基于灰色关联分析的雷达外测数据精度评估方法  

Accuracy Evaluation of Radar Measurement Data by Grey Correlation Analysis

作　　者：蔡铤彬 冯书兴 张雪芹 赵越 代佳玲 CAI Ting-bin;FENG Shu-xing;ZHANG Xue-qin;ZHAO Yue;DAI Jia-ling(No.63766 Unit of PLA,Xishuangbanna 666200,China;Space Safety Research Centre,University of Space Engineering,Beijing 101416,China;Strategic Support Force Staff,Beijing 100120,China)

机构地区：[1]63766部队,西双版纳666200 [2]航天工程大学太空安全研究中心,北京101416 [3]战略支援部队参谋部,北京100120

出　　处：《科学技术与工程》2021年第10期4272-4277,共6页Science Technology and Engineering

摘　　要：测量精度是评估雷达是否满足试验要求的关键指标。首先针对少量的雷达外测定轨评估数据,结合灰色系统理论提出一种基于改进灰色关联分析理论的方法,研究雷达外测数据与精轨数据的关联性;而后根据半监督学习方法,对灰色关联度进行聚合聚类,构建可靠的精度评估方法;最后将有标记的外测数据结合无标记的"天平"一号卫星测量数据进行评估。结果表明,该方法符合任务需求,可为快速评估雷达测量数据的精度提供参考。The measurement accuracy is the key index to evaluate whether the radar meets the test requirements. Aiming at a small amount of radar orbit evaluation data, a method based on improved grey correlation analysis theory was proposed by which the correlation between radar measurement data and accurate orbit data was studied. In addition, according to the semi-supervised learning method, using agglomerative clustering based on grey correlation degree, a reliable evaluation method was constructed, and the measurement data with labels and the measurement data of TP-1 satellite without labels were evaluated. In is concluded that the method met the requirements of the task, which provided a method of quick accuracy evaluation for radar measurement data.

关 键 词：精度评估 灰色关联分析 “天平”一号卫星 聚合聚类 

分 类 号：V556.6[航空宇航科学与技术—人机与环境工程]