Abstract: Current evaluation process for bridge technical condition included the problem of uncertainty and the neglect of the influence of the main structural components on the overall safety of the bridge, so anomaly recognition was introduced into evaluation process. The cloud model theory specialized in uncertainties was combined with the traditional fuzzy comprehensive analysis method, and an evaluation method of bridge technical condition based on cloud model and anomaly recognition was proposed. Firstly, the weight of each component and structure was determined, then the remark set and evaluation data were transformed into cloud model, and the evaluation results before and after introducing anomaly recognition in the evaluation process were acquired. Finally, the proposed model was applied to evaluate an actual bridge. The results show that anomaly recognition is introduced in the traditional cloud model evaluation to consider the impact of main structural components on the overall safety of the bridge. The evaluation method of the proposed model is more in line with the actual engineering situation, the process is more objective and the results are more comprehensive.

Key words: bridge engineering, bridge technical condition, cloud model, anomaly recognition, uncertainty

摘要： 针对现行桥梁技术状态评定过程中存在的不确定性以及忽略了主要结构部件对桥梁整体安全使用的影响，提出在评定过程中引入异常识别，并将专门研究不确定性问题的云模型理论与传统的模糊综合分析法相结合，提出一种基于云模型与异常识别的桥梁技术状态评定方法；首先确定各个部件以及结构的权重，然后将评语集、评估数据转换为云模型，并获取在评定过程中引入异常识别前、后的评定结果；最后将该模型应用于某实桥评定中。研究表明：在传统云模型评定中引入异常识别考虑了主要结构部件对桥梁整体安全使用的影响，该模型评价方式更为符合实际工程情况，过程更为客观，结果更为全面。

关键词: 桥梁工程, 桥梁技术状态, 云模型, 异常识别, 不确定性