ISSN 2071-8594

Russian academy of sciences

Editor-in-Chief

Gennady Osipov

M.V. Bobyr The method of non-linear learning the neuro-fuzzy inference system

Abstract.

The new method of learning the neuro-fuzzy inference system is considered in the article. A feature of this method is the use of a non-linear function in the model of the differential areas. The generalized model of fuzzy inference whit using the linear and nonlinear functions is structured. The feature of the generalized model is use of various t-norms (Mamdani implication, the implication of the algebraic product, Lukasevic's implication, the implication of the bounded difference, and soft implication). The results of modeling the learning process of the neuro-fuzzy inference system using the linear and non-linear functions in the method of the differential areas are presented. Evaluation of the working of the neurofuzzy inference system was carried out on the basis of calculating the RMSE and the time required for its learning. The proposed nonlinear model of the differential areas increases the accuracy during learning fuzzy systems. This conclusion is confirmed by the results of simulation modeling presented in the article.

Keywords:

method of difference areas, soft computing, defuzzification, learning, adaptive neuro-fuzzy inference system, RMSE.

PP. 67-75.

References

1. Bashlykov A.A. Obraznoe predstavlenie sostojanija slozhnyh tehnologicheskih obektov upravlenija // Iskusstvennyj intellekt i prinjatie reshenij. 2012. № 3. S. 9-18.
2. Karpov V.Je., Val'cev V.B. Dinamicheskoe planirovanie povedenija robota na osnove seti «intellektual'nyh» nejronov // Iskusstvennyj intellekt i prinjatie reshenij. 2009. № 2. S. 58-69.
3. Emaletdinova L.Ju., Strunkin D.Ju. Modelirovanie diagnosticheskoj dejatel'nosti vracha na osnove nechetkoj nejronnoj seti // Iskusstvennyj intellekt i prinjatie reshenij. 2010. № 3. S. 73-78.
4. Abdulshahed A.M., Longstaff A.P., Fletcher S. The application of ANFIS prediction models for thermal error compensation on CNC machine tools // Applied Soft Computing. 2015. Vol. 27. p. 158-168.
5. Aliev M.I., Isaeva Je.A., Aliev I.M. Teorija verojatnostej i teorija nechetkih mnozhestv L. Zade: razlichija i shodstvo // Iskusstvennyj intellekt i prinjatie reshenij. 2012. № 3. S. 19-25.
6. Tenetko O.Ju., Peskova M.I. Provedenie jeksperimenta po obosnovaniju vybora nechjotkoj implikacii, prigodnoj dlja reshenija zadach klassifikacii riskov i vyrabotki nailuchshih rekomendacij po riskam // Izvestija JuFU. Tehnicheskie nauki. 2012. № 12 (137). S. 117-121.
7. Mandal S., Jayaram B. SISO fuzzy relational inference systems based on fuzzy implications are universal approximators // Fuzzy Sets and Systems. 2015. Vol. 277, Issue 15. p. 1-21.
8. Vemuri N.R. Mutually exchangeable fuzzy implications // Information Sciences. 2015. Vol. 317, Issue 1. p. 1-24.
9. Pegat A. Nechetkoe modelirovanie i upravlenie / A. Pegat, per. s angl. – M.: BINOM. Laboratorija znanij. 2012. 798 s.
10. Bobyr' M.V., Kulabuhov S.A. Defazzifikacija vyvoda iz bazy nechetkih pravil na osnove metoda raznosti ploshhadej // Vestnik komp'juternyh i informacionnyh tehnologij. 2015. № 9 (135). S. 32-41.
11. Bobyr' M.V., Kulabuhov S.A., Milostnaja N.A. Obuchenie nejro-nechetkoj sistemy na osnove metoda raznosti ploshhadej // Iskusstvennyj intellekt i prinjatie reshenij. 2016. № 4. S. 15-26.
12. Bobyr' M.V., Kulabuhov S.A., Titov D.V. Ocenka vlijanija chisla obuchaemyh tochek na additivnost' nechetkih sistem // Promyshlennye ASU i kontrollery. 2014. № 10. S. 30-35.
13. Bobyr' M.V., Milostnaja N.A. Nechetkaja model' intellektual'noj sistemy upravlenija mobil'nym robotom // Problemy mashinostroenija i avtomatizacii. 2015. № 3. S. 57-67.
14. Bobyr' M.V. Adaptacija sistemy upravlenija mobil'nym robotom na osnove nechetkoj logiki // Mehatronika, avtomatizacija, upravlenie. 2015. T. 16. № 7. S. 449-455.
15. Van Broekhoven E., De Baets B. Only Smooth Rule Bases Can Generate Monotone Mamdani–Assilian Models Under Center-of-Gravity Defuzzification // IEEE Transactions on fuzzy systems, 2009. Vol. 17. №. 5, p. 1157-1174.