Delay estimation models for signalized intersections using differential evolution algorithm

Abstract

Delay is an important parameter in the optimization of traffic signals and the determination of the level of service (LOS) of a signalized intersection since it directly reflects lost travel time and fuel consumption.  The accurate estimation of delay is, therefore, an important issue.  However, it is not easy to estimate delay precisely because of stochastic nature of traffic, the behavior of pedestrians and drivers and geometric characteristic of the intersections.  Some delay models have been presented in the literature to obtain accurate estimation of delay by using both heuristic and analytical approaches. This paper proposes new models to estimate the delay. In this research, three types of differential evolution delay estimation models (DEDEM), i.e. linear, exponential and quadratic, are developed using differential evolution (DE) approach. In developing of the delay models, we considered green ratio (g/C effective green to cycle length) and degree of saturation (x=v/c; volume to capacity). The first one changed from 0.35 to 0.60, the second one varied between 0.7 and 1.4. While some of data taken from CORSIM simulation were employed for the development of the models, the rests of them were utilized to verify for coherence of the proposed models. The model outputs were compared analytically with HCM and Australian delay models.  The study results illustrated that R², Mean Square Error (MSE) and Mean Absolute Error (MAE) values of DEDEM were better than those of existing delay models.