电子测量与仪器学报

谢锋云,陈惠航,牛康,潘圳锴,王书蕾,孙浩然,谢源威.多域特征图神经网络的变速器复合故障诊断研究[J].电子测量与仪器学报,2025,39(12):53-63

多域特征图神经网络的变速器复合故障诊断研究

Research on transmission compound fault diagnosis basedon multi-domain feature graph neural network

DOI：

英文关键词:multi-domain features graph compound fault diagnosis graph neural networks rolling bearing transmission

基金项目:国家自然科学基金（52265068）、江西省自然科学基金（20224BAB204050）项目资助

作者	单位
谢锋云	1.华东交通大学机电与车辆工程学院南昌330013；2.华东交通大学智能交通装备全寿命技术创新中心南昌330013
陈惠航	华东交通大学机电与车辆工程学院南昌330013
牛康	华东交通大学机电与车辆工程学院南昌330013
潘圳锴	华东交通大学机电与车辆工程学院南昌330013
王书蕾	华东交通大学机电与车辆工程学院南昌330013
孙浩然	华东交通大学机电与车辆工程学院南昌330013
谢源威	华东交通大学机电与车辆工程学院南昌330013

Author	Institution
Xie Fengyun	1.School of Mechanical Electronical and Vehicle Engineering, East China Jiaotong University, Nanchang 330013, China; 2.China Life-cycle Technology Innovation Center of Intelligent Transportation Equipment, East China Jiaotong University, Nanchang 330013, China
Chen Huihang	School of Mechanical Electronical and Vehicle Engineering, East China Jiaotong University, Nanchang 330013, China
Niu Kang	School of Mechanical Electronical and Vehicle Engineering, East China Jiaotong University, Nanchang 330013, China
Pan Zhenkai	School of Mechanical Electronical and Vehicle Engineering, East China Jiaotong University, Nanchang 330013, China
Wang Shulei	School of Mechanical Electronical and Vehicle Engineering, East China Jiaotong University, Nanchang 330013, China
Sun Haoran	School of Mechanical Electronical and Vehicle Engineering, East China Jiaotong University, Nanchang 330013, China
Xie Yuanwei	School of Mechanical Electronical and Vehicle Engineering, East China Jiaotong University, Nanchang 330013, China

摘要点击次数: 596

全文下载次数: 509

中文摘要:

变速器在旋转机械中有着广泛的应用，对其复合故障诊断有利于机械设备的健康运行。为了提高变速器复合故障诊断的准确度和泛化性，提出了一种基于多域特征图神经网络(MDFGNN)的变速器复合故障诊断方法。首先，分别在时域、频域、熵中提取振动信号的多个特征，得到丰富的变速器多特征状态信息，并构建节点特征矩阵，再利用K-近邻算法（k-nearest neighbor，KNN）提取节点特征的序列规律性和有序性，并构建边索引矩阵；其次将节点特征矩阵与边索引矩阵组合来构建特征图, 将特征图输入到图神经网络（graph neural networks，GNN）模型，来进行分类识别；最后通过向原始数据中添加不同信噪比的高斯白噪声和公开的数据集检验所提模型的准确度和泛化性。为了验证所提方法的有效性，搭建了变速器振动实验平台，通过压电式加速度传感器采集5种状态的变速器数据。研究结果表明，多域特征图能够对变速器复合故障状态进行充分且全面的故障信息挖掘，克服复合故障信号微弱，非线性，复杂的问题，获取更敏锐的变速器运行状态信息，提高原始数据的利用率和模型的稳定性，相较于现有其他变速器故障诊断方法正确率可提高4.75%~12.26%，准确度相差波动区间介于0.07%~1.28%，泛化性检验可达96.25%。

英文摘要:

Transmission systems are widely applied in rotating machinery, and the diagnosis of their composite faults is crucial for ensuring the healthy operation of mechanical equipment. In order to improve the accuracy and generalization of transmission compound fault diagnosis, a method of transmission compound fault diagnosis based on multi-domain feature map neural network (MDFGNN) is proposed. Firstly, multiple features of vibration signals are extracted from time domain, frequency domain and entropy to obtain rich multi-feature status information of the transmission, and a node feature matrix is constructed. Then k-nearest neighbor (KNN) algorithm is used to extract the sequence regularity and order of node features, and an edge index matrix is constructed. Secondly, the node feature matrix and the edge index matrix are combined to build the feature map, and the feature map is input into the Graph Neural Networks (GNN) model for classification and recognition. Finally, the accuracy and generalization of the proposed model were tested by adding Gaussian white noise with different signal-to-noise ratios to the original data and the HUST Bearing dataset. In order to verify the effectiveness of the proposed method, a transmission vibration test platform was built, and transmission data of five states were collected by piezoelectric acceleration sensors. The results show that: The multi-domain feature map can fully and comprehensively mine the fault information of the compound fault state of the transmission, overcome the weak, non-linear and complex problems of the compound fault signal, obtain more sensitive information of the transmission operation state, improve the utilization rate of the original data and the stability of the model. Compared with other existing transmission fault diagnosis methods, the accuracy rate can be increased by 4.75%~12.26%, the accuracy difference fluctuation range is 0.07%~1.28%, and the generalization test can reach 96.25%.

查看全文查看/发表评论下载PDF阅读器