Abstract
As an alternative to the Kalman filter and the particle filter, the particle flow filter has recently attracted interest for solving the curse of dimensionality of the particle filter. Compared with the particle filter, the particle flow filter can obtain a better performance in high-dimensional state spaces with fewer samples. However, for some unobservable state dimensions, the flow operation wastes computational resources. In this paper, we propose a marginalized particle flow filter to handle the unobservable sub-state estimation. In contrast to the standard particle flow filter, we only migrate those observable dimensions of each particle according to homotopy theory and estimate the unobservable dimensions using the Kalman filter. The proposed algorithm can enhance the estimation quality of the unobservable state space and reduce the runtime of the particle flow filter. We evaluate the performance of the proposed algorithm through a multi-target tracking simulation.
Similar content being viewed by others
References
M.S. Arulampalam, S. Maskell, N. Gordon, T. Clapp, A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking. IEEE Trans. Signal Process. 50(2), 174–188 (2002)
T. Bengtsson, P. Bickel, B. Li et al., Curse-of-dimensionality revisited: collapse of the particle filter in very large scale systems, in Probability and Statistics: Essays in Honor of David A. Freedman, ed. by D. Nolan, T. Speed (Institute of Mathematical Statistics, Beachwood, 2008), pp. 316–334
A. Beskos, D. Crisan, A. Jasra et al., On the stability of sequential Monte Carlo methods in high dimensions. Ann Appl Probab 24(4), 1396–1445 (2014)
F. Beutler, M.F. Huber, U.D. Hanebeck, Gaussian filtering using state decomposition methods, in 12th International Conference on Information Fusion, 2009, FUSION’09 (IEEE, 2009), pp. 579–586
P. Bickel, B. Li, T. Bengtsson et al., Sharp failure rates for the bootstrap particle filter in high dimensions. In Pushing the Limits of Contemporary Statistics: Contributions in Honor of Jayanta K. Ghosh (Institute of Mathematical Statistics, 2008), pp. 318–329
P. Bunch, S. Godsill, Approximations of the optimal importance density using Gaussian particle flow importance sampling. J. Am. Stat. Assoc. 111(514), 748–762 (2016)
B. Chen, G. Hu, D.W.C. Ho, L. Yu, A new approach to linear/nonlinear distributed fusion estimation problem. IEEE Trans. Autom. Control (2018). https://doi.org/10.1109/TAC.2018.2849612
S. Choi, P. Willett, F. Daum, J. Huang, Discussion and application of the homotopy filter, in SPIE Defense, Security, and Sensing (International Society for Optics and Photonics, 2011), p. 805021
R. Costa, T.A. Wettergren, Computationally efficient angles-only tracking with particle flow filters, in SPIE Defense+ Security (International Society for Optics and Photonics, 2015), p. 947404
F. Daum, J. Huang, Nonlinear filters with log-homotopy, in Proc. SPIE, vol. 6699, p. 669918- (2007)
F. Daum, J. Huang, Particle flow for nonlinear filters with log-homotopy, in SPIE Defense and Security Symposium (International Society for Optics and Photonics, 2008), p. 696918
F. Daum, J. Huang, Generalized particle flow for nonlinear filters, in SPIE Defense, Security, and Sensing (International Society for Optics and Photonics, 2010), p. 76980I
F. Daum, J. Huang. Particle flow with non-zero diffusion for nonlinear filters, in Proceedings of SPIE: Signal processing, Sensor Fusion and Target Tracking XXII (2013), p. 87450P
F. Daum, J. Huang, A. Noushin, Exact particle flow for nonlinear filters, in SPIE Defense, Security, and Sensing (International Society for Optics and Photonics, 2010), p. 769704
