WO2022052548A1 - 数据分析方法及装置、电子设备、存储介质 - Google Patents

数据分析方法及装置、电子设备、存储介质 Download PDF

Info

Publication number
WO2022052548A1
WO2022052548A1 PCT/CN2021/100431 CN2021100431W WO2022052548A1 WO 2022052548 A1 WO2022052548 A1 WO 2022052548A1 CN 2021100431 W CN2021100431 W CN 2021100431W WO 2022052548 A1 WO2022052548 A1 WO 2022052548A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
stacking
measurement
yield problem
target
Prior art date
Application number
PCT/CN2021/100431
Other languages
English (en)
French (fr)
Inventor
李玉坤
Original Assignee
长鑫存储技术有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 长鑫存储技术有限公司 filed Critical 长鑫存储技术有限公司
Priority to US17/447,410 priority Critical patent/US20220077005A1/en
Publication of WO2022052548A1 publication Critical patent/WO2022052548A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30148—Semiconductor; IC; Wafer

Definitions

  • the present disclosure relates to the field of semiconductor technology, and in particular, to a data analysis method, a data analysis device, an electronic device, and a computer-readable storage medium.
  • semiconductor device As a kind of electronic device with conductivity between good conductors and insulators, uses the special electrical properties of semiconductor materials to complete specific functions, because it can be used to generate, control, Various functions such as receiving, transforming, amplifying signals and performing energy conversion have been paid more and more attention by people.
  • Wafer refers to the silicon wafers used to make silicon semiconductor integrated circuits. Due to the influence of the process of processing wafers, the yield of the resulting wafers is sometimes high and sometimes low. The yield problem is analyzed, so that the process can be adjusted according to the analysis result of the yield problem, and the wafer yield can be improved.
  • the purpose of the embodiments of the present disclosure is to provide a data analysis method, a data analysis device, an electronic device, and a computer-readable storage medium, thereby at least to a certain extent overcome the relatively low efficiency of screening target measurement parameters that lead to the problem of target yield in related solutions. low, low accuracy problems.
  • a data analysis method comprising: acquiring a target yield problem stacking pattern corresponding to a wafer group having a target yield problem, and acquiring data of the wafer group under different types of tests Measure the data stacking pattern; perform graphic matching between the target yield problem stacking pattern and each of the measurement data stacking patterns, and obtain the matching degree corresponding to the target yield problem stacking pattern and each of the measurement data stacking patterns data; calculate the correlation data corresponding to each of the measurement data stacking graphs and the target yield problem stacking graphs; perform weighted calculation on the matching degree data and the correlation data, and determine according to the weighted calculation results that the Target measurement parameters for the target yield problem.
  • a data analysis device comprising: a pattern acquisition module, configured to acquire a target yield problem stacking pattern corresponding to a wafer group having a target yield problem, and acquire the wafer group Measurement data stacking graphs under different types of tests; a correlation degree calculation module, used to calculate the matching degree data and correlation data corresponding to the target yield problem stacking graph and each of the measurement data stacking graphs, and for all The matching degree data and the correlation data are weighted and calculated to obtain a weighted calculation result; a target measurement parameter analysis module is configured to analyze the weighted calculation result to determine the target measurement parameter that causes the target yield problem.
  • an electronic device comprising: a processor; and a memory, the memory having computer-readable instructions stored thereon, the computer-readable instructions when executed by the processor implement a first The data analysis method described in the aspect.
  • a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, implements the data analysis method described in the first aspect.
  • FIG. 1 schematically shows a schematic flowchart of a data analysis method according to an embodiment of the present disclosure
  • FIG. 2 schematically shows a schematic flowchart of generating a target yield problem stacking graph according to an embodiment of the present disclosure
  • FIG. 3 schematically shows a schematic flowchart of calculating matching degree data according to an embodiment of the present disclosure
  • FIG. 4 schematically shows a schematic flowchart of calculating correlation data according to an embodiment of the present disclosure
  • FIG. 5 schematically shows a schematic flowchart of determining a target weighting calculation method according to an embodiment of the present disclosure
  • FIG. 6 schematically shows an application schematic diagram of determining target measurement parameters according to an embodiment of the present disclosure
  • FIG. 7 schematically shows a schematic diagram of a data analysis apparatus according to an embodiment of the present disclosure
  • FIG. 8 schematically shows a schematic structural diagram of a computer system of an electronic device according to an embodiment of the present disclosure
  • FIG. 9 schematically shows a schematic diagram of a computer-readable storage medium according to one embodiment of the present disclosure.
  • Example embodiments will now be described more fully with reference to the accompanying drawings.
  • Example embodiments can be embodied in various forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art.
  • FIG. 1 schematically shows a schematic diagram of the flow of a data analysis method according to some embodiments of the present disclosure.
  • the data analysis method may include the following steps:
  • Step S110 acquiring the stacking pattern of the target yield problem corresponding to the wafer group with the target yield problem, and acquiring the measurement data stacking pattern of the wafer group under different types of tests;
  • Step S120 performing graphic matching between the target yield problem stacking graph and each of the measurement data stacking graphs to obtain matching degree data corresponding to the target yield problem stacking graph and each of the measurement data stacking graphs;
  • Step S130 calculating the correlation data corresponding to each of the measurement data stacking graphs and the target yield problem stacking graph;
  • Step S140 Perform weighted calculation on the matching degree data and the correlation data, and determine a target measurement parameter that causes the target yield problem according to the weighted calculation result.
  • the target yield problem stacking pattern is automatically matched with each measurement data stacking pattern, and the matching degree data and correlation data obtained by pattern matching are automatically determined.
  • the target measurement parameters of the target yield problem do not need to be analyzed manually, which effectively improves the analysis efficiency of the target measurement parameters that cause the target yield problem.
  • the matching degree data and correlation data of the stacked graphs of each measurement data determine the target measurement parameters, which can improve the analysis efficiency while ensuring the accuracy of the analysis results and effectively improving the product quality.
  • step S110 a stacking pattern of target yield problems corresponding to a wafer group having a target yield problem is obtained, and a stacking pattern of measurement data of the wafer group under different types of tests is obtained.
  • the yield may be a wafer yield
  • the wafer yield refers to the difference between the number of qualified chips after all process steps are completed and the number of valid chips on the entire wafer Ratio
  • the yield problem can be caused by various process problems that cause the wafer yield to not meet the standard yield requirements.
  • the yield problem can be caused by a defect problem (Defect) or an electrical problem (WAT). ), it can also be caused by a structural problem (MET), and of course, it can also be caused by other problems in various process steps that cause the wafer yield to not meet the standard yield requirements, which is not specially made in this example embodiment. limited.
  • defect parameters corresponding to defect problems may include measurement parameters such as line width parameters, depth parameters, film thickness parameters, uniformity parameters, height difference parameters, and curvature parameters in each process; Parameters can include on-current parameters of various semiconductor components, leakage current parameters in all directions, threshold voltage parameters, breakdown voltage parameters, capacitance parameters between structures, inductance parameters, resistance parameters and other measurement parameters;
  • the structural parameters may include structural abnormal parameters that occur in each process.
  • this is only a schematic illustration, and this exemplary embodiment is not limited thereto.
  • step S110 may include steps S210 to S220, and a target yield problem stacking graph may be generated through steps S210 to S220:
  • step S210 a wafer group with a target yield problem is obtained, and each wafer in the wafer group has a target yield problem distribution pattern;
  • Step S220 stacking the target yield problem distribution patterns corresponding to each of the wafers to obtain a target yield problem stacking pattern.
  • the wafer group may refer to a collection of wafers with the same target yield problem
  • the wafer group may include the target yield problem distribution pattern of each wafer
  • the target yield problem distribution pattern may be It is obtained by obtaining the feature graph containing various measurement parameters with the same target yield problem that is input by the user immediately through the input interface, or it can be pre-collected and stored in different time periods under the process flow with the same target yield problem.
  • the characteristic patterns containing various measurement parameters and having the same target yield problem may also be obtained in other ways, which are not specifically limited in this exemplary embodiment.
  • the target yield problem stacking pattern may refer to the pattern including the target yield problem distribution characteristics obtained by stacking the target yield problem distribution patterns corresponding to each wafer in the wafer group.
  • the distribution graph of the yield problem is superimposed to obtain the graph feature corresponding to the target yield problem, so as to automatically perform graph feature matching on the stacked graph of the target yield problem.
  • the measurement data stacking graph can also be generated by the following steps:
  • Each of the measurement data distribution patterns is stacked to obtain a measurement data stack pattern corresponding to the wafer group.
  • the measurement parameters may refer to parameters corresponding to various process problems that may cause the target yield problem of the wafer group under different types of tests.
  • the measurement parameters may be the defect parameters corresponding to the defect problems, or the The electrical parameters corresponding to the performance problems can also be the structural parameters corresponding to the structural problems.
  • it can also be the measurement parameters corresponding to various process problems that can cause the target yield problem of the wafer group. This example embodiment There is no special restriction on this.
  • the measurement data set may refer to a collection of measurement data corresponding to each wafer in the wafer group under different types of tests, through which measurement data distribution graphs corresponding to each measurement parameter can be generated, and then the measurement data distribution graph corresponding to each measurement parameter can be generated.
  • the measurement data distribution patterns corresponding to the measurement parameters are stacked to obtain multiple measurement data stacking patterns of the wafer group under different types of tests.
  • DC testing verifies voltage and current parameters
  • functional testing verifies the correctness of a series of logic functions within the chip Wait.
  • sub-categories of yield under these categories which are generally expressed by Bin. Different types of Bin represent different test types.
  • the present disclosure aims to correlate the yield problem with the process problem corresponding to the wafer group when the wafer yield problem pattern is obtained, and the process problem is reflected by measurement data. A variety of measurements will be performed after the end of each process. These measurement data are indicators that reflect the quality of the process. In each measurement process, the corresponding Map will also be obtained.
  • the technical purpose of the present disclosure is to find a certain measurement pattern that has the greatest impact on the yield problem of the wafer group through the judgment of pattern matching and pattern correlation, and then use this measurement pattern. Locate the process that is most related to the yield problem of the wafer group, and then adjust the process steps to improve the yield of wafers and improve product quality.
  • step S120 the target yield problem stacking graph and each of the measurement data stacking graphs are subjected to graph matching to obtain matching degree data corresponding to the target yield problem stacking graph and each of the measurement data stacking graphs .
  • the matching degree data may refer to data representing the degree of matching between the measurement data stacking pattern and the pattern features in the target yield problem stacking pattern, for example, the matching degree data (between 0.0-1.0 or 0%-100%) can be 0.1 or 10%, can also be 0.9 or 90%, of course, can also be other data that can characterize the degree of matching between the measured data stacking pattern and the pattern features in the target yield problem stacking pattern , which is not particularly limited in this exemplary embodiment.
  • the target measurement data stacking graph corresponding to the input measurement parameters under different types of tests can be determined in the measurement data stacking graph, and then the target yield problem stacking graph and the target quantity corresponding to the measurement parameter can be determined. Stacked graphs of measured data for graph matching.
  • the target measurement data stacking graph may refer to the measurement data stacking graph corresponding to the measurement parameters input in the measurement data stacking graph.
  • the wafers under the measurement parameters are The measurement data distribution graphs corresponding to the groups are stacked to obtain the measurement data stacking graph, and then the target measurement data stacking graph corresponding to the measurement parameter is matched with the target yield problem stacking graph.
  • step S120 may include steps S310 to S330, and matching degree data may be calculated through steps S310 to S330:
  • step S310 extracting the first graphic feature in the target yield problem stacking graph, and extracting the second graph feature in each of the measurement data stacking graphs;
  • Step S320 calculating the first feature vector corresponding to the first graphic feature and the second feature vector corresponding to the second graphic feature
  • Step S330 calculating according to the first feature vector and the second feature vector, matching degree data corresponding to the target yield problem stacking graph and each of the measurement data stacking graphs.
  • the first graphic feature may be a graphic feature corresponding to the target yield problem distribution in the target yield problem stacking graph
