A distributionally robust optimization method for the coordination of train and passenger flows in regional rail transit system considering train skip-stop strategy

Aiming at the application risks caused by the performance degradation of the existing deterministic optimization methods under the condition of passenger flow uncertainty, a distributionally robust optimization method for the coordination of train and passenger flows in regional rail transit system, incorporating train skip-stop strategy, is proposed. Initially, suitable skip-stop lines and patterns are selected based on the passenger flow throughput, leading to the construction of a deterministic optimization model that considers the coordination of train and passenger flows with skip-stop operation. A distributionally robust optimization approach is then employed, utilizing ambiguity sets to describe the uncertainty of passenger demand, resulting in the development of a distributionally robust optimization model. This model is further refined with prior holiday information to adjust the ambiguity sets. The optimization model is solved using the genetic algorithm, and validation is conducted using desensitized real data from the Chongqing regional rail transit system. Experimental results indicate that the skip-stop strategy can further reduce the global transport capacity risk of the system. Compared to the deterministic model, the distributionally robust optimization model exhibits superior performance under conditions of uncertainty, and the optimization performance is further enhanced by the introduction of prior information.