Application Of Terahertz Technology In The Detection Of The Difference Between Forage Species And Species

Identification of forage species and varieties is an important content in forage seed quality inspection.And spectrum detection is a nondestructive detection method,which is important to reveal the properties of materials.In this study,terahertz time-domain spectroscopy(THz-TDS)combined with multivariate stoichiometry was used to identify and discriminate the 3 categroy of 22 forage species.The main work are summarized as follows:Firstly,The time-domain spectra of 14 species of alfalfa forage in terahertz band were obtained by experiments,and their spectral parameters such as refractive index and absorption coefficient in the 0.1~1.4THz effective band were calculated.It can be seen from the atlas that the time delay and amplitude fading of different species of alfalfa are different,and the refractive index spectrum has obvious difference.On the basis of this,the methods of chemical analysis,cluster analysis(CA)and principal component analysis(PCA)were used to assist the verification,According to the clustering results,it was proved that the protein content was an important factor affecting the refractive index spectra of alfalfa,and the PC1 value in PCA could be used to explain the difference among forages.The result show that terahertz spectroscopy can be used to qualitatively analyze the similarity and dissimilarity between forage seeds by combining two statistical methods,CA and PCA.Secondly,Using terahertz time-domain spectroscopy to test the five leguminous forage samples(Lotus corniculatus,Caragana,Mao Tiaozi,Yang Chai and Astragalus adsurgens),and by calculating to obtain the absorption coefficient and refractive index of the sample.Experiments show that,the absorption spectrum of samples with different slope monotonously increasing and the refractive index of the sample varies greatly.At the same time,By using partial least squares regression(PLSR)analysis,it is found that the PLS model constructed with refractive index spectra has better prediction accuracy,which has a higher R-value(0.9983)and lower RMSE values(0.3157).Therefore,it is more accurate to identify the species of unknown samples.In addition,the terahertz time-domain spectroscopy and Fourier transform infrared spectroscopy(FTIR)of the three kinds of gramineous forage were compared and analyzed,and in order to explore the potential applications of the THz-TDS as an auxiliary method of infrared spectroscopy in identification of forage,thus further illustrate the feasibility of THz-TDS technology for identification of forage varieties.