A root cause analysis method, device, electronic device and medium

Abstract

The embodiment of the present application provides a root cause analysis method, device, electronic device and medium, which relates to the technical field of manufacturing industry, and is used to improve the efficiency and accuracy of root cause analysis. In the method, firstly, defect information of multiple products is obtained; multiple products flow through multiple devices in the first process; according to the defect information, the number of positive samples and negative samples in the multiple products is determined; the number of positive samples and negative samples in the products flowing through each of the multiple devices is determined; according to the number of positive samples and negative samples in the products flowing through the multiple devices, and the number of positive samples and negative samples in the products flowing through each device, multiple groups of multiple devices are obtained; wherein each group includes a first group and a second group; the probability value of each group in the multiple groups is determined; the probability values ​​are sorted from small to large, and the N groups with the highest probability values ​​in the multiple groups are determined; the device in the first group or the device in the second group in the N groups is the root cause.

Description

Technical Field

[0001] The present application relates to the technical field of manufacturing industry, and in particular to a root cause analysis method, device, electronic device and medium. Background Art

[0002] With the development and progress of science and technology, the manufacturing process of products such as display panels and organic light-emitting diode (OLED) displays is becoming more and more complicated. Usually, the manufacture of a mobile phone screen or computer screen and other similar products requires hundreds of processes from raw material processing to completion, such as coating, cleaning, etching, etc. In order to improve efficiency, multiple production equipment work simultaneously under each process. Therefore, when a product fails, it is very difficult to find the root cause of the product failure from these hundreds of equipment.

[0003] Root cause analysis is a structured problem-solving method that is used to gradually identify the root cause of a problem and solve it, rather than just focusing on the symptoms of the problem. Different industries have different root cause analysis methods. In the root cause analysis of defects in high-end manufacturing industries such as display panels and OLEDs, the traditional root cause analysis process is lengthy and complicated, and manual analysis is inefficient. It often takes weeks to get results on the root cause of a product defect, and the quality of the results cannot be guaranteed. If the analysis direction is wrong, it needs to be re-analyzed.

[0004] Therefore, the current root cause analysis methods have low accuracy and efficiency. Summary of the invention

[0005] The embodiments of the present application provide a root cause analysis method, device, electronic device, and medium to improve the efficiency and accuracy of root cause analysis.

[0006] In a first aspect, an embodiment of the present application provides a root cause analysis method, comprising:

[0007] Defect information of a plurality of products is obtained; the plurality of products flow through a plurality of devices in a first process; the defect information is determined according to the size of the product and the size of the defective area of ​​the product; according to the defect information, the number of positive samples and the number of negative samples in the plurality of products are determined; wherein the positive sample is a sample whose defect information is greater than or equal to a first threshold, and the negative sample is a sample whose defect information is less than the first threshold; the number of positive samples and the number of negative samples in the product flowing through each of the plurality of devices is determined; according to the number of positive samples and the number of negative samples in the products flowing through the plurality of devices, and the number of positive samples and the number of negative samples in the products flowing through each device, a plurality of groups of the plurality of devices are obtained; wherein each group includes a first group and a second group, and the sum of the number of devices in the first group and the number of devices in the second group is equal to the number of the plurality of devices; the probability value of each group in the plurality of groups is determined; the probability value represents the probability of the group error; the probability values ​​are sorted from small to large, and the N groups with the highest probability values ​​in the plurality of groups are determined; the device in the first group or the device in the second group in the N groups is the root cause.

[0008] Most existing root cause analysis methods analyze the parameters of a single device in a process based on all the history information of the defective product. This root cause analysis method is inefficient. Moreover, this root cause analysis method performs root cause analysis based on a single device, without considering the possibility that multiple devices are the root cause. Based on the above scheme, the efficiency and accuracy of the root cause analysis can be improved by performing root cause analysis on the defect information of the defective product and considering the situation where multiple devices are combined as the root cause.

[0009] A possible implementation method, wherein obtaining defect information of multiple products includes: obtaining historical information of the multiple products; the historical information includes process identification of the products, equipment identification of equipment through which the products flow during the production process, and defect information of the products; and obtaining the defect information of the multiple products based on the historical information.

[0010] Based on the above solution, defect information of defective products is obtained, which can reduce the amount of data used for root cause analysis.

[0011] In a possible implementation, the defect information includes: the number of defects of the product or the defect ratio of the product, and the defect identification of the product; the number of defects is used to indicate the number of first areas where defects occur in the product, and the size of the first areas is predefined; the defect ratio is used to indicate the ratio of the number of first areas where defects occur in the product to the total number of the first areas in the product, and the size of the first areas is predefined.

[0012] Based on the above solution, defective products can be screened according to defect identification and defect ratio or defect number, which can further reduce the amount of data for root cause analysis.

[0013] A possible implementation method is to determine the number of positive samples and the number of negative samples in the multiple products based on the defect information, including: determining the number of positive samples and the number of negative samples in the multiple products through a clustering algorithm based on the defect information.

[0014] Based on the above scheme, the positive and negative samples are divided by clustering method, and the equipment of the same process can be sorted accordingly.

[0015] In a possible implementation, determining the number of positive samples and the number of negative samples in the multiple products includes: determining the ratio of the number of positive samples to the number of negative samples in the multiple products; determining the ratio of the number of positive samples to the total number of samples in the multiple products; the total number of samples is the total number of the multiple products; determining the number of positive samples and the number of negative samples in the products of each device in the multiple devices includes: determining the ratio of the number of positive samples to the number of negative samples in the products of each device; determining the ratio of the number of positive samples to the total number of samples.

