Hybrid Deep Reinforcement Learning – Complex Network Theory Approach for Self-healing Restoration Strategy to Minimize Energy Not Supplied (ENS) in Transmission System

Read full text Summarise with VIRAT
Authors: Ontoseno Penangsang, Rony Seto Wibowo, Ni Ketut Aryani, Muhammad Haidar Syahab, Omar Fabian, Keisya Fatika Trisnawati
Cited by: 0
References: 29

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Transmission system restoration following major blackouts requires complex sequential decision-making under strict operational constraints.Traditional heuristic approaches lack adaptability, while pure learning-based methods struggle with large combinatorial action spaces.This paper proposes a Hybrid Deep Reinforcement Learning-Complex Network Theory (DRL-CNT) framework integrating Double Deep Q-Network (DDQN) with electrical betweenness centrality for self-healing restoration.The restoration problem is formulated as a Markov Decision Process with Energy Not Supplied (ENS) minimization as the objective.Electrical betweenness centrality guides DDQN action selection, reducing exploration burden while a stage ladder strategy ensures operational feasibility.The method is validated on IEEE 39-bus and SulutGo 150 kV systems through Python-DIgSILENT PowerFactory cosimulation.DDQN-CNT achieves mean ENS of 5993.85MWh and 963.62 MWh respectively, outperforming DQN-CNT by 2.30% and 12.28%, and CNT-only by 4.94% and 7.71%.Sensitivity analyses confirm robust performance under diverse operational conditions, with advantages more pronounced on larger real-world systems.

Fields of Study & Subjects

Power System Optimization and Stability Optimal Power Flow Distribution Electric Power System Optimization

Publication details

Journal / source
International journal of intelligent engineering and systems
Volume
19
Issue
4
Pages
670–685
Publication date
23 March 2026
Language
EN