Xiongwei Peng

Concentration estimation of formaldehyde using metal oxide semiconductor gas sensor array-based e-noses

Purpose – The purpose of this paper is to present a novel concentration estimation model for improving the accuracy and robustness of low-cost electronic noses (e-noses) with metal oxide semiconductor sensors in indoor air contaminant monitoring and overcome the potential sensor drift. Design/methodology/approach – In the quantification model, a piecewise linearly weighted artificial neural network ensemble model (PLWE-ANN) with an embedded self-calibration module based on a threshold network is studied. Findings – The nonlinear estimation problem of sensor array-based e-noses can be effectively transformed into a piecewise linear estimation through linear weighted neural networks ensemble activated by a threshold network. Originality/value – In this paper, a number of experimental results have been presented, and it also demonstrates that the proposed model has very good accuracy and robustness in real-time indoor monitoring of formaldehyde.

Time series estimation of gas sensor baseline drift using ARMA and Kalman based models

Purpose – The purpose of this paper is to present a novel and simple prediction model of long-term metal oxide semiconductor (MOS) gas sensor baseline, and it brings some new perspectives for sensor drift. MOS gas sensors, which play a very important role in electronic nose (e-nose), constantly change with the fluctuation of environmental temperature and humidity (i.e. drift). Therefore, it is very meaningful to realize the long-term time series estimation of sensor signal for drift compensation. Design/methodology/approach – In the proposed sensor baseline drift prediction model, auto-regressive moving average (ARMA) and Kalman filter models are used. The basic idea is to build the ARMA and Kalman models on the short-term sensor signal collected in a short period (one month) by an e-nose and aim at realizing the long-term time series prediction in a year using the obtained model. Findings – Experimental results demonstrate that the proposed approach based on ARMA and Kalman filter is very effective in ti...

On-line Calibration of Semiconductor Gas Sensors Based on Prediction Model

In this study, an Electronic nose (Enose) instrument used indoor for monitoring formaldehyde is designed. In mass production of this instrument, because of the inherent variability in the sensor manufacturing process, the Enose instruments give different outputs. It is impossible to train an individual prediction model on each instrument to have uniform output. A new on-line calibration method based on prediction model without real master instrument is proposed. This method avoids the problem that if the real master instrument behaves drift, the calibration of the other batch of instruments would lose its effect. In this paper, the prediction model is radial basis function (RBF) neural network and particle swarm optimization (PSO) is used to determine the parameters in RBF. The results show that the responses of the same type sensors are uniform after calibration, and this new method is easy and robust.