WO2023045305A1

WO2023045305A1 - Process parameter root cause positioning method and related device

Info

Publication number: WO2023045305A1
Application number: PCT/CN2022/086785
Authority: WO
Inventors: 方育柯; 薛晓明; 孙崇敬
Original assignee: 成都数之联科技股份有限公司
Priority date: 2021-09-22
Filing date: 2022-04-14
Publication date: 2023-03-30
Also published as: CN113552856A; CN113552856B

Abstract

A process parameter root cause positioning method and a related device. The method comprises: separately dividing a process parameter and a sample output time into multiple subintervals to obtain a plurality of first subintervals and a plurality of second subintervals (step 201); for each first subinterval and each second subinterval, determining the median of each first subinterval to generate first parameter data and determining the median of each second subinterval to generate first time data (step 202); after performing first processing on a sample label and the first parameter data, obtain a correlation coefficient of parameter trend fluctuation (step 203); performing second processing on the sample label, process parameter data and the first time data to obtain a correlation coefficient of parameter time series trend fluctuation (step 204); and on the basis of the correlation coefficient of parameter trend fluctuation and the correlation coefficient of parameter time series trend fluctuation, performing a calculation to obtain a process parameter comprehensive index (step 205). The correlation analysis between parameter trend fluctuation, parameter time series trend fluctuation and label fluctuation is taken into consideration, thereby effectively recognizing the suspicious parameter.

Description

Process parameter root cause location method and related device

Cross References to Related Applications

This application claims the priority of the Chinese patent application with the application number CN202111104580.5 and titled "Process Parameter Root Cause Location Method and Related Devices" filed with the China Patent Office on September 22, 2021, the entire contents of which are hereby incorporated by reference In this application.

technical field

The invention relates to the field of data processing, in particular to a process parameter root cause location method and a related device.

Background technique

In the industrial scene, automatic glass production has been realized. During the production process, the process equipment will automatically save the actual state value of the process parameters corresponding to the glass production process. For large batches of glass with the same process, the parameter settings of the process equipment remain consistent, but fluctuations to varying degrees may lead to poor glass output.

Based on real-time state records of production parameters and well-designed algorithms, effectively mining the correlation between parameter fluctuations and failures is the key to locating the root cause of poor equipment parameters.

At present, one way to determine the root cause is based on the classic Pearson\Kendall correlation coefficient formula to calculate the correlation between the actual state value of the process parameter of the sample and the label value. The other is to do smooth curve fitting on the parameter time series scatterplot, including exponential smoothing method and smoothing spline interpolation fitting, and calculate the sum of squares of the fitting residuals. If the trend of the scatterplot is smoother, the fitting residuals will be smaller , the stronger the correlation between the fluctuation trend of the parameter and the bad glass, the greater the possibility that the corresponding parameter is the root cause.

The factors affecting the root cause in the above two methods are not fully considered, which leads to the inaccurate root cause of the final calculation.

Contents of the invention

The purpose of the present invention includes, for example, providing a process parameter root cause location method and a related device, which can take into account the correlation analysis of parameter trend fluctuations and parameter time series trend fluctuations and label fluctuations, and effectively identify suspicious parameters.

Embodiments of the present invention can be realized like this:

In the first aspect, an embodiment of the present invention provides a method for locating the root cause of a process parameter, the method comprising:

Dividing the process parameters and the sample output time into multiple sub-intervals respectively to obtain multiple first sub-intervals and multiple second sub-intervals;

For each of the first subintervals and each of the second subintervals, determine the median of each of the first subintervals, generate first parameter data, and determine the median of each of the second subintervals Number, generate the first time data;

After performing the first processing on the sample label and the first parameter data, the correlation coefficient of the parameter trend fluctuation is obtained;

performing a second process on the sample label, the process parameter data, and the first time data to obtain a correlation coefficient of time-series trend fluctuation of parameters;

Based on the correlation coefficient of the trend fluctuation of the parameter and the correlation coefficient of the time series trend fluctuation of the parameter, the comprehensive index of the process parameter is calculated.

Through the above technical solution, taking into account the correlation analysis of parameter trend fluctuations and parameter time series trend fluctuations and label fluctuations, thereby effectively identifying suspicious parameters.

In an optional implementation manner, the step of dividing the process parameters and sample output time into multiple subintervals to obtain multiple first subintervals and multiple second subintervals includes:

extracting process parameters and sample labels corresponding to the process parameters;

Using the regression number optimal binning algorithm, segmenting the process parameters into intervals to obtain a plurality of first sub-intervals;

Obtaining the process parameters, corresponding sample labels and corresponding sample output time;

The sample production time is divided into a plurality of second subintervals based on the process parameters and the corresponding sample labels.

