基于极限学习机的多传感器旋转机械故障诊断

为了更好的对旋转机械故障进行识别与分类,文章提出了一类基于极限学习机的多传感器融合故障识别方法.首先,利用FFT对数据进行预处理,并对多传感器的预处理结果进行加权融合,以单传感器历史数据识别得到的正确率为融合系数.然后,对极限学习机进行训练和测试.结果表明基于融合数据特征的识别率表现优于基于单传感器数据特征的识别率.

In order to better identify and classify the faults of rotating machinery, a kind of multi-sensor fusion fault recognition method based on limit learning machine is proposed in this paper. First of all, the FFT is used to preprocess the data, and the weighted fusion of the preprocessing results of multi-sensor is carried out, and the correct rate of historical data recognition of single sensor is taken as the fusion coefficient. Then, the extreme learning machine is trained and tested. The results show that the recognition rate based on fusion data feature is better than that based on single sensor data feature.