Machine learning-based multi-criteria framework for renewable energy microgrid design

This paper presents a data-driven framework for designing microgrid that incorporate photovoltaic systems, wind turbines, diesel generators and battery storage. The framework identifies the optimal microgrid configuration by evaluating economic, energy, and environmental (3E) sustainability performance indicators (3E-SPI). It integrates HOMER pro with a custom Python-based tool integrating extreme gradient boosting (XGBoost) machine learning algorithm and thirteen 3E-SPI calculations for the community microgrid systems. A multi-objective optimization model is formulated for multi-criteria decision-making (MCDM) microgrid evaluation based on thirteen 3E-SPI. A two-layer MCDM approach is proposed to support stakeholders in the decision-making process. In the first layer, the Best Worst Method was used to determine the weighing of the thirteen 3E-SPI. In the second layer, the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) is proposed for ranking the microgrids. The XGBoost model outperformed deep neural network (DNN), achieving R2 value exceeding 0.95. Additionally, the MCDM of microgrids shows that hybrid photovoltaic/wind/battery/diesel microgrid is the optimal solution to supply the energy demand of the studied community. It achieves a total net present cost of approximately $1.3 million, a levelized cost of energy of $0.29/kWh, and annual CO2 emissions of only 169.11 kg/year.