The Improved Particle Filter Algorithm Based on Weight Optimization

Abstract

Particle filter algorithm is to achieve recursive Bayesian filter through the simulation method of non-parameter Monte Carlo, It based on sequential importance sampling, and can not avoid particle degeneration problem, a way to overcome the particle degradation is re-sampling, However sample impoverishment will appear in the process of re-sampling, This paper proposes an improved particle filter method based on optimized weight. To some extent, the method solves the particle impoverishment problem, According to simulation results, we can confirm that the improved particle filter algorithm proposed in the paper can effectively improve the estimation precision of particle filter algorithm.