Study of Particle Filter Algorithm Based on Monte Carlo Methods

This paper analyzes the Bayesian theory and its development in detail,focusing on the nonlinear or non-Gaussian problems.The Kalman filter provides a classic approach to linear-Gaussian estimation problem.However,due to the nonlinear or non-Gaussian in real world,people have to search for a better kind of filter.Based on stochastic filtering theory,Bayesian theory and Monte Carlo methods,the particle filter theory is developing more and more,and applies in the fields such as digital communication,target tracking,computer vision and fault detection.