摘要
针对滚动轴承在噪声环境中故障诊断准确率低、诊断模型稳定性差等问题,提出了一种DARTS-CNN-BiLSTM故障诊断模型。首先,使用可微架构搜索(differentiable architecture search, DARTS)算法对卷积神经(convolutional neural network, CNN)网络进行结构寻优,以提升CNN对原始振动信号的深层特征挖掘能力和抗噪性能。然后,引入双向长短时记忆(bidirectional long-short term memory, BiLSTM)网络进一步提取信号的时序特征,提升模型的稳定性和鲁棒性。最后,通过全局平均池化和Softmax分类器完成故障分类。分别使用西储大学和渥太华大学公开数据集进行实验,结果表明,此模型平均诊断准确率可达98.38%以上,在添加不同信噪比大小的额外噪声条件下,该模型仍能保持较高的诊断准确率,与其他模型相比,具备更好的抗噪性和稳定性。
To solve the problem of the low accuracy of fault diagnosis and poor stability of the diagnosis model for rolling bearings in noisy environments,a DARTS-CNN-BiLSTM fault diagnosis model was proposed.Firstly,the differentiable architecture search algorithm(DARTS)was used to optimize the structure of convolutional neural network(CNN),so as to improve CNN′s deep feature mining ability and anti-noise performance of original vibration signals.Secondly,a bidirectional long-short-term memory network(BiLSTM)was introduced to further extract the timing features of the signal to improve the stability and robustness of the model.Finally,the fault classification was completed by global average pooling and Softmax classifier.Using the public datasets of Western Reserve University and the University of Ottawa respectively,the results show that the average diagnostic accuracy of this model can reach over 98.38%.Under the condition of adding additional noise with different signal-to-noise ratios,the model still maintains high diagnostic accuracy and better anti-noise performance and stability compared with other models.
作者
陈露萌
李一鸣
黄民
CHEN Lumeng;LI Yiming;HUANG Min(Mechanical Electrical Engineering School,Beijing Information Science&Technology University,Beijing 100192,China)
出处
《北京信息科技大学学报（自然科学版）》
2023年第2期23-31,共9页
Journal of Beijing Information Science and Technology University
基金
工信部2021年高档数控系统及伺服电机项目(TC210H03A-05)
北京市教委科研计划科技一般项目(KM202011232011)。
关键词
可微架构搜索
卷积神经网络
双向长短时记忆网络
轴承故障诊断
抗噪声
differentiable architecture search(DARTS)
convolutional neural network(CNN)
bidirectional long-short term memory(BiLSTM)
bearing fault diagnosis
anti-noise
