| 基于BP-ANN的红麻生物酶脱胶预测模型的构建 |
| |
| 引用本文： | 郑来久,李志庆.基于BP-ANN的红麻生物酶脱胶预测模型的构建[J].中国麻业,2007(4). |
| |
| 作者姓名： | 郑来久 李志庆 |
| |
| 作者单位： | 大连轻工业学院天然纤维研究中心，大连轻工业学院天然纤维研究中心 辽宁大连116034，东华大学纺织学院，上海200051，辽宁大连116034 |
| |
| 基金项目： | 国家科技部(2005EC000108)，国家教育部(03028)，辽宁省自然科学基金(2050866) |
| |
| 摘 要： | 由于红麻脱胶环境是一个典型的非线性、强耦合、时变的复杂被控对象,难以实现产业化所要求的实时跟踪控制,其模型难以通过机理建模的方式用简单数学公式或传递函数来描述。将人工神经网络理论应用到红麻韧皮纤维生物酶脱胶过程当中,分析了温度、时间、pH值、浴比和酶浓度等五项主要脱胶工艺参数对脱胶效果的影响水平,经过MATLAB仿真程序的设计和实验,确定了最佳的BP神经网络拓扑结构,建立了红麻韧皮纤维脱胶的神经网络预测模型。
|
| 关 键 词： | BP神经网络 红麻脱胶 生物酶 预测模型 |
Constructing the Forecasting Model of Kenaf Bio-enzymatic Degumming Based on BP-ANN |
| |
| ZHENG Lai-jiu,LI Zhi-qing.Constructing the Forecasting Model of Kenaf Bio-enzymatic Degumming Based on BP-ANN[J].Plant Fibers and Products,2007(4). |
| |
| Authors: | ZHENG Lai-jiu LI Zhi-qing |
| |
| Institution: | ZHENG Lai-jiu1,2,LI Zhi-qing1 |
| |
| Abstract: | Degumming environment of kenaf bast fiber is a complex controlling object being typically nonlinear,strongly coupled and time variation,so it is difficult to realize real-time tracking control needed by industrialization.It is also hard to be described by simple mathematical formulas and transfer function of mechanism modeling.The artificial neural networks(ANN) theory is applied to the bio-enzymatic degumming process of kenaf bast fiber.The effect of five main parameters of degumming as temperature,time,PH value,bath ratio and enzyme concentration was analyzed.The best Back Propagation(BP) ANN topological architecture was defined by designing and testing MATLAB simulative program,and finally forecasting Model of kenaf bio-enzymatic degumming was constructed. |
| |
| Keywords: | BP-ANN kenaf degumming bio-enzyme forecasting model |
| 本文献已被 CNKI 等数据库收录！ |
