An intelligent monitoring method and system for computer hardware production and assembly

Abstract

The present application relates to an intelligent monitoring method and system for computer hardware production and assembly, which includes: receiving production progress data and production quality data at intervals; dividing the production quality data into several homologous quality data groups; performing time series analysis on the homologous quality data groups and production progress data; inputting the features obtained by the time series analysis into the feature analysis model; responding to abnormal features based on the traceability rules and their corresponding abnormal response mechanisms. The present application achieves refined management and monitoring of the production process by conducting in-depth analysis and time series processing of production progress and production quality data, and identifies abnormal features in production data through a feature analysis model, and traces the root causes of abnormal problems in combination with preset traceability rules and triggers a response mechanism for processing. It has the effect of real-time tracking and intelligent adjustment of the assembly process of computer hardware production, reducing the waste of production resources and improving production efficiency.

Description

Technical Field

[0001] The present application relates to the technical field of assembly progress monitoring, and in particular to an intelligent monitoring method and system for computer hardware production and assembly. Background Art

[0002] With the advent of the Industrial 4.0 era, the degree of automation in the computer hardware production field has been continuously improving, and the efficient operation and precise management of the entire production process have become the key to the competitiveness of enterprises. However, when facing the complex automated production process with multiple processes, the existing production monitoring systems often focus on data collection and simple display of a single link, lacking the ability of comprehensive, real-time monitoring and in-depth intelligent analysis of the entire production process. Therefore, it is difficult for the existing production monitoring systems to effectively integrate and analyze the production progress and production quality of each process node.

[0003] More critically, the existing monitoring systems lack effective real-time tracking and intelligent adjustment mechanisms for core indicators such as scrap rate and efficiency in the production process. This means that when an abnormality occurs in a certain production link, the system cannot identify and automatically trigger the corresponding response mechanism in time, resulting in the problem not being processed in time, and further causing waste of production resources and low production efficiency.

[0004] In summary, the existing production monitoring systems have obvious deficiencies in comprehensive monitoring, intelligent analysis, and abnormal response, and cannot meet the requirements of efficient and precise management in the computer hardware production field in the Industrial 4.0 era. Summary of the Invention

[0005] In order to solve the above technical problems, the present application provides an intelligent monitoring method and system for computer hardware production and assembly.

[0006] The first invention object of the present application is achieved through the following technical solutions:

[0007] An intelligent monitoring method for computer hardware production and assembly, comprising the steps of:

[0008] Receiving production data sent by a monitoring terminal at intervals, the production data including production progress data and production quality data;

[0009] Obtaining historical production quality data, and based on the historical production quality data and production progress data, dividing the production quality data associated with each process node into several homologous quality data groups;

[0010] Performing time series analysis on the homologous quality data groups and production progress data respectively to obtain efficiency trend characteristics, quality trend characteristics, efficiency periodic change characteristics, and quality periodic change characteristics;

[0011] Input the efficiency trend feature, quality trend feature, efficiency cycle change feature, and quality cycle change feature into a pre-trained feature analysis model;

[0012] Based on pre-set traceability rules and their corresponding abnormal response mechanisms, respond to the abnormal features output by the feature analysis model.

[0013] By adopting the above technical solution, receive the production data monitored by the monitoring terminal at a pre-set interval. The production data includes production progress data representing the progress of each process node and production quality data representing the quality of the produced products. Obtain the product quality data in the historical production process, and based on the historical production quality data and production progress data, integrate the production quality data associated with each process node and divide it into several different types of homologous quality data groups. Perform time series analysis on the divided homologous quality data groups and production progress data respectively to reveal the trends of production quality and production efficiency changing over time and their possible periodic fluctuation characteristics. Input the extracted efficiency trend feature, quality trend feature, efficiency cycle change feature, and quality cycle change feature into a pre-trained feature analysis model for feature analysis and identify abnormal features. When receiving the abnormal features output by the feature analysis model, trace the root cause of the problem according to the pre-set traceability rules, and take corresponding measures based on the pre-set corresponding abnormal response mechanism for processing; Through in-depth analysis and time series processing of production progress and production quality data, this application realizes the refined management and monitoring of the production process, and identifies abnormal features in production data through a pre-trained feature analysis model. Combining the pre-set traceability rules, it traces the root cause of abnormal problems and triggers corresponding response mechanisms for processing, having the effect of real-time tracking and intelligent adjustment of the assembly process of computer hardware production, thereby optimizing the process, reducing waste of production resources, and improving production efficiency.

