Physics-Guided Data-Driven Modeling for Control in Additive Manufacturing
This event is free and open to the publicAdd to Google Calendar
Abstract: Additive Manufacturing (AM) techniques have become attractive for fabricating parts ranging from biological tissues to aircraft components. A key challenge in these techniques is controlling the quality of parts to ensure high accuracy, throughput and repeatability. Given the fast and high-dimensional multi- scale dynamics associated with many AM processes, it is often difficult to physically model these processes for control. In this talk, I examine how data obtained from these processes may be used to implement feed-forward and feedback control. In particular, I discuss the implementation of geometry control in a droplet-based AM process, namely, Inkjet 3D Printing, which is commonly used for fabricating polymer parts. Because of the complex droplet dynamics at play in Inkjet 3D printing, it is challenging to model the evolution of parts at scale in this process. I show that by imbuing domain knowledge (physical insights) into a graph-based neural network model structure, the complex dynamics can be learned in an interpretable manner with few data points. Notably, the learned model can predict evolution of part geometries other than those it has encountered in training. Finally, the interpretability and small data requirement of the model allows for certification of vital properties such as system stability; and enables offline and online predictive control.
BIO: Uduak Inyang-Udoh is an Assistant Professor in the Department of Mechanical Engineering at the University of Michigan. He was previously a Postdoc Research Associate at Purdue University. Prior to his postdoc at Purdue, Uduak completed his PhD at Rensselaer Polytechnic Institute in 2021. His research interests lie at the intersection of control theory, graph theory and machine learning, especially for application in data-rich manufacturing, thermal and energy storage systems. Uduak was named a Rising Star by the ASME Dynamic Systems and Control Division at the 2022 Modeling, Estimation and Control Conference for his research in physics- guided machine learning for controls.
***Event will take place in hybrid format. The location for in-person attendance will be room 1500 EECS. Attendance will also be possible via Zoom. The Zoom link and password will be distributed to the Controls Group e-mail list-serv.
To join this list-serv, please send an (empty) email message to [email protected] with the word “subscribe” in the subject line. To cancel your subscription, send an (empty) email to [email protected] with the word “unsubscribe” in the subject line. Zoom information is also available upon request to Shelly Feldkamp at [email protected].