Recursive Elimination Method in Moving Horizon Estimation for a Class of Nonlinear Systems and Non-Gaussian Noise

Abstract

This paper proposes a recursive elimination method for optimal filtering problems of a class of discrete-time nonlinear systems with non-Gaussian noise. By this method, most of the computations to solve an optimal filtering problem can be carried out off-line by using symbolic computation based on the results from algebraic geometry. This property is suitable for moving horizon estimation, where a certain optimal filtering problem must be solved for different measurement sequences in each sampling interval. A numerical example is provided to compare the proposed method with other state estimation methods including the particle filter, and the efficiency of the proposed method is shown.

Extended moving horizon estimation for chemical processes under non‐Gaussian noises

“…1,2 Nevertheless, probability density functions (PDFs) of random variables present in most of the chemical processes often follow non-Gaussian distributions. 3,4,5 This inconsistency between the MHE framework and the actual process operating conditions may lead to an inaccuracy in estimation.…”