Through the above technical solution, it is ensured that the process parameters are divided into small enough, so as to ensure that the sample labels corresponding to different first sub-intervals are sufficiently accurate.

In an optional implementation manner, for each of the first subintervals and each of the second subintervals, determine the median of each of the first subintervals, generate first parameter data, and determine The median of each of the second sub-intervals, the step of generating the first time data includes:

for each of said first subintervals, determining a first median in said first subinterval;

Using the first median as a matching mapping value falling within the range parameter value of the first subinterval, and using the matching mapping value as the transformed first parameter data;

for each of said second subintervals, determining a second median in said second subintervals;

The second median is used as the matching mapping value of the time falling within the range of the second subinterval, and the matching mapping value is used as the transformed first time data.

Through the above technical solution, it is avoided that calculation data exists in the first sub-interval and the second sub-interval, which affects the accuracy of the final result.

In an optional implementation manner, after performing the first processing on the sample label and the first parameter data, the step of obtaining the correlation coefficient of parameter trend fluctuations includes:

grouping the first parameter data;

determining a group value for each of said groupings;

calculating the first average value of the labels corresponding to each of the groups;

calculating the first pearson correlation coefficient of the first average value and the group value;

centralizing the group values of the first parameter data group;

Take the first absolute value group of the first group value after centralization;

Calculating a second pearson correlation coefficient of the first absolute value group and the first average value of the label corresponding to each group.

Through the above technical solution, by calculating the first pearson correlation coefficient between the first average value and the group value, and the second pearson correlation coefficient between the first absolute value group and the first average value of the label corresponding to each group, it can be accurately determined The strength of the correlation between the value range of the process parameter and the quality of the sample label.

In an optional implementation manner, the step of performing the second processing on the sample label, the process parameter data and the first time data to obtain the correlation coefficient of parameter time series trend fluctuations includes:

Sorting the first time data, the process parameters, and the sample labels in chronological order to obtain sorted second time data, second process parameters, and second sample labels;

grouping the second process parameter according to the second time data, and determining the second average value of the process parameter and the third average value of the label for each group;

calculating a third pearson correlation coefficient of said second average value and said third average value;

centering the second average value of the process parameters for each group;

Take the second absolute value group of the second mean value after centralization;

Computing a fourth pearson correlation coefficient of the second set of absolute values and the third mean value.

Through the above technical solution, the correlation strength between the time series trend fluctuation of the parameter and the sample label is calculated.

In an optional embodiment, the step of calculating the comprehensive index of process parameters based on the correlation coefficient of the parameter trend fluctuation and the correlation coefficient of the parameter time series trend fluctuation includes:

The comprehensive index of process parameters is calculated by the following formula:

TPP=max{P*TP,0} ^1/2 ;

Wherein, the P1 is the first pearson correlation coefficient, P2 is the second pearson correlation coefficient; TP1 is the third pearson correlation coefficient, TP2 is the fourth pearson correlation coefficient, and TPP is the comprehensive index of process parameters.

Through the above technical solution, taking into account the correlation analysis of parameter trend fluctuations and parameter time series trend fluctuations and label fluctuations, suspicious parameters can be effectively identified.

In an optional embodiment, the method also includes:

When the process parameter includes multiple types, calculate the type parameter comprehensive index of each type of process parameter;

Sort the size of multiple type parameter comprehensive indicators in descending order.

Through the above-mentioned technical solution, for different types of process parameters, the parameter comprehensive index is calculated and sorted. According to the sorting, it is possible to accurately know which types of process parameters are suspicious parameters, and point out the direction for the subsequent correction of the parameters of the process equipment.

In an optional implementation manner, after the step of sorting the sizes of multiple type parameter comprehensive indicators in descending order, the method further includes:

Obtain the comprehensive index of parameters sorted first;

It is determined that the process parameter corresponding to the obtained parameter comprehensive index is suspicious.

Through the above-mentioned technical solution, it is determined that the process parameters corresponding to the first parameter comprehensive indicators are suspicious, and point out the direction for the subsequent correction of the parameters of the process equipment.

