CN114511028A - Method and device for obtaining normal state control range of control diagram and computer readable medium - Google Patents
Method and device for obtaining normal state control range of control diagram and computer readable medium Download PDFInfo
- Publication number
- CN114511028A CN114511028A CN202210110885.5A CN202210110885A CN114511028A CN 114511028 A CN114511028 A CN 114511028A CN 202210110885 A CN202210110885 A CN 202210110885A CN 114511028 A CN114511028 A CN 114511028A
- Authority
- CN
- China
- Prior art keywords
- graph
- center point
- evaluation value
- peak
- statistical distribution
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- General Factory Administration (AREA)
Abstract
The method for acquiring the normal state control range of the management and control diagram comprises the following steps: receiving original parameter data; acquiring an index map; acquiring a statistical distribution curve graph of the evaluation value in the index graph; respectively obtaining the left central point evaluation value c of the leftmost peak in the statistical distribution curve chartlRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakc(ii) a Estimate c by left center pointlCalculating an evaluation value c less than or equal to the left center point for the mean valuelLeft standard deviation s of all evaluated values oflAnd evaluating the value c with the right center pointrCalculating an evaluation value c greater than or equal to the right center point as an average valuerRight standard deviation s of all evaluated valuesr(ii) a Acquiring a normal state control range [ c ] of the control diagram corresponding to the index diagraml―αsl―ρ,cr+αsr+ρ]Wherein alpha is more than or equal to 3.5 and less than or equal to 4, and rho is a minimum value and is used for adjusting the error of the normal control range. The inventionThe accuracy and the reliability of the setting of the normal state control range of the control diagram are improved.
Description
Technical Field
The present invention relates to the field of semiconductor manufacturing technologies, and in particular, to a method and an apparatus for obtaining a normal state control range of a management and control map, and a computer readable medium.
Background
In the manufacture of semiconductor devices, wafers are typically processed using production stations. In order to monitor the processing of the wafers, one or more sensors are provided in the production machine. The sensor is used for collecting various raw parameter data (raw trace data) of the production machine in the production process. Through the cooperation of the sensor and the software tool, the change curve of various raw parameter data along with time can be collected. According to a plurality of preset Key Steps' windows, KSW, the evaluation values such as the mean value, the standard deviation, the maximum value, the minimum value and the like of various original elaboration data in each main step range can be calculated by applying a statistical method, namely, various evaluation values of each wafer in the specified step range are calculated. According to the time sequence of processing a plurality of wafers by the production machine, a time sequence signal diagram, namely an Indicator diagram (Indicator Chart), which is composed of the single evaluation values of the plurality of wafers can be obtained. And forming a control graph by setting a normal state control range in the index graph. FDC (Fault Detection and Classification) analysis can be performed according to the control graph, so that the health condition of the production machine is monitored.
However, currently, the normal state control range of the control diagram is mainly obtained by means of the rule of thumb, the method needs high labor cost, and the accuracy of the normal state control range obtained by means of the rule of thumb is low.
Therefore, how to improve the accuracy and reliability of obtaining the normal state control range of the management and control diagram and reduce the labor cost is a technical problem to be solved urgently at present.
Disclosure of Invention
The invention provides a method and a device for acquiring a control diagram normal state control range and a computer readable medium, which are used for solving the problems of low accuracy and reliability of the acquisition of the control diagram normal state control range in the prior art and reducing the labor cost of the acquisition of the control diagram normal state control range.
In order to solve the above problem, the present invention provides a method for obtaining a normal state control range of a management and control map, including the following steps:
receiving original parameter data of a production machine in the process of carrying out semiconductor processing on a wafer;
acquiring an index map according to the original parameter data, wherein the index map is a variation curve of the single evaluation value of the production machine along with time;
acquiring a statistical distribution curve graph of the evaluation value in the index graph;
respectively obtaining the left central point evaluation value c of the leftmost peak in the statistical distribution curve chartlRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakc；
Evaluating the value c with the left center pointlCalculating the evaluation value c less than or equal to the left center point as a mean valuelLeft standard deviation s of all the evaluation valueslAnd evaluating the value c with the right center pointrCalculating the evaluation value c of the right center point or more as a mean valuerOf all the evaluation values ofr；
Acquiring a normal state control range [ c ] of a control diagram corresponding to the index diagraml―αsl―ρ,cr+αsr+ρ]Wherein alpha is more than or equal to 3.5 and less than or equal to 4, and rho is a minimum value and is used for adjusting the error of the normal state control range.
Optionally, the specific step of obtaining the statistical distribution graph of the evaluation values in the indicator graph includes:
acquiring an initial statistical distribution curve graph according to data of all the evaluation values in the index graph;
and denoising the initial statistical distribution curve graph to obtain the statistical distribution curve graph.
Optionally, the specific step of performing denoising processing on the initial statistical distribution graph includes:
and removing the data of one or two end preset proportion ranges of the initial statistical distribution curve graph.
Optionally, the specific step of obtaining the statistical distribution graph of the evaluation values in the indicator graph includes:
judging whether the index map has multiple distributions, if so, performing differential processing on the index map;
and acquiring a statistical distribution curve graph of the index graph after the differential processing.
