基于神经网络的民用飞机重着陆诊断技术研究  

Study of Aircraft Hard Landing Diagnosis Based on Nerual Network

作　　者：曹海鹏[1] 舒平[2] 黄圣国[1] 

机构地区：[1]南京航空航天大学民航学院,江苏南京210016 [2]中国民用航空总局航空安全技术中心,北京100028

出　　处：《计算机测量与控制》2008年第7期906-908,共3页Computer Measurement &Control

基　　金：国家自然科学基金资助项目(79870032);航空科学基金资助项目(98I52091)

摘　　要：针对飞行品质监控工作中发现的国内航空公司对于重着陆的判断方法存在很多不足,一线飞行员和机务人员对此满意度不高,研究从造成重着陆的相关因素入手,利用QAR记录的多个飞行参数的信息,采用人工神经网络建立重着陆的诊断模型;以航空公司的B737机型的实际数据为样本对模型进行训练和验证,结果显示基于神经网络的模型能够准确判断重着陆事件,为重着陆的诊断提供了一条行之有效的途径,具有较强的工程实用价值和通用性。In the flight operational quality assurance of airways, the hard landing diagosis method is insufficient and not satisfyed by pilots and aircraft crew. Related factors are studyed in the research, the information of flight data recorded by QAR is used, the diagnosis model is established based on nerual network.. The model is trained and certified by the data sample from the B737, the results show that the model based on nerual network could distinguish the hard langing effectively, which is feasible and intelligent, The new method has important engineering application value.

关 键 词：重着陆 飞行品质监控 诊断模型 BP神经网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]