摘要
以电镦工艺参数作为学习样本,选取了镦粗速度、砧子速度、镦粗压力作为输入参数,加热电流作为输出参数,运 用神经网络BP模型中的Levenberg-Marguardt优化算法对学习样本优化建模,将优化网络的预测结果与实测结果进行 比较,结果表明,BP网络可以很好实现电镦机中加热电流的预测。
The process parameters of electric upsetting acted as studying samples. Velocities of upsetting cylinder and anvil cylinder, pressure of upsetting cylinder used as input parameters, heating current was as target parameter. Studying samples were modeled and optimized with Levenberg-Marguardt algorithm of BP neutral network models. The prediction results of the neutral network were compared with the determined results indicated that the BP network can be satisfactorily used to realize prediction of heating current in the electric upsetting.
出处
《现代制造工程》
CSCD
北大核心
2004年第3期10-11,共2页
Modern Manufacturing Engineering
基金
广东自然科学基金资助项目(990141)