Optionally, the specific step of performing differential processing on the indicator graph includes:
and judging whether a position with amplitude potential difference larger than a first preset value exists in the index map, and if so, performing first-order differential processing on the index map.
Optionally, the specific step of performing differential processing on the indicator graph includes:
and judging whether the position with the amplitude surge larger than a second preset value exists in the index map, and if so, performing second-order differential processing on the index map.
Optionally, the specific step of performing differential processing on the indicator graph includes:
and judging whether the position with the amplitude turning larger than a third preset value exists in the index map, and if so, performing three-order differential processing on the index map.
Optionally, the left center point evaluation values c of the leftmost peak in the statistical distribution graph are obtained respectivelylRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakcThe method comprises the following specific steps:
judging whether the statistical distribution curve graph is a biased peak distribution curve or not, if so, acquiring a peak central point evaluation value in the statistical distribution curve graph, and simultaneously using the peak central point evaluation value as a left central point evaluation value clRight center point evaluation value crAnd an intermediate center point evaluation value cc。
Optionally, the left center point evaluation values c of the leftmost peak in the statistical distribution graph are obtained respectivelylRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakcThe method comprises the following specific steps:
respectively acquiring the leftmost of the statistical distribution curves by adopting a K-mean clustering methodLeft center point evaluation value c of side peaklRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakc。
In order to solve the above problem, the present invention further provides an apparatus for obtaining a normality control range of a management and control map, including a processor, where the processor includes:
the receiving circuit is used for receiving original parameter data of a production machine in the process of carrying out semiconductor process treatment on the wafer;
the storage circuit is used for storing an index map obtained according to the original parameter data, and the index map is a change curve of the single evaluation value of the production machine along with time;
the first calculation circuit is used for acquiring a statistical distribution curve graph of the evaluation value in the index graph;
a second calculating circuit for respectively obtaining the left central point evaluation values c of the leftmost peaks in the statistical distribution graphlRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakc；
A third calculation circuit for evaluating the value c from the left center pointlCalculating the evaluation value c less than or equal to the left center point as a mean valuelLeft standard deviation s of all the evaluation valueslAnd evaluating the value c with the right center pointrCalculating the right central point evaluation value c or more as the mean valuerOf all the evaluation values ofr；
An acquisition circuit for acquiring a normal state control range [ c ] of the control diagram corresponding to the index diagraml―αsl―ρ,cr+αsr+ρ]Wherein alpha is more than or equal to 3.5 and less than or equal to 4, and rho is a minimum value and is used for adjusting the error of the normal state control range.
Optionally, the first computing circuit is configured to obtain an initial statistical distribution graph according to data of all the evaluation values in the indicator diagram, and perform denoising processing on the initial statistical distribution graph to obtain the statistical distribution graph.
Optionally, the first computing circuit is configured to remove data in a preset proportion range at one or two end portions of the initial statistical distribution graph, so as to implement denoising processing on the initial statistical distribution graph.
Optionally, the first calculation circuit is further configured to determine whether the indicator graph has multiple distributions, and if yes, perform differentiation processing on the indicator graph and obtain a statistical distribution curve graph of the indicator graph after the differentiation processing.
Optionally, the first calculating circuit is configured to determine whether a position where the amplitude potential difference is greater than a first preset value exists in the indicator graph, and if so, perform first-order differential processing on the indicator graph.
Optionally, the first computing circuit is configured to determine whether a position where the amplitude spike is greater than a second preset value exists in the indicator graph, and if so, perform second order differential processing on the indicator graph.
Optionally, the first calculating circuit is configured to determine whether a position where the amplitude transition is greater than a third preset value exists in the indicator graph, and if so, perform third-order differential processing on the indicator graph.
Optionally, the second computing circuit is configured to determine whether the statistical distribution graph is a biased peak distribution graph, and if so, obtain a peak center point evaluation value in the statistical distribution graph, and use the peak center point evaluation value as a left center point evaluation value c at the same timelRight center point evaluation value crAnd an intermediate center point evaluation value cc。
Optionally, the second computing circuit is configured to respectively obtain left center point estimated values c of leftmost peaks in the statistical distribution graph by using a K-mean clustering methodlRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakc。
In order to solve the above problem, the present invention also provides a computer readable medium having stored thereon a computer program which, when executed by a processor, implements the method as defined in any one of the above.
According to the method and the device for acquiring the normal state control range of the control graph and the computer readable medium, the normal state control method of the control graph is automatically acquired in an objective calculation mode, and the problems of high labor cost and high subjectivity caused by the fact that the normal state control range is set by manpower subjectively in the prior art are solved. According to the method, the normal state control range is calculated by converting the non-normal distribution statistical distribution curve chart into the normal distribution curve charts, so that on one hand, the loose setting of the normal state control range is avoided, the FDC missing report rate is reduced, and an engineer can know the abnormal condition of a machine table in time; on the other hand, the situation that the normal state control range is set tightly is avoided, and the false alarm rate of the FDC is reduced, so that the manual inspection cost is reduced, and the problem of missing inspection caused by omitting manual inspection due to high false alarm rate of an engineer is also reduced.
