Session 5: Exploration of PV-Energy Storage Optimal Control and Computing-Electricity Coordination Optimization Technologies in Multi-Energy Microgrids

Exploration of PV-Energy Storage Optimal Control and Computing-Electricity Coordination Optimization Technologies in Multi-Energy Microgrids
“多能源微电网中光伏储能优化控制与算电协同优化技术探索”

With the rapid development of artificial intelligence technology and the explosive growth of global computing power demand, and under the demand for deep integration of new energy and artificial intelligence, the Multi-Energy Microgrid oriented towards Computing-Electricity Coordinated Operation is driving a profound systemic change as the forefront direction for high-quality development of energy systems in the new era. However, as a coupled system of heterogeneous resources of electricity and computing power, microgrids face dual challenges of strong random fluctuations in new energy generation and spatiotemporal migration characteristics of computing load. Therefore, this special session takes the integration of computing and electricity into microgrids as the starting point, aiming to explore a full chain solution from bottom level prediction, planning and configuration to upper level collaborative optimization control. Focus on predicting the power generation of new energy sources; Photovoltaic energy storage collaborative optimization; And precise modeling and flexibility analysis technology for computing power load. On this basis, the optimization design technology of multi-source energy supply system in computing power center and the intelligent optimization technology of source grid load storage for coordinated operation of computing power in time and space are discussed, ultimately focusing on the safety optimization control technology of microgrid in the scenario of computing power coordination. This session advocates the integration of physical mechanisms and data-driven approaches, exploring the adaptive decision-making of technologies such as model predictive control, deep reinforcement learning, and multi-agent systems in strong stochastic scenarios, forming a theoretical and technical solution that integrates "prediction-design-optimization-control". The session involves multiple interdisciplinary fields such as control, electrical engineering, and artificial intelligence, and is of great significance in promoting China's green energy transformation, low-carbon computing infrastructure, and highly elastic autonomous operation of microgrids.

随着人工智能技术的快速发展与全球算力需求的爆发式增长，在新能源与人工智能深度融合的需求下，面向算力-电力协同运行的多能源微电网作为新时代能源系统高质量发展的前沿方向，正推动着一场深彻的系统性变革。然而，作为电力与算力异质资源的耦合系统，微电网面临着新能源发电的强随机波动性与算力负荷时空迁移特性的双重考验。为此，本分论坛以算电融合微电网为切入点，旨在探索从底层预测、规划配置到上层协同优化控制的全链路解决方案。重点关注新能源发电功率预测；光伏—储能协同优化；以及算力负荷精准建模与灵活性分析技术。在此基础上，探讨算力中心多源供能系统优化设计技术与算电时空协同运行的源网荷储智能优化技术，最终聚焦于算电协同场景下的微电网安全优化控制技术。本论坛倡导物理机理与数据驱动相融合，探索模型预测控制、深度强化学习及多智能体等技术在强随机场景下的自适应决策，形成集“预测-设计-优化-控制”于一体的理论和技术方案。论坛涉及控制、电气、人工智能等多个学科交叉，对促进我国能源绿色转型、推动算力基础设施低碳化及微电网高弹性自治运行具有重要意义。

Topics (Including but not limited to)

• 1. Renewable Energy Power Forecasting, Grid-connection Control, and High-efficiency Accommodation Technologies;
  新能源发电功率预测、并网控制与高效消纳技术
• 2. Coordinated Optimal Scheduling and Capacity Configuration Methods for Photovoltaic-Energy Storage Systems;
  光伏—储能系统协同优化调度与容量配置方法
• 3. Accurate Forecasting, Modeling, and Flexibility Analysis Technologies for Computing Loads;
  算力负荷精准预测、建模与灵活性分析技术
• 4. Optimal Design Technologies for Multi-source Energy Supply Systems in Computing Centers;
  算力中心多源供能系统优化设计技术
• 5. Secure Optimal Control of Microgrids under Computing-Electricity Coordination;
  算电协同场景下的微电网安全优化控制技术
• 6. Intelligent Optimization Technologies for Source-Grid-Load-Storage in Spatiotemporal Computing-Electricity Coordinated Operation.
  算电时空协同运行的源网荷储智能优化技术