A novel nonlinear filter through constructing the parametric Gaussian regression process

In this paper, a new variational Gaussian regression filter (VGRF) is proposed by constructing the linear parametric Gaussian regression (LPGR) process including variational parameters. Through modeling the measurement likelihood by LPGR to implement the Bayesian update, the nonlinear measurement function will not be directly involved in the state estimation. The complex Monte Carlo computation used in traditional methods is also avoided well. Hence, in PVFF, the inference of state posteriori and variational parameters can be achieved tractably and simply by using variational Bayesian inference approach. Secondly, a filtering evidence lower bound (F-ELBO) is proposed as a quantitative evaluation rule of different filters. Compared with traditional methods, the higher estimation accuracy of VGRF can be explained by the F-ELBO. Thirdly, a relationship between F-ELBO and the monitored ELBO (M-ELBO) is found, i.e., F-ELBO is always larger than M-ELBO. Based on this finding, the accuracy performance improvement of VGRF can be theoretically explained. Finally, three numerical examples are employed to illustrate the effectiveness of VGRF.

[1]  Simon Haykin,et al.  Cubature Kalman smoothers , 2011, Autom..

[2]  Junxing Zhang,et al.  Object tracking using Langevin Monte Carlo particle filter and locality sensitive histogram based likelihood model , 2018, Comput. Graph..

[3]  Quan Pan,et al.  A Gaussian approximation recursive filter for nonlinear systems with correlated noises , 2012, Autom..

[4]  Hugh F. Durrant-Whyte,et al.  A new method for the nonlinear transformation of means and covariances in filters and estimators , 2000, IEEE Trans. Autom. Control..

[5]  Chunguang Li,et al.  Distributed Variational Bayesian Algorithms Over Sensor Networks , 2016, IEEE Transactions on Signal Processing.

[6]  Quan Pan,et al.  Design and implementation of Gaussian filter for nonlinear system with randomly delayed measurements and correlated noises , 2014, Appl. Math. Comput..

[7]  V. Šmídl,et al.  The Variational Bayes Method in Signal Processing , 2005 .

[8]  Ming Xin,et al.  High-degree cubature Kalman filter , 2013, Autom..

[9]  S. Andrew Gadsden,et al.  Gaussian filters for parameter and state estimation: A general review of theory and recent trends , 2017, Signal Process..

[10]  Juan M. Corchado,et al.  Approximate Gaussian conjugacy: parametric recursive filtering under nonlinearity, multimodality, uncertainty, and constraint, and beyond , 2017, Frontiers of Information Technology & Electronic Engineering.

[11]  Henry Leung,et al.  Variational Bayesian Adaptive Cubature Information Filter Based on Wishart Distribution , 2017, IEEE Transactions on Automatic Control.

[12]  Niels Kjølstad Poulsen,et al.  New developments in state estimation for nonlinear systems , 2000, Autom..

[13]  K. Fujimoto,et al.  On variational Bayes for identification of nonlinear state-space models with linearly dependent unknown parameters , 2017, 2017 56th Annual Conference of the Society of Instrument and Control Engineers of Japan (SICE).

[14]  Fu-Chuang Chen,et al.  Optimal solution of the two-stage Kalman estimator , 1995, Proceedings of 1995 34th IEEE Conference on Decision and Control.

[15]  David M. Blei,et al.  Variational Inference: A Review for Statisticians , 2016, ArXiv.

[16]  Markus Flierl,et al.  Bayesian estimation of Dirichlet mixture model with variational inference , 2014, Pattern Recognit..

[17]  Abhinoy Kumar Singh,et al.  Transformed cubature quadrature Kalman filter , 2017, IET Signal Process..

[18]  Zoubin Ghahramani,et al.  Second-Order Latent-Space Variational Bayes for Approximate Bayesian Inference , 2008, IEEE Signal Processing Letters.

