Premium
A Multiple Model Tracking Algorithm Based on an Adaptive Particle Filter
Author(s) -
Chen Zhimin,
Qu Yuanxin,
Xi Zhengdong,
Bo Yuming,
Liu Bing,
Kang Deyong
Publication year - 2016
Publication title -
asian journal of control
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.769
H-Index - 53
eISSN - 1934-6093
pISSN - 1561-8625
DOI - 10.1002/asjc.1275
Subject(s) - particle filter , tracking (education) , resampling , algorithm , particle swarm optimization , inertia , filter (signal processing) , computer science , control theory (sociology) , kernel adaptive filter , radar tracker , radar , mathematical optimization , artificial intelligence , mathematics , filter design , computer vision , psychology , telecommunications , pedagogy , physics , control (management) , classical mechanics
The interacting multiple model based on a particle filter fails to meet the requirements of real‐time performance when manoeuvring target tracking by radar due to deficiencies in its high calculation complexity. An improved particle filter based on landscape adaptive particle swarm optimization is proposed. This filter adopts the method of updating inertia weight, using not only local information and global information, but also preventing algorithm trapping in a local optimum, so the filter can find the optimal solution with less iteration. Additionally, an improved tracking model is presented. With the help of systematic resampling, the model can figure out the model index of particles. The experimental results prove that the new tracking algorithm not only improves manoeuvring target tracking accuracy, but also decreases computing complexity.