A Novel Wind Turbine Fault Diagnosis Based on Pca-Cs and Lenet-R with Small-Scale Data

A method based on PCA-CS algorithm and LeNet-R network is proposed to solve these problems. The PCA-CS algorithm combines the advantages of principal component analysis (PCA) and compressed sensing (CS). Firstly, principal component analysis is adopted to reduce the dimension of multi-dimensional raw data and preliminarily eliminate the interference. Then, the properties of compressed sensing technology are used to sparse the reduced-dimensional signal and reconstruct the signal to improve the data quality and minimize the impact of interfering signals. The signal processed by the PCA-CS algorithm not only has the characteristics of small data size, but also has a clear distinction, which helps to improve the speed and accuracy of model fault diagnosis. Experiments show that the method has high accuracy and superior diagnosis speed in the case of small-scale data compared with traditional fault diagnosis methods