基于三阶累积量奇异值分解的结构损伤识别

Structural Damage Identification Based on Singular Value Decomposition of Third-Order Cumulants

为了在钢筋混凝土结构损伤识别过程中有效地考虑微裂缝引起的非线性特征,提出了一种基于奇异值分解的三阶累积量方法,通过对结构损伤过程中振动加速度信号的三阶累积量进行奇异值分解,根据主奇异值的变化来判断混凝土结构的损伤程度和定性评价结构性状。为了验证该方法,对两根钢筋混凝土简支梁进行了损伤试验,试验数据分析结果与试验中裂缝开展的观察结果比较吻合,表明三阶累积量的最大主奇异值能够有效地反映钢筋混凝土梁在各级荷载作用下结构损伤的变化情况,说明该方法是令人满意和有效的。

A method of third-order cumulants(TOC) based on singular value decomposition(SVD) is presented to effectively consider the nonlinear characteristic caused by micro-cracks in damage identification of reinforced concrete structures.Firstly,The singular values of the TOC of vibration acceleration signals are calculated,then the singular values of TOC are applied to describe and assess the severity of structure damage and structure state.The efficiency of the method is verified by an experimental test of two reinforced concrete simply supported beams which are gradually damaged.The results clearly show that the change trend of singular values of TOC are identical to the cracks propagation of experiment records,the singular values of TOC can effectively indicate the damage degree of beams,the results are satisfying and indicate that the method is effective.