[0016] Based on the above scheme, by calculating the ratio of the number of positive samples to the number of negative samples in multiple products, the ratio of the number of positive samples to the total number of samples in multiple products, the ratio of the number of positive samples to the number of negative samples in the products of each device, and the ratio of the number of positive samples to the number of total samples in the products of each device, the situation of defective products produced in the production process of each device can be obtained. Based on this situation, the subsequent grouping of devices will be more accurate.

[0017] A possible implementation method, obtaining multiple groups of the multiple devices according to the number of positive samples and the number of negative samples in the products flowing through the multiple devices, and the number of positive samples and the number of negative samples in the products flowing through each device, includes: determining that the ratio of the number of positive samples in the products of each device to the total number of samples is greater than the product of the ratio of the number of positive samples in the multiple products to the total number of samples and a second threshold; dividing the first i devices of the multiple devices into a first group, and the i+1th to Xth devices into a second group; X is the total number of the multiple devices, and X is an integer greater than or equal to 1; i is an integer from 1 to X-1.

[0018] Based on the above solution, grouping devices in this way can take into account the possibility that each group is a root cause, thereby improving the accuracy of the root cause analysis result.

[0019] A possible implementation method, wherein the probability values ​​are sorted from small to large to determine the top N groups in the multiple groups with the highest probability values, includes: for any group in the N groups, performing the following operations: determining the probability value according to a hypothesis testing method; determining that the grouping is accurate when the probability value is less than a third threshold; the third threshold is a significance level determined according to the hypothesis testing method; calculating a correlation factor α according to the probability value P; calculating a discrete factor β according to the variance d1 of the ratio of the number of positive samples in the products of each device in the first group to the number of total samples, the variance d2 of the ratio of the number of positive samples in the products of each device in the second group to the number of total samples, and the variance d3 of the ratio of the number of positive samples in multiple products to the number of total samples; calculating a length factor γ according to the total number S of samples of all devices in any group; calculating the score of the group according to the correlation factor α, the discrete factor β and the length factor γ; arranging the multiple groups in descending order according to the scores of the groups; determining that multiple devices in the first group among the groups with the N highest scores among the multiple groups are the root cause.

[0020] Based on the above solution, multiple statistical factors are constructed to calculate scores to determine the possibility of device grouping as a root cause, thereby improving the accuracy of root cause analysis.

[0021] A possible implementation includes:

[0022] The correlation factor α satisfies the following formula: α=1-P;

[0023] The discrete factor β satisfies the following formula: β=1-(d1+d2) / d3;

[0024] The length factor γ satisfies the following formula: γ=sigmoid(S);

[0025] The score of the group satisfies the following formula:

[0026] Based on the above solution, calculation according to the above formula shows that the higher the score of the device group, the greater the possibility that it is the root cause.

[0027] A possible implementation method is to obtain historical information of the multiple products; the historical information includes device identifications of the devices through which the products flow during the production process; deduplicate the device identifications; and arrange the device identifications in descending order based on the ratio of the number of positive samples to the number of negative samples in the products of each device.

[0028] Based on the above solution, statistical values ​​of multiple different devices can be obtained by deduplication, and the devices in the first group of groups can be identified as the root cause by arranging them in descending order according to the ratio of the number of positive samples to the number of negative samples in the products of each device.

[0029] In a second aspect, an embodiment of the present application provides a root cause analysis device, including:

[0030] An acquisition unit, configured to acquire defect information of a plurality of products; the plurality of products flow through a plurality of devices in a first process; the defect information is determined based on the size of the products and the size of the area where the defects occur in the products;

[0031] A processing unit, used to perform the following operations: determine the number of positive samples and the number of negative samples in the multiple products according to the defect information; wherein the positive samples are samples whose defect information is greater than or equal to a first threshold, and the negative samples are samples whose defect information is less than the first threshold; determine the number of positive samples and the number of negative samples in the products flowing through each of the multiple devices; obtain multiple groups of the multiple devices according to the number of positive samples and the number of negative samples in the products flowing through the multiple devices, and the number of positive samples and the number of negative samples in the products flowing through each device; wherein each group includes a first small group and a second small group, and the sum of the number of devices in the first small group and the number of devices in the second small group is equal to the number of the multiple devices; determine the probability value of each group in the multiple groups; the probability value represents the probability of the group error; sort the probability values ​​from small to large, and determine the N groups with the highest probability values ​​in the multiple groups; the device in the first small group or the device in the second small group among the N groups is the root cause.

[0032] In a possible implementation, the acquisition unit acquires defect information of multiple products, for:

[0033] Acquire historical information of the plurality of products; the historical information includes process identification of the product, equipment identification of the equipment through which the product flows during the production process, and defect information of the product;

[0034] Defect information of the plurality of products is acquired based on the historical information.

[0035] In a possible implementation, the defect information includes: the number of defects of the product or the defect ratio of the product, and the defect identification of the product; the number of defects is used to indicate the number of first areas where defects occur in the product, and the size of the first areas is predefined; the defect ratio is used to indicate the ratio of the number of first areas where defects occur in the product to the total number of the first areas in the product, and the size of the first areas is predefined.

[0036] In a possible implementation, the processing unit determines the number of positive samples and the number of negative samples in the multiple products based on the defect information, and is specifically used to: determine the number of positive samples and the number of negative samples in the multiple products through a clustering algorithm based on the defect information.

