Image by Phonlamai Photo/Shutterstock

Solar power generation largely depends on weather conditions. When generation deviates from the planned output, the electricity market imposes penalty fees called "imbalance penalties." Researchers at University of Tsukuba have developed an artificial intelligence (AI)-based method that optimizes the operation of solar power generation and battery storage systems, reducing imbalance penalties by up to 47% compared to conventional methods.

Tsukuba, Japan—The growing penetration of distributed renewable energy resources necessitates more intelligent and adaptive energy management strategies than are currently available. In electricity markets, transactions are based on the generation volumes planned for the following day, which are submitted by power producers. However, solar power generation is highly susceptible to weather conditions. Discrepancies between the planned and actual supply volumes disrupt the overall market supply-demand balance, leading to penalty fees known as "imbalance penalties." Although computational methods can control this balance to some extent, they cannot adequately reflect real-world uncertainties such as sudden weather changes and complex market dynamics.

Researchers at University of Tsukuba have developed a method that optimizes the operation of solar power generation and battery storage systems while conforming to market rules. The method relies on deep reinforcement learning-based AI, which can handle problems involving uncertainty. In simulation results on actual market data, this method reduced the imbalance penalties by approximately 47% and 26% compared to conventional control methods and other deep reinforcement learning models, respectively. Furthermore, it maintained stable profits throughout the four seasons.

This research will contribute to a mechanism that improves profitability, avoids imbalance penalties, and provides a stable supply of renewable energy to the market. Furthermore, it may lay the foundation for a system that treats aggregated household power sources—such as storage batteries and electric vehicles—as a new power source, delivering societal benefits such as stabilized electricity prices and a reduced risk of power outages.

###
This work was supported by JSPS KAKENHI, Grant Number JP23K13306 and the Power Data Utilization Support Program (2025), the Green Investment Promotion Organization.

Original Paper

Title of original paper:

Imbalance-Aware Scheduling for PV Battery Storage Systems Using Deep Reinforcement Learning

Journal:

IEEE ACCESS

DOI:

10.1109/ACCESS.2025.3615960

Correspondence

Assistant Professor KODAIRA Daisuke
Institute of Systems and Information Engineering, University of Tsukuba

OSONE Yuki
Degree Programs in Systems and Information Engineering, Graduate School of Science and Technology

Related Link

Institute of Systems and Information Engineering