TY - GEN
T1 - New Environmental Dependent Modelling with Gaussian Particle Filtering Based Implementation for Ground Vehicle Tracking
AU - Yu, Miao
AU - Xue, Yali
AU - Ding, Runxiao
AU - Oh, Hyondong
AU - Chen, Wen Hua
AU - Chambers, Jonathon
N1 - Publisher Copyright:
© 2016 IEEE.
PY - 2016/10/13
Y1 - 2016/10/13
N2 - This paper proposes a new domain knowledge aided Gaussian particle filtering based approach for the ground vehicle tracking application. Firstly, a new form of modelling is proposed to reflect the influences of different types of environmental domain knowledge on the vehicle dynamic: i) a non-Markov jump model is applied with multiple models while transition probabilities between models are environmental dependent ii) for a particular model, both the constraints and potential forces obtained from the surrounding environment have been applied to refine the vehicle state distribution. Based on the proposed modelling approach, a Gaussian particle filtering based method is developed to implement the related Bayesian inference for the target state estimation. Simulation studies from multiple Monte Carlo simulations confirm the advantages of the proposed method over traditional ones, from both the modelling and implementation aspects.
AB - This paper proposes a new domain knowledge aided Gaussian particle filtering based approach for the ground vehicle tracking application. Firstly, a new form of modelling is proposed to reflect the influences of different types of environmental domain knowledge on the vehicle dynamic: i) a non-Markov jump model is applied with multiple models while transition probabilities between models are environmental dependent ii) for a particular model, both the constraints and potential forces obtained from the surrounding environment have been applied to refine the vehicle state distribution. Based on the proposed modelling approach, a Gaussian particle filtering based method is developed to implement the related Bayesian inference for the target state estimation. Simulation studies from multiple Monte Carlo simulations confirm the advantages of the proposed method over traditional ones, from both the modelling and implementation aspects.
UR - http://www.scopus.com/inward/record.url?scp=85009739657&partnerID=8YFLogxK
U2 - 10.1109/SSPD.2016.7590608
DO - 10.1109/SSPD.2016.7590608
M3 - Conference article published in proceeding or book
AN - SCOPUS:85009739657
T3 - 2016 Sensor Signal Processing for Defence, SSPD 2016
BT - 2016 Sensor Signal Processing for Defence, SSPD 2016
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 6th Conference of the Sensor Signal Processing for Defence, SSPD 2016
Y2 - 22 September 2016 through 23 September 2016
ER -