一种用于中药色谱指纹图谱相似性判别的分析仪  

A Novel Analyzer of Traditional Chinese Medicines Fingerprint

作　　者：沈晋慧[1] 曹辉[1] 马栋萍[1] 

机构地区：[1]北京联合大学生物化学工程学院信息与控制工程系,北京100023

出　　处：《生物医学工程学杂志》2010年第5期1011-1015,共5页Journal of Biomedical Engineering

基　　金：北京市教委科研基金资助项目(KM200711417014)

摘　　要：中药指纹图谱是对中药物质基础的一种整体表达方式,能够很好地反映中药复杂的化学成分特征,是目前被国内外广泛接受的一种有效的中药质量分析手段。研究基于BP神经网络的液相色谱指纹图谱相似性判别算法,该套算法重注实时及并行运算能力,能够移植到现场可编程门阵列(FPGA)硬件架构上,实验结果表明,该分析仪具备较高的可靠性,可用于中药炮制现场进行质量控制等。The chromatographic fingerprint of traditional Chinese medicine is a way of comprehensive expression to show its complicated chemical composition.It is a generally-accepted method for analyzing the quality of traditional Chinese medicines.In this article,we address an analysis algorithm of traditional Chinese medicines fingerprint and its FPGA circuit method.With the use of a method based on combined GA training BP-NN,and by virtue of high-speed operation and parallel computing,the system is implemented successfully with FPGA which can be used in field control.The result shows that this is an efficient and superior method for use in the process of preparing raw Chinese medicines.

关 键 词：色谱指纹图谱 现场可编程门阵列 遗传算法 BP神经网络 

分 类 号：R318.6[医药卫生—生物医学工程]