End User
SWW
Description
The AI-State Estimation service delivers high-resolution near-real-time representations of the operational state of LV/MV distribution grids. By combining AI-generated load profiles for unmetered assets with measured feeder-level SCADA data and a full power-flow solver, the service computes voltage magnitudes, angles, active/reactive power flows, and currents for every node and branch including those without direct measurement.
Core Capabilities
Monitoring & Anomaly Detection
Predictive & Prescriptive Analytics
Business Need
Regulatory demands such as Redispatch 2.0 require continuous monitoring of grid states, but sparse LV metering infrastructure means most nodes lack direct observability. Without accurate state estimation, DSOs risk missing voltage-quality issues and inaccurate flexibility dispatch. The service bridges this observability gap by creating virtual sensor estimates for unmetered nodes, enabling grid operators to monitor voltage stability and validate constraints across the full network.
Key Performance Indicators
Voltage magnitude MAE and power-flow MAPE at held-out validation meters
Confidence-level calibration of estimated vs. measured node values
Solver convergence rate: >95% of cycles converging successfully
Median response time: 1-4 seconds; 95th percentile latency <8 seconds
Data Provided
GridState JSON: NodeStates (voltage magnitude, angle, confidence, measured/AI-estimated flag) and BranchFlows (active/reactive power, current, direction)
QualityIndicators: measurement coverage percentage, solver convergence status
Inputs: GIS topology files, SCADA/PQM measurements, smart meter data, AI asset profiles, weather and calendar data
TEF
TEF DSO