基于卷积神经网络和迁移学习的电动泵故障诊断方法研究

Fault Diagnosis Method for Electric Pump Based on Convolutional Neural Network and Transfer Learning

提出了一种基于卷积神经网络和迁移学习的电动泵故障诊断方法,其特点是只利用少量样本数据就能自动从振动信号中提取有效的故障特征并完成诊断。设计了基于迁移学习的卷积神经网络模型训练方法,给出了利用其进行故障诊断的方法步骤,采用齿轮和电机在正常和不同故障状态时的振动数据对方法的有效性进行了测试。结果表明,所提出的方法对不同的故障状态有较高的识别精度,具有良好的实用性。

Electric pumps are widely used in energy,power,chemical and other fields.A fault diagnosis method for electric pump based on convolutional neural network(CNN)and transfer learning was proposed.The method can extract valid fault features automatically from vibration signal and complete fault diagnosis with a small amount of sample data.The training method for CNN based on transfer learning was designed,and the fault diagnosis steps were given.The method was tested with vibration data of gears and electric motor in normal and different fault states.The results show that the proposed method has high recognition accuracy for different fault states and has good practicality.