A Constrained Route Choice Model for Metro Passengers

Route choice analysis in metro system is the key to extract passengers’ route choice preference and master the spatiotemporal distribution pattern of passenger flow, especially under networked condition. Recently, in order to exactly analyze passengers’ route choice behavior, logit-based semi-compensatory route choice models have received wide attention for that the existence of non-compensatory choice behavior is proved in individual trip’s decision-making process. A Constrained Multinomial Logit (CMNL) route choice model is proposed to analyze metro passengers’ route choice behavior in this paper. The CMNL route choice model’s utility function includes compensatory, non-compensatory and error components. The compensatory component represents the trade-offs among the factors concerning to level of services and network topology, and a revised angular cost is proposed to describe the effect of network topology. The non-compensatory component is a cut-off function consisting of spatiotemporal constrained attributes, referring to travel time and number of transfers. Then, in order to estimate all parameters endogenously and identifiably which is not well solved in previous research, the threshold of a specific constrained attribute is designed as a function instead of a constant. Finally, based on the surveyed data in Guangzhou Metro, numerical analysis is conducted and the results show that the proposed model successfully avoids the generation of route choice set, which is usually viewed as a difficult point in route choice modeling. Meanwhile, the proposed revised angular cost is superior to traditional one and all parameters can be estimated endogenously and identifiably in the proposed model.