In the second aspect, the embodiment of the present invention also provides a device for locating the root cause of a process parameter, the device comprising:

A division module, configured to divide the process parameters and the sample output time into multiple sub-intervals to obtain multiple first sub-intervals and multiple second sub-intervals;

A generating module, for each of the first subintervals and each of the second subintervals, determine the median of each of the first subintervals, generate first parameter data, and determine the median of each of the second subintervals. The median of the subinterval, generating the first time data;

The first processing module is used to obtain the correlation coefficient of the parameter trend fluctuation after performing the first processing on the sample label and the first parameter data;

The second processing module is configured to perform a second processing on the sample label, the process parameter data and the first time data to obtain a correlation coefficient of time series trend fluctuation of parameters;

The calculation module is used to calculate the comprehensive index of the process parameter based on the correlation coefficient of the parameter trend fluctuation and the correlation coefficient of the parameter time series trend fluctuation.

In a third aspect, an embodiment of the present application provides an electronic device, including a memory and a processor, the memory stores a computer program, and the processor implements the steps of the process parameter root cause location method when executing the computer program .

In a fourth aspect, the embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the process parameter root cause location method are implemented.

The application has the following beneficial effects:

In this application, the process parameters and the sample output time are divided into multiple sub-intervals to obtain multiple first sub-intervals and multiple second sub-intervals. For each first sub-interval and each second sub-interval, determine The median of each first subinterval generates the first parameter data, determines the median of each second subinterval, generates the first time data, performs the first processing on the sample label and the first parameter data, and obtains the parameter trend The correlation coefficient of fluctuations, the sample label, the process parameter data and the first time data are processed for the second time to obtain the correlation coefficient of the parameter time series trend fluctuation, based on the correlation coefficient of the parameter trend fluctuation and the correlation coefficient of the parameter time series trend fluctuation, the process is calculated Comprehensive index of parameters. This application takes into account the correlation analysis between parameter trend fluctuations and parameter time series trend fluctuations and label fluctuations, so as to effectively identify suspicious parameters.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present invention more clearly, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention, and thus It should be regarded as a limitation on the scope, and those skilled in the art can also obtain other related drawings based on these drawings without creative work.

FIG. 1 is a schematic block diagram of an electronic device provided by an embodiment of the present invention;

Fig. 2 is one of the schematic flow charts of a process parameter root cause location method provided by an embodiment of the present invention;

Fig. 3 is the second schematic flow diagram of a process parameter root cause location method provided by an embodiment of the present invention;

Fig. 4 is the third schematic flow diagram of a process parameter root cause location method provided by an embodiment of the present invention;

Fig. 5 is the fourth schematic flow diagram of a process parameter root cause location method provided by an embodiment of the present invention;

Fig. 6 is a schematic structural diagram of a process parameter root cause locating device provided by an embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. The components of the embodiments of the invention generally described and illustrated in the figures herein may be arranged and designed in a variety of different configurations.

Accordingly, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

It should be noted that like numerals and letters denote similar items in the following figures, therefore, once an item is defined in one figure, it does not require further definition and explanation in subsequent figures.

In the description of the present invention, it should be noted that if the orientation or positional relationship indicated by the terms "upper", "lower", "inner" and "outer" appear, it is based on the orientation or positional relationship shown in the drawings, or It is the orientation or positional relationship that the invention product is usually placed in use, and it is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation , and therefore cannot be construed as a limitation of the present invention.

In addition, terms such as "first" and "second" are used only for distinguishing descriptions, and should not be understood as indicating or implying relative importance.

It should be noted that, in the case of no conflict, the features in the embodiments of the present invention may be combined with each other.

However, after a lot of research by the inventors, it is found that the factors affecting the root cause are considered incompletely in the way of the prior art, resulting in inaccurate final calculation of the root cause.

In view of the discovery of the above problems, this embodiment provides a process parameter root cause location method and related devices, which can take into account the correlation analysis between parameter trend fluctuations and parameter timing trend fluctuations and label fluctuations, and effectively identify suspicious parameters. The scheme provided by the embodiment is described in detail.

This embodiment provides an electronic device capable of locating the root cause of a process parameter. In a possible implementation manner, the electronic device may be a user terminal. For example, the electronic device may be, but not limited to, a server, a smart phone, a personal computer (PersonalComputer, PC), a tablet computer, a personal digital assistant (Personal Digital Assistant, PDA), Mobile Internet Device (Mobile Internet Device, MID), etc.

The electronic device may have a device capable of locating the root cause of the process parameter, for example, a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphic Processing Unit, GPU), etc., so as to execute the process parameter provided in this embodiment Root cause location method.

