论文部分内容阅读
以改进型遗传神经网络为基础,建立焊接变形预测模型,用于寻找焊接工艺参数和焊接变形之间的复杂非线性函数关系;再用该变形预测模型代替目标函数,建立以焊接工艺参数为优化变量、以横向收缩变形和角变形为优化目标的多目标优化模型。利用该多目标优化模型,在基于正交设计原理所构建的焊接工艺参数组合备选集中,优化得到符合条件的最优工艺参数组合。用Matlab语言和GUI技术开发了SMAW工艺参数优化系统,经实例验证,结果表明:本参数优化系统可有效应用于SMAW焊接工艺参数优化的工程实际中。
Based on the improved genetic neural network, the welding deformation prediction model is established to find the complex nonlinear function relationship between the welding process parameters and the welding deformation. The deformation prediction model is used to replace the objective function, and the welding process parameters are optimized Variable, multi-objective optimization model with lateral shrinkage deformation and angle deformation as optimization objectives. By using the multi-objective optimization model, the optimum process parameters combinations that meet the requirements are optimized based on the orthogonal selection of welding process parameters. The SMAW process parameter optimization system was developed with Matlab language and GUI technology. The experimental results show that this parameter optimization system can be effectively used in engineering practice of SMAW welding process parameter optimization.