基于NP准则的属性关联度量及门限确定方法  

Solution of attribute association degree and threshold based on NP rule

作　　者：井沛良[1] 徐世友[1] 李贤[2] 陈曾平[1] 

机构地区：[1]国防科学技术大学自动目标识别重点实验室,湖南长沙410073 [2]中南大学信息科学与工程学院,湖南长沙410083

出　　处：《系统工程与电子技术》2014年第3期429-433,共5页Systems Engineering and Electronics

摘　　要：在目标空域密集环境下,属性辅助运动状态数据关联是目前公认的解决数据关联性能严重下降问题的潜在有效途径。从Neyman-Pearson准则出发,在分析经典运动状态信息关联度量及门限确定方法合理性的基础上,给出属性辅助数据关联中属性度量及对应门限的确定方法。仿真结果表明,与惯用的固定门限相比,该方法确定的度量及门限具有稳定的关联性能。In the closely spaced targets environments, attribute aided kinematic state data association nas currently been universally recognized as the potential method to improve data association performance. First the rationality of classical association degree and corresponding threshold about kinematic state information in the Neyman-Pearson （NP） sense are analyzed. Then, a new method is presented, which allows an NP test for the solution of association degree and threshold about attribute information. Simulation results indicate that the proposed method could have stable association performance compared with the usual method using constant threshold.

关 键 词：数据关联 属性信息 NEYMAN-PEARSON准则 关联度量 关联门限 

分 类 号：TP212.9[自动化与计算机技术—检测技术与自动化装置]