Parameter control limit audit method, device, and computer equipment

Abstract

The present application relates to a parameter control limit audit method, device, computer equipment, storage medium and computer program product. The method includes: obtaining multiple valid data points of a target parameter within a first time interval, the target parameter being a monitoring parameter of a target product at a target site; based on multiple valid data points, calculating the actual ratio of a preset deviation difference to the deviation difference of the target parameter within the first time interval, the deviation difference includes a standard deviation or a variance, and the control limit of the target parameter used within the first time interval is determined based on the preset deviation difference; analyzing the normal distribution compliance of multiple valid data points; according to different normal distribution compliances, comparing the actual ratio with different reference ratios to determine whether the control limit of the target parameter used within the first time interval is converged. The use of this method can improve the convergence audit accuracy of parameter control limits.

Description

Technical Field

[0001] The present application relates to the field of semiconductor process control technology, and in particular to a parameter control limit audit method, apparatus, computer equipment, storage medium and computer program product. Background Art

[0002] In semiconductor process engineering, statistical principles can usually be used to control process stability based on normal distribution. Generally, 3 times the standard deviation (3σ) can be used as the control limit to detect abnormal process fluctuations when the actual value of the target parameter fluctuates more than 3σ relative to its target value.

[0003] At the same time, in the continuous production process of semiconductor products, the process will be continuously optimized, and the fluctuation of various parameters of the product will gradually decrease, so that the values ​​of various parameters will be more convergent. Therefore, it is necessary to regularly review whether the control limits of various parameters are reasonable and whether they need to converge.

[0004] Then, existing parameter control limits were audited with poor accuracy. Summary of the invention

[0005] Based on this, it is necessary to provide a parameter control limit audit method, apparatus, computer equipment, computer-readable storage medium and computer program product that can improve the accuracy of convergence audit of parameter control limits in response to the above technical problems.

[0006] A parameter control limit audit method, the method comprising:

[0007] Acquire a plurality of valid data points of a target parameter within a first time interval, wherein the target parameter is a monitoring parameter of a target product at a target site;

[0008] Based on the multiple valid data points, calculating an actual ratio of a preset deviation difference to a deviation difference of the target parameter in the first time interval, the deviation difference including a standard deviation or a variance, and a control limit of the target parameter used in the first time interval is determined based on the preset deviation difference;

[0009] Analyzing the normal distribution conformity of the plurality of valid data points;

[0010] According to different conformance conditions of the normal distribution, the actual ratio is compared with different reference ratios to determine whether the control limit of the target parameter used in the first time interval is converged.

[0011] In one embodiment, the step of obtaining a plurality of valid data points of the target parameter within the first time interval includes:

[0012] Acquire a plurality of measurement data points of a target parameter within a first time interval;

[0013] Abnormal points among the plurality of measured data points are filtered out to obtain the plurality of valid data points.

[0014] In one embodiment, the quartile method is used to filter out abnormal points among the plurality of measurement data points.

[0015] In one embodiment, the analyzing the normal distribution conformity of the plurality of valid data points comprises:

[0016] Calculating the normal distribution P value of the multiple valid data points;

[0017] According to different normal distribution compliance situations, the actual ratio is compared with different reference ratios to determine whether the control limit of the target parameter used in the first time interval is converged, including:

[0018] When the normal distribution P value is greater than 0.05, comparing the actual ratio with the first ratio to determine whether the control limit of the target parameter used in the first time interval is converged;

[0019] When the normal distribution P value is less than or equal to 0.05, the actual ratio is compared with the second ratio to determine whether the control limit of the target parameter used in the first time interval is converged, and the second ratio is less than the first ratio.

[0020] In one embodiment, the deviation includes a standard deviation, the first ratio is determined by a variance homogeneity analysis value, and the first ratio is 1.414, and / or the second ratio is 1.

[0021] In one embodiment, the first time interval includes a second time interval and a third time interval, and the second time interval is later than the third time interval.

[0022] After obtaining a plurality of valid data points of the target parameter within the first time interval, the method further includes:

[0023] Obtaining the number of valid data points in a second time interval;

[0024] When the number of valid data points in the second time interval is greater than or equal to the preset number of points, the width of the control limit used in the first time interval is compared with the preset width, the preset width being 1 / N times the width of the product specification limit, where N is less than 1;

[0025] When the width of the control limit to be reviewed is smaller than the preset width, it is determined that the control limit is not converged.