[0014] In a preferred example, this application can be further configured as follows: The step of obtaining the historical production quality data and dividing the production quality data associated with each process node into several homologous quality data groups based on the historical production quality data and production progress data includes the steps:

[0015] Construct a process network model based on the production progress data and the pre-set process sequence;

[0016] Based on the historical production quality data, perform key mapping and correlation analysis on the production quality data of each process node in the process network model to obtain several process quality feature matrices;

[0017] Divide several process quality feature matrices into several homologous quality data groups through cluster analysis.

[0018] By adopting the above technical solution, a network model capable of representing production processes and their interrelationships is constructed based on production progress data and a preset process sequence; based on the historical production quality data on the basis of the process network model, in-depth analysis is carried out on the production quality data of each process node, and the internal connections between each process node are revealed through key mapping and correlation analysis, and the analysis results are sorted into several process quality feature matrices for subsequent division of homologous quality data groups; the process quality feature matrices are divided by clustering analysis technology into several homologous quality data groups with common features or common sources, and the data within the divided homologous quality data groups are closer in quality features, facilitating subsequent analysis and processing.

[0019] In a preferred example of the present application, it can be further configured that: the step of performing key mapping and correlation analysis on the production quality data of each process node of the process network model based on the historical production quality data to obtain several process quality feature matrices includes the steps of:

[0020] Perform key analysis on the historical production quality data, and identify the key factors contained in the production quality data corresponding to each process node based on the key analysis results;

[0021] Calculate the key scores of each process node based on the identified key factors and their associated key weight coefficients;

[0022] Construct a key mapping for each process node based on the key scores of each process node;

[0023] Based on the constructed key mapping, identify the association types between each process node and quantify their association strengths based on preset quantization rules, and the association types include no association, serial association, parallel association, and quality transfer association;

[0024] Fill the key scores, association relationships, and their corresponding association strengths of each process node into the matrix to form several process quality feature matrices.

[0025] By adopting the above technical solution, critical analysis is carried out on historical production quality data, and then based on the results of the critical analysis, the production quality data of each process node is analyzed to identify the key factors affecting product quality. A weight reflecting the importance of each factor to product quality is assigned to each identified key factor, and the criticality score of each process node is calculated based on the assigned weight; a criticality mapping corresponding to showing the importance of each process node in terms of product quality is created based on the criticality score, and the association type between process nodes is determined based on the criticality mapping. At the same time, the strength of the determined association type is quantified; the criticality scores, association relationships, and their corresponding association strengths of each process node in the foregoing steps are filled into a matrix to form a number of process quality characteristic matrices for comprehensively describing the quality characteristics of the process network model.

[0026] In a preferred example of the present application, it can be further configured that: the step of identifying the association type between each process node and quantifying its association strength based on a preset quantification rule based on the constructed criticality mapping, where the association type includes non-association, serial association, parallel association, and quality transfer association, includes the steps:

[0027] Perform standardization processing on the association strength of each association type;

[0028] Quantify the superposition state of the association types by superimposing the standardized association strengths.

[0029] By adopting the above technical solution, the association strengths of different association types are converted into a standardized unified scale, and the standardized association strengths are superimposed to quantify the superposition state of the complex association types with multiple association types superimposed.

[0030] In a preferred example of the present application, it can be further configured that: the step of respectively performing time series analysis on the homologous quality data group and production progress data to obtain efficiency trend characteristics, quality trend characteristics, efficiency cycle change characteristics, and quality cycle change characteristics, includes the steps:

[0031] Based on time series, the homologous quality data group and production progress data are converted into a homologous quality data sequence and a production progress data sequence;

[0032] Perform time series decomposition on the homologous quality data sequence and production progress data sequence to obtain a number of time sub-sequences, where the time sub-sequences include long-term trend sequences, seasonal variation sequences, cyclic fluctuation sequences, and irregular fluctuation sequences;

[0033] Input the time sub-sequences into a pre-trained time series model for feature extraction.

[0034] By adopting the above technical solution, the original homologous quality data and production progress data are sorted in chronological order based on time series to form corresponding time series (homologous quality data series and production progress data series), and the homologous quality data series and production progress data series are decomposed into several different time sub-series based on time series decomposition technology. The time sub-series include long-term trend series, seasonal variation series, cyclic fluctuation series, and irregular fluctuation series, which respectively represent different components in the data. The time sub-series are input into a pre-trained time series model to extract efficiency trend features, quality trend features, efficiency cycle change features, and quality cycle change features for subsequent analysis and decision-making.

