Component quality problem monitoring method and related equipment

Abstract

The invention provides a component quality problem monitoring method which is characterized by comprising the following steps: calculating a first statistical value corresponding to each sample batch of acquired to-be-monitored data; calculating a second statistical value corresponding to each quality risk factor of the to-be-monitored data; performing quality early warning by using an early warning threshold value determined by pre-calculation and the first statistical value; calculating a risk probability value corresponding to each quality risk factor based on the second statistical value; tracing the quality problem based on the quality risk factor corresponding to the maximum risk probability value; and monitoring the quality problem based on quality early warning and quality problem tracing. According to the embodiment of the invention, the quality problem factors in the whole life cycle of the component are sorted, support is provided for subsequent component quality risk early warning and traceability, the Bayesian network is introduced into quality early warning, traceability is carried out through the maximum risk probability value, accurate traceability and timely feedback of the product quality problem are realized, and the quality of the component is improved. And the problem positioning and solving accuracy is improved.

Description

Technical Field

[0001] The present application relates to the technical field of component quality control, and in particular to a component quality problem monitoring method and related equipment. Background Art

[0002] Component manufacturers still have many deficiencies in quality control and quality management. Some companies lack a sound quality management system and strict quality control standards, which may lead to different types of quality problems in the entire life cycle of components, including production, selection, testing, installation and after-sales. In the traditional model, there is a serious lag in the discovery of component quality problems, which makes it difficult for companies to promptly discover and solve potential quality risks, which may lead to more serious equipment failures, performance degradation and even safety accidents, and have an adverse impact on the company's production and operation and brand image. Currently, more and more component manufacturers have begun to realize the importance of early warning of quality problems, and have established independent early warning systems, committed to using big data and artificial intelligence technologies to predict potential quality problems, and to achieve real-time collection, online monitoring, quality early warning, etc. through information and intelligent means.

[0003] Components have quality risks at different stages throughout their life cycle, and there is currently a lack of a comprehensive and systematic quality risk assessment system and early warning measures. Summary of the invention

[0004] In view of this, the purpose of this application is to propose a component quality problem monitoring method and related equipment.

[0005] Based on the above objectives, the present application provides a method for monitoring component quality problems, including:

[0006] Calculate a first statistical value corresponding to each sample batch of the acquired data to be monitored;

[0007] Calculate a second statistical value corresponding to each quality risk factor of the acquired data to be monitored;

[0008] Using the pre-calculated and determined warning threshold and the first statistical value to issue a quality warning;

[0009] Calculate the risk probability value corresponding to each of the quality risk factors based on the first statistical value;

[0010] Tracing the quality problem of the data to be monitored based on the quality risk factor corresponding to the maximum risk probability value;

[0011] Based on the quality warning and the tracing of the quality problem, the quality problem of the data to be monitored is monitored.

[0012] In a possible implementation, the warning threshold is determined by the following steps:

[0013] Screen the quality risk factors to be screened and form an indicator system;

[0014] Constructing a Bayesian network based on the causal relationship between each quality risk factor in the indicator system, and determining an edge weight matrix in the Bayesian network;

[0015] Based on the edge weight matrix and the sample parameters, a second statistical value is calculated;

[0016] The warning threshold is determined according to the second statistical value.

[0017] In a possible implementation, tracing the quality problem of the data to be monitored based on the quality risk factor corresponding to the maximum risk probability value includes:

[0018] Determine all the quality risk factors corresponding to all the maximum risk probability values;

[0019] Determine the risk chain corresponding to the quality problem of the data to be monitored based on all the quality risk factors and the Bayesian network;

[0020] The quality problems of the data to be monitored are traced based on the risk chain.

[0021] In a possible implementation manner, the second statistical value is calculated by the following formula:

[0022]

[0023] in, represents the second statistical value of the jth quality risk factor of the i-th batch of data samples to be monitored, x ij represents the value of the jth quality risk factor of the i-th data sample batch to be monitored, μ j represents the mean value of the data to be monitored for the jth quality risk factor, Σ ij Represents the covariance of the jth quality risk factor of the i-th batch of data samples to be monitored.

[0024] In a possible implementation manner, calculating the risk probability value corresponding to each of the quality risk factors based on the first statistical value includes:

[0025] Integrating the chi-square distribution density function within the range from zero to the first statistical value to obtain a first probability value;

[0026] The difference between 1 and the first probability value is used as the risk probability value.