  • the second graphic feature may be a graphic feature corresponding to different measurement parameter distributions in the measurement data stacking graph
  • the pre-built convolutional neural network model can be used to extract the first graph feature in the stack graph of the target yield problem and the second graph feature in the stack graph of the measurement data corresponding to each measurement parameter.
  • the feature transform algorithm Scale-invariant feature transform, SIFT
  • SIFT Scale-invariant feature transform
  • HOG directional gradient histogram algorithm
  • the first feature vector corresponding to the first pattern feature and the first feature vector corresponding to the second pattern feature can be calculated. Two eigenvectors, and further determining the matching degree data between the target yield problem stacking pattern and the measurement data stacking pattern according to the first eigenvector and the second eigenvector.
  • first and second in this exemplary embodiment are only used to distinguish different graphic features and feature vectors corresponding to different graphic features, and have no special meaning here, which should not be used in this exemplary embodiment. cause any special qualifications.
  • the matching degree data can be calculated based on the feature vector through the following steps:
  • the similarity data is used as matching degree data corresponding to the stacked graph of the target yield problem and each of the stacked graphs of the measurement data.
  • the similarity data may refer to data representing the first feature vector and the second feature vector.
  • the similarity data may be the cosine similarity data (CosineSimilarity) corresponding to the first feature vector and the second feature vector, or may be The Pearson Correlation Coefficient corresponding to the first eigenvector and the second eigenvector may also be the Euclidean Distance (Euclidean Distance) corresponding to the first eigenvector and the second eigenvector.
  • Euclidean Distance Euclidean Distance
  • the similarity data of the first feature vector and the second feature vector can be used as data representing the matching degree between the target yield problem stacking pattern and the measurement data stacking pattern.
  • the matching data between the stacked graph of the target yield problem and the stacked graph of each measurement data can also be obtained in other ways.
  • the stacked graph of the target yield problem and the target measurement can be determined by a trained deep learning model or a random forest model.
  • the matching degree data of the data stacking graph is not limited to this example embodiment.
  • step S130 the correlation data corresponding to each of the measurement data stacking patterns and the target yield problem stacking patterns is calculated.
  • the correlation data may refer to a quantity that can characterize the degree of correlation between the measurement data stacking pattern and the target yield problem stacking pattern, for example, the main pattern area in the target yield problem stacking pattern may be
  • the wafer yield in the correlation coordinate system is taken as the horizontal axis of the correlation coordinate system, and the average value of the measurement parameters corresponding to each wafer in the main pattern area in the measurement data stack pattern can be taken as the vertical axis of the correlation coordinate system.
  • the curve in the correlation coordinate system calculates the correlation data between the measured data stacking graph and the target yield problem stacking graph.
  • there may also be other methods capable of characterizing the correlation data between the measurement data stacking pattern and the target yield problem stacking pattern which is not specifically limited in this exemplary embodiment.
  • step S130 may include steps S410 to S420.
  • the measurement data stacking graph and the target yield problem stacking graph may be graph partitioned through steps S410 to S420:
  • step S410 the target yield problem stacking pattern and the measurement data stacking pattern are partitioned to determine a main pattern area;
  • Step S420 calculating correlation data between the target yield problem stacking graph and each of the measurement data stacking graphs based on the main graph area.
  • the partition may refer to the process of dividing the area according to the target yield problem stacking graphics or the distribution of graphics features in the measurement data stacking graphics. For example, it may be based on a fixed-size graphics area (such as a 10*10 graphics area) Partition the target yield problem stacking graph or the measurement data stacking graph, you can also partition the target yield problem stacking graph or the measurement data stacking graph according to the density of the pattern feature distribution, or you can select the entire target yield problem
  • the stacking pattern or the measurement data stacking pattern is partitioned, and the partitioning is specifically performed according to the distribution of the pattern features of the target yield problem stacking pattern or the measurement data stacking pattern, which is not specifically limited in this exemplary embodiment.
  • the main pattern area may refer to the main feature distribution area formed by partitioning the target yield problem stack pattern or the measurement data stack pattern.
  • the main pattern area in the target yield problem stack pattern may be the same as the measurement data stack pattern.
  • the main graphics area in corresponds to the main graphics area in (such as graphics area shape, graphics area location, graphics area size, etc.).
  • the correlation data corresponding to the measurement data stacking pattern and the target yield problem stacking pattern can be calculated through the following steps: the wafer yield data corresponding to the main pattern area in the target yield problem stacking pattern can be determined; Statistical feature data of the measurement parameters corresponding to the main pattern area in the measurement data stacking pattern; Calculate the correlation between the target yield problem stacking pattern and each measurement data stacking pattern through the wafer yield data and statistical feature data in the main pattern area data.
  • the main pattern area has a total of 100 chip dies, and 10 of them fail
  • the The yield rate of the pattern area is 90%, so the yield value corresponding to the pattern area can be used as the ordinate of the correlation curve, and then the selected 100 wafer dies have the WAT A parameters (that is, the measurement parameters at this time are The average value of the WAT A value in the measurement data stacking graph corresponding to the WAT A parameter) is taken as the abscissa of the correlation curve, and then the correlation between the measurement data stacking graph and the target yield problem stacking graph is determined according to the constructed correlation curve.
  • Sexual data is taken as the abscissa of the correlation curve.
  • step S140 weighted calculation is performed on the matching degree data and the correlation data, and a target measurement parameter causing the target yield problem is determined according to the weighted calculation result.
  • the target measurement parameter may refer to a parameter obtained through analysis that causes the target yield problem in the target yield problem stacking pattern, for example, it is assumed that the target yield problem stacking pattern and the WAT A parameter If the weighted calculation result of the corresponding measurement data stack pattern is higher, it is considered that the target yield problem stack pattern matches the WAT A parameter. At this time, the WAT A parameter can be considered to be the target yield problem of the stack pattern that causes the target yield problem. target measurement parameters.
  • the measurement data stacking graph corresponding to each measurement parameter may be sorted according to the weighted calculation result, the measurement data stacking graph corresponding to the largest weighted calculation result is determined as the maximum correlated measurement data stacking graph, and the maximum correlation quantity
  • the measurement parameters corresponding to the measurement data stacking pattern are used as the target measurement parameters that cause the target yield problem.
  • the matching degree data or the correlation data are relatively high, but the final weighted calculation result is low. At this time, it does not mean that the target yield problem has nothing to do with the measurement parameters.
  • the accuracy of the obtained target measurement parameters can be combined with matching data, correlation data and weighted calculation results to jointly decide the most relevant measurement parameters:
  • the matching data and correlation data respectively determine the measurement data stacking graph with the highest matching degree and the measurement data stacking graph with the highest correlation with the target yield problem stacking graph, and then stack the measurement data stacking graph based on the highest matching degree.
  • the graph and the highest correlation measurement data stack graph determine the highest matching metric parameter and highest correlation measurement parameter corresponding to the target yield problem stack graph; and then the highest matching metric parameter and highest correlation measurement parameter.
  • the parameters and target measurement parameters and target yield problems perform a rationality judgment process, so as to determine the measurement parameters most relevant to the target yield problem according to the rationality judgment results.
  • the measurement data stacking graph with the highest matching degree may refer to the measurement data stacking graph corresponding to the highest matching degree data after sorting the matching degree data, and the measurement parameter corresponding to the measurement data stacking graph with the highest matching degree is The highest matching metric parameter.
  • the measurement data stacking graph with the highest correlation may refer to the measurement data stacking graph corresponding to the highest correlation data after sorting the correlation data, and the measurement parameter corresponding to the measurement data stacking graph with the highest correlation is the highest correlation data stacking graph.
  • the rationality judgment process may refer to a preset value for judging the highest matching measurement parameter (the one corresponding to the measurement data stacking graph of the highest matching degree) when the values of the matching degree data, the correlation data and the weighted calculation result differ greatly.
  • measurement parameter the highest correlation measurement parameter (the measurement parameter corresponding to the highest correlation measurement data stacking graph), and the target measurement parameter (the measurement parameter corresponding to the most relevant measurement data stacking graph corresponding to the maximum weighted calculation result) ).
  • the determined highest matching measurement parameter, the highest correlation measurement parameter and the target measurement are sent to the management control terminal, so that the relevant management personnel of the management control terminal can manually further judge the most relevant measurement parameters that lead to the target yield problem; the results can also be further judged by the preset judgment conditions, such as in When the matching degree data or correlation data is greater than 90% and the weighted calculation result is less than 20%, the highest matching measurement parameter or the highest correlation measurement parameter corresponding to the matching degree data or correlation data is used as the final analysis result , of course, this is only a schematic example, which is not specifically limited in this exemplary embodiment.
  • the matching degree data and the correlation data may be weighted by a preselected target weighted calculation method.
  • the target weighted calculation method may refer to a weighted calculation method in which the weighted calculation result determined in advance according to the experimental data can correctly reflect the analysis result.
  • the target weighted calculation method can be determined through the steps in FIG. 5 :
  • step S510 the sample yield problem stacking graph and the sample measurement data stacking graph pre-stored in the database are obtained; wherein, the database further includes the sample yield problem stacking graph that causes the yield problem of the sample yield problem. sample measurement parameters;
  • Step S520 obtaining preset multiple weighted calculation methods
  • Step S530 according to multiple weighted calculation methods, sequentially weighted calculation of the sample yield problem stacking graph and the sample matching degree data and sample correlation data corresponding to the sample measurement data stacking graph;
  • Step S540 Determine a target weighting calculation method from the multiple weighting calculation methods according to the weighting calculation results corresponding to the multiple weighting calculation methods and the sample measurement parameters.
  • the sample target yield problem stacking graph and the sample measurement data stacking graph may refer to the sample data used for testing in the pre-established standard database.
  • the sample target yield problem stacking graph and the sample measurement data stacking graph may be 100% corresponding relationship.
  • the above method is used to perform weighted calculation on the sample target yield problem stacking graph and the matching degree data and correlation data corresponding to the sample measurement data stacking graph. If the weighted calculation result is 80%, it means that the corresponding weighted calculation method is poor. Other weighted calculation methods can be selected for recalculation until the weighted calculation result is infinitely close to 100%, and finally the weighted calculation method with the highest weighted calculation result is used as the target weighted calculation method.
  • Multiple weighted calculation methods may include a weighted calculation method in which matching degree data and correlation data are multiplied, or a weighted calculation method in which weights are respectively configured and added to matching degree data and correlation data.
  • the specific weighting calculation method may be: It is optimized and constructed by continuously transforming different mathematical formulas during the experiment. Therefore, the present exemplary embodiment does not specifically limit various weighting calculation methods.
  • sample target yield problem stacking graph and sample measurement data stacking graph in the pre-established standard database, select the target weighting calculation method from a variety of weighted calculation methods, and further improve the target yield problem stacking graph and measurement data stacking graph
  • the accuracy of the weighted calculation results of the corresponding matching data and correlation data improves the accuracy of the target measurement parameters, and determines the process flow that causes the target yield problem based on the obtained target measurement parameters to improve the wafer yield. .
  • FIG. 6 schematically shows an application diagram of determining target measurement parameters according to an embodiment of the present disclosure.
  • step S610 obtain the input wafer group and the target yield problem distribution pattern corresponding to each wafer in the wafer group (for example, LotA, waferB, LotC, LotD, waferE%), and obtain the difference in the input Measurement parameters under type test:
  • Step S620 stacking the target yield problem distribution patterns to obtain the target yield problem stacking patterns, and stacking the measurement data distribution patterns under different types of tests to obtain the measurement data stacking patterns. Specifically, based on the input measurement parameters under different types of tests, stack the target yield problem distribution patterns to obtain the target yield problem stacking pattern 601 of the target yield problem distribution characteristics.
