Performance degradation evaluation of harmonic reducer based on damage sources-outliers accumulation
Taking LSHG-20-50-C-I harmonic reducer as the research object and acoustic emission as the monitoring means,the damage source-outliers accumulation algorithm is proposed to complete the performance degradation evaluation of harmonic reducer.Firstly,the AE signals were analyzed,and the wear and crack types of AE sources in the process of performance degradation were identified by the experience graph analysis of ringing count.The accuracy of damage source classification was verified by comparing the results of accelerated life experiment.On this basis,a crack growth model was established.According to the crack growth law and the changing trend of the average energy and average level,the performance degradation evaluation based on the damage source was carried out,and the four main stages of the crack growth in the performance degradation process were distinguished.To solve the problem of poor trend and monotonicity of single AE parameters,a performance degradation evaluation method based on PCA outliers accumulative index was proposed.Compared with the conventional One-Class SVM algorithm,PCA algorithm was more sensitive to early outliers and the threshold alarm time was 650 h earlier,which was more suitable for the detection of outliers and performance degradation evaluation of harmonic speed reducers.
NETL
NSTL
万方数据
