Research On Fast Recognition Technology Of Apple Freshness Features Based On Gas Sensor Array Optimization

Apples are favored by consumers for their taste and high nutrients.However,due to problems such as mechanical damage and fungal infection,the appearance and internal physiological characteristics of apples will change during the process of growth,maturation,storage,transportation,and processing.A series of problems such as false ripening and rot of the fruit will cause huge problems.The economic loss also brings immeasurable harm to consumers.Therefore,the research on the rapid detection technology of apple freshness characteristics has become an important issue that needs to be solved at present.In this paper,a set of odor recognition system is developed for the detection of the freshness characteristics of Red Fuji apples,which can be used to quickly evaluate the quality of apples.By investigating and studying the changes in the concentration of ethylene gas during apple growth and storage,as well as the effect of changes in ethylene concentration on the concentration of other gases emitted by the apple,the ethylene gas sensor was added to the sensor array for the first time,and the effect was significant.A series equivalent model of mixed media of apples and air is established,and an experimental platform built with TH2822 A handheld LCR meters,shielding boxes,and self-made parallel plates,etc.,is used to test the change trend of the dielectric constant characteristic parameters of the apple during the entire storage process.The result of the analysis of the dielectric constant of the apple is used as the standard for the classification and evaluation of the freshness of the apple.The continuous projection algorithm(SPA)is used to optimize the sensor array to solve the problems of collinearity and overlap and eliminate abnormal and redundant sensors.Use the median absolute difference algorithm(MAD)and least squares filtering to carry out data preprocessing.The BP neural network algorithm optimized by the hybrid frog leaping algorithm(SFLA)is used to perform pattern recognition on the gas data.The characteristic value within 1 minute is optimized through principal component analysis and SPA algorithm.Under the premise of ensuring accuracy,the detection time is optimized from 5 minutes to 1minute,which shortens the detection time as much as possible,and finally completes the detection within 1 minute.The freshness characteristic detection of Red Fuji apples realizes the rapid detection of the freshness characteristic of apples.The experimental results show that the unoptimized sensor array test accuracy rate reaches 94%,the optimized sensor array test accuracy rate reaches 95.33%,and the 1-minute apple freshness feature detection accuracy rate reaches 84.67%.It is compared with the commercial electronic nose on the experimental device.The detection accuracy is equivalent,which can quickly and comprehensively identify the freshness of apples.