CN101782763B - Method for monitoring statistical process control - Google Patents
Method for monitoring statistical process control Download PDFInfo
- Publication number
- CN101782763B CN101782763B CN200910045976XA CN200910045976A CN101782763B CN 101782763 B CN101782763 B CN 101782763B CN 200910045976X A CN200910045976X A CN 200910045976XA CN 200910045976 A CN200910045976 A CN 200910045976A CN 101782763 B CN101782763 B CN 101782763B
- Authority
- CN
- China
- Prior art keywords
- process control
- statistical process
- alarm
- sample data
- steady state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- General Factory Administration (AREA)
Abstract
The invention discloses a method for monitoring statistical process control. The correlation detection of sample data and a false alarm rate estimated value relative to the statistical process control are used as standards to judge whether the statistical process control is in a false stable state or a stable state; and by classifying alarm information acquired by the statistical process control, the unstable states of the statistical process control are classified into different types, so the accuracy of the statistical process control is improved, the calculated amount is reduced, manpower and energy are saved and the product efficiency is improved.
Description
Technical field
The present invention relates to statistical process control technology, the method for particularly statistical Process Control being monitored.
Background technology
In the production division of enterprise, all will produce or process large-tonnage product every day, and performance of products and quality are related to the lifeblood of enterprise, product quality monitored with check and analysis timely seem particularly important.Statistical Process Control (SPC) is exactly a kind of by means of statistics, the instrument that production run is added up and controlled.In process of production; The statistical Process Control instrument is widely used in the parameter in the production run is carried out Collection and analysis, and the data that will pass through analysis offer the slip-stick artist with the form of statistical signal, such as control circle to sample data calculation control figure collected in the production run; Make control chart, histogram etc.; Perhaps use control circle in the control chart, the fluctuation situation of control chart mid point in the observation production run is with monitoring industrial processes; And help technology and producers to take appropriate measures, realize the Continual Improvement of production run.
In general, when warning information appears in statistical Process Control, need to consider to take to comprise the various measures of inspection of stopping production, ascertain the reason and with its eliminating, to return to normal production.Yet this only is in the stable situation of statistical Process Control, in production reality, has the unsettled situation of statistical Process Control, that is to say, because statistical Process Control self, said warning information can not reflect the real condition of production.Because production run or sampling process or statistical Process Control all possibly cause warning information, if can't confirm whether statistical Process Control is stable, directly any possibility factor that causes warning information investigated, and will spend plenty of time and energy.
Freescale company once proposed a kind of method that is used to assess statistical Process Control stability, and promptly with all incidents that produces warning information, with the ratio of whole incident quantity, index is monitored statistical Process Control as a reference.Application to this reference index mainly comprises: through this reference index and False Alarm Rate are compared, grasp the effective degree of statistical Process Control, thereby realize monitoring of production flow.For example: when this reference index during much smaller than False Alarm Rate, think that then the control circle is provided with too loosely, statistical Process Control is too stable; When this reference index during, explain that statistical Process Control plays pendulum much larger than False Alarm Rate.But in fact, the effect in practical application, played of this method is very limited.For example, in the method, adopted approximate be equal to when controlling the boundary and be 3 σ the False Alarm Rate that possibly occur, promptly the estimated in theory value of False Alarm Rate is a standard, compare with this reference index.3 σ are the estimated in theory value on control circle; The control of being adopted in actual production circle is often little than 3 σ; Therefore the actual False Alarm Rate that obtains the also estimated in theory value of more said False Alarm Rate is little; And since only with the estimated in theory value of False Alarm Rate as comparative standard, in practical application, this reference index is much smaller than the situation of this False Alarm Rate estimated in theory value seldom.In addition, this method can only show whether statistical Process Control is in unsteady state, statistical Process Control is not in unsteady state and provides further information, also just can not play any help to the adjustment and the control of follow-up statistical Process Control.
Summary of the invention
The problem that the present invention solves provides a kind of method for supervising, statistical Process Control is effectively monitored, and calculating is simple, spends manpower, energy is less.
