CN117280286A - Method, apparatus and storage medium for analyzing correlation of production process data - Google Patents
Method, apparatus and storage medium for analyzing correlation of production process data Download PDFInfo
- Publication number
- CN117280286A CN117280286A CN202280000648.1A CN202280000648A CN117280286A CN 117280286 A CN117280286 A CN 117280286A CN 202280000648 A CN202280000648 A CN 202280000648A CN 117280286 A CN117280286 A CN 117280286A
- Authority
- CN
- China
- Prior art keywords
- parameter data
- data
- characteristic parameter
- test parameter
- test
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 130
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 103
- 238000003860 storage Methods 0.000 title claims abstract description 18
- 238000012360 testing method Methods 0.000 claims abstract description 206
- 238000012545 processing Methods 0.000 claims abstract description 40
- 238000010219 correlation analysis Methods 0.000 claims abstract description 39
- 230000008569 process Effects 0.000 claims description 52
- 238000004364 calculation method Methods 0.000 claims description 31
- 238000009826 distribution Methods 0.000 claims description 14
- 238000010586 diagram Methods 0.000 claims description 8
- 238000007781 pre-processing Methods 0.000 claims description 7
- 230000006870 function Effects 0.000 claims description 6
- 230000004927 fusion Effects 0.000 claims description 6
- 238000007689 inspection Methods 0.000 claims description 6
- 239000010408 film Substances 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 4
- 239000010409 thin film Substances 0.000 claims description 4
- 239000011229 interlayer Substances 0.000 claims description 2
- 239000011800 void material Substances 0.000 claims description 2
- 238000001514 detection method Methods 0.000 description 27
- 230000002159 abnormal effect Effects 0.000 description 10
- 230000007547 defect Effects 0.000 description 10
- 239000000523 sample Substances 0.000 description 9
- 238000012423 maintenance Methods 0.000 description 5
- 238000012352 Spearman correlation analysis Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 4
- 239000011521 glass Substances 0.000 description 4
- 239000011159 matrix material Substances 0.000 description 4
- 238000012795 verification Methods 0.000 description 4
- 230000005856 abnormality Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000000758 substrate Substances 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 230000008439 repair process Effects 0.000 description 2
- 238000012313 Kruskal-Wallis test Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 239000010410 layer Substances 0.000 description 1
- 238000007620 mathematical function Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
Abstract
A production process data correlation analysis method, device and storage medium, the production process data correlation analysis method includes: acquiring production process data, wherein the production process data comprises a plurality of characteristic parameter data and a plurality of test parameter data, the characteristic parameter data comprises a first position coordinate and a characteristic parameter value corresponding to the first position coordinate, and the test parameter data comprises a second position coordinate and a test parameter value (101) corresponding to the second position coordinate; interpolation processing is carried out on the characteristic parameter data and/or the test parameter data to obtain n characteristic parameter data and n test parameter data, the distance between a first position coordinate in the ith characteristic parameter data and a second position coordinate in the ith test parameter data is within a preset distance range, n is a natural number greater than 1, and i' is a natural number between 1 and n (102); based on the n feature parameter data and the n test parameter data, a correlation between the feature parameter and the test parameter is calculated (103).
Description
The embodiment of the disclosure relates to the technical field of intelligent systems, and in particular relates to a method, equipment and a storage medium for analyzing the relativity of production process data.
In the production process of the display panel, the display panel is required to be sequentially processed by a plurality of process equipment. Meanwhile, there is a certain probability that various defects are inevitably present in the final display panel product, and the defects are essentially caused by the process equipment. Therefore, the method for determining the correlation between the process equipment and the defects has important significance in positioning the defects, adjusting the production flow and the like.
Disclosure of Invention
The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.
The embodiment of the disclosure provides a production process data correlation analysis method, which comprises the following steps:
acquiring production process data, wherein the production process data comprises a plurality of characteristic parameter data and a plurality of test parameter data, the characteristic parameter data comprises a first position coordinate and a characteristic parameter value corresponding to the first position coordinate, and the test parameter data comprises a second position coordinate and a test parameter value corresponding to the second position coordinate;
interpolation processing is carried out on the characteristic parameter data and/or the test parameter data to obtain n characteristic parameter data and n test parameter data, wherein the distance between a first position coordinate in the ith characteristic parameter data and a second position coordinate in the ith test parameter data is within a preset distance range, n is a natural number greater than 1, and i' is a natural number between 1 and n;
And calculating the correlation between the characteristic parameters and the test parameters according to the n characteristic parameter data and the n test parameter data.
In some exemplary embodiments, the interpolation processing is performed on the characteristic parameter data and/or the test parameter data to obtain n pieces of characteristic parameter data and n pieces of characteristic parameter data, including one of the following:
when n pieces of characteristic parameter data exist and n1 pieces of test parameter data exist, and the distance between a second position coordinate in the ith piece of test parameter data and one first position coordinate in the characteristic parameter data is within a preset distance range, carrying out interpolation processing on the test parameter data according to the (n-n 1) pieces of characteristic parameter data, wherein the second position coordinate in the interpolated test parameter data is the first position coordinate in the (n-n 1) pieces of characteristic parameter data, and i' is more than or equal to 1 and less than or equal to n1< n;
when n pieces of test parameter data exist and n1 pieces of characteristic parameter data exist, and the distance between a first position coordinate in the ith' piece of characteristic parameter data and one second position coordinate in the test parameter data is within a preset distance range, carrying out interpolation processing on the characteristic parameter data according to the (n-n 1) pieces of test parameter data, wherein the first position coordinate in the interpolated characteristic parameter data is the second position coordinate in the (n-n 1) pieces of test parameter data;
When n1 pieces of the characteristic parameter data and n2 pieces of the test parameter data exist, and the presence of n3 pieces of the characteristic parameter data and n3 pieces of the test parameter data satisfy that a distance between a first position coordinate in the ith' piece of the characteristic parameter data and one second position coordinate in the test parameter data is within a preset distance range, interpolation processing is performed on the characteristic parameter data according to (n-n 1) pieces of the test parameter data, interpolation processing is performed on the test parameter data according to (n-n 2) pieces of the characteristic parameter data, and a first position coordinate in the interpolated characteristic parameter data is a second position coordinate in the (n-n 1) pieces of the test parameter data, and a second position coordinate in the interpolated test parameter data is a first position coordinate in the (n-n 2) pieces of the characteristic parameter data, wherein 1 i ' < n1< n3< n,1 is less than or equal to i ' < n2< n.
In some exemplary embodiments, in the acquired production process data, the characteristic parameter data or the test parameter data includes m record values: { z (x) 1 ，y 1 )，z(x 2 ，y 2 )，z(x 3 ，y 3 )，...，z(x m-1 ，y m-1 )，z(x m ，y m ) Wherein m is a natural number less than or equal to n, (x) i ，y i ) Representing the first bitSet coordinates or second position coordinates, z (x i ，y i ) Representing a characteristic parameter value or a test parameter value, i being a natural number between 1 and m, the first or second position coordinates to be interpolated being (x 0 ，y 0 ) Pair (x) 0 ，y 0 ) The interpolation processing is carried out on the characteristic parameter data or the test parameter data, and the method comprises the following steps:
calculating the distance and half variance between the m recorded values;
drawing the calculated distance and the half variance into a scatter diagram, and searching for the relation between the fitting distance of the fitting curve and the half variance to obtain a functional relation;
from the obtained functional relation, a calculation (x 0 ，y 0 ) To (x) i ，y i ) Half variance between;
substituting the calculated half variance into a pre-constructed Lagrangian equation set to solve the optimal weight coefficient;
substituting the solved optimal weight coefficient into an interpolation calculation formula to obtain (x) 0 ，y 0 ) An estimate of the position.
