A robust unscented fusion filter using fuzzy adaptation rule
Chul Woo Kang, Chan Gook Park · 2014
This paper presents a new robust estimation approach for nonlinear systems. Former approaches to robust nonlinear estimation such as the unscented H ∞ filter(UHF) [1] perform well on in disturbed nonlinear systems. However, with regard to undisturbed systems, the performance of robust nonlinear filters has proven inferior to that of conventional nonlinear filters. In this paper, a new filter is proposed which performs well on both disturbed and undisturbed systems by integrating a UHF and an unscented Kalman filter (UKF). The proposed filter uses a hybrid filter structure for the proper integration of the two local filters; a fuzzy-based mode adaptation rule is also implemented to improve performance.