改进2D-CNN的永磁同步电机故障诊断分析

Improved 2D-CNN-Based Fault Diagnosis Method for Permanent Magnet Synchronous Motor

针对卷积神经网络在噪声环境中特征提取能力不足,导致永磁同步电机故障诊断出现准确率低、泛化能力差的问题,提出了一种改进二维卷积神经网络与混合注意力机制的故障诊断方法。该方法首先将采集的一维时序信号转换为二维灰度图;其次,改进多尺度特征提取模块,将该模块部分普通卷积替换为空洞卷积,以最大程度地提取到数据信号中的有效信息;然后引入混合注意力机制动态更新权重参数,强化故障特征,抑制噪声的干扰;最后使用分类器进行匝间短路故障诊断。实验结果表明:与其他方法相比,文中模型具有更好的准确性和鲁棒性;在各噪声背景下,本模型的准确率均在96%以上,表明所提方法具有较强的抗噪性能和泛化能力。

In order to solve the problem of low accuracy and poor generalization ability of permanent magnet synchronous motor fault diagnosis due to insufficient feature extraction ability of convolutional neural network in noisy environment,a fault diagnosis method with improved two-dimensional convolutional neural network and mixed attention mechanism is proposed.Firstly,the one-dimensional time sequence signal is converted into a two-dimensional gray level map.Secondly,the multi-scale feature extraction module is improved,and some ordinary convolution is replaced by empty convolution in this module,so as to extract effective information from data signals to the maximum extent.Then,the hybrid attention mechanism is introduced to dynamically update the weight parameters,strengthen the fault characteristics,and suppress the interference of noise.Finally,the classifier is used to diagnose the inter-turn short circuit fault.Experimental results show that the proposed model in this paper is more accurate and robust than other methods.Under each noise background,the accuracy of the model is above 96%,which indicates that the proposed method has strong anti-noise performance and generalization ability.