Title :
A robust hybrid neural architecture for an industrial sensor application
Author :
Djahanshahi, H. ; Ahmadi, M. ; Jullien, G.A. ; Miller, W.C.
Author_Institution :
Dept. of Electr. Eng., Windsor Univ., Ont., Canada
Abstract :
A programmable hybrid neural network architecture has been used to implement a smart optical sensor with focal-plane pattern classification for a flexible manufacturing cell environment. The network contains an integrated photosensitive array based on modified photo BJTs as input to a fully-connected multilayer feedforward (MLFF) neural classifier. The architecture features a distributed neuron realization that employs a number of active nonlinear resistor circuits operating in parallel. It minimizes the effect of parameter variations due to non-uniform device fabrication over the die surface. Moreover, due to the modularity of the architecture and locality of interconnections, synaptic density has been doubled in comparison with a conventional realization. A photosensor-classifier chip consisting of a 2-D array of 64 neural-based smart pixels and additional neural network circuits has been fabricated. The proposed architecture has been implemented in both CMOS and BiCMOS process technologies as part of a sensor optimization study
Keywords :
BiCMOS integrated circuits; CMOS integrated circuits; VLSI; feedforward neural nets; image processing equipment; image sensors; industrial control; intelligent sensors; neural chips; neural net architecture; pattern classification; phototransistors; 2D neural-based smart pixel array; BiCMOS process technolog; CMOS process technolog; active nonlinear resistor circuits; distributed neuron realization; flexible manufacturing cell environment; focal-plane pattern classification; industrial sensor application; integrated photosensitive array; modified photo BJTs; modular architecture; multilayer feedforward neural classifier; photosensor-classifier chip; programmable hybrid neural network architecture; robust hybrid neural architecture; smart optical sensor; synaptic density; Circuits; Flexible manufacturing systems; Intelligent sensors; Manufacturing industries; Multi-layer neural network; Neural networks; Neurons; Optical sensors; Pattern classification; Robustness;
Conference_Titel :
Circuits and Systems, 1998. ISCAS '98. Proceedings of the 1998 IEEE International Symposium on
Conference_Location :
Monterey, CA
Print_ISBN :
0-7803-4455-3
DOI :
10.1109/ISCAS.1998.703891