Drawings
Fig. 1A is a flowchart of a method for obtaining a normality control range of a management and control diagram according to an embodiment of the present invention;
FIG. 1B is a schematic diagram of a manufacturing tool according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of the conversion of a biased peak distribution to two normal distributions in accordance with an embodiment of the present invention;
FIG. 3 is a schematic representation of the conversion of a multimodal distribution to two normal distributions in an embodiment of the present invention;
FIGS. 4A-4B are schematic diagrams of multiple distributions of index maps in accordance with embodiments of the present invention;
FIGS. 5A-5C are schematic diagrams of differential processing of index maps of different features in accordance with embodiments of the present invention;
FIG. 6 is a graph of the correspondence between the index maps of different differential processes and the statistical distribution curve in the exemplary embodiment of the present invention;
fig. 7 is a block diagram of a device for acquiring a normal state control range of a management and control diagram according to an embodiment of the present invention.
Detailed Description
The following describes specific embodiments of a method and an apparatus for obtaining a normal state control range of a management and control diagram, and a computer readable medium according to the present invention in detail with reference to the accompanying drawings.
The current method for acquiring the normal state control range in the control graph is based on an empirical rule, namely the normal state control range of each control graph is limited to [ mean value-alpha multiplied by standard variance, mean value + alpha multiplied by standard variance ], alpha is a constant and is determined according to artificial experience; or, the normality control range of each control graph is limited to [ average value-average value × k%, average value + average value × k% ], wherein k is a constant and is determined according to manual experience. However, the current method of setting the normal state control range based on the rule of thumb is only suitable for the case where the statistical distribution curve corresponding to the index graph is normally distributed. When the statistical distribution curve graph is in off-peak distribution, due to the fact that the interval definition of a normal value and an off-value cannot be accurately obtained, the normal state control range defined by the empirical rule causes that one side of the control graph is limited too tightly and the false alarm rate is high, and the other side of the control graph is limited too loosely and the false alarm rate is high. When the statistical distribution curve chart is in multimodal distribution, due to the fact that the interval definition of a normal value and a deviation value cannot be accurately obtained, the normal state control range defined by the empirical rule causes that the limits on two sides of the control chart are too loose, and the missing report rate is high. When the index map is in multiple distribution, the normal control range of the control map cannot be directly limited by an empirical rule.
In order to improve the accuracy and reliability of the normal state control range limitation in the control diagram, the present embodiment provides a method for obtaining the normal state control range in the control diagram, and fig. 1A is a flowchart of the method for obtaining the normal state control range in the control diagram in the embodiment of the present invention. As shown in fig. 1A, the method for obtaining the normal state control range of the management and control map includes the following steps:
in step S11, the original parameter data of the wafer processed by the production machine is received.
Step S12, obtaining an index map according to the original parameter data, where the index map is a time-dependent variation curve of the single evaluation value of the production machine.
Specifically, the abscissa of the index map is the wafer numbers arranged in time sequence, and the ordinate is the evaluation value. The evaluation value in the index map in the present embodiment may be an average value, a standard deviation, a maximum value, a minimum value, or the like. The production tool may be a tool for performing any process on a wafer, such as a photolithography tool, a film deposition tool, and the like. Fig. 1B is a schematic structural diagram of a manufacturing tool according to an embodiment of the present invention, in which a solid arrow in fig. 1B represents a transmission path of a wafer entering the manufacturing tool for semiconductor process, and a dotted arrow represents a transmission path of the wafer exiting the manufacturing tool after the semiconductor process is completed. Taking the manufacturing machine shown in fig. 1B as an example, the transfer structure transfers the wafer 20 to the load port 21 of the manufacturing machine, and then enters the transfer chamber 22 inside the manufacturing machine through the load port 21. The wafers 20 entering the transfer chamber 22 are aligned in the orienter 23 inside the production tool, and then are transferred to the processing chamber 24, and the processing chamber 24 is used for performing semiconductor process on the wafers 20. The wafers are transferred out of the manufacturing tool after being processed in the processing chamber 24. One or more sensors are disposed in the processing chamber 24 in the production tool for acquiring various raw parameter data of the production tool during the production process (i.e., during the semiconductor process on the wafer). And sequentially enabling a plurality of wafers to enter the production machine for semiconductor process treatment, and obtaining the index map according to original parameter data of the plurality of wafers acquired by the sensor in the semiconductor process treatment process. And after a normal state control range (including the upper limit of the normal state control range and the lower limit of the normal state control range) is set in the index map, a control map of the production machine can be obtained. FDC analysis can be carried out on the production machine through the control graph, so that the health condition of the production machine is monitored.
Step S13, obtaining a statistical distribution graph of the evaluation values in the indicator graph.
Specifically, the statistical distribution graph is obtained by analyzing the evaluation value data in the index map. The abscissa of the statistical distribution graph is an evaluation value, and the ordinate is the number of occurrences of the evaluation value in the index map or the ratio of the number of occurrences of the evaluation value in the index map. The statistical distribution profile may be a normal distribution curve, a biased peak profile, a multi-peak profile, or a multi-histogram profile.