[0027] In a possible implementation manner, the first statistical value is calculated by the following formula:

[0028]

[0029] in, represents the first statistical value of the i-th batch of data samples to be monitored, x ij represents the value of the jth quality risk factor of the i-th data sample batch to be monitored, μ j represents the mean value of the data to be monitored for the jth quality risk factor, Ω -1 Represents the inverse of the edge weight matrix.

[0030] Based on the same inventive concept, the embodiment of the present application also provides a device for monitoring component quality problems, including:

[0031] A first statistical value calculation module is configured to calculate a first statistical value corresponding to each sample batch of the acquired data to be monitored;

[0032] A second statistical value calculation module is configured to calculate a second statistical value corresponding to each quality risk factor of the acquired data to be monitored;

[0033] An early warning module is configured to use a pre-calculated early warning threshold and the first statistical value to perform a quality early warning;

[0034] a calculation module, configured to calculate a risk probability value corresponding to each of the quality risk factors based on the second statistical value;

[0035] A tracing module is configured to trace the quality problem of the data to be monitored based on the quality risk factor corresponding to the maximum risk probability value;

[0036] The monitoring module is configured to monitor the quality problem of the data to be monitored based on the quality warning and the tracing of the quality problem.

[0037] Based on the same inventive concept, an embodiment of the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a component quality problem monitoring method as described in any one of the above items is implemented.

[0038] Based on the same inventive concept, an embodiment of the present application further provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute any of the above-mentioned component quality problem monitoring methods.

[0039] Based on the same inventive concept, an embodiment of the present application further provides a computer program product, which includes computer program instructions, and the computer instructions are used to enable the computer program product to execute any of the above-mentioned component quality problem monitoring methods.

[0040] From the above, it can be seen that the component quality problem monitoring method and related equipment provided by the present application, by calculating the first statistical value corresponding to each sample batch of the data to be monitored; calculating the second statistical value corresponding to each quality risk factor of the data to be monitored; using the pre-calculated warning threshold and the first statistical value to perform quality warning; calculating the risk probability value corresponding to each quality risk factor based on the second statistical value; tracing the quality problem of the data to be monitored based on the quality risk factor corresponding to the maximum risk probability value; monitoring the quality problem of the data to be monitored based on the quality warning and the tracing of the quality problem. The embodiment of the present application constructs a component quality risk assessment system covering important stages such as component production, design selection, testing, installation and after-sales through correlation analysis of quality data of different batches of components throughout their life cycle, and deduces the target warning level through the correlation between nodes fitted by the Bayesian probability model, so as to achieve a level warning of component quality risk. Through the risk probability value of each node in the Bayesian network, accurate backtracing of quality risk nodes can be achieved. The above Bayesian network can represent the uncertainty and correlation of random quality risks, achieve accurate prediction and early warning of component quality risk levels, achieve accurate tracing and timely feedback of product quality problems, improve the accuracy of problem location and resolution, and can significantly improve the quality control level of aerospace products. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the present application or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 A schematic diagram of a method for monitoring component quality problems according to an embodiment of the present application;

[0043] Figure 2 A cause-and-effect network diagram of components in an embodiment of the present application;

[0044] Figure 3 A code version component cause-and-effect network diagram of an embodiment of the present application;

[0045] Figure 4 This is a high- and medium-risk ROC curve diagram of an embodiment of the present application;

[0046] Figure 5 This is a low- to medium-risk ROC curve diagram of an embodiment of the present application;

[0047] Figure 6 A schematic diagram showing the statistical results of the Bayesian network test set of an embodiment of the present application;

[0048] Figure 7 A schematic diagram of the traceability process of an embodiment of the present application;

[0049] Figure 8 This is a schematic diagram of the structure of a device for monitoring component quality problems according to an embodiment of the present application;

[0050] Fig. 9 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0051] In order to make the objectives, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.

[0052] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present application should be the usual meanings understood by people with ordinary skills in the field to which the present application belongs. The "first", "second" and similar words used in the embodiments of the present application do not represent any order, quantity or importance, but are only used to distinguish different components. "Including" or "comprising" and similar words mean that the elements or objects appearing in front of the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0053] It is understandable that before using the technical solutions of each embodiment of the present disclosure, the type, scope of use, usage scenarios, etc. of the personal information involved will be informed to the user in an appropriate manner, and the user's authorization will be obtained.

[0054] For example, in response to receiving an active request from a user, a prompt message is sent to the user to clearly remind the user that the operation requested to be performed will require obtaining and using the user's personal information. Thus, the user can independently choose whether to provide personal information to software or hardware such as an electronic device, application, server, or storage medium that performs the operation of the technical solution of the present disclosure according to the prompt message.

