Toward the realization of an intelligent gas sensing system utilizing a non-linear dynamic response

Kato, Kimiko; Kato, Yutaka; Takamatsu, Kazuko; Udaka, Toshihiro; Nakahara, Takeshi; Matsuura, Yoshinobu; Yoshikawa, Kenichi

doi:10.1016/s0925-4005(00)00604-3

Cited by 34 publications

(23 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The redundancy, the multicollinearity and the assumption of linearity of the sensor responses are the main problems of the regression models, and they should be taken into account before applying regression procedures. The non-linear property of sensors, for example, is due to the saturation during the adsorption process, the inhibition or competition between gas species, and the hysteresis or ageing of the sensors (Kato et al 2000).…”

Section: Regression Analysismentioning

confidence: 99%

Sensors: From biosensors to the electronic nose

García‐González

Aparicio

2002

Grasas y Aceites

View full text Add to dashboard Cite

Section: Regression Analysismentioning

confidence: 99%

Sensors: From biosensors to the electronic nose

García‐González

Aparicio

2002

Grasas y Aceites

View full text Add to dashboard Cite

“…Kato et al [47] extracted Fourier transform features in the frequency domain from sinusoidally heated metal oxide sensors in order to discriminate ethanol, methanol, diethyl ether, acetone, ethylene, ammonia, isobutane, and benzene at a relatively low concentration level of 100 ppm. Joo et al [48] selected impedance characteristics from these same types of metal oxide Preprocessing of chemical sensor information using feature extraction techniques.…”

Section: Other Featuresmentioning

confidence: 99%

Array Optimization and Preprocessing Techniques for Chemical Sensing Microsystems

et al. 2002

View full text Add to dashboard Cite

Physical sensors, defined by their direct chemical interaction with the sensing environment, are a valuable and often essential contribution to the solution of stringent chemical sensing problems. Methods for conditioning signals from these sensors to optimize their presentation to subsequent decision-making models in the signal processing flow are presented. The assembly of sensors into an array and the preprocessing of these signals are the two primary techniques for conditioning a chemical image for concentration detection, chemical discrimination, and, in some cases, odor localization. Array optimization involves locating the point at which adding additional sensors to an array generates more noise than information and is highly dependent on the application and number of analytes to be sensed or differentiated. Signal preprocessing techniques include noise reduction, feature extraction, the reduction of array inputs, and the scaling of individual sensor signals and provide a means by which to reconstruct an array of chemical sensor signals into a subset of information that enables a decisionmaking model to do its job with greater efficiency and accuracy. In this chapter, surveys of methods are accompanied by representative examples employing a wide variety of sensors including ChemFETs, chemiresistors, acoustic wave devices, and surface plasmon resonance sensors.

show abstract

“…Mielle et al [9] demonstrated that the odors of different fruit juices afforded different time-resolved patterns when measured at a succession of different temperatures by use of a metal oxide gas sensor. Kato et al [10] also demonstrated that different gases give different dynamic sensor responses if the signal is recorded during variation of the temperature of tin oxide sensors. Besides these single-sensor approaches several publications report the use of arrays of fiber-optic sensors with immobilized dye in polymer matrices for qualitative and quantitative determination of organic vapors in mixtures [11,12,13].…”

Section: Frank Dieterlementioning

confidence: 99%

Time-resolved sensor arrays

Dieterle¹

2004

Analytical and Bioanalytical Chemistry

View full text Add to dashboard Cite

tection and determination of a variety of substances has gained increasing popularity not only in analytical chemistry but also in our daily life. Most commercially available sensor systems, for example exhaust-gas sensors for automobiles or CO-sensors, are based on single-sensor devices. Such sensors are supposed to be as selective as possible for the analyte of interest, rendering data analysis rather simple.Despite this, the problem of interfering cross-reactive analytes and the lack of highly specific sensors for many analytes have resulted in the development of so-called sensor arrays in which several different cross-reactive sensors are combined. Multivariate analysis of the signal patterns from the sensors enables simultaneous determination of several analytes. In the field of gas sensing this array approach is well known, the devices being known as "electronic noses", whereas the approach is rather new in the field of biosensors, in which the cross-reactivity of antibodies is exploited for multi-analyte determination and for detection of whole classes of substance [1, 2]. Comparing sensor approaches with classical analytical techniques, sensor arrays are equivalent to multi-wavelength spectrometers whereas single-sensor approaches can be compared with single-wavelength photometry (Table 1). An advantage of the sensor-array approach is the possibility of adapting the set-up to many different combinations of analytes by re-calibrating the device without the need to change the hardware [3]. A severe limitation of almost all the devices available for quantification of analytes in mixtures is, however, the requirement that the number of sensors exceeds the number of analytes in the samples, because usually only a single property per sensor response, for example the height, area, or slope of the sensor signal, is used for data analysis. This renders the array approach hardware-and cost-intensive for many applications.A very recent trend in sensor research is, therefore, exploitation of differences between the kinetics of interaction of the sensor with different analytes. These differences are visible as sensor responses of different shapes during exposure to different analytes. When such sensors are combined with data evaluation which recognizes and exploits the different shapes of the signals, the number of analytes quantified per sensor is limited only by the similarity of the interaction kinetics of analytes and interfering substances. This approach, which combines the sensory principle with the chromatographic principle of separating analytes in space or time, thus opens the door to multi-analyte determinations based on single-sensor setups. Consequently, this approach is comparable with simple chromatographic set-ups (Table 1). All publications using this time-resolved sensor approach can be defined by three fundamental characteristics:1. interaction principle, 2. feature extraction, and 3. data analysis.The choice of the interaction principle mainly determines the feasible sensor materials and set-ups. Se...

show abstract

Toward the realization of an intelligent gas sensing system utilizing a non-linear dynamic response

Cited by 34 publications

References 6 publications

Sensors: From biosensors to the electronic nose

Sensors: From biosensors to the electronic nose

Array Optimization and Preprocessing Techniques for Chemical Sensing Microsystems

Time-resolved sensor arrays

Contact Info

Product

Resources

About