The Study Of Electronic Tongue Based On Virtual Instrument

Electronic tongue is an electronic equipment that simulates human tongue to identify tastes. It is based on the sensor array combined with pattern recognition for a qualitative analysis of liquid. And has wide application in food quality and environment detection areas. It is widely used in testing food quality and detecting environments. Scientists have paid great attention to the research and development of compositive electronic tongue system.In this thesis, we developed an electronic tongue system independently. The electronic tongue consists of three parts, which are the array of ion-selective electrodes (ISE), signal adjustment and acquisition module and data analysis module. Followings are the main contents of the paper.Theoretically we, introduce the common electronic tongue sensor and detection means us mainly introduce the ion-selective electrodes array and electrochemical voltage theory. Sensors array is made up of ion-selective electrodes and reference electrode. The main function of electrodes is responsible for liquid samples.In terms of the signal adjustment and acquisition we, design four-channeled high precision amplified circuit. It can transform the micro-current signals from sensors array to data acquisition card and match the input parameters. A data acquisition card is used to acquire data transformed by signal process module and send them into computer.In the design of the software of electronic tongue system dialogue, we used the software Virtual Instrument. It can achieve man-machine dialogue and reach the goal of instrument control.In light of data analysis, we use multi-algorithm in the pattern-recognition, such as Principal Component Analysis(PCA), Linear Discriminant Analysis(LDA), Artificial Neural Network(ANN), Genetic Algorithm(GA) and there combinational algorithm. They analyze and identify the data from the sensor array, after which the result is displayed by the labVIEW.We use the system of Electronic Tongue to distinguish different brands of mineral water. By using the aforementioned pattern-recognition algorithm, the system can distinguish different brands of water and make comparisons of results between all the algorithms. Therefore, this experiment proves that this electronic tongue has good recognition ability in mineral test. With the adaptive genetic algorithm improving BP neural network, its discrimination can reach 100 percent in five different brands of mineral water in forecasting. In comparison with other classification algorithms, it has a higher aggregate classification rate. As we can see that the feasibility and practicality of this electronic tongue system is verified by the above forecast, we can use it in other liquid identifications.