论文部分内容阅读
在介绍了人工神经网络模型特点的基础上，结合堆焊过程的复杂性，对用人工神经网络预测 CO2气体保护堆焊药芯焊丝熔敷金属化学成分的方法进行了研究。研究发现用人工神经网络可以行之有效地解决焊接过程中的一系列复杂的、无法用数学解析式表达的非线性问题。为解决焊接冶金非平衡过程问题探索了一条新的途径。
Based on the characteristics of artificial neural network model and the complexity of surfacing process, the method of artificial neural network for predicting chemical composition of deposited metal in flux-cored wire of CO2 gas-shielded welding was studied. It is found that artificial neural network can effectively solve a series of complex nonlinear problems that can not be expressed mathematically in welding process. In order to solve the problem of welding metallurgy unbalanced, a new way is explored.