首页 | 本学科首页   官方微博 | 高级检索  
     检索      

一种跟踪机动目标的新方法
引用本文:张伯彦,蔡庆宇,袁曾任.一种跟踪机动目标的新方法[J].系统工程与电子技术,1990(1).
作者姓名:张伯彦  蔡庆宇  袁曾任
作者单位:航空航天工业部二院二十三所 (张伯彦,蔡庆宇),清华大学计算机系(袁曾任)
摘    要:本文应用控制理论分析了R.A.Singer零均值模型在跟踪机动目标方面的局限性;提出了一种目标加速度的截断正态概率密度模型,并将目标加速度模拟为非零均值、时间相关的随机过程;借助于切比雪夫(CHEBYSHEV) 不等式来确定目标加速度的均值和方差之间的关系,并引用卡尔曼滤波器所包含的目标机动的统计信息来实现方差自适应算法。 理论分析表明,本文的模型和算法跟踪恒加速目标时稳态偏差为零,从而消除了Singer方法在跟踪恒加速目标时所存在的稳态偏差。 计算机仿真结果证实了理论分析的结论,表明本文提出的模型和采用的算法能够较好地跟踪高度机动目标。

关 键 词:目标跟踪  活动目标  卡尔曼滤波  自适应控制  正态密度函数  模拟  估计理论  算法

A NEW METHOD FOR TRACKING MANEUVERING TARGETS
Zhang Boyan and Cai QingyuBeijing Institute of Radio MeasuremeniYuan Zengren.A NEW METHOD FOR TRACKING MANEUVERING TARGETS[J].System Engineering and Electronics,1990(1).
Authors:Zhang Boyan and Cai QingyuBeijing Institute of Radio MeasuremeniYuan Zengren
Institution:Zhang Boyan and Cai QingyuBeijing Institute of Radio MeasuremeniYuan ZengrenDepartment of Computer Science and Fechnology Tsinghua University
Abstract:A new method for the three-dimensional airborne maneuvering target trucking problem is presented. In this method the unknown target acceleration is considered as a time-correlated random process with nor-zero mean values and the density function of the acceleration is assumed to be truncated normal. The re-(?) between the random acceleration's mean value and its variance is determined by Chebyehev inequality. With this method, the steady-state values of dynamic errors of Singer's model and algorithm in response to a unit-step acceleration input are eliminated; the improved tracking accuracy is achieved with a normal acceleration maneuvering motion input under the measurement matrix to be nonlinear functions of the state variables.
Keywords:Target tracking  Moving target  Kalman filtering  Adaptive control  Normal density function  Simulation  Estimation theory  Algorithm    
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号