From classical to AI-driven load frequency control: Addressing smart grid challenges with renewable energy sources and EVs integration

This research investigates advancements in Load Frequency Control (LFC) by examining AI-based and optimization-driven strategies aimed at improving frequency stability in Multi-Area Power Systems (MAPSs). The study offers a broad evaluation, transitioning from early classical approaches to emerging modern and AI-driven techniques. A key methodological contribution is the structured comparison of traditional, intelligent, and optimization-based controllers evaluated by using standardized performance indices such as ITSE under realistic system conditions, including nonlinearities (GRC, GDB, CTD) and communication delays. It assesses various control strategies, emphasizing the adaptability and robustness required for effective LFC in dynamic environments. Particular attention is also paid to optimization methods and artificial controllers. The study also evaluates the impacts of energy storage technologies, electric vehicles (EV), and renewable energy sources on the system frequency stability. Moreover, it also investigates machine learning and reinforcement learning frameworks with the potential to address LFC issues. The paper contains critical smart grid issues like potential cybersecurity threats, controller failures, data transmission errors and communication boundaries. It also introduces a systematic Scopus-based literature filtration process, resulting in a publication trend analysis that reveals evolving research directions. Providing practical insights and novel approaches, this work aims to divert future studies toward more stable, flexible, and robustness-based LFC solutions in a more practical complex power system realization.