支持向量机结合近红外光谱法测定杉木木质素的含量

摘要

Via near-infrared spectroscopy (NIR)combined with support vector machine (SVM),a model for determining lignin content of Chinese fir wood was established.Forty-seven Chinese fir samples were used as experimental material.Lignin content of samples was measured according to wet-chemical method.The spectra of Chinese fir samples were recorded by near infrared spectrometer.The spectral data were pretreated by smoothing,derivating,compressing and scaling.Radial basis function ( RBF) was used as kernel function to establish a model for determing lignin content.The square sum of relative calibration error and relative prediction error were 0.007433 and 0.001219,respectively,which demonstrated that this method is precise for determination of lignin content of Chinese fir wood.%采用支持向量机(SVM)结合近红外光谱(NIR)技术建立测定杉木中木质素的定量分析模型.以47个杉木样品作为实验材料,用常规方法测定了样品中木质素的含量,用近红外光谱仪采集相应的光谱,对光谱数据进行平滑、求导、小波压缩以及归一化,结合支持向量机,以径向基(RBF)作为核函数,建立了测定杉木中木质素含量的模型.校正相对误差的平方和为0.007433,预测相对误差的平方和为0.001219.结果表明,该方法测量比较准确,可以用于杉木中木质素含量的预测.