质量控制图模式识别方法研究--以SSKJ为例
首发时间:2012-11-29
摘要:在企业中,有效的质量管理是获得市场竞争力、提高经济效益的关键因素。质量管理的主要工具是质量控制图,它是通过产品质量特性对产品生产过程进行实时监控的主要工具。由于质量控制图在企业提高市场竞争力中发挥了巨大的作用,控制图模式识别逐步受到了企业的关注和重视。在对质量控制图模式识别研究的基础上,关于提高模式识别准确率的研究将因企业以及社会的需要而成为必然趋势。本文将蒙特卡罗和BP神经网络结合,借助Minitab软件和Matlab平台,进行初始模式识别;最后,以均方误差(MSE, Mean Squared Error)为根据,对输出值进行转化,将初始识别结果与控制图模式分别对比,并转化为最为类似的控制图模式。本文将最终的均方误差与初始的均方误差进行对比,分析得出本文提出的模型能够大大提高了模式识别准确率的结论,对原有的模式识别方法进行了改进。
For information in English, please click here
Research on the Method of Quality Control Chart Pattern Recognition--Taking SSKJ Company as an Example
Abstract:Effective quality management in the enterprise is the key factor to improve market competitiveness and economic benefits. The main tool of quality management is quality control chart, which is the main tool for real-time monitoring of product quality characteristics of the production process. Because of quality control charts play a huge role in improving mark competitiveness, the control chart pattern recognition gain the concern and attention by enterprises gradually. On the basis of the quality control chart pattern recognition researches, pattern recognition accuracy will become an inevitable trend due to the needs of the enterprise and social. This thesis combined Monte Carlo with BP neural network, using Minitab software and Mat lab platform to exertive initial patter recognition. Finally, according to the mean squared error(MSE), converting the output value, comparing the initial recognition the results with control chart pattern, and converting to the most similar control chart patterns. this thesis compared the final mean square error with the initial mean square error and analyzed it. Then making a conclusion that proposed model can greatly improve the accuracy of pattern recognition, the original pattern recognition been improved.
Keywords: Quality Control Charts Pattern Recognition BP Neural Network MSE
基金:
论文图表:
引用
No.****
同行评议
共计0人参与
勘误表
质量控制图模式识别方法研究--以SSKJ为例
评论
全部评论0/1000