采用星点设计-效应面法优化暴马丁香枝条中紫丁香苷的超声提取工艺。以乙醇体积分数、超声提取时间、溶媒比为自变量，紫丁香苷提取率为因变量，通过对自变量各水平的多元线性回归及二项式拟合，用岭嵴分析和效应面法优化出最佳的提取工艺，并进行预测分析。确定紫丁香苷最佳提取工艺为：体积分数为54.64%乙醇17.05倍量，超声提取3次，每次提取85.23 min，超声功率为80 W，温度为30 ℃。最佳工艺验证结果与模型预测值相差1.26%，二项式拟合的相关系数为0.9047，说明二次多项式模型可以较好地描述紫丁香苷提取率与各影响因素之间的关系，可信度大。星点设计-效应面法优化的暴马丁香枝条中紫丁香苷的提取工艺，方法简便，精密度高，质量稳定。因此，该方法可以在实际生产中予以推广使用。

The ultrasonic process to extract syringin from the branches of Syringa reticulata was optimized using a central composite design and response surface methodology. With the ethanol volume fraction, ultrasonic extraction time, and solvent ratio as the independent variables and the extraction rate of syringin as the dependent variable, multiple linear regression analysis and binomial fitting were performed with the independent variables. Ridge analysis and response surface methodology were used to optimize the extraction process, and predictive analysis was performed. The analysis indicated that the optimum conditions to extract syringin were as follows: 17.03-fold amount of 54.64% ethanol, and ultrasonic extraction for three times. Each extraction was carried out for 85.23 min with an ultrasonic power of 80 W at 30 ℃. The deviation between the observed verification results obtained under optimal conditions and the values predicted by the model was 1.26%, and the correlation coefficient of the binomial fitting complex model was 0.9047. The results showed that the quadratic polynomial model could accurately describe the relationships between the extraction rate of syringin and the influencing factors and had a high reliability. Moreover, these findings demonstrated that the extraction process optimized by central composite design and response surface methodology was simple and stable for extracting syringin from the branches of S. reticulata, with high precision. Thus, this method can be used for actual production.