摘要
随着板带轧机的轧制速度和精度不断提高,轧机在生产过程中经常出现电机振动。本文把盲源分离Fast ICA算法应用到轧机振动故障诊断中,对轧机振动的混合信号进行分离,得到分离前后的振动加速度数据曲线以及功率谱,最终诊断出轧机出现振动的主导频率,说明盲源分离法在轧机振动故障诊断中是一种有效的信号诊断方法,为进一步抑制轧机扭振提供了有力依据。
In this paper, the blind source separation Fast ICA algorithm is applied to the mill vibration fault di- agnosis, the vibration mill mixed signal is separated by above method, the vibration acceleration data obtained power spectrum curve after the separation, the final diagnosis of the dominant frequency of vibration mill oe- eurred, descript that blind source separation in fault diagnosis of roiling mill vibration signal is an effective diag- nostic method, and the torsional vibration mill provides a strong basis for the further suppress.
出处
《重型机械》
2014年第6期28-30,共3页
Heavy Machinery
基金
河北省自然科学基金资助项目(E3013402030)
关键词
轧机振动
盲源分离
FAST
ICA算法
故障诊断
主导频率
mill vibration
blind source separation
fast ICA algorithm
fault diagnosis
dominant frequency
