共享电动汽车选择行为分析

Analysis of the selection behavior of shared electric vehicles

为探索共享电动汽车选择行为,构建出行行为选择模型,基于南京市出行者选择行为调查数据,采用计划行为理论、技术接受理论和结构方程模型,综合考虑出行者个人社会属性、出行特征及潜变量,分别构建不包括潜变量的多项Logit模型和包括潜变量的混合选择模型,并对这些因素进行显著性和敏感性分析.结果表明:不包括潜变量的多项Logit模型中,模型拟合系数(伪R2)为0.275>0.2,表明模型拟合效果较好,年龄、收入、拥有机动车、受教育程度、对共享电动汽车了解程度等因素存在显著影响;在包括潜变量的混合选择模型中,模型拟合系数(伪R2)为0.354>0.275,表明模型拟合效果优于多项Logit模型,对共享电动汽车了解程度、拥有机动车、舒适性、快捷性、行为态度、使用偏好、行为意向、使用障碍、站点障碍、车辆障碍、个人障碍等因素存在显著影响.

To explore the effects of electric vehicle timeshare rentals(EVTR)on choice behavior,and construct the travel choice behavior model,the planned behavior theory,technology acceptance theory,and structural equation model are used based on the data in Nanjing.The multinomial Logit models without latent variables and the mixed choice model with latent variables are constructed based on the socioeconomic,travel attributes and latent attributes of travelers.The significance and sensitivity of these factors are analyzed.The results show that the fitting coefficient(pseudo R2)is 0.275>0.2 in the multinomial Logit models without latent variables,and the effect is good.The factors,such as age,income,ownership of vehicles,education and familiarity with EVTR,have significant influence.The fitting coefficient(pseudo R2)is 0.354>0.275 of the mixed choice model with latent variables,indicating that the accuracy of the mixed choice model is better than that of the multinomial Logit model.The factors,such as familiarity with EVTR,car ownership,comfort attribute,efficient attribute,consciousness,use preference,behavioral intention,use obstacles,station obstacles,vehicle obstacles,and personal obstacles,have a significant influence.