摘要
为更精确地预测航班过站时间,将全国机场按照规模差异及不同地理位置所导致的客流量差异和天气差异对航班过站时间造成的不同影响进行分类,基于各类机场航班数据,构建混合轻量级梯度提升机算法(LightGBM)模型对航班过站时间分类预测。引入自适应鲁棒损失函数(adaptive robust loss function,ARLF)改进LightGBM模型损失函数,降低航班数据中存在离群值的影响;通过改进的麻雀搜索算法对改进后的LightGBM模型进行参数寻优,形成混合LightGBM模型。采用全国2019年全年航班数据进行验证,实验结果验证了方法的可行性。
To predict the flight turnaround time more accurately,airports across the country were classified according to the differences in passenger flow caused by scale differences and the different impacts of weather differences caused by different geographical locations on flight turnaround time.Based on various airport flight data,a hybrid LightGBM model was constructed to classify and predict flight turnaround time.The LightGBM model loss function was improved by introducing adaptive robust loss function(ARLF)to reduce the impact of outliers in flight data.An improved sparrow search algorithm was used to optimize the parameters of the improved LightGBM model,a hybrid LightGBM model was constructed.The feasibility of the method is verified using national flight data for the entire year of 2019,and the experimental results demonstrate its feasibility.
作者
李国
王伟倩
曹卫东
LI Guo;WANG Wei-qian;CAO Wei-dong(College of Computer Science and Technology,Civil Aviation University of China,Tianjin 300300,China)
出处
《计算机工程与设计》
北大核心
2025年第2期633-640,F0003,共9页
Computer Engineering and Design
基金
国家自然科学基金重点基金项目(U2033205、U2233214)。
