神经网络-遗传算法优化细脚拟青霉多糖超声提取工艺

Study on the optimization of ultrasonic assisted conditions of Paecilomyces tenuipes polysaccharide using ANN-GA

目的优化超声波法提取细脚拟青霉胞内多糖。方法采用单因素法和基于遗传算法的神经网络法(ANN-GA)考察超声功率、超声时间和固液比与细脚拟青霉多糖提取率之间的非线性关系。结果最终最佳提取条件为:固液比1∶73(g∶m L),376 W提取385 s;多糖得率最高达2. 37%,比未优化之前提取率提高了3倍。结论 ANN-GA法优化细角拟青霉超声波法提取多糖,是有效可行的。

Objective To optimize the ultrasonic extractitve conditions of Paecilomyces tenuipes polysaccharides.Methods The single-factor test,artificial neural network(ANN)and genetic algorithm(GA)were used to investigate the extracting conditions,respectively.Results The best extraction conditions were extraction power 376 W,extraction time 385 s and solid/solvent ratio 1∶73(g∶mL).The yield of polysaccharide was 2.37%,which was enhanced by three times.Conclusion Optimization of ultrasonic extraction of Paecilomyces tenuipes polysaccharide using ANN-GA is feasible.