CN115329144A - Root cause determination method and device for product defects - Google Patents
Root cause determination method and device for product defects Download PDFInfo
- Publication number
- CN115329144A CN115329144A CN202210949969.8A CN202210949969A CN115329144A CN 115329144 A CN115329144 A CN 115329144A CN 202210949969 A CN202210949969 A CN 202210949969A CN 115329144 A CN115329144 A CN 115329144A
- Authority
- CN
- China
- Prior art keywords
- sample data
- defect
- root
- node
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000007547 defect Effects 0.000 title claims abstract description 224
- 238000000034 method Methods 0.000 title claims abstract description 68
- 238000004519 manufacturing process Methods 0.000 abstract description 16
- 230000007613 environmental effect Effects 0.000 description 25
- 239000002994 raw material Substances 0.000 description 21
- 238000010586 diagram Methods 0.000 description 12
- 239000013618 particulate matter Substances 0.000 description 10
- 239000002245 particle Substances 0.000 description 9
- 238000004458 analytical method Methods 0.000 description 4
- 238000003860 storage Methods 0.000 description 3
- 230000002411 adverse Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 239000012080 ambient air Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000739 chaotic effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000003344 environmental pollutant Substances 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 231100000719 pollutant Toxicity 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/901—Indexing; Data structures therefor; Storage structures
- G06F16/9024—Graphs; Linked lists
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/90335—Query processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/906—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06395—Quality analysis or management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Primary Health Care (AREA)
- General Health & Medical Sciences (AREA)
- Manufacturing & Machinery (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Software Systems (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The application discloses a method and a device for determining root cause of product defects, which can be applied to the financial field or other fields. The method comprises the following steps: the method comprises the steps of obtaining a plurality of sample data and a plurality of defect influence factors, determining an information gain value of each defect influence factor according to the plurality of sample data, arranging all the defect influence factors from large to small according to the information gain values corresponding to the defect influence factors, taking the defect influence factor corresponding to the maximum value in all the information gain values as a root node, forming a plurality of sub-nodes by all the defect influence factors except the root node according to an arrangement sequence, and determining the root cause of product defects according to the root node, the sample data and the sub-nodes. The root cause of the product defect is determined by constructing the production tree graph, namely, the root cause of the product defect can be determined according to the root node, the sample data and the plurality of branch nodes, the method is different from an artificial root cause determination method in the prior art, the root cause determination of various defect influence factors is realized, and the root cause determination efficiency is improved.
Description
Technical Field
The present application relates to the field of product defects, and in particular, to a method and an apparatus for determining a root cause of a product defect.
Background
With the rapid development of the product manufacturing industry, more and more products are produced in batches by using automatic industrial machines. However, defects in product quality often occur during the production of the product. Quality defects can cause waste of raw materials, reduced capacity, reduced customer satisfaction, and the like.
In the prior art, in order to solve the quality defect, the root cause of the defect needs to be determined, the traditional root cause determination method mainly depends on the experience of engineers and statistical data analysis, a large amount of data needs to be manually processed, and only one influence factor can be analyzed when the analysis is manually performed. The traditional root cause determination method has the defects of long root cause analysis time, low efficiency of determining the root cause, single defect factor of determining the root cause and the like.
Therefore, how to solve the problems of low efficiency of determining the root cause and single defect factor of determining the root cause is an urgent problem to be solved by the technical personnel in the field.
Disclosure of Invention
Based on the above problems, the application provides a method and a device for determining root cause of product defects, so as to solve the problems that the efficiency of determining the root cause is low and the defect factor for determining the root cause is single.
The embodiment of the application discloses the following technical scheme:
in a first aspect, an embodiment of the present application provides a method for determining a root cause of a product defect, where the method includes:
obtaining a plurality of sample data of the same product, and obtaining a plurality of defect influence factors related to the product;
determining an information gain value of each defect influence factor according to the plurality of sample data;
arranging all defect influence factors from large to small according to the corresponding information gain values;
taking the defect influence factor corresponding to the maximum value in all the information gain values as a root node, and forming a plurality of sub-nodes by all the defect influence factors except the root node according to an arrangement sequence;
and determining the root cause of the product defect according to the root node, the plurality of sample data and the plurality of branch nodes.
Optionally, the determining the root cause of the product defect according to the root node, the plurality of sample data and the plurality of partial nodes comprises:
classifying the sample data according to the root node and a classification rule corresponding to the root node to obtain a plurality of main data sets;
taking the subnode corresponding to the maximum information gain value in the plurality of subnodes as a first subnode;
classifying each main data set according to the first sub-node and a classification rule corresponding to the first sub-node to obtain a plurality of first sample data sub-sets corresponding to each main data set;
for all the subnodes except the first subnode, sequentially classifying the plurality of main data sets according to the classification process of the first subnode on the plurality of main data sets;
a plurality of sample data obtained after classification the subset is used as a plurality of target subsets;
taking a target subset with the most sample data in the plurality of target subsets as a target root factor subset;
and determining the root cause of the product defect by using the target root cause set.
