Transformers-Based Encoder Model for Forecasting Hourly Power Output of Transparent Photovoltaic Module Systems

Solar power generation is usually affected by different meteorological factors, such as solar radiation, cloud cover, rainfall, and temperature. This variability has shown a negative impact on the large-scale integration of energy into supply systems. For successful electrical grid, it necessary to predict accurate panels. In this work, forecasting for two types system (non-transparent transparent panels) was configured smart artificial intelligence (AI) modelling. deep learning models, dataset obtained from target value electricity in kWh other features, weather conditions, radiance, insolation. PV values non-transparent panels were collected 1 January 31 December 2021 with an hourly interval. To prove efficiency proposed model, several approaches RNN LSTM, GRU, transformers implemented. Transformers model expressed best lower error rates MAE 0.05 0.04, RMSE 0.24 0.21, respectively. The showed efficient performance proved effective time-series data.