摘要
利用人工神经网络可以逼近任何一个有理函数的特点[1],运用人工神经网络来建立6061铝合金的FSW专家系统,通过利用实验数据训练网络,得到预测准确的神经网络,并利用网络寻找6061铝合金FSW最优的工艺参数,得到工艺参数FSW接头。
Friction stir welding process parameter has importantly influence of the strength of FSW joint.The relationship of process parameters and the the quality of welding is very complex,as artificial neural network can approximate any function.The prediction system of 6061 friction stir welding was established by artificial neural network.This system can predict the tensile property accurately,and optimize the welding parameters advisably.
出处
《电焊机》
北大核心
2014年第12期86-89,共4页
Electric Welding Machine
