As railway transportation systems become increasingly complex, the demand for cross-system data fusion continues to grow. This paper proposes a cross-system dynamic data fusion method based on train operation trajectories, aimed at enhancing the integration and analytical capabilities of data within the railway industry. First, we provide an overview of the current state of research on data fusion, highlighting the limitations of existing methods and laying the groundwork for this study. Next, we clarify the objectives of data fusion and the scope of the data involved, focusing on the core elements of spatiotemporal fusion based on train operation trajectories. Building on this foundation, we construct a knowledge graph centered on these core fusion elements to effectively link and organize data from various systems. Through this method, we generate a series of data fusion results that offer deeper business insights. Additionally, the practical effectiveness and value of this method are demonstrated through specific case studies. Finally, we summarize the main findings of the research and emphasize the potential of this method in promoting data-driven decision-making in the railway industry, providing a reference for future research and practice.