  • the target yield problem stacking pattern 601 can be Including the failed wafer area 602, and the qualified wafer area except the failed wafer area 602 in the target yield problem stacking pattern 601, of course, the target yield problem stacking pattern 601 is only a schematic diagram, and should not cause any special limited;
  • Step S630 calculating matching degree data and correlation data.
  • the image features in the target yield problem stacking pattern 601 and the measurement data stacking patterns under different types of tests such as the stacking pattern corresponding to the wafer group under the MET A measurement parameter, the wafer group under the MET B measurement parameter
  • the stacking pattern, the stacking pattern corresponding to the wafer group under the Defect B measurement parameter, etc. perform feature matching to obtain the matching degree data, and construct a correlation curve to obtain the correlation data;
  • Step S640 calculate the weighted score. Specifically, weighted calculation is performed on the obtained matching degree data and correlation data to obtain a weighted score (weighted calculation result), and finally the target measurement parameters are determined based on the weighted score, or the corresponding parameters of the weighted score, matching degree data and correlation data are determined.
  • the rationality of the target measurement parameters, the highest matching measurement parameters, and the highest correlation measurement parameters determine the most relevant measurement parameters that cause the target yield problem of the target yield problem stacking pattern 601, and finally analyze the results. (and the analysis process) for visualization.
  • the data analysis apparatus 700 includes: a graph acquisition module 710 , a correlation degree calculation module 720 and a target measurement parameter analysis module 730 . in:
  • the pattern acquisition module 710 is configured to acquire the target yield problem stacking pattern corresponding to the wafer group with the target yield problem, and acquire the measurement data stacking pattern of the wafer group under different types of tests;
  • the correlation degree calculation module 720 is configured to calculate the matching degree data and correlation data corresponding to the target yield problem stacking graph and each of the measurement data stacking graphs, and weight the matching degree data and the correlation data Calculate the weighted calculation result;
  • the target measurement parameter analysis module 730 is configured to analyze the weighted calculation results to determine target measurement parameters that cause the target yield problem.
  • the association degree calculation module 720 further includes:
  • a matching degree data calculation unit configured to perform graphic matching between the target yield problem stacking graph and the measurement data stacking graph, and obtain a match corresponding to the target yield problem stacking graph and each of the measurement data stacking graphs degree data;
  • the correlation data calculation unit is configured to calculate the correlation data corresponding to each of the measurement data stacking graphs and the target yield problem stacking graphs.
  • the correlation data calculation unit further includes:
  • a pattern partition sub-unit for partitioning the target yield problem stack pattern and the measurement data stack pattern to determine a main pattern area
  • a wafer yield determination subunit used for determining wafer yield data corresponding to the main pattern area in the target yield problem stack pattern
  • a statistical feature determination subunit configured to determine the statistical feature data of the measurement parameters corresponding to the main graphics area in each of the measurement data stacked graphics;
  • a correlation data calculation subunit configured to calculate the correlation between the measurement data stacking graph and each of the target yield problem stacking graphs through the wafer yield data and the statistical feature data in the main graph area Sexual data.
  • the graph obtaining module 710 further includes a target yield problem stack graph obtaining unit, where the target yield problem stack graph obtaining unit is configured to:
  • the target yield problem distribution patterns corresponding to each of the wafers are stacked to obtain a target yield problem stacking pattern.
  • the graph obtaining module 710 further includes a measurement data stacking graph obtaining unit, and the measurement data stacking graph obtaining unit is configured to:
  • Each of the measurement data distribution patterns is stacked to obtain a measurement data stack pattern corresponding to the wafer group.
  • the matching degree data calculation unit further includes:
  • a pattern feature extraction subunit configured to extract the first pattern feature in the stacked pattern of the target yield problem, and extract the second pattern feature in each of the measurement data stacked patterns
  • a feature vector calculation subunit used for calculating the first feature vector corresponding to the first graphic feature and the second feature vector corresponding to the second graphic feature;
  • a matching degree calculation subunit configured to calculate and obtain matching degree data corresponding to the target yield problem stacking graph and each of the measurement data stacking graphs according to the first feature vector and the second feature vector.
  • the matching degree calculation subunit is further configured to:
  • the similarity data is used as matching degree data corresponding to the stacked graph of the target yield problem and each of the stacked graphs of the measurement data.
  • the target measurement parameter analysis module 730 further includes a first analysis unit, and the first analysis unit is configured to:
  • the measurement parameter corresponding to the most relevant measurement data stacking pattern is used as the target measurement parameter that causes the target yield problem.
  • the target measurement parameter analysis module 730 further includes a second analysis unit, and the second analysis unit is configured to:
  • the matching degree data and the correlation data respectively determining a measurement data stacking graph with the highest matching degree and a measurement data stacking graph having the highest correlation with the target yield problem stacking graph;
  • the data analysis apparatus further includes a weighted calculation method screening unit, and the weighted calculation method screening unit is configured to:
  • the target weighted calculation method is determined from the multiple weighted calculation methods.
  • the measurement parameters include one or more combinations of defect parameters, electrical parameters and structural parameters.
  • modules or units of the data analysis device are mentioned in the above detailed description, this division is not mandatory. Indeed, according to embodiments of the present disclosure, the features and functions of two or more modules or units described above may be embodied in one module or unit. Conversely, the features and functions of one module or unit described above may be further divided into multiple modules or units to be embodied.
  • an electronic device capable of implementing the above data analysis method is also provided.
  • aspects of the present disclosure may be implemented as a system, method or program product. Therefore, various aspects of the present disclosure can be embodied in the following forms: a complete hardware embodiment, a complete software embodiment (including firmware, microcode, etc.), or a combination of hardware and software aspects, which may be collectively referred to herein as an embodiment "circuit", “module” or "system”.
  • FIG. 8 An electronic device 800 according to such an embodiment of the present disclosure is described below with reference to FIG. 8 .
  • the electronic device 800 shown in FIG. 8 is only an example, and should not impose any limitation on the function and scope of use of the embodiments of the present disclosure.
  • electronic device 800 takes the form of a general-purpose computing device.
  • Components of the electronic device 800 may include, but are not limited to: the above-mentioned at least one processing unit 810 , the above-mentioned at least one storage unit 820 , a bus 830 connecting different system components (including the storage unit 820 and the processing unit 810 ), and a display unit 840 .
  • the storage unit stores program codes, and the program codes can be executed by the processing unit 810, so that the processing unit 810 executes various exemplary methods according to the present disclosure described in the above-mentioned “Exemplary Methods” section of this specification. Example steps.
  • the processing unit 810 may perform step S110 as shown in FIG.
  • Step S120 the target yield problem stacking graph and each of the measurement data stacking graphs are matched graphically to obtain the target yield problem stacking graph and each of the measurement data stacking graphs Corresponding matching degree data; step S130, calculating the correlation data corresponding to each of the measurement data stacking graphs and the target yield problem stacking graphs; step S140, weighting the matching degree data and the correlation data calculating, and determining the target measurement parameter that causes the target yield problem according to the weighted calculation result.
  • the storage unit 820 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 821 and/or a cache storage unit 822 , and may further include a read only storage unit (ROM) 823 .
  • RAM random access storage unit
  • ROM read only storage unit
  • the storage unit 820 may also include a program/utility 824 having a set (at least one) of program modules 825 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, An implementation of a network environment may be included in each or some combination of these examples.
  • the bus 830 may be representative of one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local area using any of a variety of bus structures bus.
  • the electronic device 800 may also communicate with one or more external devices 870 (eg, keyboards, pointing devices, Bluetooth devices, etc.), with one or more devices that enable a user to interact with the electronic device 800, and/or with Any device (eg, router, modem, etc.) that enables the electronic device 800 to communicate with one or more other computing devices. Such communication may take place through input/output (I/O) interface 850 . Also, the electronic device 800 may communicate with one or more networks (eg, a local area network (LAN), a wide area network (WAN), and/or a public network such as the Internet) through a network adapter 860 . As shown, network adapter 860 communicates with other modules of electronic device 800 via bus 830 . It should be understood that, although not shown, other hardware and/or software modules may be used in conjunction with electronic device 800, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives and data backup storage systems.
  • the exemplary embodiments described herein may be implemented by software, or may be implemented by software combined with necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure may be embodied in the form of software products, and the software products may be stored in a non-volatile storage medium (which may be CD-ROM, U disk, mobile hard disk, etc.) or on a network , including several instructions to cause a computing device (which may be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to an embodiment of the present disclosure.
  • a computing device which may be a personal computer, a server, a terminal device, or a network device, etc.
  • a computer-readable storage medium on which a program product capable of implementing the above-described method of the present specification is stored.
  • various aspects of the present disclosure may also be implemented in the form of a program product including program code for causing the program product to run on a terminal device when the program product is run on a terminal device.
  • the terminal device performs the steps according to various exemplary embodiments of the present disclosure described in the above-mentioned "Example Method" section of this specification.
  • a program product 900 for implementing the above-mentioned data analysis method according to an embodiment of the present disclosure is described, which can adopt a portable compact disk read only memory (CD-ROM) and include program codes, and can be displayed on a terminal devices such as personal computers.
  • CD-ROM compact disk read only memory
  • the program product of the present disclosure is not limited thereto, and in this document, a readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.
  • the program product may employ any combination of one or more readable media.
  • the readable medium may be a readable signal medium or a readable storage medium.
  • the readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or a combination of any of the above. More specific examples (non-exhaustive list) of readable storage media include: electrical connections with one or more wires, portable disks, hard disks, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, portable compact disk read only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.
  • a computer readable signal medium may include a propagated data signal in baseband or as part of a carrier wave with readable program code embodied thereon. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing.
  • a readable signal medium can also be any readable medium, other than a readable storage medium, that can transmit, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
  • Program code embodied on a readable medium may be transmitted using any suitable medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
  • Program code for performing the operations of the present disclosure may be written in any combination of one or more programming languages, including object-oriented programming languages—such as Java, C++, etc., as well as conventional procedural Programming Language - such as the "C" language or similar programming language.
  • the program code may execute entirely on the user computing device, partly on the user device, as a stand-alone software package, partly on the user computing device and partly on a remote computing device, or entirely on the remote computing device or server execute on.
  • the remote computing device may be connected to the user computing device through any kind of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (eg, using an Internet service provider business via an Internet connection).
  • LAN local area network
  • WAN wide area network
  • the exemplary embodiments described herein may be implemented by software, or may be implemented by software combined with necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure may be embodied in the form of software products, and the software products may be stored in a non-volatile storage medium (which may be CD-ROM, U disk, mobile hard disk, etc.) or on a network , which includes several instructions to cause a computing device (which may be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiment of the present disclosure.
  • a computing device which may be a personal computer, a server, a touch terminal, or a network device, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Business, Economics & Management (AREA)
  • Quality & Reliability (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Economics (AREA)
  • Databases & Information Systems (AREA)
  • Multimedia (AREA)
  • Software Systems (AREA)
  • Development Economics (AREA)
  • Medical Informatics (AREA)
  • Educational Administration (AREA)
  • General Health & Medical Sciences (AREA)
  • Game Theory and Decision Science (AREA)
  • Computing Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Testing Or Measuring Of Semiconductors Or The Like (AREA)

