人工神经网络紫外光谱方法测定牙周康胶囊的含量  

Quantitative determination of contents of Yazhoukang capsule by artificial neural network spectrophotometry

作　　者：刘迎春[1] 金杰[1] 赵兵[1] 曾令欢[1] 

机构地区：[1]沈阳药科大学制药工程学院,辽宁沈阳110016

出　　处：《化学试剂》2005年第12期732-734,共3页Chemical Reagents

摘　　要：应用人工神经网络误差反向传播的方法对紫外吸收光谱重叠的牙周康胶囊进行组分不经分离的含量测定,网络隐蔽层的节点数为5,输入节点数为10时,甲硝唑和芬布芬的平均回收率分别为99.7%和99.9%,RSD分别为0.42%和0.45%。测定方法结果准确,操作简单、方便。对紫外吸收光谱重叠的药物来说,该法提供了一种含量测定的新途径。To determine the contents of Yazhoukang capsule in which UV spectra are overlapped. The error back-propagation method of artificial neural network has been used to determine Metronidazole and Fenbufen without separating them. When hidden layer node of ANN was 5 and input layer node equaled to 10,the calculation results showed that the average recoveries of Metronidazole and Fenbufen were 99.7% and 99.9%, RSD were 0.42% and 0.45% respectively. The results showed that the method was accurate and the network performed satisfactorily. For those drugs with UV spectra overlapped, this method supplied a new way of content analysis.

关 键 词：人工神经网络 误差反向传播 紫外吸收光谱 牙周康胶囊 含量测定 

分 类 号：R917[医药卫生—药物分析学]