具有固定滞后平滑的变结构机动目标跟踪  

Fixed lag smoothing variable structure multiple model algorithm for maneuvering target tracking

作　　者：毛少锋 冯新喜[1] 文莎[2] 危璋 

机构地区：[1]空军工程大学信息与导航学院,陕西西安710077 [2]南京政治学院上海分院,上海200433

出　　处：《重庆邮电大学学报（自然科学版）》2015年第2期189-196,共8页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

基　　金：陕西省自然科学基金(2011JM8023)~~

摘　　要：针对变结构多模型算法在目标机动性较强时,因模型切换不及时而出现较大的误差峰值问题,提出了一种固定滞后平滑的变结构多模型(lag smoothing variable structure multiple model,LS-VSMM)跟踪算法。建立一个完备的模型集,根据模型概率利用有向图切换规则实现模型子集与目标运动模式的匹配;引入固定滞后平滑算法,通过对状态向量扩维,将平滑问题转化为滤波问题,使得原特定时刻的目标状态在系统中停留的时间更长,增加更多的状态量测信息,让状态估计变得更加准确,并且以延迟一定时间输出来改善滤波性能;对算法进行了仿真实验分析。仿真结果表明,与基于无味有向图切换的多模型算法以及基于有向图切换的变结构多模型(digraph switch variable structure multiple model,DS-VSMM)算法相比,LS-VSMM算法在有效降低误差峰值的同时,提高了目标的跟踪精度。Models can＇t switch timely when the target maneuverability becomes stronger,which leads to larger peak error appearing. A fixed lag smoothing variable structure multiple model algorithm is proposed for tracking a maneuvering target.According to model probability,the algorithm can firstly match between the target motion model and the subset model by establishing a complete set of models. After that,fixed lag smoothing algorithm is developed by a state-augmented system,which makes the smoothing problem transform into the filtering problem. At the same time,original particular moment state remains more long time in system,and increases more state measurement information to make state estimation more accurate,which is at the cost of a time delay to increase the performance. Finally,the algorithm is analyzed by experiment simulation. Simulation results show the proposed algorithm is better in reducing peak error at maneuvering turn of target than unscented digraph switch and the digraph switch variable structure multiple mode algorithm,which improves tracking accuracy more effectively.

关 键 词：固定滞后平滑 有向图切换 变结构多模型 机动目标跟踪 

分 类 号：TN953[电子电信—信号与信息处理]