The design of gas sensor and the approach of artificial neural networks based mathematical model

In this thesis, graphene oxide (GO) thin films were prepared by Langmuir-Blodgett (LB) thin film technique to use as chemical sensor element. Quartz crystal microbalance (QCM) was employed to characterize the properties of GO nano LB thin film layers. Also, selectivity of the nano thin film sensor to various harmful organic vapors was investigated by QCM technique. Fick's Equations were used for analyzing the swelling process of nano thin film sensor. Diffusion coefficients (D) values of organic vapors for swelling were also obtained. The initial experiments have revealed that GO materials exhibited an effective chemical sensor characteristic with a large response to dichloromethane vapor. Then, by using experimental data that obtained during this process, NARX-ANN based mathematical model of frequency shift of the quartz resonator was designed. The results of NARX-ANN verify the efficiency of the designed model.