人工神经网络建模用于辣椒籽油提取工艺优化  

Optimization of extraction on Capsicum oil with artificial neural networks

作　　者：罗乐[1] 孙益民[2] 

机构地区：[1]安徽师范大学皖江学院,芜湖241008 [2]安徽师范大学化学与材料科学学院,芜湖241000

出　　处：《食品科技》2017年第4期176-182,共7页Food Science and Technology

基　　金：安徽省高等学校省级优秀青年人才基金项目(2012SQRL264)

摘　　要：运用人工神经网络建模预报实验因素对指标的影响规律,并针对产品特性进行工艺优化。分别从热力学、动力学及索氏提取溶度3个不同的角度对酸值和产率的影响规律进行了较为系统的分析。研究解决了提取工艺参数对提取效果高度非线性映射的复杂问题。不仅应用人工神经网络预报出提取工艺最佳参数,而且提出根据产品酸值或产率的靶向优化概念,计算了成套优化工艺。该模型既可进行正向预报,也可对体系进行逆映射从而完成反向预报,这一特点为辣椒籽油提取工业化和目标化生产提供了科学依据,利于加快生产周期,节约生产成本。The study predicted the influence of target indexes from process factors by the model built with artificial neural networks（ANN） and the special optimizations related to the specification of the products. The relationship between three main factors such as extracting temperature, extracting time and solidliquid ratio and acid value and productivity was more systematically studied with the built model. The study investigated and solved the complex and nonlinear mapping problem involved in extracting efficiency and extracting process parameters. The study couldn＇t only calculate and predicte the better range of process but also made the primary investigation where targeting optimizations for different products with various acid value and productivity were proposed and put in practice. The study offered the scientific basis for the extraction of capsicum seed oil in industrialized and objectified style to speed up produce and save cost.

关 键 词：辣椒籽油 均匀设计 人工神经网络 靶向优化 

分 类 号：TS255.19[轻工技术与工程—农产品加工及贮藏工程]