离心法制备Al-16%SiFGM中初晶硅分布BP算法的预测  

Forecast of Primary Crystals Silicons Distribution in Al-16%Si FGM Created by Centrifugal Casting Based on BP Algorithm

作　　者：张新平[1] 于思荣[1] 何镇明[1] 

机构地区：[1]吉林大学,长春130025

出　　处：《材料工程》2002年第8期26-29,共4页Journal of Materials Engineering

基　　金：国家自然科学基金资助项目 (5 97710 5 5 )

摘　　要：采用正交试验法及改进的 BP神经网络考察了型温、浇温及转速对离心法制备 Al- 16 % Si梯度功能材料中初晶硅分布厚度的影响 ,得到了本研究条件下影响初晶硅分布相对厚度程度大小的顺序为型温 >转速 >浇温。 BP神经网络预测结果与正交分析及实验结果一致 ,但神经网络更方便 ,且可预测由于条件限制而无法进行的实验。The influences of mould temperature, pouring temperature and rotating speed on primary crystal silicons distribution in Al 16%Si FGM created by centrifugal casting were studied by orthodoxy experiment and improved BP algorithm Under this paper condition, the influence of mould temperature on primary crystal silicons distribution is lager than that of rotating speed, and that of rotating speed is lager than of pouring temperature The forecasts by BP algorithm and experiment results are in good agreement BP algorithm is more convenient than experiment and it can forecast the results of experimentation can′t be performed

关 键 词：离心法 梯度功能材料 改进BP算法 正交试验 

分 类 号：TG319[金属学及工艺—金属压力加工]