Interval-Valued Reduced Ensemble Learning Based Fault Detection and Diagnosis Techniques for Uncertain Grid-Connected PV Systems

One of the most promising renewable energy technologies is photovoltaics (PV). Fault detection and diagnosis (FDD) becomes more important in order to guarantee high reliability PV systems. FDD systems using machine learning technique aims develop effective models that can provide a better rate accuracy. Recently, numerous based ensemble have been applied different combination techniques. Ensemble method tool merges several base produce one optimal predictive model. In this study, we propose six Leaning (EL)-based paradigms for uncertain Grid-Connected First, EL-based interval centers ranges upper lower bounds techniques are proposed deal with system uncertainties (current/voltage variability, noise, measurement errors, $\ldots$ ). Next, improve abilities, two kernel PCA (IKPCA)-based EL classifiers developed. The IKPCA-EL addressed so features extraction selection phases performed IKPCA sensitive significant interval-valued characteristics transmitted model classification purposes. Finally, number observations training data set reduced Hierarchical K-means overcome problem computation time storage cost. Therefore, KPCA-EL proposed. study demonstrated feasibility efficiency fault