面向机场时段差异的航班延误免疫预测算法  

Immune prediction algorithm of flight delay for hub airport in different periods

作　　者：丁建立[1] 王曼[1] 曹卫东[1] 胡海生[2] 黄威[2] 

机构地区：[1]中国民航大学计算机科学与技术学院,天津300300 [2]中国民航信息网络股份有限公司,北京100010

出　　处：《计算机工程与设计》2015年第4期1037-1041,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(60879015);中国民航科技重大专项基金项目(MHRD201241);中央高校发展基金项目(3122014P004)

摘　　要：为提高枢纽机场航班延误预测的准确度,提出面向机场繁忙时间段差异的航班延误免疫预测算法。基于生物免疫机制,结合机场实时运行情况,对之前提出的免疫否定选择算法进行改进。根据机场在一天内不同时间段繁忙程度的差异性自适应地调整算法中检测器的参数,充分利用机场在各个时间段的不同繁忙程度对延误产生的影响,得到更加准确合理的预测结果。仿真结果表明,与原方法相比,采用改进后方法预测的结果误差率减小,自适应性更好,与实际情况的拟合度更高。To improve the accuracy of prediction for flight delays of hub airport,a immune prediction algorithm of airport flight delays based on difference at different busy time was proposed.Based on biological immune mechanism and combining the realtime operation of the airport,the immune negative selection algorithm that previously proposed was improved.According to difference of airport busy degree in different times of a day,the parameters of the algorithm of the detector were adjusted adaptively,and the different effects on delay which resulted from the busy degree of airport in different times of a day were mode full use of,which made the prediction results more accurate and reasonable.The simulation results show that compared with the original method,the error rate using the improved method was reduced,and it has better adaptability.

关 键 词：枢纽机场 生物免疫机制 否定选择算法 检测器参数 航班延误预测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]