International Conference on Sensing, Measurement & Data Analytics in the era of Artificial Intelligence (ICSMD2023)

143 / 2023-09-20 23:53:06

Demagnetization Fault Diagnosis Based on Feature Extraction and Convolutional Neural Network for Permanent Magnet Generator

permanent magnet synchronous generator,demagnetization fault,feature extraction,convolutional neural network

Special Sessions > SS18 - Advanced Sensing, Monitoring and Diagnosis in Smart Grid

During the operation of permanent magnet wind turbines, demagnetization faults of magnetic steel may occur, which directly affects the normal operation of wind turbines and has adverse effects on wind power generation. This paper proposes a comprehensive diagnosis method for demagnetization faults of permanent magnet generators based on feature extraction and convolutional neural network. A permanent magnet generator with a power of 25kW was used for demagnetization fault simulation experiments, which can truly simulate the operation of the permanent magnet generator under different demagnetization conditions. Collect the current signal of the experimental generator during operation, perform multiple feature extraction on it, and train the extracted different feature components through convolutional neural networks to achieve pattern recognition of the feature signal, in order to determine the operating status of the generator and achieve demagnetization fault diagnosis of permanent magnet wind turbines.