other information
Release date:1985-01-29
Review date:2004-10-14
drafter:Liu Wen, Sun Jing, Ma Yilin, Li Qin, Xiao Hui
Drafting unit:China Standards Research Center, School of Economics and Management, Tsinghua University, Institute of Mathematics and Systems Science, Chinese Academy of Sciences, Institute of Mechanical Science
Focal point unit:National Technical Committee for Standardization of Statistical Methods (CSBTS/TC21)
Proposing unit:China Standards Research Center
Publishing department:General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China
competent authority:National Standardization Administration
Some standard content:
rCS C3. 120- 30
National Standard of the People's Republic of China
GB/T48862002
HIvISO 7873:1993
Control charts for arithmetir average with warning limits2002-D7-18 Issued
General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China
2002-12-01Implementation
CB/T48862002
ISOForeword
Cited Standards
3 Terms and Definitions
Application Conditions
5 Method Description
Identification of Parameter Values in Statistical Control Schemes
7: Selection of Parameter Values in Statistical Control Schemes Appendix (Suggestive Appendix) Conformity and Its Meaning Appendix B of the standard)
Determination of the mean value of the extremely unexpected plastic operation process according to the defective product rate (shown in the time requirement)
Application examples of this standard
Huai Lai suggests the original intention of the six standards)
GB/T4886-2002
This standard is equivalent to the international standard JS gate 7%73.1993 with warning stimulus mean control medical 3 (Cnnlolrhrifurarithmcticaverage with arning units).
The contents of the international standard 1787:1993 are as follows: Table 1, Table 2. Table 3 and Table 4 are also new, and the calculation errors in Table 5 in each table have been corrected. - Added Appendix A, and changed the original Appendix A to Appendix B, and the original Appendix B to Appendix (the original Appendix (. The new Appendix C table adds the standard to the situation to record the average chain length (AR) value L. List, according to 7. (. 1. The new Appendix shows the sampling statistical process control scheme in the heart and is modified to Scheme 1: The overall control limit related to the selection of flow control scheme is re-entered, and the new Appendix C1 World Bank "adjusts. The new double-sided control chart average energy length (ARL> formula is increased by 3. K = 4 when the expression. This standard replaces GB/T48%-1% :. Compared with R/T4886-1985, the main changes in the technical proposal of this standard are: the format of the winter average test length (4KL) table is adjusted, the original tables are integrated into 4 tables, and the calculation results in the original tables are slightly corrected. The expression of the average chain length (AR) formula of the double-case drawing in the standard appendix is re-derived and corrected when the average chain length (AR) formula is 4.
This standard is the correct appendix. Appendix A, Appendix C and Appendix K are informative appendices. This standard is proposed by the China Standards Research Center and issued by the China National Committee for Standardization of Statistical Methods (CSBTS) /TC21) is under the jurisdiction of the Chinese Academy of Standards. The main institutions of this standard are: China Standards Research Center, China University of Economics and Management, China University of Science and Technology, Institute of System Science, Institute of Mechanical Science. This standard mainly involves the following: Liu Wen, Environmental Protection, Ma Pinlin, Hand Break, and Consideration. This standard is the first revision of G/T4G-1 with the average value of the required or limited light production period. G814886-2002. ISo Foreword [S] (International Organization for Standardization) is a worldwide federation composed of national standardization partners (IS member bodies). The work of formulating international standards is carried out by ISU technical After the completion of the technical committee, all member groups have the right to participate in the work of the technical committee if they are interested in the work of the technical committee: International organizations that maintain relations with ISO (non-official organizations) may also participate in related work, and maintain a planned cooperative relationship with the International Committee for Electronic Standardization (EC) in the field of electronic technical standardization. The draft international standards adopted by the technical committees shall be submitted to the technical committees of each or member group in accordance with the requirements of the technical requirements. They shall be reviewed by the member groups with at least 75% or more votes and shall be officially released as international standards. International standard 1S07873 was developed by the Technical Committee of ISO/TC69 for the application of statistical methods. GR/T 4886—2002wwW.bzxz.Net
