摘要： 多传感数据融合方法能够提升轴承故障诊断的准确性，为解决电机轴承故障诊断中现有大多数多传感数据融合方法存在传感数据类型单一以及难以充分挖掘不同模态数据间冗余性、互补性的问题，提出了一种基于多约束模态不变图卷积融合网络(MCMI-GCFN)的轴承故障诊断方法。首先，通过卷积自编码器(CAE)和挤压-激励模块(SE block)对原始电流、振动信号进行特征提取；其次，引入源域分类器和域鉴别器在域对抗训练基础之上捕获不同模态数据间的模态不变性，充分挖掘多模态数据间的冗余性以及互补性；然后，利用图卷积神经网络(GCN)的空间聚合特性捕获电流、振动模态相近时间步特征之间的依赖关系以精确融合其上下文语义信息；最后，在德国帕德伯恩大学公开轴承损伤电流、振动数据集上进行验证，实验结果表明所提融合方法达到了99.6%的轴承故障诊断精度，优于非融合方法约9%~11.4%，结果验证了所提模型的有效性。

Abstract: Multisensor data fusion method can improve the accuracy of bearing fault diagnosis, in order to address the problems of single-sensor data types and the insufficient exploration of redundancy and complementarity between different modal data in most existing multisensor data fusion methods for bearing fault diagnosis, this paper proposes a bearing fault diagnosis method based on a multiple-constraint modal-invariant graph convolutional fusion network (MCMI-GCFN). Firstly, the method uses a convolutional autoencoder (CAE) and squeeze-and-excitation block (SE block) to extract features of raw current and vibration signals. Secondly, the model introduces source domain classifiers and domain discriminators to capture modal invariance between different modal data based on domain adversarial training, making use of the redundancy and complementarity between multimodal data. Then, the method utilizes the spatial aggregation property of graph convolutional neural networks (GCN) to capture the dependency relationship between current and vibration modes with similar time step features for accurately fusing contextual semantic information. Finally, the paper conducts validation on the public bearing damage current and vibration dataset from Paderborn University. The experimental results show that the deliveried fusion method achieves a bearing fault diagnosis accuracy of 99.6%, which is about 9%~11.4% better than that with nonfusion methods.