Electronic component fault detection and analysis system and method

Abstract

The invention discloses an electronic component fault detection and analysis system and method, and relates to the technical field of electronic component fault detection.The method comprises the steps that detection indexes in an electronic component and the key degree of fault detection of the electronic component are evaluated, and key index data are obtained; obtaining product fault information of the electronic component, generating a mark key group of the electronic component in combination with the key index data, and analyzing the influence key degree of different mark key groups on the fault of the electronic component in the current period; feature key groups are obtained from the fault key data, and the data influence degree between different feature key groups in the electronic component in the current period is analyzed in combination with feature historical fault detection records of the electronic component; and according to the influence degree data and the fault key data, fault causes in the production process of the electronic components are analyzed, and processing and production of the electronic components are optimized and adjusted.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic component fault detection, and specifically to an electronic component fault detection and analysis system and method. Background Technique

[0002] Electronic components refer to the basic components used to achieve specific functions in electronic circuits. They are the basis for the production of electronic products and are widely used in multiple fields such as communication, computers, and medical treatment. During the production process of electronic components, timely fault detection of electronic components can analyze and eliminate electronic components with unqualified performance, ensuring that only electronic components that meet quality standards can be put on the market, reducing the outflow rate of defective products. In addition, qualified electronic components can also greatly reduce the repair rate of products.

[0003] In the fault analysis of electronic components, traditional methods mainly rely on the experience and intuition of engineers. The interpretation of faults is likely to be affected by subjective judgments, and engineers also need to have strong work experience and work ability. In addition, with the increase in production scale and the complexity of electronic components, the amount of data required for fault analysis is also constantly increasing. Traditional fault analysis methods are difficult to effectively handle the fault problems of electronic components, not only consuming a large amount of manpower and material resources, but also causing the faults of electronic components to be unable to be corrected in time, having a huge negative impact on the production of electronic components. Summary of the Invention

[0004] The purpose of the present invention is to provide an electronic component fault detection and analysis system and method to solve the problems raised in the prior art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An electronic component fault detection and analysis method, the method includes:

[0006] Step S100: Obtain the historical fault detection records of electronic components, evaluate the detection indicators in the electronic components, and obtain the key index data for the critical degree of the fault detection of the electronic components;

[0007] Step S200: Obtain the product fault information of the electronic components, and combine the key index data to generate the marked key group of the electronic components. Obtain the characteristic historical fault detection records of the electronic components, and analyze the critical degree of the influence of different marked key groups on the faults of the electronic components in the current period to obtain the fault key data;

[0008] Step S300: Obtain the characteristic key group from the fault key data, and combine the characteristic historical fault detection records of the electronic components to analyze the data influence degree between different characteristic key groups in the electronic components in the current period to obtain the influence degree data;

[0009] Step S400: Analyze the fault causes in the production process of electronic components based on the impact degree data and the fault key data, and optimize and adjust the processing and production of electronic components.

[0010] Further, step S100 includes:

[0011] Step S101: Obtain the historical fault detection records of electronic components, where the electronic components in the historical fault detection records are detected as faulty;

[0012] Obtain the models of the electronic components, obtain the historical fault detection records of each electronic component with the obtained model, and obtain the values of each detection index in the electronic components from the historical fault detection records;

[0013] Step S102: Calculate the characteristic values of each detection index in the historical fault detection records. Among them, the characteristic value C of the a-th detection index in the historical fault detection records a =(B a -μ a ) / σ a , where B a is the value of the a-th detection index in the historical fault detection records, μ a is the mean value of the a-th detection index in the historical fault detection records of each electronic component, and σ a is the standard deviation of the a-th detection index;

[0014] Step S103: Obtain the characteristic values of each detection index in the historical fault detection records of each electronic component, and construct the data matrix D of the electronic components of the model in the current cycle;

[0015] Calculate the covariance matrix F of the data matrix = 1 / (n - 1)·D T ·D, perform eigenvalue decomposition on the covariance matrix F, and obtain the eigenvalues λ and eigenvectors V of the main components in the covariance matrix. The specific formula is: F·V = λ·V, where V = {v1, v2,..., v j}, v1, v2,..., v j respectively represent the eigenvectors of the 1st, 2nd,..., j-th main components, and λ = {λ1, λ2,..., λ j}, λ1, λ2,..., λ j respectively represent the eigenvalues of the 1st, 2nd,..., j-th main components;

