Abstract: Abstracts: Aiming at the shortcomings of traditional pedestrian evacuation models in depicting individual behavioral differences and group emotion transmission mechanisms, the efficiency of pedestrian emergency evacuation was studied against the background of sudden public safety incidents in subway stations. Based on the cellular automaton model, environmental familiarity was introduced to reflect individual differences in pedestrian evacuation behaviors, and an escalator speed factor was incorporated to enhance the model’s applicability to real scenarios; simultaneously, combined with the SIR (susceptible-infective-removed) infectious disease model, the dynamic spread of panic emotion within the crowd was simulated, thereby constructing an evacuation simulation model that integrated the effects of environmental familiarity and panic emotion. Taking the platform of Nanjing Metro Station A as an example, a simulation was conducted to analyze the respective mechanisms of the two aforementioned factors and their interactive effects on the overall evacuation process. The results indicate that the evacuation process is significantly slowed down by the spread of panic, with the total evacuation time increasing by approximately 35%. Environmental familiarity is observed to have a significant positive impact on evacuation efficiency, as the evacuation time of highly familiar individuals is reduced by an average of 15% to 40%. Furthermore, as the panic threshold increases, the evacuation time generally shows a decreasing trend, stabilizing after reaching a certain critical value. By integrating cognitive familiarity and psychological emotional factors, this study addresses the insufficient characterization of individual behavioral heterogeneity in traditional models, providing a theoretical basis for the refined development of emergency evacuation plans in subway stations.