论文部分内容阅读
应用ＡＲＴ２人工神经网络算法，使采集到的焊缝横截面方向上的灰度分布数据自组织形成若干种空间模式，并把它们作为典型空间模式存储在ＡＲＴ２人工神经网络的ＬＴＭ中．对实时采样到的灰度分布进行空间模式匹配程度检验，根据模式分布情况确定出焊缝位置．文中对梯度法检测结果进行了分析和比较，结果表明基于ＡＲＴ２人工神经网络的焊缝位置检测方法具有更强的噪声抑制能力，因而检测结果更准确、可靠．
The ART2 artificial neural network algorithm was used to self-organize the grayscale distribution data collected in the direction of the cross-section of the weld to form several spatial patterns and store them as typical spatial patterns in the ARTM artificial neural network LTM. Real-time sampling to the gray-scale distribution of spatial pattern matching test, according to the pattern distribution to determine the location of the weld. In this paper, the results of gradient method are analyzed and compared. The results show that the method of weld position detection based on ART2 artificial neural network has stronger noise suppression ability, and the result is more accurate and reliable.