Modeling Dynamics of Vehicle-Based Performance Measures of High-Occupancy Vehicle and General-Purpose Traffic Systems

Abstract

This study aimed at modeling dynamics of traffic systems with integrated and cointegrated stochastic processes using multiple time series of traffic flows generated by general purpose (GP) and high-occupancy vehicle (HOV) systems. Traffic flows on these systems are likely to exhibit trending behavior due to subsystems interactions both in the short run and long run. If they share a common trend, analytic technique is required to model common underlying stochastic trend that in the long run induces equilibrium behavior in a traffic system. Vector-error correction model (VECM) is used for multiple time series variables with persistent common trends generated by integrated variables and that the linear combination of these variables reflects cointegrated processes. Using VECM model, the results showed that traffic flows on high-occupancy vehicle (HOV) and general purpose (GP) traffic systems are integrated and cointegrated, and there are short run dynamics and long run linear equilibrium relationships between system variables. These findings are useful in studying the effectiveness of implemented traffic operation strategies and/or future planning of alternative strategies depending on equilibrium state of the traffic system.