Analytical models of delay estimation at signalized intersections for variable demand and time conditions

Average delay experienced by vehicles at signalized intersections is one of the essential parameters used to evaluate the performance of an intersection. As a measure of intersection performance, delay is used for both the design and the timing of the intersection because it reflects fuel consumption, driver frustration and the amount of lost travel time.; Delay at a signalized intersection can be measured directly in the field, but this is not always feasible because of the cost, effort or other uncontrollable factors at the intersections. An analytical model, therefore, is an alternative method for estimating of delay at signalized intersections.; In this research, the literature on the current analytical models for estimating delay at signalized intersections is reviewed, and the reliability of these models and their limitations are evaluated.; The primary objective of this research is to develop time dependent analytical delay models for signalized intersections that are suitable for variable demand and time conditions. In the proposed delay models, the delay parameter k used to describe the arrival and departure pattern is expressed as a function of degree of saturation (x), and time period for analysis (T). This delay parameter k has a fixed value in the Australian, the 1985 Highway Capacity Manual (H.C.M) and Canadian delay models. The last two models use a fixed value of 0.5 for the delay parameter k, which is based on queuing theory and describes a queuing model with random arrivals and fixed service time.; The TRAF-NETSIM microscopic simulation model is used to verify the proposed delay models. The simulation is conducted in three consecutive time periods. The demand and time period for analysis are varied for each case and 48 cases are investigated to evaluate the performance of the proposed delay models. Then, the delays generated by the simulation and estimated by the proposed models are statistically compared. The results from this research indicate that there is a good agreement between delays generated by the simulation and delays estimated by the proposed models.; It is concluded therefore that the proposed delay models are an improvement on existing delay models, and provide a good estimate for delays in the field.