Optionally, the classifying the sample data according to the root node and a preset classification rule corresponding to the root node to obtain a plurality of main data sets includes:
forming a plurality of first to-be-filled sets according to the types and/or value ranges of the defect influence factors corresponding to the root nodes;
and filling the sample data into the corresponding first to-be-filled sets respectively according to the forming basis of the first to-be-filled sets to obtain a plurality of main data sets.
Optionally, the classifying each main data set according to the first sub-node and the classification rule corresponding to the first sub-node to obtain a plurality of first sample data subsets corresponding to each main data set includes:
forming a plurality of second to-be-filled sets by using the defect influence factors corresponding to the first sub-nodes according to types and/or value ranges;
and filling a plurality of sample data in the main data set into corresponding second to-be-filled sets respectively according to the forming basis of the plurality of second to-be-filled sets to obtain a plurality of first sample data subsets.
Optionally, the determining the root cause of the product defect by using the target root cause set comprises:
determining a sub-node corresponding to the target root factor set as an end node;
determining all nodes between the root node and the end node as intermediate nodes;
and taking the defect influence factors corresponding to the root node, the defect influence factors corresponding to the intermediate node and the defect influence factors corresponding to the end node as root factors of the product defects.
Optionally, the method further comprises:
if the main data set or the sample data subset only contains one sample data, the classification is not continued, and the main data set or the sample data subset is used as a target subset;
if non-factor sample data exists in all sample data in the main data set or the sample data subset, the classification is not carried out, and the main data set or the sample data subset is used as a target subset, wherein the non-factor sample data is sample data which does not contain defect influence factors corresponding to the nodes to be classified.
In a second aspect, an embodiment of the present application provides an apparatus for determining root cause of a product defect, where the apparatus includes: the system comprises an acquisition module, an information gain value determination module, an arrangement module, a node determination module and a defect root cause determination module;
the acquisition module is used for acquiring a plurality of sample data of the same product and acquiring a plurality of defect influence factors related to the product;
the information gain value determining module is used for determining the information gain value of each defect influence factor according to the plurality of sample data;
the arrangement module is used for arranging all the defect influence factors from large to small according to the corresponding information gain values;
the node determining module is used for taking the defect influence factors corresponding to the maximum value in all the information gain values as root nodes and forming a plurality of sub-nodes by all the defect influence factors except the root nodes according to the arrangement sequence;
the defect root cause determining module is used for determining the root cause of the product defect according to the root node, the plurality of sample data and the plurality of branch nodes.
Optionally, the defect root cause determining module includes:
a main data set determining unit, configured to classify the sample data according to the root node and a classification rule corresponding to the root node to obtain multiple main data sets;
a first subnode determining unit, configured to use a subnode corresponding to a maximum information gain value in the multiple subnodes as a first subnode;
a first sample data subset determining unit, configured to classify each main data set according to the first sub-node and a classification rule corresponding to the first sub-node, so as to obtain a plurality of first sample data subsets corresponding to each main data set;
the set classification unit is used for sequentially classifying all the subnodes except the first subnode according to the classification process of the first subnode on the plurality of main data sets;
the target subset determining unit is used for taking a plurality of sample data subsets obtained after classification as a plurality of target subsets;
a target root factor combination determining unit, configured to use a target subset with the highest sample data in the multiple target subsets as a target root factor combination;
and the defect root cause determining unit is used for determining the root cause of the product defect by utilizing the target root cause combination.
Optionally, the master data set determining unit includes:
a first to-be-filled set determining unit, configured to form a plurality of first to-be-filled sets according to the type and/or value range of the defect affecting factor corresponding to the root node;
and the main data set determining subunit is used for respectively filling the plurality of sample data into the corresponding first to-be-filled sets according to the forming basis of the plurality of first to-be-filled sets to obtain a plurality of main data sets.
Optionally, the first sample data subset determining unit includes:
a second to-be-filled set determining unit, configured to form a plurality of second to-be-filled sets according to types and/or value ranges for the defect influencing factors corresponding to the first sub-nodes;
and the first sample data subset determining subunit is used for respectively filling the plurality of sample data in the main data set into the corresponding second to-be-filled sets according to the forming basis of the plurality of second to-be-filled sets to obtain a plurality of first sample data subsets.
Compared with the prior art, the method has the following beneficial effects:
the method comprises the steps of obtaining a plurality of sample data and a plurality of defect influence factors, determining an information gain value of each defect influence factor according to the plurality of sample data, arranging all the defect influence factors from large to small according to the information gain values corresponding to the defect influence factors, taking the defect influence factor corresponding to the maximum value in all the information gain values as a root node, forming a plurality of branch nodes by all the defect influence factors except the root node according to an arrangement sequence, and determining the root of a product defect according to the root node, the sample data and the branch nodes. In the method, a plurality of defect influence factors are considered, the influence of the defect influence factors on the product defect is determined by calculating the information gain value of each defect influence factor, and the root of the product defect is determined by constructing the production tree diagram, namely, the root of the product defect can be determined according to the root node, the sample data and the branch nodes. Different from the artificial root cause determination method in the prior art, the method realizes the determination of the root cause by various defect influence factors and improves the efficiency of the determination of the root cause.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive exercise.
