Artificial Intelligence-Based Methods for Economic Power Dispatch in Smart Grids

In this paper, a wide-ranging appraisal has been made for using artificial intelligence (AI) techniques in economic power dispatch of smart grids. One of essential functions is economic dispatch (ED) in power systems, which assigns the optimal generation resources to meet demand with minimum operational cost. The modern power systems exhibit a non-linear, non-convex feature making the traditional optimization methods insufficient to cope up with progress in smart grid technology and transition of conventional energy resources towards renewable. These challenges can be dealt with effectively by AI-based methods. The review covers artificial neural networks, fuzzy logic systems and evolutionary algorithms as well as hybrid possibilities. Then, it compares these methods considering the effectiveness and computational efficiency in managing uncertainty of a smart grid. In addition, it discusses the implementation of these methods in addressing smart grid-level challenges related to renewable energy integration and demand response. Discussion: It concludes by presenting the common challenges and future issues in this area as well as recent trends of AI (deep reinforcement learning) due to a potential advent of new class tools using Explainable AI.