Study On Wood Physical And Chemical Decay Properties And Grading Based On Near Infrared Spectroscopy

Wood decay caused great waste, especially rot decay, according to statistics, every year, about7million cubic metres wood has been lost due to rot decay in China, so research on wood rot decay was important. Near Infrared Spectroscopy testing as time-saving and low power consumption testing method has been widely applied in agriculture, medical section and chemical industry. Wood is Macromolecular organic compounds. It can be absorbed by near Infrared. The developing of chemometrics make the near infrared spectroscopy can be employed to detect the quality of wood.Researching the decayed process of Manchurian Walnut and Fir which vaccinated by white rot and brown rot, analyzing the degradation ability to wood crystallinity^lignin and holocellulose of Manchurian Walnut and Fir in different decayed stages and different strains;It builds the Manchurian Walnut and Fir crystallinity prediction models by using the principle partial least square method; It builds the lignin and holocellulose prediction models of Manchurian Walnut and Fir by using the principal component analysis method and principle partial least square method.At last, based on wood crystallinity Fir and Manchurian Walnut were predicted and graded accurately with NIR combined with PCA-BP neural network.The results showed that:(1)The two kinds of strains both have the strong degradation ability to Manchurian Walnut and Fir.To Fir, from0day to120days, the loss rate of wood crystallinity vaccinated by white rot values of26.7%, and21.9%to the brown rot; as to Manchurian Walnut,22.5%for the Fir and20.9%for the Manchurian Walnut. Meanwhile, the loss rate of Fir holocellulose content vaccinated by white rot is49.89%, and41.67%to the brown rot, in the same condition,45.65%and43.53%to Manchurian Walnut. However, to lignin, the degradation ability of these two strains is very different from crystallinity and holocellulose, to Fir, the loss rate of lignin vaccinated by white rot reaches to55.3%, and58.1%to Manchurian Walnut. The Contrast, brown rot has little degradation ability to lignin of Manchurian Walnut and Fir.(2) After the comprehensive analysis and assessment of Manchurian Walnut and Fir crystallinity prediction models built based on the original spectra and the spectral data preprocessed by smoothing method, baseline calibrated method, the firat derivative method and multivariate scattering correction. The results showd that the firat derivative method shows a better result than that by other preprocessing methods.(3) The results show that the principal component analysis method and principle partial least square method can get good prediction results for lignin and holocellulose. Howerer, the principle partial least square method can get a better prediction result than the principal component analysis method.(4) Based on wood crystallinity Fir and Manchurian Walnut was predicted and graded accurately with NIR combined with PCA-BP neural network. For Fir,the grade right rate was86.7%, for Manchurian Walnut, the grade right rate was80%.