[0026] A parameter control limit audit device, the device comprising:

[0027] A first acquisition module, used to acquire a plurality of valid data points of a target parameter within a first time interval, wherein the target parameter is a monitoring parameter of a target product at a target site;

[0028] a calculation module, configured to calculate, based on the plurality of valid data points, an actual ratio of a preset deviation difference to a deviation difference of the target parameter within the first time interval, the deviation difference comprising a standard deviation or a variance, and a control limit of the target parameter used within the first time interval being determined based on the preset deviation difference;

[0029] An analysis module, used for analyzing the normal distribution conformity of the plurality of valid data points;

[0030] The first comparison module is used to compare the actual ratio with different reference ratios according to different normal distribution compliance situations to determine whether the control limit of the target parameter used in the first time interval is converged.

[0031] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0032] A computer-readable storage medium stores a computer program, which implements the steps of the above method when executed by a processor.

[0033] A computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0034] The above-mentioned parameter control limit audit method, device, computer equipment, storage medium and computer program product calculate the actual ratio of the preset deviation difference to the deviation difference of the target parameter in the first time interval based on multiple valid data points of the target parameter in the first time interval, that is, calculate the actual ratio of the old deviation difference to the new deviation difference in the first time interval. In addition, a normality test is added to analyze the normal distribution compliance of multiple valid data points. Based on the hypothesis test principle, the actual ratio is compared with different reference ratios under normal distribution and non-normal distribution. At this time, it can be more accurately judged whether the new process fluctuation after optimization is compared with the original process fluctuation (old process fluctuation) to see whether it is truly converged, so as to more accurately determine whether the control limit of the target parameter used in the first time interval is converged, thereby reducing the false convergence caused by machine error, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1A schematic diagram of a process flow of a parameter control limit audit method in one embodiment;

[0036] Figure 2 is a schematic flow chart of step S10 in one embodiment;

[0037] Figure 3 A schematic diagram of a flow chart of a parameter control limit audit method in another embodiment;

[0038] Figure 4 A partial flow chart of a parameter control limit audit method in yet another embodiment;

[0039] Figure 5 is a schematic diagram of distribution of measurement data points of a target parameter within a first time interval in one embodiment;

[0040] Figure 6 It is a structural block diagram of a parameter control limit audit device in one embodiment. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0042] In one embodiment, see Figure 1 , a parameter control limit audit method is provided, comprising the following steps:

[0043] Step S10, obtaining a plurality of valid data points of a target parameter within a first time interval, where the target parameter is a monitoring parameter of a target product at a target site.

[0044] The first time interval is a time interval for which data analysis is required, for example, the first time interval is a time interval of the past 6 months starting from the current time.

[0045] At least one semiconductor product can be processed in the first time interval. When processing the same product, it needs to be processed at multiple processing stations. For the same processing station, at least one parameter needs to be monitored.

[0046] The target product is any semiconductor product processed in the first time interval. The target site is any processing site in the semiconductor product processing process. The target parameter is any parameter that needs to be monitored at the target site.

[0047] For a semiconductor product, it can be processed in batches at each processing site. After processing, for the same monitoring parameter, a measurement data can be obtained for each batch, thereby forming a data point. The measurement data of each batch can be obtained, for example, based on the average of the measurement data of each wafer in the batch.

[0048] Based on this, each valid data point of the target parameter may be a data point about the target parameter of a corresponding batch of target products at the target site.

[0049] Step S30, calculating the actual ratio of the preset deviation difference to the deviation difference of the target parameter in the first time interval, the deviation difference includes the standard deviation or variance, and the control limit of the target parameter used in the first time interval is determined based on the preset deviation difference.

[0050] As an example, the deviation includes a standard deviation.

[0051] According to a plurality of valid data points of the target parameter in the first time interval, a standard deviation σ1 of the target parameter in the first time interval may be calculated.

[0052] At the same time, the preset deviation difference may be a preset standard deviation σ2. The preset standard deviation σ2 may be obtained, for example, by calculation based on data in a time interval before the first time interval. The control limit CL1 of the target parameter used in the first time interval may be obtained by floating 3σ2 above and below the target value of the target parameter. It can be understood that the control limit of the target parameter used in the first time interval is the control limit of the target parameter when the semiconductor product is processed and prepared in the first time interval.