Retrograde statistical control of a process using a control chart with warning limits is an improvement over tight control (known as the Shewhart control chart). Mean control charts that use both warning limits and action limits are only highly sensitive in revealing process excursions: because the points in the warning area provide additional information, the mean control chart with warning limits shows the most direct excursion of the controlled pressure: in addition, if the sample mean falls outside the action limit, a large excursion from the process level is detected. Compared with the standard control chart, the mean control chart with warning limits is more sensitive when the quality metric is small and the standard deviation of the quality metric is not exceeded. 1 Range
National Standard of the People's Republic of China
Control charts for arithmetic average with warning limits
Control charts for arithmetic average with warning limits
This standard specifies the procedures for calculating sample and value control charts with limit values and action limits to implement statistical control of the process. For production batches of products, the quality is a random variable with a random distribution from the beginning to the end, but the control points are the sample means obtained from samples composed of four or more products. The assumption of normal distribution is not necessary (see 4.2). This standard is applicable to the control of process means under the condition that the process standard is known. 2 Reference standard | |tt||The relevant texts contained in the following standards constitute the provisions of this standard through reference in this standard. When this standard was published, the ones shown are valid: All standards are subject to revision. All parties using this standard should use the following standards to determine the validity of the latest version. GI/3355.11993 Statistical terminology Part 1 Statistical terminology (B/T 3353.2-1013 Statistical terminology Part 2 Statistical quality management terminology 3 Terms and definitions
15! The definitions in 3358.1 and (/3358.2) are applicable to this standard: + The definitions used in the standard are not applicable to the reference in the specification. 4 Application conditions ||tt| |4.1 The statistical analysis of the process to achieve a strong control strategy is the first step to lay the foundation for the statistical analysis of the process - the relationship between the operation and product quality, and provide suggestions for process adjustment. If the statistical analysis shows that the process is out of order and the process capacity does not meet the specified or required requirements, it is necessary to determine the source of the horizontal shift and the force of the process:
4.2 In order to distinguish the use of standard methods, first clarify the following:) The sample mean is also approximately normal distribution. Except for the case of abnormal bands, the central limit theorem is used. The detailed mean obtained from four or more samples is approximately normal distribution, even if the individual measured values are Not satisfied with the state distribution! b! In order to obtain the effect of danger transmission, when using a measuring scale instrument to measure a single value to find the average value, the graduation of the measuring instrument shall not exceed /2;
) The current level of the group average value is defined. If the single value does not shift, the deviation occurs. At this time, it is recommended to adjust the process level:
d: Use the concentration side, only mark the level, and use the center value of the tolerance range of the quality box specified in the document; "level" or heart,
Approved by the General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China on July 18, 2002 2C2.12C1 implementation
GB/T 48B6-2002
Assume that the standard deviation of the quality standard is constant and fully variable. This assumption must be confirmed by the standard profit control chart or range control letter
[) standard single case standard production or production in the following cases: its output is the standard, only the deviation of concern is of interest. When the deviation is considered to be out of control in the deviation of concern, it needs to be corrected: send it to the standard production, or! The caution indicates that the process excursion is as far as possible or such a shift should be found in an extreme way, which is called an extremely unexpected level, and the corresponding research requirements are not met. In the case of the test criteria > and - ", each case is related to the past. When the process is out of control on either side, it is required to reach T;
is determined by the product. And (just) -. Carefully determine the standard value (p, pn) of the typical value of the process under the out-of-control situation
(+--
When it is a band number, when it changes under the influence of the original minister, the process may be disturbed. 5 Method Description
5:1Use mean control limits with monitoring limits to monitor the statistical control of the process. The control chart uses a control chart to represent the level and variation of the process. The sample mean of the current quality is plotted on the control chart, as shown in Figure C1.