In some exemplary embodiments, the set of preconfigured lagrangian equations is:
wherein lambda is i For optimal weight coefficients, φ is the Lagrangian multiplier, r i0 Is (x) 0 ，y 0 ) To (x) i ，y i ) Half variance of r ij Is (x) i ，y i ) And (x) j ，y j ) Half variance between;
the interpolation calculation formula is as follows:z i is (x) i ，y i ) A characteristic parameter value or a test parameter value, c is the mean value of the m recorded values,is (x) 0 ，y 0 ) An estimate of the position.
In some exemplary embodiments, the set of preconfigured lagrangian equations is:
wherein lambda is i For optimal weight coefficients, φ is the Lagrangian multiplier, r i0 Is (x) 0 ，y 0 ) To (x) i ，y i ) Half variance of r ij Is (x) i ，y i ) And (x) j ，y j ) Half variance between;
the interpolation calculation formula is as follows:z i is (x) i ，y i ) At characteristic parameter values or test parameter values,is (x) 0 ，y 0 ) An estimate of the position.
In some exemplary embodiments, (X) 1 ，X 2 ，...，X n ) For n of said characteristic parameter values, (Y) 1 ，Y 2 ，...，Y n ) For n values of the test parameter, the correlation between the characteristic parameter and the test parameter is calculated by the following formula:
wherein ρ is S Is the correlation coefficient, d i′ Is X i′ And Y i′ Level differences between.
In some exemplary embodiments, the method further comprises:
sequentially arranging n characteristic parameter values and n test parameter values;
calculating the rank sum of n characteristic parameter values and n test parameter values in the arrangement;
calculating a sample statistic according to the calculated rank sum:
calculating a significance level by using a probability density function of chi-square distribution and the calculated sample statistics;
verifying whether a correlation calculation result between the characteristic parameter and the test parameter is valid according to the calculated significance level;
And outputting the correlation calculation result when the correlation calculation result is verified to be effective.
In some exemplary embodiments, the sample statistic is calculated according to the following formula:
where H is the sample statistic, n= 2*n, R x And R is y Representing the rank sum of the n said characteristic parameter values and the n said test parameter values in the permutation, respectively;
the significance level was calculated according to the following formula:
wherein pvalue is the level of significance,for the gamma distribution, k=2,
in some exemplary embodiments, the characteristic parameter is a process parameter and the test parameter is an electrical parameter.
In some exemplary embodiments, the process parameters include at least one of: the surface resistance, the involution precision between the thin film transistor and the color film, the line width, the film thickness and the interlayer registration precision; the electrical parameter includes at least one of: threshold voltage, mobility, operating current and reverse off current.
In some exemplary embodiments, the production process data further comprises a plurality of historical parameter data, the historical parameter data comprising at least one of: production time, production plant, product model, product type, inspection station, and process station.
In some exemplary embodiments, the method further comprises at least one of:
preprocessing the production process data, wherein the preprocessing comprises void removal data, duplicate removal data and useless field removal;
and carrying out data fusion on the characteristic parameter data, the test parameter data and the resume parameter data.
In some exemplary embodiments, the method further comprises: receiving a first input of a user, and determining the range of the acquired production process data according to the first input; the first input includes at least one of:
production time, production factory, product type, product model, inspection station, electrical parameters, process station, process parameters.
In some exemplary embodiments, the method further comprises: and outputting a correlation trend graph of the electrical parameters and the process parameters in the first input in a preset time period.
The embodiment of the disclosure also provides a production process data correlation analysis device, which comprises a memory; and a processor coupled to the memory, the processor configured to perform the steps of the production process data correlation analysis method as set forth in any one of the above, based on instructions stored in the memory.
The presently disclosed embodiments also provide a storage medium having stored thereon a computer program which, when executed by a processor, implements a method of production process data correlation analysis as described in any of the above.
Other aspects will become apparent upon reading and understanding the accompanying drawings and detailed description.
The accompanying drawings are included to provide a further understanding of the disclosed embodiments and are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the description serve to explain, without limitation, the embodiments of the disclosure. The shapes and sizes of various components in the drawings are not to scale true, and are intended to be illustrative of the present disclosure.
FIG. 1 is a flow chart of a method for process data correlation analysis provided in an exemplary embodiment of the present disclosure;
FIG. 2 is a schematic diagram of a user interface provided by an exemplary embodiment of the present disclosure;
FIG. 3 is a flow chart of another method for process data correlation analysis provided by an exemplary embodiment of the present disclosure;
fig. 4 is a schematic structural view of a production process data correlation analysis apparatus according to an exemplary embodiment of the present disclosure.
For the purposes of making the objects, technical solutions and advantages of the present disclosure more apparent, embodiments of the present disclosure will be described in detail hereinafter with reference to the accompanying drawings. It should be noted that, without conflict, the embodiments of the present disclosure and features of the embodiments may be arbitrarily combined with each other.
Unless otherwise defined, technical or scientific terms used in the disclosure of the embodiments of the present disclosure should be given the ordinary meaning as understood by one of ordinary skill in the art to which the present disclosure pertains. The terms "first," "second," and the like, as used in embodiments of the present disclosure, do not denote any order, quantity, or importance, but rather are used to distinguish one element from another. The word "comprising" or "comprises", and the like, is intended to mean that elements or items preceding the word encompass the elements or items listed thereafter and equivalents thereof without precluding other elements or items.
In recent years, with the rapid development of intelligent learning and computer software and hardware, computer intelligent learning technology has been widely applied in various fields of manufacturing industry, such as equipment maintenance, intelligent monitoring, and bad detection.
In the display panel production process, the display panel product may be damaged due to the failure of equipment at a certain production site. Because the equipment is numerous, can't confirm by which equipment or which equipment is responsible for in the first time, consequently when display panel produces the trouble in the production process, need take a large amount of time to look for the specific equipment that leads to producing the trouble, just so reduced display panel's production efficiency, improved the defective rate.
As shown in fig. 1, an embodiment of the present disclosure provides a method for analyzing correlation of production process data, including:
step 101, acquiring production process data, wherein the production process data comprises a plurality of characteristic parameter data and a plurality of test parameter data, the characteristic parameter data comprises a first position coordinate and a characteristic parameter value corresponding to the first position coordinate, and the test parameter data comprises a second position coordinate and a test parameter value corresponding to the second position coordinate;
the method for analyzing the correlation of the production process data in the embodiment of the disclosure can be applied to the production process of the display panel production line product, and can also be applied to any other production process of the product production line, and the disclosure is not limited to this.
In some example embodiments, the production process data may be stored on an external storage device (e.g., a distributed storage device) or directly on a production process data correlation analysis device, which is not limiting of the present disclosure.
