Abstract
In this study, we chose two isolated intersections to collect traffic volume per minute, queue length and signal timing in Pitesti City traffic control network. There are two queue-length estimation equations, the system method of vehicle scanning and the regression method. In order to verify the above models, another data set was collected to compare with the traffic performance of the above models. From the results of a specific test, it is found that the regression model performs better than the other one in traffic estimation efficiency based on the existing traffic control system framework in Pitesti City.