遗传算法在极端顶点混料设计的处方优化中的应用  

The Application of Genetic Algorithm in the Formulation Optimization of Extreme Vertices Mixture Design

作　　者：任雯[1] 刘春艳[1,2] 赵磊[1] 王祥[1] 王华芳[1] 魏珍[1] 仇丽霞[1] 

机构地区：[1]山西医科大学公共卫生学院卫生统计学教研室,030001 [2]太原铁路局疾病预防控制所

出　　处：《中国卫生统计》2016年第1期52-55,共4页Chinese Journal of Health Statistics

基　　金：山西省自然科学基金项目(2013011059-3)

摘　　要：目的研究单目标遗传算法对极端顶点混料设计处方优化的效果。方法选用自微乳化释药系统极端顶点混料设计的试验数据,采用全局优化的遗传算法进行混料配比优化,并与原文献中传统的等高线图法优化效果进行比较。结果经单目标遗传算法优化后,自微乳化释药系统的三种成分:油相、表面活性剂、助表面活性剂所占比例分别为33.0%、10.0%、57.0%时,总黄酮类化合物的溶解度为49.12mg/g,比等高线法图求得的值增大1.32mg/g,增加了2.76%;微乳粒径大小为22.92nm,比等高线图法求得的值减小2.02nm,降低了8.10%。结论遗传算法求解的最优解,达到了较好的效果,寻优过程中避免了传统方法存在的主观性和局部最优的缺陷,证明了单目标遗传算法在极端顶点混料设计的处方优化中具有实用价值。Objective To study the application of single-objective genetic algorithm in the formulation optimization of extreme vertices mixture design. Methods The literature results designed by the extreme vertices mixture design will be opti- mized using the genetic algorithm to select the best formulation, which will be compared with ones optimized by the conventional methods of the original text. Results The formulation was optimized by single objective genetic. It was showed that when pro- portion of three components of oil, surfactant and co-surfactant derived from micro-emulsion delivery system was 33.0%, 10. 0% ,57.0% ,respectively,the total flavonoids solubility in the delivery system was 49. 12mg/g more 1.32mg/g, which in- creased 2. 76% than one using the contour diagram strategy and smaller droplet size reached 22. 92nm lowering 1.32nm, which decreased 8. 10% than the contour diagram value. Conclusion Formulation optimization based on genetic algorithm shows the good performance of the algorithm. In the process of optimization, it avoids some shortcomings of classical methods in the appli- cation such as a great deal of subjectivity and local optimum to some extent, so single-objective optimization might be used in the formulation optimization of extreme vertices mixture design.

关 键 词：遗传算法 极端顶点混料设计 单目标优化 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] R911[自动化与计算机技术—控制科学与工程]