Travel Time Prediction for BRT Buses by Using Support Vector Machine (SVM) and Principal Component Analysis

Travel time and delay prediction plays a significant role in prevention of engaging road users with heavy traffics and subsequently increasing their confidence in traffic facilities. In the present study first of all by using the data obtained from data collection system of Tehran Traffic Control Organization, the main database to predict BRT bus travel time is established. Therefore Frhngsra-Azadi BRT line has used as a sample. In the conducted modeling, the support vector machine’s regression analysis along with the principal component analysis is used. By using the stated method, travel time in sample is predicted and the results are discussed and compared. The results show that support vector machine’s regression analysis along with principal component analysis has a high accuracy in predicting the intended purpose and has been able to improve the related results for prediction of transit travel time. Therefore the presented method in this study can be used in practical applications and play an effective role in travel time programming of BRT.