摘要
矿渣微晶玻璃以废渣和工业尾矿为主要原料,是一种对环境友好的建筑材料。但是由于矿渣微晶玻璃组成、工艺和性能之间关系仍不明确,其新材料的开发仍然采用“炒菜法”,效率低下。我们根据材料设计思想,建立了基于人工神经网络的矿渣微晶玻璃材料设计专家系统。本文详细介绍了矿渣微晶玻璃神经网络专家系统进行材料设计的总体运行过程,并具体说明了系统主界面、神经网络模块和优化模块的界面组成和应用。用实例检验了系统的运行效果,结果表明系统中已经建立的各种模型和知识库已经可以满足一般矿渣微晶玻璃材料设计问题的需求。
The environment-friendly slag glass-ceramics(SGC)take slag and mine tailing as raw materials and could be used in thebuilding industry.However,a new type of SGC is usually developed with“trial and error”method because of little understanding of therelationship between composition,processing and properties.Based on the principles of material design,an expert system(ES)forSGC design was established by adopting artificial neural network(ANN).The applications of the ES were introduced in details and theoperation of main-interface,ANN model and optimization model were also demonstrated.The effect of the system for material design istested to be satisfying.This result indicates that the models and databases in this expert system are reliable for development of slagglass-ceramics materials.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2004年第5期777-780,共4页
Computers and Applied Chemistry
基金
广东省自然科学基金(021289)
广西科学基金(0135020)
