E-nose And E-tongue Based On QCM Sensor Array:Research And Application In Detection Of Chinese Liquors

Electronic nose(E-nose)and Electronic tongue(E-tongue)are machine perception devices,which are engineered to mimic the biological olfactory system and biological gustatory system,respectively.The two most important parts in the E-noses or E-tongues are the sensor array and the pattern recognition system.A sensor array consists of dissimilar sensors,which can convert the unmeasurable information to measurable physical signals.A pattern recognition system can be used to mine comprehensive information from the physical signals and present a detecting result.Thus,E-noses and E-tongues can perform those monitoring,identifying,and analysis tasks in a fast,systematic,and scientific way.While,some essential issues need to be addressed to current E-noses and E-tongues.Firstly,E-noses and E-tongues are cost intensive,which against their commerical promotion.Secondly,E-noses and E-tongues are usually low-accuracy in qualitative and/or quantitative detection comparing to some traditional technologies.Thirdly,some environmental factors usually lead to worse performances of the E-noses and the E-tongues.We solved the issues from three points:1)preparing high-accuracy and low-cost sensors for the E-noses and E-tongues,2)establishing an application-driven pattern recognition system in the E-nose and E-tongue to improve their performances in particular detections,3)analyzing the influence of the working environments to suppress the signal drift of the E-nose and E-tongue.In this thesis,we studied three topics on the issues,and got some achievements:(1)We provided a facile method for orientation control of the Metal-organic Frameworks(MOFs)coating on self-assembly monolayer(SAM)foundation by the Layer-by-Layer(LbL)synthetic approach at room temperature.A QCM sensor with high-accuracy and low-cost characteristics was prepared based on the presented method.After investigating the growth mechanism of the high-oriented MOFs coating,we found that:1)the SAM-functional foundation leads to oriented nucleus of the MOFs crystals;2)nucleus with high density grow on the foundation under the condition of the high concentration of the precursor,and the coordination equilibrium is modulated by the concentration;3)the precursor with high concentration will avoid orientation defect units on the surface of the MOFs by removing twining sites of the disordered units.Thus,the proposed method can be used for preparing sensitive coatings with controlled morphology,orientation,and roughness,which is helpful for the reality and efficiency of the data from the sensors.It is a way to improve the performance of the QCM E-nose and E-tongue.(2)We developed three pattern recognition systems for the QCM E-nose/E-tongue in Chinese liquor detections of the brand identification,flavor testing,and products testing with same brand.Organic compound components in Chinese liquors are various in species and disparate in quantities,which leads to much more challenges for Chinese liquor detections.For the poor performance of some current E-nose and E-tongue in Chinese liquor detections,in this thesis,Chinese liquors were detected by using the E-nose and the E-tongue designed by our group,and three pattern recognition systems based on the MDS-SVM(Multidimensional Scaling and Support Vector Machine)algorithm,RF(Random Forest)algorithm,and BPNN(Back-propagation neural network)algorithm were established in the E-nose and E-tongue.This research provides us an simple,real-time,and accurate method for Chinese liquor detections of the brand identification,flavor testing,and products testing with same brand,based on the proposed high-accuracy and good-generalization pattern recognition systems and the E-nose and E-tongue,(3)We investigated the signal drift of the sensors in the E-nose with the temperature change.For Chinese liquor detection,the flow-induced force on each sensor in the sensor box vary apparently due to the temperature change of the volatile gases of Chinese liquor,which is one of important sources of the signal drift of the E-nose.The velocity strealines distribution of the airflow inside the sensor box were presented,and the interactions between the gases and the sensors were studied at different temperatures.By analyzing the velocity streamline distributions,the static pressure distributions,and the flow-induced forces on eight sensors under different temperatures over time,we got the change rules of the flow-induced force on each sensors under the condition of the temperature change.The results of this study provided us quantization parameters to suppress the signal drift of the QCM sensors,which can imporve the stability of the E-nose.