Fight sample degeneracy and impoverishment in particle filters: A review of intelligent approaches
作者：
Highlights：
• The cause, essence and interrelationship of sample degeneracy and impoverishment in the particle filter are analyzed.
• Approaches that are effective in alleviating sample degeneracy and impoverishment is presented.
• Approaches that are effective for dealing with high-dimensionality are reviewed.
• A comprehensive survey of artificial intelligence algorithms and machine learning techniques applied in particle filters is presented.
• A common principle to enhance the particle filter, termed as Particle Distribution Optimization (PDO), is established.
摘要
•The cause, essence and interrelationship of sample degeneracy and impoverishment in the particle filter are analyzed.•Approaches that are effective in alleviating sample degeneracy and impoverishment is presented.•Approaches that are effective for dealing with high-dimensionality are reviewed.•A comprehensive survey of artificial intelligence algorithms and machine learning techniques applied in particle filters is presented.•A common principle to enhance the particle filter, termed as Particle Distribution Optimization (PDO), is established.
论文关键词：Particle filter,Sequential Monte Carlo,Markov Chain Monte Carlo,Impoverishment,Artificial intelligence,Machine learning
论文评审过程：Available online 28 December 2013.
论文官网地址：https://doi.org/10.1016/j.eswa.2013.12.031
