سال انتشار: ۱۳۸۶

محل انتشار: پانزدهیمن کنفرانس مهندسی برق ایران

تعداد صفحات: ۶

نویسنده(ها):

Rahideh – School of Electrical and Electronics Engineering, Shiraz University of Technology, Shiraz, I.R.IRAN, Department of Engineering, Queen Mary, University of London, London E 1 4NS, UK
Safavi – School of Engineering, Shiraz University, Shiraz, 1.R.IRAN
Shaheed – Department of Engineering, Queen Mary, University of London, London E 1 4NS, UK

چکیده:

A nonlinear dynamic modelling approach based on neural network (NN) for a real Twin Rotor hIlM0 System (TlbMS), in terms of its 2 degree of
freedom (DOF) dynamics, is presented in this paper. The TRMS is a highly nonlinear system with signiJicant cross-coupling between its horizontal and vertical axes. It is perceived as an aerodynamic test rig representing the control challenges of modern air vehicles. Accurate
dynamic modelling is a prerequisite to address such challenges satisfactorily. A feedforward neural network has been trained using Scaled Conjugate Gradient (SCG) learning algorithm. The trained NN based modelshave been tested with a set of data that are difeerent fromthose used for trainingpurpose. For more validation thepower spectral densiv (PSD) of the model is comparedwith that of the real TRhfS and also the correlationvalidations of the test results are presented in order toshow the eflectiveness ofthe proposed model. The resultsshow that the developed model can adequately representthe highly nonlinear features ofthe system