Optimal Dynamic Network Reconfiguration Using Hybrid Quantum Deep Q-Networks
This paper introduces the application of a Hybrid Quantum Deep Q-Network (HQDQN) to solve dynamic network reconfiguration problem in distribution networks. Integrating quantum computing with deep reinforcement learning, the HQDQN is tested on the classical IEEE 33-node test feeder. It significantly outperforms traditional Deep Q-Network (DQN) models in energy efficiency. The results demonstrate the potential of quantum-enhanced machine learning algorithms to improve the operational efficiency of power grids.
Authors
- Wenzhuo Shi (ORCID: https://orcid.org/0000-0003-2311-4927)
- Yuhua Du (ORCID: https://orcid.org/0000-0001-9265-560X)
- Yigeng Huangfu (ORCID: https://orcid.org/0000-0003-0961-1214)
- Zhao Xu (ORCID: https://orcid.org/0000-0003-4480-7394)
- Huayi Wu (ORCID: https://orcid.org/0000-0003-0588-8064)
- Junyu Chen (ORCID: https://orcid.org/0000-0002-3407-5815)
- Jiaqi Ruan (ORCID: https://orcid.org/0000-0003-2584-0738)
- Yibo Ding (ORCID: https://orcid.org/0000-0003-4111-9713)
Institutions
- Hong Kong Polytechnic University (HK)
- Northwestern Polytechnical University (CN)
- Sichuan University (CN)
Publication Details
- Published
- 2024-12-10
- DOI
- https://doi.org/10.1109/icscge64239.2024.11064031
- Primary Topic
- Quantum Computing Algorithms and Architecture
- Type
- article
- Field-Weighted Citation Impact
- 0.00