Data merging

Merging FCD data to static detector data: a generalized framework and its application in Yokohama

Travel time estimation is important for today's traffic information systems, so it is for the Japanese VICS system. Travel time estimation is very precise when FCD measurements can be used. Unfortunately floating cars are mostly sparse information sources.

Alternatively, interval detector data can be used. Travel time estimation on the base of detector data tends to parameter uncertainty.

The framework presented here uses primarily cumulative detector data. It provides results even without FCD information. The additional FCD information is used for correcting the calculated travel time and O/D estimates and for re-calibrating the calculation parameters.

The interval detector data algorithm needs a minimum detector density because of calculation parameter uncertainties. Additional FCD information makes the estimates more precise. As FCD information frequency increases, detector density may decrease. This can compensate for broken detectors.

Download the paper from here: Bordeaux_ITS-1336  [PDF, 858 KB]


Merging FCD and detector data: testing and improving the new VICS algorithm

The VICS organization in Japan uses a detector-based reliable and stable travel time esti-mation algorithm for off-highway street segments. However this algorithm tends to travel time estimations that are too short, due to missing information about the downstream traffic light queues. This problem can be overcome by adding FCD (Floating Car Data) information. The paper introduces the algorithms developed and used within a generalized data merging environment introduced at the previous World Congress and shows field results from the test area in Yokohama. Because FCD is still sparse, merged travel time information ages quite quickly and tends to have the algorithm fall back to short travel time estimations. A simple intersection phasing estimation is discussed for the times where no FCD is available.

Download the paper from here: Melbourne_ITS-AP-TP0070  [PDF, 735 KB]