基于期望系统噪声模型的自适应机动目标跟踪  

An Adaptive Method of Tracking Multiple Maneuvering Targets Based on Estimation of System Noise Model

作　　者：殷业飞[1] 黄树彩[1] 

机构地区：[1]导弹学院空军工程大学,陕西三原713800

出　　处：《航空计算技术》2009年第6期6-9,共4页Aeronautical Computing Technique

基　　金：国家高技术研究发展计划(863计划)(***046)

摘　　要：针对防空作战中的机动目标跟踪问题,分析了使用IMM算法跟踪目标所存在的缺陷,提出了一种基于期望系统噪声模型的自适应机动目标跟踪算法(MIMM),改变IMM算法使用固定模型集合的方法,IMM算法中作用权重较大的只是少部分更接近于实际系统的模型,通过对部分系统噪声模型进行自适应辨识,计算出最接近于系统实际噪声水平的模型——期望系统噪声模型(ES-NM)。这些模型作为IMM多模型集的子集,是在一定模型框架内所寻求出的一个(或多个)最优模型(集)。仿真结果证明了MIMM算法能够更好地描述目标机动,达到更理想的跟踪性能,其跟踪精度优于使用固定模型集的IMM算法。For the problem of tracking maneuvering targets in background of air- defense and anti- ballistic missile, the defects of the IMM method in tracking maneuvering targets is analyzed, then a new algorithm named MIMM which is based on the estimation of system noise model is presented. In the MIMM method,the fixed model set is not about to be used again,instead, some models with greater affecting index are selected from the IMM model set and form a new submodel set in which models are more close to the real system model. By making adaptive recognition for some system noise models, the one （model set） which is （are） most close to the real system noise model is （are） calculated, that is estimation system noise model （ESNM）. As the submodel set of the IMM model set, they are optimized as the best one. The simulation results proved that the MIMM algorithm could better describe the motion of maneuvering target and acquire ideal tracking performance, and its tracking precision is better than the IMM algorithm using fixed model set.

关 键 词：IMM 机动目标跟踪 ESNM EMA 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]