Fault diagnosis of photovoltaic array based on gated residual network with multi-head self attention mechanism

Effective fault identification and diagnosis in photovoltaic (PV) arrays is vital for improving the effectiveness, and safety of solar energy systems. While various artificial intelligence methods have successfully established fault detection and diagnosis models, introducing inefficiencies and potentially overlooking useful features. Moreover, these methods often employ neural networks with limited performance capabilities. In response to these challenges, this paper introduces an innovative intelligent model that integrates a combination of a gated residual neural network (GRN) and a multi-head self-attention mechanism (MHSA). To evaluate the proposed fault diagnosis model, the small-scale PV grid system is implemented, and fault simulation experiments, including arc faults, maximum power tracking failures, line-to-line, open circuit, degradation, and partial shading with normal conditions, are conducted to acquire simulation datasets. Additionally, widely used neural network models, including artificial neural networks, recurrent neural networks, convolutional neural networks, and the proposed model without an attention mechanism, are employed for comparison. Furthermore, common machine learning approaches found in the literature for diagnosing faults of PV arrays, optimized by Bayesian technique are implemented and compared. Simulation results highlight that the proposed approach attains superior performance across key metrics, including accuracy, precision, recall, f1-score, and training efficiency. Notably, the proposed model achieves an impressive testing accuracy of 99.71%, surpassing alternative methods. This highlights its effectiveness as a robust and efficient solution for fault diagnosis in PV arrays.