用三聚氰胺生产中排放的固体废物制备三聚氰酸的模型  

Model on the preparation of cyanuric acid using the solid waste from melamine manufacturing

作　　者：任保增[1] 李晨[1] 李玉[1] 雒廷亮[1] 

机构地区：[1]郑州大学化工学院,河南郑州450002

出　　处：《计算机与应用化学》2003年第4期427-430,共4页Computers and Applied Chemistry

基　　金：河南省杰出青年科学基金(9909)

摘　　要：用三聚氰胺生产过程中排放出的固体废物在自生压力下酸解制备三聚氰酸进行资源化利用固体废物研究。结果表明,添加酸式硫酸盐在低硫酸浓度下水解几乎能达到高硫酸浓度下水解的效果。以反应温度、反应时间、NaHSO_4用量为输入矢量,三聚氰酸的质量收率为输出矢量,建立具有两层隐居的BP神经网络能较好地体现酸解反应的规律,训练集和预测集总的平均相对误差为1.2％,对水解反应具有较高的预测能力。反应温度提高,NaHSO_4用量增大可提高三聚氰酸的转化率,反应时间延长对提高三聚氰酸质量收率影响不大。Resource utilization of the solid waste from melamine manufacturing has been studied. The result indicates that cyanuric acid can be prepared by acid digestion under lower acid concentration by adding NaHSO4 in autogenous pressure, and the cyanuric acid mass yield is approximately the same as high H2SO4 concentration. Using reaction temperature, time and NaHSO4 use as input vectors, and cyanuric acid mass yield as output vector to set up BP artificial networks with two hidden layers, the networks can well reflect the acid digestion rule. The mean relative deviation of train and test aggregates is 1.2% , the expect capacity is higher. The high reaction temperature, high NaHSO4 use can increase the mass yield of cyanuric acid, and reaction time rise is insensitive.

关 键 词：三聚氰胺 固体废物 制备 三聚氰酸 资源化利用 神经网络 

分 类 号：TQ323.3[化学工程—合成树脂塑料工业] X78[环境科学与工程—环境工程]