滚动轴承的工作状况关系到使用滚动轴承的机械能否正常运行,预测轴承的剩余使用寿命(RUL)是避免机械系统失效的关键。针对传统的轴承使用寿命预测方法无法自适应调节特征权重、提取有用特征,造成预测值误差过大的问题,提出了一种带有卷...滚动轴承的工作状况关系到使用滚动轴承的机械能否正常运行,预测轴承的剩余使用寿命(RUL)是避免机械系统失效的关键。针对传统的轴承使用寿命预测方法无法自适应调节特征权重、提取有用特征,造成预测值误差过大的问题,提出了一种带有卷积块注意力模块(CBAM)的动态残差网络(Dy Res Net)用于预测轴承RUL。对振动信号进行快速傅里叶变换求得频域累积幅值特征,在动态残差网络中加入CBAM模块,并利用压缩激励模块进行特征细化得出预测结果,使用公开轴承数据集对所提模型进行评估。实验表明:与其他模型相比,Dy Res Net-CBAM模型能够充分提取特征信息,对轴承RUL预测的准确度高于其他模型。展开更多
The reliability assessment problem for products subject to degradation and random shocks is investigated. Two kinds of probabilistic models are constructed, in which the dependent competing failure process is consider...The reliability assessment problem for products subject to degradation and random shocks is investigated. Two kinds of probabilistic models are constructed, in which the dependent competing failure process is considered. First, based on the assumption of cumulative shock, the probabilistic models for hard failure and soft failure are built respectively. On this basis, the dependent competing failure model involving degradation and shock processes is established. Furthermore, the situation of the shifting-threshold is also considered, in which the hard failure threshold value decreases to a lower level after the arrival of a certain number of shocks. A case study of fatigue crack growth is given to illustrate the proposed models. Numerical results show that shock has a significant effect on the failure process; meanwhile, the effect will be magnified when the value of the hard threshold shifts to a lower level.展开更多
文摘滚动轴承的工作状况关系到使用滚动轴承的机械能否正常运行,预测轴承的剩余使用寿命(RUL)是避免机械系统失效的关键。针对传统的轴承使用寿命预测方法无法自适应调节特征权重、提取有用特征,造成预测值误差过大的问题,提出了一种带有卷积块注意力模块(CBAM)的动态残差网络(Dy Res Net)用于预测轴承RUL。对振动信号进行快速傅里叶变换求得频域累积幅值特征,在动态残差网络中加入CBAM模块,并利用压缩激励模块进行特征细化得出预测结果,使用公开轴承数据集对所提模型进行评估。实验表明:与其他模型相比,Dy Res Net-CBAM模型能够充分提取特征信息,对轴承RUL预测的准确度高于其他模型。
基金The National Natural Science Foundation of China(No.50405021)Graduate Training Innovative Projects Foundation of Jiangsu Province(No.CXLX12_0081)
文摘The reliability assessment problem for products subject to degradation and random shocks is investigated. Two kinds of probabilistic models are constructed, in which the dependent competing failure process is considered. First, based on the assumption of cumulative shock, the probabilistic models for hard failure and soft failure are built respectively. On this basis, the dependent competing failure model involving degradation and shock processes is established. Furthermore, the situation of the shifting-threshold is also considered, in which the hard failure threshold value decreases to a lower level after the arrival of a certain number of shocks. A case study of fatigue crack growth is given to illustrate the proposed models. Numerical results show that shock has a significant effect on the failure process; meanwhile, the effect will be magnified when the value of the hard threshold shifts to a lower level.
