2019 Pittsburgh AISTech

Temperature Setpoint Optimization in Steel Reheat Furnaces Using Open Architecture Neural Network and Modeling Software (Room 306)

07 May 19
11:30 AM - 12:00 PM

Tracks: Hot Sheet Rolling

Speaker(s): Justin Isaacs, Neundorfer Inc.; Ricky Vesel, Griffin Open Systems

Reheat furnaces are energy intensive operations that require control finesse to maintain the proper temperatures across a range of variables to meet proper processing specifications. Energy consumption and process rejects represent significant cost savings opportunities for steel mills. This will discuss how a fully implemented neural net with real-time modeling increased quality by 90% and reduced the cost of energy consumption millions of dollars. The topic will focus on the process used to identify the efficiency opportunities, model implementation and testing, how results were sustained, as well as how software captured knowledge to offset resource loss.