自动代客泊车接受度及停车选择行为影响因素分析  

Influence Factors of Automated Valet Parking Acceptance and Parking Choice Behavior

作　　者：赵传林 靳思缘 武海娟 丁力 ZHAO Chuan-lin;JIN Si-yuan;WU Hai-juan;DING Li(School of Civil and Transportation Engineering,Beijing University of Civil Engineering and Architecture,Beijing 102616,China)

机构地区：[1]北京建筑大学土木与交通工程学院,北京102616

出　　处：《科学技术与工程》2023年第35期15259-15268,共10页Science Technology and Engineering

基　　金：国家自然科学基金(71971014);北京市教委科技计划(KM202010016007);北京建筑大学金字塔人才培养工程项目(JDJQ20200301);北京建筑大学研究生创新项目(PG2023053)。

摘　　要：自动代客泊车是无人驾驶系统的重要组成部分,对于缓解城市交通拥堵具有重要作用。为了探讨自动代客泊车接受度及选择行为影响因素,首先基于技术接受模型,选取感知易用性、感知有用性、行为态度、行为意向、感知风险、信任六个心理潜变量构建自动代客泊车结构方程模型。随后基于线上调查问卷数据,采用随机森林算法构建考虑心理潜变量模型与不考虑心理潜变量模型,研究发现考虑心理潜变量的随机森林模型拟合优度更高,更适合用于分析自动代客泊车选择行为的影响因素。通过袋外估算误差率对考虑心理潜变量的随机森林模型进行变量筛选,结果表明当变量个数为9时,该模型的袋外估算误差率最小,筛选后的变量分别为停车费用、受教育程度、感知风险、出行总成本、行为意向、信任、月收入水平、感知易用性、年龄。最后将筛选后的9个变量纳入逻辑回归中进一步分析各变量对选择行为的影响程度,参数标定结果表明停车费用、受教育程度、感知风险、出行总成本、行为意向、信任、月收入水平、年龄变量均对选择行为具有显著影响。Automated valet parking(AVP)is an important part of driverless systems and plays an important role in alleviating urban traffic congestion.To investigate the influencing factors of acceptance and parking choice behavior of automatic valet parking,the structural equation model based on the technology acceptance model was established,using six psychological latent variables,i.e.,perceived ease of use,perceived usefulness,behavioral attitude,behavioral intention,perceived risk,and trust.The random forest algorithm was used to build acceptance models considering psychological latent variables and ignoring psychological latent variables based on online survey questionnaire data.It is found that the random forest model considering psychological latent variables has higher fitting accuracy,making it more suitable for analyzing the influencing factors of automatic valet parking choice behavior.The random forest model with psychological latent variables is used to screen the variables through the out-of-bag estimation error rate,and it is found that the model has the smallest out-of-bag estimation error rate when the number of variables is 9.The selected variables are parking cost,education level,perceived risk,total travel cost,behavioral intention,trust,monthly income level,perceived ease of use,and age.These 9 variables are then included in the logistic regression to further analyze the degree of influence of each variable on the choice behavior.The parameter calibration results show that parking cost,education level,perceived risk,total travel cost,behavioral intention,trust,monthly income level,and age have significant effects on the choice behavior.

关 键 词：自动代客泊车 结构方程模型 接受度 停车选择行为 随机森林 逻辑回归 

分 类 号：U491[交通运输工程—交通运输规划与管理]