%0 Journal Article
	%A Ahmad Forouzantabar
	%D 2010
	%J International Journal of Mechanical and Mechatronics Engineering
	%B World Academy of Science, Engineering and Technology
	%I Open Science Index 48, 2010
	%T Neural Network Control of a Biped Robot Model with Composite Adaptation Low
	%U https://publications.waset.org/pdf/7091
	%V 48
	%X this paper presents a novel neural network controller
with composite adaptation low to improve the trajectory tracking
problems of biped robots comparing with classical controller. The
biped model has 5_link and 6 degrees of freedom and actuated by
Plated Pneumatic Artificial Muscle, which have a very high power to
weight ratio and it has large stoke compared to similar actuators. The
proposed controller employ a stable neural network in to approximate
unknown nonlinear functions in the robot dynamics, thereby
overcoming some limitation of conventional controllers such as PD
or adaptive controllers and guarantee good performance. This NN
controller significantly improve the accuracy requirements by
retraining the basic PD/PID loop, but adding an inner adaptive loop
that allows the controller to learn unknown parameters such as
friction coefficient, therefore improving tracking accuracy.
Simulation results plus graphical simulation in virtual reality show
that NN controller tracking performance is considerably better than
PD controller tracking performance.
	%P 1461 - 1466