Research on Grape Storage Time Identification Method Based on Filtering and Improved Grey Wolf Algorithm
A method of grape storage time identification based on electronic nose system is studied.In the experiment,the physical and chemical indexes of grapes with different storage time are detected,and the volatile odor information of grape samples with different stor-age time is obtained by using electronic nose.The principal component analysis method,kernel principal component analysis method and improved grey wolf optimized support vector machine based on least squares smoothing filter are used to extract and analyze the vol-atile odor information of grapes,and the detection and classification model of grapes with different storage time is established.The re-sults show that nuclear principal component analysis is more effective in distinguishing grapes with different storage time.The accuracy,precison,recall and F1 value of the improved GWO-SVM classification method based on the least squares smoothing filter reach 97.56%,97.59%,97.58%and 97.59%respectively.Its classification performance index is higher than those of other classification methods,and has good stability,which can provide a reference for the identification of grape storage time in practical applications.
electronic noseprincipal component analysisfilteringGrey Wolf Algorithmsupport vector machine
万方数据
维普