Abstract

一种数据分析方法及装置、电子设备、存储介质，涉及半导体技术领域。该数据分析方法包括：获取晶圆组对应的目标良率问题堆叠图形，并获取晶圆组在不同类型测试下的量测数据堆叠图形(S110)；将目标良率问题堆叠图形与量测数据堆叠图形进行图形匹配，得到目标良率问题堆叠图形与各量测数据堆叠图形对应的匹配度数据(S120)；计算量测数据堆叠图形与目标良率问题堆叠图形对应的相关性数据(S130)；对匹配度数据与相关性数据进行加权计算，根据加权计算结果确定导致目标良率问题的目标量测参数(S140)。本方法可以在筛选导致目标良率问题的目标量测参数时，提高筛选的效率，提升确定的目标量测参数的准确度。

Description

数据分析方法及装置、电子设备、存储介质
相关申请的交叉引用
本申请要求于2020年09月09日提交的申请号为202010938941.5、名称为“数据分析方法及装置、电子设备、存储介质”的中国专利申请的优先权，该中国专利申请的全部内容通过引用全部并入本文。
技术领域
本公开涉及半导体技术领域，特别是涉及一种数据分析方法、数据分析装置、电子设备以及计算机可读存储介质。
背景技术
随着科学技术的飞速发展，半导体器件(Semiconductor Device)作为一种导电性介于良导电体与绝缘体之间，利用半导体材料特殊电特性来完成特定功能的电子器件，由于可用来产生、控制、接收、变换、放大信号和进行能量转换等多种功能，越来越得到人们的重视和关注。
晶圆(Wafer)是指制作硅半导体积体电路所用的硅晶片，由于受加工晶圆的工艺制程的影响，生成的晶圆的良率时高时低，因此需要对导致晶圆良率的良率问题进行分析，以便于根据良率问题分析结果调整工艺制程，提升晶圆良率。
目前，在对导致晶圆的良率问题的工艺参数进行分析时，一般是通过人工进行分析比对得到结论，但是这种方案使良率问题的分析效率较低，分析周期较长，并且人工分析得到的分析结果准确率较低。
需要说明的是，在上述背景技术部分公开的信息仅用于加强对本公开的背景的理解，因此可以包括不构成对本领域普通技术人员已知的现有技术的信息。
发明内容
本公开实施例的目的在于提供一种数据分析方法、数据分析装置、电子设备以及计算机可读存储介质，进而至少在一定程度上克服相关方案中筛选导致目标良率问题的目标量测参数效率较低，准确率较低的问题。
根据本公开的第一方面，提供了一种数据分析方法，包括：获取具有目标良率问题的晶圆组对应的目标良率问题堆叠图形，并获取所述晶圆组在不同类型测试下的量测数据堆叠图形；将所述目标良率问题堆叠图形与各所述量测数据堆叠图形进行图形匹配，得到所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据；计算各所述量测数据堆叠图形与所述目标良率问题堆叠图形对应的相关性数据；对所述匹配度数据与所述相关性数据进行加权计算，根据所述加权计算结果确定导致所述目标良率问题的目标量测 参数。
根据本公开的第二方面，提供了一种数据分析装置，包括：图形获取模块，用于获取具有目标良率问题的晶圆组对应的目标良率问题堆叠图形，并获取所述晶圆组在不同类型测试下的量测数据堆叠图形；关联程度计算模块，用于计算所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据与相关性数据，并对所述匹配度数据与所述相关性数据进行加权计算得到加权计算结果；目标量测参数分析模块，用于分析所述加权计算结果以确定导致所述目标良率问题的目标量测参数。
根据本公开的第三方面，提供了一种电子设备，包括：处理器；以及存储器，所述存储器上存储有计算机可读指令，所述计算机可读指令被所述处理器执行时实现第一方面所述的数据分析方法。
根据本公开的第四方面，提供一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现第一方面所述的数据分析方法。
应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，并不能限制本公开。
附图说明
此处的附图被并入说明书中并构成本说明书的一部分，示出了符合本申请的实施例，并与说明书一起用于解释本公开的原理。显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。在附图中：
图1示意性示出了根据本公开一个实施例的数据分析方法的流程示意图；
图2示意性示出了根据本公开一个实施例的生成目标良率问题堆叠图形的流程示意图；
图3示意性示出了根据本公开一个实施例的计算匹配度数据的流程示意图；
图4示意性示出了根据本公开一个实施例的计算相关性数据的流程示意图；
图5示意性示出了根据本公开一个实施例的确定目标加权计算方式的流程示意图；
图6示意性示出了根据本公开一个实施例的确定目标量测参数的应用示意图；
图7示意性示出了根据本公开一个实施例的数据分析装置的示意图；
图8示意性示出了根据本公开一个实施例的电子设备的计算机系统的结构示意图；
图9示意性示出了根据本公开一个实施例的计算机可读存储介质的示意图。
具体实施方式
现在将参考附图更全面地描述示例实施方式。然而，示例实施方式能够以多种形式实施，且不应被理解为限于在此阐述的范例；相反，提供这些实施方式使得本公开将更加全面和完整，并将示例实施方式的构思全面地传达给本领域的技术人员。
此外，所描述的特征、结构或特性可以以任何合适的方式结合在一个或更多实施例中。在下面的描述中，提供许多具体细节从而给出对本公开的实施例的充分理解。然而，本领域技术人员将意识到，可以实践本公开的技术方案而没有特定细节中的一个或更多，或者可以采用其它的方法、组元、装置、步骤等。在其它情况下，不详细示出或描述公知方法、装置、实现或者操作以避免模糊本公开的各方面。
此外，附图仅为示意性图解，并非一定是按比例绘制。附图中所示的方框图仅仅是功能实体，不一定必须与物理上独立的实体相对应。即，可以采用软件形式来实现这些功能实体，或在一个或多个硬件模块或集成电路中实现这些功能实体，或在不同网络和/或处理器装置和/或微控制器装置中实现这些功能实体。
在本示例实施例中，首先提供了一种数据分析方法，该数据分析方法可以应用于终端设备，也可以应用于服务器，本示例实施例对此不做特殊限定，下面以服务器执行该方法为例进行说明。图1示意性示出了根据本公开的一些实施例的数据分析方法流程的示意图，参考图1所示，该数据分析方法可以包括以下步骤：
步骤S110，获取具有目标良率问题的晶圆组对应的目标良率问题堆叠图形，并获取所述晶圆组在不同类型测试下的量测数据堆叠图形；
步骤S120，将所述目标良率问题堆叠图形与各所述量测数据堆叠图形进行图形匹配，得到所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据；
步骤S130，计算各所述量测数据堆叠图形与所述目标良率问题堆叠图形对应的相关性数据；
步骤S140，对所述匹配度数据与所述相关性数据进行加权计算，根据所述加权计算结果确定导致所述目标良率问题的目标量测参数。
根据本示例实施例中的数据分析方法，一方面，自动将目标良率问题堆叠图形与各量测数据堆叠图形进行图形匹配，并根据图形匹配得到的匹配度数据以及相关性数据自动确定导致晶圆组目标良率问题的目标量测参数，不需要以人工的方式进行人工分析，有效提升导致目标良率问题的目标量测参数的分析效率；另一方面，结合目标良率问题堆叠图形与各量测数据堆叠图形的匹配度数据以及相关性数据确定目标量测参数，能够在提升分析效率的同时，保证分析结果的准确度，有效提升产品质量。
下面，将对本示例实施例中的数据分析方法进行进一步的说明。
在步骤S110中，获取具有目标良率问题的晶圆组对应的目标良率问题堆叠图形，并获取所述晶圆组在不同类型测试下的量测数据堆叠图形。
在本公开的一个示例实施例中，良率可以是晶圆良率(Wafer Yield)，晶圆良率是指完成所有工艺步骤后测试合格的芯片的数量与整片晶圆上的有效芯片的比值，良率问题可以是导致晶圆良率不满足标准良率要求的各种工艺制程上的问题，例如，良率问题可以是缺陷问题(Defect)导致的，也可以是电性问题(WAT)导致的，还可以是结构问题(MET)导致的，当然，还可以是其他导致晶圆良率不满足标准良率要求的各种工艺步骤上的问题， 本示例实施例对此不做特殊限定。
具体的，缺陷问题对应的缺陷参数可以包括各道制程中的线宽参数、深度参数、膜厚参数、均匀性参数、高低差参数、弯曲度参数等量测参数；电性问题对应的电性参数可以包括各类半导体元件的导通电流参数、各向漏电流参数、阈值电压参数、击穿电压参数、各结构间的电容参数、电感参数、阻值参数等量测参数；结构问题对应的结构参数可以包括各道制程中出现的结构异常参数，当然，此处仅是示意性举例说明，本示例实施例不以此为限。
在本公开的一个示例实施例中，步骤S110可以包括步骤S210至步骤S220，可以通过步骤S210至步骤S220生成目标良率问题堆叠图形：
参考图2所示，步骤S210，获取具有目标良率问题的晶圆组，所述晶圆组中的各晶圆具有目标良率问题分布图形；
步骤S220，将各所述晶圆对应的所述目标良率问题分布图形进行堆叠，得到目标良率问题堆叠图形。
其中，晶圆组(Lot/Wafer list)可以是指具有同一目标良率问题的晶圆的集合，晶圆组可以包括各晶圆具有的目标良率问题分布图形，目标良率问题分布图形可以是通过获取用户通过输入接口即时输入的具有同一目标良率问题的包含各种量测参数的特征图形，也可以是预先收集存储的不同时间段的工艺流程下具有同一目标良率问题的包含各种量测参数的特征图形，当然，还可以是以其他方式获取的具有同一目标良率问题的包含各种量测参数的特征图形，本示例实施例对此不做特殊限定。
目标良率问题堆叠图形可以是指将晶圆组中各晶圆对应的目标良率问题分布图形进行堆叠后得到的包含目标良率问题分布特征的图形，通过将具有目标良率问题的目标良率问题分布图形进行叠加，得到该目标良率问题对应的图形特征，以便于后续自动对该目标良率问题堆叠图形进行图形特征匹配。
在本公开的一个示例实施例中，还可以通过以下步骤生成量测数据堆叠图形：
获取在不同类型测试下所述晶圆组中各所述晶圆的量测数据集；