For addressing the above problem, the invention provides a kind of method for supervising of statistical Process Control, comprising: gather sample data,, obtain warning information through said sample data is carried out statistical Process Control; According to the correlativity of warning information and sample data, judge whether said statistical Process Control is in apparent steady state: when said statistical Process Control is in apparent steady state, adjust the control circle scope of said statistical Process Control, finish this monitoring; When said statistical Process Control is not in apparent steady state,, judge whether said statistical Process Control is stable: when said statistical Process Control is in steady state (SS), finish this monitoring according to the estimated value of said warning information and sample data false alarm; When said statistical Process Control plays pendulum, the warning information that is obtained is classified; Estimated value according to said warning information, said classification results and said False Alarm Rate; The non-steady state of said statistical Process Control is divided into the first category and second classification: when the non-steady state of said statistical Process Control is first category; According to said classification results; Production and sampling process are investigated, adjusted the control circle scope of said statistical Process Control, finish this monitoring; When the non-steady state of said statistical Process Control is second classification, finish this monitoring.
Optional, said sample data is carried out correlation detection, obtain the correlation ratio of said sample data; Obtain all alarm rates, said all alarm rates account for the ratio of all sample data quantity for the corresponding sample data quantity of all alarms; The correlation ratio of more said all alarm rates and said sample data: when said all alarm rates during much smaller than the correlation ratio of said sample data; Said statistical Process Control is in apparent steady state; When said all alarm rates during, then be not in apparent steady state greater than the correlation ratio of said sample data.
Optional, the said warning information that is obtained is classified comprises: with the warning information that is obtained be divided into false alarm, can the trace back true alarm and the true alarm of can not tracing back.
Optional, the said true alarm of tracing back comprises: the said true alarm of having found its root and having solved; The said true alarm of can not tracing back comprises: do not find root as yet, or found the but still unsolved said true alarm in source.
Optional, said estimated value according to warning information, classification results and False Alarm Rate is divided into first category and second classification comprises with the non-steady state of statistical Process Control: the acquisition true alarm rate that can trace back; Obtain all alarm rates and the difference of the true alarm rate that can trace back; The estimated value of this difference and said False Alarm Rate relatively: when this difference during much larger than the estimated value of said False Alarm Rate, the non-steady state of statistical Process Control is a first category; When this difference approached said False Alarm Rate estimated value, the non-steady state of this statistical Process Control was second classification.
Optional, the said true alarm rate that traces back is the ratio that the corresponding sample data quantity of the said true alarm of tracing back accounts for all sample data quantity; The said true alarm rate that can not trace back is the ratio that the corresponding sample data quantity of the said true alarm of can not tracing back accounts for all sample data quantity.
Optional, the acquisition true alarm rate that can trace back comprises: obtain the to trace back quantity of true alarm of the experiment through limited number of time.
Optional, the non-steady state of said statistical Process Control is a first category, comprising: its true alarm rate that can not trace back is more.
Optional, the non-steady state of said statistical Process Control is a first category, comprising: in said all warning informations, true alarm is mainly the true alarm of can tracing back.
Optional, said estimated value according to warning information and sample data false alarm is judged statistical Process Control stable comprising whether: the estimated in theory value of calculating the sample data false alarm; The estimated in theory value of more said all alarm rates and said false alarm: when said all alarm rates during much larger than the estimated in theory value of said false alarm, said statistical Process Control plays pendulum; When said all alarm rates approached the estimated in theory value of said false alarm, said statistical Process Control was in steady state (SS).
Compared with prior art; The present invention has the following advantages: respectively with the correlation detection of sample data with respect to the False Alarm Rate estimated value to this statistical Process Control; As the standard of judging whether this statistical Process Control is in apparent steady state and whether is in steady state (SS), and through the warning information that this statistical Process Control obtained is classified, be divided into the non-steady state of statistical Process Control different classes of; Thereby improve the accuracy of statistical Process Control; Reduce calculated amount, save manpower and energy, enhance productivity.
Description of drawings
Fig. 1 is the schematic flow sheet of the method for supervising embodiment of statistical Process Control of the present invention;
Fig. 2 is the schematic flow sheet of step S2 embodiment among Fig. 1;
Fig. 3 is the schematic flow sheet of step S3 embodiment among Fig. 1;
Fig. 4 is the schematic flow sheet of step S5 embodiment among Fig. 1.
Embodiment
The inventor is through long-time observation to production run; Through the correlation detection of all warning informations and sample data, judge whether statistical Process Control is in too stable status, and pass through with respect to the False Alarm Rate estimated value of said statistical Process Control and the comparison of said all warning informations; Judge whether statistical Process Control is in steady state (SS); And when statistical Process Control plays pendulum, said all warning informations are classified, according to dissimilar warning informations; Non-steady state to statistical Process Control is classified; Be implemented under the different situations statistical Process Control is carried out different adjustment, thereby effectively statistical Process Control is monitored, and then improve control production run.