Fig. 1 is a flowchart of a method for determining root cause of a product defect according to an embodiment of the present application;
FIG. 2 is a production tree diagram corresponding to a product W provided in the embodiments of the present application;
fig. 3 is a schematic structural diagram of a root cause determining apparatus for a product defect according to an embodiment of the present application.
Detailed Description
As described above, the inventor found in the research on the root cause of the product defect that in the prior art, the root cause of the defect needs to be determined in order to solve the quality defect, the conventional root cause determination method mainly depends on the experience of engineers and statistical data analysis, a large amount of data needs to be manually processed, and when the analysis is manually performed, only one influencing factor can be analyzed. The traditional root cause determination method has the defects of long root cause analysis time, low efficiency of determining the root cause, single defect factor of determining the root cause and the like.
In order to solve the above problem, an embodiment of the present application provides a method for determining a root cause of a product defect. The method comprises the following steps: the method comprises the steps of obtaining a plurality of sample data and a plurality of defect influence factors, determining an information gain value of each defect influence factor according to the plurality of sample data, arranging all the defect influence factors from large to small according to the information gain values corresponding to the defect influence factors, taking the defect influence factor corresponding to the maximum value in all the information gain values as a root node, forming a plurality of sub-nodes by all the defect influence factors except the root node according to an arrangement sequence, and determining the root factor of a product defect according to the root node, the sample data and the sub-nodes.
In the method, a plurality of defect influence factors are considered, the influence of the defect influence factors on the product defect is determined by calculating the information gain value of each defect influence factor, and the root of the product defect is determined by constructing a production tree diagram, namely the root of the product defect can be determined according to the root node, sample data and a plurality of branch nodes. Different from the artificial root cause determination method in the prior art, the method realizes the determination of the root cause by various defect influence factors and improves the efficiency of the determination of the root cause.
In order to make the technical solutions of the present application better understood, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments in the present application without making any creative effort belong to the protection scope of the present application.
Fig. 1 is a flowchart of a method for determining root cause of a product defect according to an embodiment of the present disclosure, and referring to fig. 1, the present disclosure provides a method for determining root cause of a product defect, where the method may include:
s101: obtaining a plurality of sample data of the same product, and obtaining a plurality of defect influence factors related to the product.
The defects of the product refer to the situation that the design, raw materials and parts, manufacturing assembly or instruction instructions of the product and the like can not meet the necessary reasonable safety requirements for consuming or using the product.
The defect influencing factors refer to factors that can influence defects of a product, such as raw materials, a formula, an ambient temperature, ambient particles, ambient humidity, equipment faults, and the like, and include, but are not limited to, the several defect influencing factors, which are not specifically limited herein.
However, there may be a plurality of defect influencing factors or only one influencing factor in one product, and the defect influencing factors are not particularly limited herein.
The purpose of acquiring a plurality of sample data of the same product is to improve the accuracy of the determined root cause.
S102: and determining the information gain value of each defect influence factor according to the plurality of sample data.
As an implementable implementation, the information gain value of each defect influencing factor can be obtained by:
the method comprises the following steps: preprocessing a plurality of sample data.
It is necessary to convert a continuous variable into a discrete variable by setting a threshold value, for example, whether a specific temperature is greater than 100 degrees celsius. In the present embodiment, it is assumed that the defect-affecting factors are recipes (different recipes can be assigned different fixed code values, such as 650aa _prq _SORTER, 650aa _final _cut _RE, 650aa _pre _RE, etc.), raw materials (we give the same lot number to raw materials of the same lot, such as BDW0000681H0215091, RDW000068109728028, GDW000068009528009, etc.), and environmental particulates (we call pollutants such as dust in the ambient air environmental particulates as environmental particulates, and classify these environmental particulates by particle size, into small, medium, and large categories).
Step two: and calculating the information entropy according to a plurality of sample data.
