光学学报, 2014, 34 (12): 1230001, 网络出版: 2014-10-13
径向基神经网络结合近红外光谱技术分析安络小皮伞发酵组分的研究
Study on Analyzing Active Ingredient of Marasmius androsaceus via Radial Basis Function Neural Network Combining with Near Infrared Spectroscopy
光谱学 近红外光谱 蒙特卡罗偏最小二乘法 径向基神经网络 安络小皮伞 spectroscopy near infrared spectroscopy Monte Carlo partial least square radial basis function neural network Marasmius androsaceus
摘要
采用径向基神经网络(RBFNN)结合近红外光谱(NIRS)技术建立一种分析安络小皮伞发酵菌丝体中甘露醇、多糖和腺苷三种组分的定量分析模型。收集164个安络小皮伞液体发酵菌丝体样本的近红外光谱数据,采用常规方法分别测定样本中甘露醇、多糖和腺苷的含量。在应用蒙特卡罗偏最小二乘法(MCPLS)识别异常样本、确定校正集样本数量的基础上,以逼近度(Da)为评价指标,采用可移动窗口径向基神经网络(MWRBFNN)筛选特征波长变量,筛选最佳光谱预处理方法、隐含层节点数(NH)等模型参数。建立甘露醇、多糖和腺苷组分定量分析模型,最佳RBFNN-NIRS模型中校正集和预测集样本实验测定值与预测值间相关系数分别为0.9274、0.9009、0.9440和0.9354、0.9018、0.8847,表明模型具有很好的拟合度和预测性能。
Abstract
Radial basis function neural network (RBFNN) combining with near infrared spectroscopy (NIRS) is applied to develop quantitative analyzing models of mannitol, polysaccharide and adenosine in Marasmius androsaceus fermentation mycelium. Using submerge fermentation, 164 Marasmius androsaceus mycelium samples are obtained. The contents of mannitol, polysaccharide and adenosine are determined via traditional methods and the near infrared spectroscopy data of the 164 samples are collected. The outliers are removed and the number of calibration set is confirmed via Monte Carlo partial least square (MCPLS) method. Based on the values of degree of approach (Da), the moving window radial basis function neural network (MWRBFNN) is applied to optimize characteristic wavelength variables, pre-processing methods, hidden layer nodes (NH) and spreads in the models. The quantitative analyzing models of mannitol, polysaccharide and adenosine in Marasmius androsaceus fermentation mycelium are developed successfully. The correlation coefficients between the reference values and predictive values of mannitol, polysaccharide and adenosine in both of the calibration set and validation set of optimum RBFNN-NIRS models are 0.9274, 0.9009, 0.9440 and 0.9354, 0.9018, 0.8847 respectively. All the data suggest that these models possess excellent fitness and predictive ability.
宋佳, 李臣亮, 邢高杨, 孟庆繁, 逯家辉, 曹家铭, 周毓麟, 王迪, 滕利荣. 径向基神经网络结合近红外光谱技术分析安络小皮伞发酵组分的研究[J]. 光学学报, 2014, 34(12): 1230001. Song Jia, Li Chenliang, Xing Gaoyang, Meng Qingfan, Lu Jiahui, Cao Jiaming, Zhou Yulin, Wang Di, Teng Lirong. Study on Analyzing Active Ingredient of Marasmius androsaceus via Radial Basis Function Neural Network Combining with Near Infrared Spectroscopy[J]. Acta Optica Sinica, 2014, 34(12): 1230001.