摘要
基于BP网络的强大缺陷诊断功能 ,构建了灰铸铁件缺陷诊断神经网络模型 ;针对BP算法学习网络权值收敛速度慢 ,易陷入局部最优的缺点 ,采用混合遗传算法加速网络权值的学习。
A neural network model for gray iron casting fault diagnosis is constructed based on the mighty fault diagnosis function of BP neural network;To avoid the network get trapped in some local minimum and also to accelerate the weights learning,the hybrid genetic algorithms is used to learn BP neural network weights in this model;It is proved that the model can lower waster rate of gray iron casting effectively.
出处
《铸造》
CAS
CSCD
北大核心
2002年第3期177-179,共3页
Foundry
关键词
混合遗传算法
BP算法
神经网络
缺陷诊断
灰铸铁件
hybrid genetic algorithm
BP algorithm
neural network
fault diagnosis
gray iron casting