[0035] In a preferred example of the present application, it can be further configured that: the time series model includes a fitting layer and an extraction layer. The step of inputting the time sub-series into the pre-trained time series model for feature extraction includes the steps:

[0036] The fitting layer performs trend fitting on the time sub-series associated with the production progress data to obtain several fitting parameter features;

[0037] The extraction layer performs feature extraction on the fitting parameter features to obtain efficiency trend features and efficiency cycle change features.

[0038] By adopting the above technical solution, the time series model includes a fitting layer and an extraction layer. You are responsible for processing the time sub-series related to the production progress data, performing trend fitting on it to obtain several different fitting parameter features. The extraction layer receives the fitting parameter features generated by the fitting layer and further performs feature extraction to extract efficiency trend features and efficiency cycle change features. The efficiency trend features describe the change trend of production efficiency over time, while the efficiency cycle change features reveal the possible periodic change rules of production efficiency.

[0039] In a preferred example of the present application, it can be further configured that: the step of responding to the abnormal features output by the feature analysis model based on the pre-set traceability rules and their corresponding abnormal response mechanism includes the steps:

[0040] When receiving the abnormal features output by the feature analysis model, preprocess the abnormal features;

[0041] Match the preprocessed abnormal features with the pre-set traceability rule library;

[0042] Based on the matched traceability rules and abnormal features, construct an abnormal propagation path graph;

[0043] Perform risk assessment based on the abnormal propagation path graph and generate corresponding strategies according to the predefined response mechanism.

[0044] By adopting the above technical solution, the abnormal features output by the feature analysis model are preprocessed to ensure that they can be effectively matched with the pre-set traceability rule library. The preprocessed abnormal features are compared and matched with the rules in the predefined traceability rule library to identify the traceability rules corresponding to the abnormal features. An abnormal propagation path map is constructed based on the identified traceability rules and abnormal features, and a risk assessment is carried out based on the abnormal propagation path map, so as to analyze the potential impact of the abnormal situation on the production process and product quality, and corresponding strategies are generated according to the predefined response mechanism to handle the abnormal situation.

[0045] The second invention object of the present application is achieved by the following technical solutions:

[0046] An intelligent monitoring system for computer hardware production and assembly, comprising:

[0047] A data receiving module, configured to receive production data sent by a monitoring terminal at intervals, where the production data includes production progress data and production quality data;

[0048] A homologous data group division module, configured to obtain historical production quality data, and based on the historical production quality data and production progress data, divide the production quality data associated with each process node into several homologous quality data groups;

[0049] A time series analysis module, configured to perform time series analysis on the homologous quality data groups and production progress data respectively to obtain efficiency trend features, quality trend features, efficiency cycle change features, and quality cycle change features;

[0050] An input module, configured to input the efficiency trend features, quality trend features, efficiency cycle change features, and quality cycle change features into a pre-trained feature analysis model;

[0051] A response module, configured to respond to the abnormal features output by the feature analysis model based on the pre-set traceability rules and their corresponding abnormal response mechanisms.

[0052] By adopting the above technical solution, a data receiving module is configured to intermittently receive production data sent by a monitoring terminal, where the production data includes production progress data and production quality data; a homologous data group division module is configured to obtain historical production quality data, and based on the historical production quality data and the production progress data, divide the production quality data associated with each process node into a plurality of homologous quality data groups; a time series analysis module is configured to perform time series analysis on the homologous quality data groups and the production progress data respectively to obtain an efficiency trend feature, a quality trend feature, an efficiency cycle change feature, and a quality cycle change feature; an input module is configured to input the efficiency trend feature, the quality trend feature, the efficiency cycle change feature, and the quality cycle change feature into a pre-trained feature analysis model; a response module is configured to respond to abnormal features output by the feature analysis model based on a preset traceability rule and its corresponding abnormal response mechanism.

[0053] In summary, the present application includes at least one of the following beneficial technical effects:

[0054] 1. By deeply analyzing and performing time series processing on production progress and production quality data, the present application realizes refined management and monitoring of the production process, and identifies abnormal features in production data through a pre-trained feature analysis model, traces the root cause of abnormal problems in combination with a preset traceability rule, and triggers a corresponding response mechanism for processing, having the effects of real-time tracking and intelligent adjustment of the assembly process of computer hardware production, thereby optimizing the process, reducing waste of production resources, and improving production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 is a flowchart of an embodiment of an intelligent monitoring method for computer hardware production and assembly according to the present application;

[0056] Figure 2 is an implementation flowchart of step S20 in an embodiment of an intelligent monitoring method for computer hardware production and assembly according to the present application;

[0057] Figure 3 is an implementation flowchart of step S22 in an embodiment of an intelligent monitoring method for computer hardware production and assembly according to the present application;

[0058] Figure 4 is an implementation flowchart of step S30 in an embodiment of an intelligent monitoring method for computer hardware production and assembly according to the present application;

[0059] Figure 5 is an implementation flowchart of step S50 in an embodiment of an intelligent monitoring method for computer hardware production and assembly according to the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] The following will further elaborate on this application in conjunction with the accompanying drawings. Figures 1 - 5 A further detailed description of this application will be given below.