[0055] As an optional but non-limiting implementation, in response to receiving the user's active request, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0056] It is understandable that the above notification and the process of obtaining user authorization are merely illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that meet relevant laws and regulations may also be applied to the implementation of the present disclosure.

[0057] As mentioned in the background technology section, component manufacturers still have many deficiencies in quality control and quality management. Some companies lack a complete quality management system and strict quality control standards, which may lead to different types of quality problems in the entire life cycle of components, such as production, selection, testing, installation and after-sales. In the traditional model, there is a serious lag in the discovery of component quality problems, which makes it difficult for companies to promptly discover and solve potential quality risks, which may lead to more serious equipment failures, performance degradation and even safety accidents, and have an adverse impact on the company's production and operation and brand image. At present, more and more component manufacturers have begun to realize the importance of early warning of quality problems, and have established independent early warning systems, committed to using big data and artificial intelligence technologies to predict potential quality problems, and realize real-time collection, online monitoring, quality early warning, etc. through information and intelligent means.

[0058] Components have quality risks at different stages throughout their life cycle, and there is currently a lack of a comprehensive and systematic quality risk assessment system and early warning measures.

[0059] Based on the above considerations, the embodiment of the present application proposes a component quality problem monitoring method and related equipment, which calculates the first statistical value corresponding to each sample batch of the acquired data to be monitored; calculates the second statistical value corresponding to each quality risk factor of the acquired data to be monitored; uses the pre-calculated warning threshold and the first statistical value to perform quality warning; calculates the risk probability value corresponding to each of the quality risk factors based on the second statistical value; traces the quality problem of the data to be monitored based on the quality risk factor corresponding to the maximum risk probability value; monitors the quality problem of the data to be monitored based on the quality warning and the tracing of the quality problem. The embodiment of the present application constructs a component quality risk assessment system covering important stages such as component production, design selection, testing, installation and after-sales through correlation analysis of quality data of different batches of components throughout their life cycle, and deduces the target warning level through the correlation between nodes fitted by the Bayesian probability model, thereby realizing a level warning for component quality risks. Through the risk probability value of each node in the Bayesian network, accurate backtracing of quality risk nodes can be achieved. The above Bayesian network can represent the uncertainty and correlation of random quality risks, achieve accurate prediction and early warning of component quality risk levels, achieve accurate tracing and timely feedback of product quality problems, improve the accuracy of problem location and resolution, and can significantly improve the quality control level of aerospace products.

[0060] The technical solutions of the embodiments of the present application are described in detail below through specific examples.

[0061] refer to Figure 1 The component quality problem monitoring method of the embodiment of the present application comprises the following steps:

[0062] Step S101, calculating a first statistical value corresponding to each sample batch of the acquired data to be monitored;

[0063] Step S102, calculating a second statistical value corresponding to each quality risk factor of the acquired data to be monitored;

[0064] Step S103, using the pre-calculated warning threshold and the first statistical value to perform quality warning;

[0065] Step S104, calculating the risk probability value corresponding to each of the quality risk factors based on the second statistical value;

[0066] Step S105, tracing the quality problem of the data to be monitored based on the quality risk factor corresponding to the maximum risk probability value;

[0067] Step S106: Based on the quality warning and the tracing of the quality problem, the quality problem of the data to be monitored is monitored.

[0068] In the embodiment of the present application, before calculating the first statistical value corresponding to each sample batch of the acquired data to be monitored in step S101, the following steps are first required to determine the warning threshold.

[0069] Specifically, in some embodiments, the warning threshold is determined by the following steps: screening the quality risk factors to be screened to form an indicator system; constructing a Bayesian network based on the causal relationship between each quality risk factor in the indicator system, and determining the edge weight matrix in the Bayesian network; calculating a second statistical value based on the edge weight matrix and sample parameters; and determining the warning threshold according to the second statistical value.

[0070] In the embodiment of the present application, it is first necessary to establish an indicator system, which requires a total of three rounds of Delphi method to establish.

[0071] The first round of Delphi method was conducted. According to the results of the first round of investigation, the expert group believed that there were 21 component quality risk factors.

[0072] Table 1 Results of the first round of survey

[0073]

[0074]

[0075] After that, the second round of evaluation Delphi method was conducted. The second round of evaluation survey needs to be conducted on the premise that the background information of components is provided to the experts. Based on the first round of survey forms, the names of quality risk factors were revised, and "equipment quality risk" was changed to "component design level risk", "material quality risk" was changed to "raw material quality risk", "production control risk" was changed to "manufacturing process level risk", "technical indicator risk" was changed to "component reliability risk (inherent)", and "key indicator risk" was changed to "key process control risk"; a total of 21 risk factors were obtained.

