一种新的机动目标跟踪的多模型算法  

A New IMM Method for Tracking Maneuvering Target

作　　者：范小军[1] 刘锋[1] 

机构地区：[1]北京航空航天大学电子信息工程学院,北京100083

出　　处：《电子与信息学报》2007年第3期532-535,共4页Journal of Electronics & Information Technology

基　　金：航天支撑基金;航空科学基金(03F15002)资助课题

摘　　要：采用带渐消因子的“当前”统计模型与匀速运动模型进行交互,设计了一种新的机动目标跟踪的交互式多模型算法。“当前”统计模型具有对一般机动目标跟踪精度高的特点,通过渐消因子的引入增强了该模型对突发机动的自适应跟踪能力,同时通过与CV模型的交互保证了对非机动目标的跟踪性能。仿真结果表明,在跟踪一般机动目标时,其误差和“当前”统计模型与CV模型交互的IMM算法相当;在跟踪突发机动目标时,该文算法的误差明显小于“当前”统计模型与CV交互的IMM算法。A new IMM algorithm for tracking maneuvering target is proposed, which uses a CS（Current Statistical） model with a fading factor and a CV（Constant Velocity） model for interaction. The CS model itself has a high precision when tracking common maneuvering target, its ability of tracking a sudden maneuvering target has been improved by introducing a fading factor of Strong Tracking Filter （STF）, and the high precision to track non maneuvering targets can be ensured by interacting with the CV model. Simulation results show that, when there is only common maneuver the performance of the algorithm is the same with the IMM-CS/CV, when there is a sudden maneuver, the performance of this algorithm is much better than the IMM-CS/CV.

关 键 词：机动目标跟踪 交互式多模型 卡尔曼滤波 强跟踪滤波器 “当前”统计模型 

分 类 号：TN953[电子电信—信号与信息处理]