基于低碳心理潜变量Logit模型的出行方式预测模型

发布时间：2018-09-11 10:12

【摘要】：为深入解析低碳出行行为机理,提出了将低碳心理变量引入离散选择模型的研究方法。基于计划行为理论考虑低碳出行心理因素,建立了多原因多指标潜变量模型。将潜变量模型预测后的潜变量代入多项Logit模型,构建了带低碳出行心理变量的混合选择模型。以镇江市居民通勤调查样本为研究对象,实证结果表明:相对传统离散选择模型,所建的混合选择模型预测精度整体提高了2.45%;低碳心理变量对于出行方式选择的影响各不相同,低碳出行方式"行为态度"对出行方式选择没有显著性影响,而高碳出行方式"行为态度"对低碳出行有显著性负影响。通过对月收入的敏感性分析检验模型性能,结果显示:随着收入的增长,低碳出行方式占比逐渐降低,其他两种方式占比变化相反,在考虑低碳心理因素后,高碳出行方式占比上升幅度降低。
[Abstract]:In order to analyze the mechanism of low carbon travel behavior, a method of introducing low carbon psychological variables into discrete selection model is proposed. Based on the theory of planning behavior, the multiple factors of low carbon travel psychology are considered, and the latent variable model of multiple factors is established. A mixed selection model with low carbon travel psychological variables was constructed by inserting the latent variables predicted by the latent variable model into multiple Logit models. The empirical results show that compared with the traditional discrete selection model, the prediction accuracy of the established hybrid selection model is improved by 2.45 percent, and the influence of low carbon psychological variables on the choice of travel mode is different. "Behavioral attitude" has no significant influence on the choice of travel mode, while "Behavioral attitude" has a significant negative effect on low carbon travel mode. According to the sensitivity analysis of monthly income, the results show that: with the increase of income, the proportion of low carbon travel mode decreases gradually, the other two ways have the opposite change, after considering the low carbon psychological factors, The proportion of high-carbon travel mode increased by a decrease.
【作者单位】：江苏大学汽车与交通工程学院;江苏大学财经学院;
【基金】：国家自然科学基金项目(71373105,61573171) 江苏省“六大人才高峰”项目(2015-JY-025)
【分类号】：U491.14

【相似文献】