响应面法-NSGA-II算法双目标优化金银花罗汉果茶泡制工艺

Bi-objective optimization of brewing process of Lonicera japonica and Siraitia grosvenorii tea using response surface methodology-NSGA-II algorithm

中文关键词:  响应面法  NSGA?II 算法  双目标优化

英文关键词:response surface methodology  non-dominated sorting genetic algorithm II algorithm  bi-objective optimization  Lonicera japonica and Siraitia grosvenorii tea

中文摘要:

      目的 对金银花罗汉果茶进行感官和绿原酸含量的双目标优化, 获得最优工艺。方法 采用单因素实验、Box-Behnken响应面设计对金银花提取工艺进行实验, 再用响应面法和带精英策略的非支配排序遗传(non-dominated sorting genetic algorithm II, NSGA-II)算法对金银花罗汉果茶进行感官和绿原酸含量的双目标优化, 构建预测模型Y1和预测模型Y2, 并对最优条件下制作的茶汤中的绿原酸含量、总酚含量和1,1-二苯基-2-三硝基苯肼(1,1-diphenyl-2-picrylhydrazyl, DPPH)自由基清除率进行检测。结果 最优工艺为罗汉果添加量1.3 g/500 mL, 泡制时间22.3 min、泡制温度88.1 ℃、金银花添加量2.3 g/500 mL, 所得金银花罗汉果茶的感官评分为9.4分, 其绿原酸含量为162.6 μg/g。模型预测结果: 金银花罗汉果茶的感官评分为9.72分, 其绿原酸含量为159.25 μg/g, 模型误差均在5%以内, 验证了模型的预测有效性。最优条件下制作的茶汤中绿原酸含量在前3 d降低较缓, 超过第3 d后降低较快, 总酚及自由基清除率随着贮存时间的延长呈现先增加后降低的趋势。结论 本研究实现了活性成分含量与感官评价的同步提升, 为金银花罗汉果茶工艺优化提供了系统化方法。

英文摘要:

      Objective To conduct bi-objective optimization for the sensory properties and chlorogenic acid content of Lonicera japonica and Siraitia grosvenorii tea, determine the optimal processing parameters. Methods Single-factor experiments and Box-Behnken response surface design were firstly applied to explore the extraction process of Lonicera japonica. On this basis, response surface methodology combined with non-dominated sorting genetic algorithm II (NSGA-II) algorithm was adopted to perform the bi-objective optimization of Lonicera japonica and Siraitia grosvenorii tea in terms of sensory quality and chlorogenic acid content. The 2 prediction models (Y1 and Y2) were established correspondingly. Furthermore, the chlorogenic acid content, total phenolic content and 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical scavenging rate of the tea infusion prepared under optimal conditions were determined and analyzed. Results The optimal processing conditions were determined as follows: Siraitia grosvenorii addition of 1.3 g/500 mL, brewing time of 22.3 min, brewing temperature of 88.1 ℃ and Lonicera japonica addition of 2.3 g/500 mL. Under the optimized conditions, the sensory score of the prepared tea reached 9.4, and the chlorogenic acid content was 162.6 μg/g. The predicted sensory score and chlorogenic acid content of the models were 9.72 and 159.25 μg/g, respectively. The relative errors of all models were less than 5%, which verified the reliable predictive validity of the established models. In the tea infusion prepared under optimal conditions, the chlorogenic acid content decreased slightly within the first 3 days and decreased significantly after the third day. In addition, the total phenolic content and radical scavenging rate showed a trend of first increasing and then decreasing with the prolongation of storage time. Conclusion This study realizes the synchronous improvement of active ingredient content and sensory quality of Lonicera japonica and Siraitia grosvenorii tea, which provides a systematic method for the process optimization and quality improvement of the tea product.