Explainable AI for intelligent green energy forecasting: deep learning with iHow optimization algorithm (iHOW).

Area of Science:

• Energy Systems
• Artificial Intelligence
• Renewable Energy

Background:

• Accurate forecasting of green energy is vital for managing power grids and ensuring a stable electricity supply from fluctuating sources like solar and wind.
• The inherent variability of renewable energy necessitates advanced methods for reliable prediction and grid integration.

Purpose of the Study:

• To improve the accuracy of green energy forecasting by integrating Dynamic Temporal Convolutional Networks (DTCNs) with feature selection and metaheuristic optimization.
• To develop a robust model capable of handling the complexities of renewable energy generation patterns.

Main Methods:

• Implementation of Dynamic Temporal Convolutional Networks (DTCNs) for time-series forecasting.
• Application of feature selection techniques to identify and retain relevant input variables for enhanced model performance.
• Utilization of the novel iHow Optimization Algorithm (iHOW) for optimizing the forecasting model.

Main Results:

• Initial DTCN model achieved an MSE of 0.0845 and R-squared of 0.7265.
• Feature selection reduced MSE to 0.0022 and improved R-squared to 0.9005.
• Optimization with iHOW further enhanced performance, resulting in an MSE of [Formula: see text] and R-squared of 0.9804.

Conclusions:

• The integration of DTCNs, feature selection, and the iHOW metaheuristic optimizer significantly enhances green energy forecasting accuracy.
• This advanced approach provides a practical tool for energy professionals, supporting more reliable renewable energy production and grid management.
• The study demonstrates a substantial contribution to precise forecasting, aiding in the efficient integration of renewable energy sources for a sustainable future.