N.V. Andrievskaya - Ph.D. (Eng.), Associate Professor, Department of Automation icroprocessors, Perm National Research Polytechnic University. E-mail: nataly-anv@mail.ru

Identification of nonlinear systems is one of the most difficult tasks. The article reveals the shortcomings of the classical approach to system identification. Given the non-linear model of the hydromechanical system is described. The problems of estimating the model parameters are revealed. The neural network approach for solving the problem of estimating the parameters of nonlinear systems is proposed. The schemes of an artificial neural network implementation are considered. Results of artificial neural network simulation for different operating modes are described. The analysis of the simulation results is described.

 

1. Mikhajjlov A.S., Staroverov B.A. Problemy ispolzovanija iskusstvennykh nejjronnykh setejj dlja identifikacii i diagnostiki tekhnicheskikh obektov // Vestnik IGEHU. 2013. № 3. S. 64-68.
2. Andrievskaja N.V., Reznikov A.S., CHeranev A.A. Osobennosti primenenija nejjro-nechetkikh modelejj  dlja zadach sinteza sistem avtomaticheskogo upravlenija // Fundamentalnye issledovanija. 2014. № 11. S. 1445-1449.
3. Andrievskaja N.V. Ispolzovanie nejjrosetevogo podkhoda dlja reshenija zadachi sovmestnogo ocenivanija parametrov i sostojanija // Nejjrokompjutery. 2014. № 12. S. 41-44. 
4. Djageterev A.V., Zaporozhec O.V., Ovcharova T.A. Identifikacija nelinejjnojj funkcii preobrazovanija s pomoshhju iskusstvennojj nejjronnojj seti // Ukraїnskijj metrologіchnijj zhurnal. 2013. № 2. S. 4-8.
5. Fajjzrakhmanov R.A., Lipatov I.N., Dolgova E.V. Reshenie zadachi identifikacii nelinejjnogo diskretnogo dinamicheskogo obekta s ispolzovaniem mnogoslojjnojj nejjronnojj seti // Nejjrokompjutery. 2013. № 11. S. 89-95.
6. Andrievskaja N.V., Matushkin N.N., JUzhakov A.A. Identifikacija sistem upravlenija: ucheb. posobie dlja vuzov. Perm: Izd-vo Perm. nac. issled. politekhn. un-ta. 2012. 170 s.