CN120598438B - Quality information management method and system based on multi-mode AI - Google Patents
Quality information management method and system based on multi-mode AIInfo
- Publication number
- CN120598438B CN120598438B CN202511100022.XA CN202511100022A CN120598438B CN 120598438 B CN120598438 B CN 120598438B CN 202511100022 A CN202511100022 A CN 202511100022A CN 120598438 B CN120598438 B CN 120598438B
- Authority
- CN
- China
- Prior art keywords
- operation step
- quality
- grade
- data
- multimodal
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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 OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Game Theory and Decision Science (AREA)
- Probability & Statistics with Applications (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- General Factory Administration (AREA)
Abstract
The invention provides a quality information management method and system based on a multi-mode AI, which belong to the technical field of information management and comprise the steps of carrying out multi-mode detection on a set target of a corresponding operation step according to a multi-mode detection mode under each operation step in an operation flow, carrying out data analysis on multi-mode data under each operation step, evaluating the quality grade of the set target under the corresponding operation step, storing the multi-mode data under the corresponding operation step into a database I when the quality grade accords with a set standard of the corresponding operation step, otherwise, determining a storage address of the multi-mode data under the corresponding operation step in the database II according to the difference between the quality grade of the corresponding operation step and the set standard and combining the comprehensive grade of a target item determined based on all the quality grades, and carrying out storage management. The quality data of passing and failing is convenient to store and manage, the subsequent retrieval and use are convenient, and the problem checking efficiency is improved.
Description
Technical Field
The invention relates to the technical field of information management, in particular to a quality information management method and system based on multi-mode AI.
Background
In the robot manufacturing process, quality information management is important to guaranteeing product quality and improving production efficiency. Along with the rapid development of the robot manufacturing technology, the manufacturing process is increasingly complex, a plurality of operation steps such as part processing, assembly, debugging and the like are covered, and each operation step has a key influence on the final quality of the product.
Because of the existence of excessive information data in the robot manufacturing process, a unified storage strategy is generally adopted at present, the generated multi-mode data are stored in the same position no matter the quality level of the operation steps, when the quality problem occurs and the retrospective analysis is needed, workers are difficult to quickly locate key data, the problem investigation efficiency is low, and the production progress and the quality improvement flow are seriously influenced.
Therefore, the invention provides a quality information management method and system based on multi-mode AI.
Disclosure of Invention
The invention provides a quality information management method and system based on multi-mode AI (advanced technology attachment), which are used for solving the technical problems.
The invention provides a quality information management method based on multi-mode AI, comprising the following steps:
Step 1, determining a working flow of a target project, and carrying out multi-mode detection on a set target of a corresponding working step according to a multi-mode detection mode of each working step in the working flow to obtain multi-mode data;
step 2, carrying out data analysis on the multi-mode data in each operation step, and evaluating the quality grade of the set target in the corresponding operation step;
step 3, when the quality grade accords with the setting standard of the corresponding operation step, storing the multi-mode data in the corresponding operation step into a database I;
and 4, when the quality grade does not accord with the set standard of the corresponding operation step, determining the storage address of the multi-mode data of the corresponding operation step in the database II according to the difference between the quality grade of the corresponding operation step and the set standard and combining the comprehensive grade of the target item determined based on all the quality grades, and carrying out storage management.
Preferably, the process of data analysis of the multi-mode data under each operation step includes:
Filling each mode data into a mode blank table matched with the corresponding operation step respectively, determining whether a missing object exists in the filled table, and locking a missing position point of the missing object if the missing object exists;
acquiring a first reference value of a first non-missing point nearest to the missing position point, a missing point nearest to the missing position point and a second reference value of a second non-missing point nearest to the nearest missing point according to the representation form of the filled table;
If the number of the nearest missing points is 0, performing first alignment on missing data of the missing position points based on the first reference value;
if the number of the nearest missing points is 1, performing second filling on the data of the missing position points based on the first reference value and the second reference value;
if the number of the nearest missing points is 2, third filling is performed on the data of the missing position points based on the first reference value and the two second reference values;
and obtaining complete data of the corresponding modes according to the filling result.
