Research On Dynamic Detection Method Of Semiconductor Gas Sensor Based On Temperature Modulation

At the same time as the modern industry continues to develop,the emissions of waste gas from all walks of life have also increased dramatically.These exhaust gases have caused irreparable damage to the healthy survival of humans and other animals and plants.Some of these exhaust gases are explosive and flammable,such as natural gas,liquefied petroleum gas,etc.some may cause damage to human health.Studies have shown that long-term exposure to high VOCs can cause various diseases,such as respiratory diseases,neurological diseases,cancer,and even genetic damage and death.It is important to study stable,high performance gas sensors.Based on a large number of researches on the research progress of metal oxide semiconductor gas sensors at home and abroad,it is found that there are always problems of selectivity and stability.This paper avoids the traditional methods of studying sensitive materials,focusing on testing methods to improve the selectivity and stability of the sensor.The information that can be obtained by the traditional static test method is only the resistance change and response time of the sensor’s sensitive film and the final state,but the chemical information in the sensitive process cannot be obtained,and the sensitive mechanism is unfavorable.Therefore,this thesis proposes a dynamic test method,which systematically studies the principle,rationale and various influencing factors of dynamic testing,and uses six VOCs characteristic signals to identify and classify data sets.First,the development history of gas sensors,the classification of gas sensors,the research status of dynamic testing,and the research progress of universities with related subjects are reviewed.The selectivity problem of current gas sensors is proposed.Based on this,the research direction and content of the thesis are determined,and the method of changing the working temperature of the sensor,that is,dynamic testing,is adopted to improve its selectivity.VOCs are determined as the research object.Then the gas sensing mechanism of the sensor is analyzed from the grain boundary barrier model,and the effect of temperature on the semiconductor conductance itself is analyzed from the direction of semiconductor physics.The nano-zinc oxide gas sensor was fabricated by the two-solvent method,and its characteristics were analyzed and tested.The test system used and the software system written in Labview are introduced.Then the concept and principle of dynamic testing are discussed.Based on dynamic tests,a series of VOCs were tested,and the data selection method in this paper was introduced.The response signals of triangle waves with different periods,the same duty cycle and amplitude square waves and trapezoidal waves were compared.The effects of each interval,humidity,temperature rise and fall rate,and interfering gas on the gas response signal were analyzed.Then the data pre-processing and the basic concepts and construction methods of generalized regression neural network(GRNN)are discussed.The smoothing factor was selected with the minimum values of MAE and MSE as the optimal conditions,and a GRNN neural network model was constructed.The results of gas identification using the gas-sensitive signals obtained from the dynamic test model used in this paper as the data set were analyzed.Finally,the full text is summarized and prospected.The characteristics and advantages of dynamic testing in the sensor testing of VOCs gas are summarized.The improvement direction of this research is prospected,including the improvement of sensor performance,modulation method and data feature extraction algorithm.