[0037] In a possible implementation, the processing unit determines the number of positive samples and the number of negative samples in the multiple products, specifically for: determining the ratio of the number of positive samples to the number of negative samples in the multiple products; determining the ratio of the number of positive samples to the total number of samples in the multiple products; the total number of samples is the total number of the multiple products; the determination of the number of positive samples and the number of negative samples in the products of each device in the multiple devices includes: determining the ratio of the number of positive samples to the number of negative samples in the products of each device; determining the ratio of the number of positive samples to the number of total samples in the products of each device.

[0038] In a possible implementation, the processing unit obtains multiple groups of the multiple devices according to the number of positive samples and the number of negative samples in the products flowing through the multiple devices, and the number of positive samples and the number of negative samples in the products flowing through each device, and is specifically used to: determine that the ratio of the number of positive samples to the total number of samples in the products of each device is greater than the product of the ratio of the number of positive samples to the total number of samples in the multiple products and a second threshold; divide the first i devices of the multiple devices into a first group, and the i+1th to Xth devices into a second group; X is the total number of the multiple devices, and X is an integer greater than or equal to 1; i is an integer from 1 to X-1.

[0039] In a possible implementation, the processing unit sorts the probability values ​​from small to large, determines the N groups with the top probability values ​​among the multiple groups, and is used to: for any group among the N groups, perform the following operations: determine the probability value according to a hypothesis testing method; determine that the grouping is accurate when the probability value is less than a third threshold; the third threshold is a significance level determined according to the hypothesis testing method; calculate the correlation factor α according to the probability value P; calculate the discrete factor β according to the variance d1 of the ratio of the number of positive samples in the products of each device in the first group to the number of total samples, the variance d2 of the ratio of the number of positive samples in the products of each device in the second group to the number of total samples, and the variance d3 of the ratio of the number of positive samples in multiple products to the number of total samples; calculate the length factor γ according to the total number S of samples of all devices in any group; calculate the score of the group according to the correlation factor α, the discrete factor β and the length factor γ; arrange the multiple groups in descending order according to the scores of the groups; determine that multiple devices in the first group among the groups with the N highest scores among the multiple groups are the root cause.

[0040] A possible implementation method is specifically used for:

[0041] The correlation factor α satisfies the following formula: α=1-P;

[0042] The discrete factor β satisfies the following formula: β=1-(d1+d2) / d3;

[0043] The length factor γ satisfies the following formula: γ=sigmoid(S);

[0044] The score of the group satisfies the following formula:

[0045] A possible implementation method is also used to: obtain historical information of the multiple products; the historical information includes device identifications of the devices through which the products flow during the production process; deduplicate the device identifications; and arrange the device identifications in descending order based on the ratio of the number of positive samples to the number of negative samples in the products of each device.

[0046] In a third aspect, an embodiment of the present application provides an electronic device, the electronic device comprising a processor and a memory;

[0047] The memory is used to store computer programs or instructions;

[0048] The processor is used to execute the computer program or instructions in the memory to utilize the hardware resources in the controller to perform the operation steps of the method in any possible implementation of the first aspect.

[0049] In a fourth aspect, the present application provides a computer-readable storage medium, in which instructions are stored, and when the computer-readable storage medium is run on a computer, the computer executes the above-mentioned methods.

[0050] For each aspect from the second to the fourth aspect and the technical effects that may be achieved by each aspect, please refer to the above description of the technical effects that can be achieved by the first aspect or various possible schemes in the first aspect, and no further details will be given here. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application.

[0052] Figure 1 A schematic diagram of the product production process provided in the embodiment of the present application;

[0053] Figure 2 One of the exemplary flow charts of a root cause analysis method provided in an embodiment of the present application;

[0054] Figure 3 A schematic diagram of the division of equipment groups provided in the embodiment of the present application;

[0055] Figure 4 A root cause analysis device provided in an embodiment of the present application;

[0056] Figure 5 An electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the technical solution of the present application, rather than all of the embodiments. Based on the embodiments recorded in the application documents, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the technical solution of the present application.

[0058] The terms "first" and "second" in the embodiments of the present application are used to distinguish different objects, rather than to describe a specific order. In addition, the term "comprising" and any of their variations are intended to cover non-exclusive protection. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices. "Multiple" in the present application can mean at least two, for example, two, three or more, and the embodiments of the present application are not limited.

[0059] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally indicates that the associated objects before and after are in an "or" relationship.

[0060] See also Figure 1 , is a schematic diagram of the product production process provided in the embodiment of the present application. Figure 1 As shown, it is assumed that the production of a product requires three processes, and each process has three devices working simultaneously. Then the product flows through process 1, process 2 and process 3 in sequence during the production process. First enter process 1, and multiple products are assigned to equipment 1-1, equipment 1-2 and equipment 1-3 for simultaneous processing. After the processing is completed, enter process 2. Similarly, multiple products are assigned to equipment 2-1, equipment 2-2 and equipment 2-3 for simultaneous processing. After the processing is completed, enter process 3. Similarly, multiple products are assigned to equipment 3-1, equipment 3-2 and equipment 3-3 for simultaneous processing.

[0061] At present, most root cause analysis methods analyze the parameters of a single device in a process based on the historical information of defective products. This root cause analysis method is based on all the historical information of defective products, and requires a large amount of data. The root cause analysis process is time-consuming and labor-intensive, and the efficiency is low. Moreover, this root cause analysis method performs root cause analysis based on a single device, without considering the possibility that multiple devices are the root cause, that is, without analyzing the correlation of multiple devices. Therefore, how to improve the efficiency and accuracy of root cause analysis needs to be solved urgently.

