Aiming at the problem that feature extraction relies too much on human experience in the fault diagnosis of wind turbine gearboxes and the accuracy is not high,a method based on the combination of long short-term memory (LSTM) and support vector machine (SVM) was proposed.The FFT transformation was done on the original time-domain vibration signal, and the advantages of adaptive intelligent extraction of features by LSTM neural network were used, combined with the classification function of SVM, to achieve more accurate fault diagnosis of wind turbine gearbox. The simulation results show that the accuracy of the network model can reach 100% after 16 rounds of training, and the accuracy using test set data can also reach 99.1%.
参考文献
相似文献
引证文献
引用本文
王璞,孙洁,张怡.基于LSTM-SVM的风电机组齿轮箱故障诊断[J].机床与液压,2023,51(16):211-214. WANG Pu, SUN Jie, ZHANG Yi. Fault Diagnosis of Wind Turbine Gearbox Based on LSTM-SVM[J]. Machine Tool & Hydraulics,2023,51(16):211-214