摘要
针对轴承故障检测中用一般的谱分析法难以实现故障的精确判定的问题,提出一种新的轴承故障检测方法。考虑到复倒谱法对周期性异常振动特征提取的有效性,利用频谱分析结合复倒谱分析对采集的轴承振动信号进行处理,先通过频域分析获取高、中、低频带的均值,再通过复倒谱方法获取异常振动产生的周期性激励信号提取振动信号特征参数,分离出噪声中的“异音”信号,并结合模糊算法,初步实现了轴承故障的智能定位。实验结果表明,该方法是有效的。在积累足够量的样本数后,可望建立相应的专家库,实现轴承故障的快速智能诊断。
It is usually difficult to detect the bearing faults accurately by using general spectrum analysis. To improve the detection accuracy, a new detection method was proposed. Due to the validity of the complex cepstrum analysis in characteristic extraction of the periodic abnormal vibration, the method combined the spectrum analysis with the complex cepstrum analysis to extract the characteristic parameters of the bearing vibration. It could be realized like this: the average magnitudes of the high, medium and low frequency vibration waves were extracted respectively by the spectrum analysis. And then the complex cepstrum analysis was used to extract its characteristic parameters. Finally, treating by the fuzzy algorithm, the specific noise was distinguished from the noise spectrum and the fault origins could be detected. The experimental results were consistent with the expectations well. Together with expert knowledge, intelligent bearing fault detection will be expected.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2005年第5期515-517,521,共4页
Chinese Journal of Scientific Instrument
基金
浙江省科技厅资助项目。
