Evaluation of Damage Severity of Tea Plants Using Electronic Nose

Abstract

Abstract. In this study, an electronic nose (E-nose) was used to evaluate the damage severity of tea plants, and a new evaluation index (mass loss) was introduced to reflect damage severity. Gas chromatography-mass spectrometer (GC-MS) was employed for proving the potential of the E-nose to detect tea plants with different damage severities. The number of pests attacking tea plants and the time under attack are two traditional evaluation indexes that are widely applied. The prediction performance of mass loss was compared with the number of pests and time under attack based on partial least squares regression (PLSR) according to the correlation coefficient (R2) and root mean square error (RMSE), and the results showed that the prediction performance of mass loss was better than that of the other two indexes. Three regression algorithms, namely PLSR, extreme learning machine for regression (ELMR), and support vector regression (SVR), were applied to predict mass loss, and their performances were compared. The results indicated that these three algorithms all had good performances, and SVR was the best. It could be concluded that E-nose is a feasible technique for evaluating the damage severity of tea plants, and mass loss is an appropriate evaluation index for damage severity. Keywords: Damage severity, Electronic nose, Mass loss, Regression algorithm.