Marginalized adaptive particle filtering for nonlinear models with unknown time-varying noise parameters

Knowledge of the noise distribution is typically crucial for the state estimation of general state-space models. However, properties of the noise process are often unknown in the majority of practical applications. The distribution of the noise may also be non-stationary or state dependent and that prevents the use of off-line tuning methods. For linear Gaussian models, Adaptive Kalman filters (AKF) estimate unknown parameters in the noise distributions jointly with the state. For nonlinear models, we provide a Bayesian solution for the estimation of the noise distributions in the exponential family, leading to a marginalized adaptive particle filter (MAPF) where the noise parameters are updated using finite dimensional sufficient statistics for each particle. The time evolution model for the noise parameters is defined implicitly as a Kullback-Leibler norm constraint on the time variability, leading to an exponential forgetting mechanism operating on the sufficient statistics. Many existing methods are based on the standard approach of augmenting the state with the unknown variables and attempting to solve the resulting filtering problem. The MAPF is significantly more computationally efficient than a comparable particle filter that runs on the full augmented state. Further, the MAPF can handle sensor and actuator offsets as unknown means in the noise distributions, avoiding the standard approach of augmenting the state with such offsets. We illustrate the MAPF on first a standard example, and then on a tire radius estimation problem on real data.

[1]  Joris De Schutter,et al.  Adaptive Kalman filter for noise identification , 2000 .

[2]  R. Mehra.  Approaches to adaptive filtering , 1972 .

[3]  R. Maine,et al.  Formulation and implementation of a practical algorithm for parameter estimation with process and measurement noise , 1980 .

[4]  Miroslav Kárný,et al.  Optimized bayesian dynamic advising : theory and algorithms , 2006 .

[5]  Y. Bar-Shalom,et al.  A recursive multiple model approach to noise identification , 1994 .

[6]  A. Jazwinski.  Stochastic Processes and Filtering Theory , 1970 .

[7]  Adrian E. Raftery,et al.  Prediction under Model Uncertainty Via Dynamic Model Averaging : Application to a Cold Rolling Mill 1 , 2008 .

[8]  E. Jaynes.  Information Theory and Statistical Mechanics , 1957 .

[9]  Fredrik Gustafsson,et al.  Marginalized particle filters for Bayesian estimation of Gaussian noise parameters , 2010, 2010 13th International Conference on Information Fusion.

[10]  Marcelo G. S. Bruno,et al.  Bayesian blind equalization of time-varying frequency-selective channels subject to unknown variance noise , 2008, 2008 IEEE International Conference on Acoustics, Speech and Signal Processing.

[11]  Arnaud Doucet,et al.  A survey of convergence results on particle filtering methods for practitioners , 2002, IEEE Trans. Signal Process..

[12]  H. Jeffreys,et al.  Theory of probability , 1896 .

[13]  Edwin T. Jaynes,et al.  Prior Probabilities , 1968, Encyclopedia of Machine Learning.

[14]  R. Mehra.  On the identification of variances and adaptive Kalman filtering , 1970 .

[15]  Richard A. Tapia,et al.  An Extension of the Karush-Kuhn-Tucker Necessity Conditions to Infinite Programming , 1994, SIAM Rev..

[16]  A. Quinn,et al.  Bayesian estimation of non-stationary AR model parameters via an unknown forgetting factor , 2004, 3rd IEEE Signal Processing Education Workshop. 2004 IEEE 11th Digital Signal Processing Workshop, 2004..

[17]  Jacob Benesty,et al.  A Robust Variable Forgetting Factor Recursive Least-Squares Algorithm for System Identification , 2008, IEEE Signal Processing Letters.

[18]  Simo Särkkä,et al.  Recursive Noise Adaptive Kalman Filtering by Variational Bayesian Approximations , 2009, IEEE Transactions on Automatic Control.

[19]  Petar M. Djuric,et al.  Sequential particle filtering in the presence of additive Gaussian noise with unknown parameters , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[20]  M. Kárný,et al.  Approximate Bayesian Recursive Estimation On Approximation Errors No . 2317 January 26 , 2012 , 2012 .

[21]  Geir Storvik,et al.  Particle filters for state-space models with the presence of unknown static parameters , 2002, IEEE Trans. Signal Process..

[22]  Eleanor Jaynes,et al.  Information Theory and Statistical Mechanics (Notes by the lecturer) , 1963 .

[23]  Jean-Michel Marin,et al.  On Particle Learning , 2010, 1006.0554.

[24]  V. Peterka.  BAYESIAN APPROACH TO SYSTEM IDENTIFICATION , 1981 .

[25]  A. Doucet,et al.  On-Line Parameter Estimation in General State-Space Models , 2005, Proceedings of the 44th IEEE Conference on Decision and Control.

[26]  Dan Simon,et al.  Kalman Filtering with Uncertain Noise Covariances , 2004 .

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

[28]  Quan Pan,et al.  A finite-horizon adaptive Kalman filter for linear systems with unknown disturbances , 2004, Signal Process..

[29]  B. Tapley,et al.  Adaptive sequential estimation with unknown noise statistics , 1976 .

[30]  Michael A. West,et al.  Combined Parameter and State Estimation in Simulation-Based Filtering , 2001, Sequential Monte Carlo Methods in Practice.

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

[32]  R. Kulhavý,et al.  On a general concept of forgetting , 1993 .

[33]  Nicholas G. Polson,et al.  Particle Learning and Smoothing , 2010, 1011.1098.