故障诊断分类技术在工业上已经被广泛的使用，在工业设备维护起到了关键性作用，但是自动故障诊断分类技术目前还存在不足，要求精确地对设备机械进行自动诊断，准确地分析出设备故障产生的原因，从而确定故障发生的部位。针对工业上旋转机械设备的特殊性和复杂性，引入了深度学习算法来提高设备故障分类的准确率。首先对旋转机械设备建立数据集，通过深度学习算法对数据进行特征提取，由多个网络层迭代学习设备故障特征，最终优化深度学习算法模型输出不同设备故障类型，提高系统分类的准确率。本文还对故障诊断分类等技术进行总结与分析，然后重点分析了深度学习故障诊断技术在工业上机械旋转类的应用；最后提出了现有深度学习故障诊断分类技术研发方法的不足，希望深度学习领域在故障诊断技术有很好的发展。
Fault diagnosis classification
technology has been widely used in the industry and has played a key role in
the maintenance of industrial equipment. However, the automatic fault diagnosis
classification technology currently has shortcomings. It requires accurate
automatic diagnosis of equipment and machinery, and accurate analysis of
equipment failures, the cause of the occurrence so as to determine the location
of the fault. Aiming at the particularity and complexity of industrial rotating
machinery and equipment, deep learning algorithms are introduced to improve the
accuracy of equipment fault classification. Firstly, a data set is established
for rotating machinery and equipment, and features are extracted through deep
learning algorithms. Multiple network layers are used to iteratively learn
equipment fault features. Finally, the deep learning algorithm model is
optimized to output different types of equipment faults to improve the accuracy
of system classification. This paper also summarizes and analyzes fault
diagnosis and classification technologies, and then focuses on the application
of deep learning fault diagnosis technology in industrial mechanical rotation.
Finally, it puts forward the shortcomings of the existing deep learning fault
diagnosis classification technology research and development methods, and hopes
the deep learning field has a very good development in fault diagnosis
technology.
Lou, X. and Loparo, K.A. (2004) Bearing Fault Diagnosis Based on Wavelet Transform and Fuzzy Inference. Mechanical Systems & Signal Processing, 18, 1077-1095. https://doi.org/10.1016/S0888-3270(03)00077-3