[0053] At this time, the actual ratio of the preset deviation difference to the deviation difference of the target parameter in the first time interval is σ2 / σ1.

[0054] As another example, the deviation includes the variance.

[0055] According to multiple valid data points of the target parameter in the first time interval, the variance σ1 of the target parameter in the first time interval can be calculated 2 .

[0056] At the same time, the preset deviation can be the preset variance σ2 2 . The default variance is σ2 2 For example, the control limit CL1 of the target parameter used in the first time interval can be obtained by calculating the data in the time interval before the first time interval. The control limit CL1 of the target parameter used in the first time interval can be obtained by floating 3σ2 above and below the target value of the target parameter.

[0057] At this time, the actual ratio of the preset deviation difference to the deviation difference of the target parameter in the first time interval is:

[0058] σ22 / σ1 2 .

[0059] Step S40, analyzing the normal distribution conformity of multiple valid data points.

[0060] Based on the characteristics of normal distribution, we can analyze and determine whether the distribution of multiple valid data points conforms to normal distribution.

[0061] Step S50, comparing the actual ratio with different reference ratios according to different normal distribution compliance situations, to determine whether the control limits of the target parameters used in the first time interval are converged.

[0062] Specifically, the situation where the distribution of multiple valid data points conforms to the normal distribution may be the first situation. The situation where the distribution of multiple valid data points does not conform to the normal distribution may be the second situation. Different reference ratios may be selected in the first situation and the second situation. According to different reference ratios, it may be determined whether the control limit CL1 adopted in the first time interval is converged.

[0063] In this embodiment, based on multiple valid data points of the target parameter in the first time interval, the actual ratio of the preset deviation difference to the deviation difference of the target parameter in the first time interval is calculated, that is, the actual ratio of the old deviation difference to the new deviation difference in the first time interval is calculated. In addition, a normality test is added to analyze the normal distribution compliance of multiple valid data points. Based on the hypothesis test principle, the actual ratio is compared with different reference ratios under normal distribution and non-normal distribution. At this time, it can be more accurately judged whether the new process fluctuation after optimization is compared with the original process fluctuation (old process fluctuation) to see whether it is truly converged, so as to more accurately determine whether the control limit CL1 used in the first time interval is converged, thereby reducing the false convergence caused by machine error, etc.

[0064] In one embodiment, step S10 includes:

[0065] Step S11, obtaining a plurality of measurement data points of a target parameter within a first time interval.

[0066] In the first time interval, multiple batches of target products can be processed at the target site of the target product. For each batch, a measurement data point can be obtained for the target parameter, thereby obtaining multiple measurement data points.

[0067] Step S12, filtering out abnormal points from a plurality of measured data points to obtain a plurality of valid data points.

[0068] By filtering out the data of abnormal points, valid data points that meet the preset requirements can be obtained. The preset requirements can be set according to actual needs.

[0069] As an example, the quartile method may be used to filter out abnormal points among a plurality of measurement data points.

[0070] Specifically, the obtained multiple measurement data points can be arranged to obtain data points Q1, Q2 and Q3 with increasing values. Q1, Q2 and Q3 can divide all the measurement data points into four parts. Among them, Q2 can be the median of all the measurement data points, Q1 is the median of the measurement data points less than Q2, and Q3 is the median of the measurement data points greater than Q2.

[0071] Based on Q1 and Q3, the data upper limit and data lower limit can be determined. The data upper limit can be expressed as: Q1-n (Q3-Q1); the data lower limit can be expressed as: Q3+n (Q3-Q1). The value of n can be set according to actual needs. Data points greater than the data upper limit and less than the data lower limit can be regarded as abnormal points, so they can be filtered out. The smaller the n value, the more abnormal points are filtered out.

[0072] For example, the value of n may be between 1.5 and 3, such as 1.5 or 2 or 3.

[0073] The traditional method of filtering out abnormal points usually uses product specification limits to filter out data. Product specification limits can usually only filter out significant outlier data, but for data points that deviate from the overall performance due to experiments or batch operations or other production anomalies, they cannot be filtered out when they do not reach the product specification limits. At this time, these data points may expand the standard deviation calculated from multiple valid data points of the target parameter in the first time interval, resulting in a wider new control limit CL2.

