameter estimation for nonlinear systems

A novel method for combined state and parameter estimation is presented. The method is based on the recursive prediction error method for nonlinear system and is able to estimate states and parameters from nonlinear state-space models simultaneously. Moreover, because of the application of the deterministic sampling approach, no model linearization is required. A simple example (real-time parameter estimation for a three-tank system) is included to demonstrate the performance of the algorithm.