论文部分内容阅读
研究建立同时测定钢铁中的铬、锰的神经网络-分光光度法。在酸性介质中,以硝酸银作催化剂,过硫酸铵作氧化剂,将溶液中的Cr(Ⅲ)和Mn~(2+)氧化成Cr_2O_7~(2-)和MnO_4~-。Cr_2O_7~(2-)和MnO_4~-的质量溶液分别在20～80μg/mL和4～14μg/mL范围内遵守比耳定律,并且二者的吸光度的加和性很好。据此提出不经分离,直接用多层前传误差反向传输人工神经网络(BP-ANN)法和分光光度法测定铬和锰。方法用于测定PCrNi_3 Mo,GCr_(15)和86CrMoV_7三种钢铁样品中的铬和锰,结果与国家标准法相符。
The establishment of a simultaneous determination of chromium and manganese in steel by neural network - spectrophotometry. In acidic media, Cr (Ⅲ) and Mn ~ (2+) were oxidized into Cr_2O_7 ~ (2-) and MnO_4 ~ - with silver nitrate as catalyst and ammonium persulfate as oxidant. Beer ’s law was obeyed in the mass range of 20 ~ 80μg / mL and 4 ~ 14μg / mL for Cr_2O_7 ~ (2-) and MnO_4 ~ -, respectively, and the sum of absorbance of both was good. Based on this, the method of direct reverse error propagation neural network (BP-ANN) and spectrophotometry for the determination of chromium and manganese directly without separation were proposed. The method was applied to the determination of chromium and manganese in three steel samples of PCrNi_3Mo, GCr_ (15) and 86CrMoV_7, and the results were consistent with the national standard method.