This study develops a positive approach to travel demand analysis, which does not assume perfect rationality (i.e. complete information and utilitymaximization) in travel decision-making. Instead, the process through which travellers learn the characteristics of the transportation system, accumulate spatial knowledge, search fro alternatives under imperfect information, and employ subjective beliefs and heuristic rules in decision making is theorised. The proposed SILK theory is able to produce quantitative models of individual travel behaviour in which learning follows Bayes principles and behavioural rules are empirically derived by knowledge acquisitionmethods. System-level demand patterns that emerge from individual behaviour are then obtained using agent-based aggregation techniques. This positive approach is demonstrated by the subsequent development of a route choice model for Minneapolis-St Paul, Minnesota, USA metropolitan area, including the quantification of spatial knowledge and Bayesian learning process, collection of process data for model estimation and validation, the derivation of route search and choice heuristics as production (if-then) rules, ans the prediction of aggregate flow patterns. The traffic equilibrium under the adopted positive assumptions is defined as the behavioural user equilibrium (BUE) at which the subjective search gain is lower than the perceived search cost for all users. Results suggest that normative assumptions, such as perfect information and unlimited human abilities to maximise utility, can produce significant prediction biases. The proposed positive approach, consisting of the SILK theoretical framework , practical data collection techniques, and unconventional travel modelling methods , serves asan alternative framework (to rational behaviour theory) for developing travel demand models. For the covering abstract see ITRD E144727. Reprinted with permission of Elsevier.
Abstract