[0062] In view of this, an embodiment of the present application provides a root cause analysis method. The method includes: first, obtaining defect information of multiple products flowing through a first process in a database. Secondly, according to the defect information of the multiple products, multiple devices in the first process are grouped to obtain multiple groups. Then, N relatively reliable groups are selected from the multiple groups, and the devices in the N groups are used as root causes.

[0063] See also Figure 2 , which is one of the exemplary flow charts of a root cause analysis method provided in an embodiment of the present application, may include the following process. The method may be applied to a device having a root cause analysis function, such as a processor, a root cause analysis device or a chip.

[0064] S201, the device obtains defect information of multiple products.

[0065] The above-mentioned multiple products flow through multiple devices in the first process. During the production process of the products, the device will record the product identification, product model, process identification of the process through which the product flows, equipment identification of the equipment through which the product flows, and the time when the product flows through each equipment as product history information and save it in the factory database.

[0066] In one example, if a product is found to be defective during or after production, its defect information is added to the product's history information in the database. Among them, the defect information may include the number of defects or the defect ratio and the defect identification. The number of defects or the defect ratio is determined based on the size of the product and the size of the area where the defects occur. For example, the product can be divided into a areas according to its size. When there are b areas with defects in these a areas, the number of defects of the product is b, and the defect ratio of the product is b / a. When the defect ratio is less than the threshold, the product can be used normally although it is defective. When the defect ratio is greater than the threshold, the product cannot be used normally. The defect identification is used to indicate what specific defects the product has and the classification of the defects, which can be represented by an identifier.

[0067] Each product has its own product identification, and the history information of the product can be obtained from the factory database according to the identification. The device can obtain the history information of multiple products that have passed through the first process from the defective product according to the process identification of the first process, and obtain the defect information from the history information.

[0068] Optionally, the device can establish a defective product database by establishing a big data storage database. The defective product database contains the historical information of all defective products. For example, it can be a distributed database Hadoop or a non-relational database Hbase and other databases, which are not limited in this application. The historical information of the defective product is found in the factory database according to the product identification of the defective product, and then the historical information of the defective product is extracted into the defective product database through data warehouse technology (Extract-Transform-Load, ETL) at regular intervals.

[0069] Before performing root cause analysis, the device must screen the product history information in the defective product database. The device can screen the defective product history information according to the production time and defect identification of the defective product to obtain the history information of all products containing a certain specific defect in a fixed time period, and then perform root cause analysis based on the history information of the screened products.

[0070] Based on the above solution, the analysis efficiency can be improved by establishing a defective product database, and the extraction of defective product history information through big data components will not affect the actual production of the factory. Before conducting root cause analysis, information screening should be carried out to narrow the scope of root cause analysis and reduce workload.

[0071] S202, the device determines the number of positive samples and the number of negative samples in multiple products according to the defect information.

[0072] Among them, the positive sample is a sample whose defect information is greater than or equal to the first threshold, and the negative sample is a sample whose defect information is less than the first threshold.

[0073] The device may input the number of defects or defect ratio in the defect information of multiple products containing a certain defect identification into a clustering algorithm to obtain the number of positive samples and the number of negative samples in the multiple products and a first threshold for dividing positive samples and negative samples. For example, the defect ratio of multiple products containing scratch defects may be input into the clustering algorithm to obtain the first threshold for dividing positive samples and negative samples, and the products with a defect ratio greater than the first threshold are positive samples, and the products with a defect ratio less than the first threshold are negative samples.

[0074] Optionally, the clustering algorithm may adopt DBSCAN or mean shift clustering algorithm, etc., which is not limited in this application.

[0075] After determining the number of positive samples and the number of negative samples in the multiple products, the device calculates the ratio of the number of positive samples to the number of negative samples in the multiple products and the ratio of the number of positive samples to the total number of samples.

[0076] S203 , the apparatus determines the number of positive samples and the number of negative samples in the product flowing through each of the multiple devices.

[0077] The device can calculate the number of positive samples and the number of negative samples in the products of each device among multiple devices, and then calculate the ratio of the number of positive samples to the number of negative samples in the products of each device based on the number of positive samples and the number of negative samples in the products of each device, and finally calculate the ratio of the number of positive samples in the products of each device to the total number of samples.

[0078] The device obtains the history information of multiple products and performs a deduplication operation on the device identifiers in the history information of the multiple products, and then arranges the device identifiers of each device in descending order according to the ratio of the number of positive samples to the number of negative samples in the products of each device, that is, arranges each device in order from large to small according to the ratio of the number of positive samples to the number of negative samples in the products of each device. For example, the ratio of the number of positive samples to the number of negative samples of device A is 1, the ratio of the number of positive samples to the number of negative samples of device B is 1.5, and the ratio of the number of positive samples to the number of negative samples of device C is 0.67. Then, device A, device B, and device C are arranged in the order of device B, device A, and device C, and recorded as device 1, device 2, and device 3.

[0079] S204, the apparatus obtains a plurality of groups of the plurality of devices according to the number of positive samples and the number of negative samples in the products passing through the plurality of devices, and the number of positive samples and the number of negative samples in the products passing through each device.

