Pattern Recognition for Urban Traffic Speed Prediction
Traffic speed or simply congestion prediction algorithm for urban road traffic networks can be achived by pattern recognition. The motivation of the prediction is to provide short time forecast in order to support ITS functionalities, such as traveler information systems, route guidance (navigation) systems, as well as adaptive traffic control systems.
A potential and efficient solution to this problem is the application of a soft computing method. Namely, an artificial neural network (ANN) is used for the forecast by involving the measured speed patterns.
The ANN is trained by using data produced by realistic microscopic road traffic simulator (Vissim). The proposed algorithm is developed and analyzed on a real-word test network (part of downtown in Budapest).
The test network for the case study
The input of the model is calculated from mean speed data on links of the road network during 5 minutes sampling intervals. The output of the model is defined as 4 different speed categories.
The applied MuliLayer Perceptron model for ANN training
For the analysis a train data set with 2500 records and a test data set with 1000 records were applied.
Slightly decreasing recognition rate over the 5, 15 and 30 minutes forecasts
T. Tettamanti, A. Csikós, Zs. Viharos, K.B. Kis, I. Varga: Traffic speed prediction method for urban networks – an ANN approach, 4th International Conference on Models and Technologies for Intelligent Transportation Systems (MT-ITS 2015), Paper 11, IEEE Xplore ISBN: 978-963-313-142-8.
T. Tettamanti, A. Csikós, K. B. Kis, Zs. J. Viharos, I. Varga: Pattern recognition based speed forecasting methodology for urban traffic network, Transport, ACCEPTED