Dynamic structural operational safety assessment and decision support for concrete arch dams via multi-source monitoring data fusion and interval-valued intuitionistic fuzzy reasoning

Structural operational safety assessment of high concrete arch dams remains challenging because the importance of monitoring indicators varies with operating conditions, while the evaluation process is affected by multiple sources of uncertainty. To address these issues, this study develops a dynamic data-driven evaluation framework for structural operational safety assessment. Initially, a hierarchical indicator system and a refined grading scheme are established using multi-source structural response sequences. An adaptive weighting mechanism is then developed by coupling the cross-correlation function with Apriori-based association rule mining, enabling indicator weights to be dynamically updated according to the correlation and association characteristics between reservoir water-level variations and structural responses. Furthermore, Interval-Valued Intuitionistic Fuzzy Sets (IVIFSs) are introduced to explicitly represent membership, non-membership and hesitation information, allowing randomness, fuzziness, and expert cognitive uncertainty to be jointly characterized within a unified reasoning framework. The proposed method is validated using monitoring data from the Jinping-I arch dam. Comparative results demonstrate that it provides more stable and discriminative safety-state assessment than conventional static-weight and fuzzy evaluation approaches, while retaining greater adaptability to time-varying structural behavior. The framework further transforms heterogeneous monitoring information into interpretable safety states and spatially explicit diagnostic information, providing a practical basis for real-time safety assessment, targeted inspection and intelligent operation and maintenance of high arch dams.