摘要
根据旋转机械碰摩故障的特征频率,利用小波变换多分辨率特性,提取碰摩故障特征信号;利用支持向量机原理解决小样本、非线性、高维模式识别问题的优势,建立支持向量机故障诊断模型,进行旋转机械碰摩故障诊断。仿真结果表明,该算法有较好的精度和泛化性,并且运算量小,非常适合旋转机械故障诊断的在线识别。
According to the characteristic frequency of rotating machinery rub - impact fault, the acquisition of fault charaeteristie signal can be obtained by multi - resolution characteristics of wavelet transformation. Further more, as the advantages of support vector machine in sol - ving small samples, non - linearity and higher - dimension pattem recognition problem, fault diagnosis model with support vector machine is established. The simulation results prove that this method is accurate and can be spread, suitable for on- line identification.
出处
《工业安全与环保》
北大核心
2010年第3期1-3,14,共4页
Industrial Safety and Environmental Protection
基金
国家自然基金资助项目(50475169
50535010)
关键词
声发射
碰摩
小波分解
故障诊断
支持向量机
AE rub- impact wavelet decomposition fault diagnosis support vector machine（SVM）