Optionally, the specific step of obtaining the statistical distribution graph of the evaluation values in the indicator graph includes:
acquiring an initial statistical distribution curve graph according to data of all the evaluation values in the index graph;
and denoising the initial statistical distribution curve graph to obtain the statistical distribution curve graph.
Specifically, since there may be a small peak in the initial statistical distribution graph caused by a deviation value (i.e., an evaluation value generated due to various abnormal conditions, such as an evaluation value generated by restarting a production machine after the production machine stops operating due to preventive maintenance of the machine, manual shutdown, downtime, and the like), in order to avoid that such a small peak affects accurate setting of a normal state control range of the control map, the initial statistical distribution graph may be denoised to remove the small peak caused by the deviation value, so as to form the statistical distribution graph. The specific denoising method can be selected by those skilled in the art according to actual needs, as long as the denoising effect can be achieved.
Optionally, the specific step of performing denoising processing on the initial statistical distribution graph includes:
and removing the data of one or two end preset proportion ranges of the initial statistical distribution curve graph.
Specifically, since the normal state control range in the control map is continuous, small peaks generated by deviation values appear at two ends of an initial statistical distribution curve, and in order to simplify the denoising operation, data in a preset proportion range at one or two ends of the initial statistical distribution curve may be removed, so as to form the statistical distribution curve. Wherein, the preset proportion range can be but is not limited to 5%.
Optionally, the specific step of obtaining the statistical distribution graph of the evaluation values in the indicator graph includes:
judging whether the index map has multiple distributions, if so, performing differential processing on the index map;
and acquiring a statistical distribution curve chart of the index graph after the differential processing.
FIGS. 4A-4B are schematic diagrams of multiple distributions of indices in accordance with embodiments of the present invention. The mean value of the evaluation values of different production periods of the same production machine will produce a shift of higher or lower, and the obtained index map is shown in fig. 4A. In addition, when the production tool is used for a long time, the evaluation value of the same type will be displayed in an upward or downward trend as a whole, and the obtained index graph is shown in fig. 4B. The method includes the steps that firstly, differential processing is carried out on the index graphs of the multiple distributions, and then a statistical distribution curve chart of the index graphs after the differential processing is obtained, so that the statistical distribution graph of the multiple distributions is obtained, in order to convert the statistical distribution graph of the multiple distributions into biased peak distribution or multimodal distribution and remove influences of mean shift of the evaluation values caused by different time periods in the index graphs, normal state control ranges of the control graphs are conveniently set subsequently.
Optionally, the specific step of performing differential processing on the indicator graph includes:
and judging whether a position with amplitude potential difference larger than a first preset value exists in the index map, and if so, performing first-order differential processing on the index map.
Optionally, the specific step of performing differential processing on the indicator graph includes:
and judging whether the position with the amplitude surge larger than a second preset value exists in the index map, and if so, performing second-order differential processing on the index map.
Optionally, the specific step of performing differential processing on the indicator graph includes:
and judging whether the position with the amplitude turn larger than a third preset value exists in the index map, and if so, performing third-order differential processing on the index map.
Fig. 5A to 5C are schematic diagrams of differential processing performed on the index maps with different characteristics in the embodiment of the present invention, and fig. 6 is a corresponding relationship diagram between the index maps with different differential processing and a statistical distribution curve chart in the embodiment of the present invention. Aiming at the multiple distribution index maps with different characteristics, different differentiation modes are adopted, so that the normal state control range of the control map is better limited. For example, as shown in fig. 6, when the original index map composed of the original data is statistically analyzed, the original index map has multiple distributions, and the obtained statistical distribution graph has multiple distributions. When there is a relatively large amplitude position difference position (for example, a position where the amplitude position difference is greater than a first preset value) in the original index map, a first order differential process is performed on the original index map, as shown in fig. 5A. The statistical distribution curve graph of the multi-peak or off-peak distribution is obtained by performing statistical analysis on the indicator graph after the first-order differential processing, as shown in fig. 6. When there is a relatively large amplitude surge position (for example, a position where the amplitude surge is greater than the second preset value) in the original index map, the second order differential processing is performed on the original index map, as shown in fig. 5B. The statistical distribution curve graph of the multi-peak or off-peak distribution is obtained by performing statistical analysis on the indicator graph after the second-order differential processing, as shown in fig. 6. When there is a relatively large amplitude turning position (for example, a position where the amplitude turning is greater than a third preset value) in the original indicator graph, a third order differential process is performed on the original indicator graph, as shown in fig. 5C. By performing statistical analysis on the indicator graph after the third-order differential processing, the statistical distribution curve graph of the multi-peak or partial peak distribution is obtained, as shown in fig. 6. Specific numerical values of the first preset value, the second preset value and the third preset value can be set by a person skilled in the art according to actual needs.
