[1]史荣珍,史国生.小波在自相关过程监控中的应用[J].南京师范大学学报(工程技术版),2010,10(04):031-34.
　Shi Rongzhen,Shi Guosheng.Applying Wavelets to Autocorrelated Process Monitoring[J].Journal of Nanjing Normal University(Engineering and Technology),2010,10(04):031-34.
点击复制

小波在自相关过程监控中的应用
分享到：

南京师范大学学报（工程技术版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
10卷
期数:
2010年04期
页码:
031-34
栏目:
出版日期:
2010-12-31

文章信息/Info

Title:
Applying Wavelets to Autocorrelated Process Monitoring
作者:
史荣珍1;  史国生2
1. 南京师范大学泰州学院机电工程系, 江苏泰州225300; 2. 南京师范大学电气与自动化工程学院, 江苏南京210042
Author(s):
Shi Rongzhen1; Shi Guosheng2
1.Department of Electromechanical Engineering,Nanjing Normal University Taizhou College,Taizhou 225300,China;2.School of Electrical and Automation Engineering,Nanjing Normal University,Nanjing 210042,China
关键词:
多尺度小波分析;  自相关过程;  平均运行链长;  控制图
Keywords:
mu lt-i scale wave lets analysis;  autocorre lated process;  averag e run leng th;  contro l cha
分类号:
TB114.2
摘要:
常规控制图的应用以观测值之间相互独立且同分布为基本前提.然而由于系统内在因素的影响,实际工业过程的观测值往往呈现自相关的特征.此时,常规控制图将不能有效实现对自相关过程的监控.现将多尺度小波变换引入到自相关过程分析中,并与Shewhart控制图相结合,实现对过程的监控.最后以ARMA(1,1)过程为例,分别对过程发生阶跃、趋势、周期和交替模式干扰时进行大量的Monte-Carlo模拟实验,并对该监控方法各扰动下的失控平均运行链长(ARL)做初步分析,验证了该方法的有效性.
Abstract:
T raditiona l contro l charts a re based on the statistical assum ption tha tm easurements are independent and identica lly distributed. In industry applications, how ever, observa tions a re au to corre la ted due to the inherent cause o f the process. Thus trad itiona lm ethods w ill be inappropr ia te for au to co rre la ted process m onitor ing. In this paper, mu lt-i scale w avelets analysis is introduced to autocorre la ted processes. Process mon ito ring is reached by integrating Shewhart contro l charts w ith mu lt-i scale wave lets analysis. Fina lly, take ARMA ( 1, 1) pro cess as an examp le. M onte-Car lo simu lations about step-type, trend- type, cyc les-type, and a lternating-type disturbances in autocorrelated processes are perform ed to expla in the av erage run leng th ( ARL) property o f the contro l charts. The sim ulation resu lts show that the m ethod is e-f fective.

参考文献/References:

[ 1] 贾新章, 李京苑. 统计过程控制与评价, Cp k、SPC 和PPM 技术[M ]. 北京: 电子工业出版社, 2004. Jia Xinzhang, Li Jingyuan. Statistica lProcess Contro l and Eva luation cpk , SPC, and PPM Techno logy[M ]. B eijing: E lectron ic Industry Press, 2004. ( in Chinese)
[ 2] Jun Du, ZhangWu, Roger J Jiao. An improved adaptive CUSUM con tro l chart fo rmon ito ring process m ean[ C] / / Cybernetics and Inte lligent System s. Chengdu: IEEE Press, 2008: 849-854.
[ 3] Danie lW. Apley and hyun cheo l lee design of exponentia lly we ighted mov ing av erage contro l charts fo r autoco rre lated processes w ith mode l uncerta inty[ J]. Technom etr ics, 2003, 45( 3): 187-198.
[ 4] M a Y izhong, Zhao Fengyu. A new approach to d iagnos ing signa ls from m ultivariate EWMA contro l cha rt[ J]. Inte rnational Jou rnal of Plant Eng ineer ing andM anagem ent, 2003, 8( 4): 193-198.
[ 5] Yang Jennhwa ,i Yang M iinshen. A contro l chart pa ttern recogn ition system using a statistical correlation coefficient m ethod [ J] . Com puter and Industria l Eng ineering, 2005, 48: 205-221.
[ 6] 王振龙, 胡永宏. 应用时间序列分析[M ]. 北京: 科学出版社, 2007. W ang Zhen long, Hu Yonghong. App lied T im e Ser ies Ana lys is [M ]. Be ijing: Sc ience Press, 2007. ( in Chinese)
[ 7] YousefA lA ssa .f Recogn ition of contro l chart patterns us ing mu lt-i reso lu tion wave le ts ana lysis and neura l ne tw orks[ J]. Compu-t e rs and Industria Eng ineer ing, 2004, 47: 17-29.
[ 8] Zheng Chen, Susan Lu, Sarah Lam. A hybr id system for SPC concurrent pa ttern recogn ition[ J]. Advanced Eng ineer ing Inform atics, 2007, 21: 303-310.

备注/Memo

备注/Memo:
基金项目: 江苏省高校自然科学基金( 08KJB510005) . 通讯联系人: 史荣珍, 副教授, 研究方向: 工业过程的监测与诊断. E-mail:shirongzhen2007@ yakoo. cn
更新日期/Last Update: 2013-04-02