Research on Fault Diagnosis for TCT Photovoltaic Array Based on BA-KELM Model

As photovoltaic arrays are mostly installed in harsh outdoor environments, faults occur frequently during operation, resulting in reduced power generation efficiency and service life of photovoltaic arrays. In this paper, aiming at four common types of PV array faults (open circuit, short circuit, degradation, partial shadow), a bat algorithm-kernel extreme learning machine (BA-KELM) model is proposed for fault diagnosis of TCT-connected PV arrays. On the basis of analyzing the output characteristics of the PV array under typical faults, the appropriate feature quantities are selected to form the sample data. The RBF radial basis kernel function is selected to optimize the KELM model structure, and the bat algorithm is used to optimize the kernel parameters and penalty factors in the KELM model to improve the diagnostic accuracy of the model. Through the comparative analysis with PSO-KELM and PSO-ELM models, it is found that the BA-KELM model converges faster in the process of parameter optimization, and the training accuracy and testing accuracy of the optimized model are also higher.