Estimation of traffic recovery time for different flow regimes on freeways

This study attempts to estimate post-incident traffic recovery time along a freeway using Monte Carlo simulation techniques. It has been found that there is a linear relationship between post-incident traffic recovery time, and incident time and traffic intensity. For purposes of this paper, the post-incident recovery time is defined as that time beyond the clearing of an incident when pre-incident traffic conditions are achieved and traffic has returned to normalcy or steady state. The research supports Objective 2.1 of the SHA Business Plan, which seeks to develop measures to enhance the Maryland State Highway Administration's (SHA) ability to quantify the impact of congestion and delay on the highway network. In addition, the SHA understands that the capability to reasonably estimate the traffic recovery time for a given duration of incident is crucial in qualifying the cost-effectiveness of current/future traffic management programs involving detection and clearance of incident on freeways. A total of 121 traffic scenarios of traffic intensity (Rho), incident duration, and proportion of lane blockage were simulated resulting in a total of 726 experiments. The VISSIM simulation platform was used to derive values for output flow, density, and speed to determine the post-incident traffic recovery times. The analysis of simulated data showed that for a given incident duration and lane blockage scenario, the recovery time of the traffic increases non-linearly with traffic intensity. The traffic recovery time becomes uniform (stable) for low and moderate traffic intensity values. A set of linear regression models were developed to reasonably estimate the post-incident traffic recovery time using traffic intensity, incident duration, and proportion of lane blockage as exogenous variables.