[0080] Each group includes a first subgroup and a second subgroup, and the sum of the number of devices in the first subgroup and the number of devices in the second subgroup is equal to the number of the plurality of devices.

[0081] The device determines whether the ratio of the number of positive samples in the products of each device to the total number of samples is greater than the product of the ratio of the number of positive samples in multiple products to the total number of samples and the second threshold. If the ratio of the number of positive samples in the products of any device to the total number of samples is less than the product of the ratio of the number of positive samples in multiple products to the total number of samples and the second threshold, the calculation is exited. If the device determines that the ratio of the number of positive samples in the products of each device to the total number of samples is greater than the product of the ratio of the number of positive samples in multiple products to the total number of samples and the second threshold, the devices are grouped.

[0082] The second threshold Threshold may satisfy the following formula 1.

[0083]

[0084] Among them, r i It represents the ratio of the number of positive samples in the product of the ith device to the total number of samples, n represents the number of devices in the first process, and 0.25 is a constant obtained based on experience, which is not limited in this application.

[0085] The grouping method is: divide the first i devices into the first group in turn, and the remaining devices into the second group to obtain multiple grouping methods, where i is an integer from 1 to n-1.

[0086] See also Figure 3, is a schematic diagram of the equipment group division provided in the embodiment of the present application. Assuming that the first process includes four devices working simultaneously, when the ratio of the number of positive samples in the products of these four devices to the total number of samples is greater than the product of the ratio of the number of positive samples in multiple products to the total number of samples and the second threshold, three groups are obtained. When i=1, device 1 is divided into the first group, and device 2, device 3 and device 4 are divided into the second group. When i=2, device 1 and device 2 are divided into the first group, and device 3 and device 4 are divided into the second group. When i=3, device 1, device 2 and device 3 are divided into the first group, and device 4 is divided into the second group.

[0087] S205, the device determines a probability value for each of the multiple groups; the probability value represents the probability of a group error.

[0088] The device performs a hypothesis test on the first group device combination and the second group device combination for each group to determine the rationality of the grouping and obtain a probability value indicating a grouping error. The hypothesis test method may be a chi-square test, which is not limited in this application.

[0089] S206, the device sorts the probability values ​​from small to large, and determines N groups with top probability values ​​among the multiple groups; the device in the first group or the device in the second group among the N groups is the root cause.

[0090] Since a smaller probability value indicates a smaller probability that an error may occur in a group, multiple groups are sorted in ascending order according to the probability values, and the device in the first group or the device in the second group among the top N groups with the highest probability values ​​is the root cause.

[0091] Based on the above solution, by taking into account the grouping of different devices, the accuracy of the root cause analysis result can be improved.

[0092] One possible implementation method is to set the ratio of the number of positive samples in multiple products to the total number of samples as R, and then determine whether r1 is greater than R*Threshold, whether r2 is greater than R*Threshold, and so on until r n Is it greater than R*Threshold? If any device has a situation where r is less than or equal to R*Threshold, the calculation is exited. n If all are greater than R*Threshold, then set i=1, that is, the first device is divided into the first group, and the remaining devices are the second group. Perform the following operations for this grouping method:

[0093] The device calculates the probability value P of grouping error by a hypothesis testing method, and determines that the grouping is accurate when P is less than a third threshold value. The third threshold value is a significance level, and the significance level can be 0.05, etc., which is not limited in this application.

[0094] Specifically, firstly, the correlation factor α is calculated according to the probability value P of the grouping error, as shown in Formula 2:

[0095] α=1-P (Formula 2)

[0096] Then, the variance d1 of the ratio of the number of positive samples to the total number of samples in the products of each device in the first group, the variance d2 of the ratio of the number of positive samples to the total number of samples in the products of each device in the second group, and the variance d3 of the ratio of the number of positive samples to the total number of samples in multiple products are calculated respectively. The discrete factor β is calculated by d1, d2 and d3, as shown in Formula 3:

[0097] β=1-(d1+d2) / d3 (Formula 3)

[0098] The length factor γ is obtained by performing a sigmoid transformation on the total number of product samples S in all devices in the group. The calculation formula is shown in Formula 4:

[0099] γ=sigmoid(S) (Formula 4)

[0100] The score of the group is calculated based on the above correlation factor, discrete factor and length factor. The calculation formula is shown in Formula 5:

[0101]

[0102] After obtaining the score when i=1, continue to set i=2 and repeat the above operations of calculating the probability value of packet error and calculating the score until i takes all integers between 1 and n-1.

[0103] After obtaining the scores of all the groups in the first process, the device sorts the groups from large to small according to the scores, and then determines that multiple devices in the first group or the second group of the top N groups in the multiple groups in the first process are the root causes.

[0104] Since each device is arranged in descending order according to the ratio of the number of positive samples to the number of negative samples of the device in the embodiment of the present application, the device determines in the first process in the first process that multiple devices in the first group of the top N groups are the root cause.

[0105] Optionally, if each device is arranged in ascending order according to the ratio of the number of positive samples to the number of negative samples of the device, the second group in the group with the highest score among the multiple groups can be obtained as the root cause.

[0106] Another possible implementation method is to first determine r1, r2, ... nAre they all greater than R*Threshold? If any device has a situation where r is less than or equal to R*Threshold, the calculation is exited. n If all are greater than R*Threshold, all the equipment of the first process are grouped according to the grouping method of S204, and then the scores of each group of the first process are calculated, and then the groups are sorted from large to small according to the scores. Finally, the device determines that multiple equipment in the first group of the N groups with the highest scores among the multiple groups in the first process are the root causes. The score calculation method is as described above and will not be repeated here.