5,2The mean control chart with monitoring limits has a monthly line (center), which corresponds to the expected value of the quality of the particular process, that is, the daily standard corresponds to ", and the action limit corresponds to
where is the sample early. The total cost is used to calculate the statistical uniqueness R between the various planning limits, and is the appropriate value of the action limit and warning limit in the control chart. In the 6th chapter, the selection principles of E. and, 5.3 control letters can be placed on the table, displayed on the bulletin board, in the computer, and the industry should display them in other appropriate forms. 5.4 The whole drawing should be as close to the actual as possible. 1. The input and drawing points should be clear. 55 As a method of measuring process variation, a standard operating procedure should be established to include the definition, preparation, application, maintenance and use of control charts: the received documents should be marked on the control chart in a timely manner. 5.6 Single-state statistical control can use the mean control chart with operating limits to track the effectiveness. 5.6.1 When using the efficiency plan method to perform statistical control on the process, use the 5-quality garbage (see Figure 1). The following is the area method, which is between the upper and lower limits of the mean! 6) In the penalty area W+ and W, the sample mean is located between the upper warning limit and the upper front limit or between the lower alternative limit and the lower action limit: in the action area A+ and F, the sample mean is located outside the upper action limit or outside the lower action limit respectively. [a.
-Ra/VA
4-820//n
upper action limit
lower action limit
distance from the statistical control area of 1 double stroke
5.6.2 When the force method of the single criterion is used to control the overshoot, the three quality areas are reduced (see Figures 2 and 3). Describe as follows: 2
E4886—2002
a) and the standard wind slope T: the sample mean is below the upper warning limit or above the lower warning limit! b) Good area, sample and value are located between the action limit and the action limit) Action area A sample mean is located outside the action limit A+
Line limit
Pu Cheng Yin
Center Entertainment
Figure 2 Statistical control quality area of single case criterion -…upper limit-Ba/Vg
a-tfiei Y
Center double
Pu I resistance
Action use
Figure: Statistical control quality area of single case criterion lower limit
Figure 2 shows the situation when the value of the quality metric of interest increases, and Figure 3 shows the situation when the mean value of the quality metric of interest decreases. 5.! The sample mean of the quality metric is plotted on the control side diagram with integral limit. For each sample point on the control chart, select the code (which can be expressed in order of speed, time, etc.), and the frequency coordinates are the corresponding sample averages (Figure C1) 6 Statistical Control of the Process
6.1 When a sample point is in the upper action zone (A) or the lower action zone (A), it indicates that the process is out of control. When this happens, it is necessary to determine the cause of the out-of-control situation and correct it so that the process accuracy is at an appropriate level. 6.2 When consecutive points appear in the same warning zone (i.e., in the upper warning zone W, or in the lower warning zone W_), it is also a signal that the process is out of control and needs to be adjusted:
According to the sequence given in the first table, select the values of each parameter. 7 Selection of numerical values in process control plans 7.1 When selecting a statistical process control plan, the following values must be determined: 1) Sample ( )
1) Sampling interval x (see 7.3):
2) Number of consecutive points K (see S.2);
3) Determine the rate R, and R, of the action and warning limits in the control chart (see 7.2.2 and 7.4.12 e) Corrective process decision making principles.
The following are the options for selecting a statistical process control plan: or Chapter
1. and are the control process and failure level and chain length AR respectively (see 7.2). 1: + and are the requirements of the company:
GB/T4886-2002
7.2 The efficiency of the statistical process control plan is expressed in terms of average chain length. 7.2.1 When the inner number is too strong: Process three uses the chain length (ARI.) which is half the mean of the sample mean points plotted before the out-of-control signal is issued. When the process level is at half the monthly standard (the ARL returns to the maximum value as it deviates from the daily standard), the ARL will drop. When the process average is at the daily standard level (A), the design of the control chart should provide a large average chain length of 1., that is, provide a low false equalization rate. Similarly, when the process average is at or above , the design of the control chart should provide a small average chain length +, that is, provide a rapid detection of process dissatisfaction.
