Micro interaction modeling with the recursive structure for land transaction and activity chain in urban networks




The microscopic land and transportation interaction model framework and the estimation method for it are proposed. The proposed model consists of a land transaction by landowner based on random utility theory and an activity chain based on a dynamic discrete choice model. As an estimation method, we proposed a new recursive algorithm in which the results of each sub-model become explanatory variables. The proposed model is based on economic theory as well as previous LUTI models. They have only dealt with the interaction between land and transportation on a zonal basis so far. In contrast, the proposed model is formulated on a link level, and the submodels for both land and transportation are disaggregated. The empirical model estimation results show that the proposed method can remove the parameter bias in a reasonable time. We obtained this result by comparing algorithms that assume multiple behavioral equilibria with input land transaction data and activity-chain obtained from a survey of visitors in a local Japanese city. This paper explicitly shows how to a micro land-transport interaction model from three perspectives: formulation, data used, and empirical estimation methods. The proposed recursive model framework and algorithm have demonstrated their practical applicability to street-level policymaking.