Estimating the share of the main components of travel time variability using analysis of variance

Fluctuations in travel demand and uncertainty in transportation supply make travel time uncertain. Uncertainty (variability) in travel time can be attributed to three main components including vehicle-to-vehicle variability, period-to-period variability and day-to-day variability. Several studies developed indexes to measure travel time variability. However, no methods have been developed to estimate the contribution of the main components of variability. In this paper, we employ the analysis of variance method to determine the relative contributions of the aforementioned components to the total variability in travel time. To examine the applicability of the proposed method, real travel time data collected over 3 months is used. The results of the present study show that all three components play a significant role in the variability of travel time. Based on the travel time data used in this study, the share of vehicle-to-vehicle, period-to-period and day-to-day variability components are 15.8, 36.1, and 22.3 percent, respectively. Moreover, 25.8 percent of total variability is contributed to the interaction of day-to-day and period-to-period components. In addition, according to the results of this research, the probability distribution of travel time is multimodal, especially in peak hour periods. Therefore, mixture distributions can model travel time variability more accurately than single distributions.