基于所述量测数据集生成不同类型测试对应的量测数据分布图形；
将各所述量测数据分布图形进行堆叠，得到所述晶圆组对应的量测数据堆叠图形。
其中，量测参数可以是指不同类型测试下可能导致晶圆组的目标良率问题的各种工艺问题对应的参数，例如量测参数可以是导致缺陷问题对应的缺陷参数，也可以是导致电性问题对应的电性参数，还可以是导致结构问题对应的结构参数，当然，还可以是其他能够导致晶圆组的目标良率问题的各种工艺问题对应的量测参数，本示例实施例对此不做特殊限定。
量测数据集可以是指不同类型测试下晶圆组中各晶圆对应的量测数据的集合，通过该量测数据集可以生成与各量测参数对应的量测数据分布图形，进而对各量测参数对应的量测数据分布图形进行堆叠，得到晶圆组在不同类型测试下的多个量测数据堆叠图形。
举例而言，对于晶圆组的良率测试的类型有很多，如DC测试、功能测试、AC测试等，其中DC测试验证电压及电流参数；功能测试验证芯片内部一系列逻辑功能操作的正确性等。这些类别下面还有很多良率的子类，一般用Bin来表述，不同种Bin代表不同的测试类型，例如可以采用“！”、“@”、“#”、“$”、“％”、“&”、“[”、“+”等符号标记出直流参数测试失效(DC fail，电流不能进入)的晶粒(die)，可以采用“a”、“F”等符号标记出读写功能完全失效的晶粒(die)，可以采用“X”、“Y”、“s”等符号标记出读写速度功能较慢的(Margin fail)的晶粒(die)等，当然，此处仅是示意性举例说明，由于半导体各厂家使用的不同导致编号有区别，并不应对本示例实施例造成任何特殊限定。除了Bin以外还有失效形态分析(FSA，Fail shape analysis)，其也有对应的测试项编号，也就是说不同的Bin或不同的FSA代表不同的失效原因，在良率测试时，一般一片wafer会得到一个Map图形，该Map图形反应了各晶粒的失效情况。
本公开旨在获得了晶圆良率问题图形时，如何将该良率问题与晶圆组对应的工艺问题关联起来，而工艺问题则是通过量测数据来反映的。在每一道工艺流程结束后都会进行多种量测，这些量测数据则是反映该道工艺好坏的指标，在每一道量测过程中也会得到相应的Map图形，但是由于导致良率问题的工艺问题有很多种，本公开的技术目的正是为了通过图形匹配和图形相关性的判断找到对晶圆组的良率问题影响最大相关的某一量测图形，进而由这一量测图形定位到与晶圆组的良率问题最大相关的那道工艺，进而对该工艺步骤进行调整，以提升晶圆的良率，提升产品质量。
在步骤S120中，将所述目标良率问题堆叠图形与各所述量测数据堆叠图形进行图形匹配，得到所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据。
在本公开的一个示例实施例中，匹配度数据可以是指表征量测数据堆叠图形与目标良率问题堆叠图形中的图形特征匹配程度的数据，例如，匹配度数据(0.0-1.0之间或者0％-100％时)可以是0.1或者10％，也可以是0.9或者90％，当然，还可以是其他能够表征量测数据堆叠图形与目标良率问题堆叠图形中的图形特征匹配程度的数据，本示例实施例对此不做特殊限定。
具体的，可以在量测数据堆叠图形中确定与输入的不同类型测试下的量测参数对应的目标量测数据堆叠图形，进而可以将目标良率问题堆叠图形与该量测参数对应的目标量测数据堆叠图形进行图形匹配。
其中，目标量测数据堆叠图形可以是指量测数据堆叠图形中输入的量测参数对应的量测数据堆叠图形，首先基于输入的不同类型测试下的量测参数对该量测参数下晶圆组对应的量测数据分布图形进行堆叠得到量测数据堆叠图形，进而将该量测参数对应的目标量测数据堆叠图形与目标良率问题堆叠图形进行图形匹配。
在本公开的一个示例实施例中，步骤S120可以包括步骤S310至步骤S330，可以通过步骤S310至步骤S330计算匹配度数据：
参考图3所示，步骤S310，提取所述目标良率问题堆叠图形中的第一图形特征，并 提取各所述量测数据堆叠图形中的第二图形特征；
步骤S320，计算所述第一图形特征对应的第一特征向量以及所述第二图形特征对应的第二特征向量；
步骤S330，根据所述第一特征向量以及所述第二特征向量计算得到所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据。
其中，第一图形特征可以是目标良率问题堆叠图形中目标良率问题分布对应的图形特征，第二图形特征可以是量测数据堆叠图形中的不同的量测参数分布对应的图形特征，例如，可以通过预构建的卷积神经网络模型提取目标良率问题堆叠图形中的第一图形特征和提取各量测参数对应的量测数据堆叠图形中的第二图形特征，也可以通过尺度不变特征变换算法(Scale-invariant feature transform，SIFT)提取目标良率问题堆叠图形中的第一图形特征和提取量测数据堆叠图形中的第二图形特征，还可以通过方向梯度直方图算法(Histogram of oriented gradient，HOG)提取目标良率问题堆叠图形中的第一图形特征和提取目标量测数据堆叠图形中的第二图形特征，当然，还可以是其他能够提取图形特征的方式，本示例实施例对此不做特殊限定。
在提取得到目标良率问题堆叠图形中的第一图形特征和目标量测数据堆叠图形中的第二图形特征之后，可以计算第一图形特征对应的第一特征向量以及第二图形特征对应的第二特征向量，进而根据第一特征向量以及第二特征向量确定目标良率问题堆叠图形与量测数据堆叠图形的匹配度数据。
需要说明的是，本示例实施例中的“第一”、“第二”仅用于区分不同的图形特征以及不同图形特征对应的特征向量，此处并没有特殊含义，不应对本示例实施例造成任何特殊限定。
进一步的，可以通过下述步骤基于特征向量计算匹配度数据：
计算所述第一特征向量以及所述第二特征向量的相似度数据；
将所述相似度数据作为所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据。
其中，相似度数据可以是指表征第一特征向量和第二特征向量的数据，例如，相似度数据可以是第一特征向量和第二特征向量对应的余弦相似度数据(CosineSimilarity)，也可以是第一特征向量和第二特征向量对应的皮尔逊相关系数(Pearson Correlation Coefficient)，还可以是第一特征向量和第二特征向量对应的欧几里德距离(Euclidean Distance)，当然，还可以是其他能够表征第一特征向量和第二特征向量的数据，本示例实施例对此不做特殊限定。
由于第一特征向量以及第二特征向量分别是目标良率问题堆叠图形的目标良率问题分布的图形特征和量测数据堆叠图形中的不同的量测参数分布的图形特征对应的特征向量，因此第一特征向量和第二特征向量的相似度数据可以作为表征目标良率问题堆叠图形和量测数据堆叠图形的匹配程度的数据。
当然，还可以通过其他方式得到目标良率问题堆叠图形和各量测数据堆叠图形的匹配度数据，例如可以通过训练好的深度学习模型或者随机森林模型确定目标良率问题堆叠图形和目标量测数据堆叠图形的匹配度数据，本示例实施例不以此为限。
在步骤S130中，计算各所述量测数据堆叠图形与所述目标良率问题堆叠图形对应的相关性数据。
在本公开的一个示例实施例中，相关性数据可以是指能够表征量测数据堆叠图形与目标良率问题堆叠图形之间相关程度的量，例如可以将目标良率问题堆叠图形中主图形区域内的晶片良率作为相关性坐标系的横轴，可以将量测数据堆叠图形中的主图形区域内每个晶片对应的量测参数的平均值作为相关性坐标系的纵轴，并根据得到的相关性坐标系中的曲线计算量测数据堆叠图形与目标良率问题堆叠图形的相关性数据。当然，还可以是其他能够表征量测数据堆叠图形与目标良率问题堆叠图形之间相关性数据的方式，本示例实施例对此不做特殊限定。
具体的，步骤S130可以包括步骤S410至步骤S420，在计算相关性数据之前，可以通过步骤S410至步骤S420对量测数据堆叠图形与目标良率问题堆叠图形进行图形分区：
参考图4所示，步骤S410，对所述目标良率问题堆叠图形以及所述量测数据堆叠图形进行分区，确定主图形区域；
步骤S420，基于所述主图形区域计算所述目标良率问题堆叠图形与各所述量测数据堆叠图形的相关性数据。
其中，分区可以是指根据目标良率问题堆叠图形或者量测数据堆叠图形中的图形特征分布情况进行区域划分的处理过程，例如，可以根据固定大小的图形区域(如10*10的图形区域)对目标良率问题堆叠图形或者量测数据堆叠图形进行分区，也可以根据图形特征分布的密集程度对目标良率问题堆叠图形或者量测数据堆叠图形进行分区，还可以是选择整个目标良率问题堆叠图形或者量测数据堆叠图形进行分区，具体根据实际情况中目标良率问题堆叠图形或者量测数据堆叠图形的图形特征分布情况进行分区，本示例实施例对此不做特殊限定。主图形区域可以是指对目标良率问题堆叠图形或者量测数据堆叠图形进行分区后形成的主要特征分布区域，优选的，目标良率问题堆叠图形中的主图形区域可以与量测数据堆叠图形中的主图形区域相对应(如图形区域形状、图形区域位置、图形区域大小等)。
优选的，可以通过以下步骤计算量测数据堆叠图形与目标良率问题堆叠图形对应的相关性数据：可以确定目标良率问题堆叠图形中的主图形区域对应的晶圆良率数据；确定各量测数据堆叠图形中主图形区域对应的量测参数的统计特征数据；通过主图形区域中的晶圆良率数据以及统计特征数据计算目标良率问题堆叠图形与各量测数据堆叠图形的相关性数据。
举例而言，假设目标良率问题堆叠图形的图形特征主要集中在某一个图形区域(即主图形区域)内，该主图形区域一共有100个晶片die，其中有10个晶片die失效，则该图 形区域其良率就为90％，因此可以将该图形区域对应的良率值作为相关性曲线的纵坐标，然后选定的这100个晶片die在WAT A参数(即此时量测参数为WAT A参数)对应的量测数据堆叠图形中的WAT A值的平均值作为相关性曲线的横坐标，进而根据构建的该相关性曲线确定量测数据堆叠图形与目标良率问题堆叠图形的相关性数据。
在步骤S140中，对所述匹配度数据与所述相关性数据进行加权计算，根据所述加权计算结果确定导致所述目标良率问题的目标量测参数。
