基于Skip-LSTM的机场群延误预测模型

Airport Group Delay Prediction Based on Skip-LSTM

针对目前机场群发展不平衡,国际枢纽机场的延误率居高不下,航班时刻短缺,资源紧张,而区域枢纽机场却存在资源空闲的问题,提出一种基于跳过门的长短时记忆网络(Skip-LSTM,Skip Long Short Term Memory)的机场群延误预测模型。该模型首先将机场群中各个机场的信息,机场群航班信息以及机场群地区的气象信息进行融合及处理,然后搭建Skip-LSTM网络对融合后的数据信息进行特征提取,最后利用Softmax分类器对机场群的延误状况进行分类预测。Skip-LSTM网络在传统的长短时记忆网络(LSTM,Long Short Term Memory)的基础上增加了Skip门,能更加充分地提取机场群数据信息的时间相关性,获得更高的准确率。实验结果表明,基于Skip-LSTM的机场群延误预测模型的准确率可达95.35%,预测性能优于传统的网络模型,能对机场群的延误状况进行有效的预测。

In view of the current imbalance in the development of the airport group,the international hub airport has a high delay rate,the flight time is short,the resources are tight,but the regional hub airport has the problem of idle resources.An airport group delay prediction model based on Skip-LSTM is proposed.Firstly the information of each airport in the airport group,the flight information and the weather information of the airport group area are integrated and processed in the model,then the feature information of the merged data is extracted in the Skip-LSTM network model.Finally,the Softmax classifier is used to classify and predict.The Skip gate is added to the Skip-LSTM based on the traditional LSTM,which can more fully extract the time correlation of data information.The higher accuracy is obtained in the model.The experimental results show that the accuracy of the airport group delay prediction model based on Skip-LSTM can reach 95.35%,and the prediction performance is better than the traditional network model,which can effectively predict the delay of the airport group.