Developing a Multiday Travel Demand Modelling System

Abstract

This thesis presents a framework for developing a multiday travel demand modelling system utilizing single-day travel data, addressing the limitations of traditional travel demand models. The framework includes three core methods: a pseudo-panel approach to construct longitudinal datasets from single-day surveys, a multiple sequence alignment technique to derive representative weekly activity patterns, and a simulation-based optimization model to generate weekly schedules within an integrated urban modelling system. This thesis advances the understanding of temporal dynamics of travel behaviour without the high cost of multi-day data collection, demonstrating that single-day data can effectively approximate weekly travel patterns. The iTLE WeekSim module developed in this thesis provides detailed simulations of weekly travel schedules for a synthetic population, including crucial information related to in-home and out-of-home activity durations, travel origin-destinations, and travel modes. The outcomes of this research can significantly benefit policymakers by offering a decision support tool for adaptive urban transportation planning. The iTLE WeekSim module can enable policymakers to analyze weekly travel patterns and design data-driven strategies to improve transportation efficiency, reduce emissions, and manage congestion. Besides, this framework lays the methodological foundation for multi-day travel demand modeling with limited data, an approach rarely explored previously by the travel behaviour researchers.