In recent years there has been increasing interest in universal model-free
controllers. These controllers using neural networks learn about the system
s they are controlling online, and thus automatically improve their perform
ance. There has been a good deal of research on the use of neural networks
for control, although most of the articles have been ad hoc discussions lac
king theoretical proofs and repeatable design algorithms. In this paper exp
erimental results on the control of robotic manipulator using neural networ
ks have been provided and it has been demonstrated that neural networks do
indeed fulfill the promise of providing model-free learning controllers for
robotic systems and provide an excellent alternative for the control of ro
botic manipulators. (C) 1999 Elsevier Science Ltd. All rights reserved.