Empirical verification of modeled queue lengths

The purpose of left turn lanes is to separate left-turning vehicles from the through-traffic stream. This tends to increase capacity by adding another lane to the approaches of an intersection and also improves safety and reduces delay. Capacity, delay and operational issues are usually related to a single shared lane between left-turning vehicles and through-vehicles which are expected to be eliminated with an exclusive left turn lane. If left turn queues are not accurately estimated, they may be the source of these safety issues counteracting the benefits of the left turn lane. The accurate determination of left turn queues is very critical for the safety and efficient operation of an intersection. If a left turn lane is not adequately estimated to store the longest expected queue during a signal cycle with a high probability, left-turning vehicles may back up into the adjacent through lane or the through traffic may block the entrance to the left turn lane preventing left-turning vehicles from entering their lane. These effects may lead to additional intersection delay and also rear-end collisions which compromise the safety of the intersection. Currently, the state-of-art in left turn lane modeling is the use of various software packages which are either analytically or simulation based. A survey carried out as part of this study indicated that traffic engineers are not very confident in the results reported by these methods. Some traffic engineers surveyed stated that they felt the models reported shorter queue lengths than those actually observed in the field while others stated that they felt the models reported longer queues than actual queue lengths. The survey also indicated that most of the engineers surveyed did not compare model queue lengths to any field values to ascertain the performance of the models. The objective of this study is to determine the most reliable traffic software package used in left turn modeling. Several models have been developed and are available to the traffic engineer for estimating left turn lane lengths. The engineer is thus faced with selecting the most reliable model. Several studies have been conducted to determine the performance of models. The results have shown that microscopic models which are simulation-based are the more reliable and report queue lengths comparable to observed field queue lengths. To add to the knowledge of model reliability, this study evaluated the most frequently used models in traffic analysis-- Synchro, TEAPAC, HCS+ (all macroscopic models) and SimTraffic (a microscopic model). These models were selected based on results from the survey carried out as part of the study. Four intersections operating under varying traffic conditions were modeled. Results showed that SimTraffic, a simulation based model, was the most accurate in estimating left turn queues.

Name thesis:degree_name

Civil Engineering

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Civil Engineering

Year

2012

Author dc:creator

• Ababio, Gloria

Contributors dc:contributor

• Brogan, Jim
• Hall, Jerome
• Zhang, Guohui

• Left-turn lanes--Computer simulation
• Traffic congestion--Management--Computer simulation
• Queuing theory.

Language dc:language

English

Identifier

https://digitalrepository.unm.edu/ce_etds/56

OAI identifier oai:identifier

oai:digitalrepository.unm.edu:ce_etds-1057