[0076] Table 2 Risk factors

[0077]

[0078]

[0079] Twenty-one risk factors were retained.

[0080] The third round of re-examination Delphi method. The third round of re-examination survey, the third round of questionnaires were issued to the expert group, including the name of the risk factor, the expert's second round of quality risk level score (median, upper quartile, lower quartile). Experts whose scores in the second round were outside the upper and lower quartiles were asked to judge whether to modify the score. If they insisted on not modifying the score, they should explain the reason. Among them, the manufacturing process level risk, key process control risk, and raw material quality risk scores became higher in the third round.

[0081] The results of the three rounds of surveys were collated, and the risk level scores were summarized as medians to obtain the final survey results, as shown in Table 3. The third round of interpretation is shown in the following table:

[0082] Table 3 Final survey results

[0083]

[0084]

[0085] As can be seen from the table, the expert group retained 21 quality risk factors.

[0086] Furthermore, an indicator system is established. According to the cycle, it is divided into four stages: production, design selection, testing, installation and after-sales, and finally the risk factors are formed.

[0087] Table 4 Risk factors

[0088]

[0089]

[0090] refer to Figure 2 , which is a cause-effect network diagram of components in an embodiment of the present application.

[0091] refer to Figure 3 , is the code version component cause-and-effect network diagram of the embodiment of the present application.

[0092] like Figure 2 and Figure 3 As shown in the figure, based on the causal relationship among the risk factors in the “indicator system”, a Bayesian causal network diagram was constructed through expert interview method.

[0093] The causal relationship questionnaire based on the expert survey method is an important method used to construct the Bayesian network in this application. The questionnaire design aims to explore the causal relationship between variables through the subjective judgment and empirical knowledge of experts. As a qualitative research method, the expert survey method can help obtain a deep understanding and insight into the causal relationship, and provide theoretical guidance and research hypothesis construction for subsequent quantitative analysis.

[0094] The design of the questionnaire includes the participation of experts with professional background and rich experience in the research field. Through systematic interviews or surveys of experts, their views and opinions on the possible causal relationship between variables are collected. In the questionnaire design, this application adopts a combination of open-ended questions or semi-structured questions so that experts can give free play and provide detailed and rich answers.

[0095] By analyzing the questionnaire data obtained by the expert survey method, this application can obtain qualitative and quantitative descriptions of causal relationships, including potential causal chains, key variables, and causal probability estimates between variables. It provides valuable information and insights for a deeper understanding of the causal relationship between variables, and plays an important supporting role in the formation of the overall framework and conclusions of this application.

[0096] The cause-effect relationship diagram is established by the expert group in the preliminary quality survey based on past experience and discussed by the expert group. In order to overcome the subjective factors of expert scoring and ensure the validity of the data source, the expert scoring is composed of 5 reviewers. During the quality audit process, the quality of each risk factor will be assessed and corresponding questions will be raised. According to the degree of satisfaction of the quality with the questions, each reviewer will give a separate score based on past experience, and finally the mode of the scores of the 5 experts will be taken. Based on the opinions of the 5 experts, the probability value of the causal relationship is proposed and marked on the cause-effect relationship diagram, such as Figure 2 shown.

[0097] Figure 2 and Figure 3 The corresponding relationships are: raw material quality A2, manufacturing process level A3, key process control A4, component design level A1, usage design level A5, assembly adaptability A6, environmental adaptability A7, component inherent reliability B1, component usage reliability B2, DPA conclusion C2, re-inspection and screening conclusion C1, quality problem occurrence stage C9, key / important parts usage C3, application environment C4, work failure rate C5, device-level verification conclusion C6, board-level verification conclusion C7, single-machine-level verification conclusion C8, quality problem impact D1, quality problem quantity D2, quality zero E1, quality risk F1.

[0098] After obtaining the edge weight value estimated by the expert, further refinement training of the edge weight is required to ensure the accuracy of the edge weight. In the embodiment of the present application, the hierarchical multinomial Dirichlet model based on maximum likelihood estimation is used to complete parameter estimation.

[0099] In this embodiment, nodes (random variables) and directed edges (conditional dependencies) are first determined. Data is then collected, which will be used to train and adjust edge weights. Parameter estimation is then performed using the collected data to estimate the edge weight matrix Ω in the network, i.e., the parameters of the conditional probability distribution.