In the production process of the display panel, different stations correspond to different detection parameters, and the detection parameters can reflect the positions and reasons of the occurrence of the defects. The detection parameters are generally divided into characteristic parameters and test parameters, and the correlation analysis method for the production process data provided by the embodiment of the disclosure is used for mining the relation between the characteristic parameters and the test parameters by calculating the correlation between the characteristic parameters and the test parameters and converting the relation into quantized judgment indexes, so that when the test parameters are abnormal, process stations and related equipment under the characteristic parameters with larger correlation with the test parameters are quickly found. In the embodiment of the disclosure, the "abnormal occurrence of the test parameter" means that the detected value of the test parameter is not within a preset reasonable range, and the abnormal test parameter may cause a quality defect of the product, and the quality defect may cause a decrease or even rejection of the product, and may also cause that the product needs to be reworked or repaired. "the process station and the related equipment under the characteristic parameter with larger correlation with the test parameter" means that the participation of the process station and the related equipment has larger influence on the probability of occurrence of the abnormality of the test parameter of the product.
In some exemplary embodiments, the characteristic parameter may be a process parameter and the test parameter may be an electrical parameter. However, the present disclosure is not limited thereto.
In some exemplary embodiments, the process parameters include at least one of: parameters such as the overlay accuracy (Total Pitch, TP), line width (Criticlal dimension, CD), film Thickness (THK), and Layer-to-Layer overlay accuracy (OL) between the surface resistance (surface resistance, rs), thin film transistor (Thin Film Transistor, TFT), and Color Filter (CF).
In some exemplary embodiments, the electrical parameter includes at least one of: threshold voltage (Threshold voltage, VTH), mobility (MOB), operating current (ION), reverse off current (IOFF), and the like.
In some exemplary embodiments, the production process data may also include a plurality of historical parameter data, which may include, for example, parameters of production time, production plant, product model, product type, inspection site, process site, and the like. Illustratively, a history parameter data table is shown in Table 1, and a detection parameter table is shown in Table 2.
| No. | Glass ID | Step | Eqp | End Time | Product ID | Record Index |
| 1 | Glass ID1 | 655N | CC06 | time1 | BF06 | Index1 |
| 2 | Glass ID2 | 862G | BQ11 | time2 | AH76 | Index2 |
| 3 | Glass ID3 | 989F | DR02 | time3 | NU28 | Index3 |
| 4 | ... |
TABLE 1
| No. | Item | x | y | Value | Record Index |
| 1 | RS | 12.75 | 24.64 | 124 | Index1 |
| 2 | TP | 115.68 | 126.83 | 53 | Index2 |
| 3 | THK | 58.65 | 67.34 | 85 | Index3 |
| 4 | ... |
TABLE 2
In some exemplary embodiments, the production process data correlation analysis method further comprises:
Preprocessing the resume parameter data and the detection parameter data, wherein the preprocessing comprises removing blank data, removing repeated data, removing unnecessary fields and the like, so as to obtain effective resume parameter data and detection parameter data.
In some exemplary embodiments, the production process data correlation analysis method further comprises:
and carrying out data fusion on the resume parameter data and the detection parameter data to obtain basic data.
In this embodiment, when data fusion is performed on the history parameter data and the detection parameter data, the detection parameter data and the history parameter data (i.e. history data or production equipment information data, including panel ID, site name, etc.) are subjected to data fusion according to a specific field (e.g. record index number record_index), so as to obtain basic data as shown in table 3, and table 3 records related information of two sets of parameters VTH (threshold voltage, electrical parameter) and RS (surface resistance, process parameter) of Glass ID 1.
TABLE 3 Table 3
102, performing interpolation processing on the characteristic parameter data and/or the test parameter data to obtain n characteristic parameter data and n test parameter data, wherein the distance between a first position coordinate in the ith characteristic parameter data and a second position coordinate in the ith test parameter data is within a preset distance range, n is a natural number greater than 1, and i' is a natural number between 1 and n;
In the embodiment of the disclosure, since the detection devices and probes of the same product with different detection parameters are different, the distribution and the number of the detection points are also different, so that the same interpolation is required to be performed on the position before the correlation analysis is performed on the two parameters detected by the same product. Embodiments of the present disclosure may interpolate using simple Kriging interpolation or common Kriging interpolation.
In some exemplary embodiments, detecting whether n first position coordinates and n second position coordinates corresponding to the first position coordinates are present, wherein a distance between the first position coordinates in the ith characteristic parameter data and the second position coordinates in the ith test parameter data is within a preset distance range, and when n first position coordinates and n second position coordinates corresponding to the first position coordinates are not present, performing interpolation processing on the characteristic parameter data and/or the test parameter data to obtain n characteristic parameter data and n test parameter data.
In some exemplary embodiments, the preset distance range may be 10 -6 Mu m to 10 -3 um. In the embodiment of the present disclosure, the preset distance range may be set as required, which is not limited by the present disclosure.
In some exemplary embodiments, n may range from 20 to 180.
In the disclosed embodiment, interpolation processing is performed on the feature parameter data and/or the test parameter data to obtain n feature parameter data and n test parameter data, which means that the obtained production process data includes at least part of the feature parameter data and/or at least part of the test parameter data, and after the interpolation processing, another part of the feature parameter data and/or another part of the test parameter data, which are interpolated, and the obtained at least part of the feature parameter data and/or at least part of the test parameter data together form n feature parameter data and n test parameter data, that is, the n feature parameter data and the n test parameter data are not all obtained through the interpolation processing.
In some exemplary embodiments, when n pieces of characteristic parameter data exist but n pieces of test parameter data corresponding to the n pieces of characteristic parameter data do not exist, only interpolation processing is needed for the test parameter data, for example, at this time, n pieces of characteristic parameter data and n1 pieces of test parameter data may exist, and the n1 pieces of characteristic parameter data and the n1 pieces of test parameter data satisfy that a distance between a second position coordinate in the i 'th test parameter data and one first position coordinate in the characteristic parameter data is within a preset distance range, wherein 1.ltoreq.i'. Ltoreq.n 1< n; or, at this time, there may be n pieces of test parameter data and n pieces of feature parameter data, but only n1 pieces of feature parameter data and n1 pieces of test parameter data satisfy that the distance between the second position coordinate in the ith' piece of test parameter data and one first position coordinate in the feature parameter data is within the preset distance range, in which case, then, the test parameter data is interpolated according to the (n-n 1) pieces of feature parameter data, that is, the number of test parameter data to be interpolated is (n-n 1), and the second position coordinate in the interpolated test parameter data may be the first position coordinate in the (n-n 1) pieces of feature parameter data. When there are n pieces of characteristic parameter data and there are 0 pieces of test parameter data corresponding to the n pieces of characteristic parameter data (i.e., there are no pieces of test parameter data corresponding to any one piece of characteristic parameter data), we can also interpolate the test parameter data according to the n pieces of characteristic parameter data, and at this time, the number of test parameter data to be interpolated is n.
