Novel AI, machine, deep learning, and optimization-based computing for energy systems

Abstract

Dramatically increasing the integration of renewable distributed energy resources and new infrastructures of control in power systems, advanced communication devices, and novel monitoring system makes it essential to build an efficient and resilient power system. Also, due to the large volume of data, model complexity, and various uncertainties in modern power systems, traditional methods have some difficulties to reach the acceptable efficacy. However, these situations require reasonable solution/response time as well as acceptable accuracy to allow the controllers of system and operators of network to take actions that prevent and/or correct the disturbances. So, there is a basic need to choose optimal, fast, and accurate techniques for detection and identification in power systems. Recently, mathematical programming, artificial intelligence (AI), machine learning (ML), and deep learning (DL) techniques have been applied to achieve a reliable energy system.In this chapter, first, a quick overview of the modern energy systems and their main components is performed. Next, a brief overview of AI, ML, and DL techniques and their application in modern power systems are provided. Moreover, advanced techniques in the field the energy systems have been reviewed. Furthermore, some real-world applications of intelligent systems on the modern power systems are investigated. This section provides an overview of various study horizons from short-term to long-term and different control modes. Also, dynamic security assessment and stability control issues have been addressed with a focus on computational intelligence applications. Finally, future perspectives of dynamic security assessment by ML methods are given.