摘要: |
Over the last 5 to 10 years rising levels of traffic flow on motorways have lead to the development of increasingly advanced intelligent vehicle-highway systems dedicated to improving capacity, stability, and safety. Many of these advances are to be achieved by modifying behavior or introducing a degree of vehicle control. To fully understand the effects of these mechanisms, however, it is necessary to develop and use appropriate microscopic simulation models. Accurate modeling of microscopic driver behavior presents several difficulties, ranging from the requirement for high-quality (dynamic) data on real driving behavior through the methodology used for cross-checking between simulated and observed behavior to the philosophical basis of modeling itself. These issues are discussed, the degree of shortfall present in current simulation and analysis techniques is examined, and potential guidelines for increasing the validity of microscopic simulation models are suggested. |