In some exemplary embodiments, when there are n pieces of test parameter data but there are no n pieces of feature parameter data corresponding to the n pieces of test parameter data, we need to perform interpolation processing only on the feature parameter data, for example, at this time, there may be n pieces of test parameter data and n1 pieces of feature parameter data, and the n1 pieces of test parameter data and the n1 pieces of feature parameter data satisfy that a distance between a first position coordinate in the i "th piece of feature parameter data and one second position coordinate in the test parameter data is within a preset distance range, where n1< n; or, at this time, there may be n pieces of test parameter data and n pieces of feature parameter data, but only n1 pieces of test parameter data and n1 pieces of feature parameter data satisfy that the distance between the first position coordinate in the i "th piece of feature parameter data and one second position coordinate in the test parameter data is within the preset distance range, in which case, then, the feature parameter data is subjected to interpolation processing according to the (n-n 1) pieces of test parameter data, that is, the number of feature parameter data to be interpolated is (n-n 1), and the first position coordinate in the interpolated feature parameter data is the second position coordinate in the (n-n 1) pieces of test parameter data. When there are n pieces of test parameter data and there are 0 pieces of feature parameter data corresponding to the n pieces of test parameter data (i.e., there are no feature parameter data corresponding to any one piece of test parameter data), we can also perform interpolation processing on the feature parameter data according to the n pieces of test parameter data, where the number of feature parameter data to be interpolated is n.
In some exemplary embodiments, when there are no n pieces of test parameter data and no n pieces of feature parameter data, we need to perform interpolation processing on the feature parameter data and the test parameter data, respectively, for example, at this time, n1 pieces of feature parameter data and n2 pieces of test parameter data may exist, and n3 pieces of test parameter data and n3 pieces of feature parameter data satisfy that a distance between a first position coordinate in the ith' "piece of feature parameter data and one second position coordinate in the test parameter data is within a preset distance range; in this case, the interpolation processing is performed on the feature parameter data from (n-n 1) pieces of test parameter data, the interpolation processing is performed on the test parameter data from (n-n 2) pieces of feature parameter data, that is, the number of feature parameter data to be interpolated is (n-n 1), the number of test parameter data to be interpolated is (n-n 2), and the first position coordinates in the interpolated feature parameter data may be the second position coordinates in the (n-n 1) pieces of test parameter data, and the second position coordinates in the interpolated test parameter data may be the first position coordinates in the (n-n 2) pieces of feature parameter data, 1.ltoreq.i '".ltoreq.n3 < n1< n, 1.ltoreq.i'" is.ltoreq.n3 < n2< n.
In some exemplary embodiments, when there are n pieces of feature parameter data and n pieces of test parameter data, and the distance between the first position coordinate in the i 'th piece of the feature parameter data and the second position coordinate in the i' th piece of the test parameter data is within the preset distance range, the process proceeds directly to step 103.
In some exemplary embodiments, the characteristic parameter data or test parameter data comprises m record values { z (x 1 ，y 1 )，z(x 2 ，y 2 )，z(x 3 ，y 3 )，...，z(x m-1 ，y m-1 )，z(x m ，y m ) Wherein m is a natural number less than or equal to n, (x) i ，y i ) Representing the first or second position coordinates, z (x i ，y i ) Representing a characteristic parameter value or a test parameter value, i being a natural number between 1 and m, the first or second position coordinates to be interpolated being (x 0 ，y 0 ) Pair (x) 0 ，y 0 ) The interpolation processing is carried out on the characteristic parameter data or the test parameter data, and the method comprises the following steps:
calculating the distance and half variance between the m recorded values;
drawing the calculated distance and the half variance into a scatter diagram, and searching for the relation between the fitting distance of the fitting curve and the half variance to obtain a functional relation;
from the obtained functional relation, a calculation (x 0 ，y 0 ) To (x) i ，y i ) Half variance between;
substituting the calculated half variance into a pre-constructed Lagrangian equation set to solve the optimal weight coefficient;
Substituting the solved optimal weight coefficient into an interpolation calculation formula to obtain (x) 0 ，y 0 ) An estimate of the position.
In some exemplary embodiments, the characteristic parameter data or test parameter data comprises m record values { z (x 1 ，y 1 )，z(x 2 ，y 2 )，z(x 3 ，y 3 )，...，z(x m-1 ，y m-1 )，z(x m ，y m ) For (x) 0 ，y 0 ) The interpolation processing is carried out on the characteristic parameter data or the test parameter data, and the method comprises the following steps:
calculating the distance between m recorded valuesAnd half variance
To calculate d ij And r ij Drawing a scatter diagram, and searching for a relation between fitting distance of a fitting curve and half variance to obtain a functional relation r=r (d);
from the obtained functional relation, a calculation (x 0 ，y 0 ) To all known points (x i ，y i ) Half variance r of (2) i0 ；
According toSolving for optimal weight coefficient lambda i Wherein, phi is Lagrangian multiplier;
according to the following formula:obtaining the unknown point z 0 Estimate of (2)Where c is the average of the m recorded values.
In other exemplary embodiments, the characteristic parameter data or test parameter data comprises m recorded values { z (x 1 ，y 1 )，z(x 2 ，y 2 )，z(x 3 ，y 3 )，...，z(x m-1 ，y m-1 )，z(x m ，y m ) For (x) 0 ，y 0 ) The interpolation processing is carried out on the characteristic parameter data or the test parameter data, and the method comprises the following steps:
calculating the distance between m recorded valuesAnd half variance
To calculate d ij And r ij Drawing a scatter diagram, and searching for a relation between fitting distance of a fitting curve and half variance to obtain a functional relation r=r (d);
From the obtained functional relation, a calculation (x 0 ，y 0 ) To all known points (x i ，y i ) Half variance r of (2) i0 ；
According toSolving for optimal weight coefficient lambda i ；
According to the following formula:obtaining the unknown point z 0 Estimate of (2)
Simple Kriging interpolation generates an advanced statistical process of the estimated surface from a set of discrete points with z values, assuming that the distance or direction between the sampled points can reflect the spatial correlation available to account for surface variations, a mathematical function can be fitted to a specified number of points or all points within a specified radius to determine the output value for each location. Compared with the common Kriging interpolation method, the simple Kriging interpolation method can meet the unbiased condition without additional calculation, and the simple Kriging interpolation method has the following calculation formula:
wherein,is a point of%x 0 ，y 0 ) Estimated value of lambda i Is the weight coefficient, z i Is (x) i ，y i ) The attribute value at C is the mathematical expectation of the attribute value, i.e., E (Z) =c.
Here the weight coefficient lambda i Can satisfy the point (x) 0 ，y 0 ) Estimated value atAnd the true value z 0 A set of optimal coefficients with the smallest difference, i.e
(x 0 ，y 0 ) Representing a position on the display panel having a corresponding parameter estimateNext, for (x) 0 ，y 0 ) Estimate of (2) whereThe calculation is carried out, and the calculation steps are as follows:
(1) For recorded data, the distance is calculated pairwiseAnd half varianceWhere r is ij Express similarity of attributes, d ij Expressing the similarity of the space, i.eBy r ij It can be seen that the process parameter value and the electrical parameter value have similar degrees, by d ij The similarity of the process parameter points and the electrical parameter points in positions can be seen. Kriging simple interpolation hypothesis r ij And d ij There is a functional relationship and to confirm this we first apply to the record dataset { z (x 1 ，y 1 )，z(x 2 ，y 2 )，z(x 3 ，y 3 )，...，z(x m-1 ，y m-1 )，z(x m ，y m ) D of arbitrary two points is calculated ij And r ij At this time, m is obtained 2 m 2 And (d) ij ，r ij ) Is a data pair of (a) and (b).