Below in conjunction with embodiment and accompanying drawing, embodiment of the present invention is done further to specify.
With reference to figure 1, the invention provides a kind of method for supervising of statistical Process Control, comprising: step S1, gather sample data, through said sample data is carried out statistical Process Control, obtain warning information; Step S2; According to the correlativity of warning information and sample data, it is too stable to judge whether said statistical Process Control is in: when said statistical Process Control is in apparent steady state, get into step S7; When said statistical Process Control is not in apparent steady state, get into step S3; Step S3 according to the estimated value of said warning information and sample data false alarm, judges whether said statistical Process Control is stable; When said statistical Process Control is in steady state (SS); Get into step S8, when said statistical Process Control plays pendulum, get into step S4; Step S4 classifies the warning information that is obtained, and gets into step S5; Step S5; Estimated value according to said warning information, said classification results and said False Alarm Rate; The non-steady state of said statistical Process Control is divided into the first category and second classification, when the non-steady state of said statistical Process Control is first category, gets into step S6; When the non-steady state of said statistical Process Control is second classification, get into step S8; Step S6 according to the warning information of said classification, investigates production and sampling process, gets into step S8; Step S7 adjusts the control circle scope of said statistical Process Control, gets into step S8; Step S8 finishes this monitoring.
Embodiment of the present invention can comprise the statistical Process Control enforcement monitoring that semiconductor production process is carried out.Specifically, sample data described in the step S1 can be the parameter of various processing procedures in the semiconductor production process, for example, for etching process, can gather etching depth, it is carried out statistical Process Control, thereby etching process is monitored.Can at interval identical acquisition time to said sample data collection, concrete acquisition mode can comprise multiple mode according to the heterogeneity of collection sample data.
In statistics; Suppose that production run only receives stochastic factor; So, the qualitative character of product is a stochastic variable that obey to confirm probability distribution, and its distribution can be through sample data under the steady state (SS) is gathered and statistical computation obtains to being in a period of time.When statistical parameters such as the mean value of product quality characteristic and variance all during kept stable, claim that production run is in controllable state.
Because sample data character has nothing in common with each other, therefore can adopt the different statistic process control to different sample datas, the concrete statistical Process Control instrument that is adopted does not cause restriction to the enforcement of step S1.
With the control chart is example, and control chart is a kind of statistical Process Control instrument commonly used, is used to monitor production run and whether is in state of a control.Every at a distance from the set time; On production line, gather the sample data of fixed number; It is calculated; Result of calculation is represented with the form of control chart whether exceed the scope on control circle in the control chart through statistical parameters such as sample data mean value and variances, and then judge whether this production run is in controllable state.When production run is out of hand, for example, when existence in the control chart exceeds the sample data of controlling the boundary, will send a warning; Each exceed control the boundary sample data corresponding to a warning information.
After obtaining warning information, get into step S2.Wherein, with reference to figure 2, step S2 can further comprise: step S11, said sample data is carried out correlation detection, and obtain the correlation ratio of said sample data; Step S12 obtains all alarm rates; Step S13; The correlation ratio of more said all alarm rates and said sample data: when said all alarm rates during much smaller than the correlation ratio of said sample data; Said statistical Process Control is in apparent steady state; When said all alarm rates during, then be not in apparent steady state greater than the correlation ratio of said sample data.Wherein, said all alarm rates can be the ratio that the corresponding sample data quantity of all alarms that in the control circle is detected, obtain accounts for all sample data quantity.
Among the step S11, said correlation detection comprises whether being relative to each other between the inhomogeneity sample data that detection gathers and whether same type of sample data self exists autocorrelation.
Statistical Process Control is to meet certain statistics hypothesis based on the sample data of being gathered, and for example, said sample data does not have on the basis such as correlativity, the data processing that said sample data is carried out.When having correlativity between the said sample data, the warning information of said statistical Process Control can not detect it, need carry out anticipating of decorrelation to said sample data after, again the data after handling are carried out statistical Process Control.
