The research mainly focuses on the prediction and optimization of energy data with improved machine learning models, it contains two essential and highly intertwined parts:

Part 1: Energy Production Prediction

This component focuses on advanced forecasting methodologies using the energy dataset as a case study. The key objectives center on developing an accurate and interpretable prediction model for energy power generation that accounts for multiple variables (including temperature, humidity, etc.) and their complex interactions, while improving the reliability of short-term and medium-term energy power forecasting. The research approach integrates innovative multivariable time series analysis and evolutionary algorithms to advance existing research.

Part 2: Infrastructure Optimization

The second component addresses the strategic placement of energy infrastructure through a comprehensive decision-making framework. This research advances the field by developing an integrated approach that combines traditional location analysis with novel prediction-driven optimization techniques.