中文摘要：

对某实际水电站机组及其支承结构实测振动数据分析基础上,针对大型水电站机组连续运行,无法施加有效激振,油膜只在机组运行时存在等特点,基于多信号分类法,将模态参数时域识别引入到水电站机组轴系统的自振特性研究中;同时针对传统油膜动态特性参数识别方法的不足,提出了可以不受测点少,工况少等实际情况限制的遗传识别方法,对机组导轴承油膜动态特性参数进行了识别,为机组轴系统自振特性及振动响应的研究提供了基础。根据油膜动态特性识别结果进行了轴系统自振特性的有限元数值反馈计算,验证了两种识别方法及结果的正确性。

英文摘要：

Prototype tests were conducted for a generator set and the results were evaluated.Since the operation of this large generator set is continuous and shaft oil-film exists only when the unit is operating,application of an effective excitation to the shaft system is difficult.In this work,a time-domain identification method of modal parameters was adopted to study the shaft self-vibration features.This is a multiple signal classification method for the identification,while the traditional method has limitations in many cases such as few test cases or few test points.In the new method,the dynamic parameters of shaft oil-film are used for identification.The identification results were verified by FEM numerical simulation,and the method offers an effective tool for vibration response analysis.