Energy Study of Drying
Using Machine Learning to Predict the Energy Consumption of an Industrial Powder Drying Process
M. El Ouasgiri (TU Delft - Electrical Engineering, Mathematics and Computer Science)
Piao Chen – Mentor (TU Delft - Statistics)
Antonis Papapantoleon – Graduation committee member (TU Delft - Applied Probability)
Robin Verhoek – Coach (Abbott)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
In this thesis, we use data science / statistical techniques to better understand the energy consumption behind a powder drying facility located in Zwolle, as part of Abbott's initiative to better manage its energy consumption. As powder drying is by far the facility's most energy intensive process, this project therefore focuses exclusively on powder drying.
The primary goal is to develop an accurate predictive model for the energy consumption, which can be used as a baseline to assess whenever or wherever the largest changes in energy efficiency took place. This can in turn be used to influence the energy policy, and may for example be used to assess what kind of retrofits can have the largest positive effects on efficiency.
Hereby we consider a variety of predictive models in increasing order of complexity. We also make our own contribution by describing a predictive model based on cluster-analysis, where we fit separate models on each cluster in order to better capture their specific patterns. It turns out that the latter approach is capable of drastically improving predictive performance.