Travel Time Measurement of Dynamic Sequence Considering Headway on the Arterial Roads

要旨

In this paper, we proposed an algorithm measuring the travel time through the handy web camera with color sensor that collect the data for vehicle classification and sequence as well. Proposed algorithm measured a travel time of vehicle for searching the best matching sequences passing from a certain point of the upstream to the downstream. If the sequence is minimized the change of sequence in passage section, the reliability would be improved for probe vehicles. In the previous study, sequence-qualifying method was fixed value, but we proposed the dynamic sequence qualifying by adopting the headway. Headway is a time interval between leading vehicle and following vehicle. The dynamic sequence algorithm is able to minimize the effect of the traffic signal in measuring the travel time in the arterial road. We verify adopting a real data set from an AVI system. The dynamic sequence algorithm is the higher data reliability than fixed sequence algorithm about 20%. And It’s the available rate was improved about17% than that of the fixed sequence. The travel time result was verified by the MARE (Mean Absolute Relative Error) and MAE (Mean Absolute Error) as the comparison between AVI data and sequence matching data. As an improvement of rate of in and out vehicles, travel time of the fixed sequence algorithm’s MARE and MAE was improved. But dynamic sequence algorithm was not. It resulted in an equally MARE and MAE values distribution at graph. Both algorithms can measure the correct travel time, dynamic sequence algorithm was better than fixed sequence algorithm about data reliability to 20% , available rate to 20%, travel time.