| Energy is the most important resource for human survial and social development.The provision of safe,reliable,efficient and sanitary energy is one of the national energy strategies which is the pre-requisit for the development of dynamic economy.Biomass energy,with a abundant of biamss,is considered to be the most potential power in the 21st century.Miscanthus sinensis is one of the biomass energy plants that plays an important role in the energy sustainable development.It has high photosynthesis,high cellulose content.high-yield of biomass,high stress resistance,and minimal impact on other food crops.Miscanthus mainly composed of cellulose,hemicellulose,lignin and ash conent.The high content of cellulose,hemicellulose and the low content of lignin,ash are conductive to the production of ethanol.Since the 70’s of last century,near-infrared spectroscopy technology was an efficient and fast modern analytical techniques,so that it has been widely used in agricultural,food industry,textile,tobacoo,pharmaceutical,petrochemical and other fields.The objective of this work was to determine the content of hemicellulose,cellulose,lignin and ash with chemical method,and to establish accurate and stable calibration models based on near-infrared spectroscopy.The main results were achieved as follows:(1)Determination of lignocellulose and ash content from the stem of M.sinensis,the content of cellulose was higher than others.The average contents of cellulose,hemicellulose.lignin and ash,were 43.78%,32.97%,32.97%and 0.33%respectively,which ranged from 25.96%-52.41%,25.24%～39.29%,1.12%～19.68%,and 0.03%～4.26%respectively,indicating that lignocellulose and ash content had abundant genetic diversity.Among four Miscanthus origin,the US samples had the higher hemicellulose content about 35.57%,the samples from China had the higher cellulose content about 45.60%and lower ash content of 0.28%,and the Japanese samples had the higher lignin content about 35.57%respectively.(2)In correlation analysis,the hemicellulose and cellulose,the hemicellulose and lignin were significantly negative,the correlation coefficient was-0.3807 and-0.5215 respectively.There were no significant differences between other components.(3)Near-infrared spectroscopy technology was applied to build the quantitative relationship between spectral data and the content of lignicellulose.Five kinds of pretreatment methods were compared to build a partial least squares(PLS)model of full spectrum.The results of these models showed that the optimal pretreatmeant methods in hemicellulose,cellulose,and lignin,was smoothing pretreatmeant.The model of cellulose was the best and the R-Square of prediction(Rp2)is 0.943,and the R-Square of prediction(Rp2)of hemicellulose and lignin were 0.938 and 0.863 respectively.(4)A linear method of partial least squares(PLS),nonlinear method of support vector machine(LS-SVM)and radial basis function neural network(RBF-NN)were used to perform models.The results of these models showed that all of the nonlinear method models in the hemicellulose,cellulose and lignin were better than those using linear method.The nonlinear models based on characteristic spectrum of hemicellulose,and cellulose the R-Square of prediction(Rp2)was above 0.920,and for the lignin model was above 0.845.(5)Compared with full spectrum and characteristic spectrum both with partial least squares(PLS),least-squares support vector machine(LS-SVM),radial basis function neural network(RBF-NN).The results of these models showed that the best model of hemicellulose was based on characteristic spectrum of RBF-NN model.The best models of cellulose and lignin were the characteristic spectrum of LS-SVM model.The overall results demonstrated that it was feasible to detect the contents of hemicellulose,cellulose and lignin of M.sinensis using near-infrared spectroscopy,which supplied a new approach for detection the cellulose,hemicellulose and lignin of M.sinensis. |
PDF Full Text Request