基于信息融合的滚动轴承寿命状态识别研究

针对滚动轴承寿命状态识别过程中,单一传感器蕴含的信息不能全面反映寿命状态的问题,文章提出了一种基于信息融合的滚动轴承寿命状态识别方法。该方法首先采用多路卷积层提取不同传感器的数据特征信息,克服单一信息源的局限性;然后采用多层卷积、池化交替级联的方式,实现多源信息的特征值深度融合,最后采用全连接和多分类函数,实现动轴承的寿命状态识别。通过不同方法的对比实验,结果表明了所提方法能够提高滚动轴承寿命状态识别率,具有较好的可行性。

Study on Life State Identification of Bolling Bearing Based on Information Fusion

In view of the problem that the information contained in a single sensor cannot fully reflect the life state of rolling bearing in the process of life state identification,a method of life state identification based on information fusion is proposed in this paper.Firstly,the method uses the multi-channel convolutional layer to extract the data characteristic information of different sensors to overcome the limitation of a single information source.Then,the multi-layer convolution and pooling alternate cascade are adopted to realize the deep fusion of the characteristic values of multi-source information.Finally,the full-connection and multi-classification functions are adopted to realize the life status recognition of the dynamic bearing.Through the comparison experiment of different methods,the results show that the proposed method can improve the identification rate of rolling bearing life state and has good feasibility.