Showing 1–1 of 1 results for author: Spielberg, S
-
Deep Reinforcement Learning for Process Control: A Primer for Beginners
Authors:
Steven Spielberg,
Aditya Tulsyan,
Nathan P. Lawrence,
Philip D Loewen,
R. Bhushan Gopaluni
Abstract:
Advanced model-based controllers are well established in process industries. However, such controllers require regular maintenance to maintain acceptable performance. It is a common practice to monitor controller performance continuously and to initiate a remedial model re-identification procedure in the event of performance degradation. Such procedures are typically complicated and resource-inten…
▽ More
Advanced model-based controllers are well established in process industries. However, such controllers require regular maintenance to maintain acceptable performance. It is a common practice to monitor controller performance continuously and to initiate a remedial model re-identification procedure in the event of performance degradation. Such procedures are typically complicated and resource-intensive, and they often cause costly interruptions to normal operations. In this paper, we exploit recent developments in reinforcement learning and deep learning to develop a novel adaptive, model-free controller for general discrete-time processes. The DRL controller we propose is a data-based controller that learns the control policy in real time by merely interacting with the process. The effectiveness and benefits of the DRL controller are demonstrated through many simulations.
△ Less
Submitted 11 April, 2020;
originally announced April 2020.
