Fundamental Principles and Applications of Particle Filters
Ying Liu, Benping Wang, He Wang, Jing Zhao, Zhi Jun Ding · 2006
This paper introduces the key principles and applications of particle filtering. Particle Filters are a class of modern sequential Monte Carlo Bayesian methods based on point mass representation of posterior probability density. They are highly useful in parameter estimation when dealing with nonlinear system models and non-Gaussian noise. After summarizing the basic algorithms used in particle filters, two application examples will be given. The examples are given to demonstrate the application of particle filters for time delay estimation as well as in estimating and tracking signal angle of arrival at an antenna array. We present a new method to get the importance density for time delay and angle of arrival estimation. The relevant conclusions are got and verified by the simulation experiment.