Photovoltaic string fault optimization using multi-layer neural network technique

If catastrophic failures in photovoltaic (PV) arrays are not identified promptly, then the output power decreases, which is an undesirable performance and further it prompts sparking in the arrays. To prevent the damage and identification of faults in PV arrays, herein a Multilayer Neural Network (MLNN) technique is proposed. Shortcircuited modules and disconnected string faults are the results of using this approach to analyze the system. The innovative aspect of this study is its flexibility, in that a long-term dataset has been employed in the proposed MLNN technique training and validation phase, and circumstances including datasets polluted with random noise have also been explored. An accuracy of 98.76 % is achieved by testing the input to the proposed MLNN on a PV system with a capacity of 2.5 kW, and the results of this testing were supplied by the results acquired. In this study, machine learning confusion matrices are employed to examine the significant number of errors present in the PV string. Faults in the PV array that are line-to-ground (L-G) and line-to-line (L-L) faults are detected, classified, and localized with the help of the proposed technique. The proposed MLNN technique requires just one current sensor to be installed for each string. However, it can identify problems in photovoltaic arrays of any size or degree of mismatch. To validate the efficacy of the algorithm, it has been subjected to stringent testing inside an experimental environment.