Metro Traffic Prediction and Service-Oriented Scheduling Optimization Under Influence of Train Arrivals

Increasing in urban traffic brings heavy burden on metro transportations, especially for metro hubs which are crossing metro stations with railways. It calls for special scheduling optimization for such metro hubs by taking consideration of train arrivals. First, a detailed procedure is proposed on how to predict the traffic in such metro hubs under influence of train arrivals, including metros train and metro stations. Moreover, a service-oriented metro scheduling optimization procedure is proposed. Further analysis is then conducted when there is heavy traffic such as special holidays. A practical case on a major metro hub is studies for validation, the Chongqingbei (Chongqing North) metro hub in the city of Chongqing, China. The Chongqingbei metro hub is with four metro lines (Lines 3/4/10/Loop), three metro stations (Longtousi, Chongqingbei North Square, Chongqingbei South Square), and two non-connected railway stations (Chongqingbei North and Chongqingbei South). Passengers on three metro stations over four lines are calculated. Case study results show that: (1) when there is normal traffic, a hybrid schedule is more effective in comparison with regular shifts for Line 3 while Lines 4/10/Loop require no additional shifts, (2) when there is heavy traffic, an optimized metro schedule with a smaller interval (original 6-minute interval is reduced to 4-minute) is more effective for Line 3 whereas hybrid shifts and Line 10 requires hybrid shifts and Line 4 and the Loop Line require no additional shifts.