F. Daum, J. Huang, A. Noushin, Generalized Gromov method for stochastic particle flow filters, in SPIE Defense+ Security (International Society for Optics and Photonics, 2017), p. 102000I
F. Daum, J. Huang, Nonlinear filters with particle flow induced by log-homotopy, in SPIE Defense, Security, and Sensing (International Society for Optics and Photonics, 2009), p. 733603
T. Ding, M. Coates, Implementation of the Daum–Huang exact-flow particle filter. 2012 IEEE Statistical Signal Processing Workshop (SSP) (2012), pp. 257–260
A. Doucet, N. De Freitas, N. Gordon, Sequential Monte Carlo Methods in Practice (Springer, Berlin, 2001)
G. Hendeby, R. Karlsson, F. Gustafsson, The Rao-Blackwellized particle filter: a filter bank implementation. EURASIP J. Adv. Signal Process. 2010(1), 724087 (2010)
S.J. Julier, J.K. Uhlmann, Unscented filtering and nonlinear estimation. Proc. IEEE 92(3), 401–422 (2004)
R. Karlsson, T. Schon, F. Gustafsson, Complexity Analysis of the Marginalized Particle Filter (IEEE Press, New York, 2005)
M.A. Khan, M. Ulmke. Improvements in the implementation of log-homotopy based particle flow filters, in 2015 18th International Conference on Information Fusion (Fusion) (IEEE, 2015), pp. 74–81
Y. Li, M. Coates, Particle filtering with invertible particle flow. IEEE Trans. Signal Process. 65(15), 4102–4116 (2016)
S. Maskell, N. Gordon, A tutorial on particle filters for on-line nonlinear/non-Gaussian Bayesian tracking. Target Track. Algorithms Appl. (Ref. No. 2001/174), 2, 21–215 (2001)
M. Morzfeld, D. Hodyss, C. Snyder, What the collapse of the ensemble Kalman filter tells us about particle filters. Tellus A Dyn. Meteorol. Oceanogr. 69(1), 1283809 (2017)
N. Moshtagh, J. Chan, M. Chan, Homotopy particle filter for ground-based tracking of satellites at GEO, in Advanced Maui Optical and Space Surveillance Technologies Conference (2016)
N. Moshtagh, M.W. Chan, Multisensor fusion using homotopy particle filter, in 2015 18th International Conference on Information Fusion (Fusion) (IEEE, 2015), pp. 1641–1648
M.K. Pitt, N. Shephard, Filtering via simulation: auxiliary particle filters. J. Am. Stat. Assoc. 94(446), 590–599 (1999)
T. Schon, F. Gustafsson, P.J. Nordlund, Marginalized particle filters for mixed linear/nonlinear state-space models. IEEE Trans. Signal Process. 53(7), 2279–2289 (2005)
C. Snyder, T. Bengtsson, P. Bickel, J. Anderson, Obstacles to high-dimensional particle filtering. Mon. Weather Rev. 136(12), 4629–4640 (2008)
R. Van Der Merwe, A. Doucet, N. De Freitas, E.A. Wan, The unscented particle filter, in Advances in Neural Information Processing Systems (2001), pp. 584–590
Acknowledgements
We are grateful to the referees for their clarifying suggestions, which have improved the presentation of this material, and in articular to Jeremie Houssineau for his comments. This study was supported by the National Natural Science Foundation of China (NSFC, Grant No. 61305013).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
In the appendices below, we provide the proofs of Lemma 2 in this article. This procedure consists of two parts. The first part is a Kalman update using the artificial measurement. Moreover, the second step consists of a time update using the result from the first step. Let us start by Eq. (20):
where \({\hat{z}}_{k} = x_{k+1,a} - F^{a,a}x_{k,a}\) is a “virtual” measurement. Analogously to the update of Kalman filter, we obtain
The time update is to compute
where
Rights and permissions
About this article
Cite this article
Wang, J., Zhao, L. & Su, X. Marginalized Particle Flow Filter. Circuits Syst Signal Process 38, 3152–3169 (2019). https://doi.org/10.1007/s00034-018-1007-1
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00034-018-1007-1