Deep Reinforcement Learning-Based Operation of Distribution Systems Using Surrogate Model

Article Engineering, Electrical & Electronic

A Dynamic Internal Trading Price Strategy for Networked Microgrids: A Deep Reinforcement Learning-Based Game-Theoretic Approach

Van-Hai Bui et al.

Summary: In this study, a novel two-step optimization model is developed to maximize the internal power trading in a distribution network comprising multiple microgrids. The proposed model combines soft actor-critic method with entropy-regularized reinforcement learning to handle large state and action spaces. Experimental results demonstrate the superiority of the proposed model in increasing internal power trading and improving retailer's profit.

IEEE TRANSACTIONS ON SMART GRID (2022)

Add to Collection

Proceedings Paper Engineering, Electrical & Electronic

Deep Reinforcement Learning in Power Distribution Systems: Overview, Challenges, and Opportunities

Yuanqi Gao et al.

Summary: This paper reviews the growing literature on applications of reinforcement learning in power distribution systems, outlining challenges such as safety, robustness, interpretability, and sample efficiency. It provides research opportunities to address these challenges.

2021 IEEE POWER & ENERGY SOCIETY INNOVATIVE SMART GRID TECHNOLOGIES CONFERENCE (ISGT) (2021)

Add to Collection

Article Engineering, Electrical & Electronic

An Enhanced IEEE 33 Bus Benchmark Test System for Distribution System Studies

Sarineh Hacopian Dolatabadi et al.

Summary: The proposed enhanced test system for distribution systems planning and operation studies provides a comprehensive benchmark for future researches, by incorporating both balanced and unbalanced three-phase power systems and detailed information on distributed and renewable generation, reactive power compensation, and load profiles.

IEEE TRANSACTIONS ON POWER SYSTEMS (2021)

Add to Collection

Article Computer Science, Information Systems

Surrogate-Assisted Multi-Objective Probabilistic Optimal Power Flow for Distribution Network With Photovoltaic Generation and Electric Vehicles

Chitchai Srithapon et al.

Summary: This paper proposes a surrogate-assisted multi-objective probabilistic optimal power flow method that reduces computational time by using a deep learning surrogate model, effectively addressing uncertainties in solar, electric vehicle, and home appliance loads in distribution networks.

IEEE ACCESS (2021)

Add to Collection

Article Engineering, Electrical & Electronic