摘要
针对传统特征提取方法难以从海量、高维的快速存取记录器(QAR)数据中提取有效特征,且QAR数据缺乏足够的标记等问题,提出一种以Transformer网络为核心的QAR2Vec模型,将QAR数据与位置信息、飞行阶段信息共同编码,作为QAR2Vec模型的输入;通过构建自回归预测的预训练任务以自监督的方式来学习海量QAR数据中的深层特征;保存预训练好的QAR2Vec模型权重,并在飞行状态预测和着陆异常天气识别任务上,微调预训练模型,使模型适应不同的下游任务;将QAR2Vec模型与2种没有预训练步骤的深度学习算法CNN-LSTM、MTL-LSTM进行对比。结果表明:QAR2Vec模型能够更有效地从QAR数据中提取特征,在飞行状态预测和着陆异常天气识别任务上的预测误差更低、识别准确度更高。
In order to address difficulties that traditional extraction methods have in extracting effective features from massive and high-dimensional QAR data which lack sufficient labeled data,a QAR2Vec model was proposed with Transformer as its core.Firstly,QAR data were co-encoded with location and flight phase information as input of the model.Secondly,pre-training task of autoregressive prediction was constructed to learn deep features of massive QAR data in a way of self-supervision.Finally,scale of pretrained QAR2Vec model was recorded and slightly adjusted on flight state prediction and landing abnormal weather recognition tasks,and performance of QAR2Vec model was evaluated by comparing it with two deep learning algorithms—CNN-LSTM and MTL-LSTM without pre-training steps.The results show that QAR2Vec can extract features from QAR data more effectively,with lower errors and higher accuracy on flight status prediction and landing abnormal weather recognition tasks.
作者
段照斌
杜海龙
张鹏
DUAN Zhaobin;DU Hailong;ZHANG Peng(Engineering Techniques Training Center,Civil Aviation University of China,Tianjin 300300,China;College of Airworthiness,Civil Aviation University of China,Tianjin 300300,China)
出处
《中国安全科学学报》
CAS
CSCD
北大核心
2021年第1期145-152,共8页
China Safety Science Journal
基金
国家自然科学基金青年基金资助(61703406)
天津市教学成果奖重点培育项目(PYGJ-006)。
