【摘要】 将交叉模型交叉模态方法 (CMCM方法)应用到了梁式结构的损伤识别中,重点解决了该方法在实际应用中所面临的两大关键难题:测量噪声下"病态"系统的求解问题和实测模态不完备问题。提出了一种基于TSVD正则化求解的模型修正过程,该过程可利用有限低阶含噪声模态来识别结构的损伤。采用Guyan扩阶法和基于频率的迭代修正法处理了实测模态空间不完备问题,并通过探讨方法的优缺点,提出了一种两阶段损伤识别过程:首先利用Guyan扩阶和基于TSVD正则化求解的模型修正过程进行损伤定位,然后采用基于频率的迭代修正法进行损伤程度的识别。数值计算和试验研究表明:TSVD正则化方法可有效地改善CMCM方法的鲁棒性;当实测模态不完备且包含噪声时,利用两阶段损伤识别过程可准确地识别结构的损伤。更多还原

【Abstract】 The cross-model cross-mode( CMCM) method is applied to identify the damages in beam structures,and 2 key problems in practical application of the method are mainly solved,including the solution problem of "ill conditioned"system with measurement noise and the incomplete problem of measured modes. A TSVD regularization based model updating process is proposed,which can identify the damages in structures using limited low order and noise-polluted modes. The Guyan expansion method and the frequency based iterative updating method are used to deal with the spatial incompleteness of the measured modes.Through discussing their pluses and minuses,a 2-step damage identification procedure is proposed,in which the damages are first located with Guyan expansion and TSVD regularization solving based model updating process,and then the severities of the damages are identified with the frequency based iterative updating method. The numerical calcalation and experimental study demonstrate that( 1) the TSVD regularization method can effectively improve the robustness of the CMCM method;( 2) when the measured modes are incomplete and noise-polluted,the damage in the structure can be identified accurately using the 2-step damage identification procedure.更多还原