Intelligent Load Forecasting and Renewable Energy Integration for Enhanced Grid Reliability

Abstract

The integration of Renewable Energy Resources (RERs) into electrical grids introduces significant challenges concerning the reliability and stability of the grid. This paper focuses on these challenges, particularly the issues of real-time load forecasting and adaptive inertia control in renewable integrated grids. A data-driven, deep learning-based approach is proposed to dynamically forecast real-time load and renewable energy generation, using the New England IEEE 39-Bus Power System as a case study. To enhance the dynamic performance of the microgrid, the paper introduces an enhanced fractional extended state observer-based linear active disturbance rejection control mechanism coupled with a feedback architecture. This control scheme aims to provide adaptive inertia to the system, thus improving its ability to handle fluctuations and intermittencies inherent in RERs. The effectiveness of the proposed controller is rigorously compared with existing approaches through simulation studies, validating its superior performance for the IEEE 39-Bus Power System under examination. To further substantiate the findings, a hardware-in-loop real-time experimental analysis is conducted using OPAL-RT hardware. This hardware-based analysis serves as a functional validation of the proposed data-driven forecasting algorithm confirming its viability to improve the grid reliability.