[0016] Step S104: Sort according to the magnitudes of the eigenvalues of each main component, and select the main components corresponding to the eigenvectors of the first k eigenvalues for retention, where k is a preset value;

[0017] Based on the eigenvectors corresponding to the first k eigenvalues, a feature projection matrix Q for the electronic components of the model is formed, and the data matrix is projected into the principal component space to obtain matrix Z = D·Q;

[0018] Obtain several principal components in matrix z, calculate the detection indicators of the electronic components of the model, and calculate the importance scores of the detection indicators in several principal components in matrix z. Among them, the importance score Y of the detection indicator in the h-th principal component in matrix z h = w h 2 ×λ′ h where w h is the weight of the detection indicator in the h-th principal component, and λ′ h is the eigenvalue of the detection indicator in the h-th principal component;

[0019] Accumulate the importance scores of the detection indicators of the electronic components of the model in several principal components in matrix z to obtain the key score of the detection indicator;

[0020] Obtain the key scores of each detection indicator of the electronic components of the model, and obtain the p detection indicators with the largest key scores to obtain the key detection indicators of the electronic components of the model. Among them, p is a preset value, and it is determined that the key detection indicators play a key role in the fault detection of the electronic components of the model;

[0021] Collect the key detection indicators of the electronic components of the model to obtain the key indicator data.

[0022] Further, step S200 includes:

[0023] Step S201: Obtain the key indicator data of the electronic components of the model, obtain the product fault information of the electronic components of the model, and obtain the preset various fault causes of the electronic components of the model;

[0024] Based on the product fault information, obtain the key detection indicators corresponding to each fault cause, and collect them to obtain the marked key group of the fault cause. Obtain the marked key group corresponding to each fault cause, so as to obtain each marked key group of the electronic components of the model;

[0025] Step S202: Obtain the historical fault detection records of the electronic components of the model in the current period, and record them as the characteristic historical fault detection records, and obtain the preset ranges of each key detection indicator of the electronic components of the model;

[0026] When several key detection indicators in a certain marked key group are all outside the corresponding ranges, record the characteristic historical fault detection records as the abnormal characteristic historical fault detection records of a certain marked key group;

[0027] Step S203: Analyze each marked key group for the degree of critical impact on the failure of the electronic components of the model in the current cycle. Among them, for the α-th marked key group in each marked key group, the specific analysis process for the degree of critical impact on the failure of the electronic components of the model in the current cycle is as follows:

[0028] Calculate the failure impact score γ of the α-th marked key group α =R α / R sum , where R α is the total number of historical failure detection records of abnormal characteristics of the α-th marked key group, and R sum is the total number of historical failure detection records of each characteristic of the electronic components of the model;

[0029] When the failure impact score γ α is greater than the preset working impact score threshold, it is determined that the α-th marked key group has an impact on the failure of the electronic components of the model in the current cycle, and the α-th marked key group is recorded as the characteristic key group;

[0030] Step S204: Obtain each characteristic key group of the electronic components of the model in the current cycle and gather them to obtain the failure critical data.

[0031] Furthermore, step S300 includes:

[0032] Step S301: Obtain each characteristic key group of the electronic components of the model from the failure critical data, and obtain each historical failure detection record of the characteristics of the electronic components of the model;

[0033] Analyze the degree of data impact among each characteristic key group in the electronic components of the model in the current cycle. Among them, the specific process for analyzing the degree of data impact of the s-th characteristic key group on the x-th characteristic key group in the current cycle is as follows:

[0034] Obtain the preset value m, where m > 1, and obtain the marked values of the s-th characteristic key group and the x-th characteristic key group in each historical failure detection record;

[0035] When a certain historical failure detection record is the abnormal historical failure detection record of the s-th characteristic key group, the marked value P of the s-th characteristic key group in a certain historical failure detection record s is m, otherwise, it is P s is 0;

[0036] Step S302: Obtain each abnormal historical failure detection record of the x-th characteristic key group, and calculate the characteristic impact coefficient L of the s-th characteristic key group on the x-th characteristic key groups,x = (β × m) / (∑ t＝β t＝1 P x s,t + 1), where β is the total number of historical fault detection records of each abnormal feature, and P x s,t is the marked value of the s-th feature key group in the t-th historical fault detection record of each abnormal feature;

[0037] Step S303: Accumulate the absolute values of the differences between the marked values of the s-th feature key group and the x-th feature key group in each feature historical fault detection record to obtain the synchronous change value G s,v ;

[0038] Calculate the data influence value U s,x = [1 / (G s,v + 1)] × L s,x , when the data influence value U s,x is greater than the preset data influence threshold, it is determined that the s-th feature key group in the current cycle has an impact on the data of the x-th feature key group, and the x-th feature key group is recorded as the influencing feature key group of the s-th feature key group;

[0039] Step S304: Obtain and collect the influencing feature key groups of several feature key groups of the electronic components of the model to obtain the influence degree data.

