Enhancing Power Grid Resilience through Real-Time Fault Detection and Remediation Using Advanced Hybrid Machine Learning Models
Ensuring a reliable and uninterrupted supply of electricity is crucial for sustaining modern and advanced societies. Traditionally, power systems analysis was mostly dependent on formal commercial software, mathematical models produced via a mix of data analysis, control theory, and statistical methods. As power grids continue to grow and the need for more efficient and sustainable energy systems arises, attention has shifted towards incorporating artificial intelligence (AI) into traditional power grid systems, making their upgrade imperative. AI-based prediction and forecasting techniques are now being utilized to improve power production, transmission, and distribution to industrial and residential consumers. This paradigm shift is driven by the development of new methods and technologies. These technologies enable faster and more accurate fault prediction and detection, leading to quicker and more effective fault removal. Therefore, incorporating AI in modern power grids is critical for ensuring their resilience, efficiency, and sustainability, ultimately contributing to a cleaner and greener energy future. This paper focuses on integrating artificial intelligence (AI) in modern power generation grids, particularly in the fourth industrial revolution (4IR) context. With the increasing complexity and demand for more efficient and reliable power systems, AI has emerged as a possible approach to solve these difficulties. For this purpose, real-time data are collected from the user side, and internal and external grid faults occurred during a time period of three years. Specifically, this research delves into using state-of-the-art machine learning hybrid models at end-user locations for fault prediction and detection in electricity grids. In this study, hybrid models with convolution neural networks (CNN) have been developed, such as CNN-RNN, CNN-GRU, and CNN-LSTM. These approaches are used to explore how these models can automatically identify and diagnose faults in real-time, leading to faster and more effective fault detection and removal with minimum losses. By leveraging AI technology, modern power grids can become more resilient, efficient, and sustainable, ultimately contributing to a cleaner and greener energy future.