Preferably, the third filling of the data of the missing position point based on the first reference value and the two second reference values includes:
;
wherein, the Respectively a first reference value, a second reference value and another second reference value; Respectively based on the first reference value Slope of adjacent points to corresponding pointBased on a second reference valueSlope of adjacent points to corresponding pointSlope of adjacent point based on corresponding point of another second reference value C3;Representing a first reference valueCorresponding point and a second reference valueSlope between corresponding points; representing a first reference value Corresponding point and another second reference valueSlope between corresponding points; representing a sign function; indicating the total number of non-missing points present in the filled table; representing a third fill value.
Preferably, evaluating the quality level of the set target at the corresponding working step includes:
converting the data analysis result under each mode into a corresponding feature vector;
according to the characteristics and requirements of the set target, determining a plurality of quality evaluation indexes related to each operation step, and giving corresponding weight to the quality evaluation indexes under each operation step, wherein the quality evaluation indexes comprise installation accuracy, operation standardization and set running stability;
performing matching analysis on the feature vector and each quality evaluation index to construct a multi-mode result matrix , wherein,1 Respectively represents the installation accuracy, operation standardization and set running stability in the 1 st operation step; respectively representing the installation accuracy, operation standardization and set running stability in the nth operation step; Respectively representing the installation accuracy, operation standardization and weight of the set running stability in the 1 st working step; respectively representing the installation accuracy, operation standardization and weight of the set running stability in the nth operation step;
extracting the oblique quantity of the multi-mode result matrix J, and mapping each element in the oblique quantity with a grade comparison table to obtain the initial grade of the corresponding element;
determining a change amount according to the connection position of the operation step;
And adjusting the initial grade according to the change amount to obtain a quality grade.
Preferably, determining the change amount according to the engagement position of the operation step includes:
when the operation step is the 1 st step, the initial grade is regarded as a quality grade;
When the operation step is not the 1 st step, acquiring the influence of the complete quality of the last step on the corresponding operation step;
wherein, the Representing the influence coefficient generated for the ith operation step; indicating the sensitivity of the ith step in the connection position affected by the ith-1 th step; represents the degree of association closeness between the ith step and the i-1 th step, n represents the total number of operation steps; Indicating the direction judgment coefficient for the ith working step, and when the direction is positive, The value is 1, when the direction is the negative direction,The value is-1;
And determining an adjustment grade according to the influence coefficient and the influence of a preset unit, and adjusting the initial grade to obtain a quality grade.
Preferably, the adjusting the initial grade to obtain the quality grade includes:
wherein ZD represents a quality grade, CD represents an initial grade, TD represents an adjustment grade; representing rounding down symbols.
Preferably, determining the storage address of the multimodal data corresponding to the job step in the database two includes:
According to the step weight of each operation step, combining the quality grade of each operation step to obtain a comprehensive grade, and obtaining the comprehensive difference between the comprehensive grade and a comprehensive standard;
obtaining a difference array Cs=according to the quality grade and the set standard , wherein,Indicating a level difference of the ith job step;
counting symbol combinations { E1, E2} of each remaining operation step and the next operation step except the nth operation step, wherein E1 is a symbol corresponding to the remaining operation step;
if the corresponding residual steps do not meet the set standard, the storage address in the database II is obtained by matching from the step-symbol combination-double difference-address comparison table according to the step weight, the symbol combination, the grade difference and the comprehensive difference of the corresponding residual steps.
The invention provides a quality information management system based on multi-mode AI, comprising:
the data acquisition module is used for determining the operation flow of a target project, and carrying out multi-mode detection on the set target of the corresponding operation step according to the multi-mode detection mode of each operation step in the operation flow to obtain multi-mode data;
The grade evaluation module is used for carrying out data analysis on the multi-mode data in each operation step and evaluating the quality grade of the set target in the corresponding operation step;
the standard storage module is used for storing the multi-mode data in the corresponding operation step into the database I when the quality grade accords with the set standard of the corresponding operation step;
And the nonstandard storage module is used for determining the storage address of the multi-mode data of the corresponding operation step in the database II according to the difference between the quality grade of the corresponding operation step and the set standard when the quality grade does not meet the set standard of the corresponding operation step and combining the comprehensive grade of the target item determined based on all the quality grades, and carrying out storage management.