Wherein, said correlation ratio reflected is not detected by set control circle but know that in follow-up product testing process or according to manufacturing technique requirent its in-problem these products account for the ratio of all products sums.For example; In etching process, problem causes product undesirable because too wide or the too narrow or film of etching live width is too thick etc., and such product is undetected in the control circle is detected; But not through follow-up detections such as product quality; Add up the quantity of such product, calculate the ratio of itself and all products sums, can obtain said correlation ratio.
Among the step S13, the correlation ratio of more said all alarm rates and said sample data: when said all alarm rates during much smaller than the correlation ratio of said sample data, said statistical Process Control is in apparent steady state; When said all alarm rates during greater than the correlation ratio of said sample data, then said statistical Process Control is not in apparent steady state.Because said correlation ratio has reflected the substandard product quantity that detects through the control circle and has accounted for the ratio of all product quantities; When said correlation ratio is very big; Mean that a large amount of substandard products are not detected in the control circle is detected; Therefore set control circle is too wide in range, and said statistical Process Control is in apparent steady state.
When said statistical Process Control is in apparent steady state, with reference to figure 1, execution in step S7.Said apparent steady state that is to say that set control circle scope is too wide in range, thereby can't report to the police to the error information that is occurred.Therefore, corresponding, the adjustment of said statistical Process Control is comprised: adjust the control circle scope of said statistical Process Control, it is tightened up, make said statistical Process Control to report to the police the error that is occurred in the sample data.
When said statistical Process Control is not in apparent steady state, judge whether said statistical Process Control is in steady state (SS), i.e. step S3.Specifically, with reference to figure 3, can comprise: step S21 calculates the estimated value of sample data false alarm; Step S22, the estimated value of more said all alarm rates and said false alarm: when said all alarm rates during much larger than the estimated value of said false alarm, said statistical Process Control plays pendulum; When said all alarm rates approached the estimated value of said false alarm, said statistical Process Control was in steady state (SS).
Wherein, owing to there is random disturbance, some is a false alarm in the warning information that is obtained.Said false alarm certainly exists, existing problem and producing in the production run no thanks to, and its size has reflected the stability of statistical Process Control.For specific statistical Process Control, the corresponding sample data quantity of its false alarm accounts for the ratio of all sample data quantity, i.e. False Alarm Rate.Because random disturbance exists uncertain, and False Alarm Rate possibly receive other factors and disturb, and the actual value of the False Alarm Rate of therefore concrete statistical Process Control has no way of counting.But, because false alarm is relevant with the character of the sample data of statistical Process Control that is adopted and collection, according to the statistical Process Control that is adopted; Specifically; According to the size of control circle that it adopted, can estimate False Alarm Rate, obtain the estimated value of False Alarm Rate.The estimated value of said sample data false alarm, specifically, can be according to the acquisition of tabling look-up of control circle size.
When said statistical Process Control plays pendulum, get into step S4.Specifically, can with the warning information that is obtained be divided into three kinds dissimilar: first kind warning information is the true alarm of can tracing back; Second type of warning information is for tracing back true alarm; The 3rd type of warning information is false alarm.
Wherein, obtain in the warning information, except false alarm, all the other are true alarm.Existing abnormal data has been indicated in these true alarms, has reflected the problem that occurs or exist in the production run.These true alarms are further classified, specifically can comprise: according to whether having found out its root and solved, with said true alarm be divided into can trace back true alarm with can not trace back true alarm.
For those said true alarms of having found its root and having solved, after the problem that causes these alarms was solved, the true alarm that is originally brought by these problems just can not occur again.That is to say that through the experiment repeatedly of limited number of times or rule of thumb found out its root, and after its root was solved, this alarm no longer occurred.The just above-mentioned first kind warning information of true alarm like this is the true alarm of can tracing back.
And do not find root as yet, perhaps found the still unsolved true alarm in source for those, owing to get rid of the root problem that causes alarm as yet, therefore, this type of true alarm still can occur repeatedly.With the true alarm of can not tracing back of being of this type of unresolved its root, its corresponding sample data quantity accounts for the ratio of all sample data quantity, is the true alarm rate that can not trace back.Owing to can't confirm to cause the root of true alarm of can not tracing back, thereby the quantity of this type of the true alarm true alarm rate that even can not trace back, also just have no way of counting.
