Research Projects

The overarching goal of this project is to advance behavioural realism in travel demand modelling by developing a novel framework that incorporates socio-psychological influences and perception errors. Specifically, the proposal aims to address four methodological shortcomings of traditional ABMs. First, traditional ABMs lack mechanisms to integrate socio-psychological factors, such as risk perception and social influences, which play a significant role in determining travel choices. Second, they fail to account for perception errors in travel attributes like travel time and travel cost, leading to discrepancies between predicted and actual behaviour. Third, existing ABMs rely heavily on objective measures, assuming that all travellers assess travel conditions similarly, which overlooks individual biases and preferences. Fourth, most ABMs struggle to capture the variability of travel preferences across different population segments (e.g., age and socioeconomic status), leading to overly generalized predictions. These shortcomings raise the following research questions: 1. How can socio-psychological factors, such as risk perception and social influence, be integrated within the ABM framework to improve behavioural realism? 2. How can perception errors, specifically in travel time and travel cost evaluations, be modelled to reflect individual biases in travel choices? 3. What methodological advancements are needed to incorporate subjective assessments of travel conditions within an ABM framework? 4. How can we tailor ABMs to better capture travel behaviour variability across different population segments? This proposal addresses these questions through a novel, methodologically rigorous framework. This proposal solves the first question by developing a new ABM extension that includes socio-psychological variables as latent factors influencing travel choices. This extension will use a hybrid discrete choice framework, incorporating behavioural constructs. The second question will be addressed by introducing perception error modelling in ABMs. We will incorporate perception biases for travel time, cost, and comfort by using perceptual distortion functions, allowing the model to capture individual differences between perceived and objective measures. The third question involves adapting the ABM framework to reflect subjective assessments of travel conditions by incorporating latent variables that represent these subjective assessments. These will be calibrated using latent class modelling to understand distinct perceptual categories and apply them across various travel contexts. Finally, the fourth question will be addressed by tailoring the ABM framework to population-specific preferences through a segmentation approach. This will involve grouping individuals with similar socio-psychological profiles, allowing the model to capture variability across population segments, thereby enhancing predictive power.