Premium
Identification and Tracking‐Control for an Optomechatronical Image Derotator Using Neural Networks
Author(s) -
Altmann Bettina,
Rohloff Benjamin,
Pape Christian,
Reithmeier Eduard
Publication year - 2016
Publication title -
pamm
Language(s) - English
Resource type - Journals
ISSN - 1617-7061
DOI - 10.1002/pamm.201610386
Subject(s) - control theory (sociology) , artificial neural network , nonlinear system , inertia , linearization , position (finance) , vibration , tracking (education) , feedback linearization , computer science , moment of inertia , term (time) , artificial intelligence , control engineering , engineering , control (management) , physics , acoustics , classical mechanics , psychology , pedagogy , finance , quantum mechanics , economics
In this paper an optomechatronical image derotator is used for vibration measurements on rotating objects. First of all, the concept of the derotator is explained and it is shown that the phase position and the rotational velocity of the derotator and the measurement object have to be aligned. Therefore, a highly dynamic tracking‐control is needed. Considering the nonlinear friction of the synchronous motor, a model of the system which considers this non‐linearity is evolved. This is accomplished by using neural networks for the approximation of the friction term. In this case General Regression Neural Networks (GRNN) are used for the learning algorithm. Moreover, the system's parameters, eg. the friction term and the inertia, are identified based on the nonlinear model. Then a feedback control is designed by using the controllable canonical form through feedback linearization. Finally, the results of vibration measurements on a rotating blisk using the nonlinear control concept for the derotator are shown. (© 2016 Wiley‐VCH Verlag GmbH & Co. KGaA, Weinheim)