人工神经网络在航迹关联中的应用  

Application of neural network in track-to-track correlation problem

作　　者：张池平[1] 崔平远[2] 张英俊[1] 

机构地区：[1]哈尔滨工业大学理学院,黑龙江哈尔滨150001 [2]哈尔滨工业大学深空探测基础研究中心,黑龙江哈尔滨150001

出　　处：《黑龙江大学自然科学学报》2006年第1期38-41,共4页Journal of Natural Science of Heilongjiang University

基　　金：国家863项目(2002AA731215)

摘　　要：针对在密集目标环境下和运动出现交叉或分岔等难于处理的情况,将导致较多的错、漏相关航迹的问题,提出一种利用神经网络进行航迹关联的方法。该方法有效地应用了神经网络的泛化能力以及自组织自适应的学习功能,且通过对网络输入结点的设计,能够很好地解决复杂航迹关联问题。仿真实验验证了该算法在多传感器系统的航迹关联问题中的有效性。An algorithm of track - to - track correlation based on neural network is presented to these questions of inaccurate and lost correlation resulted by intersectant or furcate conditions in complex and multi goals＇ environment. This algorithm solves the problem of complex correlation by the node＇ s design of neural network＇ s input based on application of extensive ability effectively and effective study ability of itself. The simulation shows the efficiency of the way in track - to - track correlation problem of muti - sensor＇ s system.

关 键 词：多传感器 航迹关联 神经网络 

分 类 号：TN95[电子电信—信号与信息处理]