[0100] For example, the edge weight matrix of a directed graph is:

[0101]

[0102] The above matrix represents a Bayesian network graph where parameter 1 points to parameter 2 and the edge weight is 0.2.

[0103] Specifically, in the learning parameter part of the Bayesian network, the present application uses a hierarchical model to estimate the Bayesian network in a set of discrete random variables X = {X1,...,X i}. Assume that the Bayesian network structure is known and apply the hierarchical model to each single node X i ∈X. For i∈1,...,I ​​and Y, this application selects X=X i As a discrete variable, X i The joint state of the parent set Pai. For y∈Y, the parameter vector θ X|y Corresponding to the variable X i Columns of the Conditional Probability Table (CPT) associated with Pai.

[0104] In this application, an innovative method for Bayesian network parameter estimation is introduced by relaxing the local independence assumption when the parameters are estimated by hierarchical models or traditional models. Combining hierarchical structure learning with hierarchical parameter learning will further improve the performance of Bayesian network classifiers.

[0105] For each node i, according to the structure of the Bayesian network and the data set D, the conditional probability distribution parameter θ of the node is calculated i In a Bayesian network, the conditional probability table of each node is one of the parameters that need to be estimated. MLE is a parameter estimation method that aims to estimate parameters by maximizing the likelihood function of a given data set. For the CPT of each node, MLE calculates the most likely parameter value based on the observed data, maximizing the likelihood of the observed data under the parameter setting.

[0106] Further, after constructing the Bayesian network, determine the edge weight matrix in the Bayesian network, as described in the above steps, and determine the edge weight matrix based on Figure 2 The edge weight matrix represented by is then used to calculate the second statistical value based on the edge weight matrix and the sample parameters.

[0107] In some embodiments, the second statistical value is calculated by the following formula:

[0108]

[0109] in, represents the second statistical value of the i-th batch of data to be monitored, x ij represents the value of the jth quality risk factor of the i-th data batch to be monitored, μ j represents the mean value of the data to be monitored for the jth quality risk factor, Ω -1 Represents the inverse of the edge weight matrix.

[0110] In this embodiment, the T of each training set sample is calculated. 2 value.

[0111] Dataset:

[0112]

[0113] Directed graph edge weight matrix:

[0114]

[0115] Calculate parameter means:

[0116]

[0117] Calculate T for each sample 2 .

[0118] Component batch 1:

[0119]

[0120] Component batch 2:

[0121]

[0122] After the second statistical value is calculated, a warning threshold is determined.

[0123] All batches are calculated, and then supervised learning is used to determine high, medium and low thresholds based on the ROC curve.

[0124] The ROC curve is a curve that reflects the relationship between sensitivity and specificity. The horizontal axis X-axis is 1-specificity, also known as the false positive rate (false alarm rate). The closer the X-axis is to zero, the higher the accuracy; the vertical axis Y-axis is called sensitivity, also known as the true positive rate (sensitivity). The larger the Y-axis is, the better the accuracy. According to the position of the curve, the entire graph is divided into two parts. The area under the curve is called AUC, which is used to indicate the prediction accuracy. The higher the AUC value, that is, the larger the area under the curve, the higher the prediction accuracy. The closer the curve is to the upper left corner (the smaller the X, the larger the Y), the higher the prediction accuracy.

[0125] In this example, the training set has a total of 1050 batches, with each T 2 The value is used as the threshold to calculate the recall rate and precision rate, and the T with the highest recall rate and precision rate is found. 2 The value is used as the final warning threshold.

[0126] refer to Figure 4 , which is a high-risk ROC curve diagram of an embodiment of the present application.

[0127] refer to Figure 5 , which is a low- to medium-risk ROC curve diagram of an embodiment of the present application.

[0128] like Figure 4 As shown in Figure 1, the high-risk threshold is selected as 113.631, at which point the recall rate is 97.1% and the overall accuracy is 99.0%. Figure 5 As shown in Table 5, the medium-low risk threshold is selected as 22.293, at which time the recall rate is 90.9% and the overall accuracy is 97.0%. Therefore, the high, medium and low risk classification is shown in Table 5.

[0129] Table 5 Online traceability assessment quality risk level

[0130] Risk Level Risk Range High risk 113.631≤x Medium risk 22.293≤x<113.631 Low risk 0≤x<22.293

[0131] In some embodiments, after the warning threshold is obtained, the warning threshold calculated above can also be tested using a test set.

