Mechanical Engineering
Samotylova S., Ryabaya O., Torgashov A.
SVETLANA A. SAMOTYLOVA, Postgraduate Student, e-mail: samotylova.sa@gmail.com
OLGA O. RYABAYA, Student of Master’s Degree Courses, e-mail: olgaryabaya@mail.ru
ANDREI YU. TORGASHOV, Professor, е-mail: andrei.torgashov@mail.ru
Department of Industrial Technology, School of Engineering, Far Eastern Federal University, Vladivostok.
8 Sukhanova St., Vladivostok, Russia, 690950
The paper deals with the issue of the identification of prognostic models with the use of least squares, robust regression, and ridge regression. The models may be used in technological processes to stabilise catalytic cracking benzene and produce methyl tert-butyl ether. It has been demonstrated that the method of robust regression with a weight function ω6 and that of least squares make it possible to obtain prognostic models which describing best the corresponding technological processes under research.
Key words: prognostic models, technological unit, identification, gas separation, stabilisation, benzene, methyl tert-butyl ether, regression method.
REFERENCES
1. Nurmuhametova E.R., Akhmetov A.F., Rahmatullin A.R. Study of catalytic cracking benzene. Oil and gas business. 2014;2:181-193. (in Russ.). [Nurmuhametova Je.R., Ahmetov A.F., Rahmatullin A.R. Issledovanie benzina kataliticheskogo krekinga // Neftegazovoe delo. 2014. № 2. S. 181–193].
2. Tarasenko R.Yu., Burlyaeva E.V. Functional simulation production of methyl tert-butyl ether. Mathematical methods and information technologies in chemistry and chemical technology. MITHT Bulletin. 2013(13);5:136-140. (in Russ.). [Tarasenko R.Ju., Burljaeva E.V. Funkcional'noe modelirovanie proizvodstva metil-tret-butilovogo jefira // Matematicheskie metody i informacionnye tehnologii v himii i himicheskoj tehnologii. Vestnik MITHT. 2013. T. 13, № 5. S. 136–140.]
3. Holland P.W., Welsch R.E. Robust regression using iteratively reweighted least-squares. Communications in Statistics – Theory and Methods. 1977(6);9:813-827.
4. Khalaf G., Shukur G. Choosing Ridge Parameter for Regression Problems. Communications in statistics. Theory and methods. 2005(34):1177-1182.