Scheduled System Maintenance:
Some services will be unavailable Sunday, March 29th through Monday, March 30th. We apologize for the inconvenience.
By Topic

Simultaneous Classification and Concentration Estimation for Electronic Nose

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

The purchase and pricing options are temporarily unavailable. Please try again later.
2 Author(s)
Dongliang Huang ; Dept. of Electr. & Comput. Eng., Calgary Univ., Alta. ; Leung, H.

By virtue of the electronic nose (E-nose), detection and estimation of gases become feasible in many fields without resorting to complicated specific instruments. Detection is generally casted as a classification problem and concentration estimation is subsequently performed using conventional statistical techniques. In this paper, we develop a polynomial-based optimization method to perform classification and estimation simultaneously to improve the intelligence of an E-nose. The proposed method employs a parametric polynomial with user-defined order to describe sensor characteristics. Classification and concentration estimation can then be formulated as a standard convex optimization problem. The convex optimization is solved either by a typical gradient descent method for an unconstrained case or a NLS trust-region method for a constrained case. The main advantages of the proposed method are the flexibility and significant reduced computation cost as well as simple implementation. Moreover, the global minimum of the optimization is readily achieved. Experimental data analysis demonstrates the efficiency of the proposed method

Published in:

Sensors Journal, IEEE  (Volume:7 ,  Issue: 5 )