GridForge research group
Cutting-edge research for on the low-voltage grid.
A standing research group at Monash, building cumulative tools and methods for the low-voltage distribution grid. Each cohort picks up where the last left off — so individual theses compound into one body of work, hosted on a shared platform.
The methods.
Five problems on the LV grid, attacked from smart-meter data alone — no extra hardware, no extra site visits. Each method is built into the shared platform so the next cohort starts where the last left off.
01 / Topology estimation
Who is connected to whom?
Inferring the connectivity structure of an LV network — which customers share conductors, where branches join, where boundaries lie — from smart-meter data alone.
02 / Phase grouping
Which customers share a phase?
Clustering customers by inferred phase membership from voltage time-series correlation — without measurements at the substation or prior knowledge of the wiring.
03 / Impedance estimation
How does current flow?
Recovering conductor resistance and reactance from observed voltages and powers. Underpins load flow, loss accounting, and cross-validation of topology.
04 / Transformer monitoring
What is the transformer doing?
Estimating transformer loading and stress from downstream smart-meter readings — virtual sensing, no instrumentation at the transformer itself.
05 / Situation analysis
What if things change?
Stress-testing networks against rooftop solar uptake, EV charging patterns, and customer-mix shifts. Voltages and loadings under hypothetical futures.
Each year picks up where the last left off.
GridForge runs on a rolling cohort. The 2025 founding cohort built the platform from scratch — a Streamlit GUI wired to a Python analysis core and an OpenDSS simulation backend. The 2026 cohort is extending it with topology evaluation, database-backed storage, and new metrics for comparing topology estimates.
A shared workbench means new students see what's already been built, and outputs from one method become inputs to the next.
Meet the cohorts- 2027 Roadmap · impedance estimation, real-feeder pilots Planned
- 2026 Digital twin · AI topology evaluation · topology metrics Active · 3 groups
- 2025 Platform pilot · Streamlit + OpenDSS · phase grouping Founding
The grid most analytics ignore.
The medium- and high-voltage grid is well-instrumented. The low-voltage grid — where customers live — is mostly dark. We work on the dark side.