混凝土坝裂缝转异诊断的云模型法

Cloud model diagnosis method of concrete dam crack behavior abnormality

针对现有混凝土坝裂缝转异诊断方法多是基于裂缝开度单个测点进行分析这一不足,提出利用多个测点综合信息量诊断裂缝性态的思路.根据这一思路,首先研究了将裂缝开度变化转化为裂缝信息量的方法,其次利用投影寻踪法研究了裂缝开度各测点权重的确定方式,然后将裂缝开度多个单测点的信息量归纳为综合信息量,最后根据云模型理论,建立了裂缝转异诊断的两种判据:基于裂缝信息量总体数据序列的判据,以及考虑裂缝信息量历史数据和当前数据差异性的判据,建立了裂缝转异诊断的云模型法.实例分析表明,裂缝转异诊断的云模型法能够较客观地分析历史数据和当前数据贡献以及对结果的影响,该方法具有可行性.

Because the existing diagnosis methods of concrete dam crack behavior abnormality are mostly based on a single measurement point of crack opening. The idea of diagnosing concrete dam crack behavior by use of comprehensive amount of crack information of multiple measurement points was proposed. First, the method of transforming crack opening variation into amount of crack information was studied according to the idea above. Then the determining method of weight of each crack opening measurement point making use of Projection Pursuit Method was researched. Next, amount of crack information of multiple single measurement point was synthesized into comprehensive amount of crack information. At last, on the basis of cloud model theory, two kinds of crack behavior abnormality diagnosis criterion including criterion of all the data sequence of amount of crack information and criterion of considering the difference between historical data and current data of amount of crack information were established. Accordingly, cloud model diagnosis method of crack behavior abnormality was built. The case study shows that the cloud model diagnosis method can objectively analyze the contribution and impact of historical and current data on the results, and it is reasonable and feasible.