Optimizing the Integration of Wind and Solar Power for Hybrid Electrical Energy in High-rise Building Energy Systems
Abstract
The optimization of wind and solar energy utilization in high-rise building energy systems represents an innovative approach to sustainable energy management. This study examines the environmental factors influencing energy production from wind turbines and photovoltaic panels, utilizing machine learning models including bidirectional long short-term memory (Bi-LSTM), stacked long short-term memory (stacked LSTM), convolutional neural network-long short-term memory (CNN-LSTM), and attention LSTM to forecast energy output. The findings indicate that CNN-LSTM outperforms other models with the lowest mean absolute error (MAE) of 23.72, mean squared error (MSE) of 896.10, and Root mean squared error (RMSE) of 29.93 for PV power prediction, alongside the highest R-squared (R²) value of 0.9971. For wind power prediction, CNN-LSTM also achieved an MAE of 11.40, MSE of 213.28, RMSE of 14.60, and an R² of 0.9988. The analysis shows that wind power output is exponentially correlated with wind speed, while photovoltaic (PV) power output has a linear relationship with solar irradiation. The highest recorded energy production occurred in August, with PV generating 720 kWh, whereas wind power remained relatively stable, averaging 310 kWh throughout the year. The monthly and daily energy production patterns clearly demonstrate the benefits of integrating wind and solar systems for a more reliable energy supply. These findings emphasize the significance of optimizing energy systems based on location and showcase the potential of artificial intelligence in enhancing the accuracy of renewable energy forecasting.