Spatial Transferability of Normalized Time-Discounted Recursive Logit Model in Excursion of Districts around Stations

Abstract

In recent years, pedestrian circulation has become a major theme as station and town spaces are being modified by the insertion of new transportation modes. However, the existing models assume the same travel time between all intersections and do not correctly evaluate the reduction of future value, and do not correctly evaluate the sequential decision making that is characteristic of pedestrian migration. The future value reduction is not properly evaluated. In this study, we formulated a recursive migration model with a generalized time discount rate for link length that can be applied to networks with different link lengths, and verified the validity of the model and the transferability of parameters by conducting parameter estimation and validation comparisons for eight station networks. The model is validated by parameter estimation and validation comparison in eight station networks, and its superiority is demonstrated by comparing its likelihood with that of a conventional model.