Short-Term Photovoltaic Power Plant Output Forecasting Using Sky Images and Deep Learning

With the steady increase in use of renewable energy sources sector, new challenges arise, especially unpredictability these sources. This uncertainty complicates management, planning, and development systems. An effective solution to is short-term forecasting output photovoltaic power plants. In this paper, a novel method for production prediction was explored which involves continuous photography sky above plant. By analyzing series images, patterns can be identified help predict future generation. A hybrid model that integrates both Convolutional Neural Network (CNN) Long Short-Term Memory (LSTM) developed tested. effectively detects spatial temporal from images data, displaying considerable accuracy. particular, 74% correlation found between model’s predictions actual values, demonstrating efficiency. The results paper suggest CNN-LSTM offers an improvement accuracy practicality compared traditional methods. highlights potential Deep Learning improving practices, particularly prediction, contributing overall sustainability