摘要
在锆(钛)一对氯苯基荧光酮-CTMAB显色体系中,应用三层神经网络解析锆和钛的吸收光谱,不经分离分光光度法同时测定锆和钛。使用改进的BP算法,避免了神经网络在学习过程中可以产生的麻痹现象。用于合成样品和钢样中锆和钛的测定,效果较好。
A three-layer artificial neural networks was applied to simultaneous determination of zirconium and titanium by spectrophotometry. The conditons of simultaneous determination by color reaction of Zr(IV) ,Ti( IV ) with p-chlorophenyl fluorone in the presence of CTMAB were researched. The paralysis in the procedure of training was avoided with an improved error back propagation algorithms. This method was applied to the determination of synthetic samples and steel sample with satisfactory results.
出处
《鞍山科技大学学报》
2003年第1期52-53,共2页
Journal of Anshan University of Science and Technology
基金
校董事会青年基金(2002-18)
