TY - JOUR
ID - 542597
TI - Vibration Suppression of Simply Supported Beam under a Moving Mass using On-Line Neural Network Controller
JO - Journal of Solid Mechanics
JA - JSM
LA - en
SN - 2008-3505
AU - Rezaei, S
AU - Pourseifi, M
AD - University of Applied Science and Technology, Center of Mammut, Tehran, Iran
AD - Faculty of Engineering, The University of Imam Ali, Tehran, Iran
Y1 - 2018
PY - 2018
VL - 10
IS - 2
SP - 387
EP - 399
KW - Vibration control
KW - Neural network controller
KW - Euler–Bernoulli beam theory
KW - Moving mass
DO -
N2 - In this paper, model reference neural network structure is used as a controller for vibration suppression of the Euler–Bernoulli beam under the excitation of moving mass travelling along a vibrating path. The non-dimensional equation of motion the beam acted upon by a moving mass is achieved. A Dirac-delta function is used to describe the position of the moving mass along the beam and its inertial effects. Analytical solution the equation of motion is presented for simply supported boundary condition. The hybrid controller of system includes of a controller network and an identifier network. The neural networks are multilayer feed forward and trained simultaneously. The performance and robustness of the proposed controller are evaluated for various values mass ratio of the moving mass to the beam and dimensionless velocity of a moving mass on the time history of deflection. The simulations verify effectiveness and robustness of controller.
UR - http://jsm.iau-arak.ac.ir/article_542597.html
L1 - http://jsm.iau-arak.ac.ir/article_542597_726946e939564f3d26d0e0a60b18d4e8.pdf
ER -