In another possible implementation manner, the electronic device may also be a server capable of communicating with a user terminal. The server can divide the process parameters and sample output time into multiple sub-intervals to obtain multiple first sub-intervals and multiple second sub-intervals; for each of the first sub-intervals and each of the second sub-intervals respectively Sub-intervals, determining the median of each first sub-interval, generating first parameter data, determining the median of each second sub-interval, and generating first time data; performing a first step on the sample label and the first parameter data After processing, the correlation coefficient of the parameter trend fluctuation is obtained; the second processing is performed on the sample label, the process parameter data and the first time data to obtain the correlation coefficient of the parameter time series trend fluctuation; based on the parameter trend fluctuation The correlation coefficient and the correlation coefficient of the time series trend fluctuation of the parameters are calculated to obtain the comprehensive index of the process parameters.

Referring to the schematic structural diagram of the electronic device 100 shown in FIG. 1 . The electronic device 100 includes a process parameter root cause location device 110 , a memory 120 , and a processor 130 .

The memory 120, the processor 130, and various components are electrically connected to each other directly or indirectly to realize data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines. The process parameter root cause location device 110 includes at least one software function module that can be stored in the memory 120 in the form of software or firmware (Firmware) or solidified in the operating system (Operating System, OS) of the service electronic device 100 . The processor 130 is configured to execute executable modules stored in the memory 120 , such as software function modules and computer programs included in the process parameter root cause location device 110 . When the computer-executable instructions in the process parameter root cause locating device 110 are executed by the processor, the process parameter root cause locating method is realized.

Wherein, the memory 120 can be, but not limited to, random access memory (Random Access Memory, RAM), read-only memory (Read Only Memory, ROM), programmable read-only memory (Programmable Read-OnlyMemory, PROM), can Erasable Programmable Read-Only Memory (EPROM), Electric Erasable Programmable Read-Only Memory (EEPROM), etc. Wherein, the memory 120 is used to store a program, and the processor 130 executes the program after receiving an execution instruction.

The processor 130 may be an integrated circuit chip with signal processing capabilities. The above-mentioned processor can be a general-purpose processor, including a central processing unit (Central Processing Unit, referred to as CPU), a network processor (Network Processor, referred to as NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit ( ASIC), Field Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. Various methods, steps, and logic block diagrams disclosed in the embodiments of the present application may be implemented or executed. A general-purpose processor can be a microprocessor, or the processor can be any conventional processor.

Please refer to FIG. 2. FIG. 2 is a flow chart of a process parameter root cause location method applied to the electronic device 100 shown in FIG.

Step 201: Divide the process parameters and the sample output time into multiple sub-intervals to obtain multiple first sub-intervals and multiple second sub-intervals.

Step 202: For each first subinterval and each second subinterval, determine the median of each first subinterval, generate first parameter data, determine the median of each second subinterval, and generate a first time data.

Step 203: After the first processing is performed on the sample label and the first parameter data, the correlation coefficient of the parameter trend fluctuation is obtained.

Step 204: Perform a second process on the sample label, the process parameter data and the first time data to obtain the correlation coefficient of the time series trend fluctuation of the parameters.

Step 205: Based on the correlation coefficient of the parameter trend fluctuation and the correlation coefficient of the parameter time series trend fluctuation, calculate the comprehensive index of the process parameter.

The process parameter is that when the process equipment produces samples, the process equipment has different process parameters based on different samples, wherein the process parameters may include temperature, air pressure, oxygen concentration, and the like. The output time of the sample is that when the process equipment produces the sample, each sample corresponds to a different output time.

The process parameters are divided into a plurality of first subintervals, and the sample output time is divided into a plurality of second subintervals.

The process parameters are divided into intervals, specifically: the process parameters and the sample labels corresponding to the process parameters are extracted; and the process parameters are segmented into intervals through a preset algorithm to obtain multiple first sub-intervals.

Among them, the sample label represents the quality of the produced sample. For example, when the sample label is 0, it indicates that the quality of the sample is good; when the sample label is 1, it indicates that the quality of the sample is poor.

Extract the process parameters and the sample labels corresponding to the process parameters, and divide the process parameters into intervals based on the preset algorithm. It should be noted that there are many kinds of preset algorithms, such as optimal binning algorithm for regression number, chi-square binning, and WOE binning, etc. The embodiment of the present invention adopts the optimal binning algorithm for regression number to perform interval analysis on process parameters. divided.

