A Simulation-Based Study on Flexible Seat Capacity Allocation Policy for Railway Systems

Journal : Journal of the Korea Society for Simulation
Title : A Simulation-Based Study on Flexible Seat Capacity Allocation Policy for Railway Systems
Authors : Sehee Choi, Jaeyeon Jeon, Bosung Kim, Soondo Hong
Abstract : The railway seat capacity allocation problem suffers from mismatches between forecasted and realized demand, resulting in lost sales and residual seats. Conventional fixed-allocation policies assign seat capacity to each origin–destination (OD) pair without exceeding total train capacity, failing to respond effectively to demand uncertainty. To address this limitation, this study proposes a flexible seat capacity allocation policy for railway systems based on the concept of safety stock. Under the proposed policy, seat allocations for individual OD pairs can exceed physical train capacity in advance to accommodate demand fluctuations. However, the number of confirmed bookings is always constrained by the actual seat capacity of the train, ensuring physical capacity is never exceeded. The policy is optimized using a discrete-event simulation model integrated with a genetic algorithm (GA). Experimental results show that the flexible policy performs best under high demand variability. Compared with the conventional fixed-allocation policy, total revenue increased by 2.08%, while lost sales and residual seats decreased by 13.79% and 18.39%, respectively. These findings indicate that the proposed policy can be effective for railway seat capacity allocation under high demand variability.
Leave a comment