Optimizing Solar Energy Forecasting Using Deep Learning Techniques: A Study on Predictive Models in Forested Environments

Precise information about the efficiency of harvesting solar energy is an important requirement in the application of the renewable energy on practical projects, especially where climate is characteristically changeable. This paper focuses on the use of deep learning methods namely LSTM to predict the efficiency of solar power in forested and other regions of the land. This seems to be more efficient than conventional neural network like Feedforward Neural Network (FNN) and Convolutional Neural Network (CNN) in terms of accuracy with the MAPE of 22% than FNN and 15% than CNN. Currently, based on the results of the LSTM model the authors got 7.45% and 20.35% improvements in terms of computational time and memory respectively when compared with the CNN models. Based on these results, it can be concluded that deep learning techniques, especially LSTM, are suitable when it comes to the solar power predictionconsidering accuracy and practicability. The further developments of the work will be the inclusion of other environmental data, as well as enhancing the real time prediction –leading to better sustainable energy solutions.