Grid Monitoring with Synchro-Waveform and AI Foundation Model Technologies

TL;DR

The paper tackles resilient grid monitoring in inverter-dominated systems by introducing a physics-based AI foundation model built on the Innovation Representation Model (IRM) to process high-resolution synchro-waveform data. By extracting sufficient statistics through the innovation sequence $\nu_t$ and pre-training with a Wiener-Kallianpur-Rosenblatt framework, the approach enables rapid anomaly detection, data-driven over-current protection, and efficient WMU data compression. It demonstrates substantial gains in fault-detection speed, protection reliability, and bandwidth efficiency through simulations on field data, including an IEEE 13-bus scenario with stochastic distributed generation. The work highlights the necessity of high-resolution WMU data and AI-enabled, adaptable monitoring for future grids, while acknowledging practical challenges like location-specific pre-training and transfer learning to manage complexity. Overall, the proposed IRM-based foundation model offers a cohesive pathway to enhanced situational awareness, fast fault response, and robust protection in dynamic, low-inertia power systems.

Abstract

Purpose:This article advocates for the development of a next-generation grid monitoring and control system designed for future grids dominated by inverter-based resources. Leveraging recent progress in generative artificial intelligence (AI), machine learning, and networking technology, we develop a physics-based AI foundation model with high-resolution synchro-waveform measurement technology to enhance grid resilience and reduce economic losses from outages. Methods and Results:The proposed framework adopts the AI Foundation Model paradigm, where a generative and pre-trained (GPT) foundation model extracts physical features from power system measurements, enabling adaptation to a wide range of grid operation tasks. Replacing the large language models used in popular AI foundation models, this approach is based on the Wiener-Kallianpur-Rosenblatt innovation model for power system time series, trained to capture the physical laws of power flows and sinusoidal characteristics of grid measurements. The pre-trained foundation model causally extracts sufficient statistics from grid measurement time series for various downstream applications, including anomaly detection, over-current protection, probabilistic forecasting, and data compression for streaming synchro-waveform data. Numerical simulations using field-collected data demonstrate significantly improved fault detection accuracy and detection speed. Conclusion:The future grid will be rich in inverter-based resources, making it highly dynamic, stochastic, and low inertia. This work underscores the limitations of existing Supervisory-Control-and-Data-Acquisition and Phasor-Measurement-Unit monitoring systems and advocates for AI-enabled monitoring and control with high-resolution synchro-waveform technology to provide accurate situational awareness, rapid response to faults, and robust network protection.

Figures (10)

• Figure 1: Frequency contents of various grid events, each represented by a bar covering the range of the necessary Nyquist sampling rates on the left axis. The right axis shows the necessary reporting frequency of uncompressed measurement and the required compression ratio (relative to the 256 Hz PMU reporting rate.) This Figure is adapted from WangLiuTong:21TPSSilversteinFollum:20AndersonAgrawalNess:99Perez:10.
• Figure 2: The foundation model pre-training and adaptations with innovation representation model (IRM).
• Figure 3: A Machine Learning Architecture for Innovation Representation Model.
• Figure 4: Neyman's Smooth Test (NST).
• Figure 5: Two types of erroneous protection actions caused by stochastic distributed generation (SDG). Left: Protection blinding. Right: Sympathetic tripping.