Step S14, obtaining the left center point evaluation value c of the leftmost peak in the statistical distribution curve graphlRightmost side, aEvaluation value c of right center point of peakrAnd the center point estimate c of the center point of the center peakc。
Optionally, the left center point evaluation values c of the leftmost peak in the statistical distribution graph are obtained respectivelylRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakcThe method comprises the following specific steps:
respectively acquiring left central point evaluation values c of leftmost peaks in the statistical distribution curve chart by adopting a K-mean clustering methodlRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakc。
Optionally, the left center point evaluation values c of the leftmost peak in the statistical distribution graph are obtained respectivelylRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakcThe method comprises the following specific steps:
judging whether the statistical distribution curve graph is a biased peak distribution curve or not, if so, acquiring a peak central point evaluation value in the statistical distribution curve graph, and simultaneously using the peak central point evaluation value as a left central point evaluation value clRight center point evaluation value crAnd the center point evaluation value cc。
FIG. 2 is a schematic diagram of the conversion of a biased peak distribution to two normal distributions in an embodiment of the present invention. The statistical distribution graph is described as a biased peak distribution graph. Since there is only one peak in the partial peak profile, as shown in fig. 2, it can be considered that the leftmost peak, the rightmost peak and the middle peak in the partial peak profile coincide with each other, that is, the left center point evaluation value c of the leftmost peaklRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakcThe peak central point evaluation values c in the biased peak distribution curve are all equal. In this embodiment, a normal distribution can also be considered as a subset of a biased peak distribution.
FIG. 3 is an embodiment of the present inventionTo convert a multimodal distribution into two normal distributions. In the following description, the statistical distribution graph is a multimodal distribution graph, and μ in fig. 3 represents a mean value of all evaluation values in the index graph. As shown in fig. 3, the small peak at the right side of the statistical distribution graph of the multi-peak distribution (i.e., the peak indicated by the dotted arrow in fig. 3) is the small peak due to the deviation value, i.e., noise, and therefore, the small peak due to the deviation value at the right end of the statistical distribution curve in the multi-peak distribution shown in fig. 3 can be excluded. Then, respectively obtaining the left central point estimated value c of the leftmost peak in the statistical distribution curve chart by a K-mean clustering method or other statistical algorithmslRight center point evaluation value c of rightmost peakrAnd the evaluation value c of the center point of the middle peakc(not shown in fig. 3).
Step S15, evaluating the value c with the left center pointlCalculating the evaluation value c less than or equal to the left center point as a mean valuelLeft standard deviation s of all the evaluation valueslAnd evaluating the value c with the right center pointrCalculating the evaluation value c of the right center point or more as a mean valuerOf all the evaluation values ofr。
Specifically, all raw data in the statistical distribution curve is X. Based on all the evaluation values c which are less than or equal to the left central point in the statistical distribution curve graphlIs Y, and its evaluation value c is obtained from the left center point by the following formulalLeft side standard deviation s as meanl：
Wherein i is a positive integer.
Based on the statistical distribution curveAll of the right center point evaluation values c in the line graph are greater than or equal torIs Z, and the right side center point evaluation value c thereof is obtained by using the following formularLeft side standard deviation s as meanr：
Wherein i is a positive integer.
For example, for the statistical distribution graph of the biased peak distribution in fig. 2, the peak center point evaluated value c in the statistical distribution graph is obtained, and the peak center point evaluated value is simultaneously used as the left center point evaluated value clRight center point evaluation value crAnd an intermediate center point evaluation value cc. And regarding the data on the left side of the peak center point as one normal distribution, and regarding the data on the right side of the peak center point as the other normal distribution. The solid line in the normal distribution curve of fig. 2 represents actual data in the index graph, and the dotted line in the normal distribution curve represents dummy data (which is not actually present in the index graph) supplemented when calculating data to the left of the peak center point as one normal distribution and calculating data to the right of the peak center point as another normal distribution. When the partial peak distribution curves are regarded as the two normal distribution curves, the proportion of the actual data in each normal distribution curve reaches 99.73%, so that the accuracy and the reliability of the normal control range obtained by regarding the partial peak distribution curves as the two normal distribution curves are high.
For another example, for the statistical distribution graph having a multi-peak distribution as in fig. 3, after determining the leftmost peak, the rightmost peak, and the middle peak, the left center point evaluation values c of the leftmost peak in the statistical distribution graph are directly obtained respectivelylRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakc. Combining the part at the left side of the leftmost peak and the part at the right side of the rightmost peak into a partial peak distribution curve, and then regarding the combined partial peak distribution curve as two normal distributions, namely, regarding the estimated value c of the center point at the left sidelRegarding the left part as a biased peak distribution, and evaluating the right center point value crThe right part is considered as another off-peak distribution. Evaluating the value c with the left center pointlCalculating the evaluation value c less than or equal to the left center point as a mean valuelLeft standard deviation s of all the evaluation valueslAnd evaluating the value c with the right center pointrCalculating the evaluation value c of the right center point or more as a mean valuerOf all the evaluation values ofr. The solid line in the normal distribution curve of FIG. 3 represents actual data according to the index chart, and the dotted line in the normal distribution curve represents the estimated value c for the left center pointlThe data on the left side is calculated as a normal distribution, and the right center point evaluation value c is calculatedrThe data on the right side is regarded as dummy data (which is not actually present in the index map) supplemented when another normal distribution is calculated. The proportion of actual data in each normal distribution curve reaches 99.73%, so that the accuracy and the reliability of a normal state control range obtained by regarding the biased peak distribution curve as two normal distribution curves are high.
