转盘轴承的相似性寿命预测方法研究

Research on Similarity Life Prediction Method of Slewing Bearing's Residual Life

提出了基于相似性的转盘轴承寿命预测方法,首先,在时域内提取服役样本和参照样本的温度、力矩、振动信号的特征值,并对其作主成分分析(PCA),得到各自的温度、力矩和振动PCA值,通过PCA融合形成服役样本和参照样本的综合寿命性能指标。然后,根据归一互相关(NCC)算法计算服役样本最近时间区间内的轨迹与参照样本寿命全程轨迹的相似度,寻找出二者最相似的轨迹区间,以此预测服役样本的剩余寿命。最后,通过试验验证了此方法的有效性。

The life prediction method of slewing bearing＇s residual life based on similarity is proposed. First, the feature values of temperature, torque and vibration signal of the service sample and the reference sample are extracted in the time domain, and the principal component analysis is applied on these feature value, obtaining the PCA value of temperature, torque and vibration, and the life comprehensive performance indicators of the service sample and the reference sample are formed by PCA fusion. Then in order to predict the remaining life of the service sample, the similarity between the service sample track on time interval recently and the reference sample track of the whole life is calculated according normalized cross correlation arithmetic to find out the most similar interval of track with the service track. The effectiveness of the method is demonstrated through the experiment at last.