A power-dominated computer-based coordinated control strategy and system

By constructing a unified model of computer-computer coupling and multi-timescale collaborative optimization, the problem of the disconnect between power grid and computing system regulation was solved, enabling precise regulation of computing load and efficient utilization of resources, thereby improving the power grid's renewable energy absorption capacity and the economic efficiency of computing infrastructure.

CN122292548APending Publication Date: 2026-06-26LIAOYANG POWER SUPPLY COMPANY OF STATE GRID LIAONING ELECTRIC POWER SUPPLY +2
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202610401129.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-30
Publication Date
2026-06-26

AI Technical Summary

Technical Problem

In existing technologies, there are information and physical barriers to the regulation of power grids and computing power systems, which makes it impossible to accurately and quickly regulate computing loads and effectively transform them into high-quality regulation resources for the power grid. There is also a lack of standardized coupling models and cross-domain collaborative optimization algorithms that uniformly describe the information attributes and physical energy consumption attributes of computing tasks.

Method used

A unified model of power grid and computing coupling is constructed. By establishing power baseline, regulation capacity, response rate and quality of service constraint parameters, the power grid operation status information and computing node information are obtained. A multi-timescale collaborative optimization problem is constructed to generate the optimal collaborative control strategy. Deep collaborative control between the power grid and computing system is realized through a central collaborative decision-making platform, regional computing power aggregation agents and local resource control units.

Benefits of technology

It enables dynamic collaboration between the power grid and the computing network, transforming computing load into a virtual, flexible resource that the power grid can perceive in real time and predict accurately. This enhances the new power system's ability to absorb new energy sources and its real-time power balance capabilities, reduces the operating costs of computing infrastructure, and promotes the deep integration of the energy and information industries.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122292548A_ABST
    Figure CN122292548A_ABST
Patent Text Reader

Abstract

This invention discloses a power-led computer-computer collaborative control strategy method and system. First, a unified computer-computer coupling model is established, standardizing and quantifying the information attributes and energy consumption characteristics of computing tasks. Based on this model, through multi-timescale collaborative optimization, a control strategy is generated to guide the computing load through time-shifting and spatial migration, enabling it to proactively adapt to the grid operating state, achieving peak shaving and valley filling, promoting the consumption of new energy sources, and achieving rapid frequency response. The corresponding system adopts a three-layer collaborative architecture of "center-edge-local," including a central collaborative decision-making platform, regional computing power aggregation agents, and local resource control units, forming a closed-loop control from global optimization decision-making to precise local execution. This invention significantly improves the safety and economy of grid operation while ensuring the quality of computing power services, and provides an innovative technological path for the computing power industry to reduce energy costs and improve its green level.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • A multi-element power grid information interaction system and method based on edge computing architecture

    CN109638964B

  • Multi-node computing network energy collaborative planning method for computing power task multi-time-space-dimension scheduling

    CN120031302A