GB/T 17989.6-2022 Active National standards

GB/T 17989.6-2022 Statistical method of quality control in production process—Control charts—Part 6: EWMA control charts

Publish Date: 2022-03-09 Implement Date: 2022-10-01 For services related to genuine standard inquiry, procurement, translation, and other related services in China, please Contact Us

Basic Information

Standard Code: GB/T 17989.6-2022

Standard Type: National standards

Standard Status: Active

is_force_gb: no

CCS Name: Mathematics

ICS Name: The application of statistical methods

Publish Date: 2022-03-09

Implement Date: 2022-10-01

Pages: 30 pages

Scope

This document provides guidance on understanding and applying the Exponentially Weighted Moving Average (EWMA) control chart for statistical process control. The EWMA control chart is a statistical process control technique used to monitor minor fluctuations in the process mean, enabling timely detection of minor and moderate fluctuations in the process mean. The EWMA control chart evaluates the process mean by calculating the exponential weighted moving average of the sample mean of all historical data. The EWMA weights the samples in a geometrically decreasing manner from closest to farthest from the current position, with samples closer to the current position having greater weight and impact, and samples farther away having lesser weight and impact. The specific weight depends on the smoothing parameter (λ).
This document is applicable to situations where the control chart is used to monitor minor fluctuations in the process mean.
Note 1: The main objective of the EWMA chart is the same as that of the conventional control chart. The relevant content of the conventional control chart is described in GB/T 17989.2. The following situations are suitable for using the conventional control chart:
——Low productivity;
——The process of sampling and inspection is complex and time-consuming;
——The cost of inspection is too high;
——There are safety hazards.
Note 2: Quality workers can use a single observation value obtained from the production line to create a single-value control chart, without the need to use a sample of multiple observation values to create an average control chart. This option is necessary when the cost of inspecting a sample of multiple observation values is too high or difficult to achieve. For example: The number of customer complaints or product returns is often statistically analyzed on a monthly basis, and quality managers need to use them to create charts to reflect quality issues. This type of control ch