Joint estimation of state and sensor systematic error in hybrid system

Author

Zhou, Lin ; Pan, Quan ; Liang, Yan ; Jin, Zhen-lu

Author_Institution

Sch. of Autom., Northwestern Polytech. Univ., Xi´´an, China

fYear

2012

fDate

9-12 July 2012

Firstpage

969

Lastpage

975

Abstract

Consider the hybrid systems with nonlinear property, and the sensor measurements with unknown and time-varying systematic error in this paper. In order to obtain the joint least square (LS) estimation of state and systematic error, a new method - JE-EM (joint estimation-expectation maximization) is proposed. In this paper, the relationship between the sensor systematic error estimation and state estimation is derived, which can be described by the framework of EM. Due to the character of the hybrid system, the target state is estimated by the IMM with PF filter. Based on the above relationship, systematic error is iteratively estimated by the framework of EM. Simulation results with a maneuvering target tracking scenario show the effectiveness of the proposed method.

Keywords

least squares approximations; measurement errors; particle filtering (numerical methods); target tracking; EM; PF filter; hybrid system; joint estimation-expectation maximization; joint least square estimation; nonlinear property; sensor measurements; sensor systematic error; state error; target tracking scenario; time-varying systematic error; Computational modeling; Error analysis; Measurement uncertainty; State estimation; Systematics; Target tracking; Expectation maximization (EM); Hybrid system; Interacting multiple model (IMM); Particle filter (PF); Systematic error;

fLanguage

English

Publisher

ieee

Conference_Titel

Information Fusion (FUSION), 2012 15th International Conference on

Conference_Location

Singapore

Print_ISBN

978-1-4673-0417-7

Electronic_ISBN

978-0-9824438-4-2

Type

conf

Filename

6289907