摘要
目的探讨遗传算法在均匀试验设计最优条件选择中的应用。方法利用甘草苷提取工艺均匀试验设计结果,以中心化二次回归模型为目标函数,用遗传算法搜索最优试验条件。结果甘草苷中心化二次回归模型有统计学意义,遗传算法确定的最优试验条件:52%的甲醇溶液冷浸11 h,超声提取77 min,甘草苷提取量预测值达到了29.65 mg/g,比2号试验最高提取量增加了2.67 mg/g,提高了10%。结论以二次回归模型为目标函数,利用遗传算法确定的均匀试验最优条件客观性强、精度高,为均匀试验设计优化分析提供了合理的新方法。
Objective To explore the use of genetic algorithm in selecting the optimal condition for uniform experimental design. Methods Using data from liquirtin extraction uniform experiment, optimal experimental condition was searched by genetic algorithm （GA） based on the quadratic regression model as a target function. Results The quadratic regression model was found to be significant. The optimal conditions of extraction process were as follow： soaking the samples in 52% methanol for 11 h followed by an ultrasonic bath for 77 min. The estimated recovery of liquirtin was 29.65 mg/g, with 2.67 mg/g more than that in the second experiment. Conclusion The optimal experimental condition searched by GA was more precise and objective. The new method provided a technical basis for uniform experiment design.
出处
《中国药物与临床》
CAS
2009年第8期695-697,共3页
Chinese Remedies & Clinics
基金
山西医科大学学生创新项目基金培训项目(200757)
山西医科大学科技创新基金项目(01200715)
关键词
遗传算法
均匀试验设计
最优试验条件
Genetic algorithm
Uniform experimental design
The best experimental condition
