MODELLING TOLL PLAZAS USING QUEUE THEORY

摘要

This study aims to increase the operational knowledge of the toll plazas to maximize their performance, using the queuing theory. First, we describe the theory of queues and its appropriateness to the collection booths. We defined the concept of level of service for toll plazas with their performance standards for waiting time in queue and average queue length. It is shown a number of assumptions (input variables, criteria for the traffic division, and information output) and develops a model called Car Motion. It is verified the consistency of queuing theory to simulate the toll plaza, through calibration of the model, with error in the order of 11%. Analyses are made in five types of operational scenarios. It is proven the exponential relationship between the volume of traffic and waiting time in queue. The performance of exclusively manual booth is similar to mixed booth (manual and electronic). There is a linear relationship between increasing percentage of heavy vehicles and toll capacity reduction. The increasing of the percentage of electronic payments to 41% can reduce 33% the number of booths manuals, representing large cost savings. Finally, the use of 50% of prepaid ticket payments achieves a 38% gain in capacity.