In Intelligent Transportation Systems predicting arrival time of bus is a key challenge. The bus arrival time is the primary information for providing passengers with an accurate information that can reduce passenger waiting time. The accuracy of predicting arrival time is highly dependable on several factors such as data input, quality of the data collected and the accuracy of prediction algorithms. Several studies have been implemented by the developed countries in predicting the arrival time under homogeneous traffic conditions. However, the direct implementation of such systems in developing countries especially like India where the heterogeneity of the vehicles can be found on the roads, may not be suitable. The main components in predicting the arrival time is the travel time between the stops in heterogeneous traffic conditions, which need to be incorporated in the prediction scheme for better accuracy. In this paper, ANN, Kalman and Regression models are used also taking delays into account for better prediction of bus arrival time. Data were collected using smartphone having GPS sensor by installing our Android application on it. The performance of our experiment showed that regression model has the better prediction in predicting the arrival time of bus and improvement over other models.
展开▼
