| 铝钢异种材料摩擦焊接质量的人工神经网络识别 |
| |
| 引用本文： | 鄢君辉,王泓,刘小文,薛朝改.铝钢异种材料摩擦焊接质量的人工神经网络识别[J].理化检验(物理分册),2001,37(3):113-116. |
| |
| 作者姓名： | 鄢君辉 王泓 刘小文 薛朝改 |
| |
| 作者单位： | 西北工业大学材料学院 |
| |
| 基金项目： | 航空基金! (97H5 3 10 2 )的资助 |
| |
| 摘 要： | 异种材料摩擦焊接接头中二维及弥散分布的缺陷，常规无损检测方法很难可靠地检测出，因此必须建立异种材料摩擦焊接质量监控的新方法。以铝钢异种材料摩擦焊接为例，利用基于MATLAB的人工神经网络，根据纯铝－钢异种摩擦焊接过程中多个物理参量特征值实时检测结果，首次建立了异种材料摩擦焊接接头性能的分类模型，结果表明，该网络模型的运行效率高，分类成功率可达100％，可以直接用于摩擦焊接生产的可靠性评估，并可推广应用于摩擦焊接缺陷的识别和质量监控，成为一种新的控制手段。
|
| 关 键 词： | 人工神经网络 异种材料 摩擦焊接 焊头质量 二维分布 弥散分布 分类模型 缺陷识别 质量监控 |
| 文章编号： | 1001-4012(2001)03-0113-04 |
| 修稿时间： | 2000年9月29日 |
QUALITY-IDENTIFICATION OF FRICTION WELDING OF ALUMINUM WITH CARBON STEEL BASED ON ARTIFICIAL NEURAL NETWORKS |
| |
| YAN Jun hui,WANG Hong,LIU Xiao wen,XUE Chao gai.QUALITY-IDENTIFICATION OF FRICTION WELDING OF ALUMINUM WITH CARBON STEEL BASED ON ARTIFICIAL NEURAL NETWORKS[J].Physical Testing and Chemical Analysis Part A:Physical Testing,2001,37(3):113-116. |
| |
| Authors: | YAN Jun hui WANG Hong LIU Xiao wen XUE Chao gai |
| |
| Abstract: | It is difficult by traditional nondestructive testing methods to effectively detect the dispersively distributed two dimensional defects in friction welded joints of dissimilar metals, so new methods should be developed to monitor and control the quality of friction welded joints of dissimilar metals An arificial neural networks (ANN) model is presented to identify the quality of friction welded joints of aluminium with carbon steel according to the real time testing results of characteristic values of several physical parameters such as pressure, displacement and torque in the process It is shown that the ANN model based on the MATLAB circumstance works with high efficiency and classification accuracy is up to 100%, and may be further applied to assess the reliability of friction welding and form a new method for quality control |
| |
| Keywords: | Artificial neural network Dissimilar metals Friction welding Quality of joints Classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
