Fire Evacuation Safety Assessment of University Libraries Based on the Cloud Model

University libraries are critical campus venues characterized by high-density occupancy, making comprehensive fire evacuation safety assessment essential for protecting lives and cultural assets. Conventional evaluation methods often suffer from either subjective bias or a lack of engineering realism, while standalone cloud models lack robust weight allocation mechanisms. To address these deficiencies, this study proposes a fire evacuation safety evaluation method integrating combined weighting and the cloud model. An evaluation framework consisting of 21 indicators is established across four dimensions: human factors, evacuation facilities, emergency management, and venue environment. For weight determination, the Analytic Hierarchy Process (AHP) and the entropy method are synthesized to balance expert experience with data-driven objectivity. The cloud model is then employed to handle the inherent fuzziness and randomness in safety assessment, utilizing the forward and backward cloud generators to map qualitative linguistic descriptions to quantitative values based on three digital characteristics: expectation (Ex), entropy (En), and hyper-entropy (He). A case study of a university library in Wuhan is conducted. The results demonstrate that the library’s overall fire evacuation safety level is "Excellent," with a comprehensive cloud parameter set of (83.577, 0.620, 0.116). The assessment reveals that while evacuation facilities perform optimally, the complex architectural layout represents a structural vulnerability, and personnel safety literacy requires further stabilization. This integrated framework effectively overcomes the limitations of single-weighting strategies and provides a reliable, fine-grained quantitative tool for fire safety management and targeted risk mitigation in high-density public buildings.