z-logo
Premium
Design of optimal interval observers using set‐theoretic methods for robust state estimation
Author(s) -
Xu Feng,
Tan Junbo,
Raïssi Tarek,
Liang Bin
Publication year - 2020
Publication title -
international journal of robust and nonlinear control
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.361
H-Index - 106
eISSN - 1099-1239
pISSN - 1049-8923
DOI - 10.1002/rnc.4960
Subject(s) - observer (physics) , mathematics , control theory (sociology) , interval (graph theory) , separation principle , alpha beta filter , state observer , mathematical optimization , computer science , nonlinear system , statistics , control (management) , kalman filter , artificial intelligence , physics , quantum mechanics , combinatorics , extended kalman filter , moving horizon estimation
Summary This article aims to design an optimal interval observer for discrete linear time‐invariant systems. Particularly, the proposed design method first transforms the interval observer into a zonotopic set‐valued observer by establishing an explicit mathematical relationship between the interval observer and the zonoptopic set‐valued observer. Then, based on the established mathematical relationship, a locally optimal observer gain is designed for the interval observer via the equivalent zonotopic set‐valued observer structure and the Frobenious norm‐based size of zonotopes. Third, considering that the dynamics of the optimal interval observer becomes a discrete linear time‐varying system due to the designed time‐varying optimal gain, an optimization problem to obtain a coordinate transformation matrix and the locally optimal observer gain for the interval observer is formulated and handled. Finally, a theoretic comparison on the conservatism of the interval observer and the zonotopic set‐valued observer is made. At the end of this article, a microbial growth bioprocess is used to illustrate the effectiveness of the proposed method.

This content is not available in your region!

Continue researching here.

Having issues? You can contact us here