Particle Flow Gaussian Particle Filter

TL;DR

This paper introduces a novel particle filter method that uses invertible particle flow to generate proposals close to the posterior, improving state estimation in high-dimensional non-linear models.

Contribution

The paper proposes the particle flow Gaussian particle filter (PFGPF), combining invertible particle flow with Gaussian particle filtering for enhanced performance.

Findings

PFGPF retains asymptotic properties of Gaussian particle filters.

PFGPF shows improved performance in high-dimensional simulations.

Compared with existing particle flow filters, PFGPF offers better state estimation.

Abstract

State estimation in non-linear models is performed by tracking the posterior distribution recursively. A plethora of algorithms have been proposed for this task. Among them, the Gaussian particle filter uses a weighted set of particles to construct a Gaussian approximation to the posterior. In this paper, we propose to use invertible particle flow methods, derived under the Gaussian boundary conditions for a flow equation, to generate a proposal distribution close to the posterior. The resultant particle flow Gaussian particle filter (PFGPF) algorithm retains the asymptotic properties of Gaussian particle filters, with the potential for improved state estimation performance in high dimensional spaces. We compare the performance of PFGPF with the particle flow filters and particle flow particle filters in two challenging numerical simulation examples.