Title :
A full-differential analog design of an indirect inverse control law based on neural networks
Author :
Lesueur, Sébastien ; Massicotte, Daniel ; Sicard, Pierre
Author_Institution :
Dept. of Electr. & Comput. Eng., Univ. du Quebec a Trois-Rivieres, Que.
Abstract :
This paper presents a full-differential analog design of an indirect inverse control law based on dynamic back propagation neural networks developed. The on-line adaptation algorithms of the synaptic weights are modeled by means of continuous-time integration circuits. The simulation results obtained at a post-layout simulation level show a very good computing precision as well as interesting power consumption and integration area. The speed of the circuit is largely sufficient to meet real-time requirements in numerous applications of the control fields
Keywords :
VLSI; analogue integrated circuits; backpropagation; neural nets; back propagation neural networks; continuous-time integration circuits; full-differential analog design; indirect inverse control law; online adaptation algorithms; post-layout simulation; synaptic weights; Circuit noise; Circuit simulation; Computational modeling; Computer networks; Control systems; Feedforward neural networks; Neural networks; Process control; Signal processing algorithms; Very large scale integration;
Conference_Titel :
Circuits and Systems, 2006. ISCAS 2006. Proceedings. 2006 IEEE International Symposium on
Conference_Location :
Island of Kos
Print_ISBN :
0-7803-9389-9
DOI :
10.1109/ISCAS.2006.1693201
