摘要: |
It seems intuitively obvious that the more traffic that tries to use a given section of road, the slower it must move, but the precise mechanisms behind this relationship are surprisingly elusive. The availability of large amounts of data collected automatically raises the possibility of validating theories and models. This Insight Report examines the features of some actual data and speed–flow–density relationships, and “classical” models of speed, flow and density in the context of the wealth of detailed traffic data now available. Data from detectors support the idea that average speed decreases with increasing traffic density, although the data suggest that this decrease is only significant above a certain density. Below this, speed is virtually independent of density. This is not a feature of “classical” macroscopic models, suggesting that none of them fully describes traffic over the full range of possible densities. The role of speed, flow and density in queuing theory is also examined, including a case study of modelling a moving bottleneck. |