基于BP神经网络算法的香料鉴别分析系统的研制
毕业论文关键词：误差反向传播算法；人工神经网络；香精香料；相似度；人工感官评价；电子鼻；电子舌
The Designs of Flavoring Recognize and Analyze System by using BP Neural Network Algorithm
Abstract: When the manufacturer gets an unknown sample, they need to make a rapid identification. This is called quality assurance. Because ingredients are almost different. So we should give out the similarity rate between ingredients and flavorings. According to the rate, the manufacture can meet the demand of production. Nowadays, experts using hierarchical cluster method or PCA method to recognize flavoring data. In order to overcome the disadvantages of two methods, using BP neural network algorithm to instead. This achieved satisfactory results. To confirm the design and implementation. The quantitative relationships between samples are given with each category by using formula. Finally, establish the relationship between the flavoring data and artificial sensory evaluation. It can reduce the complexity of artificial taste and the result of the prediction data has high reliability.
Keywords: BP neural network; artificial neural nets; flavoring; similarity rate; artificial sensory evaluation; electronic nose; electronic tongue
目录
摘要 i
Abstract i
目录 ii
1 绪论 1
1.1 项目背景 1
1.2 研究内容 1
1.3 国内外研究方向 1
1.4 文章结构 2
2 人工神经网络算法——BP算法 3
2.1 神经网络的基本概念 3
2.1.1 什么是人工神经网络 3
2.1.2 人工神经元模型 4
2.1.3 人工神经网络模型 4
2.2 BP神经网络 5
2.2.1 BP神经网络的发展 5
2.2.2 BP神经网络学习算法——最速下降法 5
2.2.3 BP神经网络的局限 9
2.2.4 标准BP算法的改进 10
3 香料数据的分类 12
3.1 香料数据的来源——电子鼻与电子舌 12
3.1.1 电子鼻简介 12
3.1.2 电子舌简介 12
3.2 功能描述 13
3.3 算法详解 14
3.3.1 layer类 14
3.3.2 net类 14
上一篇：模糊控制在变频调速系统中的应用
下一篇：磁悬浮系统的模糊控制器设计