一种机动目标位置自适应估计模型  

An adaptive position estimation for maneuvering target tracking

作　　者：许彦伟[1] 张春华[1] 蔡惠智 王永丰[1] 刘云涛 

机构地区：[1]中国科学院声学研究所,北京100190 [2]北京中科海讯数字信号处理技术有限公司,北京100095

出　　处：《声学技术》2012年第5期459-462,共4页Technical Acoustics

摘　　要：目标运动模型是机动目标跟踪的基本要素之一。现有的机动目标运动模型有的无法自适应目标运动模式,有的模型复杂,计算量较大。根据牛顿运动定律、声纳目标加速度短时间内一般不会剧烈变化的特性和高阶马尔可夫过程,推导出了适应于加速度渐变、不变等多种机动形式的声纳机动目标位置自适应估计模型。该模型采用单一的状态变量,不存在状态耦合,因此模型简单,计算实时性强。通过对不同运动模式的目标数据进行仿真估计,证明了此估计模型的正确性和自适应性。Maneuvering target motion model is one of the basics for maneuvering target tracking.Some of the current models are not adaptive,and some are very complicated.An adaptive position estimation model for gradual change acceleration or constant acceleration sonar targets tracking is proposed in this paper,which is based on the Newton’s Laws of Motion,the acceleration of sonar objects usually cannot be changed acutely and high order Markov process.The model has only one state variable,and so the time consumption for operation is less than that of conventional Models.With processing of various motion targets data,it is shown that the position estimation model is good and adaptive.

关 键 词：马尔可夫过程 运动模型 自适应估计 

分 类 号：TN957[电子电信—信号与信息处理]