摘要
以取珠后的蚌肉为原料,采用超高压方法提取河蚌多糖。利用响应面试验设计获取神经网络所需的训练样本,利用训练成熟的神经网络进行训练和仿真,分析因素(压力强度、料液比和保压时间)以及因素间的交互作用对河蚌多糖得率的影响,并对超高压得河蚌多糖的工艺进行优化。结果表明:人工神经网络法比响应面法的优化准确性高,预测值的可信度较强;超高压提取河蚌多糖最佳工艺条件为:压力强度340 MPa、料液比1:42(g/mL)、保压时间10 min,该条件下的多糖得率预测值为7.18%,实测值为7.12%,相对误差为0.84。该工艺具有短时、高效、环保等优点,为河蚌多糖的开发利用提供技术依据。
Eased on the raw materials of mussel meat after taking out the pearl, ultra high pressure is used to extract mussel poiysaecharide. The response surface experimental design was used to get the training samples for the neutral network, while the sophisti cated neutral network was used for training and stimulating, analyzing extraction factors （pressure strength, solidliquid ratio and pressure holding time）, the interaction between the factors that affected the extraction rate of mussel polysaccharide, and optimizing the ultra high pressure extraction process of mussel polysaceharide. The results showed that artificial neutral network optimization was more accurate than response surface method optimization, and the reliability of predictive value was greater. The optimum conditions for ultra high pressure extraction of mussel polysaccharide were： the pressure strength 340 MPa, the solid-liquid ratio 42 ： 1 （mL/g）, the pressure holding time 10 min, under this condition, the predicted ex traction rate of polysaccharide was 7.18 %, while the measured value was 7.12~, with the relative error 0.84%. The process has the advantages of short time, high efficiency, environmental protection etc., and provides technical basis for the development and utilization of mussel polysaccharide.
出处
《食品与机械》
CSCD
北大核心
2016年第11期148-153,共6页
Food and Machinery
基金
安徽省教育厅自然科学研究重点项目(编号:KJ2013A182)
2016年高校领军人才引进资助项目
2014年蚌埠学院学术技术带头人及后备人选项目
安徽省质量工程卓越工程师培养计划项目(编号:2013zyjh040)
关键词
河蚌
多糖
人工神经网络
超高压
mussel
polysaccharide
artificial neural network
ultral high pressure
