基于自组织映射网络的齿轮故障分类识别  

Classification of gear fault and their identification based on self-organizing maps

作　　者：廖广兰[1] 史铁林[1] 姜南[1] 

机构地区：[1]华中科技大学机械科学与工程学院

出　　处：《华中科技大学学报（自然科学版）》2003年第7期69-71,共3页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家重大基础研究基金资助项目(G19980 2 0 3 2 0 );湖北省自然科学基金资助项目 (2 0 0 0J12 5 )

摘　　要：分析了自组织映射网络算法 ,在U 矩阵方法基础上提出了一种改进的可视化训练结果方法 ,并应用于齿轮故障的模式识别 .研究表明 ,自组织映射网络能对齿轮状态进行正确分类 ,有效识别故障模式 ,改进的可视化方法能更加清楚地显示网络训练结果 。The algorithm of Self-Organizing Maps (SOM) was intr od uced, and an improved technique based on the U-matrix method for visualizing th e trained SOM results was presented. The application of SOM in the gear fault cl assification and the proposed method for visualization were studied. The experim ent results show that SOM is effective in the gear fault classification and iden tification, and the proposed method can visualize the results more clearly than the U-matrix method.

关 键 词：齿轮故障 自组织映射 U-矩阵法 分类识别 

分 类 号：TH17[机械工程—机械制造及自动化] TP18[自动化与计算机技术—控制理论与控制工程]