Paper Title
An Application Of Adaptive Neuro-Fuzzy Inference System In Vehicle Delay Modeling

Abstract
An Adaptive Neuro-Fuzzy Inference System (ANFIS) approach is widely used to solve engineering problems. In this study, a model for estimating the average delay was developed by using ANFIS according to effective parameters, and the performance of the model was statistically evaluated. In this model, a hybrid method which is ANFIS was performed for training and test. The artificial network part of method used three input parameters representing cycle time of signalization (C), the green time (g) and the degree of saturation (x). The results of neuro-fuzzy network part of method were compared with the analytical models named HCM and Australian delay models. The performance of ANFIS was the better than other models and showed close agreement with simulated values. Keywords – Vehicle Delay, Neuro-Fuzzy, Signalized Intersection