Preferably, the storage medium comprises a memory for storing the quality information management method based on the multi-mode AI.
Preferably, the computer equipment comprises a processor, wherein the processor is used for executing the quality information management method based on the multi-mode AI.
Compared with the prior art, the application has the following beneficial effects:
On one hand, the quality management of the project process is facilitated, on the other hand, the storage management of qualified and unqualified quality data is facilitated, further the subsequent retrieval and use are facilitated, and the problem investigation efficiency is improved.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a quality information management method based on a multi-mode AI according to an embodiment of the present invention;
Fig. 2 is a block diagram of a quality information management system based on a multi-mode AI according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention provides a quality information management method based on multi-mode AI, as shown in figure 1, comprising the following steps:
Step 1, determining a working flow of a target project, and carrying out multi-mode detection on a set target of a corresponding working step according to a multi-mode detection mode of each working step in the working flow to obtain multi-mode data;
step 2, carrying out data analysis on the multi-mode data in each operation step, and evaluating the quality grade of the set target in the corresponding operation step;
step 3, when the quality grade accords with the setting standard of the corresponding operation step, storing the multi-mode data in the corresponding operation step into a database I;
and 4, when the quality grade does not accord with the set standard of the corresponding operation step, determining the storage address of the multi-mode data of the corresponding operation step in the database II according to the difference between the quality grade of the corresponding operation step and the set standard and combining the comprehensive grade of the target item determined based on all the quality grades, and carrying out storage management.
In this embodiment, the target item is a manufacturing item of an industrial robot, and the work flow includes a plurality of steps of machining mechanical parts (such as machining operations of milling, drilling, etc. of arms and joint parts), assembling electrical components (mounting motors, sensors, etc.), programming and debugging software (writing control programs and testing operation logic), assembling the machined mechanical parts and the electrical components integrally, and the like. In the machining step of the mechanical part, a plurality of detection modes such as visual detection (a high-definition camera is used for shooting the surface of the part, whether appearance problems such as cracks and flaws exist or not are detected), size measurement sensor detection (such as a laser ranging sensor is used for measuring whether the key size of the part meets design requirements or not), acoustic detection (whether noise generated by abnormal vibration exists in the machining process or not is detected through a sonar device and whether the machining equipment operates normally or not) and the like are adopted, so that a multi-mode data set covering a plurality of types of data such as a numerical value (such as a part size numerical value), an image (a part appearance image), audio (a machining equipment operating sound audio) and the like is obtained, and the state information of a setting target of the operation step is comprehensively reflected. The multi-mode detection can synthesize various data and completely present the actual conditions of the operation steps. For example, only the measurement of the guiding dimension cannot find the surface microcrack of the part, but the visual detection can be effectively supplemented, and the acoustic detection can early warn the hidden trouble of the equipment in advance. The comprehensive data acquisition provides a sufficient basis for subsequent quality evaluation, and the evaluation accuracy is improved.
In the embodiment, a machine learning algorithm, such as Convolutional Neural Network (CNN), is used for processing image data to extract appearance characteristics of parts, a signal processing method, such as Fourier transform, is used for analyzing audio data to judge whether sound frequency is normal, statistical analysis is performed on numerical data, and statistics, such as mean, standard deviation and the like, are calculated. The quality grades can be divided into 1, 2, 3, 4 and other grades, different grades are represented by corresponding numerical values, and the higher the grade is, the more qualified the corresponding quality is.
In this embodiment, the setting criterion is a quality requirement preset for each working step, that is, a class criterion, for example, a class of 4 is a setting criterion.
