Title :
Adaptive neural network control system inspired from the cerebro-cerebellar network for manipulation
Author :
Dufoss?©, Michel ; Kaladjian, Arthur ; Frolov, Alexander A. ; Bensma?¯l, Soraya ; Ouezdou, Fathi B.
Author_Institution :
Univ. Pierre et Marie Curie, Paris, France
Abstract :
The present neural network approach for controlling any arbitrary nonlinear plant is based on a model of the cerebro-cerebellar interaction during learning of the new visuomotor transformation. Four main prerequisites are required: the columnar organization of the cerebral cortex, the Marr-Albus-Ito theory of cerebellar learning, the equilibrium point theory of motor control and the differential neurocontroller theory. The convergence of the learning procedure is proved for the linear plant, and is confirmed for nonlinear plants by computer simulation for arm reaching movements in the horizontal plane.
Keywords :
adaptive control; learning (artificial intelligence); manipulators; neural nets; neurocontrollers; nonlinear systems; Marr-Albus-Ito theory; adaptive control system; arm reaching movements; cerebellar learning; cerebro-cerebellar network; columnar cerebral cortex organization; computer simulation; convergence; differential neurocontroller theory; equilibrium point theory; horizontal plane; learning procedure; neural network system; nonlinear plant; visuomotor transformation; Adaptive control; Adaptive systems; Brain modeling; Cerebral cortex; Control systems; Convergence; Motor drives; Neural networks; Neurocontrollers; Programmable control;
Conference_Titel :
Computational Intelligence in Robotics and Automation, 2003. Proceedings. 2003 IEEE International Symposium on
Print_ISBN :
0-7803-7866-0
DOI :
10.1109/CIRA.2003.1222319
