摘要
针对实际生产中摩擦焊接主要工艺参数依据试验和经验选择的局限性,采用改进的三层BP网络,建立了摩擦焊接主要工艺参数选择的神经网络模型。研究结果表明预测值与试验值吻合良好,为摩擦焊接工艺参数优化,提高焊接质量开辟了新的途径。
Aimting at to difficulty to determine the parameter correctly for dissimilar material with different thermal properties, a new method for predicting key welding parameters was introduced, which was developed on the basis of the three layers of back-propagation neural network. The result showed that the predicting values are well identical with the experimental ones.
出处
《焊接学报》
EI
CAS
CSCD
北大核心
2005年第4期33-36,共4页
Transactions of The China Welding Institution
基金
航空基金资助项目(98H53073)
关键词
摩擦焊接
顶锻压力
摩擦时间
神经网络
friction welding
forging pressure
friction period
neural network
