ZICQ
中 Log in / Sign up
Newsroom Research & Papers #Microsoft Research #AI for Power Systems #Space Weather Forecasting #Machine Learning #Grid Resilience

Microsoft Research Introduces AI System to Forecast Space Weather Risks on Power Grids

Avatar of Mr.Xu

By Mr.Xu Compiled & Reviewed by Editorial

Published: · 8 views

中文阅读 (Chinese) English Version
Co-reading Select any sentence in the article to highlight, annotate, or copy a quote card. Join discussion ↓

Summary:Microsoft Research has developed a machine learning system that forecasts space-weather risks for 66,935 substations across the continental United States, providing 30-60 minutes of advance warning before potential impacts. The system integrates solar-wind observations, Auroral Electrojet (AE) and Disturbance Storm Time (Dst) index forecasts, geological conductivity, and grid infrastructure data to generate location-specific risk estimates. This innovation enables utilities to prioritize protect


Microsoft Introduces AI System to Forecast Space Weather Risks on Power Grids

Extreme space-weather events can severely damage power systems on Earth, degrade GPS accuracy, and disrupt satellite operations. Microsoft Research has developed a novel machine learning system that can forecast space-weather risks for 66,935 substations across the continental United States, providing 30-60 minutes of advance warning before potential impacts. This system offers grid operators precise risk assessments to take preventive measures and mitigate the impact of space-weather events on power infrastructure.

Key Features and Technical Highlights

  1. End-to-End Forecasting Pipeline: The system leverages solar-wind data from the L1 Lagrange point, forecasts of the Auroral Electrojet (AE) and Disturbance Storm Time (Dst) indices, and local geological and grid infrastructure data to generate location-specific risk estimates.

  2. Multi-Stage Prediction Model:

    • Stage 1: Predict AE and Dst indices using solar-wind data and gather geological and location features for each substation.
    • Stage 2: A gradient-boosting model combines the forecasts and location-specific inputs to estimate the rate of magnetic-field change (dB/dt) associated with GIC risk.
    • Stage 3: Convert predictions into location-specific risk estimates and aggregate them into a continental-scale risk assessment.
  3. AI Agent Collaboration: The system integrates 50 AI agents to explore features, validation strategies, and model configurations, ensuring the accuracy and robustness of predictions.

  4. Performance: During the 2020-2026 evaluation period, the system detected nearly 80% of major space-weather events and outperformed traditional empirical models in most cases.

Industry Impact and Applications

  • Early Warning Tool for Grid Operators: The system provides grid operators with precise risk assessments, enabling them to prioritize engineering reviews and implement targeted protective actions, such as adjusting reactive-power reserves or reconfiguring parts of the grid.

  • AI-Powered Grid Resilience: This project is part of Microsoft Research's broader work on AI for power systems, demonstrating the potential of AI to enhance grid planning and disaster response.

Future Directions

  • Extend Forecasting Horizons: Explore temporal-transformer approaches to capture longer-term patterns in solar-wind data and extend the current 30-60-minute prediction window.

  • International Adaptation: Adapt the system to additional regions while accounting for different geological conditions and grid topologies.

  • Integration with Grid Operations: Evaluate how forecasts can support existing decision-making and engineering-review workflows before considering higher levels of automation.

  • Provide Transformer-Level Risk Assessments: Add asset-specific characteristics to move from substation-level estimates toward more granular assessments of critical equipment.

Conclusion

Microsoft's research demonstrates the significant potential of AI in space-weather forecasting, providing grid operators with more precise and timely warnings. This innovation enhances the resilience and reliability of power systems and, with further development and deployment, could offer a powerful tool for mitigating the impact of space-weather events globally.


Source: Microsoft Research AI (2026-09-30)

— END —

Tags: #Microsoft Research #AI for Power Systems #Space Weather Forecasting #Machine Learning #Grid Resilience

Editorial & Fact-Checking Note: This article is compiled from primary research, official release documentation, and source papers by the ZICQ Newsroom pipeline with automated entity verification and human editorial review. If you notice any technical inaccuracy, please submit a correction via our corrections policy or email our editorial desk directly.

What's your take on this story?
🚀 Explore & Submit Works

Community Comments

Loading live comments and annotations…