This has been one of the major projects of my PhD. I started from a low to very low base of knowledge in both methods applied, so it has taken a long span of time to get anywhere near a product of value. The goal of this article is to build models of the airflow and temperature within a clamp using historical experimental and some new experimental data and verify these against current data.
The modelling is split into two models:
- a Computational Fluid Dynamics (CFD) mechanistic model, as used in engineering sciences
- a Machine Learning (ML) statistical model, as used pretty much everywhere
The models will be compared on their accuracy, breadth of information, computational cost, ease of use, etc.
There are two ultimate use of these models. The CFD model will be used to model new designs and technologies that can be used in sugar beet storage. The ML model will hopefully be used as part of a network of temperature sensors to provide forecasts of clamp temperature and warnings of dangerous conditions.