(2) And searching a relation between the fitting distance of the fitting curve and the half variance, so that the corresponding half variance can be calculated according to any distance. Drawing all d and r into a scatter diagram, searching a best fitting curve to fit the relationship of d and r to obtain a functional relation r=r (d), and determining the point (x) of any two points (x i ，,y i ) And (x) j ，,y j ) Firstly, calculating the parameter distance d ij Then, the half variance r of the two points can be obtained according to the obtained functional relation ij 。
(3) For unknown parameter point attribute value z 0 Calculate it to all known points z i Half variance r of (2) i0 。
(4) According toSolving for optimal weight coefficient lambda i Order-makingOur goal is to find a set of lambda that minimizes J i And J is lambda i Thus directly comparing J to lambda i The partial derivative is calculated to be 0. I.e.
Written in matrix form
Wherein phi is Lagrangian multiplier and inverting the matrix can be solved to obtain the optimal weight coefficient lambda i 。
(5) Weighting and summing the attribute values of the known points by using the optimal coefficient to obtain an unknown point z 0 Is used for the estimation of the estimated value of (a).
Compared with simple Kriging interpolation, the general Kriging interpolation calculation formula is as follows:
wherein,λ i and z i As with the simple Kriging interpolation definition, in the normal Kriging interpolation, the attribute values of the unknown points are considered to be a weighted summation of the attribute values of the known points.
The interpolation form of the common Kriging interpolation method is different from the simple Kriging interpolation, but the unknown point z is solved 0 The estimation method of (2) is the same, the important difference is that the common Kriging interpolation method needs constraint conditionsCompared with a simple Kriging interpolation step, the common Kriging interpolation is written in a matrix form as follows:
in some exemplary embodiments, the product may be a display panel.
In some exemplary embodiments, the method further comprises: dividing the display panel into a plurality of grid areas, and obtaining no more than one characteristic parameter data and one test parameter data in each grid area through interpolation processing.
For example, assuming that the size of the display panel is 3370mm×2940mm, the number of detection points of different detection parameters is generally between 20 and 180, so that interpolation prediction is performed according to 16×14 grid points while considering calculation speed and accuracy, that is, the display panel is divided into 16 rows and 14 columns, a grid area is formed at the position where the rows and columns intersect, and interpolation processing is performed according to whether feature parameter data and test parameter data exist in a certain grid area, for example, when 0 feature parameter data and 0 test parameter data exist in the grid area or when one feature parameter data and one test parameter data exist in the grid area, the interpolation processing is not needed for the grid; when 1 characteristic parameter data and 0 test parameter data exist in the grid area or 0 characteristic parameter data and one test parameter data exist in the grid area, interpolation processing is carried out on the grid to obtain 1 characteristic parameter data and 1 test parameter data. In actual use, the size of each grid region may not be exactly equal, and the partial grid regions may be adaptively shifted according to the coordinates of the detection points, so long as the distance between the first position coordinates in the feature parameter data and the second position coordinates in the test parameter data in each grid region is within a preset distance range.
In the embodiment of the disclosure, the display panel is divided into a plurality of grid areas, so that matrix operation can be better performed during subsequent correlation analysis, and further, the calculation speed and the calculation precision are improved. Of course, the embodiments of the present disclosure are not limited thereto, and in some exemplary embodiments, the interpolation process may be directly performed according to the obtained feature parameter data and test parameter data in the production process data without dividing the grid area.
And 103, calculating the correlation between the characteristic parameters and the test parameters according to the n characteristic parameter data and the n test parameter data.
In some exemplary embodiments, the correlation between the characteristic parameter and the test parameter is calculated by a Spearman correlation analysis method.
In the embodiment of the disclosure, n pieces of characteristic parameter data and n pieces of test parameter data with identical or close position coordinates (i.e., within a preset distance range) are obtained according to the interpolated data, and when a certain parameter is abnormal from a service perspective, the root cause of the previous-stage process is required to be checked for verification test and maintenance. The reasons of the defects can be rapidly positioned through the analysis of the correlation among the parameters, and the efficiency of business personnel is improved. Accordingly, embodiments of the present disclosure quantify the correlation between electrical parameters and process parameters by Spearman correlation analysis methods, giving business personnel multiple references.
The Spearman correlation analysis method can use Spearman correlation coefficients to study as long as the observed values of two variables are paired rating data or rating data obtained by converting continuous variable observed data, regardless of the overall distribution form of the two variables and the size of sample capacity. Defining X and Y as two sets of data, the Spearman correlation coefficient is calculated as follows:
wherein d i′ Is X i′ And Y i′ The level difference, i', between 1 and n is a natural number.
For example, X may be expressed as GLASS ID1 interpolatedN characteristic parameter values (X 1 ，X 2 ，...，X n ) Y may be expressed as n test parameter values (Y 1 ，Y 2 ，...，Y n ) The method of calculating the level difference is shown in table 4.
TABLE 4 Table 4
From the level difference between the two sets of parameters, the Spearman correlation coefficient ρ can be obtained S ，ρ S Between-1 and 1, ρ S The closer the absolute value is to 1, the greater the correlation, and the two parameters exhibit strong correlation. The speedy positioning of the service personnel to the bad root cause of the preamble process is facilitated according to the Spearman correlation coefficient, so that verification test and maintenance can be performed.
According to the production process data correlation analysis method, the correlation between the characteristic parameters and the test parameters is calculated, so that when the characteristic parameters are abnormal, the test parameters with larger correlation with the characteristic parameters can be quickly found, and the process stations and related equipment causing the characteristic parameters to be abnormal are found according to the test parameters, so that workers can repair and adjust the equipment in time, the productivity of product manufacturing can be increased, and the economic benefit of enterprises is improved.
In some exemplary embodiments, the method further comprises: a first input is received from a user, and a range of acquired production process data is determined based on the first input.
The first input includes at least one of:
production time, production factory, product type, product model, inspection station, electrical parameters, process station, process parameters.
In some exemplary embodiments, the method further comprises: and outputting a correlation trend graph of the electrical parameters and the process parameters in the first input in a preset time period. By way of example, the preset time period may be within the last three months, the last month, the last week, etc., which is not limiting to the present disclosure.
The method for analyzing the correlation of the production process data provided by the embodiment of the disclosure can be used for calculating the correlation of a certain electrical parameter and a certain process parameter in a preset time period, can also be used for calculating the correlation of a certain electrical parameter and any one of a plurality of process parameters in a preset time period or the correlation of any one of a plurality of electrical parameters and any one of a plurality of process parameters in a preset time period, or can also be used for calculating the correlation of any one of a plurality of electrical parameters and any one of a plurality of process parameters, and is not limited in this disclosure.
Fig. 2 is a schematic diagram of a user interface according to an exemplary embodiment of the present disclosure. As shown in fig. 2, in the input box of the user interface, 2020-11-26 is a production date, 61L is a product type value (illustratively, a product type may include a plurality of product sizes, etc.), TPCN is a product type value (illustratively, a product type may include a mass production product, a test product, etc.), A8A1EPN is a detection site name, tft2_vth_sat is an electrical parameter, a320TKN is a process site name, and 6k_thk1l is a process parameter. The parameter level defect detection is real-time calculation, and a user can screen according to the required conditions of production time, production factories, product models, detection parameters and the like, so that correlation among the detection parameters is calculated by one key.