[0040] Furthermore, step S400 includes:

[0041] Step S401: Obtain the influence degree data of the electronic components of the model, and analyze the fault causes in the production process of the electronic components from the influencing feature key groups corresponding to the feature key groups in the influence degree data. The specific process is as follows:

[0042] When the ratio between the total number of influencing feature key groups caused by a certain feature key group and the feature key groups in the fault key data is greater than the preset ratio threshold, a certain feature key group is recorded as the target feature key group;

[0043] Obtain the fault cause corresponding to the target feature key group and record it as the target fault cause of the electronic components of the model in the current cycle;

[0044] Step S402: Obtain several target fault causes of the electronic components of the model in the current cycle, and check the production and processing line of the electronic components according to the several target fault causes, and optimize and adjust the production and processing line of the electronic components of the model;

[0045] Through the obtained target feature key group in the above steps, the cause of the failure of the electronic component can be quickly found. Relevant staff only need to detect the cause of the failure, and then the failure problem of the electronic component in the current cycle can be quickly solved, which not only improves the qualified rate of the electronic component, but also optimizes the production of the electronic component.

[0046] In order to better implement the above method, an electronic component failure detection and analysis system is also proposed. The system includes a key index data module, a failure impact module, an impact degree data module, and a failure analysis module;

[0047] The key index data module is used to evaluate the key degree of the detection indexes in the electronic component for the failure detection of the electronic component, and obtain the key index data;

[0048] The failure impact module is used to analyze the key degree of the marked key group on the failure of the electronic component in the current cycle, and obtain the failure key data;

[0049] The impact degree data module is used to analyze the data impact degree between different feature key groups in the electronic component in the current cycle according to the failure key data, and obtain the impact degree data;

[0050] The failure analysis module is used to analyze the cause of the failure in the production process of the electronic component, and optimize and adjust the processing and production of the electronic component.

[0051] Furthermore, the key index data module includes a key score unit and a key index data unit;

[0052] The key score unit is used to calculate the key scores of the various detection indexes of the electronic component;

[0053] The key index data unit is used to determine the key degree of the detection index for the failure detection of the electronic component of the model according to the key score, and obtain the key index data.

[0054] Furthermore, the failure impact module includes a failure impact scoring unit and a failure impact unit;

[0055] The failure impact scoring unit is used to generate the marked key group of the electronic component according to the key index data, and calculate the failure impact score of each marked key group of the electronic component;

[0056] The failure impact unit is used to analyze the key degree of the marked key group on the failure of the electronic component in the current cycle according to the failure impact score, and obtain the failure key data.

[0057] Further, the impact degree data module includes a data impact value unit and an impact degree data unit;

[0058] The data impact value unit is used to obtain the key fault data, acquire the characteristic key groups from the key fault data, and calculate the data impact values between the respective characteristic key groups;

[0059] The impact degree data unit is used to analyze the data impact degree of each characteristic key group of the electronic components in the current cycle according to the data impact value, and obtain the impact degree data.

[0060] Further, the fault analysis module includes a fault analysis unit;

[0061] The fault analysis unit is used to obtain the impact degree data and the key fault data of the electronic components, analyze the fault causes in the production process of the electronic components, and optimize and adjust the processing and production of the electronic components.

[0062] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention realizes the intelligent fault detection and fault analysis of electronic components in the production and processing process. By analyzing the historical fault detection records of electronic components, different detection indexes of electronic components are analyzed, and the key degree of each detection index in the fault detection process of electronic components is obtained, and the key detection indexes of electronic components are obtained. At the same time, through the product fault information of electronic components, the combination of key detection indexes corresponding to each fault cause is obtained, and according to the data impact degree between the characteristic key groups, the target characteristic key group is found, the fault causes in the processing and production of electronic components are analyzed, and the production line of electronic components is adjusted and optimized. It not only quickly finds the fault problems of electronic components in the production process, but also optimizes the production and processing, saves a large amount of human and material resources costs, and further improves the qualified rate of products. Description of the Drawings

