As a fundamental component of urban underground space, the metro system functions both as critical urban infrastructure and public space within a complex transportation network. This study examines the interchange efficiency of transfer metro stations in Xi'an, with a focus on Xibeigongyedaxue Station (X) and Jianzhukejidaxue·Lijiacun Station (J). Field research was conducted to collect environmental data, supporting an in-depth analysis. Utilizing AnyLogic as the primary simulation tool, highly accurate models were developed to achieve two key objectives: (1) to simulate the dynamic passenger flow variations during peak hours comprehensively, and (2) to analyze the multi-dimensional factors influencing interchange efficiency, including transfer processes, facility layouts, and passenger behaviors. Comparative results suggest that J station's direct transfer route and compact passageways contribute to concentrated passenger flow, potentially leading to frequent congestion. In contrast, X station's expansive yet complex layout results in unclear pathways and navigational inefficiencies. Based on the findings, variations in passenger behavior and external conditions influence transfer efficiency in ways not fully captured by the simulation. Nonetheless, optimization strategies are evaluated to enhance the design of underground transit nodes, considering how spatial configurations and facility arrangements impact passenger experience. The research provides case study with scenario-based simulations that contributes to an efficient and sustainable operation of urban metro systems.