A systematic review on AI-driven computational strategies for sustainable power systems

The magnitude and scope of the application of artificial intelligence (AI) and information-based computing methods to green and sustainable power generation systems have been significantly expanded to include research, initial development, implementation, and deployment. Over the past five years, this study has investigated various prominent AI modeling and optimization strategies for sustainable power systems. The following methodologies are included: cognitive neural network (NN) approaches, adaptive NNs, recurrent NNs with long-term dependencies, statistical approaches, sequential data processing networks, feature extraction networks, meta-heuristic approaches, evolutionary swarm algorithms, and hybrid optimization methodologies. The current study conducted a trend analysis of AI frameworks, with an emphasis on the performance, flexibility, and efficacy of some of the most frequently employed computational techniques for the development of sustainable power sources. Furthermore, this investigation examined the utilization of novel evaluation criteria, including computational complexity, accuracy, and ROC curve performance, in the context of power generation systems. The findings section is a systematic compilation and categorization of numerous distinct studies that were conducted over the past five years, with a focus on the methodology used and the application area. We have concluded by providing a concise summary of the results and outlining potential future research directions.