The first database is a database for storing data meeting quality standards. The method can be a relational database, such as a MySQL database, a table structure is built according to dimensions such as operation steps, data types and the like, data storage and query are facilitated, the database is a database specially used for storing data which does not meet quality standards, a distributed database such as a Hadoop Distributed File System (HDFS) is adopted to be combined with an HBase database so as to meet the requirements that the amount of unqualified data is possibly large and rapid query analysis is needed, and when the quality grade does not meet the standards, a storage address is determined by combining the difference and the comprehensive grade, so that the problem data can be stored in a classified mode more accurately. And when quality problems occur, quality management personnel can quickly locate data of related operation steps, and analyze the root cause of the problems. For example, by analyzing unqualified data in the database II, quality fluctuation of robots in a certain batch in a plurality of operation steps is found, and the problems of raw materials or production equipment can be deeply checked, and improvement measures can be timely taken.
In this embodiment, the level difference means, for example, the quality level is 2, the standard is set to 4, and at this time, the level difference is-2, and when in use, it is only required to use "-2" directly.
In this embodiment, the comprehensive quality level is calculated by means of weighted average or the like according to the quality level of each working step, so as to obtain the comprehensive quality level of the whole industrial robot manufacturing project.
The technical scheme has the beneficial effects that on one hand, the quality management of the project process is facilitated, on the other hand, the storage management of qualified and unqualified quality data is facilitated, further, the subsequent retrieval and use are facilitated, and the problem checking efficiency is improved.
The invention provides a quality information management method based on multi-mode AI, which comprises the following steps in the process of carrying out data analysis on multi-mode data in each operation step:
Filling each mode data into a mode blank table matched with the corresponding operation step respectively, determining whether a missing object exists in the filled table, and locking a missing position point of the missing object if the missing object exists;
acquiring a first reference value of a first non-missing point nearest to the missing position point, a missing point nearest to the missing position point and a second reference value of a second non-missing point nearest to the nearest missing point according to the representation form of the filled table;
If the number of the nearest missing points is 0, performing first alignment on missing data of the missing position points based on the first reference value;
if the number of the nearest missing points is 1, performing second filling on the data of the missing position points based on the first reference value and the second reference value;
if the number of the nearest missing points is 2, third filling is performed on the data of the missing position points based on the first reference value and the two second reference values;
and obtaining complete data of the corresponding modes according to the filling result.
Preferably, the third filling of the data of the missing position point based on the first reference value and the two second reference values includes:
;
wherein, the Respectively a first reference value, a second reference value and another second reference value; Respectively based on the first reference value Slope of adjacent points to corresponding pointBased on a second reference valueSlope of adjacent points to corresponding pointSlope of adjacent point based on corresponding point of another second reference value C3;Representing a first reference valueCorresponding point and a second reference valueSlope between corresponding points; representing a first reference value Corresponding point and another second reference valueSlope between corresponding points; representing a sign function; indicating the total number of non-missing points present in the filled table; representing a third fill value.
In this embodiment, the modality blank table is a blank table designed for each modality data. For example, the column of the sensor mode blank table may include a sensor number, an acquisition time, a measurement parameter name, a measurement value, and the like, and the missing object is exemplified by sensor mode data, and a certain specific sensor does not acquire data within a certain period of time, so that the sensor measurement value in the period of time is the missing object, and in the sensor mode blank table, a crossing position of a row and a column corresponding to the missing data is a missing position point, for example, data at a certain row (corresponding to a specific acquisition time) and a certain column (corresponding to a specific measurement parameter) is missing. And the data are respectively filled into the corresponding blank tables according to the modes, so that independent management and analysis of each mode data are facilitated, and the pertinence of data processing is improved. The method is characterized in that the missing position points are locked to provide a clear target for the follow-up data filling, blind processing is avoided, the accuracy and the high efficiency of filling operation are ensured, and the fact that the data filling of the part is aimed at filling on data is required, the representation form of the filled table is in a two-dimensional table form, the table is provided with a row and column structure, and the rows and columns respectively correspond to different data attributes and records;
In this embodiment, in the filled table, the non-missing data point closest to the missing position point is the first non-missing point, and the data value corresponding to the first non-missing point is the first reference value.
In this embodiment, in the filled table, the other missing data point closest to the missing position point is the closest missing point, the non-missing data point closest to the closest missing point is the second non-missing point, and the data value corresponding to the second missing point is the second reference value.
In this embodiment, when the number of nearest missing points is 0, the missing data of the missing position points is padded based on only the first reference value.
