Modeling and control of a robot manipulator using neural networks
Heng Soon Neo · DR-NTU (Nanyang Technological University) · 2000
This thesis presents a three-layer neural networks controller suitable for controlling a SCARA robot. A modified feedback error propagation learning algorithm was used to train the multi-layered neural network. A powerful graphical software tool, known as Simulink, was used to build graphical block diagrams of the entire control system, simulate dynamic response of the system, evaluate system performance and refine the design.