A Study on Underground Garage Traffic Simulation and Optimization of Flow Lines

Large-scale multi-level underground garages often experience high traffic volumes and complex traffic patterns. imposing high demands on the design and optimization of internal flowlines. To improve vehicle circulation efficiency and reduce exit delays, we propose a flowline optimization method based on traffic simulation. The road network of the garage is abstracted into an origin-destination (OD) model, with parking spaces and exits designated as origins and destinations, respectively. Time-dependent OD demand matrices are constructed and implemented within a dynamic system optimal (DSO) traffic assignment framework to simulate vehicle flow under different flowline configurations. Flowline design is evaluated based on dynamic traffic loading. Priority is given to organizing inter-level connections, particularly the directional use of ramps. Traffic conflict points near exits are reduced following the principle of minimizing interference. Road segments meeting criteria for two-way traffic are reconfigured to support bidirectional flow, enhancing overall network capacity. A case study was conducted on a two-level underground garage in a Beijing residential complex. A microscopic traffic simulation platform is developed using Simulation of Urban Mobility (SUMO), incorporating time-varying stochastic origin-destination demands generated from empirical data. Simulation results show that the proposed method significantly reduces total travel time, exit queue lengths, and congestion in critical sections. The robustness of the optimized flowline scheme is also validated under emergency conditions, including sudden changes in exit availability and unexpected increases in outbound demand. The results confirm that the proposed approach improves traffic efficiency and system resilience within underground garages.