在本公开的一个示例实施例中，目标量测参数可以是指经过分析得到的导致目标良率问题堆叠图形中的目标良率问题的参数，例如，假设目标良率问题堆叠图形与WAT A参数对应的量测数据堆叠图形的加权计算结果较高，则认为目标良率问题堆叠图形与WAT A参数较匹配，此时WAT A参数可以认为是导致目标良率问题堆叠图形的目标良率问题的目标量测参数。
具体的，可以根据加权计算结果对各量测参数对应的量测数据堆叠图形进行排序，确定最大的加权计算结果对应的量测数据堆叠图形作为最大相关量测数据堆叠图形，并将最大相关量测数据堆叠图形对应的量测参数作为导致目标良率问题的目标量测参数。
优选的，在某些情况下，匹配度数据或者相关性数据比较高，但是最终的加权计算结果较低，此时并不代表目标良率问题与该量测参数并没有关系，为了进一步保证的得到的目标量测参数的准确性，可以结合匹配度数据、相关性数据以及加权计算结果共同决策最相关的量测参数：
可以根据匹配度数据和相关性数据，分别确定与目标良率问题堆叠图形具有最高匹配度的量测数据堆叠图形和最高相关性的量测数据堆叠图形，然后基于最高匹配度的量测数据堆叠图形和最高相关性的量测数据堆叠图形，确定与目标良率问题堆叠图形对应的最高匹配度量测参数和最高相关性量测参数；进而对最高匹配度量测参数、最高相关性量测参数以及目标量测参数与目标良率问题执行合理性判断流程，以根据合理性判断结果确定与目标良率问题最相关的量测参数。
其中，最高匹配度的量测数据堆叠图形可以是指对匹配度数据进行排序后最大的匹配度数据对应的量测数据堆叠图形，该最高匹配度的量测数据堆叠图形对应的量测参数为最高匹配度量测参数。最高相关性的量测数据堆叠图形可以是指对相关性数据进行排序后最大的相关性数据对应的量测数据堆叠图形，该最高相关性的量测数据堆叠图形对应的量测参数为最高相关性量测参数。
合理性判断流程可以是指预先设置的、用于在匹配度数据、相关性数据以及加权计算结果的数值相差较大时判断最高匹配度量测参数(最高匹配度的量测数据堆叠图形对应的量测参数)、最高相关性量测参数(最高相关性的量测数据堆叠图形对应的量测参数)以及目标量测参数(最大加权计算结果对应的最相关量测数据堆叠图形的量测参数)。
例如，可以在检测到匹配度数据、相关性数据以及加权计算结果之间的差值大于预设差值阈值时，将确定的最高匹配度量测参数、最高相关性量测参数以及目标量测参数发送 给管理控制端，以使管理控制端的相关管理人员以人工的方式进一步判断导致目标良率问题最相关的量测参数；也可以通过预先设置的判断条件对结果进行进一步的判定，如在匹配度数据或者相关性数据大于90％，而加权计算结果小于20％时，此时将匹配度数据或者相关性数据对应的最高匹配度量测参数或者最高相关性量测参数作为最终的分析结果，当然，此处仅是示意性举例说明，本示例实施例对此不做特殊限定。
在本公开的一个示例实施例中，可以通过预选取的目标加权计算方式对匹配度数据与相关性数据进行加权计算。其中，目标加权计算方式可以是指预先根据实验数据确定的加权计算结果能正确反映分析结果的加权计算方式。
具体的，可以在对匹配度数据与相关性数据进行加权计算之前，可以通过图5中的步骤确定目标加权计算方式：
参考图5所示，步骤S510，获取数据库中预存储的样本良率问题堆叠图形以及样本量测数据堆叠图形；其中，所述数据库还包括导致所述样本良率问题堆叠图形的良率问题的样本量测参数；
步骤S520，获取预先设置的多种加权计算方式；
步骤S530，根据多种加权计算方式依次加权计算所述样本良率问题堆叠图形以及所述样本量测数据堆叠图形对应的样本匹配度数据以及样本相关性数据；
步骤S540，依据所述多种加权计算方式对应的加权计算结果和所述样本量测参数，从所述多种加权计算方式中确定目标加权计算方式。
其中，样本目标良率问题堆叠图形以及样本量测数据堆叠图形可以是指预先建立的标准数据库中用于测试的样本数据，此时样本目标良率问题堆叠图形以及样本量测数据堆叠图形可以是百分百对应的关系。然后采用上述方法对样本目标良率问题堆叠图形以及样本量测数据堆叠图形对应的匹配度数据以及相关性数据进行加权计算，如果加权计算结果为80％，则说明对应的加权计算方式较差，则可以选用其他的加权计算方式进行重新计算，直到加权计算结果无限接近于100％，最终将加权计算结果最高的加权计算方式作为目标加权计算方式。
多种加权计算方式可以包括将匹配度数据和相关性数据相乘的加权计算方式，也可以包括对匹配度数据和相关性数据分别配置权重并相加的加权计算方式，具体的加权计算方式可以是在实验过程中不断变换不同的数学公式来优化构建的，因此，本示例实施例对多种加权计算方式不做特殊限定。
通过预先建立的标准数据库中的样本目标良率问题堆叠图形以及样本量测数据堆叠图形，从多种加权计算方式中选择目标加权计算方式，进一步提升目标良率问题堆叠图形与量测数据堆叠图形对应的匹配度数据以及相关性数据的加权计算结果的准确性，提升目标量测参数的准确度，并基于得到的目标量测参数确定导致目标良率问题的工艺流程，提升晶圆的良率。
图6示意性示出了根据本公开一个实施例的确定目标量测参数的应用示意图。
参考图6所示，步骤S610，获取输入的晶圆组以及晶圆组中各晶圆对应的目标良率问题分布图形(例如LotA、waferB、LotC、LotD、waferE…)，并获取输入的不同类型测试下的量测参数：
步骤S620，对目标良率问题分布图形进行堆叠得到目标良率问题堆叠图形，并基于不同类型测试下的量测数据分布图形进行堆叠得到量测数据堆叠图形。具体的，基于输入的不同类型测试下的量测参数，对目标良率问题分布图形进行堆叠得到目标良率问题分布特征的目标良率问题堆叠图形601，例如，目标良率问题堆叠图形601可以包括失效晶片区域602，以及目标良率问题堆叠图形601中除失效晶片区域602之外的合格晶片区域，当然，目标良率问题堆叠图形601仅是示意图，并不应对本示例实施例造成任何特殊限定；
步骤S630，计算匹配度数据以及相关性数据。具体的，将目标良率问题堆叠图形601中的图像特征与不同类型测试下的量测数据堆叠图形(如晶圆组在MET A量测参数下对应的堆叠图形、晶圆组在MET B量测参数下对应的堆叠图形、晶圆组在WAT A量测参数下对应的堆叠图形、晶圆组在WAT B量测参数下对应的堆叠图形、晶圆组在Defect A量测参数下对应的堆叠图形、晶圆组在Defect B量测参数下对应的堆叠图形等)进行特征匹配得到匹配度数据，并构建相关性曲线得到相关性数据；
步骤S640，计算加权分数。具体的，对得到的匹配度数据以及相关性数据进行加权计算，得到加权分数(加权计算结果)，最终基于加权分数确定目标量测参数，或者根据加权分数、匹配度数据和相关性数据对应的目标量测参数、最高匹配度量测参数和最高相关性量测参数的合理性，确定导致目标良率问题堆叠图形601的目标良率问题的最相关的量测参数，最后并可以将分析结果(和分析过程)进行可视化展示。
需要说明的是，尽管在附图中以特定顺序描述了本公开中方法的各个步骤，但是，这并非要求或者暗示必须按照该特定顺序来执行这些步骤，或是必须执行全部所示的步骤才能实现期望的结果。附加的或备选的，可以省略某些步骤，将多个步骤合并为一个步骤执行，以及/或者将一个步骤分解为多个步骤执行等。
此外，在本示例实施例中，还提供了一种数据分析装置。参照图7所示，该数据分析装置700包括：图形获取模块710、关联程度计算模块720以及目标量测参数分析模块730。其中：
图形获取模块710用于获取具有目标良率问题的晶圆组对应的目标良率问题堆叠图形，并获取所述晶圆组在不同类型测试下的量测数据堆叠图形；
关联程度计算模块720用于计算所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据与相关性数据，并对所述匹配度数据与所述相关性数据进行加权计算得到加权计算结果；
目标量测参数分析模块730用于分析所述加权计算结果以确定导致所述目标良率问题的目标量测参数。
在本公开的一种示例性实施例中，基于前述方案，所述关联程度计算模块720还包括：
匹配度数据计算单元，用于将所述目标良率问题堆叠图形与所述量测数据堆叠图形进行图形匹配，得到所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据；
相关性数据计算单元，用于计算各所述量测数据堆叠图形与所述目标良率问题堆叠图形对应的相关性数据。
在本公开的一种示例性实施例中，基于前述方案，所述相关性数据计算单元还包括：
图形分区子单元，用于对所述目标良率问题堆叠图形以及所述量测数据堆叠图形进行分区，确定主图形区域；
晶圆良率确定子单元，用于确定所述目标良率问题堆叠图形中的所述主图形区域对应的晶圆良率数据；
统计特征确定子单元，用于确定各所述量测数据堆叠图形中的所述主图形区域对应的量测参数的统计特征数据；
相关性数据计算子单元，用于通过所述主图形区域中的所述晶圆良率数据以及所述统计特征数据计算所述量测数据堆叠图形与各所述目标良率问题堆叠图形的相关性数据。
在本公开的一种示例性实施例中，基于前述方案，所述图形获取模块710还包括目标良率问题堆叠图形获取单元，所述目标良率问题堆叠图形获取单元被配置为：
获取具有目标良率问题的晶圆组，所述晶圆组中的各晶圆具有目标良率问题分布图形；
将各所述晶圆对应的所述目标良率问题分布图形进行堆叠，得到目标良率问题堆叠图形。
在本公开的一种示例性实施例中，基于前述方案，所述图形获取模块710还包括量测数据堆叠图形获取单元，所述量测数据堆叠图形获取单元被配置为：
获取在不同类型测试下所述晶圆组中各所述晶圆的量测数据集；
基于所述量测数据集生成不同类型测试对应的量测数据分布图形；
将各所述量测数据分布图形进行堆叠，得到所述晶圆组对应的量测数据堆叠图形。
在本公开的一种示例性实施例中，基于前述方案，所述匹配度数据计算单元还包括：
图形特征提取子单元，用于提取所述目标良率问题堆叠图形中的第一图形特征，并提取各所述量测数据堆叠图形中的第二图形特征；
特征向量计算子单元，用于计算所述第一图形特征对应的第一特征向量以及所述第二图形特征对应的第二特征向量；
匹配度计算子单元，用于根据所述第一特征向量以及所述第二特征向量计算得到所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据。
在本公开的一种示例性实施例中，基于前述方案，所述匹配度计算子单元还被配置为：
