Previous |  Up |  Next


stochastic systems; nonlinear filtering; particle filter; sample size; adaptation
The paper deals with the particle filter in state estimation of a discrete-time nonlinear non-Gaussian system. The goal of the paper is to design a sample size adaptation technique to guarantee a quality of a filtering estimate produced by the particle filter which is an approximation of the true filtering estimate. The quality is given by a difference between the approximate filtering estimate and the true filtering estimate. The estimate may be a point estimate or a probability density function estimate. The proposed technique adapts the sample size to keep the difference within pre-specified bounds with a pre-specified probability. The particle filter with the proposed sample size adaptation technique is illustrated in a numerical example.
[1] Gordon, N., Salmond, D., Smith, A. F. M.: Novel approach to nonlinear/ non-Gaussian Bayesian state estimation. IEE Proc. F 140 (1993), 107–113, 1993.
[2] Doucet, A., Freitas, N. de, Gordon, N., eds.: Sequential Monte Carlo Methods in Practice. Springer, 2001. MR 1847783 | Zbl 0967.00022
[3] Straka, O., Šimandl, M.: Sampling densities of particle filter: a survey and comparison. In: Proc. 26th American Control Conference (ACC), AACC, New York 2007, pp. 4437–4442.
[4] Šimandl, M., Straka, O.: Nonlinear estimation by particle filters and Cramér–Rao bound. In: Proc. 15th Triennial World Congress of the IFAC, Barcelona 2002, pp. 79–84.
[5] Šimandl, M., Královec, J., Tichavský, P.: Filtering, predictive and smoothing Cramér–Rao bounds for discrete-time nonlinear dynamic filters. Automatica 37 (2001), 11, 1703–1716. DOI 10.1016/S0005-1098(01)00136-4 | MR 2108831
[6] Koller, D., Fratkina, R.: Using learning for approximation in stochastic processes. In: Proc. 15th Internat. Conference on Machine Learning, Morgan Kaufmann, San Francisco 1998, pp. 287–295.
[7] Fox, D.: Adapting the sample size in particle filters through KLD-sampling. Internat. J. Robotics Research 22 (2003), 985–1003. DOI 10.1177/0278364903022012001
[8] Soto, A.: Self adaptive particle filter. In: Proc. Internat. Joint Conference on Artificial Intelligence 2005, pp. 1398–1406.
[9] Straka, O., Šimandl, M.: Adaptive particle filter based on fixed efficient sample size. In: Proc. 14th IFAC Symposium on System Identification, Newcastle 2006.
[10] Lanz, O.: An information theoretic rule for sample size adaptation in particle filtering. In: 14th Internat. Conference on Image Analysis and Processing (ICIAP 2007), pp. 317–322.
[11] Straka, O., In, M. Šimandl. Adaptive particle filter with fixed empirical density quality.: Proc. 17th World Congress of the IFAC, 2008.
[12] Straka, O., Šimandl, M.: A survey of sample size adaptation techniques for particle filters. In: Proc. 15th Symposium on System Identification, IFAC, Saint-Malo 2009.
[13] Karlsson, R., Gustafsson, F.: Monte Carlo data association for multiple target tracking. In: IEE Workshop on Target Tracking, Eindhoven 2001.
[14] Bolic, M., Hong, S., Djuric, P. M.: Performance and complexity analysis of adaptive particle filtering for tracking applications. In: Proc. Conference on Signals, Systems and Computers, 2002.
[15] Kerridge, D. F.: Inaccuracy and inference. J. Roya. Statist. Soc. 23 (1961), 184–194. MR 0123375 | Zbl 0112.10302
[16] Hayya, J., Armstrong, D., Gressis, N.: A note on the ratio of two normally distributed variables. Management Sci. 21 (1975), 11, 338–1341. DOI 10.1287/mnsc.21.11.1338 | Zbl 0309.62011
[17] Merwe, R. van der, Wan, E. A.: Sigma-point particle filters for sequential probabilistic inference in dynamic state-space models. In: Proc. Internat.l Conference on Acoustics, Speech, and Signal Processing (ICASSP), IEEE, Hong Kong 2003, p. 4.
[18] Šimandl, M., Královec, J., Söderström, T.: Advanced point – mass method for nonlinear state estimation. Automatica 42 (2006), 7, 1133–1145. DOI 10.1016/j.automatica.2006.03.010 | Zbl 1118.93052
Partner of
EuDML logo