多点平稳随机载荷识别方法研究  

Identification method of multi-point stationary random load

作　　者：姜金辉[1] 陈国平[1] 张方[1] 

机构地区：[1]南京航空航天大学振动工程研究所,江苏南京210016

出　　处：《振动工程学报》2009年第2期162-167,共6页Journal of Vibration Engineering

摘　　要：根据矩阵谱分解的思想,提出多点任意相关的随机载荷识别方法。鉴于响应功率谱矩阵和激励功率谱矩阵具有相同的秩且为非负定,首先推导完全相干的功率谱矩阵的识别方法,将任意相干的激励功率谱矩阵进行谱分解成完全相干的功率谱矩阵之和,利用响应信息识别分解后的完全相干功率谱进而完成对激励功率谱的识别。该方法具有较高的识别精度。针对求解逆问题中的适定性进行了讨论,指出病态的原因并运用条件数权重法,该法能在一定程度上减轻病态,提高识别精度。通过实验和仿真验证上述方法的正确性。最后对提高识别精度提出了建议。The identification of multi-point arbitrary correlated stationary random load is presented in the paper in the light of the method of matrices spectrum decomposition. Based on the same rank and non negative matrix between response power spectrum matrix and stimulation power spectrum, the identification of the complete-correlated power spectrum matrices are deduced firstly, then the arbitrary correlated power spectrum matrix of stimulation is decomposed into the complete-correlated power spectrum matrices which can be identified by the response information in order to fulfill the identification of the power spectrum of stimulation. The method is of higher precision. And the paper analyzes the wellposedness of the inverse problem, presents the reason for ill-pose problem, namely, the ill-condition of matrix of frequency response function, utilizes the way of weighted condition number which can ease ill-posed problem and improve identification accuracy to some degree. The correctness of the way above is demonstrated by experiments and simulation. Finally some advice is given for improving the identification precision.

关 键 词：矩阵谱分解 非负定 随机载荷识别 适定性 病态 

分 类 号：TU311.3[建筑科学—结构工程] TU352.1