计算所述第一特征向量以及所述第二特征向量的相似度数据；
将所述相似度数据作为所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据。
在本公开的一种示例性实施例中，基于前述方案，所述目标量测参数分析模块730还包括第一分析单元，所述第一分析单元被配置为：
根据所述加权计算结果对各所述量测数据堆叠图形进行排序，确定最大的所述加权计算结果对应的量测数据堆叠图形作为最相关量测数据堆叠图形；
将所述最相关量测数据堆叠图形对应的量测参数作为导致所述目标良率问题的目标量测参数。
在本公开的一种示例性实施例中，基于前述方案，所述目标量测参数分析模块730还包括第二分析单元，所述第二分析单元被配置为：
根据所述匹配度数据和所述相关性数据，分别确定与所述目标良率问题堆叠图形具有最高匹配度的量测数据堆叠图形和最高相关性的量测数据堆叠图形；
基于所述最高匹配度的量测数据堆叠图形和所述最高相关性的量测数据堆叠图形，确定与所述目标良率问题堆叠图形对应的最高匹配度量测参数和最高相关性量测参数；
对所述最高匹配度量测参数、所述最高相关性量测参数以及所述目标量测参数与所述目标良率问题执行合理性判断流程，以根据合理性判断结果确定与所述目标良率问题最相关的量测参数。
在本公开的一种示例性实施例中，基于前述方案，所述数据分析装置还包括加权计算方式筛选单元，所述加权计算方式筛选单元被配置为：
获取数据库中预存储的样本良率问题堆叠图形以及样本量测数据堆叠图形；其中，所述数据库还包括导致所述样本良率问题堆叠图形的良率问题的样本量测参数；
获取预先设置的多种加权计算方式；
根据多种加权计算方式依次加权计算所述样本良率问题堆叠图形以及所述样本量测数据堆叠图形对应的样本匹配度数据以及样本相关性数据；
依据所述多种加权计算方式对应的加权计算结果和所述样本量测参数，从所述多种加权计算方式中确定目标加权计算方式。
在本公开的一种示例性实施例中，基于前述方案，所述量测参数包括缺陷参数、电性参数和结构参数中的一种或者多种组合。
上述中数据分析装置各模块的具体细节已经在对应的数据分析方法中进行了详细的描述，因此此处不再赘述。
应当注意，尽管在上文详细描述中提及了数据分析装置的若干模块或者单元，但是这种划分并非强制性的。实际上，根据本公开的实施方式，上文描述的两个或更多模块或者单元的特征和功能可以在一个模块或者单元中具体化。反之，上文描述的一个模块或者单元的特征和功能可以进一步划分为由多个模块或者单元来具体化。
此外，在本公开的示例性实施例中，还提供了一种能够实现上述数据分析方法的电子设备。
所属技术领域的技术人员能够理解，本公开的各个方面可以实现为系统、方法或程序 产品。因此，本公开的各个方面可以具体实现为以下形式，即：完全的硬件实施例、完全的软件实施例(包括固件、微代码等)，或硬件和软件方面结合的实施例，这里可以统称为“电路”、“模块”或“系统”。
下面参照图8来描述根据本公开的这种实施例的电子设备800。图8所示的电子设备800仅仅是一个示例，不应对本公开实施例的功能和使用范围带来任何限制。
如图8所示，电子设备800以通用计算设备的形式表现。电子设备800的组件可以包括但不限于：上述至少一个处理单元810、上述至少一个存储单元820、连接不同系统组件(包括存储单元820和处理单元810)的总线830、显示单元840。
其中，所述存储单元存储有程序代码，所述程序代码可以被所述处理单元810执行，使得所述处理单元810执行本说明书上述“示例性方法”部分中描述的根据本公开各种示例性实施例的步骤。例如，所述处理单元810可以执行如图1中所示的步骤S110，获取具有目标良率问题的晶圆组对应的目标良率问题堆叠图形，并获取所述晶圆组在不同类型测试下的量测数据堆叠图形；步骤S120，将所述目标良率问题堆叠图形与各所述量测数据堆叠图形进行图形匹配，得到所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据；步骤S130，计算各所述量测数据堆叠图形与所述目标良率问题堆叠图形对应的相关性数据；步骤S140，对所述匹配度数据与所述相关性数据进行加权计算，根据所述加权计算结果确定导致所述目标良率问题的目标量测参数。
存储单元820可以包括易失性存储单元形式的可读介质，例如随机存取存储单元(RAM)821和/或高速缓存存储单元822，还可以进一步包括只读存储单元(ROM)823。
存储单元820还可以包括具有一组(至少一个)程序模块825的程序/实用工具824，这样的程序模块825包括但不限于：操作系统、一个或者多个应用程序、其它程序模块以及程序数据，这些示例中的每一个或某种组合中可能包括网络环境的实现。
总线830可以为表示几类总线结构中的一种或多种，包括存储单元总线或者存储单元控制器、外围总线、图形加速端口、处理单元或者使用多种总线结构中的任意总线结构的局域总线。
电子设备800也可以与一个或多个外部设备870(例如键盘、指向设备、蓝牙设备等)通信，还可与一个或者多个使得用户能与该电子设备800交互的设备通信，和/或与使得该电子设备800能与一个或多个其它计算设备进行通信的任何设备(例如路由器、调制解调器等等)通信。这种通信可以通过输入/输出(I/O)接口850进行。并且，电子设备800还可以通过网络适配器860与一个或者多个网络(例如局域网(LAN)，广域网(WAN)和/或公共网络，例如因特网)通信。如图所示，网络适配器860通过总线830与电子设备800的其它模块通信。应当明白，尽管图中未示出，可以结合电子设备800使用其它硬件和/或软件模块，包括但不限于：微代码、设备驱动器、冗余处理单元、外部磁盘驱动阵列、RAID系统、磁带驱动器以及数据备份存储系统等。
通过以上的实施例的描述，本领域的技术人员易于理解，这里描述的示例实施例可以 通过软件实现，也可以通过软件结合必要的硬件的方式来实现。因此，根据本公开实施例的技术方案可以以软件产品的形式体现出来，该软件产品可以存储在一个非易失性存储介质(可以是CD-ROM，U盘，移动硬盘等)中或网络上，包括若干指令以使得一台计算设备(可以是个人计算机、服务器、终端装置、或者网络设备等)执行根据本公开实施例的方法。
在本公开的示例性实施例中，还提供了一种计算机可读存储介质，其上存储有能够实现本说明书上述方法的程序产品。在一些可能的实施例中，本公开的各个方面还可以实现为一种程序产品的形式，其包括程序代码，当所述程序产品在终端设备上运行时，所述程序代码用于使所述终端设备执行本说明书上述“示例性方法”部分中描述的根据本公开各种示例性实施例的步骤。
参考图9所示，描述了根据本公开的实施例的用于实现上述数据分析方法的程序产品900，其可以采用便携式紧凑盘只读存储器(CD-ROM)并包括程序代码，并可以在终端设备，例如个人电脑上运行。然而，本公开的程序产品不限于此，在本文件中，可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行系统、装置或者器件使用或者与其结合使用。
所述程序产品可以采用一个或多个可读介质的任意组合。可读介质可以是可读信号介质或者可读存储介质。可读存储介质例如可以为但不限于电、磁、光、电磁、红外线、或半导体的系统、装置或器件，或者任意以上的组合。可读存储介质的更具体的例子(非穷举的列表)包括：具有一个或多个导线的电连接、便携式盘、硬盘、随机存取存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、光纤、便携式紧凑盘只读存储器(CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。
计算机可读信号介质可以包括在基带中或者作为载波一部分传播的数据信号，其中承载了可读程序代码。这种传播的数据信号可以采用多种形式，包括但不限于电磁信号、光信号或上述的任意合适的组合。可读信号介质还可以是可读存储介质以外的任何可读介质，该可读介质可以发送、传播或者传输用于由指令执行系统、装置或者器件使用或者与其结合使用的程序。
可读介质上包含的程序代码可以用任何适当的介质传输，包括但不限于无线、有线、光缆、RF等等，或者上述的任意合适的组合。
可以以一种或多种程序设计语言的任意组合来编写用于执行本公开操作的程序代码，所述程序设计语言包括面向对象的程序设计语言—诸如Java、C++等，还包括常规的过程式程序设计语言—诸如“C”语言或类似的程序设计语言。程序代码可以完全地在用户计算设备上执行、部分地在用户设备上执行、作为一个独立的软件包执行、部分在用户计算设备上部分在远程计算设备上执行、或者完全在远程计算设备或服务器上执行。在涉及远程计算设备的情形中，远程计算设备可以通过任意种类的网络，包括局域网(LAN)或广域网(WAN)，连接到用户计算设备，或者，可以连接到外部计算设备(例如利用因特 网服务提供商来通过因特网连接)。
此外，上述附图仅是根据本公开示例性实施例的方法所包括的处理的示意性说明，而不是限制目的。易于理解，上述附图所示的处理并不表明或限制这些处理的时间顺序。另外，也易于理解，这些处理可以是例如在多个模块中同步或异步执行的。
通过以上的实施例的描述，本领域的技术人员易于理解，这里描述的示例实施例可以通过软件实现，也可以通过软件结合必要的硬件的方式来实现。因此，根据本公开实施例的技术方案可以以软件产品的形式体现出来，该软件产品可以存储在一个非易失性存储介质(可以是CD-ROM，U盘，移动硬盘等)中或网络上，包括若干指令以使得一台计算设备(可以是个人计算机、服务器、触控终端、或者网络设备等)执行根据本公开实施例的方法。
本领域技术人员在考虑说明书及实践这里公开的发明后，将容易想到本公开的其它实施例。本申请旨在涵盖本公开的任何变型、用途或者适应性变化，这些变型、用途或者适应性变化遵循本公开的一般性原理并包括本公开未公开的本技术领域中的公知常识或惯用技术手段。说明书和实施例仅被视为示例性的，本公开的真正范围和精神由权利要求指出。
应当理解的是，本公开并不局限于上面已经描述并在附图中示出的精确结构，并且可以在不脱离其范围进行各种修改和改变。本公开的范围仅由所附的权利要求来限制。

