Fault diagnosis and detection in photovoltaic (PV) systems is essential to ensure the reliability and efficiency of renewable energy systems. This paper examines the application of advanced machine learning techniques to demonstrate the detection, diagnosis and prediction of anomalies in PV systems. We used a set of models within the Orange platform independently evaluate the performance of each model in classifying and analyzing data to discriminate between different classes based on metrics such as: accuracy (CA), Sensitivity (Recall), precision (Prec), specificity (Spec), Area Under the Curve (AUC), F1 score (F1), Confusion Matrix and ROC-AUC(Area Under the Receiver Operating Characteristic Curve). We then applied the stacking technique by creating a new model and using the same metrics as before to determine whether the stacking technique is more effective than the individual performance of the models in improving the classification performance.