Step S16, acquiring a normal state control range [ c ] of the control map corresponding to the index mapl―αsl―ρ,cr+αsr+ρ]Wherein alpha is more than or equal to 3.5 and less than or equal to 4, and rho is a minimum value and is used for adjusting the error of the normal state control range.
In particular, cr+αsr+ ρ is the upper limit of the normal control range in the control map, cl―αslρ is the lower limit of the normal control range in the governing map. Rho is a minimum value, and the specific numerical value can be according to the actual control precision requirement or the production machine table is used for data acquisitionAccuracy determination of the sensors of the set. In an example, the value of ρ may be 0. When α is 3, the expected sample scale range is 0.9973002, occurring 1 out of 370 times of the approximate expected frequency; when α is 3.5, the expected sample scale range is 0.9995347, occurring 1 of 2149 times approximately the expected frequency; when α is 4, the expected sample ratio range is 0.9999367, approximating 1 out of 15787 expected frequencies; when α is 4.5, the expected sample scale range is 0.9999932, occurring 1 of the 147160 approximate expected frequencies.
Moreover, this embodiment further provides a device for obtaining a normal state control range of the management and control map. Fig. 7 is a block diagram of a device for acquiring a normal state control range of a management and control diagram according to an embodiment of the present invention. The apparatus for acquiring a normal state control range of a management map according to the present embodiment may acquire the normal state control range of the management map by using the method for acquiring the normal state control range of the management map shown in fig. 1A, 1B, 2 to 3, 4A to 4B, 5A to 5C, and 6. The device for acquiring the normal state control range of the management and control diagram comprises:
a receiving circuit 75, configured to receive original parameter data of a wafer during a semiconductor process performed by a production machine;
the storage circuit 70 is configured to store an index map obtained according to the original parameter data, where the index map is a time variation curve of the single evaluation value of the production machine;
a first calculation circuit 71, configured to obtain a statistical distribution graph of the evaluation values in the indicator graph;
a second calculating circuit 72 for obtaining the left center point evaluation values c of the leftmost peaks in the statistical distribution graphlRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakc；
A third calculation circuit 73 for evaluating the value c from the left center pointlCalculating the evaluation value c of the left central point or less as the mean valuelLeft standard deviation s of all the evaluation valueslAnd evaluating the value c with the right center pointrCalculating the evaluation value c of the right center point or more as a mean valuerOf all the evaluation values ofr；
An obtaining circuit 74 for obtaining a normal state control range [ c ] of the management and control map corresponding to the index mapl―αsl―ρ,cr+αsr+ρ]Wherein alpha is more than or equal to 3.5 and less than or equal to 4, and rho is a minimum value and is used for adjusting the error of the normal state control range.
Optionally, the first calculating circuit 71 is configured to obtain an initial statistical distribution graph according to data of all the evaluation values in the indicator graph, and perform denoising processing on the initial statistical distribution graph to obtain the statistical distribution graph.
Optionally, the first calculating circuit 71 is configured to remove data in a preset proportion range at one or two end portions of the initial statistical distribution graph, so as to implement denoising processing on the initial statistical distribution graph.
Optionally, the first calculating circuit 71 is further configured to determine whether the indicator graph has multiple distributions, and if yes, perform differentiation processing on the indicator graph, and obtain a statistical distribution curve graph of the indicator graph after the differentiation processing.
Optionally, the first calculating circuit 71 is configured to determine whether a position where the amplitude potential difference is greater than a first preset value exists in the indicator graph, and if so, perform first-order differential processing on the indicator graph.
Optionally, the first calculating circuit 71 is configured to determine whether there is a position in the indicator graph where the amplitude spike is greater than a second preset value, and if so, perform second order differential processing on the indicator graph.
Optionally, the first calculating circuit 71 is configured to determine whether a position where the amplitude transition is greater than a third preset value exists in the indicator graph, and if so, perform third-order differential processing on the indicator graph.
Optionally, the second calculating circuit 72 is configured to determine whether the statistical distribution graph is a biased-peak distribution graph, and if so, obtain a peak center comment in the statistical distribution graphEstimating, using the peak central point estimated value as the left central point estimated value clRight side center point evaluation value crAnd an intermediate center point evaluation value cc。
Optionally, the second calculating circuit 72 is configured to respectively obtain left center point estimated values c of leftmost peaks in the statistical distribution graph by using a K-mean clustering methodlRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakc。
The present embodiment also provides a computer readable medium, on which a computer program is stored, where the computer program is executed by a processor to implement the method for obtaining the normality control range of a management and control map according to any one of the above.
The method and the device for acquiring the normal state control range of the control diagram and the computer readable medium provided by the embodiment automatically acquire the normal state control method of the control diagram in an objective calculation mode, so that the problems of high labor cost and high subjectivity caused by the fact that the normal state control range is set by manpower subjectively in the prior art are solved. According to the method, the normal state control range is calculated by converting the non-normal distribution statistical distribution curve chart into the normal distribution curve charts, so that on one hand, the loose setting of the normal state control range is avoided, the FDC missing report rate is reduced, and an engineer can know the abnormal condition of a machine table in time; on the other hand, the situation that the normal state control range is set tightly is avoided, and the false alarm rate of the FDC is reduced, so that the manual inspection cost is reduced, and the problem of missing inspection caused by omitting manual inspection due to high false alarm rate of an engineer is also reduced.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.
