摘要
针对川东造船厂特种船舶制造使用的2205双相不锈钢,在其焊接试验的基础上分别建立了双相不锈钢焊接应力和收缩变形的支持向量机(SVM)回归预测模型,并与传统的逐步回归方法进行比较,较好地模拟了焊接残余应力和变形与板厚、焊接电流、电弧电压、焊接速度等工艺参数之间的非线性关系.计算结果表明,SVM回归比传统的逐步回归在预测精度和泛化能力上都有很大的提高.
With regard to the 2205 duplex stainless steel which is used for special shipbuilding in Chuandong Shipyard SVM regression forecast models on welding residual stress and welding shrinkage of 2205 duplex stainless steel have been respectively established based on the welding experiment of 2205 duplex stainless steel, and non-linear relationships have been better simulated between welding residual stress ＆ deformation and process parameters such as thickness, welding current ,arc voltage and welding speed compared with the traditional method of stepwise regression. The results show that SVM regression is greatly improved both in the prediction accuracy and generalization capability than the traditional stepwise regression.
出处
《船舶工程》
CSCD
北大核心
2010年第1期74-78,共5页
Ship Engineering
基金
重庆市科委科技攻关项目(项目编号CSTS2008AB3033)
关键词
双相不锈钢
焊接应力
残余应力
支持向量机
核函数
duplex stainless steel
welding stress
residual stress
support vector machine
kernel
