密集杂波环境下一种新的机动目标跟踪算法  

A new algorithm for tracking maneuvering target in dense clutter

作　　者：肖胜[1] 邢昌风[1] 隋江波[1] 

机构地区：[1]海军工程大学电子工程学院,武汉430033

出　　处：《海军工程大学学报》2010年第3期46-49,83,共5页Journal of Naval University of Engineering

基　　金：湖北省自然科学基金资助项目(2006ABA010);海军工程大学自然科学研究基金资助项目(HGDJJ08015)

摘　　要：针对密集杂波环境下的机动目标跟踪问题,在传统方法基础上结合模糊推理的思想,提出一种模糊自适应交互多模型概率数据关联算法。该算法将目标运动模式分为机动和非机动两类区别对待,无需选取大量模型来覆盖机动目标的运动模式,而是动态调节模型集合中部分模型的参数,具有更强的自适应跟踪能力。最后,给出了算法的仿真分析,结果表明:该方法能够有效地跟踪密集杂波中的机动目标,并且跟踪性能较传统方法有所提高。In order to track a maneuvering target in dense clutters,a fuzzy adaptive interactive multiple model probabilistic data association（FAIMM-PDA） algorithm was proposed based on the conventional methods and fuzzy reasoning.The algorithm divided the mode of movement into maneuvering and no-maneuvering so that there was no need to select a lot of models.Some parameters in the model-set were dynamically adjusted so that the algorithm was more adaptive than IMM-PDA.Then,a Monte Carlo simulation was used to analyze the method.The results show that the algorithm can effectively estimate a maneuvering target in dense clutters and is better than the traditional algorithm in the performance of tracking.

关 键 词：机动目标跟踪 交互式多模型 模糊推理 概率数据关联 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]