油脂降解培养基优化的遗传算法实验研究

我们运用遗传算法，首次对利用热带假丝酵母降解油脂的培养基优化问题进行了研究。通过5代实验完成了50个实验样本、6种培养基成分、64个浓度水平的优化任务。按照优化后的培养基组成，热带假丝酵母降解油脂的降解率为95．4％，比正交设计法提高了7．63％。实验结果表明：利用遗传算法可优化培养基成分含量，取得更好的效果；遗传算法是一种优于单因素试验法和正交设计法的新型、高效的培养基优化方法。

Experimental Study on Genetic Algorithms of Medium Optimization for Oil-degradation

Medium optimization of the culture for oil degradation by Candida tropicalis using genetic algorithms (GA) was studied. The 64 concentration levels of 6 medium components were optimized within 50 experiments. The results show that GA can be applied in the medium optimization and better results are obtained. Employing optimized medium composition, a final oil degradation rate of 95.4% was obtained and it was improved by 6.94% compared to the results with the medium before optimization. GA is an efficient new medium optimum method which is superior to the other optimum means such as orthogonal array method.