Neural and Neurofuzzy FELA Adaptive Robot Control Using Feedforward and Counterpropagation Networks

  • Authors:

  • S. G. Tzafestas;G. G. Rigatos

  • Affiliations:

  • Intelligent Robotics and Automation Laboratory, Department of Electrical and Computer Engineering, National Technical University of Athens, Zografou 15773, Athens, Greece;Intelligent Robotics and Automation Laboratory, Department of Electrical and Computer Engineering, National Technical University of Athens, Zografou 15773, Athens, Greece

  • Venue:
  • Journal of Intelligent and Robotic Systems
  • Year:
  • 1998

Quantified Score

Hi-index 0.04

Visualization

Abstract

In this paper, the application of neural networks and neurofuzzy systemsto the control of robotic manipulators is examined. Two main controlstructures are presented in a comparative manner. The first is a CounterPropagation Network-based Fuzzy Controller (CPN-FC) which is able toself-organize and correct on-line its rule base. The self-tuning capabilityof the fuzzy logic controller is attained by taking advantage of thestructural equivalence between the fuzzy logic controller and acounterpropagation network. The second control structure is a more familiarneural adaptive controller based on a feedforward (MLP) network. The neuralcontroller learns the inverse dynamics of the robot joints, and graduallyeliminates the model uncertainties and disturbances. Both schemes cooperatewith the computed torque control algorithm, and in that way the reduction oftheir complexity is achieved. The ability of adaptive fuzzy systems tocompete with neural networks in difficult control problems is demonstrated.A sufficient set of numerical results is included.