Using Backpropagation Algorithm for Neural Adaptive Control: Experimental Validation on an Industrial Mobile Robot
This paper presents an original method in the use of neural networks and backpropagation algorithm to learn control of robotics systems. The originality consists to express the control objective as a criterion of which the gradient is backpropagating through the network instead of the classical quadratic error used in standard backpropagation. This technic allows on-line learning that is impossible to do with standard backpropagation. Experimental validation is realised by the position and the orientation control of a faster industrial mobile robot. Results show the feasability of the method, and particularly establish that on-line learning scheme permit to refine the weights of the network in front of the kinematics constraints of the robot.
KeywordsControl Objective Real Robot Neural Controller Wheel Velocity Human Teacher
Unable to display preview. Download preview PDF.
- D.E Rummelhart, J.L. McClelland, “Parallel Distributed Processing”, MIT Press, pp. 318–362, 1986Google Scholar
- A. Guez, J. Selinsky, A neuromorphic controller with a human teacher, Proc. of IEEE International Conference on Neural Networks, pp 595–602, 1988Google Scholar
- D. Nguyen, B. Widrow “Neural Networks for Self-Learning Control Systems”,IEEE Work. on Industrial Applications of Neural Networks, pp 18–23, 1991Google Scholar
- S. Delaplace, “Navigation d Coût Minimal d’un Robot Mobile dans un Environnement Partiellement Connu”’ thèse de l’Université Paris 6 ( Pierre et Marie Curie ), France, November 1991Google Scholar
- D.M.A. Lee, W.H. ElMaraghy, “A Neural Network Solution for Biped Gait Synthesis”, Int. Joint Conf. on Neural Networks, Vol II, pp. 763–767, 1992Google Scholar
- M.I. Jordan and D.E Rummelhart, “Forwars Models: Supervised Learning with a Distal Teacher”, Cognitive Science, Vol 16, pp. 307–354, 1992Google Scholar
- C. K. Ait-Abderrahim, “Commande de Robots Mobiles”, Thèse de l’Ecole des Mines de Paris, 1993.Google Scholar
- P. Hénaff, H. Schwenk,M. Milgram “ A Neural Network Approach of the Control of Dynamic Biped Equilibrium”,International Symposium on Measurement and Control in Robotics,1993Google Scholar
- P. Hénaff, “Mises en Œuvre de Commandes Neuronales par Rétropropagation Indirecte: Application d la Robotique Mobile”, thèse de l’Université Paris 6 ( Pierre et Marie Curie ), France, june 1994Google Scholar
- P. Hénaff, “Adaptive Neural Control In mobile Robotics: Experimentation for a Wheeled Cart”, IEEE Int. Conf. on Systems, Man and Cybernetics, pp 1139–1144, Oct. 1994Google Scholar