摘要
基于人工神经网络原理,以酸性焊条配方为研究对象,在生产数据支持的基础上,建立了反映焊条配方与熔敷金属抗拉强度之间映射关系的神经网络模型。采用BP算法训练网络。研究结果表明,该模型预测的结果同生产实际值之间有很好的对应关系。根据网络估测的结果可定量地进行焊条性能预报。为焊条的设计提供了一种科学方法。
In this paper, based on the principle of Artificial Neural Network, a new approach is provided to predict the
mechanical properties of electrodes. A prediction model of mechanical properties for welding rods was built on the basis of
production data. In the research, a back propagation algorithm is used as the neural network learning rule. The results show:
there are good correlations between the predicted results and the practical production data. Based on the estimated results of
the network, the properties of electrodes can be predicted quantitatively. This research provides a more scientific method for
welding rod design.
出处
《兵器材料科学与工程》
CAS
CSCD
2004年第3期30-33,共4页
Ordnance Material Science and Engineering
关键词
人工神经网络
酸性焊条
焊条设计
性能预报
artificial neural network
welding rod design
property prediction
