灰色极限学习机在滚动轴承故障预测中的应用

Rolling Element Bearing Fault Prediction Based on Grey Sequential Extreme Learning Machine

针对较强噪声环境下的滚动轴承故障预测问题,为提高轴承故障预测的精度,提出并研究了一种新的滚动轴承预测技术;采用将灰色模型和极限学习机(ELM)相结合的方法,针对轴承运行状态值的非线性特点,先将样本数据进行灰色处理,解决数据的随机性和波动性问题,然后代入学习速度快,泛化精度高的ELM神经网络进行训练;在训练完毕后,对未来的轴承运行状态数据进行分析,将其与轴承设备的理论诊断标准相比较以达到故障预测的目的。

Aiming at the prediction of rolling element bearing fault in the strong noise environment, a novel method of prediction for roiling element bearing is proposed to improve the bearing fault prediction accuracy. This paper presents a kind of new rolling bearing prediction technology, using grey model combined with the extreme learning machine （ELM）. The sample is first grey processed to solve the randomness and volatility, and then introduced into the extremely fast learning speed and high generalization accuracy of ELM neural network training. Based on the trained model, the bearing operation state of future time points is analyzed, and the result is compared with the theoretical diagnosis standard of the bearing equipment to realize the fault prediction.