Dynamic expert project assessment for green wind energy park investments via molecular fuzzy reinforcement learning decision-making technique

Abstract

Wind energy parks play a key role in sustainable energy production and carbon emission reduction. This study proposes a novel decision-making framework to identify effective investment strategies for green wind energy park projects. A dynamic expert dataset is constructed using the Q learning algorithm, while molecular fuzzy Bayesian network and molecular fuzzy multi objective particle swarm optimization are used to weight evaluation criteria and rank strategy alternatives. The analysis focuses on a 50 MW onshore wind farm with an average wind speed of 7.7 m/s and an annual energy production of approximately 153 GWh. The project provides an annual carbon reduction of nearly 95,000 tons and demonstrates strong operational efficiency. The findings show that social compliance and ecological compliance are the most critical evaluation criteria, while balanced energy supply with energy storage integration emerges as the most effective investment strategy.