【摘要】 液压提升塔架广泛应用于大型工程吊装领域，其运行状态直接关系到工程质量与作业安全。当前，受复杂工况与多重危险源耦合作用的影响，液压提升塔架的安全评估方法仍存在不足。为此，提出了一种融合多种赋权策略与不确定性推理机制的多指标安全评估方法，以更加全面、稳健地量化设备的安全水平。针对传统赋权与推理过程中易受不确定性因素影响的问题，提出区间数博弈组合赋权方法，结合三标度层次分析法与基于层间相关性的客观赋权法，实现主客观权重信息的稳健融合。同时，基于浮动容忍度思想构建双云模型推理机制，通过上下界隶属度区间建模，显式表征数据波动对推理结果的影响。在此基础上，构建了涵盖结构与机构双层面的多层次安全评价体系，结合模糊综合评价法，对液压提升塔架的安全性能进行综合判定。将所提方法应用于某大型造船门式起重机吊装工程中的液压提升塔架实例，结果表明，该方法能够有效识别潜在风险，为塔架安全监测、风险预警与运维决策提供了有力支撑。更多还原

【Abstract】 Hydraulic lifting towers are widely used in large-scale engineering hoisting, with their operational status directly impacting project quality and safety. However, the coupling of complex working conditions and multiple hazards leads to limitations in current safety assessment methods. This paper proposes a multi-index safety assessment approach integrating diverse weighting strategies and uncertainty reasoning to achieve a more comprehensive and robust evaluation of equipment safety. To address the susceptibility of traditional weighting and inference processes to uncertainty, an interval-valued game-theoretic combination weighting method is developed, which robustly fuses subjective weights derived from a three-scale analytic hierarchy process(AHP) and objective weights based on inter-criteria correlation(CRITIC). Meanwhile, a dual-cloud inference mechanism is constructed based on the concept of tolerance intervals, enabling membership degree modeling in the upper and lower bounds to explicitly capture the influence of data fluctuations on inference results. On this basis, a hierarchical safety assessment framework covering both the structural and mechanical aspects is established. Fuzzy comprehensive evaluation is then used to integrate heterogeneous data and perform the final safety classification. The proposed method is applied to a hydraulic lifting tower in a large-scale shipyard gantry crane hoisting project. The results demonstrate that the method can effectively identify potential risks and provides strong support for safety monitoring, risk warning and maintenance decision-making.更多还原