Abstract—Cubature particle filter with Markovian Chain Monte Carlo process (CPF-MC) is proposed in order to alleviate the degeneracy and impoverishment problems existing in the particle filter, and the CPF-MC algorithm is improved from two aspects. On the one hand, CPF-MC uses the square root cubature Kalman filter (SRCKF) as proposal distribution that integrates the latest measurement into the particle filter and approximates the optimal posterior probability distribution more accurately. On the other hand, after resampling, a SRCKF based Markovian Chain Monte Carlo (MCMC) process is used to make the particles diversity and suppress the impoverish phenomenon. The CPF-MC algorithm is applied to re-entry ballistic target tracking; simulation results demonstrate that the CPF-MC achieves the better performance and is superior to generic particle filter with MCMC (GPF-MC), extended particle filter with MCMC (EPF-MC) and unscented particle filter with MCMC (UPF-MC), furthermore, the CPF-MC run faster.
Index Terms—Markovian chain Monte Carlo, nonlinear filter, particle filter, re-entry ballistic target
J. Mu is with the Xi’an Technological University, Xi’an, 710032 China (e-mail: mujing1977@163.com). Y. Cai is with the Xi’an Jiaotong University, Xi’an, 710046 China (e-mail: ylicai@mail.xjtu.edu.cn)
[PDF]
Cite: Jing Mu and Yuan liCai, "Cubature Particle Filter with MCMC and Applications to Re-entry Ballistic Target Tracking," International Journal of Engineering and Technology vol. 8, no. 1, pp. 65-69, 2016.