Based on the process parameters and the corresponding sample labels, after interval division, based on the sample labels, the process parameters using different first sub-intervals can be obtained, and samples of different qualities can be obtained. Well, the quality of the samples obtained by the process parameters corresponding to the other first sub-intervals is poor.

Assuming that the samples are sufficient, you can set the minimum number of first sub-intervals for default segmentation to 20 to ensure that the number of first sub-intervals is sufficient and that the process parameters are divided small enough to ensure that the samples corresponding to different first sub-intervals Labels are accurate enough.

There are many ways to divide the sample output time into multiple sub-intervals. For example, the process parameters, corresponding sample labels and corresponding sample output time are obtained; based on the process parameters and corresponding sample labels, the sample output time is divided into are multiple second subintervals.

Among them, there is a corresponding relationship among process parameters, sample label and sample output time, that is, a sample corresponds to sample label, sample output time and process parameters.

Process parameters and corresponding sample labels and sample output time data, check whether the time range of sample output time exceeds the first preset time length, if it exceeds, divide the sample output time into multiple A second sub-interval, if it does not exceed the first preset duration, check whether the time range exceeds the second preset duration, if it exceeds the second preset duration, divide the sample output duration into If the multiple second sub-intervals do not exceed the second preset duration, the sample output time is divided into multiple second sub-intervals in minutes.

In order to avoid the existence of calculated data in the first subinterval and the second subinterval and affect the accuracy of the final result, the median of each first subinterval is determined for each first subinterval and each second subinterval respectively , generate the first parameter data, determine the median of each second subinterval, and generate the first time data.

After the first processing is performed on the sample label and the first parameter data, a linear correlation coefficient and a nonlinear correlation coefficient of parameter trend fluctuations are obtained. The second processing is performed on the sample label, the process parameter data and the first time data to obtain the linear correlation coefficient and the nonlinear correlation coefficient of the time series trend fluctuation of the parameters. Finally, based on the linear correlation coefficient and nonlinear correlation coefficient of the parameter trend fluctuation and the linear correlation coefficient and nonlinear correlation coefficient of the parameter time series trend fluctuation, the comprehensive index of the process parameter is determined, so that the suspicious parameters of the process equipment are determined according to the comprehensive index of the process parameter. Further, the correction direction for the process equipment is determined.

In order to ensure that the extreme data existing in each first sub-interval does not affect the accuracy of subsequent calculation results, for the above step 202, in another embodiment of the present application, as shown in Figure 3, a process parameter root cause The positioning method specifically includes the following steps:

Step 202-1: For each first subinterval, determine the first median in the first subinterval.

Step 202-2: Use the first median as the matching mapping value of the parameter value falling within the first subinterval range, and use the matching mapping value as the transformed first parameter data.

Step 202-3: For each second subinterval, determine the second median in the second subinterval.

Step 202-4: use the second median as the matching mapping value of the time falling within the second subinterval range, and use the matching mapping value as the transformed first time data.

For example: if a certain first sub-interval is (1, 2, 3, 4, 5), then take the first median in the interval, which is 3, and use (3, 3, 3, 3, 3) as the falling Enter the matching mapping value of the parameter value of the first subinterval range, and use (3, 3, 3, 3, 3) as the transformed first parameter data.

For the second subinterval, the first time data corresponding to the second subinterval is determined in the same manner as above.

To determine the correlation coefficient of parameter trend fluctuations, for the above step 203, in another embodiment of the present application, as shown in Figure 4, a method for locating the root cause of process parameters is provided, which specifically includes the following steps: process parameters that need to be explained It includes multiple types, and the types of process parameters can include temperature process parameters, air pressure process parameters, and oxygen concentration process parameters. The following steps are specific calculation methods for calculating the correlation coefficient of the parameter trend fluctuation of a certain type of process parameter:

Step 203-1: Group the first parameter data into groups.

Step 203-2: Determine the group value of each group.

Step 203-3: Calculate the first average value of the labels corresponding to each group.

Step 203-4: Calculate the first pearson correlation coefficient between the first average value and the group value.

Step 203-5: Centralize the group values of the first parameter data group.

Step 203-6: Obtain the first absolute value group of the first group of values after the centralization process.

Step 203-7: Calculate the second pearson correlation coefficient of the first mean value of the first absolute value group and the label corresponding to each group.

Exemplarily, when the first parameter data is (3,3,3,4,4,4,5,5,5), after grouping the first parameter data, the grouped data is (3,3,3) , (4,4,4), (5,5,5), determine the group value of each group, the group value of group (3,3,3) is 3, the group value of group (4,4,4) The group value is 4, and the group value of group (5,5,5) is 5.

