1 College of Energy Environment and Safety Engineering, China Jiliang University, Hangzhou 310018, China
2 College of Quality and Standardization, China Jiliang University, Hangzhou 310018, China
3 Linhai Weixing New Building Materials Co., Ltd., Taizhou 317000, China
Show less
文章历史+
Received
Published
Revised
2025-07-22
2026-06-18
2025-12-16
Online Date
2026-08-04
PDF (6028K)
摘要
传统指数加权移动平均(exponentially weighted moving average,EWMA)控制图依赖于数据的正态分布假设,缺乏实时反馈和自适应能力,限制了其在实际生产过程质量监控中的应用。提出一种基于旋转门算法(sliding rotation algorithm,SRA)的Wilcoxon符号秩混合指数加权移动平均(WSR-SRA-AEWMA)非参数自适应控制图,有效克服传统控制图在分布假设与响应灵敏性方面不足的问题。通过蒙特卡洛模拟方法评估平均运行长度(average run length,ARL)和运行长度标准差(standard deviation of run length,SDRL)指标,来实现图表参数优化。最后结合聚乙烯(polyethylene,PE)管道生产企业的挤出工序数据,从3个质量参数外径t、壁厚d与不圆度r出发,将WSR-SRA-AEWMA控制图与传统EWMA图及秩和(rank-sum control chart,RSCC)控制图进行对比验证,结果表明该方法相比于传统控制图具有更强的适应性和检测性能。
Abstract
The traditional exponentially weighted moving average (EWMA) control chart is based on the assumption of normality. Its lack of real-time feedback and adaptive ability is viewed as a limitation in practical process quality monitoring. A nonparametric adaptive chart called the WSR-SRA-AEWMA chart was proposed. It is built by combining the Wilcoxon sign-rank method with the sliding rotation algorithm (SRA). The chart was designed to overcome the weaknesses of traditional charts in distributional assumptions and response sensitivity. The average run length (ARL) and the standard deviation of run length (SDRL) were evaluated by Monte Carlo simulation. These indicators were used to optimize the chart parameters. Finally, extrusion-process data from a polyethylene (PE) pipe manufacturer were analyzed. Three quality characteristics were used: outer diameter t, wall thickness d, and out-of-roundness r. The WSR-SRA-AEWMA chart was compared with the traditional EWMA chart and the rank-sum control chart (RSCC). The results demonstrate that the proposed method offers superior adaptability and detection performance compared to traditional control charts.
ZHENG Ye-wen, ZHOU Juan, WANG Qiang, ZHANG Wei-jiao.
Design and Application of WSR-SRA-AEWMA Control Chart[J].
Science Technology and Engineering, 2026, 26(17): 7123-7130 DOI:10.12404/j.issn.1671-1815.2505425
在聚乙烯PE管道生产等现代制造过程中,控制图作为统计过程控制(statistical process control,SPC)中的核心工具[1],用于监测PE管产品质量和过程稳定性。控制图利用设定的上控制限(upper control limit,UCL)和下控制限(lower control limit,LCL)[2],在图形上展示过程随时间的波动,用于判断过程是否处于受控状态。休哈特(Shewhart)控制图[3]、累计和(cumulative sum,CUSUM)控制图[4]与EWMA控制图[5]被视为三大经典控制图,构建了现代制造体系中的质量控制基础网络[6]。EWMA控制图能兼顾当前数据与历史数据,对小幅度偏移有更高的检测灵敏度[7]。故EWMA控制图在工业实践中更广泛应用[8]。然而在PE管实际生产中,挤出工序数据外径t、壁厚d和不圆度r常呈现非正态、非稳态等特性,由于EWMA控制图的自适应能力比较匮乏,经常表现出滞后响应,使得其误报率上升,灵敏度下降[9]。在此背景下,非参数控制图[10]摒弃数据分布假设结合秩、分位数或核密度估计等统计量构建控制限,展现出更强的鲁棒性。周茂袁等[11]整合Wilcoxon秩和检验和Ansari-Bradley检验到变点模型,提出了一个在探测位置和尺度漂移时优于EWMA控制图的非参数控制图。李阳等[12]基于拟合优度检验的非参数控制图能够监控过程中位置参数、尺度参数以及形状参数。此外,为增强EWMA控制图自适应性,Capizzi等[13]引入误差传递函数并提出了自适应指数加权移动平均均值(adaptive exponentially weighted moving average,AEWMA)控制图来监控过程均值的偏移,在AEWMA控制图的研究方面已经取得重大发展。Muhammad等[14]通过SVR模型预测平滑常数提出一种基于支持向量回归(support vector regression,SVR)的AEWMA控制图,Qiao等[15]提出基于中位数运行长度(median run length,MRL)和期望中位数运行长度(expected median run length,EMRL)的AEWMA控制图优化设计方法。AEWMA控制图依赖数据服从正态分布的假设,而实际运用过程中监测数据往往不遵从正态分布,导致其实际运用的准确性较低。
WuChunjie, LiHongwei. Improvement on date quality monitoring methods based on nonparametric control charts[J]. Journal of Applied Statistics and Management, 2024, 43(1): 1-19.
