拖拉机电气故障诊断方法研究——基于深度学习理论和大数据

Research on Electrical Fault Diagnosis Method of Tractor——Based on Deep Learning Theory and Big Data

基于深度学习神经网络模型,结合拖拉机在线监测数据和故障数据,提出了一种新的故障诊断方法。该方法采用神经网络学习理论和大数据分析能力,通过样本数据的训练诊断出拖拉机各种运行状态的电气故障,为工作人员决定是否对电气系统进行检修提供更多参考信息。以拖拉机电压调节器故障诊断为例,对神经网络深度学习理论故障诊断的准确性进行了验证,结果表明:采用神经网络深度学习理论,在相同条件下具有更高的故障诊断准确率。

Based on the deep learning neural network model, it proposed a new fault diagnosis method, which combines the tractor line monitoring data and fault data. This method adopts the learning theory of neural network and the ability of large data analysis. It can diagnose the electrical faults of tractors in various operating states through training of sample data, and provide more reference information for staff to decide whether to overhaul the electrical system. Finally, taking the tractor voltage regulator fault diagnosis as an example, it verified the accuracy of fault diagnosis based on neural network deep learning theory. The test results show that the neural network deep learning theory has higher fault diagnosis accuracy under the same conditions.