And when the number of the nearest missing points is 1, the data of the missing position points are complemented by combining the average value of the first reference value and the second reference value.
In this embodiment, the formula is incorporated intoThe data information is acquired from multiple dimensions, the unilateral property of only relying on a single reference value is avoided, the multiple reference values can more comprehensively reflect the characteristics and the trend of the data around the missing position points, the filling value is more in line with the overall data distribution,The introduction of the reference values considers the change trend of the adjacent points of the corresponding points of the reference values. Data often has a certain change rule in space or time, and the trend can be captured through the slope.、The change condition among the corresponding points of different reference values is considered, the description of the change trend of the data is further perfected, and the filling value is more attached to the actual change of the data.
In this embodiment of the present invention, the process is performed,According toTo adjust the positive and negative of the subsequent calculation part, balance the effect of the difference between the two second reference values on the result, ifThe absolute value of the difference of (c) is smaller than b1,And the value is 1, otherwise, the value is-1, and the value of b1 is 1.
In the embodiment, the denominator N1 performs normalization processing on the calculation result, so that the calculation result fluctuates within a reasonable range, and abnormal results caused by factors such as data scale and the like are avoided.
The technical scheme has the beneficial effects that the latest missing points with different numbers mean different conditions of missing data periphery, and the differentiated filling strategy is adopted to achieve pertinence. When the number of the nearest missing points is 0, the fact that the peripheral data of the missing points are relatively complete is indicated, the first reference value is used for filling in the missing points simply and efficiently, when the number is 1 or 2, the plurality of reference values are combined for filling in the missing points, the change trend and the association relation of the data can be considered better, and the deviation possibly caused by a single reference value is avoided.
The invention provides a quality information management method based on a multi-mode AI, which evaluates the quality grade of a set target under a corresponding operation step and comprises the following steps:
converting the data analysis result under each mode into a corresponding feature vector;
according to the characteristics and requirements of the set target, determining a plurality of quality evaluation indexes related to each operation step, and giving corresponding weight to the quality evaluation indexes under each operation step, wherein the quality evaluation indexes comprise installation accuracy, operation standardization and set running stability;
performing matching analysis on the feature vector and each quality evaluation index to construct a multi-mode result matrix , wherein,1 Respectively represents the installation accuracy, operation standardization and set running stability in the 1 st operation step; respectively representing the installation accuracy, operation standardization and set running stability in the nth operation step; Respectively representing the installation accuracy, operation standardization and weight of the set running stability in the 1 st working step; respectively representing the installation accuracy, operation standardization and weight of the set running stability in the nth operation step;
extracting the oblique quantity of the multi-mode result matrix J, and mapping each element in the oblique quantity with a grade comparison table to obtain the initial grade of the corresponding element;
determining a change amount according to the connection position of the operation step;
And adjusting the initial grade according to the change amount to obtain a quality grade.
In this embodiment, the feature vector is in a form of a numeric and structured representation of the data analysis result, for example, the result of the operation step 1 for the parameter A1 is U1, the result for the parameter A2 is U2, and the result for the parameter A3 is U3, and the obtained feature vector is [ U1U 2U 3].
In this embodiment, the weight is a value given according to the importance degree of each quality evaluation index in different operation steps, which is preset, and the importance of each quality evaluation index in different operation steps is different, so that the influence of the key index on the quality of the operation steps can be highlighted by giving the weight, for example, the weight of the operation stability is 0.5 in operation step 1, the weight of the installation accuracy is 0.3, and the weight of the operation normalization is 0.2.
In the embodiment, the matching analysis is to compare the characteristic vector of the deviation of the installation position of the part under the visual mode with the installation accuracy index, judge the influence degree of the deviation of the installation position on the installation accuracy, compare the characteristic vector of the pressure, the temperature and the like under the sensor mode with the set operation stability index, and analyze the effect of the characteristic vector on the operation stability of the equipment.
In this embodiment, the diagonal quantity refers to a vector extracted according to a main diagonal direction, the diagonal quantity elements represent comprehensive performance results of quality evaluation indexes under different operation steps, and the grade comparison table contains values of different elements and grades matched with the values according to preliminary quality grades corresponding to each element obtained by mapping the diagonal quantity elements and the grade comparison table.
