MatAstraNext-Generation PMU Synchrophasor Analytics for Stable Power Grids
MatAstra turns raw phasor measurements into located, quantified oscillations operators can act on before they escalate. Two tools and one engineering service, one discipline: total clarity.
Two Tools
01 — The Situation
Undetected oscillations cascade into blackouts.
The physics
Every large power system hums — a slow, rhythmic exchange of energy between generators. This swaying motion could be an electromechanical oscillation, forced oscillation and Inverter Based Resource (IBR) induced oscillation. Usually it damps out on its own.
The problem
Phasor Measurement Units stream thirty or more time-synchronized measurements every second. The early signatures of instability are already in that record — yet in most control rooms, most of it is never read.
The data
The record is there, second by second, across the whole interconnection. What is missing is not measurement but reading — the ability to turn that stream into a located, understood oscillation before it matters.
02 — The Method
From measurement to mitigation, in four movements.
03 — Instrument I
Power Sentinel: the watch that never sleeps.
Power Sentinel
Real-time power grid oscillation monitoring — a clean yet powerful solution for unsupervised monitoring that lets operators mitigate oscillations before they escalate.
Multi-channel mode shapes computed in real time, continuously, in the control room: a watch instrument that never blinks, reporting mode, magnitude, and source while there is still time to act. Where other tools stop at near-real-time, single-channel views, Power Sentinel reads the whole grid as it moves.
Real-Time Precision
Multi-channel mode shapes, oscillation analysis, and source location resolved in fractions of a second — continuously, without supervision.
Minimalist Interface
A clean, intuitive layout with a single-button workflow — engineered for clarity, speed, and operator confidence under pressure.
Effortless Scalability
Handles tens to hundreds of PMU channels with consistent low-latency performance across wide-area grids.
True Real-Time Oscillation Surveillance
Beyond Near-Real-Time WAMS Analytics
| Feature / Category | Power Sentinel | Other Vendors |
|---|---|---|
| Mode-Shape Calculation | Provide a solid foundation for ambient analysis, oscillation detection and source localization. | N/A |
| Ambient Listening | Eigenvalue analysis across all channel data to characterize system natural mode. | N/A |
| Oscillation Detection | Capture a full view of grid oscillation when it happens. | Relies on single-channel waveform analysis. |
| Source Localization | Based on advanced filtering techniques, fast and accurate. | Based on conventional bandpass filters or FFT methods, time consuming. |
| Scalability | Easily scalable from tens to hundreds of channels with consistent low latency. | Computational burden becomes unmanageable as the number of system PMUs grows. |
04 — Instrument II
PMUmate: the long read of an event.
PMUmate
The same engine as Power Sentinel, offline — unlock the full potential of Phasor Measurement Unit data with multi-channel eigenvalue analysis and high-resolution mode-shape identification.
Designed for rigorous post-event investigation rather than simple validation, PMUmate processes historical PMU measurement archives to reveal the power system's true dynamic behavior—including its oscillatory modes, damping characteristics, and core stability attributes. Serving as the definitive offline analytical benchmark for operational decision-making, PMUmate answers far more than what occurred. It precisely identifies where disturbances originated, at what frequency oscillations evolved, and why system behavior degraded—delivering complete, traceable, and technically rigorous grid dynamic analysis.
It is also the low-risk way to begin: run the algorithm on your own historical events and confirm it on your grid before Power Sentinel goes onto the operations floor.
Mode-Shape Analysis
Extract and visualize mode shapes from synchronized PMU data to uncover system-wide dynamic behavior that no single-channel view can reveal.
Automatic Detection
Intelligent algorithms estimate system order and identify dominant modes — no manual tuning required, no expert configuration in the loop.
Advanced Processing
Careful preprocessing isolates true oscillatory content from noise and ambient variation, then locates the source of the oscillation.
05 — We Also Provide
PMU Dynamic Analysis Service
From ambient to event forensics.
Engineering Service · On Your PMU Archive
- Ambient
- Ring-Down
- Event Forensics
Your PMU archive in. Audit-ready engineering findings out.
Key Deliverables
- 01
Modal Parameters
Frequency, damping and mode-shape from ambient and ring-down data.
- 02
Oscillation Assessment
Oscillation classification and source-localization assessment.
- 03
Long-Term Trending
System dynamic behavior tracked with the changing generation mix.
- 04
Event Forensics
Post-disturbance timeline reconstruction and ring-down forensics.
- 05
Model Validation
Measured-vs-simulation comparison for dynamic model validation (MOD-033).
- 06
Data Quality
PMU archive data-quality anomaly identification.
- 07
Technical Report
Audit-ready report with plots and engineering recommendations.
06 — Who We Are
Engineers and mathematicians.
A multidisciplinary team built to translate rigorous analysis into dependable grid intelligence.
We are mathematics and engineering veterans with decades of practice on the grid — analysts who have spent careers inside the methods, not a software team that later learned power systems.
The practice is specialized to the extreme: eigenvalue analysis and algorithm refinement pushed until the method holds under real PMU noise; wide-area power-system grid analysis; multi-channel time-series signal processing; and the software that turns those methods into instruments operators can run.
Use math tools to solve real-world problems.
Power Systems
Grid dynamics, stability, and operational context.
Signal Processing
Extracting trustworthy insight from complex PMU measurements.
Software Development
Turning advanced methods into reliable, usable instruments.
Mathematics
Models and algorithms grounded in rigorous quantitative reasoning.
07 — Questions
What operators ask before they trust an instrument.
High-fidelity synchrophasor data streamed from Phasor Measurement Units (PMUs) — typically 30 or more time-synchronized samples per second per channel, across tens to hundreds of channels covering a wide-area interconnection.
Power Sentinel runs continuously, in real time, computing multi-channel mode shapes so operators can catch and mitigate oscillations as they emerge. PMUmate applies the same analytical core offline, for deep post-event investigation of historical PMU archives with full analytical depth.
No. MatAstra estimates system order and identifies dominant modes automatically. The workflow is deliberately spare — in Power Sentinel, a single button — so the instrument informs decisions rather than demanding configuration.
Our demonstration uses Case 1 from the 2021 IEEE-NASPI Oscillation Source-Location Contest as the primary test dataset. This case features a dominant oscillatory mode at 0.82 Hz, with the true oscillation source located at Bus 1431. The dataset and supporting materials are publicly available at: https://web.eecs.utk.edu/~kaisun/Oscillation/2021Contest/
The identified oscillation modes from our pipeline have been cross-validated across multiple public reference datasets, including:
- Grid Event Signature Library (GESL) from Oak Ridge National Laboratory: https://gsl.ornl.gov/
- ISO-New-England real-world sustained-oscillation PMU event library: https://web.eecs.utk.edu/~kaisun/Oscillation/actualcases.html
Grid operators, utilities, ISOs and RTOs, and power systems engineers responsible for the stability of the interconnected grid — anyone who stakes operational decisions on measurement.
The Invitation
Read your own grid.
The data your grid produces is already telling you what it needs. Most of it simply isn't being read. Power Sentinel and PMUmate change that — one for the quiet, continuous watch; one for the careful, post-event investigation. Together they turn thirty measurements a second into the one thing that matters: time to act.
Request a demonstration