[0132] The test results are shown in Table 6:

[0133] Table 6 Confusion matrix

[0134]

[0135] The macro recall is calculated as:

[0136]

[0137] Therefore, the model passed 80% of the performance indicators, indicating that it performed well on the test set.

[0138] The TP sample number is counted as a “true positive” when the test set prediction result Y'=yes and the calibration result Y=yes. The FN sample number is counted as a “false negative” when the test set prediction result Y'=no and the calibration result Y=yes.

[0139] In the review of the domestic quality risk early warning process of component quality, four types of early warning (green, yellow, red, black) are considered, and after expert discussion and meetings with middle and senior quality management, it is worth noting that this scoring standard is a set of evaluation methods based on the current established expert experience. When the industry is different or changes over time, this range will be adjusted accordingly.

[0140] This application sets the warning completeness rate (recall rate):

[0141]

[0142] After the early warning threshold is calculated, for step S101 and step S102, a first statistical value corresponding to each sample batch of the acquired data to be monitored is calculated, and a second statistical value corresponding to each quality risk factor of the acquired data to be monitored is calculated.

[0143] In some embodiments, the first statistical value is calculated by the following formula:

[0144]

[0145] in, represents the first statistical value of the i-th batch of data samples to be monitored, x ij represents the value of the jth quality risk factor of the i-th data sample batch to be monitored, μ j represents the mean value of the data to be monitored for the jth quality risk factor, Ω -1 Represents the inverse of the edge weight matrix.

[0146] In some embodiments, the second statistical value is calculated by the following formula:

[0147]

[0148] in, represents the second statistical value of the jth quality risk factor of the i-th data batch to be monitored, x ij represents the value of the jth quality risk factor of the i-th data batch to be monitored, μ j Represents the mean of the data to be monitored for the jth quality risk factor.

[0149] In some embodiments:

[0150]

[0151] Among them, Σ represents the covariance of x and y, Σ ij represents the covariance of the jth quality risk factor of the i-th batch of data to be monitored, x ij represents the value of the jth quality risk factor of the i-th data batch to be monitored, μ j represents the mean value of the data to be monitored for the jth quality risk factor, y i represents the label value of the i-th batch of data to be monitored, represents the mean of the label values, and m represents the total number of data batches to be monitored.

[0152] In this embodiment, x is calculated by 11 node For example, the data set is as follows:

[0153]

[0154] First calculate the mean:

[0155]

[0156] Next, calculate the covariance:

[0157]

[0158] Final calculation

[0159]

[0160] After calculating the first statistical value and the second statistical value, for step S103, a quality warning is performed based on the warning threshold value and the first statistical value calculated above. The specific implementation process is to compare the first statistical value with multiple warning thresholds to see in which interval the first statistical value is located, that is, to obtain the corresponding warning level, and then perform a quality warning based on the warning level.

[0161] Further, with respect to step S104, a risk probability value corresponding to each of the quality risk factors is calculated based on the second statistical value.

[0162] In some embodiments, the risk probability value corresponding to each of the quality risk factors is calculated based on the second statistical value, including: integrating the chi-square distribution density function in the range of zero to the second statistical value to obtain a first probability value; and taking the difference between 1 and the first probability value as the risk probability value.

[0163] In this embodiment, the risk probability P of each node is calculated. ’ Value, used to evaluate the degree of association between nodes or the probability of an event. Set the risk probability value P ’=1-p. p probability passes through [0,T 2 The chi-square distribution density function is obtained by integrating the chi-square distribution density function within the range of ]. The chi-square distribution density function is as follows:

[0164]

[0165] Among them, n=1, because the parameter dimension is 1.

[0166] Furthermore, after the risk probability value is calculated, for steps S105 and S106, the quality problem of the data to be monitored is traced based on the quality risk factor corresponding to the maximum risk probability value; based on the quality warning and the tracing of the quality problem, the quality problem of the data to be monitored is monitored.

[0167] In some embodiments, the tracing of the quality problems of the data to be monitored based on the quality risk factors corresponding to the maximum risk probability values ​​includes: determining all the quality risk factors corresponding to all the maximum risk probability values; determining the risk chain corresponding to the quality problems of the data to be monitored based on all the quality risk factors and the Bayesian network; and tracing the quality problems of the data to be monitored based on the risk chain.