[0074] The quartile method is used to filter out abnormal points in multiple measurement data points, which can more accurately filter out abnormal points from a statistical perspective.

[0075] Specifically, Figure 5 As shown in the figure, Y1 and Y2 represent product specification limits. As can be seen from the figure, when filtering data through product specification limits, the two data points located in the dotted circle deviate from the overall performance, but do not exceed the lower limit of the product specification limit, so they will not be filtered out. At this time, the two data points will expand the standard deviation obtained based on the filtered multiple valid data points, thereby increasing the width of the new control limit CL2 obtained based on the standard deviation.

[0076] When filtering data based on the upper and lower limits determined by Q1 and Q3, the two data points within the dotted circle will be judged as outliers and filtered out using the quartile method. At this time, the new control limit CL2 calculated based on the standard deviation of multiple valid data points after filtering will be more accurate.

[0077] In one embodiment, step S40 includes:

[0078] Step S41, calculating the normal distribution P value of multiple valid data points.

[0079] A plurality of valid data points can be input into a relevant software model to calculate a normal distribution P value. The software model that can calculate a normal distribution P value is well known to those skilled in the art and will not be described in detail here.

[0080] According to the normal distribution P value, we can effectively judge whether the distribution of multiple valid data points conforms to the normal distribution.

[0081] Step S50 includes:

[0082] Step S51, when the normal distribution P value is greater than 0.05, the actual ratio is compared with the first ratio to determine whether the control limit of the target parameter used in the first time interval is converged.

[0083] According to statistical principles, a normal distribution P value greater than 0.05 indicates that the distribution of multiple valid data points conforms to a normal distribution. At this time, the first ratio can be used as a reference ratio.

[0084] It can be understood that the specific value of the first ratio can be adjusted according to user needs, and specifically can be adjusted according to the degree of acceptance of fluctuations in the production process.

[0085] As an example, the deviation difference includes a standard deviation, the first ratio is determined by a variance homogeneity analysis value, and the first ratio is 1.414.

[0086] The preset deviation difference can be regarded as the deviation difference obtained based on the data conforming to the normal distribution. Meanwhile, when the multiple valid data points of the target parameter in the first time interval conform to the normal distribution, the deviation difference of the target parameter is also obtained based on the data conforming to the normal distribution.

[0087] According to statistical principles, processing data that conforms to normal distribution can obtain chi-square distribution. Processing data with two chi-square distributions can obtain F distribution. For data that conforms to F distribution, when the ratio of the variances of the two groups of data is greater than 2, it can be considered that the two groups of data are truly different. Otherwise, it can be considered that there is no real difference between the two groups of data, but normal data fluctuations caused by different sample sizes. Therefore, when the ratio of the standard deviations of the two groups of data is greater than (approximately equal to 1.414), it can be considered that the two sets of data are truly different, otherwise there are no real differences between the two sets of data.

[0088] Based on this, the first ratio can be set to 1.414.

[0089] When the actual ratio of the calculated preset deviation difference (preset standard deviation) to the deviation difference (standard deviation) of the target parameter in the first time interval is greater than 1.414, it means that the data fluctuation of the target parameter in the first time interval is actually reduced relative to the previous data fluctuation, so that the control limit used in the first time interval can be converged according to the standard deviation of the target parameter in the first time interval. Specifically, the width of the control limit can be converged from the original CL1 to CL2 calculated based on the standard deviation of the target parameter in the first time interval.

[0090] When the actual ratio of the calculated preset deviation difference (preset standard deviation) to the deviation difference (standard deviation) of the target parameter in the first time interval is less than or equal to 1.414, it means that the data fluctuation of the target parameter in the first time interval has not actually decreased relative to the previous data fluctuation, so that the control limit CL1 adopted in the first time interval can be kept unchanged without converging.

[0091] As an example, the target value of the target parameter in the first time interval is 10300Å. The upper limit UCL of the control limit adopted by the target parameter in the first time interval is 10455Å, and the lower limit LCL is 10145Å. The preset deviation is the preset standard deviation σ2, which is 51.7Å. According to multiple valid data points of the target parameter in the first time interval, the calculated standard deviation σ1 of the target parameter in the first time interval is 47.9Å.

