Force feedback control of an assembly robot by neural networks
This paper presents the control of a non linear dynamic system by a neural network controller. This approach based on a feedforward neural network doesn't need any mathematical model of the system. The architecture and the stability of this controller are first analyzed, then an implementation on a flexible assembly cell including a parallel robot is presented. Experimental results of a peg in a hole insertion task show that the proposed hybrid neural controller exhibits better perfomances than the classical external hybrid force position controller.
Unable to display preview. Download preview PDF.
- 1.M. Y. Amirat. Contribution à la commande de haut niveau de processus robotisés et à l'utilisation des concepts de VIA dans l'interaction robot-environnement. PhD thesis, University Paris XII, 22 January 1996.Google Scholar
- 2.P. Antsakis. Neural networks in control systems. IEEE Control Syst. Mag., 10:3–5, 1990.Google Scholar
- 3.A. G. Barto and P. Anandan. Pattern recognizing stochastic learning automata. IEEE Transaction on Systems, Man, and Cybernetics, 15:360–375, 1985.Google Scholar
- 4.O. Khatib. A unified approach to motion and force control of robot manipulators. IEEE J. Robot Automation, 1(3):43–53, 1987.Google Scholar
- 5.V. Perdereau. Contribution à la commande hybride force-position. Master's thesis, University Paris VI, 18 Feb 1991.Google Scholar
- 6.M. H. Raibert and J. J. Craig. Hybrid position/force control of manipulators. J. Dyn. Syst. Meas. Control, 102:126–133, 1982.Google Scholar
- 7.N. Saadia, Y. Amirat, J. Pontnau, and N. K. M'Sirdi. Experimental comparison between external force position controller and new hybrid neural networks controller. ECPD'97., page In press, September 15–17 1997.Google Scholar
- 8.N. Saadia, Y. Amirat, J. Pontnau, and A. Ramdane-Cherif. Hybrid force position control of nonlinear systems using neural networks. The sec. ECPD int. Conf. on Adv. rob. Int. Aut. and Act. Sys., pages 68–73, September 26–28 1996.Google Scholar
- 9.K. Watanabe, T. Fukuda, and S. J. Tzafestas. An adaptative control for carma systems using linear neural networks. Int. J. Control, 56:483–497, 1992.Google Scholar