유클리디언 거리 기반의 단계적 소거 방법을 통한 화학센서 어레이 성능 최적화
A Step-wise Elimination Method Based on Euclidean Distance for Performance Optimization Regarding to Chemical Sensor Array
2
조회
0
다운로드
3
인용
0
참고문헌
키워드
초록
In order to prevent drink-driving by detecting concentration of alcohol from driver’s exhale breath, twenty chemical sensors fabricated. The one of purposes for sensor array which consists of those sensors is to discriminate between target gas(alcohol) and interference gases(CH3CH2OH, CO, NOx, Toluene, and Xylene). Wilks’s lambda was presented to achieve above purpose and optimal sensors were selected using the method. In this paper, step-wise sensor elimination based on Euclidean distance was investigated for selecting optimal sensors and compared with a result of Wilks’s lambda method. The selectivity and sensitivity of sensor array were used for comparing performance of sensor array as a result of two methods. The data acquired from selected sensor were analyzed by pattern analysis methods, principal component analysis and Sammon’s mapping to analyze cluster tendency in the low space (2D). The sensor array by stepwise sensor elimination method had a better sensitivity and selectivity compared to a result of Wilks’s lambda method.
Abstract
In order to prevent drink-driving by detecting concentration of alcohol from driver’s exhale breath, twenty chemical sensors fabricated. The one of purposes for sensor array which consists of those sensors is to discriminate between target gas(alcohol) and interference gases(CH3CH2OH, CO, NOx, Toluene, and Xylene). Wilks’s lambda was presented to achieve above purpose and optimal sensors were selected using the method. In this paper, step-wise sensor elimination based on Euclidean distance was investigated for selecting optimal sensors and compared with a result of Wilks’s lambda method. The selectivity and sensitivity of sensor array were used for comparing performance of sensor array as a result of two methods. The data acquired from selected sensor were analyzed by pattern analysis methods, principal component analysis and Sammon’s mapping to analyze cluster tendency in the low space (2D). The sensor array by stepwise sensor elimination method had a better sensitivity and selectivity compared to a result of Wilks’s lambda method.
참고문헌 (7)
- 김보영, "Pt, Ni, Cr이 도포된 튜브형 SnO2 나노섬유의 합성과 가스 감응특성", 센서학회지, 23(3), pp.211~215 (2014), 한국센서학회[DOI]
- 전진영, "Wilks Lamda 방법을 이용한 화학센서 어레이 최적화", 센서학회지, 23(5), pp.299~304 (2014), 한국센서학회[DOI]
- Chaudry A. N., "A method for selecting an optimum sensor array", Sensors and Actuators B : Chemical, 69(3), pp.236~242 (2000)
- 변형기, "PCA와 Sammon Mapping 분석을 통한 센서 어레이 패턴들의 실시간 가시화 방법", 센서학회지, 23(2), pp.99~104 (2014), 한국센서학회[DOI]
- Sammon Jr. J. W., "A nonlinear mapping for data structure analysis", IEEE Trans. on computers, 18(5), pp.401~409 (1969)