关键词: 卷积神经网络, 传感器数据, 发电机转子, 断裂故障, 发电机转子裂纹模型, 二维矩阵

Abstract: The fault diagnosis algorithm of new energy wind turbine rotor fault based on convolutional neural network and sensor data was proposed. Based on the correlation between the rotor running state and the axial stiffness of the crack, the generator rotor crack model was established. After integrating the generator rotor vibration data collected by multiple sensors, the original one-dimensional data was converted into the form of mn two-dimensional matrix, analyzing the sensor data information by using the multi-channel convolutional neural network, and putting the output results into the generator rotor crack model to realize the diagnosis of rotor fracture fault. In the test results, the error of the design algorithm is stable to within 0. 03 μm for the generator rotor crack fault diagnosis results under different data sources.

Key words: convolutional neural network, sensor data, generator rotor, crack fault, generator rotor crack model, two-dimensional matrix