Each data in the first parameter data corresponds to a sample label, and the sample label of the first parameter data is the sample label corresponding to the process parameter data before the first parameter data is processed. For example: the first parameter data is (3,3,3,4,4,4,5,5,5), and the corresponding sample label is (1,1,1,0,0,0,1,1,1 ), the first average value of the sample label corresponding to the group (3,3,3) is 1, the first average value of the sample label corresponding to the group (4,4,4) is 0, and the group (5,5,5) The first mean value of the corresponding sample label is 1.

calculate

The first pearson correlation coefficient of . Pearson Correlation Coefficient (Pearson Correlation Coefficient) is used to measure whether two data sets are on a line, and it is used to measure the linear relationship between fixed-distance variables.

Decentralize the group value of the first parameter data group, that is, determine the mean value of the process parameter before processing, subtract the mean value of the process parameter from the value of different groups, and take the calculated Absolute value, get the first absolute value group. Compute the second pearson correlation coefficient of the first mean of the first absolute value group and the label corresponding to each group.

To determine the correlation coefficient of the time series trend fluctuation of the parameter, for the above step 204, in another embodiment of the present application, as shown in Figure 5, a method for locating the root cause of the process parameter is provided, which specifically includes the following steps:

The following steps are specific calculation methods for calculating the correlation coefficient of the parameter trend fluctuation of a certain type of process parameter, wherein the type of process parameter involved in the calculation in step 204 is consistent with that in step 203 .

Step 204-1: Sort the first time data, process parameters and sample labels in chronological order to obtain sorted second time data, second process parameters and second sample labels.

Step 204-2: Group the second process parameters according to the second time data, and determine the second average value of the process parameters and the third average value of the labels of each group.

Step 204-3: Calculate the third pearson correlation coefficient of the second average value and the third average value.

Step 204-4: Centralize the second average value of the process parameters of each group.

Step 204-5: Take the second absolute value group of the second mean value after the centralization process.

Step 204-6: Calculate the fourth pearson correlation coefficient of the second absolute value group and the third mean value.

The processing method of determining the correlation coefficient of the time series trend fluctuation of parameters is similar to that of steps 203-1 to 203-7, the difference is that the first time data, process parameters and sample labels need to be sorted in chronological order to obtain the sorted The subsequent second time data, second process parameter and second sample label.

Specifically, the calculation methods for calculating the third pearson correlation coefficient and the fourth pearson correlation coefficient will not be repeated again.

Finally, based on the correlation coefficient of the parameter trend fluctuation and the correlation coefficient of the parameter time series trend fluctuation, the comprehensive index of the process parameter is calculated. For example, the comprehensive index of the process parameter is calculated by the following formula:

TPP=max{P*TP,0} ^1/2 ;

Among them, P1 is the first pearson correlation coefficient, P2 is the second pearson correlation coefficient; TP1 is the third pearson correlation coefficient, TP2 is the fourth pearson correlation coefficient, and TPP is the comprehensive index of process parameters.

The type parameter comprehensive index of each type of process parameter is calculated separately, and the sizes of the multiple type parameter comprehensive indexes are sorted in descending order. Acquiring the first-ranked parameter comprehensive index; determining that the process parameter corresponding to the acquired parameter comprehensive index is suspicious.

Please refer to FIG. 6 , this embodiment also provides a process parameter root cause location device 110 applied to the electronic device 100 shown in FIG. 1 , and the process parameter root cause location device 110 includes:

Dividing module 111, is used for dividing process parameter and sample output time into a plurality of sub-intervals respectively, obtains a plurality of first sub-intervals and a plurality of second sub-intervals;

The generation module 112 is used to determine the median of each of the first subintervals for each of the first subintervals and each of the second subintervals, generate first parameter data, and determine the median of each of the first subintervals. The median of the two sub-intervals to generate the first time data;

The first processing module 113 is configured to obtain the correlation coefficient of the parameter trend fluctuation after performing the first processing on the sample label and the first parameter data;

The second processing module 114 is configured to perform a second processing on the sample label, the process parameter data and the first time data to obtain a correlation coefficient of time series trend fluctuation of parameters;

The calculation module 115 is configured to calculate and obtain the comprehensive index of the process parameter based on the correlation coefficient of the parameter trend fluctuation and the correlation coefficient of the parameter time series trend fluctuation.