In this embodiment, the engagement position refers to the effect that the quality of the previous working step is unstable may have on the subsequent working step.
The technical scheme has the advantages that the data analysis result in each mode is converted into the corresponding feature vector, unified structural representation of the data is realized, subsequent processing and analysis are facilitated, a plurality of quality evaluation indexes related to each operation step are determined, corresponding weights are given to each index, a set of quantization system aiming at quality evaluation of the operation step is constructed, a multi-mode result matrix is constructed, the matching results of different mode data and quality evaluation indexes in each operation step are presented in a matrix form, quality related information of each operation step is comprehensively reflected, the diagonal vector is extracted, representative key information can be screened out from the matrix, the data is simplified, the core quality characteristics are reserved, the change amount calculated according to the connection position of the operation step is determined, and the change amount is used for subsequent adjustment of the initial grade.
The invention provides a quality information management method based on a multi-mode AI, which determines a change amount according to the connection position of the operation steps, and comprises the following steps:
when the operation step is the 1 st step, the initial grade is regarded as a quality grade;
When the operation step is not the 1 st step, acquiring the influence of the complete quality of the last step on the corresponding operation step;
wherein, the Representing the influence coefficient generated for the ith operation step; indicating the sensitivity of the ith step in the connection position affected by the ith-1 th step; represents the degree of association closeness between the ith step and the i-1 th step, n represents the total number of operation steps; Indicating the direction judgment coefficient for the ith working step, and when the direction is positive, The value is 1, when the direction is the negative direction,The value is-1;
And determining an adjustment grade according to the influence coefficient and the influence of a preset unit, and adjusting the initial grade to obtain a quality grade.
Preferably, the adjusting the initial grade to obtain the quality grade includes:
wherein ZD represents a quality grade, CD represents an initial grade, TD represents an adjustment grade; representing rounding down symbols.
In this embodiment of the present invention, the process is performed,The value of (2) is from 0 to 1, the closer the value is to 1, the more sensitive the ith step is affected by the previous step, and the closer the value is to 0, the less sensitive the ith step is affected by the previous step.
The value of (1) is from 0 to 1, the closer the value is to 1, indicating that the two steps are more closely related, and the closer the value is to 0, indicating that the more loosely related.
CD represents the initial grade, which is the quality grade determined before considering the effect of the operation step connection, and the value is a non-negative integer.
The influence of the previous step on the current step can be comprehensively and accurately measured, and in actual production, the operation steps are related to each other, and the quality of the previous step can influence the subsequent step. Regardless of this effect, the quality assessment may be inaccurate. The formula can more truly reflect the quality of the operation steps by considering the step connection influence.
In this embodiment, d is generally 1.
The technical scheme has the advantages that the quality of the operation steps can be reflected more truly by considering the step connection influence, the accurate quality grade provides a reliable basis for production decision, a scientific and reasonable quality grade determination mode is provided, complex association relations can be processed, and the requirement for accurate quality management in complex production environments is met.
The invention provides a quality information management method based on a multi-mode AI, which determines a storage address of multi-mode data corresponding to a working step in a database II, and comprises the following steps:
According to the step weight of each operation step, combining the quality grade of each operation step to obtain a comprehensive grade, and obtaining the comprehensive difference between the comprehensive grade and a comprehensive standard;
obtaining a difference array Cs=according to the quality grade and the set standard , wherein,Indicating a level difference of the ith job step;
counting symbol combinations { E1, E2} of each remaining operation step and the next operation step except the nth operation step, wherein E1 is a symbol corresponding to the remaining operation step;
if the corresponding residual steps do not meet the set standard, the storage address in the database II is obtained by matching from the step-symbol combination-double difference-address comparison table according to the step weight, the symbol combination, the grade difference and the comprehensive difference of the corresponding residual steps.
In this embodiment, the comprehensive standard is a preset standard for measuring the quality of the whole production process or the product, for example, a quality grade with a value of 5, and considering the step weight can highlight the influence of the key operation step on the whole quality, and the comprehensive grade is calculated by combining the quality grade to comprehensively reflect the quality condition of the production process. The calculation of the integrated differences provides a clear direction for quality improvement.