[0168] In this embodiment, 146 high-risk components with warnings are traced for analysis. For example, when the risk threshold is set to 0.9, if P′ A1 =1, P′ B1 =1, P′ C1 =1, P′ C2 =1, P′ C3 =1, P′ c4 =1, p′ c5 =1, p′ D1 =1, P′ D2 =1, P′ E1 =1, P′ F1 =1, a significant risk classification result can be obtained, which shows that under the set high threshold, this application can more effectively identify high-risk nodes and conduct traceability analysis on them, thereby improving the overall warning accuracy and risk assessment effect.

[0169] like Figure 6 , which is a statistical diagram of the Bayesian network test set results of an embodiment of the present application.

[0170] like Figure 7 The figure shows a schematic diagram of the traceability process of an embodiment of the present application.

[0171] Depend on Figure 6 and Figure 7As shown, the early warning chain is A1-B1-C3-D1-E1-F1, and the source node is the component design level, which can eliminate abnormal situations layer by layer.

[0172] Based on the aforementioned early warning process and traceability process, the monitoring of component quality issues can eventually be achieved.

[0173] It can be seen from the above embodiments that the component quality problem monitoring method and related equipment described in the embodiments of the present application, by calculating the first statistical value corresponding to each sample batch of the data to be monitored; calculating the second statistical value corresponding to each quality risk factor of the data to be monitored; using the pre-calculated warning threshold and the first statistical value to perform quality warning; calculating the risk probability value corresponding to each quality risk factor based on the second statistical value; tracing the quality problem of the data to be monitored based on the quality risk factor corresponding to the maximum risk probability value; monitoring the quality problem of the data to be monitored based on the quality warning and the tracing of the quality problem. The embodiments of the present application construct a component quality risk assessment system covering important stages such as component production, design selection, testing, installation and after-sales through correlation analysis of quality data of different batches of components throughout their life cycle, and derive the target warning level through the correlation between nodes fitted by the Bayesian probability model, so as to achieve a level warning of component quality risk. Through the risk probability value of each node in the Bayesian network, accurate backtracing of quality risk nodes can be achieved. The above Bayesian network can represent the uncertainty and correlation of random quality risks, achieve accurate prediction and early warning of component quality risk levels, achieve accurate tracing and timely feedback of product quality problems, improve the accuracy of problem location and resolution, and can significantly improve the quality control level of aerospace products.

[0174] It should be noted that the method of the embodiment of the present application can be performed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present application, and the multiple devices will interact with each other to complete the described method.

[0175] It should be noted that the above describes some embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0176] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a device for monitoring component quality problems.

[0177] refer to Figure 8 , the component quality problem monitoring device comprises:

[0178] A first statistical value calculation module 81 is configured to calculate a first statistical value corresponding to each sample batch of the acquired data to be monitored;

[0179] A second statistical value calculation module 82 is configured to calculate a second statistical value corresponding to each quality risk factor of the acquired data to be monitored;

[0180] An early warning module 83 is configured to perform a quality early warning using a pre-calculated early warning threshold and the first statistical value;

[0181] A calculation module 84 is configured to calculate a risk probability value corresponding to each of the quality risk factors based on the second statistical value;

[0182] The tracing module 85 is configured to trace the quality problem of the data to be monitored based on the quality risk factor corresponding to the maximum risk probability value;

[0183] The monitoring module 86 is configured to monitor the quality problem of the data to be monitored based on the quality warning and the tracing of the quality problem.

[0184] For the convenience of description, the above device is described in terms of functions divided into various modules. Of course, when implementing the present application, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0185] The device of the above embodiment is used to implement the corresponding component quality problem monitoring method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.

[0186] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments, the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the component quality problem monitoring method described in any of the above embodiments is implemented.

[0187] Fig. 9A more specific schematic diagram of the hardware structure of an electronic device provided in this embodiment is shown, and the device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other through the bus 1050 in the device.

[0188] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0189] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0190] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0191] The communication interface 1040 is used to connect a communication module (not shown) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB, network cable, etc.) or a wireless mode (such as mobile network, WIFI, Bluetooth, etc.).

[0192] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0193] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.

[0194] The electronic device of the above embodiment is used to implement the corresponding component quality problem monitoring method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.

[0195] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the component quality problem monitoring method described in any of the above embodiments.

[0196] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0197] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute the component quality problem monitoring method described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0198] Based on the same inventive concept, corresponding to the component quality problem monitoring method described in any of the above embodiments, the present disclosure also provides a computer program product, which includes computer program instructions. In some embodiments, the computer program instructions can be executed by one or more processors of a computer so that the computer and / or the processor execute the component quality problem monitoring method. Corresponding to the execution subject corresponding to each step in each embodiment of the component quality problem monitoring method, the processor that executes the corresponding step can belong to the corresponding execution subject.