[0092] In this case, if the normal distribution of multiple valid data points is not analyzed, but the standard deviation σ1 of the target parameter in the first time interval is directly converged, the width of the control limit will converge to 11.3Å, and the upper limit value UCL of the new control limit after convergence is 10443.7Å, and the upper limit value LCL of the new control limit after convergence is 10156.3Å.

[0093] According to this method, before convergence, the normal distribution of multiple valid data points is first analyzed. If multiple valid data points conform to the normal distribution, but σ2 / σ1 is less than 1.414, it means that the data fluctuation of the target parameter in the first time interval has not actually decreased relative to the previous data fluctuation. At this time, keeping the control limit CL1 used in the first time interval unchanged and not converging it can effectively improve the accuracy of the convergence review.

[0094] As described above, at least one semiconductor product can be processed in the first time interval. When processing the same product, it needs to be processed at multiple processing stations. For the same processing station, at least one parameter needs to be monitored.

[0095] When it is determined that the control limit of the target parameter used in the first time interval is converged, the target site adds a convergence item.

[0096] The method of the present application can effectively prevent the control limit from being incorrectly converged, thereby greatly reducing the number of convergence items of the processing site. For example, according to actual production data statistics, when the normal distribution conformity analysis of multiple valid data points is not performed, the sum of the convergence items of all processing sites obtained is 8282, while after the normal distribution conformity analysis of multiple valid data points by the present method, the sum of the convergence items of all processing sites obtained is 3054. The present method can automatically filter out 1-3054 / 8282=63.12% of invalid convergence items.

[0097] Step S52, when the normal distribution P value is less than or equal to 0.05, the actual ratio is compared with the second ratio to determine whether the control limit of the target parameter used in the first time interval is converged, and the second ratio is less than the first ratio.

[0098] According to statistical principles, if the normal distribution P value is less than or equal to 0.05, it means that the distribution of multiple valid data points does not conform to the normal distribution. At this time, the second ratio can be used as a reference ratio for the actual ratio.

[0099] It can be understood that the specific value of the second ratio can also be adjusted according to user needs, and specifically can be adjusted according to the degree of acceptance of fluctuations in the production process.

[0100] As an example, the second ratio takes a value of 1.

[0101] When the distribution of multiple valid data points of the target parameter in the first time interval does not conform to the normal distribution, it indicates that there is an abnormality in the production process. At this time, the standard deviation or variance is likely to be inflated.

[0102] In this case, if the actual ratio is still greater than 1, it means that the standard deviation or variance is easily inflated, and the standard deviation or variance of the target parameter in the first time interval is still smaller than the original standard deviation or variance (preset deviation).

[0103] At this time, the control limits used in the first time interval can be converged according to the standard deviation of the target parameter in the first time interval, so that it is easier to monitor production anomalies in the subsequent production process, thereby facilitating analysis and control of production anomalies. Specifically, the width of the control limit can be converged from the original CL1 to CL2 calculated based on the standard deviation of the target parameter in the first time interval.

[0104] When the actual ratio is less than or equal to 1, the control limit CL1 used in the first time interval may be kept unchanged without being converged, thereby preventing false convergence of the control limit.

[0105] In one embodiment, the first time interval includes a second time interval and a third time interval, and the second time interval is later than the third time interval.

[0106] As an example, the first time interval is a time interval of the past 6 months from the current time, and the second time interval is a time interval of the past 3 months from the current time.

[0107] After step S10, the method further includes:

[0108] Step S21, obtaining the number of valid data points in the second time interval.

[0109] For example, the number of valid data points within the past three months starting from the current time may be obtained.

[0110] Step S22, when the number of valid data points in the second time interval is greater than the preset number of points, compare the width of the control limit used in the first time interval with the preset width, the preset width is 1 / N times the width of the product specification limit, and the value of n is greater than 8.

[0111] The preset number of points and the value of N can be set according to actual needs. For example, the preset number of points can be set to 15 to 30, specifically 20. And the preset number of points can be set to 10.

[0112] When the number of valid data points in the second time interval is greater than or equal to the preset number of points, it means that the number of target products processed at the target site recently is sufficient to perform statistical analysis.

[0113] At this time, the width of the control limit CL1 used in the first time interval is compared with the preset width, so as to facilitate determining the process stability of the target site in the first time interval according to the width of the control limit CL1 used in the first time interval.

