Application of Gaussian Mixture Particle Filter on State Estimation
Xiaogang Yang · Journal of Projectiles.Rockets.Missiles and Guidance · 2007
In standard particle filter(PF),the principal problem which influences the estimation performance is sample depletion brought by resampling step.To solve the problem,this paper presents an improved PF algorithm, the posterior state density that is represented by a Gaussian mixture model is recovered from the particle set of the measurement update step by means of a weighted EM algorithm.This step replaces the resampling stage needed by most particle set and reduced computional complexity compared to other related algorithms. It is demonstrated by simulation that this new approach has an improved estimation performance and reduced computional complexity compared to other related algorithms.