Fault detection and diagnosis in grid-connected photovoltaic systems

Multivariate feature extraction is very important for multivariate statistical systems monitoring. It can reduce the dimension of modeling data and facilitate the final monitoring accuracy. However, almost all the existing monitoring models are trained based on normal data. In practice, not only a mass of normal data but also fault data are easily collected and stored by the advanced sensing and computer technology. Therefore, this paper deals with the problem of monitoring through the development of fault detection and diagnosis (FDD) approach. The developed FDD technique uses feature extraction and selection, and fault classification tools under different operating conditions. This is addressed such that, the principal component analysis (PCA) technique is used for extracting and selecting features and the machine learning (ML) classifiers are applied for faults diagnosis. Only the most relevant features are chosen to be fed to different ML classifiers. The classification performance is established via different metrics for various ML-based PCA classifiers using data extracted from different operating conditions of the grid-connected photovoltaic (GCPV) system. The obtained results confirm the feasibility and effectiveness of the proposed approaches for fault diagnosis.