基于模糊神经网络的中成药组成药材识别  

Medicine Components Recognition of Traditional Chinese Pharmaceutical Medicine Based on FNN

作　　者：李一波[1,2] 黄小原[1] 

机构地区：[1]江苏大学计算机学院,镇江212013 [2]沈阳航空工业学院,沈阳110034

出　　处：《计算机工程》2003年第22期9-10,55,共3页Computer Engineering

基　　金：辽宁省教育厅科技攻关基金资助项目(20182241);辽宁省自然科学基金项目(972147)

摘　　要：提出一种用于中成药组成药材识别的模糊推理神经网络。该网络借鉴自适应共振（ART）的思想，参考正规化模糊神经网络而构造。网络采用两阶段工作模式，首先根据中药材样本特征学习网络结构，而后再进行集中批学习网络参数。网络为中成药组成药材识别定制，也可用于中药材识别。经实际测试，达到了预期效果，为中成药药方解析、中成药质量控制和鉴定提供了一个新方法。A fuzzy neural network is proposed to recognize medicine components from traditional Chinese pharmaceutical medicine. The network is built on the idea of ART while referring to regular fuzzy neural network. The network takes on two-phase models:first, studies the network structure by samples; and then studies the network parameters in concentrated batch. The network customizes traditional Chinese medicine recognition. It can be used not only to traditional Chinese medicine but recognize the components of the Chinese pharmaceutical medicine by its chromatograph. The experimental results indicate the arrival to the prospective. It is a new method being used to analyze and control traditional Chinese pharmaceutical medicine quality.

关 键 词：中成药 模式识别 模糊神经网络 色谱 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]