Assuming that a quality defect occurs in a certain product W, and the defect code is GOA, we want to analyze the root cause of the product W (i.e. the root cause of the defect in the product W) and obtain multiple samples of the product W, where the data of three defect influencing factors corresponding to each sample is as follows:
table 1: data of three defect influencing factors corresponding to each sample
| Sample numbering | Defect code | Formulation of | Raw material | Environmental particulate matter |
| A | GOA | 650AA_PRQ_SORTER | BDW0000681H0215091 | In |
| B | OK | 650AA_PRQ_SORTER | BDW0000681H0215091 | Small |
| C | OK | 650AA_PRQ_SORTER | BDW0000681H0215091 | Small |
| D | GOA | 650AA_PRE_SORTER | RDW000068109728028 | In |
| E | GOA | 650AA_FINAL_CUT_RE | RDW000068109728028 | Big (a) |
| F | GOA | 650AA_FINAL_CUT_RE | RDW000068109728028 | Big (a) |
| G | OK | 650AA_FINAL_CUT_RE | GDW000068009528009 | In |
| H | GOA | 650AA_PRE_RE | GDW000068009528009 | In |
| I | GOA | 650AA_PRE_RE | GDW000068009528009 | Big (a) |
The defect code OK indicates that the sample data has no quality problem, the defect code GOA indicates that the sample data has a quality problem, table 1 contains a total of nine samples, and table 1 is only an example and is not taken as a basis for narrowing the protection scope of the present application.
Step three: and calculating the information entropy of each defect influence factor.
In this embodiment, an ID3 algorithm may be adopted, and the calculation formula is as follows:
wherein, P (x) i ) To classify x i The probability of occurrence, n, is the number of classifications.
The larger the information entropy is, the more chaotic the sample data is; the smaller the information entropy, the purer the sample data.
Step four: and calculating the information gain value of each defect influence factor on the sample data set.
G A (S)＝H(S)-H(S|A)
Wherein S is a sample data set, H (S | A) is the conditional entropy of the defect influencing factor A under the condition of the sample data set, H (S) is the information entropy of the defect influencing factor A, G A (S) is the information gain value of the defect contributor A on the sample set S.
Taking the environmental particulate matter as an example, the information gain value calculation process of the environmental particulate matter is as follows:
there are 2 sample data for a small type of environmental particulates, for a total of 9 sample data, so the probability of the small type is
Since no defect with a defect code of GOA occurs in any of the 2 sample data, the probability of OK is 1 and the probability of GOA is 0.
There are 4 sample data for the medium type in the environmental particulates, and 9 sample data in total, so the probability of the medium type is
There are 3 samples of the 4 samples where GOA occurs, so the probability of OK is
The probability of GOA is
The large type of the environmental particles has 3 sample data, and 9 sample data in total, so the probability of the large type is
Since there are 3 specimen data having a GOA defect, the probability of OK is 0 and the probability of GOA is 1.
The entropy, characterized by the environmental particulates, is:
therefore, the information gain value of the environmental particulate matter is:
G A (S)＝H(S)-H(S|A)＝0.918295-0.47167556＝0.44661944
in the same way, it can be calculated that: information gain value of raw material: 0.30609775; information gain value of the recipe: 0.16775195.
s103: and arranging all defect influence factors from large to small according to the information gain values corresponding to the defect influence factors.
In S102, the three influencing factors are taken as an example, and the information gain values of the 3 defect influencing factors are environmental particles, raw materials and formulas in sequence from high to low.
The smaller the information entropy is, the larger the information gain is, and the purer the sample data is. Therefore, the defect influence factor with the largest information gain is selected as the splitting criterion of the production tree diagram. That is, in this embodiment, the environmental particulates are taken as the division criteria of the production tree diagram.
S104: and taking the defect influence factor corresponding to the maximum value in all the information gain values as a root node, and forming a plurality of sub-nodes by all the defect influence factors except the root node according to the arrangement sequence.
Taking three influencing factors in S102 as an example, after sorting the information gain values in S103, taking the environmental particles as root nodes, and taking the raw materials and the formula as branch nodes.
S105: and determining the root cause of the product defect according to the root node, the plurality of sample data and the plurality of branch nodes.
Taking three influencing factors in S102 as an example, after sorting the information gain values in S103, taking the environmental particulate matter as a root node in S104, taking the raw material and the formula as branch nodes, and constructing a production tree diagram according to the environmental particulate matter as the root node, the sample data of the product W, the raw material as the branch nodes and the formula as the branch nodes, thereby determining the root cause of the product defect.
According to the method for determining the root cause of the product defect, a plurality of sample data and a plurality of defect influence factors are obtained, the information gain value of each defect influence factor is determined according to the plurality of sample data, all the defect influence factors are arranged from large to small according to the information gain values corresponding to the defect influence factors, the defect influence factor corresponding to the maximum value of all the information gain values is used as a root node, all the defect influence factors except the root node form a plurality of sub-nodes according to the arrangement sequence, and the root cause of the product defect is determined according to the root node, the sample data and the plurality of sub-nodes. In the method, a plurality of defect influence factors are considered, the influence of the defect influence factors on the product defect is determined by calculating the information gain value of each defect influence factor, and the root of the product defect is determined by constructing the production tree diagram, namely, the root of the product defect can be determined according to the root node, the sample data and the branch nodes. Different from the artificial root cause determination method in the prior art, the method realizes the determination of the root cause by various defect influence factors and improves the efficiency of the determination of the root cause.
