基于PLSR分析的香梨果汁风味品质预测模型的构建及其制备过程中香气丧失机制的研究

Korla pear is one kind of regional characteristics fruit resources, and always favored by consumers both in raw fruit and juices due to their unique fragrance, sweetness, juiciness, and special aroma. However, the significant losses of intense aroma during the processing of juice and their instable flavor have become the bottleneck of restriction the development of Korla pear industry. Based on the previous research achievements, in this project, volatiles were analyzed by solid phase micro extraction (SPME) / stir bar sorptive extraction (SBSE) with gas chromatography-mass spectrometry. Partial least squares regression (PLSR) was applied to quantify the relationships among the volatiles, electronic nose (E-nose) responses and flavor qualities to establish the quantitative predictability models. Further, changes in the aroma characteristics between Korla pear and juice samples which were prepared under different simulated processing were evaluated using principle component analysis (PCA) and PLSR combined with descriptive sensory analysis (DSA) in order to explore the key factors on Korla pear juice flavor quality. It can be also selected and found the significantly different compounds related to natural Korla pear flavor. Meanwhile, the loss or reform pathways of Korla pear characteristics flavor were performed by using isotope labeling, adding significant compounds combined with their process and law interpretation of the function structures, and further revealing the loss mechanism of Korla pear flavor during the processing. On the basis, the results of this study will serve as a theoretical basis for the improvements of stability and upgrading the flavor quality of the Korla pear juice.

香梨为新疆区域特色林果资源，具有香味浓郁、细嫩多汁等特点，除鲜食外香梨汁加工已成为其产业链延伸的主导途径，然而加工中天然果香风味的严重失真及其所导致的产品品质不稳定等问题已成为制约香梨产业发展的瓶颈。本项目基于前期研究成果拟采用SPME/SBSE-GC-MS技术分析香梨鲜果及果汁制备样本的香气成分，结合PLSR分析建立其与E-nose和果汁风味品质间的量化关系，构建香梨汁风味品质预测模型；采用PCA及PLSR结合描述性感官分析明晰鲜果及不同工序果汁样本间潜在的关键影响因素，挖掘与香梨汁样本自然果香呈显著相关的差异及特征香气成分；通过同位素示踪法、差异组分消增模拟技术及所形成的差异组分的官能结构形成历程规律的诠释，追溯香梨汁制备过程中关键风味组分丧失或转化途径，深入解析香梨汁制备过程中香气丧失机制。预期为香梨汁风味稳态化及香梨汁特征风味物质的呈香调控技术及产品品质升级提供坚实的理论基础。

库尔勒香梨是新疆特色林果资源，目前其果汁加工已成为其产业链延伸的主导途径，然而加工中呈香品质劣变导致的产品品质异化等问题已成为制约香梨产业发展的瓶颈。本项目在基于BP神经网络+GA遗传算法高效优化香梨汁挥发性成分萃取分析方法基础上，结合GC-O时间强度法和物质的OAV值筛选获得香梨汁特征香气物质；在此基础上，通过对鲜榨香梨汁和不同操作单元制备香梨汁样本中香气成分的分析鉴定，结合描述性感官分析、PCA降维分析和PLSR技术，挖掘并筛选与香梨天然果香风味和香梨汁不良风味呈现紧密相关的显著差异组分；同时采用同位素示踪技术和差异组分的消增模拟技术，研究了香梨制汁过程中香气丧失或形成的量化轨迹，初步对香梨汁制备过程中香气丧失机制进行了诠释，可为后续高品质果汁风味品质评价与改良及其调控研究技术的开展提供理论依据。研究获得成果如下：. 基于BP神经网络建立的模型具有很好的逼近能力，可以用于固相微萃取过程的预测。在选择的水平范围内，通过遗传算法成功优化了SPME萃取条件，即样品量5.33g、萃取温度45℃、萃取时间25min和NaCl添加量0.63g测出的挥发性物质含量为3.37±0.23μg/g，与最优个体的相对误差为-1.19%，表明BP神经网络结合遗传算法能准确地寻到固相微萃取的最优参数。. 香梨汁中香气物质以酯类、醛类物质为主，通过GC-O的TIM法和香气物质OAV法筛选鉴定出11种香梨特征呈香物质；通过GC-MS对不同加工工序制备的香梨汁样品中挥发性物质进行分离和鉴定，采用PCA降维分析，确定香梨制汁过程中对其风味丧失呈显著变化工艺；在此基础上，采用三种不同杀菌方式处理香梨汁（巴氏、微波和高温瞬时杀菌），通过GC-MS技术研究了杀菌前后香梨汁中香气物质变化，PLSR模型中的Jack-knife不确定检验方法，通过计算挥发性成分对感官属性估计的回归系数的显著性分析，筛选获得13种香梨鲜汁杀菌前后导致风味显著变化呈香物质，采用特定的消增模拟技术初步提出鲜果中特征或关键香气物质的转化途径，最后利用同位素示踪法对提出的特征成分转化途径进行追踪验证，明确果汁差异香气物质转化或丧失途径。