中文摘要：

目的应用星点设计-效应面法优化酸枣仁黄酮滴丸制备工艺。方法以滴距、药液温度和冷却剂温度为自变量。以丸重变异系数、圆整度和溶散时限为指标，采用SPSS软件对实验数据进行多元线性模型和二次多项式模型拟合，得出最佳数学模型，Origin软件绘制效应图和等高线图，再根据效应图优选最佳条件。结果二次多项式模型相关系数优于多元线性模型，复相关系数为0．981，为最终拟合模型；模型的理论预测值与实测值偏差较小，模型具有良好的预测性。结论通过星点设计．效应面法建立的模型预测性良好，可用于对酸枣仁黄酮滴丸制备工艺的优化。

英文摘要：

OBJECTIVE To optimize the formulations of Zizyphi Spinosi Semen flavonoid dropping pills by central composite design-response surface methodology. METHODS Central composite design-response surface methodology was applied to optimize the preparation process with the dropping distance, dropping temperature, and cooling temperature as independent variables, while dependent variables were the variance of weight, spherical degree, and dissolution time. SPSS software was used to fit multivariate linear equation and second-order polynomial equation for experimental data. Response surface and contour plot were delineated according to best-fit mathematic models by Origin software, and the optimum formulation was selected by response surface. RESULTS Quadratic muhinomial model was better than multivariate linear model, and the regression coefficient was 0. 981. The bias between the observed and predicted values of the optimum process was negligible, indicating the high predictability of the model. CONCLUSION The model established by central composite design-response surface methodology and SPSS software is accurate for prediction and can be used to optimize the preparation process of Zizyphi Spinosi Semen flavonoids dropping pills.