2026-09-21T16:53:28Zhttps://dspace.mit.edu/server/oai/requestoai:dspace.mit.edu:1721.1/1537302024-03-14T03:02:24Zcom_1721.1_7582com_1721.1_7581col_1721.1_131023
Anthony, Brian
Zhang, Yutong
Massachusetts Institute of Technology. Department of Mechanical Engineering
2024-03-13T13:30:03Z
2024-03-13T13:30:03Z
2024-02
2024-02-15T21:17:19.917Z
https://hdl.handle.net/1721.1/153730
Optical fiber has revolutionized communication, and the market has experienced rapid growth in the last ten years. It can transmit information at high speeds with minimal loss over long distances due to its structure. Fiber extrusion, a common manufacturing method in the industry, involves controlling the fiber diameter during its formation. In this thesis, a control framework for a desk-top fiber extrusion device is developed, incorporating Deep Reinforcement Learning. By improving the mechanical design of the desk-top fiber extrusion device and implementing PID controllers over the system on the Allen-Bradley PLC, the coefficient of variation in the fiber extrusion process is reduced to 0.1. A communication path is established based on open platform communication unified architecture (OPC UA), enabling the external devices to access the data in the PLC. Using a Deep Reinforcement Learning model on a separate PC, the process is controlled to have a coefficient of variation of 0.13, with the potential to reduce the response time.
M.Eng.
Massachusetts Institute of Technology
In Copyright - Educational Use Permitted
Copyright retained by author(s)
https://rightsstatements.org/page/InC-EDU/1.0/
Development of Process Control Framework
Incorporating Deep Reinforcement Learning for Desktop
Fiber Extrusion Device via PLC Implementation
Thesis
application/pdf
Master
Master of Engineering in Advanced Manufacturing and Design
Publication
unknown
unknown
<Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="f0da2ce0-be91-418f-93be-b32fa71a0972">
<Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type>
<Title>Development of Process Control Framework
Incorporating Deep Reinforcement Learning for Desktop
Fiber Extrusion Device via PLC Implementation</Title>
<PublishedIn>
<Publication>
</Publication>
</PublishedIn>
<PublicationDate>2024-02</PublicationDate>
<Authors>
<Author>
<DisplayName>Zhang, Yutong</DisplayName>
<Affiliation>
<OrgUnit>
</OrgUnit>
</Affiliation>
</Author>
</Authors>
<Editors>
</Editors>
<Publishers>
<Publisher>
<DisplayName>Massachusetts Institute of Technology</DisplayName>
<OrgUnit />
</Publisher>
</Publishers>
<License>https://rightsstatements.org/page/InC-EDU/1.0/</License>
<Abstract>Optical fiber has revolutionized communication, and the market has experienced rapid growth in the last ten years. It can transmit information at high speeds with minimal loss over long distances due to its structure. Fiber extrusion, a common manufacturing method in the industry, involves controlling the fiber diameter during its formation. In this thesis, a control framework for a desk-top fiber extrusion device is developed, incorporating Deep Reinforcement Learning. By improving the mechanical design of the desk-top fiber extrusion device and implementing PID controllers over the system on the Allen-Bradley PLC, the coefficient of variation in the fiber extrusion process is reduced to 0.1. A communication path is established based on open platform communication unified architecture (OPC UA), enabling the external devices to access the data in the PLC. Using a Deep Reinforcement Learning model on a separate PC, the process is controlled to have a coefficient of variation of 0.13, with the potential to reduce the response time.</Abstract>
<Access xmlns="http://purl.org/coar/access_right"
>
</Access>
</Publication>