Modelling and Evaluation of Work Zone Queues: Istanbul Case Study

Road/lane closures due to maintenance works adversely affect the traffic conditions, causing delays in the various locations of the whole network. In Istanbul, Metropolitan Municipality charge the companies for the planned closure of the highways for a given time interval using simple spreadsheet calculations. These calculations include assumed capacity and speed values for each type of highway. In this paper, it is aimed to establish a methodology for the work zone evaluation that considers seasonality in demand and spatiotemporal variability in the flow and capacity. This goal will be achieved by observing and assessing the characteristics of İstanbul roadways and work zone areas. After the assessment, a traffic microsimulation software, namely PARAMICS, was used to simulate various road and work zone combinations to produce traffic data, including queue length, waiting times, vehicle speeds, flow and density of work zone section. Various machine-learning models were developed and their performance on test data will be analyzed. The selected model was used as a sub-model to estimate the expected delays, speeds and accelerations on different road types, different demands and lane closure combinations. A second sub-model used the delay, speed and acceleration values to estimate the total excess user cost due to road closure with determined delay-cost parameters.