摘要
介绍了主成分分析方法及人工神经网络技术在相关因素分析和质量控制的建模与估计中的应用.以大电流MAG焊熔宽控制为例,通过对6个焊接过程参数进行主成分分析,提取出影响熔宽的4个主要因素,讨论了提取的主成分与原始过程参数间的关系.以主成分得分作为新的训练样本集,送入神经网络进行计算.结果表明,基于主成分分析的神经网络无论在收敛速度,还是在训练精度上,都远远优于基本BP神经网络.
The application of principal component analysis (PCA) and artificial neural networks (ANN) to the multivariate statistical analysis and quality control was introduced. The pool width control of MAG weld with high current was taken as an example. Through the PCA of six welding parameters, four main factors were extracted. The relationship between the main factors and the original parameters was discussed. The principal component values were taken as a new training sample set. The output results indicate that both the convergent speed and the training accuracy of PCA-based ANN are much better than those of basic BP ANN.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2003年第10期1536-1539,共4页
Journal of Shanghai Jiaotong University
关键词
焊接
质量控制
主成分分析
人工神经网络
BP算法
Backpropagation
Data processing
Neural networks
Principal component analysis
Quality control