Optionally, in some possible implementation manners, the dividing module 111 is specifically configured to:

Optionally, in some possible implementation manners, the generating module 112 is specifically configured to:

Optionally, in some possible implementation manners, the first processing module 113 is specifically configured to:

grouping the first parameter data;

determining a group value for each of said groupings;

centralizing the group values of the first parameter data group;

Optionally, in some possible implementation manners, the second processing module 114 is specifically configured to:

centering the second average value of the process parameters for each group;

Optionally, in some possible implementation manners, the computing module 115 is specifically configured to:

TPP=max{P*TP,0} ^1/2 ;

Optionally, in some possible implementation manners, the calculation module 115 is further configured to:

Obtain the comprehensive index of parameters sorted first;

In summary, this application divides the process parameters and sample output time into multiple sub-intervals to obtain multiple first sub-intervals and multiple second sub-intervals, respectively for each first sub-interval and each second sub-interval Two sub-intervals, determine the median of each first sub-interval, generate the first parameter data, determine the median of each second sub-interval, generate the first time data, and perform the first processing on the sample label and the first parameter data Finally, the correlation coefficient of the parameter trend fluctuation is obtained, and the sample label, the process parameter data and the first time data are processed for the second time to obtain the correlation coefficient of the parameter time series trend fluctuation, based on the correlation coefficient of the parameter trend fluctuation and the correlation coefficient of the parameter time series trend fluctuation The coefficient is calculated to obtain the comprehensive index of the process parameters. This application takes into account the correlation analysis between parameter trend fluctuations and parameter time series trend fluctuations and label fluctuations, so as to effectively identify suspicious parameters.

The present application also provides an electronic device 100 , and the electronic device 100 includes a processor 130 and a memory 120 . The memory 120 stores computer-executable instructions, and when the computer-executable instructions are executed by the processor 130, the process parameter root cause location method is realized.

The embodiment of the present application also provides a computer-readable storage medium. The storage medium stores a computer program. When the computer program is executed by the processor 130, the process parameter root cause location method is implemented.

In the embodiments provided in this application, it should be understood that the disclosed devices and methods may also be implemented in other ways. The device embodiments described above are only illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the architecture, functions and possible implementations of devices, methods and computer program products according to multiple embodiments of the present application. operate. In this regard, each block in a flowchart or block diagram may represent a module, program segment, or part of code that includes one or more Executable instructions. It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks in succession may, in fact, be executed substantially concurrently, or they may sometimes be executed in the reverse order, depending upon the functionality involved. It should also be noted that each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations, can be implemented by a dedicated hardware-based system that performs the specified function or action , or may be implemented by a combination of dedicated hardware and computer instructions.

In addition, each functional module in each embodiment of the present application may be integrated to form an independent part, each module may exist independently, or two or more modules may be integrated to form an independent part. If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including Several instructions are used to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disc, etc., which can store program codes. .

It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is a relationship between these entities or operations. There is no such actual relationship or order between them. Furthermore, the term "comprises", "comprises" or any other variation thereof is intended to cover a non-exclusive inclusion such that a process, method, article, or apparatus comprising a set of elements includes not only those elements, but also includes elements not expressly listed. other elements of or also include elements inherent in such a process, method, article, or device. Without further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising said element.

The above are just various implementations of the present application, but the scope of protection of the present application is not limited thereto. Anyone skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application. All should be covered within the scope of protection of this application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

A process parameter root cause location method is characterized in that the method comprises:

Dividing the process parameters and the sample output time into multiple sub-intervals respectively to obtain multiple first sub-intervals and multiple second sub-intervals;

For each of the first subintervals and each of the second subintervals, determine the median of each of the first subintervals, generate first parameter data, and determine the median of each of the second subintervals Number, generate the first time data;

After performing the first processing on the sample label and the first parameter data, the correlation coefficient of the parameter trend fluctuation is obtained;

performing a second process on the sample label, the process parameter data, and the first time data to obtain a correlation coefficient of time-series trend fluctuation of parameters;

Based on the correlation coefficient of the trend fluctuation of the parameter and the correlation coefficient of the time series trend fluctuation of the parameter, the comprehensive index of the process parameter is calculated.
The method according to claim 1, wherein the step of dividing the process parameters and sample output time into a plurality of subintervals to obtain a plurality of first subintervals and a plurality of second subintervals comprises:

extracting process parameters and sample labels corresponding to the process parameters;

Using the regression number optimal binning algorithm, segmenting the process parameters into intervals to obtain a plurality of first sub-intervals;

Obtaining the process parameters, corresponding sample labels and corresponding sample output time;