[0061] In one embodiment, as Figure 1 shown, this application discloses an intelligent monitoring method for computer hardware production and assembly, specifically including the following steps:

[0062] S10: Receive production data sent by the monitoring terminal at intervals, where the production data includes production progress data and production quality data;

[0063] In this embodiment, the monitoring terminal is a terminal device for collecting and sending production data; the production data is a comprehensive data set containing production progress data and production quality data;

[0064] Specifically, receive the production data monitored by the monitoring terminal at preset intervals, where the production data includes production progress data representing the progress of each process node and production quality data representing the quality of the produced products;

[0065] S20: Obtain historical production quality data, and based on the historical production quality data and production progress data, divide the production quality data associated with each process node into several homologous quality data groups;

[0066] In this embodiment, the historical production quality data is the product quality data stored in the historical production process; the process node is a specific process link in the production process; the homologous quality data group is a set of quality data with the same or similar characteristics, sources, or influencing factors;

[0067] Specifically, obtain the product quality data in the historical production process, and based on the historical production quality data and production progress data, integrate the production quality data associated with each process node and divide it into several different types of homologous quality data groups;

[0068] S30: Perform time series analysis on the homologous quality data groups and production progress data respectively to obtain efficiency trend characteristics, quality trend characteristics, efficiency cycle change characteristics, and quality cycle change characteristics;

[0069] In this embodiment, time series analysis is an analysis method used to extract meaningful information and characteristics from data arranged in chronological order, specifically to extract the changing trends of production efficiency and production quality over time and their possible periodic fluctuation characteristics; the efficiency trend characteristic is the overall trend of production efficiency changing over time; the quality trend characteristic is the overall trend of production quality changing over time; the efficiency cycle change characteristic is the periodic fluctuation characteristic of production efficiency presented by the production progress data over time; the quality cycle change characteristic is the periodic fluctuation characteristic of production quality over time;

[0070] Specifically, time series analysis is performed on the divided homologous quality data groups and production progress data respectively to reveal the trends of production quality and production efficiency over time and their possible periodic fluctuation characteristics;

[0071] S40: Input the efficiency trend characteristics, quality trend characteristics, efficiency periodic change characteristics, and quality periodic change characteristics into a pre-trained feature analysis model;

[0072] In this embodiment, the feature analysis model is a machine learning model for identifying and analyzing data features. Specifically, it identifies anomalies from the input efficiency trend characteristics, quality trend characteristics, efficiency periodic change characteristics, and quality periodic change characteristics and extracts the corresponding anomaly features;

[0073] Specifically, the extracted efficiency trend characteristics, quality trend characteristics, efficiency periodic change characteristics, and quality periodic change characteristics are input into a pre-trained feature analysis model for feature analysis and anomaly features are identified;

[0074] S50: Respond to the anomaly features output by the feature analysis model based on pre-set traceability rules and their corresponding anomaly response mechanisms;

[0075] In this embodiment, the pre-set traceability rules are rules for tracing and locating the root causes of abnormal situations; the anomaly response mechanism is a predefined measure taken when a preset anomaly is detected;

[0076] Specifically, when the anomaly features output by the feature analysis model are received, the root cause of the problem is traced according to the pre-set traceability rules, and corresponding measures are taken based on the preset corresponding anomaly response mechanism for processing.

[0077] In one embodiment, as Figure 2 shown, step S20 includes the steps:

[0078] S21: Construct a process network model based on the production progress data and the pre-set process sequence;

[0079] S22: Perform critical mapping and correlation analysis on the production quality data of each process node in the process network model based on historical production quality data to obtain several process quality feature matrices;

[0080] S23: Divide several process quality feature matrices into several homologous quality data groups through cluster analysis;

