摘要
航班延误预测对提高航空公司经济效益和旅客满意度具有重要意义。本研究提出了一种融合时空特征的双向注意力长短时记忆网络(Bi-ATT-LSTM)模型,旨在提升航班延误预测的准确性。该模型能够有效捕捉时间序列的动态特性及其空间依赖性。通过与随机森林模型和标准LSTM模型的对比实验,结果表明Bi-ATT-LSTM模型在复杂的时空数据背景下和多个数据集上显示出优越的性能。
Flight delay prediction is crucial for enhancing the economic benefits of airlines and passenger satisfaction.This study introduces a Bidirectional Attention Long Short-Term Memory Network(Bi-ATT-LSTM)model that incorporates spatio-temporal features to improve the accuracy of flight delay predictions.The model effectively captures the dynamic characteristics of time series and their spatial dependencies.Comparative experiments with the Random Forest model and the standard LSTM model demonstrate that the Bi-ATT-LSTM model exhibits superior performance in complex spatio-temporal data contexts and across multiple datasets.
作者
罗凤娥
郭玲玉
朱子垚
李玫
LUO Feng-e;GUO Ling-yu;ZHU Zi-yao;LI Mei(Civil Aviation Flight University of China,Deyang 618000,China)
出处
《航空计算技术》
2025年第1期17-21,27,共6页
Aeronautical Computing Technique
基金
民航教育类基金“面向共建‘一带一路’国家民航专业本科教育”项目资助(0252108)。
