Scholarlink

<p>Decentralized generation, varying demand, and the integration of renewable energy sources are all posing challenges to conventional grid control systems, making modern power grids more complicated. Due to their inability to effectively handle dynamic and unexpected grid circumstances, these traditional systems—which are built on static rule-based methodologies create instability and inefficiency. The increased issues in power grid management brought on by dispersed renewable energy sources, variable system conditions, and rising energy demand are discussed in this study. Static rule-based optimization is the foundation of traditional power grid control systems, which find it difficult to adjust to these complicated and dynamic circumstances. A dynamic map service resource adjustment technique is put forth to get over these restrictions. It combines transformer networks with Reinforcement Learning (RL) to optimize power supply network control and boost overall efficiency. The strategy incorporates the FEDformer forecasting model, which makes precise predictions about future power demand and allows the system to react proactively to variations in energy output and consumption. In order to optimize power generation, distribution, and grid stability, a RL-based resource allocation technique is used to dynamically modify grid resources. In order to give operators a real-time picture of the grid, a Geographic Information System (GIS)-based visualization dashboard is also created. It shows important metrics including resource distribution, grid status, and dynamic modifications performed by the RL agent. The suggested approach effectively combines geographic visualization, RL, and forecasting to maximize power grid management. The system’s capacity to accurately forecast power demand, dynamically modify resources, and improve grid performance is demonstrated by the results, which also show notable gains in operational efficiency, stability, and resource utilization. Important measures like success rates and cumulative incentives show that the RL adjusts well to changing grid circumstances, maximizing grid performance and reducing energy waste.</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

Application of dynamic adjustment strategy of map service resources combined with reinforcement learning in power supply network visualization

  • 期刊论文

  • 发表日期：

    2026年7月2日

查看全文
  • Zhu Meiying,
  • Zheng Sida,
  • Hu Hao

摘要

Decentralized generation, varying demand, and the integration of renewable energy sources are all posing challenges to conventional grid control systems, making modern power grids more complicated. Due to their inability to effectively handle dynamic and unexpected grid circumstances, these traditional systems—which are built on static rule-based methodologies create instability and inefficiency. The increased issues in power grid management brought on by dispersed renewable energy sources, variable system conditions, and rising energy demand are discussed in this study. Static rule-based optimization is the foundation of traditional power grid control systems, which find it difficult to adjust to these complicated and dynamic circumstances. A dynamic map service resource adjustment technique is put forth to get over these restrictions. It combines transformer networks with Reinforcement Learning (RL) to optimize power supply network control and boost overall efficiency. The strategy incorporates the FEDformer forecasting model, which makes precise predictions about future power demand and allows the system to react proactively to variations in energy output and consumption. In order to optimize power generation, distribution, and grid stability, a RL-based resource allocation technique is used to dynamically modify grid resources. In order to give operators a real-time picture of the grid, a Geographic Information System (GIS)-based visualization dashboard is also created. It shows important metrics including resource distribution, grid status, and dynamic modifications performed by the RL agent. The suggested approach effectively combines geographic visualization, RL, and forecasting to maximize power grid management. The system’s capacity to accurately forecast power demand, dynamically modify resources, and improve grid performance is demonstrated by the results, which also show notable gains in operational efficiency, stability, and resource utilization. Important measures like success rates and cumulative incentives show that the RL adjusts well to changing grid circumstances, maximizing grid performance and reducing energy waste.