有限理性视野下出行者出行方式选择分层Logit模型研究

在传统分层Logit模型的基础上，放松出行者完全理性的假设，基于有限理性满意决策准则，建立了多方式有限理性分层Logit(BRNL)模型，来描述出行者的方式选择行为.模型假设当方式间的成本差在无差异区间内时，出行者依据偏好或随机进行方式选择；否则，出行者将选择出行成本更小的交通方式.此外，提出了一种嵌套的相继平均算法对模型进行求解.最后，通过一个数值算例验证了模型的合理性和算法的有效性.研究结果表明，出行者并不总是选择出行成本最小的交通方式，其选择行为与出行者的理性程度和偏好有关，且理性程度越低，出行者对方式间的成本差越不敏感.

Nested Logit Model on Travel Mode Choice under Boundebly Rational View

Based on the classical nested Logit model, this paper develops the multi-modal boundedly rational nested Logit (BRNL) model by relaxing the assumption of perfect rationality of travelers under satisficing rule, which is used to describe travelers’mode choice behavior. Travelers are assumed to choose travel mode randomly or in accordance with the habit preference when the cost difference between travel modes is within the indifference band. On the contrary, travelers will choose the travel mode with the least travel cost. In addition, a nested method of successive average is proposed to solve the model. Finally, the rationality of the model and effectiveness of the algorithm are proved by using a numerical example. The results indicate that travelers don’t always choose the travel mode with the least travel cost. Travelers’ mode choice behavior is impacted by the rational degree and preference. Moreover, the lower rational degree implies that travelers become less sensitive to the cost difference between modes.