Scientific Machine Learning-assisted Model Discovery from Telemetry Data

TL;DR

This paper introduces Dyad Model Discovery, a semi-automated AI-assisted approach that enhances physical models with data-driven symbolic expressions, improving digital twin accuracy for HVAC and refrigeration systems.

Contribution

It presents the first AI-assisted engineering workflow that combines symbolic data-driven model augmentation with an engineer-in-the-loop process.

Findings

Improved predictive performance of a refrigeration unit digital twin.

First demonstration of AI-assisted model discovery in engineering design.

Effective integration of data-driven expressions with physical models.

Abstract

Calibration of dynamic models to data is an important step in building building digital twins of HVAC equipment, thermal loads and control systems. Sometimes, when a model fails to calibrate to data, a possible cause is that the model has made too many sim- plifying assumptions and is missing physics. In this paper we propose a semi-automated approach, called Dyad Model Discovery, that can augment the physical equations of the model with symbolic expressions discovered from the data. We demonstrate this method on a digital twin of a transportation refrigeration unit to improve its predictive performance, trained using telemetry data. An engineer-in-the-loop workflow is proposed, which provides suggestions to the user which can then be accepted or rejected. This is the first AI-assisted engineering design workflow to our knowledge.