高速公路应急救援中心选址优化模型  

Research on optimization model of expressway emergency rescue center location

作　　者：胡立伟[1] 何越人 佘天毅 孟玲 杨锦青 HU Liwei;HE Yueren;SHE Tianyi;MENG Ling;YANG Jinqing(Faculty of Transportation Engineering,Kunming University of Science and Technology,Kunming Yunnan 650500,China;Zhenjiang Transportation Administration Office,Zhenjiang Jiangsu 212000,China)

机构地区：[1]昆明理工大学交通工程学院,云南昆明650500 [2]镇江市运输管理处,江苏镇江212000

出　　处：《中国安全科学学报》2019年第5期145-150,共6页China Safety Science Journal

基　　金：国家自然科学基金资助(61863019);国家工程实验室开放研究基金资助(NELJA201605)

摘　　要：为完善高速公路交通事故应急救援体系,以昆石高速为研究对象,将遗传算法(GA)与反向传播神经网络模型(BP)相结合,运用GA-BP神经网络模型鉴定高速公路事故多发段,根据昆石高速环境条件选定应急救援中心候选点;应用双层规划理论模型分析昆石高速应急救援中心选址,并利用萤火虫算法优化求解,得到昆石高速公路应急救援中心选址的最优方案。研究结果表明:GA-BP神经网络模型鉴别高速公路事故多发段相较于传统模型更准确;运用双层规划模型和萤火虫算法求解可得到昆石高速应急救援中心最佳选址方案。In order to improve the emergency rescue system for expressway traffic accidents,Kunshi expressway was taken as an example,GA was combined with BP neural network model,and GA-BP neural network model was used to identify the highway accidents prone location. According to the Kunshi expressway environmental conditions,the candidate points for the emergency rescue base were selected. The bi - level programming theory model was applied to analyze the location of the Kunshi expressway emergency rescue base,and the firefly algorithm was used to optimize the solution. Then the optimal layout plan for the location of the emergency rescue base of Kunshi expressway was obtained. The results show that the GA-BP neural network model is more accurate in identifying the faulty section of expressway than the traditional model,and that the bi-level planning model and the firefly algorithm can be used to identify the best location of the Kunshi expressway emergency rescue center.

关 键 词：高速公路 遗传算法-反向传播神经网络(GA-BP) 交通事故 应急救援 选址模型 

分 类 号：X928.03[环境科学与工程—安全科学]