A Novel Particle Filter Based on One-Step Smoothing for Nonlinear System with Missing Measurements

This paper proposes a novel particle filter based on one-step smoothing for nonlinear systems with missing measurements. This filter iteratively employs a one-step smoother to improve the efficiency of importance sampling in bootstrap particle filtering through incorporating current measurement information into the apriori distribution. A simulation example of a target tracking model is given to illustrate that the proposed algorithm improves the sampling efficiency and estimation accuracy, which effectively limits the particle degradation phenomenon.