As an implementation manner, in order to better determine the root cause of the product defect, step S105 may specifically include:
step 1: and classifying the sample data according to the root node and a classification rule corresponding to the root node to obtain a plurality of main data sets.
As an implementation manner, the obtaining process of the plurality of main data sets may be:
forming a plurality of first to-be-filled sets according to the types and/or value ranges of the defect influence factors corresponding to the root nodes;
and filling the sample data into the corresponding first to-be-filled sets respectively according to the forming basis of the first to-be-filled sets to obtain a plurality of main data sets.
The above process of obtaining multiple main data sets may specifically take the three influencing factors in step S102 as examples, where the environmental particulate matter is a root node, see table 1, and since the types corresponding to the environmental particulate matter are small, medium, and large, the small, medium, and large are respectively used as the first to-be-filled sets, and according to the type corresponding to each sample data itself, the small, medium, and large are filled respectively.
Step 2: and taking the subnode corresponding to the maximum information gain value in the plurality of subnodes as a first subnode.
Specifically, taking the three influencing factors in step S102 as an example, the environmental particulate matter is a root node, the information gain value of the raw material is greater than the information gain value of the recipe, and the raw material is taken as a first node.
And step 3: and classifying each main data set according to the first sub-node and a classification rule corresponding to the first sub-node to obtain a plurality of first sample data subsets corresponding to each main data set.
As an implementation manner that can be implemented, the obtaining process of the plurality of first sample data subsets may specifically include:
forming a plurality of second to-be-filled sets by using the defect influence factors corresponding to the first sub-nodes according to types and/or value ranges;
and filling a plurality of sample data in the main data set into corresponding second to-be-filled sets respectively according to the forming basis of the plurality of second to-be-filled sets to obtain a plurality of first sample data subsets.
Taking the raw material as the first node in step 2 as an example, the raw materials in the same batch are given the same batch number, the raw materials in different batches are different batch numbers, a plurality of second to-be-filled sets are formed according to the different batch numbers, each main set has a corresponding plurality of second to-be-filled sets, and the sample data corresponding to each main set is filled into the different second to-be-filled sets.
And 4, step 4: and for all the subnodes except the first subnode, sequentially classifying the plurality of main data sets according to the classification process of the first subnode on the plurality of main data sets.
And forming a plurality of third to-be-filled sets according to different fixed code values by taking the root node as the environment particulate matter and the first sub-node as the raw material and taking the corresponding second sub-node as a formula, and filling the sample data in each first sample data subset into the plurality of third to-be-filled sets.
And 5: and taking a plurality of sample data subsets obtained after classification as a plurality of target subsets.
For example, the fifth node divides a plurality of sixth sample data subsets into a plurality of sixth sample data subsets, and the above contents are only used as a simple example for determining the target subset, and are not taken as a basis for limiting the protection scope of the present application.
Step 6: and taking the target subset with the most sample data in the plurality of target subsets as a target root factor subset.
In the embodiment of the present application, after a plurality of target subsets are determined, if more sample data are in the target subsets, it is proved that the influence of defect influence factors corresponding to the target subsets on product defects is larger.
And 7: and determining the root cause of the product defect by utilizing the target root cause set.
As an alternative embodiment, the process of determining the root cause may include:
determining a sub-node corresponding to the target root factor set as an end node;
determining all nodes between the root node and the end node as intermediate nodes;
and taking the defect influence factors corresponding to the root node, the defect influence factors corresponding to the intermediate node and the defect influence factors corresponding to the end node as root factors of the product defects.
In an implementation manner, when the defect influencing factors of the product defect are formula, raw material and environmental particles, and the arrangement sequence of the information gain values corresponding to the three defect influencing factors from large to small is environmental particles, raw material and formula, the corresponding production tree diagram constructed by using the root node as the environmental particles, the sample data of the product W, the branch node as the raw material and the branch node as the formula is shown in fig. 2, fig. 2 is the production tree diagram corresponding to the product W provided in the embodiment of the present application, and with reference to fig. 2, the corresponding root cause process for determining the product defect in step 7 is as follows:
the situation of poor aggregation can be obtained by obtaining the production tree diagram by the sample data according to the method, and as shown in fig. 2, the sample E, F is higher in the poor aggregation of the GOA in the target subset, so that the type of two defect influence factors, namely large in environmental particles and RDW000068109728028 in raw materials, can be presumed to have higher adverse influence on the GOA. The sample B, C is the higher aggregate in the target subset of OK, from which it can be presumed that environmental particulates are the smaller defect contributors that have a less adverse effect on GOA.
As an implementable embodiment, the method may further comprise:
if the main data set or the sample data subset only contains one sample data, the classification is not continued, and the main data set or the sample data subset is used as a target subset;
if non-factor sample data exists in all the sample data in the main data set or the sample data subset, the classification is not carried out, and the main data set or the sample data subset is used as a target subset, wherein the non-factor sample data refers to the sample data which does not contain defect influence factors corresponding to the nodes to be classified.