[0107] In one example, the device can repeat S201-S206 for the second process until all processes of the production process are traversed to obtain the root cause equipment grouping corresponding to each process, and then the first N groups determined for each process are arranged in descending order according to the root cause score, and the top M groups with the highest scores can be located, and the first group of the M groups is determined to be the root cause. Among them, N and M are both preset according to demand or experience, and N of each process can take different values ​​or the same value. This application does not limit this.

[0108] Based on the same concept of the above method, see Figure 4 , is a root cause analysis device 400 provided in an embodiment of the present application, the device 400 can execute each step in the above method, and in order to avoid repetition, it will not be described in detail here. The device 400 includes an acquisition unit 401 and a processing unit 402.

[0109] In one scenario:

[0110] The acquisition unit 401 is used to acquire defect information of a plurality of products; the plurality of products flow through a plurality of devices in a first process; the defect information is determined according to the size of the product and the size of the area where the defect occurs in the product;

[0111] The processing unit 402 is configured to perform the following operations:

[0112] Determine the number of positive samples and the number of negative samples in the multiple products according to the defect information; wherein the positive samples are samples whose defect information is greater than or equal to a first threshold, and the negative samples are samples whose defect information is less than the first threshold;

[0113] determining a number of positive samples and a number of negative samples in a product flowing through each of the plurality of devices;

[0114] According to the number of positive samples and the number of negative samples in the products flowing through the multiple devices, and the number of positive samples and the number of negative samples in the products flowing through each device, a plurality of groups of the multiple devices are obtained; wherein each group includes a first small group and a second small group, and the sum of the number of devices in the first small group and the number of devices in the second small group is equal to the number of the multiple devices;

[0115] Determine a probability value for each of the plurality of packets; the probability value represents a probability that the packet is erroneous;

[0116] The probability values ​​are sorted from small to large, and N groups with top probability values ​​among the multiple groups are determined; the device in the first group or the device in the second group among the N groups is the root cause.

[0117] In a possible implementation, the acquisition unit 401 acquires defect information of multiple products and is used to: acquire historical information of the multiple products; the historical information includes the process identification of the product, the equipment identification of the equipment through which the product flows during the production process, and the defect information of the product; and acquire the defect information of the multiple products based on the historical information.

[0118] In a possible implementation, the defect information includes: the number of defects of the product or the defect ratio of the product, and the defect identification of the product; the number of defects is used to indicate the number of first areas where defects occur in the product, and the size of the first areas is predefined; the defect ratio is used to indicate the ratio of the number of first areas where defects occur in the product to the total number of the first areas in the product, and the size of the first areas is predefined.

[0119] In a possible implementation, the processing unit 402 determines the number of positive samples and the number of negative samples in the multiple products based on the defect information, and is specifically used to: determine the number of positive samples and the number of negative samples in the multiple products through a clustering algorithm based on the defect information.

[0120] In a possible implementation, the processing unit 402 determines the number of positive samples and the number of negative samples in the multiple products, and is specifically used to: determine the ratio of the number of positive samples to the number of negative samples in the multiple products; determine the ratio of the number of positive samples to the total number of samples in the multiple products; the total number of samples is the total number of the multiple products; the determination of the number of positive samples and the number of negative samples in the products flowing through each device in the multiple devices includes: determining the ratio of the number of positive samples to the number of negative samples in the products of each device; determining the ratio of the number of positive samples to the number of total samples in the products of each device.

[0121] In a possible implementation, the processing unit 402 obtains multiple groups of the multiple devices according to the number of positive samples and the number of negative samples in the products flowing through the multiple devices, and the number of positive samples and the number of negative samples in the products flowing through each device, and is specifically used to: determine that the ratio of the number of positive samples in the products of each device to the total number of samples is greater than the product of the ratio of the number of positive samples in the multiple products to the total number of samples and a second threshold; divide the first i devices of the multiple devices into a first group, and the i+1th to Xth devices into a second group; X is the total number of the multiple devices, and X is an integer greater than or equal to 1; i is an integer from 1 to X-1.

[0122] In a possible implementation, the processing unit 402 sorts the probability values ​​from small to large, determines N groups with the highest probability values ​​among the multiple groups, and is specifically used to: for any group among the N groups, perform the following operations: determine the probability value according to a hypothesis testing method; determine that the grouping is accurate when the probability value is less than a third threshold; the third threshold is a significance level determined according to the hypothesis testing method; calculate a correlation factor α according to the probability value P; and calculate a variance d1 of the ratio of the number of positive samples to the number of total samples in the products of each device in the first group. The discrete factor β is calculated based on the variance d2 of the ratio of the number of positive samples to the total number of samples in the products of each device in the second group and the variance d3 of the ratio of the number of positive samples to the total number of samples in multiple products; the length factor γ is calculated based on the total number S of samples of all devices in any of the groups; the score of the group is calculated based on the correlation factor α, the discrete factor β and the length factor γ; the multiple groups are arranged in descending order according to the scores of the groups; and it is determined that multiple devices in the first group among the groups with the N highest scores among the multiple groups are the root cause.

