应用遗传算法优化铝合金穿孔型等离子弧立焊工艺参数

通过实验并应用基本遗传算法确定等离子孤立焊中获得良好焊缝形状的焊接工艺参数．用这种方法确定工艺参数的优化配置无须建立输入、输出变量的模型，并在较少实验量的基础上获得了工艺参数的近似最优化配置．输出变量是焊缝形状的4个参数：焊缝的正面熔宽、正面余高、焊缝的背面熔宽、背面余高；焊缝形状主要由焊接速度、焊接电流、离子气的流量和喷嘴到工件的距离4个工艺参数确定．结果表明，应用遗传算法确定铝合金穿孔型等离子孤立焊的近似最优化工艺参数是一种有效的方法。

Optimization of aluminum keyhole PAW in vertical position process parameters using a genetic algorithm

Aluminum keyhole plasma arc welding (PAW) in vertical position is a complicated process and the welding quality is affected by many technologic factors. The purpose of this study is to decide optimal settings of the keyhole PAW welding process parameters, which produce the good weld geometry. This method tries to find the optimal settings through experiments without a model between input and output variables. The output variables were front width, front height, back width and back height of welding geometry. These output variables were determined according to input variables, which are the welding speed, welding current, ion gas flux and distance between nozzle and weldment. The result shows that using a genetic algorithm is an effective method for the optimal parameters in the keyhole PAW welding process parameters.