Analysis and Modeling of Night Express Bus User's Behavior

最近，日本的许多公交企业陷入了财务困境。在这样的背景下，一些公交公司期待市内夜间快车服务成为扩大市场份额的有希望的政策之一。本研究旨在找出高府至大阪间夜间特快班车的基本特征，并利用分类回归分析及判别函数分析方法分析使用者的意识与需求。一些实证观察表明，客运量的年偏差表现为月度模式，这种模式是由非工作日和旅游用户形成的。研究还发现，公交车座椅舒适度、座椅间空间和行李空间是用户评价的主要因素。通过聚类分析，将用户的活动模式分为四类。此外，本研究旨在确定用户的评价与他们的活动和旅行模式之间的关系。对用户上、下车的日子进行了活动日记调查。LISREL模型的校准考虑了个人属性、出行目的和调度条件等决定因素。

Recently, many bus businesses in Japan have been in financial difficulty. Under such a background, some bus companies expect an intra-city night express bus service to be one of the promising policies to expand their market share. This study aims to identify the basic characteristics of night express bus between Kofu and Osaka and analyze users' consciousness and needs using a categorical regression analysis and a discriminating function analysis. Some empirical observations indicate that the annual deviations of the number of passengers are characterized as a monthly pattern, which is formed by non-workdays and excursion users. They also show that comfort of bus seat, space between seats, and baggage-space are regarded as a principal factor for user's evaluation.A cluster analysis was applied to classify user's activity pattern into four groups. Also, this study aims to identify relationship between the users' evaluation and their activity & travel patterns. An activity diary survey was conducted on the days when users riding on and getting off the bus. A LISREL model is calibrate considering determining factors such as individual attributes, the purpose of travel and scheduling conditions.