回归神经网络红外光谱法同时测定丙酮中苯、甲苯的含量  

RECURRENT NEURAL NETWORKS INFRARED SPECTROSCOPY FOR SIMULTANEOUS DETERMINATION OF BENZEN AND TOLUENE IN ACETONE

作　　者：卞莉山[1] 郭卫民[1] 张绍衡 张懋森 

机构地区：[1]郑州轻工业学院基础课部

出　　处：《郑州轻工业学院学报》1996年第4期63-66,共4页Journal of Zhengzhou Institute of Light Industry（Natural Science）

基　　金：河南省自然科学基金

摘　　要：采用一种基于回归神经网络模型的新方法，对丙酮中所含红外光谱相互重叠的苯及甲苯同时进行定量测定，与最小二乘法处理所得结果相对照。A new method based on recurrent neural network is proposed according to our purpose sinultaneous determination of multicompounds which satisfy the Lambert Beer law.The method combined with infrared spectroscopy to determine benzene and toluene in acetone whose infrared spectrums are overlapped.The results display recurrent neural network based infrared spectroscopy is better than other traditional methods such as the method of the least squares.

关 键 词：回归神经网络 红外光谱法 丙酮 苯 甲苯 

分 类 号：O623.521[理学—有机化学] O657.33[理学—化学]