Based on the specific description of the forming process of the production tree diagram provided by the embodiment, the root cause of the product defect can be determined by establishing the production tree, which is different from the artificial root cause determination method in the prior art, the root cause determination of various defect influence factors is realized, the root cause determination efficiency is improved, and the influence of various defect factors on the product defect is considered, so that the root cause of the product defect is not single.
Based on the method for determining root cause of product defect provided by the foregoing embodiment, referring to fig. 3, fig. 3 is a device for determining root cause of product defect provided by the embodiment of the present application, and as shown in fig. 3, the device 300 for determining root cause of product defect may include:
an obtaining module 301, configured to obtain multiple sample data of a same product, and obtain multiple defect influencing factors related to the product;
an information gain value determining module 302, configured to determine an information gain value of each defect affecting factor according to the plurality of sample data;
the arrangement module 303 is configured to arrange all defect influencing factors from large to small according to the information gain values corresponding to the defect influencing factors;
a node determining module 304, configured to use a defect influence factor corresponding to a maximum value of all information gain values as a root node, and form a plurality of sub-nodes according to an arrangement order for all defect influence factors except the root node;
a defect root cause determining module 305, configured to determine a root cause of the product defect according to the root node, the plurality of sample data, and the plurality of branch nodes.
As an optional implementation manner, the defect root cause determining module 305 may include:
a master data set determining unit, configured to classify the sample data according to the root node and a classification rule corresponding to the root node to obtain multiple master data sets;
a first subnode determining unit, configured to use a subnode corresponding to a maximum information gain value in the multiple subnodes as a first subnode;
a first sample data subset determining unit, configured to classify each main data set according to the first sub-node and a classification rule corresponding to the first sub-node, so as to obtain a plurality of first sample data subsets corresponding to each main data set;
the set classification unit is used for sequentially classifying all the subnodes except the first subnode according to the classification process of the first subnode on the plurality of main data sets;
the target subset determining unit is used for taking a plurality of sample data subsets obtained after classification as a plurality of target subsets;
a target root factor combination determining unit, configured to use a target subset with the highest sample data in the multiple target subsets as a target root factor combination;
and the defect root cause determining unit is used for determining the root cause of the product defect by utilizing the target root cause combination.
As an optional implementation, the main data set determining unit may include:
a first to-be-filled set determining unit, configured to form a plurality of first to-be-filled sets according to the type and/or value range of the defect affecting factor corresponding to the root node;
and the main data set determining subunit is used for respectively filling the plurality of sample data into the corresponding first to-be-filled sets according to the forming basis of the plurality of first to-be-filled sets to obtain a plurality of main data sets.
As an optional embodiment, the first sample data subset determining unit may include:
a second to-be-filled set determining unit, configured to form a plurality of second to-be-filled sets according to types and/or value ranges for the defect influence factors corresponding to the first sub-node;
and the first sample data subset determining subunit is used for respectively filling the plurality of sample data in the main data set into the corresponding second to-be-filled sets according to the forming basis of the plurality of second to-be-filled sets to obtain a plurality of first sample data subsets.
As an optional implementation manner, the defect root cause determining unit may include:
an end node determining unit, configured to determine a sub-node corresponding to the target root cause set as an end node;
an intermediate node determining unit configured to determine all nodes between the root node and the end node as intermediate nodes;
and the defect root cause determining subunit is used for taking the defect influence factors corresponding to the root node, the defect influence factors corresponding to the intermediate node and the defect influence factors corresponding to the end node as root causes of the product defects.
As an implementable embodiment, the apparatus may further comprise:
the first target subset determining unit is used for not continuing to classify if only one sample data is contained in the main data set or the sample data subset, and taking the main data set or the sample data subset as a target subset;
and the second target subset determining unit is used for not classifying if non-factor sample data exists in all sample data in the main data set or the sample data subset, and taking the main data set or the sample data subset as a target subset, wherein the non-factor sample data refers to sample data which does not contain defect influence factors corresponding to the nodes to be classified.
The embodiment of the application also provides corresponding equipment and a computer storage medium, which are used for realizing the scheme provided by the embodiment of the application.
The device comprises a memory and a processor, wherein the memory is used for storing instructions or codes, and the processor is used for executing the instructions or codes so as to enable the device to execute the root cause determination method of the product defects.
The computer storage medium has code stored therein, and when the code is executed, the apparatus for executing the code implements the root cause determination method for product defects according to any embodiment of the present application.
The method and the device for determining the root cause of the product defect can be applied to the financial field or other fields, for example, can be applied to application scenes for determining the root cause of the defect in the financial field. Other fields are any fields other than the financial field, for example, the field of product defects. The above description is only an example, and does not limit the application field of the method and the apparatus for determining the root cause of the product defect provided by the present invention.
