摘要
通地小波变换压缩数据 ,以压缩后的数据作为人工神经网络的输入 ,极大地加快了网络的收敛速度 .将该法应用于测定撒痛风注射液 ,结果良好 .3组分的平均回收率及相对标准偏差分别为 :10 0 1% ,1 43 % ;10 0 0 % ,0 5 0 % ;10 0 6 % ,1 76 % ,优于单纯人工神经网络 .
The data of UV-spectrum were compressed by wavelet transform,the compressed data were used as an input of artificial newral network.When the method was applied to determination of the Satongfen Injection,the convergent rate was improved.The mean recovery and relative standard deviation of the three components were 100 1%,1 43%,100 0%,0 50% and 100 6%,1 76% respectively.The results of this method are better than that of the artificial neural network.
出处
《沈阳药科大学学报》
CAS
CSCD
2000年第6期419-421,共3页
Journal of Shenyang Pharmaceutical University