[0114] Step S23: When the width of the control limit to be reviewed is smaller than the preset width, it is determined not to converge the control limit of the target parameter used in the first time interval.

[0115] When the width of the control limit CL1 used in the first time interval is less than the preset width, that is, the width of the control limit CL1 of the target parameter used in the first time interval is less than 1 / N times the width of the product specification limit, it means that the control limit CL1 used in the first time interval is small enough, the process stability of the target site is good enough, and there is no need to further optimize the process, so there is no need to further converge the control limit. Therefore, at this time, the non-convergent control limit can be directly determined.

[0116] When the width of the control limit CL1 used in the first time interval is not less than the preset width, steps S30 to S50 may be continued to optimize the process of the target site by controlling the control limit of the target parameter.

[0117] In one embodiment, when the number of valid data points in the second time interval is less than the preset number of points, it means that the number of target products processed at the target site recently is not large enough and has no statistical significance, so the control limit convergence analysis of the target parameters may no longer be continued.

[0118] At the same time, in the actual production process of the factory, many kinds of products are usually generated within a period of time. When the number of valid data points in the second time interval is less than the preset number of points, it means that the number of target products produced in the near future is not enough, which also indicates that the product is not the main product produced in the near future and does not need more attention. At this time, no more control limit convergence analysis is performed on the target parameters, which can effectively save production costs.

[0119] It should be understood that although Figure 1-4 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1-4 At least part of the steps may include multiple steps or multiple stages. These steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed in turn or alternately with other steps or at least part of the steps or stages in other steps.

[0120] In one embodiment, Figure 6 As shown, a parameter control limit audit device is provided, comprising: a first acquisition module 100, a calculation module 200, an analysis module 300 and a first comparison module 400, wherein:

[0121] The first acquisition module 100 is used to acquire a plurality of valid data points of a target parameter within a first time interval, where the target parameter is a monitoring parameter of a target product at a target site.

[0122] The calculation module 200 is used to calculate the actual ratio of the preset deviation difference to the deviation difference of the target parameter in the first time interval based on multiple valid data points. The deviation difference includes the standard deviation or variance. The control limit adopted by the target parameter in the first time interval is determined based on the preset deviation difference.

[0123] The analysis module 300 is used to analyze the normal distribution conformity of multiple valid data points.

[0124] The first comparison module 400 is used to compare the actual ratio with different reference ratios according to different normal distribution compliance situations to determine whether the control limit of the target parameter used in the first time interval is converged.

[0125] In one embodiment, the first acquisition module includes an acquisition unit and a filtering unit.

[0126] The acquisition unit is used to acquire a plurality of measurement data points of the target parameter within a first time interval;

[0127] The filtering unit is used to filter out abnormal points from a plurality of measured data points to obtain a plurality of valid data points.

[0128] In one embodiment, the filtering unit uses the quartile method to filter out abnormal points from the plurality of measurement data points.

[0129] In one embodiment, the analysis module is used to calculate the normal distribution P value of multiple valid data points.

[0130] At the same time, the first comparison module is used to compare the actual ratio with the first ratio when the normal distribution P value is greater than 0.05, to determine whether the control limits of the target parameters used in the first time interval are converged; when the normal distribution P value is less than or equal to 0.05, the actual ratio is compared with the second ratio to determine whether the control limits of the target parameters used in the first time interval are converged, and the second ratio is less than the first ratio.

[0131] In one embodiment, the deviation comprises a standard deviation, the first ratio is determined by a variance homogeneity analysis value, and the first ratio is 1.414, and / or the second ratio is 1.

[0132] In one embodiment, the first time interval includes a second time interval and a third time interval, and the second time interval is later than the third time interval. The parameter control limit auditing device further includes a second acquisition module and a second comparison module.

[0133] The second acquisition module is used to acquire the number of valid data points in the second time interval.

[0134] The second comparison module is used to compare the width of the control limit used in the first time interval with the preset width when the number of valid data points in the second time interval is greater than or equal to the preset number of input points. The preset width is 1 / N times the width of the product specification limit, and N is less than 1; when the width of the control limit to be reviewed is less than the preset width, it is determined that the control limit is not converged.

[0135] In one embodiment, the second comparison module is further configured to stop performing control limit convergence analysis on the target parameter when the number of valid data points in the second time interval is less than a preset number of points.

