Institute of Transportation, MOTC

       Advanced public transport services use historical average travel time data between bus stops to estimate the estimated time of bus routes, use historical data grouped by holiday and workdays, and then reassess the travel time of subsequent stations. The vehicle leaves the station based on the location of the vehicle, but the estimated time is affected by road traffic and driving behavior.
There are few studies on improving the travel time quality of on-board equipment. Therefore, this study has made reference and improvement on Bayesian regression theory.
       And apply historical data to construct the travel time mode of each bus line in different time periods. According to the peak time and the general time, it is divided into four groups of peak time and general time, and then divided into morning peak hours: 6:00 am to 9:00 am. In the afternoon, the peak time is from 4:00 pm to 8:00 pm, and the general time is from 9:00 am to 16:00 pm and from 8:00 pm to 6:00 pm. The most concentrated data is used to estimate and clear outliers, and the data is close to the actual situation.
       This study takes the Taichung Bus 53 route as the analysis object, and estimates that the estimated time is similar to the actual arrival time.