摘要
从行为分析的角度出发研究电动汽车用户的充电决策机制,是合理预测充电需求、优化充电设施布局的重要基础。利用混合选择模型框架引入态度潜变量,在选择模型部分构建兼顾类间和类内偏好异质性的潜类混合logit模型,以实现用户类别的划分,并对充电需求的产生机理展开分析;依托电动汽车用户充电选择的显示偏好(RP)和陈述偏好(SP)调查数据实现对模型的测试与标定,分析不同类别的用户特征以及影响用户充电选择行为的因素。根据标定结果将电动汽车用户划分为“风险偏好型”和“风险规避型”两类,并给出关于电动汽车公共充电桩规划管理的基础性建议。此外,本文构建传统潜类混合logit模型作为对比,验证了态度潜变量的加入能够有效提升模型的拟合优度和解释能力。
Research on charging decision-making mechanism of EV users from the perspective of behavior analysis is an important basis for reasonably predicting charging demand and optimizing the layout of charging facilities.Hybrid choice model framework was used to introduce the latent variable of attitude,and a latent class mixed-logit model was constructed in the choice model,which considered the heterogeneity of the inter-class and intra-class,to realize the classification of user categories and analyzed the generation mechanism of charging demand.The model was tested and calibrated based on the survey data of revealed preference and stated preference of EV users’charging choice,the characteristics of different types of users and the factors influencing users’charging choice behavior were analyzed.According to the results,EV users are divided into“risk preference”and“risk aversion”,and suggestions on planning of public charging piles are proposed.Moreover,a traditional latent class mixed-logit model is constructed as a comparison and verified that the latent variables of attitude can effectively improve the goodness of fit and interpretation ability of the model.
作者
麦启欣
张奕源
罗霞
MAI Qixin;ZHANG Yiyuan;LUO Xia(School of Traffic and Logistics,Southwest Jiaotong University,Chengdu 611756,China)
出处
《综合运输》
2023年第7期13-20,共8页
China Transportation Review
基金
四川省科技计划项目(2020YJ0255)。
