Improvement of Moving Horizon Estimators via Direct Virtual Sensor techniques

This paper presents a novel approach to the design of estimators for nonlinear systems. The proposed methodology is based on a combination of linear model-based Moving Horizon Estimation (MHE) and nonlinear Direct Virtual Sensor (DVS) techniques. Stability of the designed estimator is guaranteed and convex constraints on the variables to be estimated can be easily taken into account. Moreover, the optimality of the approach with respect to an “ideal” MHE (obtained by assuming exact knowledge of the system dynamics and of the global solution of the related nonlinear program) is analyzed. The approach is tested on a nonlinear mass-spring-damper system.