In step S5,, specifically can comprise: step S31, the acquisition true alarm rate that can trace back with reference to figure 4; Step S32 obtains all alarm rates and the difference of the true alarm rate that can trace back; Step S33, the relatively estimated value of this difference and said False Alarm Rate: when this difference during much larger than the estimated value of said False Alarm Rate, the non-steady state of statistical Process Control is a first category; When this difference approached said False Alarm Rate estimated value, the non-steady state of this statistical Process Control was second classification.
Specifically, owing to the true alarm of can tracing back can be got rid of through the experiment or the experience of limited number of time, thereby can add up the true alarm of can tracing back, its corresponding sample data quantity accounts for the ratio of all sample data quantity, is the true alarm rate that can trace back.All alarm rates are the False Alarm Rate actual value and can not trace back true alarm rate sum with the difference of the true alarm rate that can trace back.
All alarm rates and can the trace back difference of true alarm rate and the estimated value of False Alarm Rate are compared.Because the estimated value of False Alarm Rate is the upper limit of its actual value; That is to say; When this difference during much larger than the estimated value of said False Alarm Rate; For example greater than the twice of said False Alarm Rate estimated value or more than, the value of the true alarm rate that can not trace back that this difference is included is bigger, then in this case the non-steady state of said statistical Process Control is a first category.When this difference approached said False Alarm Rate estimated value, then in this case the non-steady state of said statistical Process Control was second classification.
Specifically, when said statistical Process Control plays pendulum, and non-steady state is when being first category; That is to say that the true alarm rate that can not trace back is more, because the root of the true alarm of can not tracing back is not investigated clear as yet; This moment is to the adjustment of statistical Process Control, with reference to figure 2, i.e. and step S6; Can comprise: production run is investigated, and the process of gathering sample data is investigated.
When said statistical Process Control plays pendulum; And when non-steady state is second classification because the difference of all alarm rates and the true alarm rate that can trace back approaches or even less than the estimated value of False Alarm Rate, that is to say; In said all warning informations; True alarm is mainly the true alarm of can tracing back, and through to causing the investigation of the true alarm root of can tracing back, can get rid of true alarm basically; Said difference is approached or even less than the estimated value of False Alarm Rate, thereby make said statistical Process Control get into steady state (SS).
Other embodiment of the present invention can be applicable to the manufacturing process that other adopts statistical Process Control, and said sample data is not limited to the parameter of semiconductor production manufacture process.
Above-mentioned embodiment provides a kind of method for supervising of statistical Process Control, and it compared with prior art has the following advantages:
Utilize the essence of statistical Process Control; Through introducing the correlation ratio that correlation detection obtained,, whether said statistical Process Control is in apparent steady state judges as criterion by sample data; And through calculating the False Alarm Rate estimated value of actual count process control; As criterion, whether said statistical Process Control is in steady state (SS) judges, thereby make judged result more meet practical condition; And then raising result's accuracy, enhance productivity.
When said statistical Process Control is in steady state (SS) following time; Warning information to being obtained is classified; Through said all alarm rates and the difference of the said true alarm rate that traces back and comparing once more of False Alarm Rate estimated value; Reduce the time that warning information is repeatedly investigated repeatedly, saved a large amount of manpowers and energy.
Though the present invention through the preferred embodiment explanation as above, these preferred embodiments are not in order to limit the present invention.Those skilled in the art is not breaking away from the spirit and scope of the present invention, should have the ability various corrections and additional are made in this preferred embodiment, and therefore, protection scope of the present invention is as the criterion with the scope of claims.
Claims (10)
1. the method for supervising of a statistical Process Control is characterized in that, comprising:
Gather sample data,, obtain warning information through said sample data is carried out statistical Process Control;
According to the correlativity of warning information and sample data, judge whether said statistical Process Control is in apparent steady state:
When said statistical Process Control is in apparent steady state, adjust the control circle scope of said statistical Process Control, finish this monitoring;
When said statistical Process Control is not in apparent steady state,, judge whether said statistical Process Control is stable: when said statistical Process Control is in steady state (SS), finish this monitoring according to the estimated value of said warning information and sample data false alarm;
When said statistical Process Control plays pendulum, the warning information that is obtained is classified; Estimated value according to said warning information, said classification results and False Alarm Rate; The non-steady state of said statistical Process Control is divided into the first category and second classification: when the non-steady state of said statistical Process Control is first category; According to said classification results; Production and sampling process are investigated, finished this monitoring;
When the non-steady state of said statistical Process Control is second classification, finish this monitoring.
