摘要
提出了一个考虑模态噪声影响的基于残余力向量法的损伤识别两步法。首先从理论上证明了由连续观测所得的模态参数计算的残余力向量样本,其均值向量可以作为损伤位置识别的标识量。在可能损伤位置确定后,通过提取刚度矩阵和质量矩阵中与损伤相关的自由度来减少计算量,用矩阵缩减后得到的残余力向量构造遗传算法优化目标函数,来识别损伤程度。数值算例证明,该方法具有较好定位和定量识别效果和抗噪声能力。
A new two-step damage identification method based on residual force vector considering the impact of measurement error （noise） in most cases is proposed. The idea that the mean vector of the residual-force-vector sample that obtained from continuous observing on structural modal testing data still can be viewed as the flag and indicator of damage localization is proposed and theoretically proved. To reduce the calculation burden effectively, a method based on the matrix condensation approach is presented to identify the severity of damage. Matrix condensation is applied to obtain the degrees-of-freedom of the stiffness matrix and mass matrix associated with the damaged elements. An objective function for GA is formulated with residual force vector calculated by matrices condensed. Then, the damage quantitative determination is performed by GA. Numerical experiments verify the effectiveness of identification and the anti-noise ability of the proposed method.
出处
《广西工学院学报》
CAS
2008年第4期1-4,17,共5页
Journal of Guangxi University of Technology
关键词
损伤识别
噪声
残余力向量法
遗传算法
两步法
damage identification
noise
residual force vector
genetic algorithm
two-step method