[0063] Figure 1 is the method flow chart of a method for detecting and analyzing faults of electronic components according to the present invention;

[0064] Figure 2 is the module schematic diagram of a system for detecting and analyzing faults of electronic components according to the present invention. Detailed Embodiments

[0065] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0066] Example: As Figure 1 - Figure 2 shown, the present invention provides a technical solution, a method for detecting and analyzing faults of electronic components, the method comprising:

[0067] Step S100: Obtain the historical fault detection records of the electronic components, evaluate the detection indicators in the electronic components, and obtain the key index data for the criticality of the fault detection of the electronic components;

[0068] Among them, step S100 includes:

[0069] Step S101: Obtain the historical fault detection records of the electronic components, wherein the electronic components in the historical fault detection records are detected as having faults;

[0070] Obtain the model numbers of the electronic components, obtain the historical fault detection records of each electronic component with the obtained model number, and obtain the values of each detection indicator in the electronic components from the historical fault detection records;

[0071] For example, for example, each detection indicator includes current, voltage, impedance, frequency response, temperature, etc.;

[0072] Step S102: Calculate the characteristic values of each detection indicator in the historical fault detection records, wherein the characteristic value C of the a-th detection indicator in the historical fault detection records a =(B a -μ a ) / σ a , where B a is the value of the a-th detection indicator in the historical fault detection records, μ a is the mean value of the a-th detection indicator in the historical fault detection records of each electronic component, and σ a is the standard deviation of the a-th detection indicator;

[0073] Step S103: Obtain the characteristic values of each detection indicator in the historical fault detection records of each electronic component, and construct the data matrix D of the electronic components of the model number in the current period;

[0074] Calculate the covariance matrix F of the data matrix = 1 / (n - 1)·D T ·D, perform eigenvalue decomposition on the covariance matrix F, and obtain the eigenvalues λ and eigenvectors V of the main components in the covariance matrix. The specific formula is: F·V = λ·V, where V = {v1, v2,..., v j}, v1, v2,..., v j respectively represent the eigenvectors of the 1st, 2nd,..., j-th main components, λ = {λ1, λ2,..., λ j}, λ1, λ2,..., λ jare respectively represented as the eigenvalues of the 1st, 2nd, ..., jth principal components;

[0075] Step S104: Sort according to the magnitudes of the eigenvalues of each principal component, and select the principal components corresponding to the eigenvectors of the top k eigenvalues for retention, where k is a preset value;

[0076] Based on the eigenvectors of the top k eigenvalues, construct the characteristic projection matrix Q corresponding to the electronic components of the model, project the data matrix into the principal component space, and obtain the matrix Z = D·Q;

[0077] Obtain several principal components in the matrix z, and calculate the importance scores of the detection indexes of the electronic components of the model in several principal components in the matrix z. Among them, the importance score Y of the detection index in the hth principal component in the matrix z h = w h 2 ×λ′ h where w h is the weight of the detection index in the hth principal component, and λ′ h is the eigenvalue of the detection index in the hth principal component;

[0078] For example, the weight w1 of the detection index in the 1st principal component is 1.2, and the eigenvalue λ′1 of the detection index in the 1st principal component is 0.8;

[0079] Calculate the importance score Y1 of the above detection index in the 1st principal component in the matrix z = 1.2 2 ×0.8 = 1.152;

[0080] Accumulate the importance scores of the detection indexes of the electronic components of the model in several principal components in the matrix z to obtain the key score of the detection index;

[0081] Obtain the key scores of each detection index of the electronic components of the model, and obtain the top p detection indexes with the largest key scores to obtain the key detection indexes of the electronic components of the model, where p is a preset value, and it is determined that the key detection indexes play a key role in the fault detection of the electronic components of the model;

[0082] Collect each key detection index of the electronic components of the model to obtain key index data;

[0083] For example, each key detection index includes, for example, each detection index includes current, voltage, impedance, frequency response, temperature, etc.;

[0084] Step S200: Obtain the product failure information of the electronic component, and combine it with the key index data to generate the marked key group of the electronic component. Obtain the characteristic historical failure detection record of the electronic component, analyze the different marked key groups, and obtain the key failure data on the influence degree of the electronic component failure in the current cycle;

[0085] Among them, step S200 includes:

[0086] Step S201: Obtain the key index data of the electronic component of the model, obtain the product failure information of the electronic component of the model, and obtain the preset failure causes of the electronic component of the model;

