基于最小模型集切换的变结构IMMPF算法

摘要

In order to enhance the tracing precision, a new variable structure IMMPF tracking algorithm is presented. Thus, variable structure is realized by switching the minimal model set composed of "current" statistic model at different times. When moving model of the target is changed, filtering model in this algorithm can match the real moving model of the target by switching the least model set. Consequently, the response time to the maneuver target is reduced. Compared with general IMM algorithm,simulation proves the superiority of the algorithm presented here.%为提高对机动目标的跟踪精度,提出了一种新的基于最小模型集切换的变结构IMMPF跟踪算法,其中以"当前"统计模型为基础的不同最小模型集在不同时刻之间的切换实现了多模型的结构变换.当目标机动方式发生改变时,通过最小模型集的切换实现算法的滤波模型与目标实际机动方式的快速匹配,减小了目标机动的响应时间.通过仿真实验,与通用的IMM估计进行了比较,证明了算法的优越性.