基于哑变量改进的NL模型在交通方式选择中的应用

Application of Improved NL Model Based on Dummy Variables in Traffic Mode Selection

摘要: 论文通过对随机效用理论进行深度剖析，并在原有NL模型的基础上改进，提出了一种基于哑变量改进的NL模型。该模型可重新对效用选择函数的指标进行选取和分类，对于不可量化的性质指标则可以以哑变量赋值的方式注入到效用方程中，该方式能够使回归模型变量解释范围更广，进而达到扩量保质的作用，在面对现实问题的描述时能够起到以一抵多的联合效果。

Abstract: Based on the in-depth analysis of the random utility theory and the improvement of the original NL model, an improved NL model based on dummy variables is proposed. The model can select and classify the index of utility selection function again. The property of not quantitative indicators can be injected in the form of dummy variable assignment to the utility equation. This way can make variable regression model to explain widely, thus expanding quantity with good quality, in the face of the description of the realistic problems to have the joint effect that one is worth more.