SHE Shiying^1,2, ZHENG Meng², XIANG Yanling³, LIU Heng³, CHEN Yanyan¹, LEI Huanyu³

　　1. Beijing University of Technology, Beijing 100124, China; 2. Wuhan Transportation Development Strategy Institute, Wuhan Hubei 430017, China; 3. Shenzhen Urban Transport Planning Center, Shenzhen Guangdong 518000, China

　　【Abstract】 With the comparison of the practical limitations of the two model frameworks of the activity chain and the trip chain, this paper proposes a new model method based on the Aggregate Tour-based Model and analyzes the core algorithm. Then the Chinese megacity Wuhan is taken as an example to verify the feasibility of this method. The results show that the model framework fully considers the time and spatial constraints and internal consistency of the outbound and return journeys of each tour in terms of time, travel mode, main destination choice and stop selection, as well as the iteration and convergence between demand and supply. The method can adapt to all kinds of tours in a unified framework for modeling. Using the basic tour as the analysis unit, the model can both effectively reduce the complexity of resident activity modeling and ensure the consistent characteristics of residents’ activity. In the case of ten million population, 3467 TAZS, and complex traffic environment, the model performs a great convergence process. Gap<0.2% and Gap<0.1% require 28.1 hours and 49.6 hours respectively, while the convergence process of 1678 TAZS only requires 9.5 hours and 17.8 hours, the model calibration and the reality and sensitivity test also fully verified the performance of the model. Finally, this paper discusses the relationship between the Four Steps Model, the Tour-Based Model, and the Activity-Based Model so as to avoid blindly falling into the trap of model refinement and complexity.

　　【Keywords】 activity-based model; aggregate tour-based model; basic tour; tour time model; destination mode choice model; empirical research; Wuhan

　　【DOI】 10.13813/j.cn11-5141/u.2021.0053-en