Advanced traveller information services (ATIS) and load-dependent user charges may reduce traffic congestion. The effectiveness of ATIS depends on several factors, such as the information quality, the utilisation of road networks, the equipment of vehicles, the rate of use, and the number of complying users to recommendations. Travellers’ reactions on road tolls especially depend on the pricing system and their own price sensitivity. It is very likely that information quality of expected traffic conditions will improve in the future. Additionally, more car drivers will use ATIS than today. Thus, the question arises to what extend transport information will be able to influence traveller behaviour and traffic conditions in road networks. This dissertation introduces a microscopic model of travel behaviour for analysing the effects of individualised, dynamic transport information and road tolls adjusted to traffic load. The presented approach integrates the choice of a route and departure time as a response to pre-trip information or load-dependent user charges. Furthermore, it considers individual route choices during the journey as a reaction to on-trip information. The model is based on an empirical study that was carried out by the Institute for Transport Studies at the University of Karlsruhe. In order to model transport information the feedback loops between traffic load, recommendations of the service provider, and decisions of travellers were taken into account. Referring to travellers’ reactions to road pricing, the model reproduces learning behaviour by iterating through the decision process. Finally, a microscopic traffic flow model simulates individual trips by car in the road network. The output of the simulation runs shows differentiated results. For example, transport information is able to reduce travel times of road users in most cases if rates of ATIS use are below 30% or road capacity decreases sharply to 50% or more because of traffic congestions. In contrast high rates of ATIS use from more than 70% together with low traffic disruptions or small distinctions between travel times of alternative routes will probably make traffic conditions worse. Generally, all drivers take advantage of transport information, whereas users realise better journey times than non-users up to a rate of ATIS use of about 30%. With higher rates travel times of all drivers become more and more similar. In addition, journey times are mostly better in simulation runs where ATIS only provide on-trip information. Hence, pre-trip information very often creates a self-destroying prophecy for traffic forecasts. Finally, loaddependent road tolls are able to increase the robustness of traffic flows by smoothing traffic load curves. But for cutting travel times, road tolls can only be applied efficiently, if the road network is close to or above its capacity. If there is no congestion, no positive effects towards reducing travel times are likely to be exploited through road charges. (Author/publisher)
Abstract