[0087] Based on the product failure information, obtain the key detection indexes corresponding to each failure cause, and conduct aggregation to obtain the marked key group of the failure cause. Obtain the marked key groups corresponding to each failure cause, so as to obtain each marked key group of the electronic component of the model;

[0088] For example, each failure cause includes capacitor failure, electrolytic capacitor aging, etc.;

[0089] For example, the product failure information is that when a failure cause occurs in the electronic component, there is a corresponding relationship between the numerical changes of the key detection indexes in the electronic component. Among them, when electrolytic capacitor aging occurs in the electronic component, the ESR of the electronic component will increase abnormally, the leakage current will increase, and the temperature stability will also decrease. Then the key detection indexes corresponding to electrolytic capacitor aging are ESR, leakage current, and temperature;

[0090] Step S202: Obtain the historical failure detection record of the electronic component of the model in the current cycle, and record it as the characteristic historical failure detection record. Obtain the preset range of each key detection index of the electronic component of the model;

[0091] When several key detection indexes in a certain marked key group are all outside the corresponding range, record the characteristic historical failure detection record as the abnormal characteristic historical failure detection record of a certain marked key group;

[0092] Step S203: Analyze the key influence degree of each marked key group on the failure of the electronic component of the model in the current cycle. Among them, for the α-th marked key group in each marked key group, the specific analysis process of the key influence degree on the failure of the electronic component of the model in the current cycle is as follows:

[0093] Calculate the failure influence score γ of the α-th marked key group α =R α / R sum where, R αR is the total number of abnormal feature historical fault detection records of the α - marked key group sum R is the total number of historical fault detection records of each feature of the electronic components of the model;

[0094] For example, the total number of abnormal feature historical fault detection records R1 of the first - marked key group is 20, and the total number of historical fault detection records R of each feature of the electronic components of the model sum is 100. Calculate the fault impact score γ1 of the α - marked key group = 20 / 100 = 0.2;

[0095] When the fault impact score γ α is greater than the preset working impact score threshold, it is determined that the α - marked key group has an impact on the faults of the electronic components of the model in the current cycle, and the α - marked key group is recorded as the feature key group;

[0096] Step S204: Obtain and collect each feature key group of the electronic components of the model in the current cycle to obtain the fault - critical data;

[0097] Step S300: Obtain the feature key groups from the fault - critical data, and combine with the historical fault detection records of the features of the electronic components to analyze the data influence degree between different feature key groups in the electronic components in the current cycle to obtain the influence - degree data;

[0098] Among them, step S300 includes:

[0099] Step S301: Obtain each feature key group of the electronic components of the model from the fault - critical data, and obtain the historical fault detection records of each feature of the electronic components of the model;

[0100] Analyze the data influence degree between each feature key group in the electronic components of the model in the current cycle. Among them, analyzing the data influence degree of the s - th feature key group on the x - th feature key group in the current cycle, the specific process is as follows:

[0101] Obtain a preset value m, where m > 1, and obtain the marked values of the s - th feature key group and the x - th feature key group in each historical fault detection record;

[0102] When a certain historical fault detection record is an abnormal feature historical fault detection record of the s - th feature key group, the marked value P of the s - th feature key group in a certain historical fault detection record s is m, otherwise, it is P s is 0;

[0103] Step S302: Obtain the historical fault detection records of each abnormal feature in the x-th feature key group, and calculate the feature influence coefficient L of the s-th feature key group on the x-th feature key group s,x =(β×m) / (∑ t＝β t＝1 P x s,t +1), where β is the total number of historical fault detection records of each abnormal feature, and P x s,t is the marked value of the s-th feature key group in the t-th historical fault detection record of each abnormal feature historical fault detection record;

[0104] Step S303: Accumulate the absolute values of the differences between the marked values of the s-th feature key group and the x-th feature key group in each feature historical fault detection record to obtain the synchronous change value G s,v ;

[0105] Calculate the data influence value U of the s-th feature key group on the x-th feature key group s,x =[1 / (G s,v +1)]×L s,x , when the data influence value U s,x is greater than the preset data influence threshold, it is determined that the data of the s-th feature key group in the current cycle has an impact on the x-th feature key group, and the x-th feature key group is recorded as the influencing feature key group of the s-th feature key group;

[0106] Step S304: Obtain and collect the influencing feature key groups of several feature key groups of the electronic components of the model to obtain the influence degree data;

[0107] Step S400: Analyze the fault causes in the production process of the electronic components according to the influence degree data and the fault key data, and optimize and adjust the processing and production of the electronic components;

