Study of traffic delay model at signalized intersection

Traffic delay at signalized intersection is one of important parts of micro-operational evaluation and macro-demand forecast of urban transportation. For without a standard delay model at signalized intersection in the past,Webster signalized intersection delay model,Highway Capacity Manual (HCM) 1985 signalized intersection delay model and HCM 2000 signalized intersection delay model have been directly used till now in China. There are many differences in characteristics between the mixed traffic flow in Chinese cities and traffic flow in the United State cities. The purpose of the study in this paper is to determine if the delay results calculated from the above mentioned delay models conform to the actual delays in the signalized intersection in Chinese cities. A methodology including three steps is used to do this research. Firstly,by analyzing the advantages and disadvantages of the above models,HCM 2000 signalized intersection delay model had better characteristics than the other models and it was chosen to calculate the theoretical delay at signalized intersection. Secondly,the actual delay survey with the License Plate method was conducted on the 31 main signalized intersections in Shenzhen. Results show that the actual survey delay is significantly different from and higher than the calculated delay. Thirdly,a dispersed dot map including the theoretical calculated delays and the actual survey delays was drawn. The figure clearly indicates that dots from the surveyed delay are higher than dots calculated from the delay at the same signalized intersection and there are similarly variable trends between the theoretical calculated delays and the survey delays. It appears that the survey delay are a multiple of the calculated delay with HCM 2000 signalized intersection delay model. The Least Squares Method was used to calibrate the coefficient between the real data and the calculated data. Results indicate that the calculated delay with HCM 2000 signalized intersection delay model multiplied by 1.269 make the actual delay at the signalized intersection in Shenzhen. The conclusions suggest that HCM 2000 signalized intersection delay model can't directly be used to analyze the delay in signalized intersection in Chinese cities.