SHORT-TERM PREDICTION OF TRAFFIC FLOW CONDITIONS IN A MULTILANE MULTICLASS NETWORK

摘要

t Improving traffic conditions on urban and motorway networks given the highly complex interplay of the multiple control instruments, requires a profound understanding of the response of the heterogeneously composed traffic flow to a variety of diverging traffic control strategies. Such insights will reveal effective control scenarios yielding desired traffic flow behaviour, thereby significantly improving network efficiency and traffic safety, considering the diverse objectives of the multiple user-classes, such as vehicle-types (trucks, busses, person-cars, vans, intelligent vehicles, etc.), traveller types (commuters, freight, recreational, etc.), paying and non-paying traffic, contra-flow traffic, and various types of supported vehicles (Autonomous Intelligent Cruise Control, Intelligent Speed Adaptation). A large number of contemporary traffic control options, such as priority-control for busses at metered on-ramps, dynamic truck overtaking prohibitions, and dynamically assigning roadway lanes to specific user-classes, recognise the importance of distinguishing these user-classes.