2. method for supervising as claimed in claim 1 is characterized in that, said correlativity according to warning information and sample data judges whether statistical Process Control is in apparent steady state and comprises:
Said sample data is carried out correlation detection, obtain the correlation ratio of said sample data;
Obtain all alarm rates, said all alarm rates account for the ratio of all sample data quantity for the corresponding sample data quantity of all alarms;
The correlation ratio of more said all alarm rates and said sample data: when said all alarm rates during much smaller than the correlation ratio of said sample data; Said statistical Process Control is in apparent steady state; When said all alarm rates during, then be not in apparent steady state greater than the correlation ratio of said sample data.
3. method for supervising as claimed in claim 1 is characterized in that, the said warning information that is obtained is classified comprises: with the warning information that is obtained be divided into false alarm, can the trace back true alarm and the true alarm of can not tracing back.
4. method for supervising as claimed in claim 3 is characterized in that, the said true alarm of tracing back comprises: the said true alarm of having found its root and having solved; The said true alarm of can not tracing back comprises: do not find root as yet, or found the but still unsolved said true alarm in source.
5. method for supervising as claimed in claim 1 is characterized in that, said estimated value according to warning information, classification results and False Alarm Rate is divided into first category and second classification comprises with the non-steady state of statistical Process Control:
The acquisition true alarm rate that can trace back;
Obtain all alarm rates and the difference of the true alarm rate that can trace back;
The estimated value of this difference and said False Alarm Rate relatively: when this difference during much larger than the estimated value of said False Alarm Rate, the non-steady state of statistical Process Control is a first category; When this difference approached said False Alarm Rate estimated value, the non-steady state of this statistical Process Control was second classification.
6. method for supervising as claimed in claim 5 is characterized in that, the said true alarm rate that traces back is the ratio that the corresponding sample data quantity of the said true alarm of tracing back accounts for all sample data quantity; The said true alarm rate that can not trace back is the ratio that the corresponding sample data quantity of the said true alarm of can not tracing back accounts for all sample data quantity.
7. method for supervising as claimed in claim 5 is characterized in that, the acquisition true alarm rate that can trace back comprises: obtain the to trace back quantity of true alarm of the experiment through limited number of time.
8. method for supervising as claimed in claim 5 is characterized in that, the non-steady state of said statistical Process Control is a first category, comprising: its true alarm rate that can not trace back is more.
9. method for supervising as claimed in claim 5 is characterized in that, the non-steady state of said statistical Process Control is second classification, comprising: in said all warning informations, true alarm is mainly the true alarm of can tracing back.
10. method for supervising as claimed in claim 2 is characterized in that, said estimated value according to warning information and sample data false alarm is judged that statistical Process Control is whether stable to comprise:
Calculate the estimated value of sample data false alarm;
The estimated value of more said all alarm rates and said false alarm: when said all alarm rates during much larger than the estimated value of said false alarm, said statistical Process Control plays pendulum; When said all alarm rates approached the estimated value of said false alarm, said statistical Process Control was in steady state (SS).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910045976XA CN101782763B (en) | 2009-01-19 | 2009-01-19 | Method for monitoring statistical process control |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910045976XA CN101782763B (en) | 2009-01-19 | 2009-01-19 | Method for monitoring statistical process control |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101782763A CN101782763A (en) | 2010-07-21 |
| CN101782763B true CN101782763B (en) | 2012-01-25 |
Family
ID=42522798
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910045976XA Expired - Fee Related CN101782763B (en) | 2009-01-19 | 2009-01-19 | Method for monitoring statistical process control |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101782763B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103246228A (en) * | 2012-02-14 | 2013-08-14 | 厦门烟草工业有限责任公司 | System for monitoring mean value and standard deviation in on-line way |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102354116B (en) * | 2011-08-05 | 2014-06-25 | 北京航空航天大学 | Method for making omega event interval control chart for high quality process statistics control |
