Fault Diagnosis of Photovoltaic Arrays Based on Support Vector Machine and t-Distributed Stochastic Neighbor Embedding

Fault diagnosis of Photovoltaic arrays becomes an interesting topic for authors due to the difficulty distinguish between faults. Many techniques have been applied for the diagnosis and classification of faults based on datasets samples like Decision Tree (DT), Random Forest (RF), and Support Vector Machine (SVM). However, to assure the best accuracy of diagnosis model must extract and prepare important data using a dimensionality reduction technique namely, Principal Component Analysis (PCA), and Independent Component Analysis (ICA). Therefore, this paper aims to build a diagnosis model using Support Vector Machine based on radial basis kernel function through prepared data using t-Distributed Stochastic Neighbor Embedding (T-SNE). The proposed model is compared with other methods for preparing data with SVM classifiers which are PCA and ICA. The simulation of faults, like partial shading, degradation, short circuit, and open circuit and diagnosis models studied are investigated and its results are reported.