Weekly line planning to improve the match between rail supply and passenger demand

Traditional railway line planning approaches are used to determine the best train lines during specified periods (e.g., 1 day, 1 h, or several hours). However, mismatches occur relative to the spatiotemporal changes in travel demands over longer time periods (e.g., 1 week). To more effectively match rail supply with passenger demand, a weekly line planning (WLP) model considering both peak and nonpeak passenger demand over an entire week and during all periods of each day is proposed. This model maximizes the defined supply-demand matching utility and minimizes train operational cost. Important constraints to ensure proper daily and hourly train repetition over multiple days are included. A customized genetic algorithm (GA) is developed, which outperforms the CPLEX solver in real-world cases of high-speed railways in China. Compared with practical and single-day planning, WLP shows clear improvement in terms of operational cost, passenger travel speed, and supply-demand matching.