摘要
将人工神经网络技术用于质量控制图的异常趋势识别，采用反向传播算法（ＢＰ算法）训练的神经网络能够很好地识别控制图中点于逐渐上升／下降、突然上升／下降、循环变化等异常趋势。
Back propagation artificial neural networkds are used to identify unnatural patternson quality control charts. These unnatural patterns, such as trend, sudden shift and cycle, whichare very difficult to be identified by normal algorithms, can be quirkly identified by a trainedneural network.
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
1995年第6期121-124,共4页
Journal of Harbin Institute of Technology
关键词
神经网络
模式识别
质量控制图
BP算法
自动识别
Artificial neural networks, statistical process control, quality control, pattern recognition
