Title :
Neural network technique for parametric testing of mixed-signal circuits
Author_Institution :
Inst. of Electron., Tech. Univ. Lodz, Poland
fDate :
2/2/1995 12:00:00 AM
Abstract :
Feedforward neural networks have been used to identify unknown parameters of an active filter excited by a voltage step. It has been demonstrated that the proposed technique is much faster and more robust in the presence of noise when compared to least square-error model fitting
Keywords :
fault diagnosis; feedforward neural nets; integrated circuit testing; mixed analogue-digital integrated circuits; network parameters; parameter estimation; IC testing; VLSI circuits; circuit parameter identification; feedforward neural networks; mixed-signal circuits; parametric testing; voltage step excitation;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19950148