[0081] In this embodiment, the process network model is a network structure used to represent production processes and their interrelationships, where nodes represent processes and edges represent the dependency or sequential relationships between processes; the criticality mapping is an analysis method, specifically for identifying and mapping the key factors that affect production quality; the association analysis is a statistical method used to reveal the internal connections and correlations between data, specifically for performing association analysis on the production quality data of each process node to reveal their connections and correlations; the process quality feature matrix is a matrix used to represent the quality features of each process node and their interrelationships; the clustering analysis is a data mining method used to divide data into several groups or clusters with similar features; the homologous quality data group is a set of quality data with the same or similar features, sources, or influencing factors, and is obtained through clustering analysis in this embodiment;

[0082] Specifically, based on the production progress data and the preset process sequence, a network model that can represent production processes and their interrelationships is constructed; based on the historical production quality data on the basis of the process network model, in-depth analysis is performed on the production quality data of each process node, and the internal connections between each process node are revealed through criticality mapping and association analysis, and the analysis results are sorted into several process quality feature matrices for subsequent homologous quality data group division; the process quality feature matrix is divided by clustering analysis technology into several homologous quality data groups with common features or common sources, and the data within the divided homologous quality data groups are closer in quality features, facilitating subsequent analysis and processing.

[0083] In one embodiment, as Figure 3 shown, step S22 includes the steps:

[0084] S221: Perform criticality analysis on the historical production quality data, and identify the key factors contained in the production quality data of each process node based on the criticality analysis results;

[0085] S222: Calculate the criticality scores of each process node based on the identified key factors and their associated criticality weight coefficients;

[0086] S223: Construct the criticality mapping of each process node based on the criticality scores of each process node;

[0087] S224: Based on the constructed criticality mapping, identify the association types between each process node and quantify their association strengths based on the preset quantization rules, and the association types include no association, serial association, parallel association, and quality transfer association;

[0088] S225: Fill the criticality scores, association relationships, and their corresponding association strengths of each process node into the matrix to form several process quality feature matrices;

[0089] In this embodiment, the criticality analysis is to deeply analyze the data to identify the factors that have a significant impact on it. Specifically, it is to deeply analyze the historical production quality data to identify and analyze the factors that have an impact on it; the critical factors are the factors that have a significant impact on product quality; the critical weight coefficient is a value representing the degree of influence of the critical factors on product quality; the critical score is a value reflecting the importance of the process nodes to product quality; the critical mapping is a carrier for displaying the relative importance of each process node in terms of product quality; the association type is the type of relationship between process nodes, which includes no association, serial association, parallel association, and quality transfer association in this embodiment; the association strength is a value representing the degree of closeness of the relationship between process nodes; the process quality characteristic matrix is a matrix containing the critical scores, association relationships, and association strengths of process nodes, and is used to comprehensively describe the quality characteristics of the process network model.

[0090] Specifically, conduct a criticality analysis on the historical production quality data, and then analyze the production quality data of each process node based on the results of the criticality analysis to identify the critical factors that affect product quality. Assign a weight to each identified critical factor that reflects the importance of the factor to product quality, and calculate the critical score of each process node based on the assigned weight; create a critical mapping corresponding to display the importance of each process node in terms of product quality based on the critical score, and determine the association type between process nodes based on the critical mapping, and at the same time quantify the strength of the determined association type; fill the critical scores, association relationships, and their corresponding association strengths of each process node in the above steps into the matrix to form several process quality characteristic matrices, which are used to comprehensively describe the quality characteristics of the process network model.

[0091] In one embodiment, step S224 includes the steps:

[0092] S2241: Standardize the association strength of each association type;

[0093] S2242: Quantify the superposition state of the association types by superimposing the standardized association strengths.

[0094] In this embodiment, the standardization process is used to convert the association strength values of different association types into a unified measurement standard; the superposition state of the association types is a state with several different association types.

[0095] Specifically, convert the association strengths of different association types into a standardized unified scale, and superimpose the standardized association strengths to quantify the superposition state of the association types with multiple association types superimposed.

[0096] In one embodiment, as Figure 4As shown, step S30 includes the steps:

[0097] S31: Based on the time sequence, transform the homologous quality data group and production progress data into a homologous quality data sequence and a production progress data sequence;

[0098] S32: Perform time series decomposition on the homologous quality data sequence and production progress data sequence to obtain a number of time subsequences, where the time subsequences include a long-term trend sequence, a seasonal variation sequence, a cyclic fluctuation sequence, and an irregular fluctuation sequence;

[0099] S33: Input the time subsequences into a pre-trained time series model for feature extraction;