7.2.2 Single criterion control charts use Tables 1 to 4. Each table gives the value of T. The value of each row in each table corresponds to the value of each row that is not zero. For multiple combinations of R, R, and BF that cannot be maintained, try to select a group of values that achieve the lowest ratio. 2.3 Double control charts use Tables 1 to 4. The ARL of the type controlled process is then in Table 1. / indicates the value of the row = 0 to determine the ARL of the out-of-regulation process, that is, L, which is determined by the values of the rows of /1 in Table 4 and the rows of 1 in Tables 1 to 3. Because the ART. of the double-sided criterion of the control process is consistent with the ARI. of the single-sided criterion in terms of numbers (see Table D1>72.4 Clothes 1~Table 1 does not have the corresponding value of the element value out of the specification. The linear integration of the AR curve is guaranteed by the sample number and the influence of the number of samples, 5 and blood. The formula in [ is shown. In addition, for the total number of measured values of the Dingshu network, the control distance can be designed to have a relatively large sample time and a relatively small plate. In the application of the industry, the chain design should be considered in the control medical design process. The consulting test period is determined: and the straight line is taken. It is recommended that you evaluate the design during the elapsed time related to the and value. In most cases, the tea individual The existing sampling scheme is the basic "test acid combination. Other test designs are recommended to compare the performance and alternatives of the basic design.
7.4 Use Tables 1-4 to select statistical control schemes. 7.4.1 If the values of and and 7 and (and other restrictions) have been determined in advance, in Tables 1-4, find the row with the specified value (if necessary, perform linear internalization) and find the columns that meet the specified conditions (in most cases 2). In the various information of the drug, or the values of the provisional conditions (in most cases there are multiple), determine the scheme from the following: : and length according to the provisions of?.2.1. If there are two process control schemes that meet the specified requirements, it is advisable to select the scheme with the larger ratio of 1 and : (see Figure 2). In this case, if the ratio is smaller (less than 40), it is recommended to select the smaller one as the scheme.
7. 4.2 If the sample size is not determined in advance, the test value can be obtained by using Tables 1 to 4. Select the columns in Tables 1 to 4 that meet the conditions in 1. Then find the first number of the given coefficients that satisfies the L value, and then find the corresponding / value. D is calculated from the given 3. If the sample size is not an integer, the result will be rounded to the nearest integer. In some cases, multiple statistical control plans may be required; the choice of the plan that improves the cost (see 7.2 and 7.4.) is necessary. This is especially important when the improvement in control is expensive. 7.5 Under the main production technology selection conditions, various changes may occur, such as the skills of the operator, the supply of materials, the change or widening of the action limit due to some technical inefficiencies, etc. It is recommended that these changes be considered in the implementation of process statistical control. The document should indicate that the control chart is updated and that a potential analysis is performed within a specified time (one inch, one year, etc.). The frequency of such analysis should be determined according to production needs. The single-sided criterion 2
=X product ARL
5-1, 4
345,?
34.1 e1.9
GD/T 4B8G- 2D02
3.2 ARL (single-sided criterion
3. 25 and A121.
12 small, 6
179.n278.c
ARL value (double measurement standard)
, ARI.dangerous
65.2\1C2.
39.91742 8
116.71136.2
367.R257.9
311. X 13.
65. 7 3G9. 8572.
*01.4 307.4 3c3. 1
190. 7 195. 3 193. H
1+7.*:1:4. 3
XR*.935.1
696.n 7.1. 2
Average chain length of the process (AKL)
GB/T 4986—2002
Appendix A
(Suggestive Appendix)
Symbols and quantity text
Average length of controlled over-control (double stroke only used in the case of over-control). Indicates the average length of the sample value: the current process level te
Process mean level
Under the one-sided criterion. The process mean is "extremely unexpected\seeking a flat,
Under the criterion, the "unexpected" level of the over-action value-the standard deviation of the quality single
Sample craftsman
Falls in the control area, and is judged as a failure. The number of consecutive points K
B, determines the position of the action limit in the control chart. The system effect of determining the control resistance and the warning limit should be set in the process deviation or the standard, the standardized form of the process mean in the case of out of control: when the supply and value are above, 3- When the two wires
are confirmed with the value below, 3=4-day
standard area, the sample mean is in the upper warning area or the lower warning area T
, the sample mean is between the upper warning area and the upper action limit, the sample mean is between the lower warning area and the lower action limit A+
upper action area, the sample mean is outside the upper action limit, the sample mean is outside the lower action limit
Tip: This standard content only shows part of the intercepted content of the complete standard. If you need the complete standard, please go to the top to download the complete standard document for free.