Chen & Roald's adaptive smart meter algorithm secures grid voltage in real time
New algorithm picks minimal smart meters to certify all voltage bounds on IEEE 123 feeder.
As distributed energy resources like rooftop solar and battery storage proliferate, distribution grids face growing variability that can cause voltage violations and phase unbalance. Yet many feeders lack the sensors needed for full real-time state awareness. In their paper "Real-time Assessment of Distribution Grid Security through Adaptive Smart Meter Measurements," Jiaqi Chen and Line A. Roald tackle this by exploiting smart meters already installed across most grids.
Their approach assumes only a limited number of smart meters transmit real-time voltage magnitude data. The authors design an iterative algorithm that (1) solves optimization problems to compute worst-case voltage magnitudes given the current measurement subset, and (2) uses sensitivity information from those solutions to swap in the most informative meters. This feedback loop quickly converges to a set that certifies all voltage magnitudes remain within bounds. In simulations on the IEEE 123-bus feeder, the algorithm consistently identifies and tracks the nodes with the highest and lowest voltage, even as loads shift over time. The work was published at the 2024 IEEE 63rd Conference on Decision and Control (CDC), pp. 6493-6500.
- Algorithm iterates between worst-case voltage optimization and sensitivity-based meter selection
- Requires only a limited subset of real-time smart meter measurements, not full grid observability
- Validated on IEEE 123-bus feeder; tracks extreme voltage nodes under changing load conditions
Why It Matters
Enables utilities to boost grid security with existing smart meters, avoiding costly new sensor deployments amid rising renewables.