YangLeijun, GuoChengjun, ZhangShunsheng. An integrity monitoring algorithm based on Shewhart control chart[J]. Science Technology and Engineering, 2014, 14(16): 67-71, 77.
XuZhiyong, ZhangZongbing, ChenGang, et al. Application of CUSUM and Shewhart control chart in the quality control of pavement construction[J]. Science Technology and Industry, 2022, 22(6): 355-360.
[9]
HaqA, WoodallW H. A critique of the use of modified and moving average-based EWMA control charts[J]. Quality and Reliability Engineering International, 2023, 39(4): 1269-1276.
JiChunfang, ZhuoXinjian. Chang point and control charts in application of data monitoring[J]. Science Technology and Engineering, 2010, 10(30): 7483-7487.
[12]
LiuZ, WangS, GuX, et al. Application of a novel EWMA-φ chart on quality control in asphalt mixtures production[J]. Construction and Building Materials, 2022, 323: 126264.
[13]
MitraA, LeeK B, ChakrabortiS. An adaptive exponentially weighted moving average-type control chart to monitor the process mean[J]. European Journal of Operational Research, 2019, 279(3): 902-911.
YangShuren, LiuShuqi, FanJiawei, et al. The study of macro-control diagram of the screw pump[J]. Science Technology and Engineering, 2010, 10(25): 6276-6278.
ZhaoYu, LiYanting. Nonparametric control chart of rank test for small samples of high-dimensional date[J]. China Mechanical Engineering, 2022, 33(9): 1104-1114.
ZhouMaoyuan, LiuZhimin, LouXiao. Nonparametric change point model monitoring mean and variance[J]. Acta Scientiarum Naturalium Universitatis Nankaiensis, 2019, 52(4): 81-85.
LiYang, DingDong. Goodness-of-fit test-based nonparametric EWMA control chart[J]. Journal of Systems Science and Mathematical Sciences, 2020, 40(5): 917-926.
[22]
CapizziG, MasarottoG. An adaptive exponentially weighted moving average control chart[J]. Technometrics, 2003, 45(3): 199-207.
[23]
MuhammadW K, MuhammadN. Adaptive EWMA control chart by using support vector regression[J]. Quality and Reliability Engineering International, 2024, 40(7): 3831-3843.
[24]
QiaoY, SunJ, CastagliolaP, et al. Optimal design of one-sided exponential EWMA charts based on median run length and expected median run length[J]. Communications in Statistics-Theory and Methods, 2022, 51(9): 2887-2907.
TangAnan, HuXuelong, XieFupeng, et al. Design of the nonparametric adaptive EWMA SR chart with variable sampling intervals[J]. Operations Research and Management Science, 2024, 33(3): 82-88.
ZhanQingqing, QiXiaoyin, ZhaoXingyu, et al. Adaptive control of lagrange systems with uncertain parameters[J]. Control Engineering of China, 2025, 32(3): 564-570.
AsifF, Noor-ul-AminM, RiazA. Accelerated failure time model based risk adjusted MA-EWMA control chart[J]. Communications in Statistics-Simulation and Computation, 2024, 53(10): 4821-4831.
XuXinyi, ZhuYongzhong, YangYiliu. Non-parametric step cumulative sum control chart based on signed on signed-rank[J]. Computing Technology and Automation, 2025, 44(1): 88-95.
ZengWenxu, KuShaoping, ZhengHao. Strategy of self-adaptive frequency conversion data acquision based on swing door trending algorithm[J]. Application Research of Computers, 2018, 35(3): 769-772.
[35]
RazaM A, TariqF, ZaaganA A, et al. A nonparametric mixed exponentially weighted moving average-moving average control chart with an application to gas turbines[J]. PLoS One, 2024, 19(8): e0307559.