The sample production time is divided into a plurality of second subintervals based on the process parameters and the corresponding sample labels.
The method according to claim 1, characterized in that, for each of the first subintervals and each of the second subintervals, determine the median of each of the first subintervals, and generate the second A parameter data, the step of determining the median of each second sub-interval and generating the first time data includes:

for each of said first subintervals, determining a first median in said first subinterval;

Using the first median as a matching mapping value falling within the range parameter value of the first subinterval, and using the matching mapping value as the transformed first parameter data;

for each of said second subintervals, determining a second median in said second subintervals;

The second median is used as the matching mapping value of the time falling within the range of the second subinterval, and the matching mapping value is used as the transformed first time data.
The method according to claim 1, wherein the step of obtaining the correlation coefficient of parameter trend fluctuation after the first processing of the sample label and the first parameter data includes:

grouping the first parameter data;

determining a group value for each of said groupings;

calculating the first average value of the labels corresponding to each of the groups;

calculating the first pearson correlation coefficient of the first average value and the group value;

centralizing the group values of the groups of the first parameter data;

Take the first absolute value group of the first group value after centralization;

Calculating a second pearson correlation coefficient of the first absolute value group and the first average value of the label corresponding to each group.
The method according to claim 4, characterized in that the step of performing the second processing on the sample label, the process parameter data and the first time data to obtain the correlation coefficient of parameter time series trend fluctuations includes :

Sorting the first time data, the process parameters, and the sample labels in chronological order to obtain sorted second time data, second process parameters, and second sample labels;

grouping the second process parameter according to the second time data, and determining the second average value of the process parameter and the third average value of the label for each group;

calculating a third pearson correlation coefficient of said second average value and said third average value;

centering the second average value of the process parameters for each group;

Take the second absolute value group of the second mean value after centralization;

Computing a fourth pearson correlation coefficient of the second set of absolute values and the third mean value.
The method according to claim 5, characterized in that the step of calculating the comprehensive index of process parameters based on the correlation coefficient of the parameter trend fluctuation and the correlation coefficient of the parameter time series trend fluctuation includes:

The comprehensive index of process parameters is calculated by the following formula:

TPP=max{P*TP,0} 1/2 ;

Among them, P1 is the first pearson correlation coefficient, P2 is the second pearson correlation coefficient; TP1 is the third pearson correlation coefficient, TP2 is the fourth pearson correlation coefficient, and TPP is the comprehensive index of process parameters.
The method according to claim 1, further comprising:

When the process parameter includes multiple types, calculate the type parameter comprehensive index of each type of process parameter;

Sort the size of multiple type parameter comprehensive indicators in descending order.
The method according to claim 7, characterized in that, after the step of sorting the sizes of multiple type parameter comprehensive indicators in descending order, the method further comprises:

Obtain the comprehensive index of parameters sorted first;

It is determined that the process parameter corresponding to the obtained parameter comprehensive index is suspicious.
A process parameter root cause location device, characterized in that the device comprises:

A division module, configured to divide the process parameters and the sample output time into multiple sub-intervals to obtain multiple first sub-intervals and multiple second sub-intervals;

A generating module, for each of the first subintervals and each of the second subintervals, determine the median of each of the first subintervals, generate first parameter data, and determine the median of each of the second subintervals. The median of the subinterval, generating the first time data;

The first processing module is used to obtain the correlation coefficient of the parameter trend fluctuation after performing the first processing on the sample label and the first parameter data;

The second processing module is configured to perform a second processing on the sample label, the process parameter data and the first time data to obtain a correlation coefficient of time series trend fluctuation of parameters;

The calculation module is used to calculate the comprehensive index of the process parameter based on the correlation coefficient of the parameter trend fluctuation and the correlation coefficient of the parameter time series trend fluctuation.
The device according to claim 9, wherein the dividing module is specifically used for:

extracting process parameters and sample labels corresponding to the process parameters;

Using the regression number optimal binning algorithm, segmenting the process parameters into intervals to obtain a plurality of first sub-intervals;

Obtaining the process parameters, corresponding sample labels and corresponding sample output time;

The sample production time is divided into a plurality of second subintervals based on the process parameters and the corresponding sample labels.
An electronic device, characterized in that it includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the process parameter root cause location method described in any one of claims 1-8 is realized A step of.
A computer-readable storage medium, on which a computer program is stored, characterized in that, when the computer program is executed by a processor, the steps of the process parameter root cause location method described in any one of claims 1-8 are realized.