In this embodiment, the symbol combination comprises "-", and "- +". The "-" indicates that the corresponding operation steps do not meet the set standard, the "+" indicates that the corresponding operation steps meet the set standard, and the symbol combination can intuitively reflect the quality change condition and the association relationship between the operation steps.
In the embodiment, the step-symbol combination-double difference-address comparison table is a pre-established mapping relation table, and the corresponding relation between the combination conditions of different operation step weights, symbol combination, level difference and comprehensive difference and two storage addresses of the database is recorded. For example, when the task weight is within a certain range, the symbol combination is in a specific form, the level difference and the comprehensive difference satisfy a certain condition, a certain storage area address of the second database is corresponding.
The technical scheme has the advantages that the comprehensive difference between the comprehensive grade and the comprehensive standard is obtained, the quantitative measurement and comparison are carried out on the whole quality, the grade difference of each operation step is presented in an array form, the quality of a single operation step is conveniently analyzed and compared, the symbol combination of each residual operation step except the nth operation step and the next operation step is obtained through statistics, the connection relation of the quality states among the operation steps is recorded, and for the residual operation steps which do not meet the set standard, the storage address of the second database is obtained through matching from the comparison table according to the relevant parameters, so that the accurate storage and positioning of the data of the problem operation step are realized.
The invention provides a quality information management system based on multi-mode AI, as shown in figure 2, comprising:
the data acquisition module is used for determining the operation flow of a target project, and carrying out multi-mode detection on the set target of the corresponding operation step according to the multi-mode detection mode of each operation step in the operation flow to obtain multi-mode data;
The grade evaluation module is used for carrying out data analysis on the multi-mode data in each operation step and evaluating the quality grade of the set target in the corresponding operation step;
the standard storage module is used for storing the multi-mode data in the corresponding operation step into the database I when the quality grade accords with the set standard of the corresponding operation step;
And the nonstandard storage module is used for determining the storage address of the multi-mode data of the corresponding operation step in the database II according to the difference between the quality grade of the corresponding operation step and the set standard when the quality grade does not meet the set standard of the corresponding operation step and combining the comprehensive grade of the target item determined based on all the quality grades, and carrying out storage management.
The technical scheme has the beneficial effects that on one hand, the quality management of the project process is facilitated, on the other hand, the storage management of qualified and unqualified quality data is facilitated, further, the subsequent retrieval and use are facilitated, and the problem checking efficiency is improved.
The invention provides a storage medium, which comprises a memory for storing the quality information management method based on the multi-mode AI.
The invention provides a computer device, which comprises a processor and a quality information management method based on multi-mode AI.
It should be noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present invention, and not for limiting the same, and although the present invention has been described in detail with reference to the above-mentioned embodiments, it should be understood by those skilled in the art that the technical solution described in the above-mentioned embodiments may be modified or some technical features may be equivalently replaced, and these modifications or substitutions do not make the essence of the corresponding technical solution deviate from the spirit and scope of the technical solution of the embodiments of the present invention.
