为了靶向利用榛子原料资源，获得更高品质的榛子露产品。从榛子的四大栽培区（北部、中部、南部、干旱及半干旱）收集了共21份榛子样品，按照相同工艺加工成榛子露植物蛋白饮料并测定其感官、理化与营养、加工品质三类指标。结合相关性、主成分分析筛选出榛子露的核心指标5项：b*、可溶性固形物、离心沉淀率、脂肪、蛋白质。层次分析确定核心指标权重依次为：7.198%、10.916%、14.271%、27.657%、39.959%，从而建立榛子露综合品质评价模型。在综合品质得分基础上通过聚类分析实现数据的最优分割。17号（南部安徽辽榛9号）以0.92综合得分排序第一，北部栽培区的辽榛3号表现较佳，推荐加工榛子露，干旱栽培区的榛子表现欠佳，可做其它榛子深加工产品的考虑。上述结果可以为后续企业生产榛子露时，在榛子原料品种选择和榛树选育上提供理论参考。

To obtain high-quality hazelnut milk through the efficient utilization of hazelnuts, a total of 21 hazelnut samples were collected from four main cultivation areas (north, central, south, arid, and semi-arid regions) and processed into hazelnut milk using standard processes and procedures. Three index types were determined (sensory, physicochemical and nutritional, and processing quality indices) and five core hazelnut milk indicators were selected by correlation and principal component analysis: b*, soluble solids, centrifugal precipitation rate, fats, and proteins. An analytical hierarchy process determined that the weights of these core indicators, which were 7.198%, 10.916%, 14.271%, 27.657%, and 39.959%, respectively. Based on these results, a comprehensive hazelnut milk quality evaluation model was established, and based on the comprehensive quality scores, optimal data segmentation was achieved using cluster analysis. Sample no. 17 (Liaozhen no. 9 from Anhui in southern China) ranked first, with a comprehensive score of 0.92. Liaozhen no. 3 from the north also showed good performance and is recommended for hazelnut milk production. Hazelnuts grown in arid cultivation areas showed poor performances, and they should be considered for further processing to produce other hazelnut products. These results provide a reference for raw material selection and hazelnut cultivation for future hazelnut milk production.

辽宁省重点研发计划项目（2020JH2/10200037）