Forecasting Solar Photovoltaic Power Production with Hybrid Approach of Multilayer CNNs-GRU-BiLSTMs

This research article introduces an innovative neural network architecture designed to address the inherent unpredictability and intermittent nature of photovoltaic (PV) power generation. This design intelligently combines the capabilities of multilayer Convolutional Neural Networks (CNNs), a Gated Recurrent Unit (GRU), and Bidirectional Long Short-Term Memory Networks (BiLSTMs), resulting in a robust deep learning model adapted to improve solar PV power forecasting. To evaluate the performance of the proposed model, we leverage real-time weather data from the year 2022, utilizing the PVlib Python library to simulate the electricity output of a PV plant. Our findings demonstrate that the developed hybrid model performs very well, particularly in short-term forecasting scenarios, achieving remarkable precision. Therefore, this research exhibits remarkable potential for practical applications in solar energy planning and management, encouraging a more sustainable and energy-efficient future.