神经网络在转子轴承系统固有频率识别中的应用  

The Application of Neural Network in Natural Frequency Identification of Rotor Bearing System

作　　者：赵虎城[1] 钱坤[1] 周正平 

机构地区：[1]西京学院机电系,西安市710123 [2]成都轴流风机厂

出　　处：《风机技术》2009年第3期45-48,共4页Chinese Journal of Turbomachinery

基　　金：国家自然科学基金资助(50276070);陕西省自然科学基金资助项目(2005A06)

摘　　要：通过神经网络建立了滚动轴承转子系统固有频率与各种参数之间的映射关系,有了确定的映射关系,降低了对于同结构转子的固有频率的计算工作量,最后通过仿真算例和试验验证了本方法的优越性和可行性。In this paper, the mapping relationship between natural frequency in rolling bearing rotor system and every parameter is established based on neural network. Thus the working load of calculating natural frequency of rotor with same structure is lowered based on the determined mapping relationship. The superiority and feasibility of this method are verified through simulation example and test at last.

关 键 词：径向基神经网络 有限元 模型修正 

分 类 号：TH113.1[机械工程—机械设计及理论]