On-line Magazine 'Sensors & Transducers' (S&T e-Digest)

    (ISSN 1726- 5479)

Vol. 72, Issue 10, October 2006, pp. 731-739

Modeling of an Aged Porous Silicon Humidity Sensor Using ANN Technique

Tarikul ISLAM², Hiranmay SAHA^1*

¹IC Design and Fabrication Center, Dept. of E.T.C.E., Jadavpur University, Kolkata 700032, W.B., India

E-mail: juicc@vsnl.com, hsaha@vsnl.net, tel:913324146217, fax: 913324146217, ^*Corresponding author

²Electrical Engineering Department, Jamia Millia Islamia University, New Delhi, India, 110025,

E-mail: tariq940@rediffmail.com

Abstract: Porous silicon (PS) sensor based on capacitive technique used for measuring relative humidity has the advantages of low cost, ease of fabrication with controlled structure and CMOS compatibility. But the response of the sensor is nonlinear function of humidity and suffers from errors due to aging and stability. One adaptive linear (ADALINE) ANN model has been developed to model the behavior of the sensor with a view to estimate these errors and compensate them. The response of the sensor is represented by third order polynomial basis function whose coefficients are determined by the ANN technique. The drift in sensor output due to aging of PS layer is also modeled by adapting the weights of the polynomial function. ANN based modeling is found to be more suitable than conventional physical modeling of PS humidity sensor in changing environment and drift due to aging. It helps online estimation of nonlinearity as well as monitoring of the fault of the PS humidity sensor using the coefficients of the model.

Keywords: Porous silicon; Humidity sensing; Aging; Drift due to aging; Modeling; ANN

Click <here> or title of paper to download the full pages article (384 K)

1999 - 2006 Copyright ©, International Frequency Sensor Association (IFSA). All Rights Reserved.

Home - News - Links - Archives - Tools - Standardization - Patents - Marketplace - Projects - Wish List - Subscribe - Search - e-Shop - Membership

 Members Area -Sensors Portal -Training Courses - S&T Digest - For advertisers - Bookstore - Forums - Polls - Sensor Jobs - Site Map