Dynamic Modeling for Metro Passenger Flows on Congested Transfer Routes

With the rapid development of urbanization, metro becomes more and important for people’s travel in big cities. To quantitatively describe passenger flows on congested transfer routes, this paper introduces a dynamic model based automated data from automatic fare collection (AFC) vehicle location (AVL) systems. An expectation maximization (EM) algorithm is proposed to compute maximum likelihood estimates unknown parameters our model. Our can yield systematic analysis one-transfer both population individual aspects. Important characteristics, including time, boarding probabilities, walking passenger-to-train assignment total be inferred using only AFC AVL data. We provide case study Beijing metro. Detailed results are given. also present cross-validation method validate with real