[0108] Among them, Step S400 includes:

[0109] Step S401: Obtain the influence degree data of the electronic components of the model, and analyze the fault causes in the production process of the electronic components from the influencing feature key groups corresponding to the feature key groups in the influence degree data. The specific process is as follows:

[0110] When the ratio of the total number of influencing feature key groups caused by a certain feature key group to the feature key groups in the fault key data is greater than the preset ratio threshold, a certain feature key group is recorded as the target feature key group;

[0111] Obtain the fault causes corresponding to the key group of target features, and record them as the target fault causes of the electronic components of the model in the current cycle;

[0112] Step S402: Obtain several target fault causes of the electronic components of the model in the current cycle, and check the production line of the electronic components according to the several target fault causes, and optimize and adjust the production line of the electronic components of the model;

[0113] In order to better implement the above method, an electronic component fault detection and analysis system is also proposed. The system includes a key index data module, a fault impact module, an impact degree data module, and a fault analysis module;

[0114] The key index data module is used to evaluate the key degree of the detection indexes in the electronic components for the fault detection of the electronic components, and obtain the key index data;

[0115] The fault impact module is used to analyze the key degree of the marked key group on the electronic component fault in the current cycle, and obtain the fault key data;

[0116] The impact degree data module is used to analyze the data impact degree between different feature key groups in the electronic components in the current cycle according to the fault key data, and obtain the impact degree data;

[0117] The fault analysis module is used to analyze the fault causes in the production process of the electronic components, and optimize and adjust the processing and production of the electronic components;

[0118] Among them, the key index data module includes a key score unit and a key index data unit;

[0119] The key score unit is used to calculate the key scores of the various detection indexes of the electronic components;

[0120] The key index data unit is used to determine the key degree of the detection index for the fault detection of the electronic components of the model according to the key score, and obtain the key index data;

[0121] Among them, the fault impact module includes a fault impact scoring unit and a fault impact unit;

[0122] The fault impact scoring unit is used to generate the marked key group of the electronic components according to the key index data, and calculate the fault impact scores of the various marked key groups of the electronic components;

[0123] The fault impact unit is used to analyze the key degree of the marked key group on the electronic component fault in the current cycle according to the fault impact score, and obtain the fault key data;

[0124] Among them, the impact degree data module includes a data impact value unit and an impact degree data unit;

[0125] The data impact value unit is used to obtain the fault key data, obtain the characteristic key groups from the fault key data, and calculate the data impact values between the respective characteristic key groups;

[0126] The impact degree data unit is used to analyze the data impact degree of each characteristic key group of the electronic components in the current cycle according to the data impact value, and obtain the impact degree data;

[0127] Among them, the fault analysis module includes a fault analysis unit;

[0128] The fault analysis unit is used to obtain the impact degree data and the fault key data of the electronic components, analyze the fault causes in the production process of the electronic components, and optimize and adjust the processing and production of the electronic components.

[0129] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present invention. Any reference signs in the claims should not be regarded as limiting the claims involved.

Claims

1. A method for detecting and analyzing faults of electronic components, characterized in that The method includes: Step S100: Obtain the historical fault detection records of the electronic components, evaluate the detection indicators in the electronic components, and determine the criticality of the fault detection of the electronic components to obtain critical indicator data; Step S200: Obtain the product fault information of the electronic components, and combine the critical indicator data to generate the marked critical group of the electronic components. Obtain the characteristic historical fault detection records of the electronic components, analyze the criticality of the influence of different marked critical groups on the faults of the electronic components in the current period to obtain fault critical data; Step S300: Obtain the characteristic critical group from the fault critical data, and combine the characteristic historical fault detection records of the electronic components to analyze the data influence degree between different characteristic critical groups in the electronic components in the current period to obtain influence degree data; Step S400: Analyze the fault causes in the production process of the electronic components according to the influence degree data and the fault critical data, and optimize and adjust the processing and production of the electronic components.

