Realization Of Fruit Quality And Variety Recognition System Based On Electronic Nose Technology

The electronic nose is an artificial olfactory device that simulates the human olfactory process.Electronic nose technology can not only make up for the subjective shortcomings of human senses,but also has the advantages of better repeatability and shorter detection time.With the rapid advancement of science and technology,electronic nose technology has been widely used in agriculture,medical care,food,security,environmental testing and other fields,and has gradually become an indispensable testing tool for people.At the same time,people’s requirements for fruit quality and variety are gradually increasing.The price of different quality fruits is different,and the price of different types of fruit varies greatly.Therefore,it is of great significance for fruit variety identification.In this paper,after introducing the development and background of electronic nose technology,the composition of the self-made electronic nose system is introduced in detail,and fruit variety and quality recognition experiments are carried out through the electronic nose.Finally,the PCA and BPNN used to distinguish the variety and quality of the fruit.The main work content is as follows:(1)According to the working principle of the electronic nose,design the electronic nose hardware system.The hardware system includes: gas sampling device,sensor array,sensor array reaction gas chamber and data acquisition system.The sensor is selected according to the gas released when the fruit is stored.The reaction gas chamber is simulated through the workbench simulation software,and a reasonable gas chamber is designed.The process of uploading the collected signals to the PC is completed through the digital-to-analog conversion chip and the microcontroller.(2)Use LABVIEW to write the upper computer program,and realize the real-time display and storage function of the response signal generated by the sensor array.(3)Using a self-made electronic nose system,the three sample fruits of fresh Liaoning grapes,fresh blackcurrants,fresh Yunnan red grapes,were identified.The BPNN pattern recognition method is compared with Method of combining PCA and BPNN.The method of combining PCA with BPNN is the best,and the accuracy can reach 83.33%.At the same time,two different qualities of grapes are also distinguished and identified,which can be successfully identified.