It should be noted that, in this specification, each embodiment is described in a progressive manner, and the same and similar parts between the embodiments are referred to each other, and each embodiment focuses on differences from other embodiments. In particular, for the apparatus and device embodiments, since they are substantially similar to the method embodiments, they are described relatively simply, and reference may be made to some of the descriptions of the method embodiments for related points. The above-described embodiments of the apparatus, device and system are merely illustrative, and units described as separate components may or may not be physically separate, and components indicated as units may or may not be physical units, may be located in one place, or may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
In the embodiments of the present application, the names "first" and "second" (if present) in the names "first" and "second" are used merely for name identification, and do not represent the first and second in sequence.
As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that all or part of the steps in the above embodiment methods can be implemented by software plus a general hardware platform. Based on such understanding, the technical solution of the present application may be embodied in the form of a software product, which may be stored in a storage medium, such as a read-only memory (ROM)/RAM, a magnetic disk, an optical disk, or the like, and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network communication device such as a router) to execute the method according to the embodiments or some parts of the embodiments of the present application.
The above description is only one specific embodiment of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present application should be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A method for determining a root cause of a product defect, the method comprising:
obtaining a plurality of sample data of the same product, and obtaining a plurality of defect influence factors related to the product;
determining an information gain value of each defect influence factor according to the plurality of sample data;
arranging all defect influence factors from large to small according to the corresponding information gain values;
taking the defect influence factor corresponding to the maximum value in all the information gain values as a root node, and forming a plurality of sub-nodes by all the defect influence factors except the root node according to an arrangement sequence;
and determining the root cause of the product defect according to the root node, the plurality of sample data and the plurality of branch nodes.
2. The method of claim 1, wherein the determining the root cause of the product defect according to the root node, the plurality of sample data, and the plurality of partial nodes comprises:
classifying the sample data according to the root node and a classification rule corresponding to the root node to obtain a plurality of main data sets;
taking the subnode corresponding to the maximum information gain value in the plurality of subnodes as a first subnode;
classifying each main data set according to the first sub-node and a classification rule corresponding to the first sub-node to obtain a plurality of first sample data sub-sets corresponding to each main data set;
for all the subnodes except the first subnode, sequentially classifying the plurality of main data sets according to the classification process of the first subnode on the plurality of main data sets;
taking a plurality of sample data subsets obtained after classification as a plurality of target subsets;
taking a target subset with the most sample data in the plurality of target subsets as a target root factor subset;
and determining the root cause of the product defect by utilizing the target root cause set.
3. The method of claim 2, wherein the classifying the sample data according to the root node and a preset classification rule corresponding to the root node to obtain a plurality of main data sets comprises:
forming a plurality of first to-be-filled sets according to the types and/or value ranges of the defect influence factors corresponding to the root nodes;
and filling the sample data into the corresponding first to-be-filled sets respectively according to the forming basis of the first to-be-filled sets to obtain a plurality of main data sets.
4. The method of claim 2, wherein the classifying each main data set according to the first branch node and the classification rule corresponding to the first branch node to obtain a plurality of first sample data subsets corresponding to each main data set comprises:
forming a plurality of second to-be-filled sets by using the defect influence factors corresponding to the first nodes according to types and/or value ranges;
and filling a plurality of sample data in the main data set into corresponding second to-be-filled sets respectively according to the forming basis of the plurality of second to-be-filled sets to obtain a plurality of first sample data subsets.
5. The method of claim 2, wherein the determining the root cause of the product defect using the target root cause combination comprises:
determining a sub-node corresponding to the target root factor set as an end node;
determining all nodes between the root node and the end node as intermediate nodes;
and taking the defect influence factors corresponding to the root node, the defect influence factors corresponding to the intermediate node and the defect influence factors corresponding to the end node as root factors of the product defects.
6. The method of determining root cause of a product defect of any one of claims 2 to 4, further comprising:
if the main data set or the sample data subset only contains one sample data, the classification is not continued, and the main data set or the sample data subset is used as a target subset;
if non-factor sample data exists in all sample data in the main data set or the sample data subset, the classification is not carried out, and the main data set or the sample data subset is used as a target subset, wherein the non-factor sample data is sample data which does not contain defect influence factors corresponding to the nodes to be classified.
7. An apparatus for determining a root cause of a product defect, the apparatus comprising: the system comprises an acquisition module, an information gain value determination module, an arrangement module, a node determination module and a defect root cause determination module;
the acquisition module is used for acquiring a plurality of sample data of the same product and acquiring a plurality of defect influence factors related to the product;
the information gain value determining module is used for determining the information gain value of each defect influence factor according to the plurality of sample data;
the arrangement module is used for arranging all the defect influence factors from large to small according to the corresponding information gain values;
the node determining module is used for taking the defect influence factors corresponding to the maximum value in all the information gain values as root nodes and forming a plurality of sub-nodes by all the defect influence factors except the root nodes according to the arrangement sequence;
the defect root cause determining module is used for determining the root cause of the product defect according to the root node, the plurality of sample data and the plurality of branch nodes.
