摘要
利用模式识别技术对影响汽车发动机缸体铸件抗拉强度和硬度的铁水化学成分以及浇注工艺参数进行了分析，给出了影响铸件抗拉强度和硬度的主要因素的优化范围．利用经已知样本集训练的人工神经网络对铸件抗拉强度和硬度进行了预报，预报结果与实测值符合较好．
By means of pattern recoghtion technique the principal factors affecting thetensile strength and hardness of automobile cylinder block cast are discussed, and their optimizedrange determined also. Aritificial neural network, trained by known data, has been used topredict the values of both tensile strength and hardness. The predicted results agree well withthe experimental ones. It has been demonstrated that the mentioned optidsation techniqueprovides a technical approach to disprove the tensile strength and hardness of cylinder block castof automobile engine.
出处
《金属学报》
SCIE
EI
CAS
CSCD
北大核心
1998年第10期1068-1072,共5页
Acta Metallurgica Sinica
基金
福特─中国研究与发展基金!9716214
关键词
模式识别
人工神经网络
铸造
缸体
汽车发动机
pattern recognition
multivariate analysis
artificial neural network
cast
tensile strength
hardness