[0199] The computer program product of the above embodiment is used to enable the computer and / or the processor to execute the component quality problem monitoring method described in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0200] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present application (including the claims) is limited to these examples. In line with the concept of the present application, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present application as described above, which are not provided in detail for the sake of simplicity.

[0201] In addition, to simplify the description and discussion, and in order not to make the embodiments of the present application difficult to understand, the known power supply / ground connection with the integrated circuit (IC) chip and other components may or may not be shown in the provided drawings. In addition, the device can be shown in the form of a block diagram to avoid making the embodiments of the present application difficult to understand, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform to be implemented in the embodiments of the present application (that is, these details should be fully within the scope of understanding of those skilled in the art). In the case of elaborating specific details (e.g., circuits) to describe exemplary embodiments of the present application, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details or when these specific details are changed. Therefore, these descriptions should be considered to be illustrative rather than restrictive.

[0202] Although the present application has been described in conjunction with specific embodiments of the present application, many replacements, modifications and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed.

[0203] The embodiments of the present application are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present application should be included in the scope of protection of the present application.

Claims

1. A method for monitoring component quality problems, characterized in that: include: Calculate a first statistical value corresponding to each sample batch of the acquired data to be monitored; Calculate a second statistical value corresponding to each quality risk factor of the acquired data to be monitored; Using the pre-calculated and determined warning threshold and the first statistical value to issue a quality warning; Calculate the risk probability value corresponding to each of the quality risk factors based on the second statistical value; Tracing the quality problem of the data to be monitored based on the quality risk factor corresponding to the maximum risk probability value; Based on the quality warning and the tracing of the quality problem, the quality problem of the data to be monitored is monitored.

2. The method according to claim 1, characterized in that The warning threshold is determined by the following steps: Screen the quality risk factors to be screened and form an indicator system; Constructing a Bayesian network based on the causal relationship between each quality risk factor in the indicator system, and determining an edge weight matrix in the Bayesian network; Based on the edge weight matrix and the sample parameters, a first statistical value is calculated; The warning threshold is determined according to the first statistical value.

3. The method according to claim 2, characterized in that The tracing of the quality problem of the data to be monitored based on the quality risk factor corresponding to the maximum risk probability value includes: Determine all the quality risk factors corresponding to all the maximum risk probability values; Determine the risk chain corresponding to the quality problem of the data to be monitored based on all the quality risk factors and the Bayesian network; The quality problems of the data to be monitored are traced based on the risk chain.

4. The method according to claim 1, characterized in that The second statistical value is calculated by the following formula: in, represents the second statistical value of the jth quality risk factor of the i-th batch of data samples to be monitored, x ij represents the value of the jth quality risk factor of the i-th data sample batch to be monitored, μ j represents the mean value of the data to be monitored for the jth quality risk factor, Σ ij Represents the covariance of the jth quality risk factor of the i-th batch of data samples to be monitored.

5. The method according to claim 1, characterized in that The calculating the risk probability value corresponding to each of the quality risk factors based on the second statistical value includes: Integrating the chi-square distribution density function within the range from zero to the second statistical value to obtain a first probability value; The difference between 1 and the first probability value is used as the risk probability value.

6. The method according to claim 1, characterized in that The first statistical value is calculated by the following formula: in, represents the first statistical value of the i-th batch of data samples to be monitored, x ij represents the value of the jth quality risk factor of the i-th data sample batch to be monitored, μ j represents the mean value of the data to be monitored for the jth quality risk factor, Ω -1 Represents the inverse of the edge weight matrix.

7. A device for monitoring component quality problems, characterized in that: include: A first statistical value calculation module is configured to calculate a first statistical value corresponding to each sample batch of the acquired data to be monitored; A second statistical value calculation module is configured to calculate a second statistical value corresponding to each quality risk factor of the acquired data to be monitored; An early warning module is configured to use a pre-calculated early warning threshold and the first statistical value to perform a quality early warning; a calculation module, configured to calculate a risk probability value corresponding to each of the quality risk factors based on the second statistical value; A tracing module is configured to trace the quality problem of the data to be monitored based on the quality risk factor corresponding to the maximum risk probability value; The monitoring module is configured to monitor the quality problem of the data to be monitored based on the quality warning and the tracing of the quality problem.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 6.

10. A computer program product, comprising computer program instructions, which, when executed on a computer, cause the computer to execute the method according to any one of claims 1 to 7.

Images