Work

Proactive disturbance detection and robust control for power grid stability using anomaly-inspired machine learning and sliding mode control

Year: 2025

Type: article

Abstract: The increasing integration of electric vehicles (EVs) into power grids introduces nonlinear disturbances from unpredictable plug-in (Grid-to-Vehicle, G2V) and plug-out (Vehicle-to-Grid, V2G) events, which challenge voltage stability, frequency regulation, and power quality. This paper proposes a novel framework that combines a proactive disturbance detection scheme, inspired by anomaly detection techniques, with a Sliding Mode Control (SMC) strategy enhanced by a Gaussian Process Regression (GPR) model. EV-induced disturbances are detected in real-time using GPR’s probabilistic capabilities to capture stochastic patterns and uncertainties. These insights into disturbances enhance the capabilities of the SMC-GPR controller design, which robustly mitigates disturbances and outperforms traditional control methods. The efficacy of the proposed control design is evaluated using the IEEE 14-bus system, achieving up to a 30 % reduction in frequency deviation, 35 % improvement in power smoothness, and 40 % lower variance in oscillations compared to standard SMC and Model Predictive Control (MPC) controllers. The approach’s generalizability to other grid disturbances (e.g., renewable intermittency, cyberattacks) enhances its potential for providing modern power grid resilience and sustainability.

Source: Results in Engineering

DOI: 10.1016/j.rineng.2025.108534

Author: Bilal Khan

Institution: King Fahd University of Petroleum and Minerals

Language: English

Topic: Microgrid Control and Optimization

Subfield: Control and Systems Engineering

Field: Engineering

Domain: Physical Sciences

SDGs: Industry, innovation and infrastructure, Affordable and clean energy

Open Access status: gold

Estimated APC paid: $1,940