Travel time reliability measure based on predictability using the Lempel-Ziv algorithm

摘要

Travel time reliability is a crucial factor affecting traveler behaviors, including choices of route, departure time, and mode. One commonly accepted definition of travel time reliability, given by the Federal Highway Administration, states that "Drivers are used to congestion and they expect and plan for some delay, but most travelers are less tolerant of unexpected delays. Travel time reliability measures the extent of this unexpected delay." However, most widely used measures of travel time reliability replace the notion of "reliability" with "variability," and there lacks measure that directly incorporates the unforeseen levels of travel time. Thus, we extended travel time reliability from the perspective of predictability based on the Lempel Ziv algorithm in the field of information theory, which provides an upper bound of travel time predictability based on historical data. To demonstrate the advantages of proposed measure and correlations with traditional metrics, we conduct thorough case studies with various data sources along with artificial simulations to explore the nature of the proposed predictability measure and its distinct features from other reliability measures.