Feedback Particle Filter With Data-Driven Gain-Function Approximation

This paper addresses the continuous discrete-time nonlinear filtering problem for stochastic dynamical systems using the feedback particle filter (FPF). The FPF updates each particle using feedback from the measurements, where the gain function that controls the particles is the solution of a Poisson equation. The main difficulty in the FPF is to approximate this solution using the particles that approximate the probability distribution. We develop a novel Galerkin-based method inspired by high-dimensional data-analysis techniques. Based on the time evolution of the particle cloud, we determine basis functions for the gain function and compute values of it for each individual particle. Our method is completely adapted to the recorded history of the particles and the update of the particles do not require further intermediate approximations or assumptions. We provide an extensive numerical evaluation of the proposed approach and show that it compares favorably compared to baseline FPF and particle filters based on the importance-sampling paradigm.

[1]  A. Doucet,et al.  A Tutorial on Particle Filtering and Smoothing: Fifteen years later , 2008 .

[2]  Sean P. Meyn,et al.  Feedback Particle Filter , 2013, IEEE Transactions on Automatic Control.

[3]  Thomas B. Schön,et al.  Marginalized particle filters for mixed linear/nonlinear state-space models , 2005, IEEE Transactions on Signal Processing.

[4]  Simon J. Godsill,et al.  An Overview of Existing Methods and Recent Advances in Sequential Monte Carlo , 2007, Proceedings of the IEEE.

[5]  Amirhossein Taghvaei,et al.  An optimal transport formulation of the linear feedback particle filter , 2015, 2016 American Control Conference (ACC).

[6]  Karl Berntorp.  Particle filter for combined wheel-slip and vehicle-motion estimation , 2015, 2015 American Control Conference (ACC).

[7]  Lingling Zhao,et al.  Particle flow auxiliary particle filter , 2015, 2015 IEEE 6th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP).

[8]  Gene H. Golub,et al.  Matrix computations (3rd ed.) , 1996 .

[9]  Fred Daum,et al.  Particle flow with non-zero diffusion for nonlinear filters , 2013, Defense, Security, and Sensing.

[10]  Karl Berntorp,et al.  Feedback particle filter: Application and evaluation , 2015, 2015 18th International Conference on Information Fusion (Fusion).

[11]  Karl Berntorp,et al.  Joint Wheel-Slip and Vehicle-Motion Estimation Based on Inertial, GPS, and Wheel-Speed Sensors , 2016, IEEE Transactions on Control Systems Technology.

[12]  F Gustafsson,et al.  Particle filter theory and practice with positioning applications , 2010, IEEE Aerospace and Electronic Systems Magazine.

[13]  Sean P. Meyn,et al.  Gain function approximation in the feedback particle filter , 2016, 2016 IEEE 55th Conference on Decision and Control (CDC).

[14]  B. Feeny,et al.  On the physical interpretation of proper orthogonal modes in vibrations , 1998 .

[15]  J. Guermond,et al.  Theory and practice of finite elements , 2004 .

[16]  Pete Bunch,et al.  Approximations of the Optimal Importance Density Using Gaussian Particle Flow Importance Sampling , 2014, 1406.3183.

[17]  Uwe D. Hanebeck,et al.  Semi-analytic Gaussian Assumed Density Filter , 2011, Proceedings of the 2011 American Control Conference.

[18]  Uwe D. Hanebeck,et al.  Progressive Bayes: a new framework for nonlinear state estimation , 2003, SPIE Defense + Commercial Sensing.

[19]  J. Zabczyk,et al.  Wong-Zakai approximations of stochastic evolution equations , 2006 .

[20]  Prashant G. Mehta,et al.  A comparative study of nonlinear filtering techniques , 2013, Proceedings of the 16th International Conference on Information Fusion.

[21]  S. Särkkä,et al.  On Unscented Kalman Filtering for State Estimation of Continuous-Time Nonlinear Systems , 2007, IEEE Transactions on Automatic Control.

[22]  Johan Dahlin,et al.  Sequential Monte Carlo Methods for System Identification , 2015, 1503.06058.

[23]  Fredrik Gustafsson,et al.  Storage efficient particle filters for the out of sequence measurement problem , 2008, 2008 11th International Conference on Information Fusion.

[24]  Sean P. Meyn,et al.  Multivariable feedback particle filter , 2016, Autom..

[25]  G. Kerschen,et al.  The Method of Proper Orthogonal Decomposition for Dynamical Characterization and Order Reduction of Mechanical Systems: An Overview , 2005 .

[26]  Simon J. Godsill,et al.  On sequential Monte Carlo sampling methods for Bayesian filtering , 2000, Stat. Comput..

[27]  A. Chatterjee.  An introduction to the proper orthogonal decomposition , 2000 .

[28]  Neil J. Gordon,et al.  A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking , 2002, IEEE Trans. Signal Process..

[29]  Tao Yang,et al.  The continuous-discrete time feedback particle filter , 2014, 2014 American Control Conference.

[30]  X. R. Li,et al.  Survey of maneuvering target tracking. Part I. Dynamic models , 2003 .

[31]  Simon J. Godsill,et al.  Particle filtering with progressive Gaussian approximations to the optimal importance density , 2013, 2013 5th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP).

[32]  Wolfram Burgard,et al.  Improved Techniques for Grid Mapping With Rao-Blackwellized Particle Filters , 2007, IEEE Transactions on Robotics.

[33]  Karl-Erik Årzén,et al.  Storage efficient particle filters with multiple out-of-sequence measurements , 2012, 2012 15th International Conference on Information Fusion.

[34]  Fredrik Gustafsson,et al.  Statistical Sensor Fusion , 2013 .

[35]  Karl Berntorp,et al.  Data-driven gain computation in the feedback particle filter , 2016, 2016 American Control Conference (ACC).

[36]  N. Gordon,et al.  Novel approach to nonlinear/non-Gaussian Bayesian state estimation , 1993 .

[37]  Sean P. Meyn,et al.  Poisson's equation in nonlinear filtering , 2014, 53rd IEEE Conference on Decision and Control.

[38]  Tao Ding,et al.  Implementation of the Daum-Huang exact-flow particle filter , 2012, 2012 IEEE Statistical Signal Processing Workshop (SSP).

[39]  Sean P. Meyn,et al.  A mean-field control-oriented approach to particle filtering , 2011, Proceedings of the 2011 American Control Conference.

[40]  Karl Berntorp,et al.  Process-noise adaptive particle filtering with dependent process and measurement noise , 2016, 2016 IEEE 55th Conference on Decision and Control (CDC).

[41]  Fred Daum,et al.  Nonlinear filters with particle flow , 2009, Optical Engineering + Applications.

[42]  Fredrik Lindsten,et al.  Particle gibbs with ancestor sampling , 2014, J. Mach. Learn. Res..