基于机器学习的FDR MoC6试验数据分析方法  

FDR MoC6 experimental data analysis method based on machine learning

作　　者：严淑芹 王伟[1] YAN Shuqin;WANG Wei(Shanghai Aircraft Design and Research Institute,Shanghai 201210,China)

机构地区：[1]上海飞机设计研究院,上海201210

出　　处：《民用飞机设计与研究》2023年第4期14-21,共8页Civil Aircraft Design & Research

摘　　要：MoC6适航符合性验证试飞是民用飞机取证过程重要的必需条件。飞行数据记录仪(FDR)的MoC6试验有一项内容是:飞机试飞结束后,下载飞行数据记录器(FDR)记录的数据,并与飞行试验测试设备(FTI)采集的数据进行比对,以判断FDR记录数据是否准确。将其他型号飞机的FDR科目MoC6试验的分析结果用机器学习方法训练得出闵可夫斯基距离参数p值(1.8)、阈值(2.05),然后计算本型号飞机的FDR数据和FTI数据的闵可夫斯基距离、判断该距离是否超过阈值,以判断FDR记录的参数值变化与FTI的记录是否一致,即判断FDR记录数据是否准确。实现将FDR与FTI记录不一致的参数都找出来,即实现了召回率100%。MoC6 airworthiness compliance verification flight test is an important and necessary condition for civil aircraft certification.Flight Data Recorder is an on-board equipment and one of the contents of its MoC6 test is to download the data recorded by flight data recorder(FDR)after the flight test,and compare it with the data recorded by flight test instrument(FTI)to determine whether the recorded data in FDR is accurate.This article in how to tell whether the data recorded in FDR is correct by means of Minkowski distance.By the way,the p parameter and the threshold value for Minkowski distance is acquired via machine learning.In this paper,the inconsistent parameters recorded by FDR and FTI were identified,achieving a recall rate of 100%.

关 键 词：飞行数据记录仪(FDR) MoC6 机器学习 闵可夫斯基距离 

分 类 号：TN915[电子电信—通信与信息系统]