【作者】 张建宇； 宫兆盛； 黄永海；

【Author】 ZHANG Jian-yu~1 GONG Zhao-sheng~2 HUANG Yong-hai~1 (1.Key Laboratory of Advanced Technology,Beijing University of Technology,Beijing 100124; 2.Gudao Oil Production Plant,Shengli Oil Field,Sinopec,Dongying Shandong 257231)

【机构】 北京工业大学 北京市先进制造技术重点实验室； 中石化胜利油田孤岛采油厂；

【摘要】 相对于单一故障诊断,混合故障的特征分离历来是诊断领域的一大难题。针对某高线轧机的一次多故障并发的振动信号,本文引入了独立分量分析(ICA)技术,以负熵最大化为分离判据,以牛顿迭代法为优化算法,实现了滚动轴承故障特征的有效分离和提取。更多还原

【Abstract】 Being compared with single fault,the feature separation of compound fault in mechanical system is more challenging in fault diagnosis field.Aiming at the vibration signal collected on a high speed wire rolling mill with multiple faults,the independent component analysis was introduced in this paper. Based on the negative entropy maximization and Newton iterative methods,the hybrid fault feature of rolling bearings was separated and identified successfully.更多还原