东南大学学报（英文版）2019，Vol.35Issue(2)：213-219,7.DOI:10.3969/j.issn.1003-7985.2019.02.011
混合纯电动汽车与传统汽油车的网络交通流演化
Network traffic flow evolution with battery electric vehicles and conventional gasoline vehicles
摘要
Abstract
In order to investigate the effect of the use of battery electric vehicles on traffic dynamics,the valid paths of electric battery vehicles are defined and a check-based method is proposed to obtain them.Then,assuming that travelers only focus on their past travel experience,a day-to-day traffic assignment model is established based on reinforcement learning and bounded rationality.In the proposed model,the Bush-Mosteller model,a reinforcement learning model,is modified to calculate path choice probability according to bounded rationality.The modified model updates the path choice probability only if the gap between expected travel time and perceived travel time is beyond the cognitive threshold.Numerical experiments validate the effectiveness of the model and show that traffic flows can converge to the equilibrium in any case of cognitive thresholds and penetration rates of battery electric vehicles.The cognitive threshold has a positive influence on the variation of traffic flows while it has a negative influence on the differences between traffic flows.The adaptation of battery electric vehicles leads to the poor performance of the traffic system.关键词
纯电动汽车/约束路径/强化学习/有限理性/交通流动态Key words
battery electric vehicles/constrained path/reinforcement learning/bounded rationality/traffic dynamics分类
交通工程引用本文复制引用
李嫚嫚,陆建,孙加辉,涂强..混合纯电动汽车与传统汽油车的网络交通流演化[J].东南大学学报（英文版）,2019,35(2):213-219,7.基金项目
The National Natural Science Foundation of China (No.51478110),Postgraduate Research & Practice Innovation Program of Jiangsu Province (No.KYCX18_0139). （No.51478110）