[0123] In a possible implementation, the processing unit 402 is specifically configured to:

[0124] The correlation factor α satisfies the following formula: α=1-P;

[0125] The discrete factor β satisfies the following formula: β=1-(d1+d2) / d3;

[0126] The length factor γ satisfies the following formula: γ=sigmoid(S);

[0127] The score of the group satisfies the following formula:

[0128] In a possible implementation, the processing unit 402 is also used to: obtain historical information of the multiple products; the historical information includes device identifications of the devices through which the products flow during the production process; deduplicate the device identifications; and arrange the device identifications in descending order according to the ratio of the number of positive samples to the number of negative samples in the products of each device.

[0129] Based on the same concept of the above method, see Figure 5 , is an electronic device provided in an embodiment of the present application, and the electronic device includes a processor 501 and a memory 502. The memory 502 is used to store computer instructions, and the processor 501 is connected to the memory and is used to execute the computer instructions in the memory, and implements the steps of any of the above methods when executing the computer instructions.

[0130] The embodiment of the present application also provides a computer-readable storage medium on which computer instructions are stored. When the computer instructions are executed on a computer, the computer executes the steps of any of the above methods.

[0131] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0132] Although the specific embodiments of the present application are described above, it should be understood by those skilled in the art that these are merely examples, and the scope of protection of the present application is limited by the attached claims. Those skilled in the art may make various changes or modifications to these embodiments without departing from the principles and essence of the present application, but these changes and modifications all fall within the scope of protection of the present application. Although the preferred embodiments of the present application have been described, once those skilled in the art are aware of the basic creative concepts, additional changes and modifications may be made to these embodiments. Therefore, the attached claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0133] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A root cause analysis method, characterized in that: include: Get defect information for multiple products; The plurality of products flow through a plurality of equipment in a first process; The defect information is determined based on the size of the product and the size of the area where the defect occurs in the product; According to the defect information, determine the number of positive samples and the number of negative samples in the multiple products, and according to the number of positive samples and the number of negative samples in the multiple products, determine the ratio of the number of positive samples to the total number of samples in the multiple products; wherein the positive samples are samples whose defect information is greater than or equal to a first threshold, the negative samples are samples whose defect information is less than the first threshold, and the total number of samples is the total number of the multiple products; Determine the number of positive samples and the number of negative samples in the product flowing through each of the multiple devices, and determine the ratio of the number of positive samples in the product of each of the multiple devices to the total number of samples based on the number of positive samples and the number of negative samples in the product of each of the multiple devices; Determine that the ratio of the number of positive samples to the total number of samples in the product of each device is greater than the product of the ratio of the number of positive samples to the total number of samples in multiple products and a second threshold; The first i devices of the plurality of devices are divided into a first group, and the i+1th to Xth devices are divided into a second group; X is the total number of the plurality of devices, X is an integer greater than or equal to 1; i is an integer from 1 to X-1; Determine a probability value for each of the plurality of packets; the probability value represents a probability that the packet is erroneous; The probability values ​​are sorted from small to large, and N groups with top probability values ​​among the multiple groups are determined; the device in the first group or the device in the second group among the N groups is the root cause.

2. The method according to claim 1, characterized in that The obtaining of defect information of multiple products includes: Acquire historical information of the plurality of products; the historical information includes process identification of the product, equipment identification of the equipment through which the product flows during the production process, and defect information of the product; Defect information of the plurality of products is acquired based on the historical information.

3. The method according to claim 1 or 2, characterized in that: The defect information includes: The number of defects of the product or the defect ratio of the product, and the defect identification of the product; the number of defects is used to indicate the number of first areas where defects occur in the product, and the size of the first areas is predefined; the defect ratio is used to indicate the ratio of the number of first areas where defects occur in the product to the total number of the first areas in the product, and the size of the first areas is predefined.

4. The method according to claim 1 or 2, characterized in that: Determining the number of positive samples and the number of negative samples in the multiple products according to the defect information includes: According to the defect information, the number of positive samples and the number of negative samples in the multiple products are determined by a clustering algorithm.

5. The method according to claim 1, characterized in that: After determining the number of positive samples and the number of negative samples in the multiple products, the method further includes: Determining, according to the number of positive samples and the number of negative samples in the multiple products, a ratio of the number of positive samples to the number of negative samples in the multiple products; After determining the number of positive samples and the number of negative samples in the product flowing through each of the plurality of devices, the method further includes: According to the number of positive samples and the number of negative samples in the product of each device among the multiple devices, a ratio of the number of positive samples to the number of negative samples in the product of each device is determined.

6. The method according to claim 5, characterized in that The step of sorting the probability values ​​from small to large to determine N groups with top probability values ​​among the multiple groups includes: For any group among the N groups, perform the following operations: Determine the probability value according to a hypothesis testing method; determine that the grouping is accurate when the probability value is less than a third threshold; the third threshold is a significance level determined according to the hypothesis testing method; Calculate the correlation factor α according to the probability value P; Calculate the discrete factor β according to the variance d1 of the ratio of the number of positive samples to the total number of samples in the product of each device in the first group, the variance d2 of the ratio of the number of positive samples to the total number of samples in the product of each device in the second group, and the variance d3 of the ratio of the number of positive samples to the total number of samples in multiple products; Calculate the length factor γ according to the total number S of samples of all devices in any one of the groups; Calculating a score of the grouping according to the correlation factor α, the dispersion factor β and the length factor γ; Arrange the multiple groups in descending order according to the scores of the groups; It is determined that a plurality of devices in a first group of the groups having the N highest scores among the plurality of groups are root causes.

