基于状态特征向量的装配式建筑钢结构损伤识别方法

Damage Identification Method for Fabricated Building Steel Structures Based on State Feature Vectors

为提高装配式建筑钢结构损伤识别精度,测定准确的损伤裂缝长度及相关数据,引入状态特征向量,开展对装配式建筑钢结构损伤识别方法的设计研究。利用状态特征向量,提取建筑钢结构参数,结合支持向量机,完成钢结构损伤磁记忆特征检测。基于检测结果,实现对装配式建筑钢结构损伤的等级判定。将新的识别方法与现有的两种识别方法应用于相同实验环境对比可知,新的识别方法可以得到更准确、精度更高的识别结果,查明所有的损伤问题,且识别误差不超过±0.1 mm。将新的识别方法应用于实际,能够进一步提高钢结构的施工质量,同时可以通过识别结果为钢结构性能优化和维护策略的提出提供重要依据。

In order to improve the accuracy of damage identification for fabricated building steel structures,measure accurate damage crack lengths and related data,introduce state feature vectors and carry out design research on damage identification methods for fabricated building steel structures.Extracting structural parameters of building steel using state feature vectors;Combined with support vector machine,the magnetic memory feature detection of steel structure damage is completed.Based on the detection results,it is possible to determine the damage level of fabricated building steel structures.By applying the new identification method and the other two existing identification methods to the same experimental environment in contrast,the new identification method can obtain more accurate and accurate identification results,achieving full identification of all damage issues,with an identification error of no more than±0.1 mm.Applying the new identification method to practice can further promote the improvement of the construction quality of steel structures,and the obtained identification results can provide an important basis for the performance optimization and maintenance strategy of steel structures.