Fault diagnosis is an important measure to ensure the safety of production, and all kinds of fault diagnosis methods are of importance in actual production process. However, the complexity and uncertainty of productio...Fault diagnosis is an important measure to ensure the safety of production, and all kinds of fault diagnosis methods are of importance in actual production process. However, the complexity and uncertainty of production process often lead to the changes of data distribution and the emergence of new fault classes, and the number of the new fault classes is unpredictable. The reconstruction of the fault diagnosis model and the identification of new fault classes have become core issues under the circumstances. This paper presents a fault diagnosis method based on model transfer learning and the main contributions of the paper are as follows: 1) An incremental model transfer fault diagnosis method is proposed to reconstruct the new process diagnosis model. 2) Breaking the limit of existing method that the new process can only have one more class of faults than the old process, this method can identify M faults more in the new process with the thought of incremental learning. 3) The method offers a solution to a series of problems caused by the increase of fault classes. Experiments based on Tennessee-Eastman process and ore grinding classification process demonstrate the effectiveness and the feasibility of the method.展开更多
Aiming at the problems of lack of fault diagnosis samples and low model generalization ability of cross-working gear based on deep transfer learning, a fault diagnosis method based on improved deep residual network an...Aiming at the problems of lack of fault diagnosis samples and low model generalization ability of cross-working gear based on deep transfer learning, a fault diagnosis method based on improved deep residual network and transfer learning was proposed. Firstly, one-dimensional signal is transformed into two-dimensional time-frequency image by continuous wavelet transform. Then, a deep learning model based on ResNet50 is constructed. Attention mechanism is introduced into the model to make the model pay more attention to the useful features for the current task. The network parameters trained by ResNet50 network on ImageNet dataset were used to initialize the model and applied to the fault diagnosis field. Finally, to solve the problem of gear fault diagnosis under different working conditions, a small sample training set is proposed for fault diagnosis. The method is applied to gearbox fault diagnosis, and the results show that: The proposed deep model achieves 99.7% accuracy of gear fault diagnosis, which is better than the four models such as VGG19 and MobileNetV2. In the cross-working condition fault diagnosis, only 20% target dataset is used as the training set, and the proposed method achieves 93.5% accuracy.展开更多
为了解决滚动轴承故障诊断中样本分布差异大、有效故障样本少以及不同故障样本数量不均衡所导致的诊断精度较低的问题;提出基于个性化联邦迁移学习(personalized federated transfer learning,PFTL)的滚动轴承故障诊断方法。在所提出的P...为了解决滚动轴承故障诊断中样本分布差异大、有效故障样本少以及不同故障样本数量不均衡所导致的诊断精度较低的问题;提出基于个性化联邦迁移学习(personalized federated transfer learning,PFTL)的滚动轴承故障诊断方法。在所提出的PFTL中,首先在预训练阶段,将不同分布的各类型故障样本作为联邦学习的各个客户端的输入,并引入贝叶斯层级模型对联邦学习的本地训练和聚合规则进行个性化调整,从而使得预训练模型在避免过拟合问题的同时具有较强的泛化能力;其次引入模型补丁,对预训练模型结构进行调整,并利用目标任务样本对模型进一步微调;最后在CWRU轴承数据集上进行故障诊断实验。实验结果证明所提方法的有效性。展开更多
文摘Fault diagnosis is an important measure to ensure the safety of production, and all kinds of fault diagnosis methods are of importance in actual production process. However, the complexity and uncertainty of production process often lead to the changes of data distribution and the emergence of new fault classes, and the number of the new fault classes is unpredictable. The reconstruction of the fault diagnosis model and the identification of new fault classes have become core issues under the circumstances. This paper presents a fault diagnosis method based on model transfer learning and the main contributions of the paper are as follows: 1) An incremental model transfer fault diagnosis method is proposed to reconstruct the new process diagnosis model. 2) Breaking the limit of existing method that the new process can only have one more class of faults than the old process, this method can identify M faults more in the new process with the thought of incremental learning. 3) The method offers a solution to a series of problems caused by the increase of fault classes. Experiments based on Tennessee-Eastman process and ore grinding classification process demonstrate the effectiveness and the feasibility of the method.
基金Supported by National Natural Science Foundation of P. R. China （60574083）, Key Laboratory of Process Industry Automation, State Education Ministry of China （PAL200514）
文摘Aiming at the problems of lack of fault diagnosis samples and low model generalization ability of cross-working gear based on deep transfer learning, a fault diagnosis method based on improved deep residual network and transfer learning was proposed. Firstly, one-dimensional signal is transformed into two-dimensional time-frequency image by continuous wavelet transform. Then, a deep learning model based on ResNet50 is constructed. Attention mechanism is introduced into the model to make the model pay more attention to the useful features for the current task. The network parameters trained by ResNet50 network on ImageNet dataset were used to initialize the model and applied to the fault diagnosis field. Finally, to solve the problem of gear fault diagnosis under different working conditions, a small sample training set is proposed for fault diagnosis. The method is applied to gearbox fault diagnosis, and the results show that: The proposed deep model achieves 99.7% accuracy of gear fault diagnosis, which is better than the four models such as VGG19 and MobileNetV2. In the cross-working condition fault diagnosis, only 20% target dataset is used as the training set, and the proposed method achieves 93.5% accuracy.
文摘为了解决滚动轴承故障诊断中样本分布差异大、有效故障样本少以及不同故障样本数量不均衡所导致的诊断精度较低的问题;提出基于个性化联邦迁移学习(personalized federated transfer learning,PFTL)的滚动轴承故障诊断方法。在所提出的PFTL中,首先在预训练阶段,将不同分布的各类型故障样本作为联邦学习的各个客户端的输入,并引入贝叶斯层级模型对联邦学习的本地训练和聚合规则进行个性化调整,从而使得预训练模型在避免过拟合问题的同时具有较强的泛化能力;其次引入模型补丁,对预训练模型结构进行调整,并利用目标任务样本对模型进一步微调;最后在CWRU轴承数据集上进行故障诊断实验。实验结果证明所提方法的有效性。
