Abstract: Taking the Beijing Sihui Hub as the research object, this paper explored a data-driven evaluation and optimization design method of the hub dynamic guidance signage system to meet passenger needs. First, the KANO passenger demand model was built, and through the desktop experiment, the optimal design scheme of the dynamic guidance sign system in terms of content, style and spatial location was formed, which is in contrast with the existing system of the Sihui Hub. Second, based on the wayfinding theory, building information modeling (BIM) + virtual reality (VR) simulation technology was used to realize the information interaction between people and the hub, and extract the characteristic parameters of the passenger wayfinding process under the action of the old and new guidance sign systems. Finally, through the analysis of subjective and objective indicators in the pathfinding experiment, it can be seen that the pathfinding time, number of lost points and number of errors of the subjects in the new version of the dynamic guidance sign system are significantly reduced, and the satisfaction of the new version of the system is better than that of the old version in terms of content, style and spatial location. This paper built a BIM + VR virtual simulation platform and formed a data-driven oriented comprehensive evaluation and optimization design method for the hub dynamic guidance signage system, providing technical and theoretical support for the design and reasonable application of the hub dynamic guidance signage system.