Multivariable feedback particle filter

DOI10.1016/J.AUTOMATICA.2016.04.019zbMATH Open1343.93088arXiv1303.1205OpenAlexW1992747726MaRDI QIDQ313150FDOQ313150

Sean P. Meyn, Prashant G. Mehta, R. S. Laugesen, Tao Yang

Publication date: 9 September 2016

Published in: Automatica (Search for Journal in Brave)

Abstract: In recent work it is shown that importance sampling can be avoided in the particle filter through an innovation structure inspired by traditional nonlinear filtering combined with Mean-Field Game formalisms. The resulting feedback particle filter (FPF) offers significant variance improvements; in particular, the algorithm can be applied to systems that are not stable. The filter comes with an up-front computational cost to obtain the filter gain. This paper describes new representations and algorithms to compute the gain in the general multivariable setting. The main contributions are, (i) Theory surrounding the FPF is improved: Consistency is established in the multivariate setting, as well as well-posedness of the associated PDE to obtain the filter gain. (ii) The gain can be expressed as the gradient of a function, which is precisely the solution to Poisson's equation for a related MCMC diffusion (the Smoluchowski equation). This provides a bridge to MCMC as well as to approximate optimal filtering approaches such as TD-learning, which can in turn be used to approximate the gain. (iii) Motivated by a weak formulation of Poisson's equation, a Galerkin finite-element algorithm is proposed for approximation of the gain. Its performance is illustrated in numerical experiments.

Full work available at URL: https://arxiv.org/abs/1303.1205

zbMATH Keywords

particle filteringestimation theorynonlinear filtering

Mathematics Subject Classification ID

Filtering in stochastic control theory (93E11) Nonlinear systems in control theory (93C10) Multivariable systems, multidimensional control systems (93C35) Estimation and detection in stochastic control theory (93E10)

Cites Work

• Sequential Monte Carlo Methods in Practice
• Title not available (Why is that?)
• Curse-of-dimensionality revisited: Collapse of the particle filter in very large scale systems
• Filtering via Simulation: Auxiliary Particle Filters
• Title not available (Why is that?)
• Stochastic differential equations. An introduction with applications.
• Mean field games
• Higher-order implicit strong numerical schemes for stochastic differential equations
• Following a moving target -- Monte Carlo inference for dynamic Bayesian models
• Title not available (Why is that?)
• Title not available (Why is that?)
• Large-Population Cost-Coupled LQG Problems With Nonuniform Agents: Individual-Mass Behavior and Decentralized $\varepsilon$-Nash Equilibria
• A Liapounov bound for solutions of the Poisson equation
• Title not available (Why is that?)
• Synchronization of Coupled Oscillators is a Game
• Conditional Markov Processes
• Mean field simulation for Monte Carlo integration
• On the spectral properties of Witten-Laplacians, their range projections and Brascamp-Lieb's inequality
• A survey of numerical methods for nonlinear filtering problems
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Multivariable feedback particle filter
• A Variational Approach to Nonlinear Estimation
• Title not available (Why is that?)
• Poisson's Equation in Nonlinear Filtering
• A dynamical systems framework for intermittent data assimilation
• On the Differential Equations Satisfied by Conditional Probablitity Densities of Markov Processes, with Applications
• Feedback Particle Filter
• Marginalized particle filters for mixed linear/nonlinear state-space models
• Concentration inequalities for mean field particle models
• Propagation of chaos and Poincaré inequalities for a system of particles interacting through their CDF

Cited In (22)

• Gradient iterative algorithm for dual-rate nonlinear systems based on a novel particle filter
• Nonlinear unknown input observer based on singular value decomposition aided reduced dimension cubature Kalman filter
• Coupled quantum harmonic oscillators and Feynman-Kac path integrals for linear diffusive particles
• On the stability of linear feedback particle filter
• Multivariate feedback particle filter rederived from the splitting-up scheme
• PF-MPC: particle filter-model predictive control
• Controlled interacting particle algorithms for simulation-based reinforcement learning
• Time-varying feedback particle filter
• Feedback Particle Filter on Riemannian Manifolds and Matrix Lie Groups
• An optimal control approach to particle filtering
• How to Avoid the Curse of Dimensionality: Scalability of Particle Filters with and without Importance Weights
• On the mathematical theory of ensemble (linear-Gaussian) Kalman-Bucy filtering
• A theoretical analysis of one-dimensional discrete generation ensemble Kalman particle filters
• Diffusion Map-based Algorithm for Gain Function Approximation in the Feedback Particle Filter
• Backward Nonlinear Smoothing Diffusions
• Multivariable feedback particle filter
• The Hitchhiker's guide to nonlinear filtering
• Feedback particle filter for collective inference
• Multilevel Particle Filters
• A Unification of Weighted and Unweighted Particle Filters
• Discrete-time linear skew-Gaussian system and its recursive fixed-dimensional exact density filtering
• On one-dimensional Riccati diffusions

This page was built for publication: Multivariable feedback particle filter

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q313150)