In the rapidly evolving landscape of energy management, a groundbreaking study published in the journal Engineering Science and Technology (工程科学与技术) is set to revolutionize how we integrate and utilize renewable energy sources. Led by CHEN Shi, this research introduces a novel approach to energy storage scheduling that promises to enhance the efficiency and economic viability of renewable energy integration.
At the heart of this innovation is a two-sided energy storage cooperative scheduling method designed for both transmission and distribution networks. The method leverages multi-agent attention-deep reinforcement learning to optimize the use of energy storage devices, addressing long-standing challenges in the energy sector.
Traditionally, energy storage devices have been constrained by geographical limitations and single dispatch methods, leading to low utilization efficiency. This inefficiency severely restricts the effective integration of renewable energy sources like wind and solar power. CHEN Shi’s research aims to change this by integrating energy storage resources from both the transmission grid’s power supply side and the distribution grid’s user side. “By coordinating efforts on both sides, we can explore more efficient energy storage utilization methods and enhance renewable energy integration capabilities,” explains CHEN Shi.
The proposed method establishes a cooperative alliance between power suppliers and energy storage providers, driven by the need for renewable energy integration and profit. An improved Shapley value is used to allocate additional income, providing a cooperative incentive. This alliance is crucial for balancing the interests of both parties and ensuring that the benefits are shared equitably.
One of the key innovations in this research is the use of the Multi-Agent Attention Noisy Twin Delayed Deep Deterministic Policy Gradient Algorithm (MAAN-TD3). This algorithm captures interdependencies among agents, enabling potential intent recognition and cooperative behavior perception. “The attention mechanism strengthens the focus among collaborators, significantly improving the convergence speed and optimization performance,” says CHEN Shi.
The practical implications of this research are immense. Using a modified IEEE transmission-distribution joint system as an example, the study demonstrates that energy storage idle time can be reduced by 7 hours per day, and renewable energy integration can increase by 6.81 MWh per day. This means more efficient use of existing infrastructure and a significant boost in the integration of renewable energy sources.
For the energy sector, this research opens up new avenues for optimizing energy storage and distribution. It promises to reduce operational costs, increase economic benefits for all stakeholders, and promote the integration of wind power and other renewable energy sources. As the world moves towards a more sustainable energy future, this innovative approach could play a pivotal role in making renewable energy more accessible and efficient.
The findings published in Engineering Science and Technology (工程科学与技术) highlight the potential of multi-agent attention-deep reinforcement learning in transforming energy storage and distribution. As we continue to explore and implement these technologies, the future of renewable energy integration looks brighter than ever. This research not only addresses current challenges but also paves the way for future developments in the field, making it a significant milestone in the journey towards a sustainable energy future.