As shown in fig. 2, the output a8a1epn-tft2_vth_sat trend graph and the a320TKN-6k_thk1l trend graph each use a Box graph to show the data of the last three months, four weeks and seven days, and the Box graph (Box-plot) is also called a Box whisker graph, a Box graph or a Box line graph, which is a statistical graph used as a data for displaying a set of data dispersion conditions. The method is mainly used for reflecting the characteristics of original data distribution and can also be used for comparing multiple groups of data distribution characteristics. The output correlation R values of the A8A1EPN-TFT2_VTH_SAT vs A320TKN-6K_THK1L and the process parameter mean trend graph are sequentially connected with each calculated value by adopting line segments, and the correlation R values of the A8A1EPN-TFT2_VTH_SAT and the A320TKN-6K_THK1L are between-0.7 and-0.8 within the range of three months recently, so that the correlation R values of the A8A1EPN-TFTV2_VTH_SAT and the A320TKN-6K_THK1L are relatively large.
In some exemplary embodiments, the production process data correlation analysis method further comprises:
sequentially arranging n characteristic parameter values and n test parameter values;
calculating the rank sum of n characteristic parameter values and n test parameter values in the arrangement;
calculating a sample statistic according to the calculated rank sum:
calculating a significance level by using a probability density function of chi-square distribution and the calculated sample statistics;
verifying whether a correlation calculation result between the characteristic parameter and the test parameter is valid according to the calculated significance level;
and outputting a correlation calculation result when the correlation calculation result is verified to be valid.
Illustratively, the method for analyzing the correlation of the production process data further comprises:
arranging n characteristic parameter values and n test parameter values in a column in increasing order;
calculating R x And R is y ，R x And R is y Representing the sum of the n characteristic parameter values and the n test parameter values in the rank of the arrangement, respectively;
according to R x 、R y And calculating a statistic H according to the following formula:
wherein n= 2*n;
the significance level pvalue is calculated by using the probability density function of chi-square distribution and the calculated statistic H, and the calculation formula is as follows:
wherein,for the gamma distribution, k=2,
And verifying whether the correlation between the calculated characteristic parameter and the test parameter is valid or not according to the calculated significance level pvalue.
In embodiments of the present disclosure, the Spearman correlation coefficient ρ is verified from the calculated significance level pvalue value S Is effective in the following. If the significance level is pvalue<0.05, spearman correlation coefficient ρ S The closer the absolute value is to 1, the greater the correlation, and the two parameters exhibit strong correlation. If the significance level pvalue is greater than or equal to 0.05, the spearman correlation coefficient ρ S The closer the absolute value is to 0, the less the correlation, and the two parameters exhibit weak correlation.
In some exemplary embodiments, as shown in fig. 3, the method for analyzing correlation of production process data provided by the embodiments of the present disclosure includes the following steps:
step one, display panel production process data (for example, production process data in an Array substrate (Array) stage) including history data and detection parameter data is acquired.
And secondly, preprocessing the history data and the detection parameter data, including removing blank data, removing repeated data, removing unnecessary fields and the like, so as to obtain effective data.
And thirdly, carrying out data fusion on the history data obtained in the second step and the detection parameter data according to a specific field (such as a record index record_index) to obtain basic data, wherein the basic data is shown in the table 3.
And step four, interpolation is carried out on the data fused in the step three, and as the detection equipment and the probes of different detection parameters of the same substrate are different, the distribution and the number of detection points are also different, so that interpolation is needed before correlation analysis is carried out on two parameters (process parameters and electrical parameters) detected by the same substrate to ensure the same position, and the interpolation can be carried out by using a simple Kriging interpolation method or a common Kriging interpolation method by way of example.
And fifthly, obtaining parameters with the same coordinate points according to the data obtained by interpolation in the step four, and checking the root cause of the previous process when a certain parameter is abnormal from the service perspective so as to carry out verification test and maintenance. The reasons of the defects can be rapidly positioned through the analysis of the correlation among the parameters, and the efficiency of business personnel is improved. Illustratively, the correlation between the electrical parameters and the process parameters can be quantified by a Spearman correlation analysis method, giving business personnel various references.
Step six, verifying the validity of the correlation coefficient by using hypothesis testing. The validity of the correlation coefficient can be verified by the Kruskal-walis (Kruskal-walis) test, for example. The Kruskal-Wallis test is a non-parametric method of testing whether two or more samples are from the same probability distribution, without knowing the distribution and overall parameters of the raw data.
The Spearman correlation coefficient ρ can be obtained from the level differences of the process parameters and the electrical parameters S ，ρ S Between-1 and 1, ρ is verified according to the significance level pvalue value S Validity, if pvalue<0.05，ρ S The closer the absolute value is to 1, the greater the correlation, and the two parameters exhibit strong correlation. The correlation coefficients according to pvaue and Spearman are convenient for business personnel to quickly locate the bad root cause of the preamble process so as to carry out verification test and maintenance.
According to the production process data correlation analysis method, the site equipment which is related to the method and causes the parameter abnormality is closer to the site equipment which causes the parameter abnormality in the actual display panel production process, the problem of lower production efficiency caused by the abnormal parameter problem in the display panel production process is avoided, and an intuitive, scientific and convenient detection parameter correlation analysis method is provided for enterprises and clients.
The embodiments of the present disclosure also provide a production process data dependency analysis apparatus, which may comprise a processor and a memory storing a computer program executable on the processor, which processor, when executing the computer program, implements the steps of the production process data dependency analysis method according to any one of the preceding claims in the present disclosure.
As shown in fig. 4, in one example, a production process data correlation analysis device may include: the device comprises a processor 910, a memory 920, a bus system 930 and a transceiver 940, wherein the processor 910, the memory 920 and the transceiver 940 are connected through the bus system 930, the memory 920 is used for storing instructions, and the processor 910 is used for executing the instructions stored in the memory 920 to control the transceiver 940 to transmit and receive signals. In particular, transceiver 940 may obtain a first input of a user from a user interface under control of processor 910; acquiring production process data according to a first input of a user, wherein the production process data comprises a plurality of characteristic parameter data and a plurality of test parameter data, the characteristic parameter data comprises a first position coordinate and a characteristic parameter value corresponding to the first position coordinate, and the test parameter data comprises a second position coordinate and a test parameter value corresponding to the second position coordinate; interpolation processing is carried out on the characteristic parameter data and/or the test parameter data to obtain n characteristic parameter data and n test parameter data, the distance between a first position coordinate in the ith characteristic parameter data and a second position coordinate in the ith test parameter data is within a preset distance range, n is a natural number greater than 1, and i is a natural number between 1 and n; and calculating the correlation between the characteristic parameters and the test parameters according to the n characteristic parameter data and the n test parameter data, and outputting the calculated correlation between the characteristic parameters and the test parameters to a user interface through a transceiver.
It is to be appreciated that the processor 910 may be a central processing unit (Central Processing Unit, CPU), and the processor 910 may also be other general purpose processors, digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
Memory 920 may include read-only memory and random access memory and provide instructions and data to processor 910. A portion of memory 920 may also include non-volatile random access memory. For example, the memory 920 may also store information of a device type.
