基于飞行QAR数据与模拟机实验的民机着陆异常事件形成机制研究

项目名称： 基于飞行QAR数据与模拟机实验的民机着陆异常事件形成机制研究

项目编号： No.61304207

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 汪磊

作者单位： 中国民航大学

项目金额： 23万元

中文摘要： 远接地和重着陆是飞行过程中高频次、高风险的两类着陆异常事件，是航空公司安全管理监控的重点，快速存取记录器（Quick Access Recorder，QAR）可以记录这些超限事件的所有飞行及操作数据。本项目旨在通过QAR数据、模拟机实验、飞行员问卷综合研究远接地和重着陆这两类着陆异常事件的外在特征及内在形成机制。首先应用数据挖掘和统计建模的方法从QAR海量数据中提取远接地和重着陆这两类着陆异常事件的飞行参数特征，再进一步通过模糊神经网络建模获取着陆异常事件的操作特征；其次通过飞行模拟机实验探索着陆操作中的高度判断机制问题；最后通过设计飞行员问卷研究影响高度判断及着陆异常操作的关键因素。最终获得着陆异常事件的发生规律、内在机理和深层次原因，项目成果预期为着陆异常事件的控制、着陆近地事故的预防提供参考依据，对减少着陆阶段不安全事件和事故、提高飞行安全水平有着积极的现实意义。

中文关键词： 着陆安全；飞行数据；人因；事故；事件

英文摘要： Long landing and hard landing are two kinds of abnormal landing events with high frequency and risk, which are also the monitoring focus by air carriers. Quick Access Recorder (QAR) could record all kinds of flight and operation parameters of abnormal and exceedence events. The aim of this project is to research the formation mechanism of these two abnormal landings based on QAR data, experiment in flight simulator and pilot questionnaires. Firstly the method of data mining and statistical modeling is used to extract flight parameter features from massive QAR data. Secondly fuzzy neural network modeling is applied to obtain operation characteristics of two abnormal landing events. Thirdly the flight simulator experiment is designed and implemented to explore the height cognition mechanism in visual landing phase. Fourthly the pilot survey is carried on to find out key factors causing abnormal landing operations. Finally the regular operation patterns and deep level reasons leading to abnormal landing are found out and the research conclusions are expected to be applied into practice for reducing unsafe operation and events in landing. The products of this projects would be helpful and meaningful for improving the whole flight safety.

英文关键词： landing safety；flight data；human factors；accident；incident