Twin Delayed Deep Deterministic Policy Gradient‐Based Load Frequency Control for Interconnected Hybrid Microgrids With Renewable Energy Integration Using a Tilt–Integral–Derivative Controller
Artificial Intelligence for Engineering · 2025
DOI: 10.1049/aie2.70006 ↗Kurz gefasst
Fachartikel in der Zeitschrift Artificial Intelligence for Engineering, 2025. Verfasst von Robert Bhadra Naorem und 2 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
This paper presents a tilt–integral–derivative (TID) controller for load–frequency regulation in a two‐area interconnected hybrid microgrid with renewable energy integration. The controller parameters are optimised using the twin delayed deep deterministic policy gradient (TD3) method, an advanced actor‐critic reinforcement learning framework designed to address function approximation errors in deterministic policy learning, especially within continuous action spaces. A centralised‐training and decentralised‐execution paradigm is adopted to enable the agent to capture global system behaviour while applying localised control within each area. The proposed TD3–TID approach is trained under a wide range of operating conditions, including random and time‐varying load disturbances, to reflect practical microgrid uncertainties. The approach employs delayed policy updates, twin critic networks, and clipped noise regularisation, ensuring better learning stability and policy reliability in continuous control tasks. The comparative results show that the TD3–TID controller provides improved suppression of frequency deviations, reduced tie‐line power oscillations and enhanced transient performance compared to conventional and optimisation‐based controllers. These results highlight its suitability for real‐time implementation in modern distributed energy systems.
Zitieren
APA
Naorem, R. B., Das, D. C., & Roy, P. (2025). Twin Delayed Deep Deterministic Policy Gradient‐Based Load Frequency Control for Interconnected Hybrid Microgrids With Renewable Energy Integration Using a Tilt–Integral–Derivative Controller. Artificial Intelligence for Engineering. https://doi.org/10.1049/aie2.70006
Vancouver
Naorem RB, Das DC, Roy P. Twin Delayed Deep Deterministic Policy Gradient‐Based Load Frequency Control for Interconnected Hybrid Microgrids With Renewable Energy Integration Using a Tilt–Integral–Derivative Controller. Artificial Intelligence for Engineering. 2025. doi:10.1049/aie2.70006
BibTeX
@article{Naorem2025,
author = {Naorem, Robert Bhadra and Das, Dulal Chandra and Roy, Prasanta},
title = {Twin Delayed Deep Deterministic Policy Gradient‐Based Load Frequency Control for Interconnected Hybrid Microgrids With Renewable Energy Integration Using a Tilt–Integral–Derivative Controller},
journal = {Artificial Intelligence for Engineering},
year = {2025},
doi = {10.1049/aie2.70006}
}
EUVIMED