High-order interacting multiple model filter based on mixture transition distribution

Gongjian Zhou; Xiaoping Ye

High-order interacting multiple model filter based on mixture transition distribution

Gongjian Zhou
, Xiaoping Ye

School of Electronics and Information Engineering, Harbin Institute of Technology

Research output: Contribution to conference › Paper › peer-review

Abstract

A novel nth-order interacting multiple model (IMMn) filter based on the mixture transition distribution (MTD) for maneuvering target tracking referred as IMMn-MTD filter is proposed here. It adopts MTD model rather than fully parameterized high-order Markov chain (MC) to approximate the model sequence transition probabilities. Since only a few independent parameters need to be fixed in MTD model, IMMn-MTD is more flexible compared with the high-order MC mothed. This approach simultaneously possesses the advantage in accuracy because high order model sequence is evaluated. Simulations are performed to demonstrate the effectiveness of the proposed approach.

Original language	English
State	Published - 2017
Externally published	Yes
Event	2017 International Conference on Radar Systems, Radar 2017 - Belfast, United Kingdom Duration: 23 Oct 2017 → 26 Oct 2017

Conference

Conference	2017 International Conference on Radar Systems, Radar 2017
Country/Territory	United Kingdom
City	Belfast
Period	23/10/17 → 26/10/17

Keywords

High-order MC
IMMn
MTD
Model switching

Cite this

@conference{ba8779814b1e49a7a176a504efc847dd,

title = "High-order interacting multiple model filter based on mixture transition distribution",

abstract = "A novel nth-order interacting multiple model (IMMn) filter based on the mixture transition distribution (MTD) for maneuvering target tracking referred as IMMn-MTD filter is proposed here. It adopts MTD model rather than fully parameterized high-order Markov chain (MC) to approximate the model sequence transition probabilities. Since only a few independent parameters need to be fixed in MTD model, IMMn-MTD is more flexible compared with the high-order MC mothed. This approach simultaneously possesses the advantage in accuracy because high order model sequence is evaluated. Simulations are performed to demonstrate the effectiveness of the proposed approach.",

keywords = "High-order MC, IMMn, MTD, Model switching",

author = "Gongjian Zhou and Xiaoping Ye",

note = "Publisher Copyright: {\textcopyright} 2017 Institution of Engineering and Technology. All rights reserved.; 2017 International Conference on Radar Systems, Radar 2017 ; Conference date: 23-10-2017 Through 26-10-2017",

year = "2017",

language = "英语",

}

TY - CONF

T1 - High-order interacting multiple model filter based on mixture transition distribution

AU - Zhou, Gongjian

AU - Ye, Xiaoping

PY - 2017

Y1 - 2017

N2 - A novel nth-order interacting multiple model (IMMn) filter based on the mixture transition distribution (MTD) for maneuvering target tracking referred as IMMn-MTD filter is proposed here. It adopts MTD model rather than fully parameterized high-order Markov chain (MC) to approximate the model sequence transition probabilities. Since only a few independent parameters need to be fixed in MTD model, IMMn-MTD is more flexible compared with the high-order MC mothed. This approach simultaneously possesses the advantage in accuracy because high order model sequence is evaluated. Simulations are performed to demonstrate the effectiveness of the proposed approach.

AB - A novel nth-order interacting multiple model (IMMn) filter based on the mixture transition distribution (MTD) for maneuvering target tracking referred as IMMn-MTD filter is proposed here. It adopts MTD model rather than fully parameterized high-order Markov chain (MC) to approximate the model sequence transition probabilities. Since only a few independent parameters need to be fixed in MTD model, IMMn-MTD is more flexible compared with the high-order MC mothed. This approach simultaneously possesses the advantage in accuracy because high order model sequence is evaluated. Simulations are performed to demonstrate the effectiveness of the proposed approach.

KW - High-order MC

KW - IMMn

KW - MTD

KW - Model switching

UR - https://www.scopus.com/pages/publications/85048653377

M3 - 论文

AN - SCOPUS:85048653377

T2 - 2017 International Conference on Radar Systems, Radar 2017

Y2 - 23 October 2017 through 26 October 2017

ER -

High-order interacting multiple model filter based on mixture transition distribution

Abstract

Conference

Keywords

Other files and links

Fingerprint

Cite this