An improved method of state estimation for Markov jump linear system with quantized measurements

This paper is concerned with the problem of state estimation for Markov jump linear system (MJLS) subject to the quantization effects due to the network communication channel constraints. After taking into account the statistical knowledge of the quantized measurements, an exact minimum mean square...

Full description

Saved in:
Bibliographic Details
Published in2016 IEEE International Conference on Industrial Technology (ICIT) pp. 1527 - 1532
Main Authors Yingjun Niu, Wei Dong, Yindong Ji
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.03.2016
Online AccessGet full text

Cover

More Information
Summary:This paper is concerned with the problem of state estimation for Markov jump linear system (MJLS) subject to the quantization effects due to the network communication channel constraints. After taking into account the statistical knowledge of the quantized measurements, an exact minimum mean square error (MMSE) estimate algorithm is derived, referred to as IMM Kalman-like particle (IMMKLP) algorithm. The proposed IMMKLP algorithm redesigns the kernel filter under the interacting multiple model (IMM) framework, based on the exact state posterior probability density conditioned on quantized measurements. A simulation example is provided demonstrating that IMMKLP achieves considerably high estimate accuracy and is computationally appealing by reducing the particle number dramatically.
DOI:10.1109/ICIT.2016.7474987