ABSTRACT
Road traffic information provision has become indispensable for advances in travel in safety and comfort on a freeway. Use of the present travel time by data from vehicle detectors equipped along a freeway is proposed as a travel time estimation technique for providing travel time information. If the traffic situation on a freeway changes by the occurrence of a delay, in particular, a large difference is apt to occur between the present travel time estimated and the real travel time. Therefore, a new model which can predict the future travel time more precisely for providing travel time information is required. This chapter presents a study that proposes a new prediction model. Before modelling a new travel time prediction model by neural network driven fuzzy reasoning, it points out some problems of conventional present travel time estimation.
