Random coefficient modeling research on short-term forecast of passenger flow into an urban rail transit station

Abstract

Taking a representative metro station in Beijing as example, this research has newly developed a random coefficient model to predict the short-term passenger flows with sudden increases sometimes into an urban rail transit station. The hierarchical Bayesian approach is iteratively applied in this work to estimate the new model and the estimation outcomes in each of the iterative calibrations are improved by sequential Bayesian updating. It has been proved that the estimation procedure is able to effectively converge to rational results with satisfying accuracies. In addition, the model application study reveals that besides sufficient preparations in manpower, devices, etc.; the information of the factors affecting the passenger flows into an urban rail transit station should be timely transferred in advance from important buildings, road intersections, squares and so on in neighborhood to this station. In this way, this station is able to cope with the unexpectedly sharp increases of the passenger flows into the station to ensure its operation safety.