AI Applications for Power Quality Challenges in Distribution Systems: Technical Barriers and Emerging Trends
Abstract
The integration of renewable energy sources and power electronic devices into modern distribution systems has intensified power quality (PQ) challenges, including voltage sags, harmonic distortions, and transient disturbances. This review presents a comprehensive analysis of artificial intelligence (AI) applications for PQ monitoring and control, addressing both technical limitations and emerging trends in distribution systems (DSS). Unlike previous studies that focus on isolated AI methods or specific PQ issues, this work provides a comprehensive synthesis encompassing detection, classification, and mitigation across multiple disturbance types. Using over 70 recent publications, the study classifies key PQ disturbances, evaluates AI techniques-including machine learning, deep learning, and hybrid models-and highlights their performance in terms of accuracy, real-time processing, and adaptability. The article illustrate the advantages of AI in diagnosing PQ issues more quickly and accurately than traditional methods. One significant finding is the limited application of AI-driven control in power converters (e.g., voltage source inverters, multilevel inverters) despite their central role in grid interfacing and stability. The findings emphasize that AI enhances PQ management by enabling predictive diagnostics and adaptive control in renewable-rich, bidirectional grids. However, real-time deployment is constrained by challenges such as data scarcity, model interpretability, and integration with existing SCADA systems. The review recommends advancing hybrid and federated learning models, as well as edge AI deployment and explainable AI frameworks, to overcome these barriers. Furthermore, this work provides practical insights for researchers and engineers seeking intelligent, scalable, and future-ready PQ solutions in decentralized power systems.