摘要
滚动轴承作为大多数旋转机械的重要零部件,其工作状态直接影响设备的工作寿命。针对传统故障诊断方法自适应性差、特征提取过于依赖人工经验的问题,提出一种基于双向门控循环单元的轴承故障诊断方法。该方法直接将原始振动信号作为模型输入,自动进行故障特征提取与故障诊断。结合轴承外圈故障、内圈故障及滚动体故障等9种故障状况,对所提方法进行了验证。实验结果表明,其故障诊断准确率可达99.56%,诊断效果优于门控循环单元、长短期记忆网络等算法,且泛化能力好。
Rolling bearing is the key part of rotating machinery,and its working state directly affects theworking life of rotating machinery.To solve the problems that the traditional fault diagnosis have poor adaptability and feature extraction depends too much on manual experience,a bearing fault diagnosis was proposed based on the bi-directional gated recurrent unit,which used the original vibration signal as the model input signal,extracted the bearing fault signal automatically and made fault diagnosis.The proposed model was used to identify nine fault states of bearing outer ring fault,inner ring fault and rolling element fault.The experimental results show that the fault diagnosis accuracy of this method can reach 99.56%,and its generalization ability are better than the results of the gated recurrent unit and long short-term memory algorithm.
作者
石静雯
侯立群
SHI Jingwen;HOU Liqun(Department of Automation,North China Electric Power University,Baoding 071003,China)
出处
《电力科学与工程》
2021年第10期64-70,共7页
Electric Power Science and Engineering
基金
河北省自然科学基金(F2016502104)。
关键词
故障诊断
双向门控循环单元
深度学习
滚动轴承
fault diagnosis
bidirectional gated recurrent unit
deep learning
rolling bearing