2. The method for detecting and analyzing faults of electronic components according to claim 1, wherein The step S100 includes: Step S101: Obtain the historical fault detection records of the electronic components, where the electronic components in the historical fault detection records are detected as faulty; Obtain the model of the electronic components, obtain the historical fault detection records of each electronic component of the model, and obtain the values of each detection indicator in the electronic components from the historical fault detection records; Step S102: Calculate the eigenvalue of each detection index in the historical fault detection record, where the eigenvalue C of the a-th detection index in the historical fault detection record a =(B a -μ a ) / σ a , where B a is the value of the a-th detection index in the historical fault detection record, μ a is the mean value of the a-th detection index in the historical fault detection records of each electronic component, and σ a is the standard deviation of the a-th detection index; Step S103: Obtain the characteristic values of each detection indicator in the historical fault detection records of each electronic component, and construct the data matrix D of the electronic components of the model in the current period; Calculate the covariance matrix F of the data matrix: F = 1 / (n - 1)·D T ·D, perform eigenvalue decomposition on the covariance matrix F to obtain the eigenvalues λ and eigenvectors V of the principal components in the covariance matrix. The specific formula is: F·V = λ·V, where V = {v1, v2,..., v j}, v1, v2,..., v j represent the eigenvectors of the 1st, 2nd,..., jth principal components respectively, and λ = {λ1, λ2,..., λ j}, λ1, λ2,..., λ j represent the eigenvalues of the 1st, 2nd,..., jth principal components respectively; Step S104: Sort according to the magnitude order of the characteristic values of each principal component, and select the principal components corresponding to the eigenvectors of the first k characteristic values to retain, where k is a preset value; Based on the eigenvectors of the first k characteristic values, construct the characteristic projection matrix Q corresponding to the electronic components of the model, project the data matrix into the principal component space, and obtain the matrix Z = D·Q; Obtain several principal components in the matrix z, and calculate the detection index of the electronic components of the model, which is the importance score of several principal components in the matrix z. Among them, the importance score Y of the detection index in the h-th principal component in the matrix z h = w h 2 × λ′ h , where w h is the weight of the detection index in the h-th principal component, and λ′ h is the eigenvalue of the detection index in the h-th principal component; Accumulate the important scores of the several principal components in the matrix z for the detection indicators of the electronic components of the model to obtain the critical scores of the detection indicators; Obtain the critical scores of each detection indicator of the electronic components of the model, and obtain the p detection indicators with the largest critical scores to obtain the critical detection indicators of the electronic components of the model, where p is a preset value, and determine that the critical detection indicators play a key role in the fault detection of the electronic components of the model; Collect the critical detection indicators of the electronic components of the model to obtain critical indicator data.

3. The method for detecting and analyzing faults of electronic components according to claim 2, characterized in that, The step S200 includes: Step S201: Obtain the critical indicator data of the electronic components of the model, obtain the product fault information of the electronic components of the model, and obtain each preset fault cause of the electronic components of the model; Based on the product failure information, obtain the key detection indicators corresponding to each failure cause, and gather them to obtain the marked key group of failure causes. Obtain the marked key group corresponding to each failure cause, so as to obtain each marked key group of the electronic components of this model; Step S202: Obtain the historical failure detection records of the electronic components of this model in the current period, and record them as characteristic historical failure detection records. Obtain the preset ranges of the key detection indicators of the electronic components of this model; When several key detection indicators in a marked key group are all outside the corresponding ranges, record the characteristic historical failure detection records as the abnormal characteristic historical failure detection records of this marked key group; Step S203: Analyze the key influence degrees of each marked key group on the failure of the electronic components of this model in the current period. Among them, for the α-th marked key group in each marked key group, the specific analysis process of its key influence degree on the failure of the electronic components of this model in the current period is as follows: Calculate the failure impact score γ of the α - marked critical group α = R α / R sum , where R α is the total number of abnormal - feature historical fault detection records of the α - marked critical group, and R sum is the total number of historical fault detection records of each feature of the electronic components of the model; When the fault impact score γ α is greater than a preset working impact score threshold, it is determined that the α-marked key group has an impact on the faults of the electronic components of the model in the current cycle, and the α-marked key group is recorded as the characteristic key group; Step S204: Obtain and gather each characteristic key group of the electronic components of this model in the current period to obtain the key failure data.

