Title :
Simultaneous measurement of temperature and strain: an artificial neural network approach
Author :
Chan, C.C. ; Jin, W. ; Rad, A.B. ; Demokan, M.S.
Author_Institution :
Dept. of Electr. Eng., Hong Kong Polytech., Hung Hom, Hong Kong
fDate :
6/1/1998 12:00:00 AM
Abstract :
In this letter, we report the use of an artificial neural network approach for simultaneous recovery of information about strain and temperature from fiber optic sensors. Simulation results show that, for a particular sensor with large cross-sensitivity, temperature and strain measurement accuracy can be increased by 12 and 3 times, respectively, when compared with the matrix inversion method.
Keywords :
fibre optic sensors; measurement errors; optical neural nets; sensitivity; strain measurement; temperature measurement; artificial neural network approach; fiber optic sensors; large cross-sensitivity; matrix inversion method; simultaneous measurement; strain measurement; strain measurement accuracy; temperature measurement; Artificial neural networks; Backpropagation; Capacitive sensors; Error correction; Intelligent sensors; Random number generation; Sensor phenomena and characterization; Strain measurement; Temperature measurement; Temperature sensors;
Journal_Title :
Photonics Technology Letters, IEEE
