Photovoltaic system fault diagnosis based on binary salp swarm and optimized support vector machine

In this study, we develop a pattern recognition method that utilizes dimensionality reduction and an optimized support vector machine (SVM) for fault diagnosis in photovoltaic systems, based on three-phase currents data. Initially, eleven (11) statistical descriptors are calculated from each phase currents. As a result, thirty-three (33) descriptors are included in the feature vector. However, not all descriptors are equally sensitive to faults. Because of this, we use the binary salp swarm optimisation algorithm (BSSA) based on the application of counter-propagation artificial neural networks classification error as fitness function to choose the most sensitive descriptors and exclude those with low sensitivity. Finally, an optimal support vector machine is adopted to ensure the classification task. The suggested approach is evaluated by using a real dataset. The obtained results demonstrate that the BSSA algorithm has a high convergence speed and can effectively select pertinent features. Furthermore, the high classification rate indicates that the suggested method can be effectively employed for photovoltaic system fault diagnosis.