[0136] For the specific definition of the parameter control limit audit device, please refer to the definition of the parameter control limit audit method above, which will not be repeated here. Each module in the above-mentioned parameter control limit audit device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0137] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0138] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0139] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0140] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0141] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0142] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A parameter control limit audit method, characterized in that: The method comprises: Acquire a plurality of valid data points of a target parameter within a first time interval, wherein the target parameter is a monitoring parameter of a target product at a target site; Based on the multiple valid data points, calculating an actual ratio of a preset deviation difference to a deviation difference of the target parameter in the first time interval, the deviation difference including a standard deviation or a variance, and a control limit of the target parameter used in the first time interval is determined based on the preset deviation difference; Analyzing the normal distribution conformity of the plurality of valid data points; According to different conformance conditions of the normal distribution, the actual ratio is compared with different reference ratios to determine whether the control limit of the target parameter used in the first time interval is converged; The first time interval includes a second time interval and a third time interval, the second time interval is later than the third time interval, After obtaining a plurality of valid data points of the target parameter within the first time interval, the method further includes: Obtaining the number of valid data points in a second time interval; When the number of valid data points in the second time interval is greater than or equal to the preset number of points, the width of the control limit used in the first time interval is compared with the preset width, the preset width being 1 / N times the width of the product specification limit, where N is less than 1; When the width of the control limit to be reviewed is smaller than the preset width, it is determined that the control limit is not converged.

2. The parameter control limit audit method according to claim 1, characterized in that: The step of obtaining a plurality of valid data points of the target parameter within the first time interval includes: Acquire a plurality of measurement data points of a target parameter within a first time interval; Abnormal points among the plurality of measured data points are filtered out to obtain the plurality of valid data points.

3. The parameter control limit audit method according to claim 2, characterized in that: The quartile method is used to filter out abnormal points in the plurality of measurement data points.

4. The parameter control limit audit method according to claim 1, characterized in that: The analyzing whether the normal distribution of the plurality of valid data points is consistent with the situation includes: Calculating the normal distribution P value of the multiple valid data points; According to different normal distribution compliance situations, the actual ratio is compared with different reference ratios to determine whether the control limit of the target parameter used in the first time interval is converged, including: When the normal distribution P value is greater than 0.05, comparing the actual ratio with the first ratio to determine whether the control limit of the target parameter used in the first time interval is converged; When the normal distribution P value is less than or equal to 0.05, the actual ratio is compared with the second ratio to determine whether the control limit of the target parameter used in the first time interval is converged, and the second ratio is less than the first ratio.

5. The parameter control limit audit method according to claim 4, characterized in that: The deviation includes a standard deviation, the first ratio is determined by a variance homogeneity analysis value, and the first ratio is 1.414, and / or the second ratio is 1.

6. The parameter control limit audit method according to claim 1, characterized in that: The first time interval is a time interval of the past 6 months from the current time, and the second time interval is a time interval of the past 3 months from the current time.

7. A parameter control limit audit device, characterized in that: The device comprises: A first acquisition module, used to acquire a plurality of valid data points of a target parameter within a first time interval, wherein the target parameter is a monitoring parameter of a target product at a target site; a calculation module, configured to calculate, based on the plurality of valid data points, an actual ratio of a preset deviation difference to a deviation difference of the target parameter within the first time interval, the deviation difference comprising a standard deviation or a variance, and a control limit of the target parameter used within the first time interval being determined based on the preset deviation difference; An analysis module, used for analyzing the normal distribution conformity of the plurality of valid data points; A first comparison module is used to compare the actual ratio with different reference ratios according to different normal distribution compliance conditions to determine whether the control limit of the target parameter used in the first time interval is converged; The first time interval includes a second time interval and a third time interval, the second time interval is later than the third time interval, and the parameter control limit audit device also includes a second acquisition module and a second comparison module. The second acquisition module is used to obtain the number of valid data points in the second time interval. The second comparison module is used to compare the width of the control limit used in the first time interval with the preset width when the number of valid data points in the second time interval is greater than or equal to the preset number of input points. The preset width is 1 / N times the width of the product specification limit, and N is less than 1; when the width of the control limit to be reviewed is less than the preset width, it is determined that the control limit is not converged.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images