交互式多模型粒子滤波算法及声源跟踪性能分析

Interacting Multiple Model Particle Filter Algorithm and Acoustic Source Tracking Performance Analysis

交互式多模型粒子滤波是一种融合了交互式多模型和粒子滤波的目标跟踪算法。为了便于对交互式多模型粒子滤波算法在声源跟踪的性能分析,首先阐述了交互式多模型粒子滤波算法的原理;然后对整个算法用到的关键参数进行了分析并对改进算法进行了归纳整理;最后对交互式多模型粒子滤波在声源跟踪当中的应用进行了仿真并指出未来的研究方向。

The interacting multiple model particle filter which combined by interacting multiple model and particle filter is applied in tracking maneuvering target. In order to analyze the performances of the interacting mul？tiple model particle filter algorithm in the acoustic source tracking, the principle of the interacting multiple model particle filter algorithm is introduced at first. And then, the key parameters used in the algorithm are analyzed and the improved algorithm is concluded. Finally, the application of interacting multiple model particle filter in acoustic source tracking is simulated and the future research directions are pointed out.