A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking Export

@article{citeulike:431128, abstract = {Increasingly, for many application areas, it is becoming important to include elements of nonlinearity and non-Gaussianity in order to model accurately the underlying dynamics of a physical system. Moreover, it is typically crucial to process data on-line as it arrives, both from the point of view of storage costs as well as for rapid adaptation to changing signal characteristics. In this paper, we review both optimal and suboptimal Bayesian algorithms for nonlinear/non-Gaussian tracking problems, with a focus on particle filters. Particle filters are sequential Monte Carlo methods based on point mass (or "particle") representations of probability densities, which can be applied to any state-space model and which generalize the traditional Kalman filtering methods. Several variants of the particle filter such as SIR, ASIR, and RPF are introduced within a generic framework of the sequential importance sampling (SIS) algorithm. These are discussed and compared with the standard EKF through an illustrative example}, author = {Arulampalam, M. S. and Maskell, S. and Gordon, N. and Clapp, T.}, booktitle = {Signal Processing, IEEE Transactions on}, citeulike-article-id = {431128}, citeulike-linkout-0 = {http://dx.doi.org/10.1109/78.978374}, citeulike-linkout-1 = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=978374}, doi = {10.1109/78.978374}, journal = {Signal Processing, IEEE Transactions on}, keywords = {filter, particle}, number = {2}, pages = {174--188}, posted-at = {2009-03-19 00:23:18}, priority = {2}, title = {A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking}, url = {http://dx.doi.org/10.1109/78.978374}, volume = {50}, year = {2002} }

TY - JOUR ID - citeulike:431128 L3 - citeulike-article-id:431128 N2 - Increasingly, for many application areas, it is becoming important to include elements of nonlinearity and non-Gaussianity in order to model accurately the underlying dynamics of a physical system. Moreover, it is typically crucial to process data on-line as it arrives, both from the point of view of storage costs as well as for rapid adaptation to changing signal characteristics. In this paper, we review both optimal and suboptimal Bayesian algorithms for nonlinear/non-Gaussian tracking problems, with a focus on particle filters. Particle filters are sequential Monte Carlo methods based on point mass (or "particle") representations of probability densities, which can be applied to any state-space model and which generalize the traditional Kalman filtering methods. Several variants of the particle filter such as SIR, ASIR, and RPF are introduced within a generic framework of the sequential importance sampling (SIS) algorithm. These are discussed and compared with the standard EKF through an illustrative example IS - 2 JF - Signal Processing, IEEE Transactions on EP - 188 TI - A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking VL - 50 T2 - Signal Processing, IEEE Transactions on SP - 174 KW - filter KW - particle AU - Arulampalam, MS AU - Maskell, S AU - Gordon, N AU - Clapp, T PY - 2002/// UR - http://dx.doi.org/10.1109/78.978374 ER -