基于四诊法的管制席安全运行状态诊断  

Diagnosis of safety operation state for ATC position based on four diagnostic methods

作　　者：岳仁田[1] 韩蒙 YUE Rentian;HAN Meng(College of Air Traffic Management,Civil Aviation University of China,Tianjin 300300,China)

机构地区：[1]中国民航大学空中交通管理学院,天津300300

出　　处：《中国安全科学学报》2021年第9期52-59,共8页China Safety Science Journal

基　　金：国家自然科学基金资助(U1533112)。

摘　　要：为保障空中交通安全运行,基于中医学四诊法,从望、闻、问、切4方面构建空中交通管制(ATC)席安全运行状态诊断指标体系,将组合赋权法、集对分析和BP神经网络相结合,提出一种具有自学习能力的组合诊断方法。首先,通过组合赋权法获得集对分析所需权重;其次,通过集对分析得到每个样本中管制席安全运行状态;最后,基于样本完成BP神经网络的训练和测试,得到安全运行状态诊断模型。结果表明:该模型简单易行,诊断结果准确度高,具有定性与定量相结合的优点。In order to ensure safety operation of air traffic,based on four diagnostic methods of traditional Chinese medicine,diagnostic index system of safety operation state for ATC position was constructed from four aspects of observation,auscultation and olfaction,inquiry,pulse feeling and palpation. Combining combination weighting method,set pair analysis and BP neural network,a combination diagnosis method with self-learning ability was proposed. Firstly,weight required by set pair analysis was obtained by combination weighting method. Secondly,operation state of ATC position in each sample was obtained by set pair analysis. Finally,BP neural network was trained and tested based on samples,and operation state diagnosis the model was obtained. The results show that the model is simple and easy to use,and diagnosis results are accurate,and it has advantages of combining qualitative and quantitative.

关 键 词：管制席 四诊法 安全运行状态诊断 集对分析 BP神经网络 组合赋权 

分 类 号：X949[环境科学与工程—安全科学] X951