Delay Modelling at Signalized Intersection Under Mixed Traffic Conditions

Delay is one of the major parameters in estimating the level of service (LOS) of a signalized intersection. Several methods are available for estimating the delay at a signalized intersection. Highway Capacity Manual (HCM) method and Webster’s method are the most popular and widely used methods among them. These methods are based on homogeneous traffic conditions. But in developing country like India, traffic condition is highly heterogeneous with almost no lane discipline. So, these conventional methods will give errors in de-lay estimation and a proper delay model is needed as per Indian traffic conditions. In the present study, a new delay model is developed by modifying Web-ster’s delay equation. Field delay was estimated according to HCM 2010 procedure. Variations of field delay with different input variables were studied. The semi-empirical adjustment term in Webster’s equation is replaced with the field delay adjustment factor. The adjustment term was developed with two methods, Artificial Neural Network (ANN) model and the regression method using SPSS software. Mean absolute percentage error (MAPE) and coefficient of determination (R2) were taken into consideration for determining a better result. Results obtained from both models were compared with the conventional method and the proposed model gives better results.