Short-Term Traffic and Travel Time Prediction Models

Road traffic is the visible result of the complex interplay between traffic demand (the amount and mix of vehicles arriving at a particular place and time) and traffic supply (e.g., capacity, prevailing speeds, and other average traffic characteristics). As a result, short-term prediction of road traffic variables is a complex nonlinear task that has been the subject of many research efforts in the past few decades. The term “short term” usually entails that the variables of interest are predicted for a period up to 1 h ahead, although the exact definition differs largely between approaches. In practical terms, short-term traffic prediction is an important if not critical component for intelligent transportation systems (ITS) and particularly in traffic control and traffic information provision. This paper briefly discusses some general aspects related to short-term traffic prediction. It then provides a taxonomy of the many different approaches reported in the literature for the general problem of short-term traffic prediction; in these cases, artificial intelligence (AI) techniques are discussed either as a complete solution or as part of a hybrid approach to short-term prediction.