Claims (19)
1. A method for obtaining a normal state control range of a management and control diagram is characterized by comprising the following steps:
receiving original parameter data of a production machine in the process of carrying out semiconductor processing on a wafer;
acquiring an index map according to the original parameter data, wherein the index map is a variation curve of the single evaluation value of the production machine along with time;
acquiring a statistical distribution curve graph of the evaluation value in the index graph;
respectively obtaining the left central point evaluation value c of the leftmost peak in the statistical distribution curve chartlRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakc；
Evaluating the value c with the left center pointlCalculating the evaluation value c less than or equal to the left center point as a mean valuelLeft standard deviation s of all the evaluation valueslAnd evaluating the value c with the right center pointrCalculating the evaluation value c of the right center point or more as a mean valuerOf all the evaluation values ofr；
Acquiring a normal state control range [ c ] of a control diagram corresponding to the index diagraml―αsl―ρ,cr+αsr+ρ]Wherein alpha is more than or equal to 3.5 and less than or equal to 4, and rho is a minimum value and is used for adjusting the error of the normal state control range.
2. The method according to claim 1, wherein the specific step of obtaining the statistical distribution graph of the evaluation values in the indicator graph includes:
acquiring an initial statistical distribution curve graph according to data of all the evaluation values in the index graph;
and denoising the initial statistical distribution curve graph to obtain the statistical distribution curve graph.
3. The method for obtaining the normality control range of the management and control graph according to claim 2, wherein the specific step of performing denoising processing on the initial statistical distribution graph comprises:
and removing the data of one or two end preset proportion ranges of the initial statistical distribution curve graph.
4. The method according to claim 1, wherein the specific step of obtaining the statistical distribution graph of the evaluation values in the indicator graph includes:
judging whether the index graph has multiple distributions, if so, carrying out differential processing on the index graph; and acquiring a statistical distribution curve chart of the index graph after the differential processing.
5. The method for obtaining the normality control range of the management and control map according to claim 4, wherein the specific step of differentiating the index map comprises:
and judging whether a position with an amplitude potential difference larger than a first preset value exists in the index map, and if so, performing first-order differential processing on the index map.
6. The method for obtaining the normality control range of the management and control map according to claim 4, wherein the specific step of differentiating the index map comprises:
and judging whether the position with the amplitude surge larger than a second preset value exists in the index map, and if so, performing second-order differential processing on the index map.
7. The method for obtaining the normality control range of the management and control map according to claim 4, wherein the specific step of differentiating the index map comprises:
and judging whether the position with the amplitude turn larger than a third preset value exists in the index map, and if so, performing third-order differential processing on the index map.
8. The method according to claim 1, wherein the method for obtaining the normalcy control range of the management and control map is characterized by respectively obtaining the normalcy control rangesLeft center point evaluation value c of leftmost peak in statistical distribution curve chartlRight center point evaluation value c of rightmost peakrAnd the evaluation value c of the center point of the middle peakcThe method comprises the following specific steps:
judging whether the statistical distribution curve graph is a biased peak distribution curve or not, if so, acquiring a peak central point evaluation value in the statistical distribution curve graph, and simultaneously using the peak central point evaluation value as a left central point evaluation value clRight side center point evaluation value crAnd an intermediate center point evaluation value cc。
9. The method according to claim 1, wherein left center point evaluation values c of leftmost peaks in the statistical distribution graph are respectively obtainedlRight center point evaluation value c of rightmost peakrAnd the evaluation value c of the center point of the middle peakcThe method comprises the following specific steps:
respectively acquiring the left central point evaluation value c of the leftmost peak in the statistical distribution curve chart by adopting a K-mean clustering methodlRight center point evaluation value c of rightmost peakrAnd the evaluation value c of the center point of the middle peakc。
10. A device for obtaining a normal state control range of a control diagram is characterized by comprising:
the receiving circuit is used for receiving original parameter data of a production machine in the process of carrying out semiconductor process treatment on the wafer;
the storage circuit is used for storing an index map obtained according to the original parameter data, and the index map is a change curve of the single evaluation value of the production machine along with time;
the first calculation circuit is used for acquiring a statistical distribution curve graph of the evaluation value in the index graph;
a second calculation circuit for respectively obtaining the left center point evaluation values c of the leftmost peaks in the statistical distribution graphlRight side of the rightmost peakCenter point evaluation value crAnd the center point estimate c of the center point of the center peakc；
A third calculation circuit for evaluating the value c from the left center pointlCalculating the evaluation value c less than or equal to the left center point as a mean valuelLeft standard deviation s of all the evaluation valueslAnd evaluating the value c with the right center pointrCalculating the right central point evaluation value c or more as the mean valuerOf all the evaluation values ofr；
An acquisition circuit for acquiring a normal state control range [ c ] of the control diagram corresponding to the index diagraml―αsl―ρ,cr+αsr+ρ]Wherein alpha is more than or equal to 3.5 and less than or equal to 4, and rho is a minimum value and is used for adjusting the error of the normal state control range.