[0100] In this embodiment, time series decomposition is a method of decomposing time series data into different components (such as long-term trends, seasonal variations, etc.); time subsequences are sequences representing different data components obtained through time series decomposition, including long-term trend sequences, seasonal variation sequences, cyclic fluctuation sequences, and irregular fluctuation sequences; the time series model is a time series model that can be used for feature extraction or prediction, specifically for extracting trend features and periodic change features from long-term trend sequences, seasonal variation sequences, cyclic fluctuation sequences, and irregular fluctuation sequences;

[0101] Specifically, based on the time sequence, sort the original homologous quality data and production progress data in chronological order to form corresponding time sequences (homologous quality data sequence and production progress data sequence), and based on time series decomposition technology, decompose the homologous quality data sequence and production progress data sequence into several different time subsequences. The time subsequences include long-term trend sequences, seasonal variation sequences, cyclic fluctuation sequences, and irregular fluctuation sequences, which respectively represent different components in the data. Input the time subsequences into a pre-trained time series model to extract efficiency trend features, quality trend features, efficiency periodic change features, and quality periodic change features for subsequent analysis and decision-making.

[0102] In one embodiment, the time series model includes a fitting layer and an extraction layer, and step S33 includes the steps:

[0103] S331: The fitting layer performs trend fitting on the time subsequence associated with the production progress data to obtain several fitting parameter features;

[0104] S332: The extraction layer performs feature extraction on the fitting parameter features to obtain efficiency trend features and efficiency periodic change features;

[0105] In this embodiment, the fitting layer is a component in the time series model, responsible for trend fitting of the time subsequence to generate fitting parameter features; the extraction layer is another component in the time series model, responsible for receiving the fitting parameter features generated by the fitting layer and performing feature extraction to obtain specific efficiency features and quality features; trend fitting is the process of using a mathematical function or model to describe the trend in data; fitting parameter features are features that describe the trend characteristics of data generated through the trend fitting process.

[0106] Specifically, the time series model includes a fitting layer and an extraction layer. You are responsible for processing the time subsequence related to the production progress data, performing trend fitting on it to obtain several different fitting parameter features. The extraction layer receives the fitting parameter features generated by the fitting layer and further performs feature extraction to extract the efficiency trend feature and the efficiency periodic change feature. The efficiency trend feature describes the change trend of production efficiency over time, while the efficiency periodic change feature reveals the possible periodic change law of production efficiency.

[0107] In one embodiment, as Figure 5 shown, step S50 includes the steps:

[0108] S51: When receiving the abnormal features output by the feature analysis model, preprocess the abnormal features;

[0109] S52: Match the preprocessed abnormal features with the pre-set traceability rule library;

[0110] S53: Based on the matched traceability rules and abnormal features, construct an abnormal propagation path diagram;

[0111] S54: Perform risk assessment based on the abnormal propagation path diagram and generate corresponding strategies according to the predefined response mechanism;

[0112] In this embodiment, the abnormal features are features output by the feature analysis model used to represent abnormalities or deviations from the normal state in the production process; preprocessing is the process of cleaning, transforming, and standardizing the abnormal features to ensure that the abnormal features can be effectively matched with the traceability rule library; the traceability rule library is a predefined set of rules that define the possible causes, propagation paths, and potential impacts of abnormalities; the abnormal propagation path diagram is a graphical representation showing the propagation path of an abnormality from the source to the current state, including possible impacts and associated other abnormalities; risk assessment is the process of analyzing the abnormal propagation path diagram to evaluate the potential impact of the abnormality on the production process or product quality.

[0113] Specifically, preprocess the abnormal features output by the feature analysis model to ensure that they can be effectively matched with the pre-set traceability rule library. Compare and match the preprocessed abnormal features with the rules in the predefined traceability rule library to identify the traceability rules corresponding to the abnormal features. Construct an abnormal propagation path diagram based on the identified traceability rules and abnormal features, and perform risk assessment based on the abnormal propagation path diagram, so as to analyze the potential impact of abnormal situations on the production process and product quality, and generate corresponding strategies to handle the abnormal situations according to the predefined response mechanism.

[0114] It should be understood that the sequence numbers of the steps in the above embodiments do not indicate the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0115] In one embodiment, an intelligent monitoring system for computer hardware production and assembly is provided. The intelligent monitoring system for computer hardware production and assembly corresponds one-to-one with the intelligent monitoring method for computer hardware production and assembly in the above embodiment. The intelligent monitoring system for computer hardware production and assembly includes:

[0116] A data receiving module, configured to receive production data sent by the monitoring terminal at intervals, where the production data includes production progress data and production quality data;

[0117] A homologous data group division module, configured to obtain historical production quality data, and based on the historical production quality data and production progress data, divide the production quality data associated with each process node into several homologous quality data groups;

