Adaptive interacting multiple model particle filter based on a dual-pattern bat algorithm for maneuvering target tracking
Mengchu Tian,Zhimin Chen,Yubin Fang
DOI: https://doi.org/10.1177/09544100241286129
2024-10-15
Proceedings of the Institution of Mechanical Engineers Part G Journal of Aerospace Engineering
Abstract:Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, Ahead of Print. The interacting multiple model particle filter (IMM-PF) is a filtering method commonly used for nonlinear non-Gaussian radar systems. Its transfer probabilities are usually fixed and dependent on prior information. When tracking maneuvering targets, the IMM-PF may cause excessive tracking errors due to the model switching lag. In addition, the particle impoverishment caused by the resampling of the particle filter seriously affects the filtering accuracy. Therefore, the IMM-PF has difficulty meeting the accuracy and speed requirements of modern high-performance radar target tracking systems. To address these problems, an adaptive interacting multiple model particle filter based on the dual-pattern bat algorithm (ADIMM-DPBA-PF) is proposed for tracking maneuvering targets under nonlinear and non-Gaussian conditions. First, the adaptive model switching mechanism is established to adjust the transition probabilities using the model probability posterior information at consecutive times, improving the model switching efficiency. Second, a particle filter based on the dual-pattern bat algorithm (DPBA-PF) is proposed. The filter exploits global search and local search strategies to optimize the particles and intelligently move the particles to the high likelihood region. Finally, a particle filter based on the dual-pattern bat algorithm is used to form the adaptive interacting multiple model method. The experimental results show that the proposed algorithm has better comprehensive performance than the IMM-PF.
engineering, mechanical, aerospace