摘要
应用三层ANN_BP网络解析苯和甲苯的吸收光谱 ,不经分离 ,用紫外分光光度法同时测定苯和甲苯 .使用改进的BP算法 ,避免了神经网络学习过程中可能产生的麻痹现象 ,提出了目标向量的简单变换方法及便于网络参数选择的收敛评价函数 .实验验证了方法的有效性 ,同时表明 ,苯在 0 - 72 0mg/L和甲苯在 0 - 84
In this paper,a three_layer artificial neural networks along with ultroviolet spectrophontometry has been applied to simultaneous determination of benzene and methyl benzene.The paralysis produced by ANN's learning has been avoided with the improved BP algorithms.A simple method for transformation of target vectors and a evaluation function for selection of networks parameters have been put forward and proved by an experiment.The experimantal result shows that benzene and methyl benzene obey Beer's law in 0-720 mg/L and 0-840 mg/L respectively.
出处
《鞍山钢铁学院学报》
2002年第3期168-170,共3页
Journal of Anshan Institute of Iron and Steel Technology
关键词
人工神经网络
紫外分光光度法
苯
甲苯
测定
artificial neural networks
ultraviolet spectrophotometry
benzene
methyl benzene