| CN103300467B (en) * | 2012-09-11 | 2015-12-02 | 张家口卷烟厂有限责任公司 | A kind of sheet cigarette formula vertical library leaf group performs formula managing and control system and method thereof |
| CN103412542B (en) * | 2013-08-14 | 2015-11-04 | 沈阳中科博微自动化技术有限公司 | A kind of integrated circuit technology unit exception early warning technology method of data-driven |
| US11062802B1 (en) * | 2015-06-04 | 2021-07-13 | Cerner Innovation, Inc. | Medical resource forecasting |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1867876A (en) * | 2003-10-16 | 2006-11-22 | Abb公司 | Detecting faults of system components in a continuous process |
| US7337034B1 (en) * | 2005-09-07 | 2008-02-26 | Advanced Micro Devices, Inc. | Method and apparatus for determining a root cause of a statistical process control failure |
| TW200815771A (en) * | 2006-09-29 | 2008-04-01 | Powerchip Semiconductor Corp | Method of estimating bound limit |
| CN101174149A (en) * | 2006-11-03 | 2008-05-07 | 力晶半导体股份有限公司 | Establishment method of control specification limit |
| CN101290517A (en) * | 2007-04-17 | 2008-10-22 | 中芯国际集成电路制造(上海)有限公司 | Method and device for statistical process control for discrete samples data |
-
2009
- 2009-01-19 CN CN200910045976XA patent/CN101782763B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1867876A (en) * | 2003-10-16 | 2006-11-22 | Abb公司 | Detecting faults of system components in a continuous process |
| US7337034B1 (en) * | 2005-09-07 | 2008-02-26 | Advanced Micro Devices, Inc. | Method and apparatus for determining a root cause of a statistical process control failure |
| TW200815771A (en) * | 2006-09-29 | 2008-04-01 | Powerchip Semiconductor Corp | Method of estimating bound limit |
| CN101174149A (en) * | 2006-11-03 | 2008-05-07 | 力晶半导体股份有限公司 | Establishment method of control specification limit |
| CN101290517A (en) * | 2007-04-17 | 2008-10-22 | 中芯国际集成电路制造(上海)有限公司 | Method and device for statistical process control for discrete samples data |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103246228A (en) * | 2012-02-14 | 2013-08-14 | 厦门烟草工业有限责任公司 | System for monitoring mean value and standard deviation in on-line way |
| CN103246228B (en) * | 2012-02-14 | 2016-01-20 | 厦门烟草工业有限责任公司 | The system of a kind of on-line monitoring average and standard deviation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101782763A (en) | 2010-07-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101782763B (en) | Method for monitoring statistical process control | |
| CN105095747B (en) | A kind of Java application health degree appraisal procedure and system | |
| CN112348237B (en) | Abnormal trend detection method for dynamic drilling data | |
| CN113518011B (en) | Abnormality detection method and apparatus, electronic device, and computer-readable storage medium | |
| CN105718715A (en) | Anomaly detection method and device | |
| EP2466495A1 (en) | Systems and/or methods for event stream deviation detection | |
| CN109741927B (en) | Intelligent prediction system for equipment faults and potential defective products of miniature transformer production line | |
| CN110601900A (en) | Network fault early warning method and device | |
| CN111865407A (en) | Intelligent early warning method, device, equipment and storage medium for optical channel performance degradation | |
| CN107276779B (en) | Monitoring method, system and equipment | |
| CN116523508B (en) | Intelligent operation and maintenance management system based on AIoT asset | |
| CN101668012A (en) | Method and device for detecting security event | |
| CN110889597A (en) | Method and device for detecting abnormal business timing sequence indexes | |
| CN115311829A (en) | Accurate alarm method and system based on mass data | |
| CN101587161A (en) | Limit value determination method for wafer test parameters | |
| CN117130332A (en) | Intelligent monitoring system for production line of military industry enterprise based on data analysis | |
| CN116681307B (en) | River four-disorder supervision traceability display method and system based on multi-terminal fusion feedback | |
| CN116764261B (en) | Execution safety supervision system for distillation flow | |
| EP2882139B1 (en) | System and method for IT servers anomaly detection using incident consolidation | |
| CN103713976B (en) | Signalling arrangement fault rootstock searching method for centralized signal supervision system | |
| CN108737164B (en) | Method and device for filtering real-time alarm of telecommunication network | |
| CN115168154B (en) | Abnormal log detection method, device and equipment based on dynamic baseline | |
| CN116432007A (en) | Optical fiber early warning mode identification method based on airspace characteristics and machine learning | |
| CN116128308A (en) | Chip distribution operation risk supervision system based on big data analysis | |
| CN114661562A (en) | Data warning method, device, equipment and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20120125 Termination date: 20200119 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |