|
Abstract:
PERSONNEL SAFETY REQUIRES DETECTION OF HAZARDOUS VAPORS AT SUB-PART-PER- MILLION CONCENTRATIONS AND FAN DETECTORS HAVE THE NECESSARY SENSITIVITY AND SELECTIVITY WITHOUT SOPHISTICATED DATA ANALYSIS SUCH AS PATTERN RECOGNITION. SURFACE ACOUSTIC WAVE DEVICES WITH DIFFERENT VAPOR SENSITIVE COATINGS ARE BEING USED TO DETECT LOW CONCENTRATIONS OF VAPORS IN AIR. INDIVIDUALLY, THE SENSORS LACK SELECTIVITY; HOWEVER, AN ARRAY OF THE SENSORS CAN PRODUCE A UNIQUE FINGERPRINT FOR EACH VAPOR OF INTEREST. A DATA MATRIX, FORMED BY EXPOSING 12 COATINGS TO SEVERAL VAPORS REPRESENTING DIFFERENT CHEMICAL CLASSES AND CONCENTRATIONS, HAS BEEN STUDIED USING PATTERN RECOGNITION METHODS. SUPERVISED LEARNING TECHNIQUES WERE USED TO REDUCE TO THREE THE NUMBER OF SENSORS NECESSARY TO SEPARATE HAZARDOUS VAPORS FROM THE OTHERS TESTED. A SECOND DATA SET OF 4 COATINGS EXPOSED TO 10 DIFFERENT VAPORS AND 2-COMPONENT MIXTURES WAS USED TO TEST THE CLASSIFICATION CAPACITY OF THE THREE SENSORS FOUND PREVIOUSLY. EIGHTY-FIVE PERCENT OF THE VAPORS WERE CORRECTLY IDENTIFIED. THE DATA SET WAS ALSO USED TO INVESTIGATE MIXTURE CLUSTERING WITH OTHER PATTERN RECOGNITION TECHNIQUES.
| Description: |
CONFERENCE PAPER |
| Pages: |
8 |
| Report Date: |
JUN 1987 |
| Report Number: |
D063057 |
Report Unavailable |
| This title is unavailable from Storming Media. We do not know when it might be available, if at all. We list the report on our site for bibliographic completeness, to help our users know what other work has been performed in this field. Please note that as with all titles on this site, we do not have contact information for any of the authors. Nor can we give any suggestions on how one might obtain this report. |
|
|
|
Email This Abstract
