聚类分析在黄霉素发酵过程中的应用

【目的】将聚类分析的方法应用于黄霉素摇瓶发酵条件的优化过程中。【方法】通过系统聚类算法、K均值聚类算法和模糊C均值聚类算法对不同批次黄霉素发酵的摇瓶数据的聚类分析进行比较,发现模糊C均值聚类算法优于其他聚类算法,确定了以模糊C均值聚类算法对黄霉素摇瓶发酵数据进行聚类分析。【结果】然后利用模糊C均值聚类算法选取优质组样本,并利用优质样本优化了黄霉素摇瓶发酵的控制参数分布范围。【结论】这充分证明了聚类分析在发酵过程的优化过程中有良好的实用性。

Application of cluster analysis on Flavomycin fermentation

[Objective] The cluster analysis was applied to the optimization of the Flavomycin flask fermentation. [Methods] The data of Flavomycin flask fermentation from different batches was clustered by using system cluster, K-means cluster and fuzzy C-means cluster. The result showed that fuzzy C-means cluster was better than the other two. [Results] Then the classes were separated and the excellent samples were distilled, moreover, the range of control parameter of Flavomycin flask fermentation was optimized. [Conclusion] The experiment results indicated that cluster analysis was effective in the optimization of fermentation process.