8. The apparatus of claim 7, wherein the defect root cause determination module comprises:
a main data set determining unit, configured to classify the sample data according to the root node and a classification rule corresponding to the root node to obtain multiple main data sets;
a first sub-node determining unit, configured to use a sub-node corresponding to a maximum information gain value in the plurality of sub-nodes as a first sub-node;
a first sample data subset determining unit, configured to classify each main data set according to the first sub-node and a classification rule corresponding to the first sub-node, so as to obtain a plurality of first sample data subsets corresponding to each main data set;
the set classification unit is used for sequentially classifying all the subnodes except the first subnode according to the classification process of the first subnode on the plurality of main data sets;
the target subset determining unit is used for taking a plurality of sample data subsets obtained by classification as a plurality of target subsets;
a target root factor combination determining unit, configured to use a target subset with the highest sample data in the multiple target subsets as a target root factor combination;
and the defect root cause determining unit is used for determining the root cause of the product defect by utilizing the target root cause combination.
9. The apparatus of claim 8, wherein the master data set determining unit comprises:
a first to-be-filled set determining unit, configured to form a plurality of first to-be-filled sets according to the types and/or value ranges of the defect affecting factors corresponding to the root node;
and the main data set determining subunit is used for respectively filling the plurality of sample data into the corresponding first to-be-filled sets according to the forming basis of the plurality of first to-be-filled sets to obtain a plurality of main data sets.
10. The apparatus of claim 8, wherein the first subset of sample data determining unit comprises:
a second to-be-filled set determining unit, configured to form a plurality of second to-be-filled sets according to types and/or value ranges for the defect influence factors corresponding to the first sub-node;
and the first sample data subset determining subunit is used for respectively filling the plurality of sample data in the main data set into the corresponding second to-be-filled sets according to the forming basis of the plurality of second to-be-filled sets to obtain a plurality of first sample data subsets.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210949969.8A CN115329144A (en) | 2022-08-09 | 2022-08-09 | Root cause determination method and device for product defects |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210949969.8A CN115329144A (en) | 2022-08-09 | 2022-08-09 | Root cause determination method and device for product defects |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115329144A true CN115329144A (en) | 2022-11-11 |
Family
ID=83920961
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210949969.8A Pending CN115329144A (en) | 2022-08-09 | 2022-08-09 | Root cause determination method and device for product defects |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115329144A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2024103436A1 (en) * | 2022-11-18 | 2024-05-23 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | Device defect data rule base construction method and device defect correlation analysis method |
-
2022
- 2022-08-09 CN CN202210949969.8A patent/CN115329144A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2024103436A1 (en) * | 2022-11-18 | 2024-05-23 | 中国南方电网有限责任公司超高压输电公司检修试验中心 | Device defect data rule base construction method and device defect correlation analysis method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115423603B (en) | Wind control model building method, system and storage medium based on machine learning | |
| CN110610482A (en) | Workpiece flaw detection method based on resnet | |
| JP2004157814A (en) | Decision tree generating method and model structure generating device | |
| CN115329144A (en) | Root cause determination method and device for product defects | |
| CN111967521B (en) | Cross-border active user identification method and device | |
| CN115756919A (en) | Root cause positioning method and system for multidimensional data | |
| CN112132187A (en) | Method for rapidly judging perovskite structure stability based on random forest | |
| CN112199407B (en) | Data packet ordering method, device, equipment and storage medium | |
| CN109492660B (en) | Complaint information processing method, apparatus, computer device and storage medium | |
| CN115796704A (en) | Goods and materials sampling inspection method and device based on LightGBM index model | |
| CN112464970A (en) | Regional value evaluation model processing method and device and computing equipment | |
| CN113641654B (en) | Marketing treatment rule engine method based on real-time event | |
| CN115660730A (en) | Loss user analysis method and system based on classification algorithm | |
| CN115470279A (en) | Data source conversion method, device, equipment and medium based on enterprise data | |
| CN108062395A (en) | A kind of track traffic big data analysis method and system | |
| CN114282598A (en) | Multi-source heterogeneous power grid data fusion method, device, equipment and computer medium | |
| CN114510405A (en) | Index data evaluation method, index data evaluation device, index data evaluation apparatus, storage medium, and program product | |
| CN112131105A (en) | Test data construction method and device | |
| US9665795B2 (en) | Method and apparatus for identifying root cause of defect using composite defect map | |
| CN115982634A (en) | Application program classification method and device, electronic equipment and computer program product | |
| CN112990246B (en) | Method and device for establishing isolated tree model | |
| CN118037121B (en) | Importance scoring method, device, equipment and medium for situation awareness data | |
| CN117094648B (en) | Visual management system of warehouse based on thing networking | |
| CN108074256B (en) | Sulfide information extraction method, device and system based on distributed processing | |
| CN114239721A (en) | Model feature selection method and device and computer readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |