摘要
采用一种新颖的基于WEB的人工神经网络-遗传算法系统对介孔立方ZrO_2球形粒子准气相反应合成的新工艺进行了建模与优化设计。结果表明,合成最小粒径ZrO_2前驱物球的最优参数为:M_([ZrOCl_2·8H_2O])=0.49 mol/L,Vs=0.035 m^3/h,P=99.80 kPa;ZrO_2前驱物球的最小平均粒径为0.75μm。通过实验验证,此最优参数下合成的ZrO_2前驱物球平均粒径为0.72μm,计算值与实验值符合的较好。将合成的最小粒径前驱物球经干燥和600℃煅烧后,得到了平均粒径为0.64μm的介孔立方ZrO_2球。
A novel web-based artificial neural network-genetic algorithm system was applied to new quasi-gaseous state reaction synthesis of mesoporous cubic ZrO2 spherical particles. The synthesis process was modeled by artificial neural network and the processing parameters were optimized by genetic algorithm to obtain minimum size ZrO2 precursor spheres. Computation results show that the optimization processing parameters are： M[ZrOCl2·8H2O]-0.49 mol/L, Vs-0.035 m^3/h, P-99,80 kPa, and the minimum size of ZrO2 precursor spheres obtained with these parameters is 0.75 μm. Correspondingly the experimental value of diameter of ZrO2 precursor spheres synthesized with the optimized parameters is 0.72μm, which is quite consistent with computation value and becomes 0.64μm after drying and calcination at 600 ℃.
出处
《稀有金属材料与工程》
SCIE
EI
CAS
CSCD
北大核心
2007年第A03期519-523,共5页
Rare Metal Materials and Engineering
基金
国家自然科学基金项目(59974026)
北京交通大学论文基金项目(PD297)资助
关键词
人工神经网络-遗传算法系统
准气相反应法
介孔立方ZrO2
artificial neural network-genetic algorithm system
quasi-gaseous state reaction method
mesoporous cubic ZrO2