[0118] A time series analysis module, configured to perform time series analysis on the homologous quality data groups and production progress data respectively to obtain efficiency trend features, quality trend features, efficiency cycle change features, and quality cycle change features;

[0119] An input module, configured to input the efficiency trend features, quality trend features, efficiency cycle change features, and quality cycle change features into a pre-trained feature analysis model;

[0120] A response module, configured to respond to the abnormal features output by the feature analysis model based on the pre-set traceability rules and their corresponding abnormal response mechanisms;

[0121] Optionally, the homologous data group division module includes:

[0122] A network model construction sub-module, configured to construct a process network model based on the production progress data and the pre-set process sequence;

[0123] A feature matrix acquisition sub-module, which is used to perform critical mapping and correlation analysis on the production quality data of each process node in the process network model based on historical production quality data, and obtain a number of process quality feature matrices;

[0124] A clustering analysis sub-module, which is used to divide a number of process quality feature matrices into several homologous quality data groups through clustering analysis;

[0125] Optionally, the feature matrix acquisition sub-module includes:

[0126] A critical factor identification sub-module, which is used to perform critical analysis on historical production quality data, and identify the critical factors contained in the production quality data of each process node based on the results of the critical analysis;

[0127] A critical score calculation sub-module, which is used to calculate the critical scores of each process node based on the identified critical factors and their associated critical weight coefficients;

[0128] A critical mapping construction sub-module, which is used to construct the critical mapping of each process node based on the critical scores of each process node;

[0129] An association strength quantification sub-module, which is used to identify the association types between each process node based on the constructed critical mapping and quantify their association strength based on a preset quantification rule, and the association types include no association, serial association, parallel association, and quality transfer association;

[0130] A matrix construction sub-module, which is used to fill the critical scores, association relationships, and their corresponding association strengths of each process node into a matrix to form a number of process quality feature matrices;

[0131] Optionally, the association strength quantification sub-module includes:

[0132] A normalization processing sub-module, which is used to perform normalization processing on the association strength of each association type;

[0133] An overlay quantification sub-module, which is used to quantify the association types in the overlay state by overlaying the association strength after normalization processing;

[0134] Optionally, the time series analysis module includes:

[0135] A time series conversion sub-module, which is used to convert the homologous quality data group and production progress data into a homologous quality data sequence and a production progress data sequence based on time series;

[0136] A time series decomposition sub-module, which is used to decompose the homologous quality data sequence and the production progress data sequence into time series to obtain several time sub-series, and the time sub-series include a long-term trend series, a seasonal variation series, a cyclic fluctuation series, and an irregular fluctuation series;

[0137] A time sub-series input sub-module, which is used to input the time sub-series into a pre-trained time series model for feature extraction;

[0138] Optionally, the time series model includes a fitting layer and an extraction layer, and the time sub-series input sub-module includes:

[0139] A fitting layer sub-module, which is used to perform trend fitting on the time sub-series associated with the production progress data to obtain several fitting parameter features;

[0140] An extraction layer sub-module, which is used to extract features from the fitting parameter features to obtain an efficiency trend feature and an efficiency cycle change feature;

[0141] Optionally, the response module includes:

[0142] A preprocessing sub-module, which is used to preprocess the abnormal features when receiving the abnormal features output by the feature analysis model;

[0143] A traceability rule matching sub-module, which is used to match the preprocessed abnormal features with a pre-set traceability rule library;

[0144] A path diagram construction sub-module, which is used to construct an abnormal propagation path diagram based on the matched traceability rules and abnormal features;

[0145] A risk assessment sub-module, which is used to perform risk assessment based on the abnormal propagation path diagram and generate corresponding strategies according to a predefined response mechanism.

[0146] For the specific limitations of an intelligent monitoring system for computer hardware production and assembly, reference can be made to the limitations of an intelligent monitoring method for computer hardware production and assembly in the above text, which will not be elaborated here. Each module in the above intelligent monitoring system for computer hardware production and assembly can be implemented in whole or in part by software, hardware, and their combination. The above modules can be embedded in the processor of the computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above modules.

