Key points are not available for this paper at this time.
The increasing global demand for energy and the environmental impact of fossil fuels have propelled the advancement of solar photovoltaic (PV) technology. This paper emphasizes the importance of accurately predicting the efficiency and cost-effectiveness of solar PV systems for informed market and investment decisions. We adopted a machine learning approach due to the inadequacy of traditional models, utilizing historical data from household PV panels. We evaluated three models: Gradient Boosted Regression Tree (GBRT), Bi-Directional Gated Recurrent Unit (BiGRU), and Random Forest (RF). The results indicated that the BiGRU model excels in prediction accuracy, while the RF model demonstrates superior stability and robustness to outliers. Furthermore, the study assesses the economic and environmental returns of household PV systems, concluding that their installation offers significant environmental and financial benefits under a fully grid-connected operation.
Qing et al. (Tue,) studied this question.