摘要
针对旋转机械中滚动轴承退化趋势预测存在健康指标构建依赖先验知识、预测精度低等问题,提出了基于卷积自编码器(convolutional auto-encodes,CAE)和融合注意力机制的门控循环单元(attention gated recurrent unit,AGRU)的滚动轴承退化趋势预测方法。首先,该方法通过快速傅里叶变换(fast Fourier transform,FFT)将滚动轴承时域信号转换为频域信号,卷积自编码器从频域信号中自适应提取特征,编码特征通过评估选择构建健康指标(health indicators,HI),在此基础上,将健康指标输入融入注意力的门控循环单元网络(gate recurrent unit,GRU)模型,剪枝算法对模型参数进行优化,完成了滚动轴承性能退化趋势预测。结果表明,所提的方法能获得更准确的滚动轴承退化趋势预测。
Aiming at problems of health indictor construction depending on prior knowledge and prediction accuracy being low for rolling bearing performance degradation trend prediction methods in rotating machinery,a prediction method for rolling bearing degradation trend based on convolutional auto-encodes(CAE)and attention gated recurrent unit(AGRU)was proposed.Firstly,the method converted the rolling bearing time domain signal into frequency domain signal with fast Fourier transform(FFT),and the features were extracted adaptively from frequency domain signal with convolutional auto-encodes.Then,the health indicators were constructed from encoding features.Finally,the health indicators were input into the attention gated recurrent unit mode,and the pruning algorithm were used to optimize the parameters to predict the performance degradation trend of rolling bearings.Results show that the proposed method can obtain more accurate prediction results for rolling bearing performance degradation trend.
作者
焦玲玲
陈捷
刘连华
JIAO Lingling;CHEN Jie;LIU Lianhua(School of Mechanical and Power Engineering,Nanjing Tech University,Nanjing 211816,China;Jiangsu Province Key Laboratory of Industrial Equipment Manufacturing and Digital Control Technology,Nanjing Tech University,Nanjing 211816,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2023年第12期109-117,共9页
Journal of Vibration and Shock
基金
国家重点研发计划(2019YFB20052004)。
作者简介
第一作者:焦玲玲,女,硕士生,1998年生;通信作者:陈捷,女,博士,教授,1971年生。