7. The method according to claim 6, characterized in that Specifically include: The correlation factor α satisfies the following formula: α=1-P; The discrete factor β satisfies the following formula: β=1-(d1+d2) / d3; The length factor γ satisfies the following formula: γ=sigmoid(S); The score of the group satisfies the following formula:

8. The method according to claim 5, characterized in that Also includes: Acquire historical information of the plurality of products; the historical information includes equipment identifications of equipment through which the products flow during the production process; Performing a deduplication operation on the device identifier; The device identifiers are arranged in descending order according to the ratio of the number of the positive samples to the number of the negative samples in the product of each device.

9. A root cause analysis device, characterized in that: include: An acquisition unit, used for acquiring defect information of multiple products; The plurality of products flow through a plurality of equipment in a first process; The defect information is determined based on the size of the product and the size of the area where the defect occurs in the product; A processing unit is used to perform the following operations: According to the defect information, determine the number of positive samples and the number of negative samples in the multiple products, and according to the number of positive samples and the number of negative samples in the multiple products, determine the ratio of the number of positive samples to the total number of samples in the multiple products; wherein the positive samples are samples whose defect information is greater than or equal to a first threshold, the negative samples are samples whose defect information is less than the first threshold, and the total number of samples is the total number of the multiple products; Determine the number of positive samples and the number of negative samples in the product flowing through each of the multiple devices, and determine the ratio of the number of positive samples in the product of each of the multiple devices to the total number of samples based on the number of positive samples and the number of negative samples in the product of each of the multiple devices; Determine that the ratio of the number of positive samples to the total number of samples in the product of each device is greater than the product of the ratio of the number of positive samples to the total number of samples in multiple products and a second threshold; The first i devices of the plurality of devices are divided into a first group, and the i+1th to Xth devices are divided into a second group; X is the total number of the plurality of devices, X is an integer greater than or equal to 1; i is an integer from 1 to X-1; Determine a probability value for each of the plurality of packets; the probability value represents a probability that the packet is erroneous; The probability values ​​are sorted from small to large, and N groups with top probability values ​​among the multiple groups are determined; the device in the first group or the device in the second group among the N groups is the root cause.

10. The device according to claim 9, characterized in that The acquisition unit acquires defect information of a plurality of products for: Acquire historical information of the plurality of products; the historical information includes process identification of the product, equipment identification of the equipment through which the product flows during the production process, and defect information of the product; Defect information of the plurality of products is acquired based on the historical information.

11. The device according to claim 9 or 10, characterized in that The defect information includes: The number of defects of the product or the defect ratio of the product, and the defect identification of the product; the number of defects is used to indicate the number of first areas where defects occur in the product, and the size of the first areas is predefined; the defect ratio is used to indicate the ratio of the number of first areas where defects occur in the product to the total number of the first areas in the product, and the size of the first areas is predefined.

12. The device according to claim 9 or 10, characterized in that The processing unit determines the number of positive samples and the number of negative samples in the multiple products according to the defect information, specifically for: According to the defect information, the number of positive samples and the number of negative samples in the multiple products are determined by a clustering algorithm.

13. The device according to claim 9, characterized in that After determining the number of positive samples and the number of negative samples in the plurality of products, the processing unit is further configured to: Determining, according to the number of positive samples and the number of negative samples in the multiple products, a ratio of the number of positive samples to the number of negative samples in the multiple products; After determining the number of positive samples and the number of negative samples in the product flowing through each of the plurality of devices, the processing unit is further configured to: According to the number of positive samples and the number of negative samples in the product of each device among the multiple devices, a ratio of the number of positive samples to the number of negative samples in the product of each device is determined.

14. The device according to claim 13, characterized in that The processing unit sorts the probability values ​​from small to large, and determines N groups with top probability values ​​among the multiple groups, for: For any group among the N groups, perform the following operations: Determine the probability value according to a hypothesis testing method; determine that the grouping is accurate when the probability value is less than a third threshold; the third threshold is a significance level determined according to the hypothesis testing method; Calculate the correlation factor α according to the probability value P; Calculate the discrete factor β according to the variance d1 of the ratio of the number of positive samples to the total number of samples in the product of each device in the first group, the variance d2 of the ratio of the number of positive samples to the total number of samples in the product of each device in the second group, and the variance d3 of the ratio of the number of positive samples to the total number of samples in multiple products; Calculate the length factor γ according to the total number S of samples of all devices in any one of the groups; Calculating a score of the grouping according to the correlation factor α, the dispersion factor β and the length factor γ; Arrange the multiple groups in descending order according to the scores of the groups; It is determined that a plurality of devices in a first group of the groups having the N highest scores among the plurality of groups are root causes.

15. The device according to claim 14, characterized in that Specifically used for: The correlation factor α satisfies the following formula: α=1-P; The discrete factor β satisfies the following formula: β=1-(d1+d2) / d3; The length factor γ satisfies the following formula: γ=sigmoid(S); The score of the group satisfies the following formula:

16. The device according to claim 13, characterized in that Also used for: Acquire historical information of the plurality of products; the historical information includes equipment identifications of equipment through which the products flow during the production process; Performing a deduplication operation on the device identifier; The device identifiers are arranged in descending order according to the ratio of the number of the positive samples to the number of the negative samples in the product of each device.

17. An electronic device, characterized in that: include: Memory, for storing computer instructions; A processor is connected to the memory, and is used to execute computer instructions in the memory, and implement the method according to any one of claims 1 to 8 when executing the computer instructions.

18. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and when the computer instructions are executed on a computer, the computer is caused to perform the method according to any one of claims 1 to 8.

Images