电弧焊焊接过程数据快速提取方法

Fast Data Extraction Methodology of Arc Welding Process

为了获取全熔透情况下建立焊缝熔透人工神经网络预测模型所需的大量数据 ,提出了一套获取焊接熔池正面几何参数和焊缝背面熔宽的方法 ,并在Windows环境下采用VC + +开发了相应的软件。在小电流薄板脉冲TIG焊接全熔透的试验条件下进行了试验 ,并获得了所需数据。试验证明这种方法高效便捷 ,是可行的 。

In order to acquire mass data for training the neural network predictive model of weld penetration under the conditions of full penetration, a set of methods were proposed. The relevant software was developed under Windows environment with Visual C++. Experiments were conducted under the conditions of TIG welding with small current and thin plates. The desired dada was extracted. Experimental results show that the methods are effective, convenient and feasible. The methods lay the foundation of establishing the artificial neural network model for penetration prediction.