异亮氨酸发酵过程中浓度近红外预测模型的建立  

Establishment of near-infrared prediction model of concentration in the isoleucine fermentation

作　　者：郭宇飞[1] 石拓[2] 徐庆阳[2] 

机构地区：[1]天津现代职业技术学院,天津300350 [2]天津科技大学生物工程学院,天津300457

出　　处：《发酵科技通讯》2017年第3期138-142,共5页Bulletin of Fermentation Science and Technology

摘　　要：利用近红外分析仪和相关近红外软件,建立谷氨酸棒杆菌发酵过程中异亮氨酸质量浓度的近红外预测模型.结合偏最小二乘法,在波数为7 400~9 100cm^(-1),减去一条直线作为光谱预处理的条件下,获得异亮氨酸质量浓度最优近红外预测模型.该模预测偏差(RPD)为3.36,交叉验证误差均方根(RMSECV)为1.73g/L,决定系数(R2)为0.997.通过外部验证,该模型能够较好地检测发酵过程中异亮氨酸的质量浓度.The near-infrared spectrum scanning analysis was carried out using near infrared analyzer and related software,and the optimal prediction model for isoleucine concentration was established using near-infrared spectroscopy combining with partial least squares method under conditions of a straight line minus at wavenumber between 7 400cm-1 and 9 100cm-1.The value of root mean square error of cross validation（RMSECV）,determination coefficient（R2）and residual prediction error（RPD）was 1.73g/L,0.997 and 3.36,respectively.Through external validation,the model could well predict the concentration of isoleucine in the fermentation process.

关 键 词：近红外 预测模型 异亮氨酸 谷氨酸棒杆菌 

分 类 号：TQ922[轻工技术与工程—发酵工程]