Scholarlink

<p>The implementation of model predictive control (MPC) with rolling optimization and feedback correction is a crucial technique for achieving multi-timescale optimal scheduling in integrated energy systems. However, the computational burden of centralized MPC often hinders its online application in large-scale IES. This paper proposes a novel distributed model predictive control (DMPC) framework that is specifically tailored for the multi-energy nature of IES. The key novelty lies in its hybrid multi- -time scale decomposition architecture, which coordinates electrical, thermal, and gaseous energy subsystems according to their distinct dynamic characteristics. Furthermore, a Nash-optimality based coordination mechanism is employed in the real-time adjustment phase to efficiently manage prediction uncertainties. Simulation results demonstrate that, compared to centralized MPC, the proposed strategy not only reduces computational time by approximately 13% but also significantly mitigates power fluctuations, leading to a 1.5% improvement in operational economy. This method advances DMPC beyond a generic decomposition tool, presenting it as a physics-informed control paradigm for complex IES.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing Grid Flexibility: Distributed MPC for Multi-Time Scale Energy Optimization

  • 期刊论文

  • 发表日期：

    2026年3月19日

查看全文
  • Fan Rui,
  • Wang Xiang,
  • Li Honghao,
  • Ding Changxin,
  • Wang Lei

摘要

The implementation of model predictive control (MPC) with rolling optimization and feedback correction is a crucial technique for achieving multi-timescale optimal scheduling in integrated energy systems. However, the computational burden of centralized MPC often hinders its online application in large-scale IES. This paper proposes a novel distributed model predictive control (DMPC) framework that is specifically tailored for the multi-energy nature of IES. The key novelty lies in its hybrid multi- -time scale decomposition architecture, which coordinates electrical, thermal, and gaseous energy subsystems according to their distinct dynamic characteristics. Furthermore, a Nash-optimality based coordination mechanism is employed in the real-time adjustment phase to efficiently manage prediction uncertainties. Simulation results demonstrate that, compared to centralized MPC, the proposed strategy not only reduces computational time by approximately 13% but also significantly mitigates power fluctuations, leading to a 1.5% improvement in operational economy. This method advances DMPC beyond a generic decomposition tool, presenting it as a physics-informed control paradigm for complex IES.