Optimization of neural network model structures for valve stiction modeling

• Main Authors: H., Zabiri, N., Mazuki
• Format: Conference or Workshop Item
• Published: 2009
• Subjects: TP Chemical technology
• Online Access: http://eprints.utp.edu.my/3733/1/S065.pdf; http://www.scopus.com/record/display.url?eid=2-s2.0-77949992585&origin=resultslist&sort=plf-f&src=s&st1=zabiri&st2=h&nlo=1&nlr=20&nls=first-t&sid=E5NmG27IsJfsII9yXMDsTvP%3a73&sot=anl&sdt=aut&sl=37&s=AU-ID%28%22Zabiri%2c+Haslinda%22+19639359300%29&relpos=0; http://eprints.utp.edu.my/3733/

Stiction is the most commonly found valve problem in the process industry. Valve stiction may cause oscillations in control loops which increases variability in product quality, accelerates equipment wear and tear, or leads to system instability. To help understand and study the behavior of sticky valve, several valve stiction models have been proposed in the literature. In this paper, a black box Neural Network-based modeling approach is proposed to model valve stiction. It is shown that with optimum model structures, performance of the developed NN stiction model is comparable to other established method. © 2009 IEEE.