论文部分内容阅读
由于正交多项式算法在频响函数模态参数拟合时,模态数目和多项式分子阶次较难确定,无法对拟合过程进行干涉。在原有算法的基础上,通过结合稳定图和模态指示函数,确定了频响函数的模态数目和极点(阻尼及固有频率);在已知极点的基础上,由线性最小二乘方法对留数进行估计。此外,该方法被扩展到频响函数的整体拟合。通过将该方法应用于模态耦合严重的频响函数模态参数估计,并与原有正交多项式算法及商业化软件M E’scopeVES进行对比,表明该算法在避免虚假模态或发生模态遗漏现象的同时,拟合结果准确、稳定,抗干扰能力强。
Since orthogonal polynomial algorithm fits modal parameters of frequency response function, it is difficult to determine the number of modes and the order of polynomial molecules, so that the fitting process can not be interfered. Based on the original algorithm, the modal number and pole (damping and natural frequency) of the frequency response function are determined by combining the stability diagram and the modal indication function. On the basis of the known poles, the linear least square method The number of remaining estimates. In addition, the method is extended to the overall fitting of the frequency response function. By applying this method to modal parameter estimation of frequency response function with severe modal coupling and comparing it with the original orthogonal polynomial algorithm and commercial software M E’scopeVES, it is shown that the proposed method avoids the false modal or modal The omission phenomenon at the same time, the fitting result is accurate, stable, anti-interference ability.