Claims (8)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202511100022.XA CN120598438B (en) | 2025-08-07 | 2025-08-07 | Quality information management method and system based on multi-mode AI |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202511100022.XA CN120598438B (en) | 2025-08-07 | 2025-08-07 | Quality information management method and system based on multi-mode AI |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN120598438A CN120598438A (en) | 2025-09-05 |
| CN120598438B true CN120598438B (en) | 2025-10-28 |
Family
ID=96893935
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202511100022.XA Active CN120598438B (en) | 2025-08-07 | 2025-08-07 | Quality information management method and system based on multi-mode AI |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN120598438B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114612934A (en) * | 2022-03-14 | 2022-06-10 | 银河水滴科技（北京）有限公司 | Gait sequence evaluation method and system based on quality dimension |
| CN118503832A (en) * | 2024-07-16 | 2024-08-16 | 杭州祐全科技发展有限公司 | Industrial intelligent detection method and system based on multi-mode large model |
| CN119149964A (en) * | 2024-11-19 | 2024-12-17 | 水利部交通运输部国家能源局南京水利科学研究院 | Buried water delivery pressure steel pipe disease inspection method and system |
| CN119760475A (en) * | 2024-12-16 | 2025-04-04 | 华电西港发电有限公司 | Behavior recognition method, apparatus, device, storage medium, and computer program product |
| CN120429586A (en) * | 2025-06-25 | 2025-08-05 | 浙江安防职业技术学院 | A multimodal fault prediction method and system for industrial equipment based on deep learning |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2531605B2 (en) * | 1984-02-24 | 1996-09-04 | 株式会社東芝 | Image registration device |
| US20260010854A1 (en) * | 2023-04-20 | 2026-01-08 | Beijing Boe Technology Development Co., Ltd. | Information processing method and apparatus, evaluation parameter output apparatus, and system and medium |
-
2025
- 2025-08-07 CN CN202511100022.XA patent/CN120598438B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114612934A (en) * | 2022-03-14 | 2022-06-10 | 银河水滴科技（北京）有限公司 | Gait sequence evaluation method and system based on quality dimension |
| CN118503832A (en) * | 2024-07-16 | 2024-08-16 | 杭州祐全科技发展有限公司 | Industrial intelligent detection method and system based on multi-mode large model |
| CN119149964A (en) * | 2024-11-19 | 2024-12-17 | 水利部交通运输部国家能源局南京水利科学研究院 | Buried water delivery pressure steel pipe disease inspection method and system |
| CN119760475A (en) * | 2024-12-16 | 2025-04-04 | 华电西港发电有限公司 | Behavior recognition method, apparatus, device, storage medium, and computer program product |
| CN120429586A (en) * | 2025-06-25 | 2025-08-05 | 浙江安防职业技术学院 | A multimodal fault prediction method and system for industrial equipment based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN120598438A (en) | 2025-09-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20220146462A1 (en) | Method for predicting remaining life of numerical control machine tool | |
| JP5539382B2 (en) | Identify faults in aero engines | |
| CN114237155A (en) | Error prediction and compensation method, system and medium for multi-axis CNC machining | |
| KR102212022B1 (en) | Method of automatically determining condition of hydro turbine in hydroelectric power plant and system for the same | |
| CN112307579B (en) | Thermal error temperature sensitive point selection method and system based on temperature comprehensive information | |
| CN116075733A (en) | Battery management system for sorting battery modules | |
| CN118403948B (en) | Optimized machining method and system for composite stamping die | |
| CN119199523B (en) | Main shaft servo motor performance evaluation method and device adopting data tracing | |
| CN120338618B (en) | Dynamic key region process inspection method and system based on real-time data driving | |
| CN115659271A (en) | Sensor anomaly detection method, model training method, system, device and medium | |
| CN118243378A (en) | Fault diagnosis method and system for cutting speed reducer of cantilever type heading machine | |
| CN111580463A (en) | Method for selecting thermal key points of machine tool | |
| CN120598438B (en) | Quality information management method and system based on multi-mode AI | |
| CN120703003A (en) | A glass production line cold end stress detection system and method | |
| CN118503884B (en) | Equipment state identification method, equipment and medium | |
| CN118277940B (en) | A method and system for safety control of assembled equipment with multiple fixed points | |
| CN119087892A (en) | A production monitoring system for aluminum alloy storage tanks | |
| US12117813B2 (en) | Component inspection system and method | |
| CN117131364A (en) | Rolling bearing process detection integration method and system | |
| CN120069494B (en) | A big data processing and analysis method, system, electronic device, and storage medium | |
| CN120765121B (en) | Information entropy calculation method and equipment for evaluating stability of wafer processing machine | |
| CN120744527B (en) | Special equipment state identification method and system based on electrical parameter pattern matching | |
| CN120901413B (en) | Adaptive wire feeding control method, platform and medium for welding robots | |
| CN120101917B (en) | Weighing monitoring and early warning method and device for smart scales | |
| CN120870842B (en) | Intelligent Chip Testing Method and System Based on ALPG Testing System |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |