INTELLIGENT GASOLINE BLENDING CONTROL SYSTEM IN REAL TIME TAKING INTO ACCOUNT PARAMETRIC UNCERTAINTY
UDC: 004.942+665.73
DOI: 10.33285/0132-2222-2021-7(576)-28-36
Authors:
GORBUNOV SERGEY SERGEEVICH 1,
KOSTANDYAN ARTUR VALERIEVICH 2,
EGOROV ALEKSANDR FEDOROVICH 3,
SIDOROV VALERY VASILIEVICH 4,
ALEKSANYAN ASHOT ARSENOVICH 1
1 MCE-Engineering, Moscow, Russian Federation
2 KSIMATIC, Moscow, Russian Federation
3 Mendeleev University of Chemical Technology of Russia, Moscow, Russian Federation
4 National University of Oil and Gas "Gubkin University", Moscow, Russian Federation
Keywords: gasoline mixing, intelligent system, optimization model, gasoline mixing recipe, mathematical description, optimality criterion, limitation, parametric uncertainty, neural network
Annotation:
The process of gasoline blending is an important final stage in the overall technological chain of gasoline production. The technological costs of compounding are determined by the efficiency of maintaining the hydrodynamic mode of mixing gasoline components, control, optimization of the fuels mixing components, optimal control and regulation. The development of a mathematical model, the formulation of the problem of optimization and control over gasoline mixing in real time (online mode) is one of the important stages in the creation of a distributed control system (DCS). The paper presents a mathematical model of gasolines mixing online optimization, taking into account the conditions of uncertainty of the technological regime parameters. The presented mathematical model takes into account the parametric uncertainty of the gasoline mixing process in real time when constructing control algorithms in the DCS. A neural network model for solving the problem of forecasting and optimizing a gasoline blending control system in real time is presented.
Bibliography:
1. White J., Hall F. Gasoline blending optimization cuts use of expensive components // Oil & Gas J. - 1992. - Vol. 90, Issue 44. - P. 81-84.
2. Michalek T.F., Nordeen K., Rys R. Using a relational database for blend optimization // Hydrocarbon Processing. - 1994. - Vol. 73, Issue 9. - P. 47-49.
3. Agrawal S.S. Integrate blending control, optimization and planning // Hydrocarbon Processing. - 1995. - Vol. 74, Issue 8. - P. 129-139.
4. Diaz A., Barsamian J.A. Meet changing fuel requirements with online blend optimization // Hydrocarbon Processing. - 1996. - Vol. 75, Issue 2. - P. 71-76.
5. Blend-control system all but eliminates reblends for Canadian refiner / P.J. Vermeer, C.C. Pedersen, W.M. Canney, J.S. Ayala // Oil & Gas J. - 1997. - Vol. 95, Issue 30. - P. 74-79.
6. Naman B.T. Linear models help refiners develop RFG recipes // Oil & Gas J. - 1999. - Vol. 97, Issue 8. - P. 64-66.
7. Chen J., Wang N., Wang S.Q. Neuro-internal model based optimal control for gasoline blending process // Control Theory and Application. - 1999. - Vol. 16, Aug. - P. 532-535.
8. Zhang Y., Monder D., Forbes J.F. Real-time optimization under parametric uncertainty: a probability constrained approach // J. of Process Control. - 2002. - Vol. 12, Issue 3. - P. 373-389. - DOI: 10.1016/S0959-1524(01)00047-6
9. Программный комплекс оптимального планирования и оптимизации рецептур смешения бензинов и мазутов / С.С. Горбунов, А.А. Алексанян, В.А. Костандян, А.Ф. Егоров // Нефтепереработка и нефтехимия. Научно-техн. достижения и передовой опыт. - 2019. - № 1. - С. 13-19.
10. Учет нелинейности рецептур смешения топлив в программном комплексе оптимального планирования и оптимизации рецептур смешения топлив / С.С. Горбунов, А.А. Алексанян, В.А. Костандян, А.Ф. Егоров // Нефтепереработка и нефтехимия. Научно-техн. достижения и передовой опыт. - 2019. - № 2. - С. 9-11.
11. Комплексная система планирования и оптимизации рецептур смешения бензинов / С.С. Горбунов, А.В. Костандян, А.А. Алексанян, А.Ф. Егоров // Математические методы в технике и технологиях: сб. тр. междунар. науч. конф. - СПб: Изд-во Политехн. ун-та, 2019. - Т. 4. - С. 91-94.
12. Hines J.W., Garvey D. Process and equipment monitoring methodologies applied to sensor calibration monitoring // Quality and Reliability Engineering. - 2007. - Vol. 23, Issue 1. - P. 123-135. - DOI: 10.1002/qre.818
13. Zhang Y., Monder D.S., Forbes J.F. Real-time optimization under parametric uncertainty: a probability constrained approach // J. of Process Control. - 2002. - Vol. 12, Issue 3. - P. 373-389. - DOI: 10.1016/S0959-1524(01)00047-6
14. Model-based real-time optimization of automotive gasoline blending operations / A. Singh, J.F. Forbes, P.J. Vermeer, S.S. Woo // J. of Process Control. - 2000. - Vol. 10, Issue 1. - P. 43-58. - DOI: 10.1016/S0959-1524(99)00037-2
15. Muller A. New method produces accurate octane blending values // Oil & Gas J. - 1992. - Vol. 90, Issue 12. - P. 80-90.
16. Zahed A.H., Mullah S.A., Bashir M.D. Predict octane number for gasoline blends // Hydrocarbon Processing. - 1993. - Vol. 72, Issue 5. - P. 85-87.
17. Бураков М.В. Нейронные сети и нейроконтроллер: учеб. пособие. - СПб.: ГУАП, 2013. - 284 с.
18. Tian G.Y., Zhao Z.X., Baines R.W. A Fieldbus-based intelligent sensor // Mechatronics. - 2000. - Vol. 10, Issue 8. - P. 835-849. - DOI: 10.1016/S0957-4158(99)00078-1
19. Kramer M.A. Nonlinear principal component analysis using autoassociative neural networks // AIChE J. - 1991. - Vol. 37, Issue 2. - P. 233-243. - DOI: 10.1002/aic.690370209
Back to issue