Claims (15)

  1. 一种数据分析方法，其特征在于，包括：
    获取具有目标良率问题的晶圆组对应的目标良率问题堆叠图形，并获取所述晶圆组在不同类型测试下的量测数据堆叠图形；
    将所述目标良率问题堆叠图形与各所述量测数据堆叠图形进行图形匹配，得到所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据；
    计算各所述量测数据堆叠图形与所述目标良率问题堆叠图形对应的相关性数据；
    对所述匹配度数据与所述相关性数据进行加权计算，根据所述加权计算结果确定导致所述目标良率问题的目标量测参数。
  2. 根据权利要求1所述的数据分析方法，其特征在于，所述获取具有目标良率问题的晶圆组对应的目标良率问题堆叠图形，包括：
    获取具有目标良率问题的晶圆组，所述晶圆组中的各晶圆具有目标良率问题分布图形；
    将各所述晶圆对应的所述目标良率问题分布图形进行堆叠，得到目标良率问题堆叠图形。
  3. 根据权利要求1所述的数据分析方法，其特征在于，所述获取所述晶圆组在不同类型测试下的量测数据堆叠图形，包括：
    获取在不同类型测试下所述晶圆组中各所述晶圆的量测数据集；
    基于所述量测数据集生成不同类型测试对应的量测数据分布图形；
    将各所述量测数据分布图形进行堆叠，得到所述晶圆组对应的量测数据堆叠图形。
  4. 根据权利要求1所述的数据分析方法，其特征在于，将所述目标良率问题堆叠图形与各所述量测数据堆叠图形进行图形匹配，生成所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据，包括：
    提取所述目标良率问题堆叠图形中的第一图形特征，并提取各所述量测数据堆叠图形中的第二图形特征；
    计算所述第一图形特征对应的第一特征向量以及所述第二图形特征对应的第二特征向量；
    根据所述第一特征向量以及所述第二特征向量计算得到所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据。
  5. 根据权利要求4所述的数据分析方法，其特征在于，根据所述第一特征向量以及所述第二特征向量计算得到所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据，包括：
    计算所述第一特征向量以及所述第二特征向量的相似度数据；
    将所述相似度数据作为所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据。
  6. 根据权利要求1所述的数据分析方法，其特征在于，计算各所述量测数据堆叠图形与所述目标良率问题堆叠图形对应的相关性数据，包括：
    对所述目标良率问题堆叠图形以及所述量测数据堆叠图形进行分区，确定主图形区域；
    基于所述主图形区域计算所述目标良率问题堆叠图形与各所述量测数据堆叠图形的相关性数据。
  7. 根据权利要求6所述的数据分析方法，其特征在于，基于所述主图形区域计算各所述量测数据堆叠图形与所述目标良率问题堆叠图形的相关性数据，包括：
    确定所述目标良率问题堆叠图形中的所述主图形区域对应的晶圆良率数据；
    确定各所述量测数据堆叠图形中的所述主图形区域对应的量测参数的统计特征数据；
    通过所述主图形区域中的所述晶圆良率数据以及所述统计特征数据计算所述目标良率问题堆叠图形与各所述量测数据堆叠图形的相关性数据。
  8. 根据权利要求1所述的数据分析方法，其特征在于，根据所述加权计算结果确定导致所述目标良率问题的目标量测参数，包括：
    根据所述加权计算结果对各所述量测数据堆叠图形进行排序，将最大的所述加权计算结果对应的量测数据堆叠图形作为最相关量测数据堆叠图形；
    将所述最相关量测数据堆叠图形对应的量测参数作为导致所述目标良率问题的目标量测参数。
  9. 根据权利要求8所述的数据分析方法，其特征在于，在获得所述最相关量测数据堆叠图形，并将所述最相关量测数据堆叠图形所对应的量测参数作为目标量测参数后，还包括：
    根据所述匹配度数据和所述相关性数据，分别确定与所述目标良率问题堆叠图形具有最高匹配度的量测数据堆叠图形和最高相关性的量测数据堆叠图形；
    基于所述最高匹配度的量测数据堆叠图形和所述最高相关性的量测数据堆叠图形，确定与所述目标良率问题堆叠图形对应的最高匹配度量测参数和最高相关性量测参数；
    对所述最高匹配度量测参数、所述最高相关性量测参数以及所述目标量测参数与所述目标良率问题执行合理性判断流程，以根据合理性判断结果确定与所述目标良率问题最相关的量测参数。
  10. 根据权利要求1所述的数据分析方法，其特征在于，对所述匹配度数据与所述相关性数据进行加权计算，包括：
    通过预选取的目标加权计算方式对所述匹配度数据与所述相关性数据进行加权计算。
  11. 根据权利要求10所述的数据分析方法，其特征在于，在对所述匹配度数据与所述相关性数据进行加权计算之前，所述方法还包括：
    获取数据库中预存储的样本良率问题堆叠图形以及样本量测数据堆叠图形；其中，所述数据库还包括导致所述样本良率问题堆叠图形的良率问题的样本量测参数；
    获取预先设置的多种加权计算方式；
    根据多种加权计算方式依次加权计算所述样本良率问题堆叠图形以及所述样本量测数据堆叠图形对应的样本匹配度数据以及样本相关性数据；
    依据所述多种加权计算方式对应的加权计算结果和所述样本量测参数，从所述多种加权计算方式中确定目标加权计算方式。
  12. 根据权利要求1所述的数据分析方法，其特征在于，所述量测参数包括缺陷参数、电性参数和结构参数中的一种或者多种组合。
  13. 一种数据分析装置，其特征在于，包括：
    图形获取模块，用于获取具有目标良率问题的晶圆组对应的目标良率问题堆叠图形，并获取所述晶圆组在不同类型测试下的量测数据堆叠图形；
    关联程度计算模块，用于计算所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据与相关性数据，并对所述匹配度数据与所述相关性数据进行加权计算得到加权计算结果；
    目标量测参数分析模块，用于分析所述加权计算结果以确定导致所述目标良率问题的目标量测参数。
  14. 根据权利要求13所述的数据分析装置，其特征在于，所述关联程度计算模块还包括：
    匹配度数据计算单元，用于将所述目标良率问题堆叠图形与所述量测数据堆叠图形进行图形匹配，得到所述目标良率问题堆叠图形与各所述量测数据堆叠图形对应的匹配度数据；
    相关性数据计算单元，用于计算各所述量测数据堆叠图形与所述目标良率问题堆叠图形对应的相关性数据。
  15. 根据权利要求14所述的数据分析装置，其特征在于，所述相关性数据计算单元还包括：
    图形分区子单元，用于对所述目标良率问题堆叠图形以及所述量测数据堆叠图形进行分区，确定主图形区域；
    晶圆良率确定子单元，用于确定所述目标良率问题堆叠图形中所述主图形区域对应的晶圆良率数据；
    统计特征确定子单元，用于确定各所述量测数据堆叠图形中所述主图形区域对应的量测参数的统计特征数据；
    相关性数据计算子单元，用于通过所述主图形区域中的所述晶圆良率数据以及所述统计特征数据计算所述目标良率问题堆叠图形与各所述量测数据堆叠图形的相关性数据。
PCT/CN2021/100431 2020-09-09 2021-06-16 数据分析方法及装置、电子设备、存储介质 WO2022052548A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/447,410 US20220077005A1 (en) 2020-09-09 2021-09-10 Data analysis method and electronic device, and storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010938941.5A CN114240817A (zh) 2020-09-09 2020-09-09 数据分析方法及装置、电子设备、存储介质
CN202010938941.5 2020-09-09

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US17/447,410 Continuation US20220077005A1 (en) 2020-09-09 2021-09-10 Data analysis method and electronic device, and storage medium

Publications (1)

Publication Number Publication Date
WO2022052548A1 true WO2022052548A1 (zh) 2022-03-17

Family

ID=80630240

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/100431 WO2022052548A1 (zh) 2020-09-09 2021-06-16 数据分析方法及装置、电子设备、存储介质

Country Status (2)

Country Link
CN (1) CN114240817A (zh)
WO (1) WO2022052548A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114398141A (zh) * 2022-03-25 2022-04-26 杭州广立微电子股份有限公司 基于wat测试的晶圆良率数据处理方法和计算机设备
CN114745294A (zh) * 2022-03-30 2022-07-12 深圳市国电科技通信有限公司 网络多节点通信质量评估方法、装置及电子设备
CN114745293A (zh) * 2022-03-30 2022-07-12 深圳市国电科技通信有限公司 网络通信质量评估方法、装置、电子设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6097887A (en) * 1997-10-27 2000-08-01 Kla-Tencor Corporation Software system and method for graphically building customized recipe flowcharts
US20110296362A1 (en) * 2009-02-04 2011-12-01 Tamao Ishikawa Semiconductor defect integrated projection method and defect inspection support apparatus equipped with semiconductor defect integrated projection function
CN103186657A (zh) * 2011-12-28 2013-07-03 敖翔科技股份有限公司 智能型缺陷良率总览接口系统与方法
CN110909968A (zh) * 2018-09-17 2020-03-24 长鑫存储技术有限公司 半导体产品良率分析方法、分析系统及计算机存储介质

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6097887A (en) * 1997-10-27 2000-08-01 Kla-Tencor Corporation Software system and method for graphically building customized recipe flowcharts
US20110296362A1 (en) * 2009-02-04 2011-12-01 Tamao Ishikawa Semiconductor defect integrated projection method and defect inspection support apparatus equipped with semiconductor defect integrated projection function
CN103186657A (zh) * 2011-12-28 2013-07-03 敖翔科技股份有限公司 智能型缺陷良率总览接口系统与方法
CN110909968A (zh) * 2018-09-17 2020-03-24 长鑫存储技术有限公司 半导体产品良率分析方法、分析系统及计算机存储介质

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114398141A (zh) * 2022-03-25 2022-04-26 杭州广立微电子股份有限公司 基于wat测试的晶圆良率数据处理方法和计算机设备
CN114745294A (zh) * 2022-03-30 2022-07-12 深圳市国电科技通信有限公司 网络多节点通信质量评估方法、装置及电子设备
CN114745293A (zh) * 2022-03-30 2022-07-12 深圳市国电科技通信有限公司 网络通信质量评估方法、装置、电子设备及存储介质
CN114745293B (zh) * 2022-03-30 2023-11-17 深圳市国电科技通信有限公司 网络通信质量评估方法、装置、电子设备及存储介质
CN114745294B (zh) * 2022-03-30 2023-12-05 深圳市国电科技通信有限公司 网络多节点通信质量评估方法、装置及电子设备

Also Published As

Publication number Publication date
CN114240817A (zh) 2022-03-25

Similar Documents

Publication Publication Date Title
WO2022052548A1 (zh) 数据分析方法及装置、电子设备、存储介质
WO2022110557A1 (zh) 一种台区户变关系异常诊断方法及装置
CN106650780B (zh) 数据处理方法及装置、分类器训练方法及系统
WO2021077841A1 (zh) 一种基于循环残差网络的信号调制识别方法及装置
US20180082215A1 (en) Information processing apparatus and information processing method
JP6179305B2 (ja) 分類装置、分類方法及び電子設備
CN109783879B (zh) 一种雷达辐射源信号识别效能评估方法及系统
CN113852603B (zh) 网络流量的异常检测方法、装置、电子设备和可读介质
WO2023109527A1 (zh) 窃电行为的检测方法、装置、计算机设备和存储介质
WO2020232898A1 (zh) 文本分类方法、装置、电子设备及计算机非易失性可读存储介质
WO2021174812A1 (zh) 用于画像的数据的清洗方法、装置、介质及电子设备
CN111833175A (zh) 基于knn算法的互联网金融平台申请欺诈行为检测方法
CN112560981A (zh) 生成对抗模型的训练方法、装置、设备、程序和存储介质
CN111738319A (zh) 一种基于大规模样本的聚类结果评价方法及装置
WO2022062968A1 (zh) 一种自训练方法、系统、装置、电子设备及存储介质
CN111863135B (zh) 一种假阳性结构变异过滤方法、存储介质及计算设备
CN114463587A (zh) 一种异常数据检测方法、装置、设备及存储介质
CN113987910A (zh) 一种耦合神经网络与动态时间规划的居民负荷辨识方法及装置
CN110909804B (zh) 基站异常数据的检测方法、装置、服务器和存储介质
CN108133234B (zh) 基于稀疏子集选择算法的社区检测方法、装置及设备
CN113591400B (zh) 一种基于特征相关性分区回归的电力调度监控数据异常检测方法
US20220077005A1 (en) Data analysis method and electronic device, and storage medium
CN114609483A (zh) 一种基于Hilbert变换的GIS局部放电信号特征提取方法
CN115963350A (zh) 一种直流配电网故障定位方法及装置
US11520831B2 (en) Accuracy metric for regular expression

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21865607

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21865607

Country of ref document: EP

Kind code of ref document: A1