سال انتشار: ۱۳۸۵

محل انتشار: چهاردهمین کنفرانس سالانه مهندسی مکانیک

تعداد صفحات: ۹

نویسنده(ها):

Aghil Yousefi-Koma – Assistant Professor, University of Tehran
Kavous Jorabchi – Undergraduate Student, University of Tehran
Zimcik – Group Leader, National Research Council Canada (NRC)

چکیده:

A neural network based control system was designed and implemented for vibration suppression of a flexible fin with piezoelectric actuators. A new approach in choosing the number of tapped delays of neural networks is introduced. Smart fin vibration was monitored through tip acceleration, and piezoactuators were employed to suppress the vibration. The controller was examined against four cases of disturbances: dwell at the first and second modes, sine sweep, and random excitation. Results indicated that the flexural vibrations were suppressed for both dynamic modes.