The bus system 930 may include a power bus, a control bus, a status signal bus, and the like in addition to a data bus. The various buses are labeled as bus system 930 in fig. 4 for clarity of illustration.
In implementation, the processing performed by the processing device may be performed by integrated logic circuits of hardware in processor 910 or by instructions in the form of software. That is, the method steps of the embodiments of the present disclosure may be embodied as hardware processor execution or as a combination of hardware and software modules in a processor. The software modules may be located in random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, and other storage media. The storage medium is located in the memory 920, and the processor 910 reads information in the memory 920 and performs the steps of the above method in combination with its hardware. To avoid repetition, a detailed description is not provided herein.
The embodiment of the disclosure further provides a computer storage medium, where executable instructions are stored, where the executable instructions when executed by a processor may implement the method for analyzing correlation of production process data provided in any of the foregoing embodiments of the disclosure, where the method for analyzing correlation of production process data may obtain production process data, where the production process data includes a plurality of feature parameter data and a plurality of test parameter data, where the feature parameter data includes a first location coordinate and a feature parameter value corresponding to the first location coordinate, and the test parameter data includes a second location coordinate and a test parameter value corresponding to the second location coordinate; interpolation processing is carried out on the characteristic parameter data and/or the test parameter data to obtain n characteristic parameter data and n test parameter data, the distance between a first position coordinate in the ith characteristic parameter data and a second position coordinate in the ith test parameter data is within a preset distance range, n is a natural number greater than 1, and i is a natural number between 1 and n; according to the n characteristic parameter data and the n test parameter data, the correlation between the characteristic parameters and the test parameters is calculated, so that when the characteristic parameters are abnormal, the test parameters with larger correlation with the characteristic parameters can be quickly found, and a process station and related equipment causing the characteristic parameters to be abnormal can be found according to the test parameters, so that workers can repair and adjust the equipment in time, the productivity of manufacturing the display panel can be increased finally, and the economic benefit of enterprises is improved. The method for performing executable instruction-driven process data correlation analysis is substantially the same as the process data correlation analysis method provided in the above embodiment of the present disclosure, and will not be described herein.
Those of ordinary skill in the art will appreciate that all or some of the steps, systems, functional modules/units in the apparatus, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. In a hardware implementation, the division between the functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed cooperatively by several physical components. Some or all of the components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Furthermore, as is well known to those of ordinary skill in the art, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.
While the embodiments disclosed in this disclosure are described above, the embodiments are only used for facilitating understanding of the disclosure, and are not intended to limit the present invention. Any person skilled in the art will recognize that any modifications and variations can be made in the form and detail of the present disclosure without departing from the spirit and scope of the disclosure, which is defined by the appended claims.
Claims (16)
- A method of process data correlation analysis, comprising:acquiring production process data, wherein the production process data comprises a plurality of characteristic parameter data and a plurality of test parameter data, the characteristic parameter data comprises a first position coordinate and a characteristic parameter value corresponding to the first position coordinate, and the test parameter data comprises a second position coordinate and a test parameter value corresponding to the second position coordinate;interpolation processing is carried out on the characteristic parameter data and/or the test parameter data to obtain n characteristic parameter data and n test parameter data, wherein the distance between a first position coordinate in the ith characteristic parameter data and a second position coordinate in the ith test parameter data is within a preset distance range, n is a natural number greater than 1, and i' is a natural number between 1 and n;And calculating the correlation between the characteristic parameters and the test parameters according to the n characteristic parameter data and the n test parameter data.
- The production process data correlation analysis method according to claim 1, wherein interpolation processing is performed on the characteristic parameter data and/or the test parameter data to obtain n pieces of the characteristic parameter data and n pieces of the characteristic parameter data, including one of:when n pieces of characteristic parameter data exist and n1 pieces of test parameter data exist, and the distance between a second position coordinate in the ith piece of test parameter data and one first position coordinate in the characteristic parameter data is within a preset distance range, carrying out interpolation processing on the test parameter data according to the (n-n 1) pieces of characteristic parameter data, wherein the second position coordinate in the interpolated test parameter data is the first position coordinate in the (n-n 1) pieces of characteristic parameter data, and i' is more than or equal to 1 and less than or equal to n1< n;when n pieces of test parameter data exist and n1 pieces of characteristic parameter data exist, and the distance between a first position coordinate in the ith' piece of characteristic parameter data and one second position coordinate in the test parameter data is within a preset distance range, carrying out interpolation processing on the characteristic parameter data according to the (n-n 1) pieces of test parameter data, wherein the first position coordinate in the interpolated characteristic parameter data is the second position coordinate in the (n-n 1) pieces of test parameter data;When n1 pieces of the characteristic parameter data and n2 pieces of the test parameter data exist, and the presence of n3 pieces of the characteristic parameter data and n3 pieces of the test parameter data satisfy that a distance between a first position coordinate in the ith' piece of the characteristic parameter data and one second position coordinate in the test parameter data is within a preset distance range, interpolation processing is performed on the characteristic parameter data according to (n-n 1) pieces of the test parameter data, interpolation processing is performed on the test parameter data according to (n-n 2) pieces of the characteristic parameter data, and a first position coordinate in the interpolated characteristic parameter data is a second position coordinate in the (n-n 1) pieces of the test parameter data, and a second position coordinate in the interpolated test parameter data is a first position coordinate in the (n-n 2) pieces of the characteristic parameter data, wherein 1 i ' < n1< n3< n,1 is less than or equal to i ' < n2< n.
- The production process data correlation analysis method according to claim 1, wherein the characteristic parameter data or the test parameter data in the obtained production process data includes m record values: { z (x) 1 ，y 1 )，z(x 2 ，y 2 )，z(x 3 ，y 3 )，…，z(x m-1 ，y m-1 )，z(x m ，y m ) Wherein m is a natural number less than or equal to n, (x) i ，y i ) Representing the first or second position coordinates, z (x i ，y i ) Representing a characteristic parameter value or a test parameter value, i being a natural number between 1 and m, the first or second position coordinates to be interpolated being (x 0 ，y 0 ) Pair (x) 0 ，y 0 ) The interpolation processing is carried out on the characteristic parameter data or the test parameter data, and the method comprises the following steps:calculating the distance and half variance between the m recorded values;drawing the calculated distance and the half variance into a scatter diagram, and searching for the relation between the fitting distance of the fitting curve and the half variance to obtain a functional relation;from the obtained functional relation, a calculation (x 0 ，y 0 ) To (x) i ，y i ) Half variance between;substituting the calculated half variance into a pre-constructed Lagrangian equation set to solve the optimal weight coefficient;substituting the solved optimal weight coefficient into an interpolation calculation formula to obtain (x) 0 ，y 0 ) An estimate of the position.
- A production process data correlation analysis method according to claim 3, wherein the set of pre-constructed lagrangian equations is:wherein lambda is i For the optimal weight coefficient, phi is the Lagrangian multiplier, r i0 Is (x) 0 ，y 0 ) To (x) i ，y i ) Half variance of r ij Is (x) i ，y i ) And (x) j ，y j ) Half variance between;the interpolation calculation formula is as follows:z i is (x) i ，y i ) A characteristic parameter value or a test parameter value, c is the mean value of the m recorded values,is (x) 0 ，y 0 ) An estimate of the position.
