Inverse prediction of gas concentration based on nonparametric model

doi:10.7523/j.issn.2095-6134.2020.06.002

Abstract

Abstract: Gas sensor array is an important and powerful technique for detecting gas and measuring gas concentrations. The conventional strategy to describe the relationship between the response of the sensor and the actual gas concentration is to use some specific nonlinear parametric models. In this work, we use the nonparametric model to depict the change in the gas sensor response with the gas concentrations, which effectively avoids model misspecification. Furthermore, we propose an inverse prediction method based on the nonparametric model to predict gas concentrations. Data-driven selection of tuning parameters is also developed. The simulation results reveal that, when the real model of the sensor array is unknown or misspecified, the nonlinear parametric model is inferior to the nonparametric model in performance. Meanwhile, we verify this with the real data analysis.

Key words: Gauss Newton method, gas concentration, inverse prediction, nonlinear parametric model, nonparametric model

CLC Number:

O212.7

WU Dong, GUO Xiao. Inverse prediction of gas concentration based on nonparametric model[J]. , 2020, 37(6): 728-735.

References

[1] Gujral P, Amrhein M, Bonvin D. Drift correction in multivariate calibration models using on-line reference measurements[J]. Analytica Chimica Acta, 2009, 642:27-36.
[2] Gujral P, Amrhein M, Wise B M, et al. Framework for explicit drift correction in multivariate calibration models[J]. Journal of Chemometrics, 2010, 24:534-543.
[3] Trincavelli M, Coradeschi S, Loutfi A. Odour classification system for continuous monitoring applications[J]. Sensors and Actuators B:Chemical, 2009, 139(2):265-273.
[4] Marcelloni F. Recognition of olfactory signals based on supervised fuzzy C-means and k-NN algorithms[J]. Pattern Recognition Letters, 2001, 22(9):1007-1019.
[5] Marco S, Gutiérrez-Ga'lvez A. Signal and data processing for machine olfaction and chemical sensing:a review[J]. IEEE Sensors Journal, 2012, 12(11):3189-3214.
[6] Llobet E, Brezmes J, Vilanova X, et al. Qualitative and quantitative analysis of volatile organic compounds using transient and steady-state responses of a thick-film tin oxide gas sensor array[J]. Sensors and Actuators B:Chemical, 1997, 41:13-21.
[7] Marco S, Pardo A, Davide F A M, et al. Different strategies for the identification of gas sensing systems[J]. Sensors and Actuators B:Chemical, 1996, 34:213-223.
[8] Clifford P K, Tuma D T. Characteristics of semiconductor gas sensors. I. Steady state gas response[J]. Sensors and Actuators, 1982, 3:233-254.
[9] Clifford P K, Tuma D T. Characteristics of semiconductor gas sensors. II. Transient response to temperature change[J]. Sensors and Actuators, 1983, 3:255-281.
[10] Madou M J, Morrison S R. Chemical sensing with solid state devices[M]. Boston:Academic Press, 1989:547-556.
[11] Chaiyboun A, Traute R, Kiesewetter O, et al. Modular analytical multicomponent analysis in gas sensor aarrays[J]. Sensors, 2006, 6(4):270-283.
[12] Faraway J J. Extending the linear model with R:generalized linear, mixed effects and nonparametric regression models[M]. Boca Raton, Florida:Chapman and Hall, 2006.
[13] Pardo A, Marco S, Samitier J. Nonlinear inverse dynamic models of gas sensing systems based on chemical sensor arrays for quantitative measurements[J]. IEEE Transactions on Instrumentation and Measurement, 1998, 47(3):644-651.
[14] Fan J, Gijbels I. Local polynomial modeling and its applications[M]. London:Chapman and Hall, 1996.
[15] Ruppert D, Wand M P. Multivariate locally weighted least squares regression[J]. The Annals of Statistics, 1994, 22:1346-1370.
[16] Björck A. Numerical methods for least squares problems[M]. Philadelphia:Society for Industrial and Applied Mathematics, 1996.
[17] Xia Y C, Li W K. Asymptotic behavior of bandwidth selected by the cross-validation method for local polynomial fitting[J]. Journal of Multivariate Analysis, 2002, 83(2):265-287.
[18] Li Q, Racine J. Cross-validated local linear nonparametric regression[J]. Statistica Sinica, 2004, 14:485-512.
[19] Racine J, Li Q. Nonparametric estimation of regression functions with both categorical and continuous data[J]. Journal of Econometrics, 2004, 119(1):99-130.
[20] Zhang C M. Calibrating the degrees of freedom for automatic data smoothing and effective curve checking[J]. Journal of the American Statistical Association, 2003, 98(463):609-628.
[21] Fonollosa J, Sheik S, Huerta R, et al. Reservoir computing compensates slow response of chemosensor arrays exposed to fast varying gas concentrations in continuous monitoring[J]. Sensors and Actuators B:Chemical, 2015, 215:618-629.