Traffic State Fuzzy Evaluation Based on Probe Vehicle Data
Fan Yao-zu · Jisuanji fangzhen · 2008
A new traffic state evaluation approach based on probe vehicle and multi-classification fuzzy pattern recognition was discussed. This approach has two key points. For the first one, probe vehicle used for achieving five congestion indexes, i.e. average speed of probe vehicle, congestion index, proportion stop time, acceleration noise and mean velocity gradient, and the value of these indexes was decided using VISSIM. For the second point, multi-classification fuzzy pattern recognition was applied in traffic state evaluation processing. The VISSIM simulation results show that this new approach can evaluate the traffic state efficiently and easily, and can also effectively reflect the punctuation of traffic state in the evaluation interval.