11. The apparatus for obtaining a normalcy control range of a management and control graph according to claim 10, wherein the first computing circuit is configured to obtain an initial statistical distribution graph according to data of all the evaluation values in the indicator graph, and perform denoising processing on the initial statistical distribution graph to obtain the statistical distribution graph.
12. The apparatus according to claim 11, wherein the first computing circuit is configured to remove data in a preset proportion range at one or two end portions of the initial statistical distribution graph, so as to implement denoising processing on the initial statistical distribution graph.
13. The apparatus according to claim 10, wherein the first computing circuit is further configured to determine whether there is multiple distribution in the indicator graph, and if so, perform differentiation processing on the indicator graph, and obtain a statistical distribution curve of the indicator graph after the differentiation processing.
14. The apparatus according to claim 13, wherein the first calculating circuit is configured to determine whether there is a position in the indicator graph where the amplitude potential difference is greater than a first preset value, and if so, perform first order differentiation processing on the indicator graph.
15. The apparatus according to claim 13, wherein the first computing circuit is configured to determine whether there is a position in the indicator graph where an amplitude spike is greater than a second preset value, and if so, perform second order differential processing on the indicator graph.
16. The apparatus according to claim 13, wherein the first calculating circuit is configured to determine whether there is a position in the indicator graph where an amplitude transition is greater than a third preset value, and if so, perform third-order differential processing on the indicator graph.
17. The apparatus according to claim 10, wherein the second computing circuit is configured to determine whether the statistical distribution graph is a biased-peak distribution curve, and if so, obtain a peak-center-point evaluation value in the statistical distribution graph, and use the peak-center-point evaluation value as a left-center-point evaluation value clRight center point evaluation value crAnd an intermediate center point evaluation value cc。
18. The apparatus according to claim 10, wherein the second computing circuit is configured to obtain left center point estimated values c of leftmost peaks in the statistical distribution graph by using a K-mean clustering methodlRight center point evaluation value c of rightmost peakrAnd the center point estimate c of the center point of the center peakc。
19. A computer-readable medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210110885.5A CN114511028A (en) | 2022-01-29 | 2022-01-29 | Method and device for obtaining normal state control range of control diagram and computer readable medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210110885.5A CN114511028A (en) | 2022-01-29 | 2022-01-29 | Method and device for obtaining normal state control range of control diagram and computer readable medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114511028A true CN114511028A (en) | 2022-05-17 |
Family
ID=81551764
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210110885.5A Pending CN114511028A (en) | 2022-01-29 | 2022-01-29 | Method and device for obtaining normal state control range of control diagram and computer readable medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114511028A (en) |
-
2022
- 2022-01-29 CN CN202210110885.5A patent/CN114511028A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI720227B (en) | System and method for tool health monitoring and matching | |
| CN107862338B (en) | Marine environment monitoring data quality management method and system based on double inspection method | |
| KR20160140474A (en) | Metrology sampling method with sampling rate decision scheme and computer program product thereof | |
| US8756028B2 (en) | Fault detection method of semiconductor manufacturing processes and system architecture thereof | |
| JP2013073414A (en) | Sensor diagnostic device and sensor diagnostic method for plant | |
| KR20140113153A (en) | Method and System for Statistical Equivalence Test | |
| CN109345060B (en) | Product quality characteristic error traceability analysis method based on multi-source perception | |
| JP2006146459A (en) | Method and system for manufacturing semiconductor device | |
| CN118051863B (en) | Health data acquisition system and method based on digital metering technology | |
| JPH07280603A (en) | Abnormality decision method for machine | |
| CN108182306B (en) | Method for determining degradation failure threshold of abrasive particle characteristic parameters of vehicle power transmission device | |
| CN114384872A (en) | Product development process quality comprehensive management and control system | |
| CN114511028A (en) | Method and device for obtaining normal state control range of control diagram and computer readable medium | |
| JPH03132899A (en) | Service life predicting method for on-line equipment | |
| TWI745912B (en) | Blast furnace abnormality determination device, blast furnace abnormality determination method and blast furnace operation method | |
| US10146215B2 (en) | Monitor system and method for semiconductor processes | |
| CN108459948B (en) | Method for determining failure data distribution type in system reliability evaluation | |
| WO2019041732A1 (en) | Evaluation method and apparatus for manufacturing process capability | |
| CN114397125B (en) | Method and device for checking running state of machine, electronic equipment and storage medium | |
| CN114764550A (en) | Operation method and operation device of failure detection and classification model | |
| CN117892093B (en) | Monitoring parameter data interaction processing method, processing system and processing equipment | |
| CN107209054B (en) | Inspection method and check device | |
| CN117787928B (en) | Automatic assembly equipment assembly efficiency analysis management system based on big data | |
| CN117609710B (en) | Method and device for preventing normal jump of monitoring data from being removed | |
| TWI855746B (en) | Testing system and testing method for sensor data of monitoring system |
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 |