[0147] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. An intelligent monitoring method for computer hardware production and assembly, characterized in that: Includes steps: receiving production data sent by the monitoring terminal at intervals, wherein the production data includes production progress data and production quality data; Obtain historical production quality data, and based on the historical production quality data and production progress data, divide the production quality data associated with each process node into several homologous quality data groups; Conduct time series analysis on homologous quality data groups and production progress data to obtain efficiency trend characteristics, quality trend characteristics, efficiency cycle change characteristics, and quality cycle change characteristics; Inputting the efficiency trend feature, the quality trend feature, the efficiency cycle change feature, and the quality cycle change feature into a pre-trained feature analysis model; Analyze the abnormal characteristics of the model output based on the preset traceability rules and their corresponding abnormal response mechanism response characteristics; The step of obtaining historical production quality data and dividing the production quality data associated with each process node into a plurality of homologous quality data groups based on the historical production quality data and the production progress data comprises the steps of: Build a process network model based on production quality data and pre-set process sequence; Based on the historical production quality data, the production quality data of each process node in the process network model is critically mapped and correlated to obtain several process quality feature matrices; Several process quality feature matrices are divided into several homogeneous quality data groups through cluster analysis.

2. The intelligent monitoring method for computer hardware production and assembly according to claim 1, characterized in that: The step of performing critical mapping and correlation analysis on the production quality data of each process node of the process network model based on the historical production quality data to obtain several process quality feature matrices includes the following steps: Conduct critical analysis on historical production quality data, and identify the critical factors corresponding to the production quality data of each process node based on the critical analysis results; Calculate the criticality score of each process node based on the identified criticality factors and their associated criticality weight coefficients; Construct a criticality map of each process node based on the criticality score of each process node; Based on the constructed criticality mapping, the association type between each process node is identified and the association strength thereof is quantified based on a preset quantification rule, wherein the association type includes no association, serial association, parallel association and mass transfer association; The criticality score, correlation relationship and corresponding correlation strength of each process node are filled into the matrix to form several process quality feature matrices.

3. The intelligent monitoring method for computer hardware production and assembly according to claim 2, characterized in that: The critical mapping based on the construction identifies the association type between each process node and quantifies its association strength based on a preset quantification rule, wherein the association type includes no association, serial association, parallel association and mass transfer association, including the steps of: The strength of association for each association type was normalized; The correlation type of the superposition state is quantified by superimposing the normalized correlation intensities.

4. The intelligent monitoring method for computer hardware production and assembly according to claim 1, characterized in that: The step of respectively performing time series analysis on the homologous quality data group and the production progress data to obtain efficiency trend characteristics, quality trend characteristics, efficiency cycle change characteristics and quality cycle change characteristics comprises the steps of: Converting homologous quality data groups and production progress data into homologous quality data sequences and production progress data sequences based on time series; Performing time series decomposition on homologous quality data series and production progress data series to obtain a number of time subsequences, wherein the time subsequences include long-term trend series, seasonal change series, cyclic fluctuation series, and irregular fluctuation series; The time subseries is input into the pre-trained time series model for feature extraction.

5. The intelligent monitoring method for computer hardware production and assembly according to claim 4, characterized in that: The time series model includes a fitting layer and an extraction layer. The step of inputting the time subsequence into the pre-trained time series model for feature extraction includes the steps of: The fitting layer performs trend fitting on the time subseries associated with the production progress data to obtain several fitting parameter features; The extraction layer extracts the fitting parameter features to obtain efficiency trend features and efficiency cycle change features.

6. The intelligent monitoring method for computer hardware production and assembly according to claim 1, characterized in that: The step of analyzing the abnormal features output by the response feature analysis model based on the preset tracing rules and their corresponding abnormal response mechanisms comprises the following steps: When receiving abnormal features output by the feature analysis model, preprocessing the abnormal features; Match the preprocessed abnormal features with the preset tracing rule base; Based on the matched tracing rules and abnormal features, an abnormal propagation path diagram is constructed; Perform risk assessment based on the anomaly propagation path diagram and generate corresponding strategies based on predefined response mechanisms.

7. An intelligent monitoring system for computer hardware production and assembly, used in the steps of an intelligent monitoring method for computer hardware production and assembly as claimed in any one of claims 1 to 6, characterized in that: include: A data receiving module, used for receiving production data sent by the monitoring terminal at intervals, wherein the production data includes production progress data and production quality data; A homologous data group division module is used to obtain historical production quality data, and divide the production quality data associated with each process node into several homologous quality data groups based on the historical production quality data and production progress data; The time series analysis module is used to perform time series analysis on the homologous quality data group and the production progress data to obtain efficiency trend characteristics, quality trend characteristics, efficiency cycle change characteristics and quality cycle change characteristics; An input module, used for inputting efficiency trend features, quality trend features, efficiency periodic change features, and quality periodic change features into a pre-trained feature analysis model; The response module is used to respond to the abnormal features output by the feature analysis model based on the preset tracing rules and their corresponding abnormal response mechanisms.

Images