一种基于流形映射与非完全信息博弈的敏捷算电协同方法及系统
By employing an agile computing-power system collaborative approach based on manifold mapping and incomplete information games, the problem of spatiotemporal mismatch and power supply instability caused by transient load changes in computing power and thermodynamic response lag of generator sets was solved. This approach enables efficient and flexible collaborative control of computing power and power systems, improving prediction accuracy and equipment lifespan.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- YANTAI UNIV
- Filing Date
- 2026-05-18
- Publication Date
- 2026-07-17
AI Technical Summary
Existing computing-power collaborative prediction and control technologies are unable to meet the demands of agile power supply and stable computing power operation when faced with high concurrency and high dynamic microsecond-level computing load changes. They suffer from bottlenecks such as physical response lag, cross-domain information silos, and conflicts of interest, leading to spatiotemporal mismatch between supply and demand and system instability.
An agile computing-power system collaborative method based on manifold mapping and incomplete information game theory is adopted. Through multi-dimensional computing power manifold reconstruction, cross-medium physical topology construction, multi-agent incomplete information Nash equilibrium game theory, and feedforward Pareto optimization, efficient collaborative control of computing power and power system is achieved.
It improved the accuracy of predicting sudden changes in computing load, enhanced the agility of system response and electromagnetic physical consistency, achieved a supply and demand consensus rate of up to 98.4% and a 43.0% extension of equipment lifespan, and reduced operation and maintenance costs.
Smart Images
Figure CN122220795B_ABST