Bus IC Card Data-Based Decision Support Research

As an important source of data and information about passenger flow, Public transit smartcard data, with its advantage in its scope of usage, continuity of time and low-price, are expected to play an essential role in public transit decision making. Not many studies can be found in this area, and there are lots of promising opportunities.In this paper, based on the discussion of smartcard data, the author proposed the framework of future research in public transit planning based on smartcard data. Prediction of short-term passenger flow with high accuracy was accomplished by Generalized Regression Neural Network. Then, by using the predicted data as the input, the departure timetable with the least severity of crowding was built. Viewing two shortcomings of some former study about the minimization of passenger transfer waiting time:the big possibility of putting the passenger in danger of longtime-waiting by treating the minimized passenger transfer waiting time and the synchronization of buses as equal as well as the negligence of actual difference of passenger transfer counting in different stops, this paper developed a model of optimization of the estimation of overall passenger transfer-time, which was based upon the timetable with least severity of crowding built previously and statistics of smartcard data. Through genetic algorithm and computer simulation, the solution of the model was given as well as further discussion.