小波分析与蚁群神经网络5′-硝基水杨基荧光酮分光光度法同时测定钨钼锑

来源期刊：冶金分析2010年第5期

论文作者：赵红霞

文章页码：19 - 23

关键词：分光光度法； 多组分； 蚁群神经网络； 小波变换

Key words：spectrophotometry； multi-component； ant colony neural network； wavelet transformation

摘    要：提出了钨、钼、锑-5′-硝基水杨基荧光酮-CTMAB的新显色体系,并用于钨、钼、锑三组分光度分析。为解析钨、钼和锑的重叠光谱,提高多组分同时测定的准确度,将蚁群神经网络(ACAN)与小波分析相结合,建立一种利用小波变换提取重叠光谱信息,再用蚁群神经网络解析的方法,此方法应用于合金钢标准样品中钨、钼、锑多组分同时测定,测定值与认定值相符,相对误差分别为0.14%、0.13%、0.013%。

Abstract: A new simultaneous determination method of tungsten,molybdenum and antimony with tungsten(molybdenum,antimony)-5′-nitrosalicylfluorone-CTMAB coloring system was developed and applied to spectrophotometry.In order to analyze the overlapped spectra of tungsten,molybdenum and antimony,and to prove the accuracy of simultaneous determination for multi-components,the ant colony neural network (ACAN) and wavelet transformation (WT) were combined to establish a new chemometric method to extract useful information.The proposed method has been applied to the simultaneous determination of tungsten,molybdenum and antimony in alloy steel standard samples.The determination results were consistent with the certified values,and the relative errors were 0.14 %,0.13 % and 0.013 %,respectively.