Tuning simulations against measurements
Could the gap between simulation and measurement close on its own?
When a simulation drifts away from measurements, we think parameters often get re-tuned without a clear record of who changed what. We want to study calibration that keeps the uncertainty and a full record. If we share those records, others can reproduce the same result.
Measurements and a simulation model
We aim to connect and clean both, then have agents calibrate the model and verify each other.
Calibrated parameters with uncertainty ranges
| Field | Physics |
|---|---|
| Status | Planned |
| Input | Measurements · simulation model |
| Output | Calibrated parameters |
| Delivery | Joint research |