摘要
探讨利用近红外光谱技术(NIRS)快速测定羊酸奶质构特性的可行性。以不同菌种和工艺条件生产出45种质构特性差异较大的羊酸奶,通过解析样品的近红外光谱图,采用二阶导数(SD)方法进行预处理,运用偏最小二乘法(PLS)建立羊酸奶硬度、黏聚性和保水力质构参数的校正模型,其校正相关系数R2分别为0.6668、0.6461、0.7269,所建模型验证R2分别为0.7577、0.7200、0.7174。对预测值与实测值进行t检验,差异不显著(P>0.05)。结果表明:NIRS适合评价羊酸奶的质构特性,具有一定的准确性与预测能力,但是建模方法值得进一步研究以提高预测的精度。
The aim of this study was to assess the feasibility of using near infrared spectroscopy （NIRS） to rapidly determine three textural properties of goat yoghurt：firmness,cohesiveness and water holding capacity. Forty five goat yoghurt samples fermented by different strains under different process conditions were subjected to near infrared spectral analysis,and their spectra were resolved,and based on pre-processing by second derivative （SD） method,prediction models for firmness,cohesive-ness and water holding capacity were developed with partial least square （PLS）,with determination coefficients （R2） of 0.6668,0.6461and 0.7269,respectively. The correlation coefficient of external-validation for 12 goat yoghurt samples was 0.7577 for firmness,0.7200 for cohesiveness and 0.7174 for water holding capacity. No significant difference was found in t tests between the predicted and experimental values of these three parameters （P 0.05）. The above results show that NIRS is applicable to assess the textural properties of goat yoghurt with good accuracy. However,further studies on statistical modeling need to be done for improving prediction precision.
出处
《食品科学》
EI
CAS
CSCD
北大核心
2010年第19期170-173,共4页
Food Science
基金
农业部公益性行业(农业)科研专项经费项目(3-45)
西北农林科技大学校青年学骨干支持计划资助项目
关键词
羊酸奶
近红外
质构特性
偏最小二乘法
goat yoghurt
near infrared spectroscopy
textural property
partial least squares