- A production process data correlation analysis method according to claim 3, wherein the set of pre-constructed lagrangian equations is:wherein lambda is i For optimal weight coefficients, φ is the Lagrangian multiplier, r i0 Is (x) 0 ，y 0 ) To (x) i ，y i ) Half variance of r ij Is (x) i ，y i ) And (x) j ，y j ) Half variance between;the interpolation calculation formula is as follows:z i is (x) i ，y i ) At characteristic parameter values or test parameter values,is (x) 0 ，y 0 ) An estimate of the position.
- The production process data correlation analysis method according to claim 1, wherein (X) 1 ，X 2 ，…，X n ) For n of said characteristic parameter values, (Y) 1 ，Y 2 ，…，Y n ) For n values of the test parameter, the correlation between the characteristic parameter and the test parameter is calculated by the following formula:wherein ρ is S Is the correlation coefficient, d i′ Is X i′ And Y i′ Level differences between.
- The production process data correlation analysis method according to claim 1, the method further comprising:sequentially arranging n characteristic parameter values and n test parameter values;calculating the rank sum of n characteristic parameter values and n test parameter values in the arrangement;calculating a sample statistic according to the calculated rank sum:Calculating a significance level by using a probability density function of chi-square distribution and the calculated sample statistics;verifying whether a correlation calculation result between the characteristic parameter and the test parameter is valid according to the calculated significance level;and outputting the correlation calculation result when the correlation calculation result is verified to be effective.
- The production process data correlation analysis method according to claim 7, wherein,the sample statistic is calculated according to the following formula:where H is the sample statistic, n= 2*n, R x And R is y Representing the rank sum of the n said characteristic parameter values and the n said test parameter values in the permutation, respectively;the significance level was calculated according to the following formula:wherein pvalue is the level of significance,for the gamma distribution, k=2,
- the method of claim 1, wherein the characteristic parameter is a process parameter and the test parameter is an electrical parameter.
- The production process data correlation analysis method of claim 9, wherein the process parameters include at least one of: the surface resistance, the involution precision between the thin film transistor and the color film, the line width, the film thickness and the interlayer registration precision; the electrical parameter includes at least one of: threshold voltage, mobility, operating current and reverse off current.
- The production process data correlation analysis method according to claim 9, wherein the production process data further includes a plurality of history parameter data including at least one of: production time, production plant, product model, product type, inspection station, and process station.
- The production process data correlation analysis method of claim 11, the method further comprising at least one of:preprocessing the production process data, wherein the preprocessing comprises void removal data, duplicate removal data and useless field removal;and carrying out data fusion on the characteristic parameter data, the test parameter data and the resume parameter data.
- The production process data correlation analysis method of claim 12, the method further comprising: receiving a first input of a user, and determining the range of the acquired production process data according to the first input; the first input includes at least one of:production time, production factory, product type, product model, inspection station, electrical parameters, process station, process parameters.
- The production process data correlation analysis method of claim 13, the method further comprising: and outputting a correlation trend graph of the electrical parameters and the process parameters in the first input in a preset time period.
- A production process data correlation analysis device comprising a memory; and a processor coupled to the memory, the processor configured to perform the steps of the production process data correlation analysis method of any one of claims 1 to 14 based on instructions stored in the memory.
- A storage medium having stored thereon a computer program which, when executed by a processor, implements the production process data correlation analysis method of any one of claims 1 to 14.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2022/084141 WO2023184252A1 (en) | 2022-03-30 | 2022-03-30 | Production process data correlation analysis method, device, and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117280286A true CN117280286A (en) | 2023-12-22 |
Family
ID=88198511
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202280000648.1A Pending CN117280286A (en) | 2022-03-30 | 2022-03-30 | Method, apparatus and storage medium for analyzing correlation of production process data |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN117280286A (en) |
| WO (1) | WO2023184252A1 (en) |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014160760A2 (en) * | 2013-03-27 | 2014-10-02 | Abb Technology Ag | Method and apparatus for using post assembly process interaction signatures to detect assembly failures |
| CN107340365B (en) * | 2017-06-19 | 2019-04-26 | 中国科学院南京地理与湖泊研究所 | A kind of three-dimensional monitoring and data digging system and method towards lake blue algae disaster |
| CN108009730B (en) * | 2017-12-05 | 2021-08-24 | 河海大学常州校区 | Photovoltaic power station system health state analysis method |
| DE102018106563B4 (en) * | 2018-03-20 | 2019-11-14 | Vega Grieshaber Kg | Method for measured value compensation for capacitive pressure cells |
| CN110826715A (en) * | 2019-11-08 | 2020-02-21 | 江西理工大学 | Improved space-time Kriging interpolation algorithm for encrypting slope monitoring data |
-
2022
- 2022-03-30 CN CN202280000648.1A patent/CN117280286A/en active Pending
- 2022-03-30 WO PCT/CN2022/084141 patent/WO2023184252A1/en unknown
Also Published As
| Publication number | Publication date |
|---|---|
| WO2023184252A1 (en) | 2023-10-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11636292B2 (en) | Dynamic outlier bias reduction system and method | |
| JP2020126601A (en) | Fault detecting method and fault detecting device | |
| Zhang et al. | Monitoring wafers’ geometric quality using an additive Gaussian process model | |
| US10948903B2 (en) | System and method that consider tool interaction effects for identifying root causes of yield loss | |
| US11257001B2 (en) | Prediction model enhancement | |
| TWI783965B (en) | Method and system of circuit yield analysis for evaluating rare failure events | |
| US20210364999A1 (en) | System and method for analyzing cause of product defect, computer readable medium | |
| CN113946986B (en) | Method and device for evaluating average time before product failure based on accelerated degradation test | |
| US10935962B2 (en) | System and method for identifying root causes of yield loss | |
| CN116128839A (en) | Wafer defect identification method, device, electronic equipment and storage medium | |
| WO2022160574A1 (en) | Failure analysis method, computer device and storage medium | |
| CN115841046B (en) | Accelerated degradation test data processing method and device based on wiener process | |
| Sogandi et al. | Monotonic change point estimation of generalized linear model-based regression profiles | |
| CN114519519A (en) | Method, device and medium for assessing enterprise default risk based on GBDT algorithm and logistic regression model | |
| Niaki et al. | Monotonic change-point estimation of multivariate Poisson processes using a multi-attribute control chart and MLE | |
| CN117280286A (en) | Method, apparatus and storage medium for analyzing correlation of production process data | |
| Woo et al. | Modern definitions in reliability engineering | |
| Nduka | EM-based algorithms for autoregressive models with t-distributed innovations | |
| Hsu et al. | Data mining for yield enhancement in TFT-LCD manufacturing: an empirical study | |
| Lin et al. | A Kernel-Density based Semi-Parametric stochastic degradation model with dependent increments | |
| CN115222114A (en) | Automobile part assembly index value prediction method, terminal device and storage medium | |
| US20210294204A1 (en) | Method of measuring critical dimension of a three-dimensional structure and apparatus for measuring the same | |
| Nazari et al. | An improved algebraic geometric solution to the identification of switched ARX models with noise | |
| CN109146840B (en) | Method and device for evaluating data geometric positioning precision based on Gaussian probability statistics | |
| CN114766023B (en) | Data processing method, device and system and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |