摘要
本文研究了无杂波和漏检的情况下三种不同类型、不同位置的传感器对数目未知的目标进行检测时的静态数据关联问题.这一问题可以通过对测量划分的联合似然函数的极大化来解决,通常可将其转化为三维匹配问题,但其求解的复杂度是NP的.本文提出了一种基于遗传算法的优化算法,来解决三维匹配问题,实验结果表明这种算法具有很高的关联成功率.
This paper deals with the static problem of associating measurements from three spatially distributed heterogeneous sensor without clutter and missed detections. This problem can be solved by maximizing the joint likelihood function of be measurement partion. Usu-ally, the data association problem may be led to a generalization of the three dimensional as-signment problem, which is known to be NP-hard. An optimazation algorithm based on ge-netic algorithm for solving 3D assignment problem is presented in the paper. Finally, higher correct association percent is illustrated by the experiment results.
出处
《系统工程与电子技术》
EI
CSCD
1997年第8期13-16,33,共5页
Systems Engineering and Electronics
关键词
多传感器
多目标跟踪
数据相关
匹配
算法
Multisensor, Multitarget, Static data association, Measurement partion, 3D assignment,Genetic algorithm.
