基于ISSA-ELM的终端区进场交通流混沌识别研究  

Chaotic Identification of Incoming Traffic Flow in Terminal Area Based on ISSA-ELM

作　　者：杨新湦[1] 刘馥嘉 张召悦[3] YANG Xinsheng;LIU Fujia;ZHANG Zhaoyue(Civil Aviation University of China,Tianjin 300300,China;Transportation research and Engineering College,Civil Aviation University of China,Tianjin 300300;Air Traffic Management College,Civil Aviation University of China,Tianjin 300300,China)

机构地区：[1]中国民航大学校办,天津300300 [2]中国民航大学交通科学与工程学院,天津300300 [3]中国民航大学空中交通管理学院,天津300300

出　　处：《综合运输》2024年第3期109-113,137,共6页China Transportation Review

基　　金：京津冀机场群“三地四场”协同发展研究(2021JWZD38)。

摘　　要：随着我国航空需求增长,航班延误频发,拥堵现象日益严重。为保障空中交通持续有序,航班正常运行,需在空中交通流混沌产生之初快速准确的识别混沌状态。为此提出了一种改进的空中交通流混沌智能识别模型,该模型基于极限学习机和改进的麻雀搜索算法,结合小波包提取出的交通流特征信息,最终以某终端区为例进行验证,结果表明,当ELM选取sin为激活函数时,较SVM运行时间更短,准确率较高,而所提出的ISSA-ELM混沌识别模型较基础ELM模型准确性提高了约4.29%,验证了改进后模型的有效性。该模型的提出丰富了混沌识别的理论研究,弥补了空中交通流混沌智能识别的空缺。With the growth of aviation demand in China,flight delays are frequent and congestion is becoming increasingly serious.In order to ensure the continuous orderly air traffic and normal flight operation,the chaotic state needs to be identified quickly and accurately at the beginning of air traffic flow chaos generation.An improved intelligent recognition model of air traffic flow chaos is proposed,which is based on the extreme learning machine and improved sparrow search algorithm,and combined with the traffic flow feature information extracted by wavelet packets,and finally validated with a terminal area as an example.ELM chaos recognition model improves the accuracy by about 4.29%compared with the ELM model,which verifies the effectiveness of the model.The proposed model enriches the theoretical study of chaos recognition and fills the gap of chaotic intelligent recognition of air traffic flow.

关 键 词：空中交通流 机场终端区 混沌识别 极限学习机 

分 类 号：U491[交通运输工程—交通运输规划与管理]