4. A method for detecting and analyzing faults of electronic components according to claim 3, characterized in that, The step S300 includes: Step S301: Obtain each characteristic key group of the electronic components of this model from the key failure data, and obtain each characteristic historical failure detection record of the electronic components of this model; Analyze the data influence degrees among each characteristic key group in the electronic components of this model in the current period. Among them, the specific process of analyzing the data influence degree of the s-th characteristic key group on the x-th characteristic key group in the current period is as follows: Obtain a preset value m, where m > 1. Obtain the marked values of the s-th characteristic key group and the x-th characteristic key group in each characteristic historical failure detection record; When a certain characteristic historical fault detection record is an abnormal characteristic historical fault detection record of the s-th characteristic key group, the marked value P of the s-th characteristic key group in the certain characteristic historical fault detection record s is m, otherwise, it is P s is 0; Step S302: Obtain the historical fault detection records of each abnormal feature of the x-th key feature group, and calculate the feature influence coefficient L of the s-th key feature group on the x-th key feature group s,x =(β × m) / (∑ t＝β t＝1 P x s,t + 1), where β is the total number of the historical fault detection records of each abnormal feature, and P x s,t is the marked value of the s-th key feature group in the t-th historical fault detection record of each abnormal feature historical fault detection record; Step S303: Accumulate the absolute value of the difference between the marked values of the s-th feature key group and the x-th feature key group in each feature historical fault detection record to obtain the synchronous change value G of the s-th feature key group and the x-th feature key group s,v ; Calculate the data influence value U of the s-th feature key group on the x-th feature key group s,x = [1 / (G s,v + 1)] × L s,x , when the data influence value U s,x is greater than the preset data influence threshold, it is determined that the s-th feature key group in the current period has an influence on the data of the x-th feature key group, and the x-th feature key group is recorded as the influence feature key group of the s-th feature key group; Step S304: Obtain and gather the influencing characteristic key groups of several characteristic key groups of the electronic components of this model to obtain the influence degree data.

5. The method for detecting and analyzing faults of electronic components according to claim 4, wherein, The step S400 includes: Step S401: Obtain the influence degree data of the electronic components of this model. Analyze the failure causes in the production process of the electronic components from the influencing characteristic key groups corresponding to the characteristic key groups in the influence degree data. The specific process is as follows: When the ratio of the total number of influencing characteristic key groups caused by a certain characteristic key group to the characteristic key groups in the key failure data is greater than the preset ratio threshold, record this characteristic key group as the target characteristic key group; Obtain the failure cause corresponding to the target characteristic key group, and record it as the target failure cause of the electronic components of this model in the current period; Step S402: Obtain several target failure causes of the electronic components of this model in the current period, and check the production and processing line of the electronic components according to the several target failure causes, and optimize and adjust the production and processing line of the electronic components of this model.

6. An electronic component fault detection and analysis system for performing the method for detecting and analyzing faults of an electronic component according to any one of claims 1-5, characterized in that, The system includes a key index data module, a fault impact module, an impact degree data module, and a fault analysis module; The key index data module is used to evaluate the detection indexes in electronic components and the key degree of fault detection of the electronic components, so as to obtain key index data; The fault impact module is used to analyze the key degree of the marked key group on the fault of the electronic components in the current cycle, so as to obtain fault key data; The impact degree data module is used to analyze the data impact degree between different characteristic key groups in the electronic components in the current cycle according to the fault key data, so as to obtain impact degree data; The fault analysis module is used to analyze the fault causes in the production process of the electronic components and optimize and adjust the processing and production of the electronic components.

7. An electronic component fault detection and analysis system according to claim 6, characterized in that, The key index data module includes a key score unit and a key index data unit; The key score unit is used to calculate the key scores of the detection indexes of the electronic components; The key index data unit is used to determine the key degree of the detection index on the fault detection of the electronic components of the model according to the key score, so as to obtain key index data.

8. An electronic component fault detection and analysis system according to claim 6, characterized in that, The fault impact module includes a fault impact scoring unit and a fault impact unit; The fault impact scoring unit is used to generate the marked key group of the electronic components according to the key index data and calculate the fault impact scores of the respective marked key groups of the electronic components; The fault impact unit is used to analyze the key degree of the marked key group on the fault of the electronic components in the current cycle according to the fault impact score, so as to obtain fault key data.

9. An electronic component fault detection and analysis system according to claim 6, wherein, The impact degree data module includes a data impact value unit and an impact degree data unit; The data impact value unit is used to obtain the fault key data, obtain the characteristic key group from the fault key data, and calculate the data impact value between the respective characteristic key groups; The impact degree data unit is used to analyze the data impact degree of the respective characteristic key groups of the electronic components in the current cycle according to the data impact value, so as to obtain impact degree data.

10. An electronic component fault detection and analysis system according to claim 6, characterized in that, The fault analysis module includes a fault analysis unit; The fault analysis unit is used to obtain the impact degree data and the fault key data of the electronic components, analyze the fault causes in the production process of the electronic components, and optimize and adjust the processing and production of the electronic components.

Images