[19]  Xinjian Xiang,et al.  Dirichlet Process Mixture Model Based Nonparametric Bayesian Modeling and Variational Inference , 2019, 2019 Chinese Automation Congress (CAC).

[20]  F. W. Cathey,et al.  The iterated Kalman filter update as a Gauss-Newton method , 1993, IEEE Trans. Autom. Control..

[21]  H. Robbins.  A Stochastic Approximation Method , 1951 .

[22]  Ángel F. García-Fernández,et al.  Posterior Linearization Filter: Principles and Implementation Using Sigma Points , 2015, IEEE Transactions on Signal Processing.

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

[24]  Quan Pan,et al.  An Iterative Nonlinear Filter Using Variational Bayesian Optimization , 2018, Sensors.

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

[26]  Zhongliang Jing,et al.  A Consensus Nonlinear Filter With Measurement Uncertainty in Distributed Sensor Networks , 2017, IEEE Signal Processing Letters.

[27]  Alfred Mertins,et al.  Boosting Black-Box Variational Inference by Incorporating the Natural Gradient , 2018, 2018 24th International Conference on Pattern Recognition (ICPR).

[28]  S. Haykin,et al.  Cubature Kalman Filters , 2009, IEEE Transactions on Automatic Control.

[29]  Nader Ghadimi,et al.  Target tracking with fast adaptive revisit time based on steady state IMM filter , 2017, Digit. Signal Process..

[30]  Zhongliang Jing,et al.  Maneuvering multi-target tracking based on variable structure multiple model GMCPHD filter , 2017, Signal Process..

[31]  Eduardo Mario Nebot,et al.  Approximate Inference in State-Space Models With Heavy-Tailed Noise , 2012, IEEE Transactions on Signal Processing.

[32]  Hong Xu,et al.  Black box variational inference to adaptive kalman filter with unknown process noise covariance matrix , 2020, Signal Process..

[33]  John W. Paisley,et al.  Nonlinear Kalman Filtering With Divergence Minimization , 2017, IEEE Transactions on Signal Processing.

[34]  Nasser M. Nasrabadi,et al.  Pattern Recognition and Machine Learning , 2006, Technometrics.

[35]  Yonggang Zhang,et al.  A Novel Adaptive Kalman Filter With Inaccurate Process and Measurement Noise Covariance Matrices , 2018, IEEE Transactions on Automatic Control.

[36]  Hedvig Kjellström,et al.  Advances in Variational Inference , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[37]  Kevin P. Murphy,et al.  Machine learning - a probabilistic perspective , 2012, Adaptive computation and machine learning series.

[38]  Aristidis Likas,et al.  Bayesian feature and model selection for Gaussian mixture models , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[39]  Behrouz Safarinejadian,et al.  A novel distributed variational approximation method for density estimation in sensor networks , 2016 .

[40]  Baiqing Hu,et al.  A Variational Bayesian-Based Unscented Kalman Filter With Both Adaptivity and Robustness , 2016, IEEE Sensors Journal.

[41]  Qi Hu,et al.  A standard PHD filter for joint tracking and classification of maneuvering extended targets using random matrix , 2018, Signal Process..

[42]  Guohua Gu,et al.  Infrared small target tracking based on sample constrained particle filtering and sparse representation , 2017 .

[43]  Ángel F. García-Fernández,et al.  Iterative Filtering and Smoothing in Nonlinear and Non-Gaussian Systems Using Conditional Moments , 2018, IEEE Signal Processing Letters.

[44]  Sean Gerrish,et al.  Black Box Variational Inference , 2013, AISTATS.

[45]  Ming Xin,et al.  Sparse-grid quadrature nonlinear filtering , 2012, Autom..

[46]  José Jaime Da Cruz,et al.  Complete offline tuning of the unscented Kalman filter , 2017, Autom..

[47]  Quan Pan,et al.  Measurement Random Latency Probability Identification , 2016, IEEE Transactions on Automatic Control.