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
In recent years,most studies of complex networks have focused on a single network and ignored the interaction of multiple networks,much less the coupling mechanisms between multiplex networks.In this paper we investig...In recent years,most studies of complex networks have focused on a single network and ignored the interaction of multiple networks,much less the coupling mechanisms between multiplex networks.In this paper we investigate synchronization phenomena in multilayer networks with nonidentical topological structures based on three specific coupling mechanisms:assortative,disassortative,and anti-assortative couplings.We find rich and complex synchronous dynamic phenomena in coupled networks.We also study the behavior of effective frequencies for layers I and II to understand the underlying microscopic dynamics occurring under the three different coupling mechanisms.In particular,the coupling mechanisms proposed here have strong robustness and effectiveness and can produce abundant synchronization phenomena in coupled networks.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Project supported by the National Natural Science Foundation of China(Grants Nos.71801066 and 71704046)the Natural Science Foundation of Anhui Province,China(Grant Nos.1808085QG225 and 1908085MA22)+1 种基金the FundamentalResearch Funds for the Central Universities,China(Grant Nos.JZ2020HGTB0021 and JZ2021HGTB0065)the Outstanding Young Talent Support Program in Universities of Anhui Province in 2020 year。
文摘In recent years,most studies of complex networks have focused on a single network and ignored the interaction of multiple networks,much less the coupling mechanisms between multiplex networks.In this paper we investigate synchronization phenomena in multilayer networks with nonidentical topological structures based on three specific coupling mechanisms:assortative,disassortative,and anti-assortative couplings.We find rich and complex synchronous dynamic phenomena in coupled networks.We also study the behavior of effective frequencies for layers I and II to understand the underlying microscopic dynamics occurring under the three different coupling mechanisms.In particular,the coupling mechanisms proposed here have strong robustness and effectiveness and can produce abundant synchronization phenomena in coupled networks.
