Research On Optimization Of Temperature Modulation Modes Of Semiconductor Gas Sensors

The temperature modulation technology has the characteristics of improving the selectivity and stability of the semiconductor gas sensor.However,the current temperature modulation modes are numerous,and the technique faces the problem of optimal selection of the modulation modes.In this paper,temperature modulation technology is used in gas detection for optimization research.A temperature modulation experimental test system is built,which includes a gas distribution module,a temperature modulation module,and an acquisition module.The gas distribution module sets the concentration of the gas to be measured and the concentration of the mixed gas.The temperature modulation module can realize 0~5V heating voltage adjustable,and can output four heating waveforms of sine wave,square wave,triangle wave and sawtooth wave,and the heating frequency is adjustable from 0 to450 kHz.The acquisition module collects the voltage response value of the sensor through the data acquisition card and displays it in real time through the display screen.A wide-spectrum commercial semiconductor gas sensor(TGS2611)was selected for the test and optimization of temperature modulation mode.The four gases of methane,carbon monoxide,carbon dioxide and ethanol were used as test objects.Four kinds of modulation waveforms,i.e.sine wave,square wave,triangular wave and sawtooth wave,were tested for methane,carbon monoxide,carbon dioxide and ethanol.Eight cycles(T=4s,10 s,20s,30 s,40s,50 s,60s,80s)were tested.The dynamic response characteristics of four different heating voltages(0 ~ 5V,1 ~ 5V,2 ~ 5V,3 ~ 5V)are also discussed.The discrete wavelet transform is used to extract the wavelet features of the dynamic response signals of the above gases,and the principal component analysis is used to reduce the dimension of the wavelet features.The quantitative analysis of the measured gases is carried out by combining the Support Vector Machine and Probabilistic Neural Network technology.By comparing and analyzing the recognition results,the temperature modulation optimization algorithm is established.Through the analysis and processing of the experimental data,the results show that the TGS2611 sensor has the highest recognition rate of the above gas under the heating voltage of 2~5V,square wave modulation and 20 s period.The data analysis and experimental results show that the proposed temperature modulation mode optimization method based on wavelet feature combined with probabilistic neural network and support vector machine is effective,which can provide new detection ideas for low-power gas detection in complex environments.