Method and system for detecting factory qualification of computer mainboard based on artificial intelligence

Abstract

The invention discloses a computer mainboard delivery qualification detection method and system based on artificial intelligence, and relates to the technical field of detection, and the method comprises the steps: obtaining delivery detection data of a plurality of continuous computer mainboards closest to the current moment, and recording the delivery detection data as sample detection data; whether the reject ratio of the sample detection data is larger than a preset value or not is judged, if yes, the sample detection data are analyzed based on a detection analysis algorithm, and if not, no response is made; wherein the detection analysis algorithm specifically comprises the following steps: constructing an unqualified defect type library; adding bad indexes; related processing steps of unqualified defect types are determined; adding an initial abnormal index to each computer mainboard processing step; and calculating an actual abnormal index of each computer mainboard processing step through an AI mining algorithm. The method has the advantages that hidden danger steps in the computer mainboard processing factory can be quickly and accurately recognized, then efficient management of the computer mainboard processing factory is achieved, and the computer mainboard processing efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of detection technology, and in particular to an artificial intelligence-based computer motherboard factory qualified detection method and system. Background Art

[0002] The motherboard, also called mainboard, systemboard, or motherboard, is one of the most basic and important components of a computer. The motherboard is generally a rectangular circuit board on which the main circuit systems that make up the computer are installed, generally including BIOS chip, I / O control chip, keyboard and panel control switch interface, indicator light connector, expansion slot, DC power supply connector for the motherboard and plug-in card, and other components.

[0003] In order to ensure the factory yield of computer motherboards, computer motherboards usually need to undergo qualification tests before leaving the factory. The test results reflect the execution status of each step in the processing of the computer motherboard. By analyzing the test results, guidance can be provided for the maintenance of computer motherboard processing equipment. However, the existing solutions lack the correlation analysis between the various computer motherboard processing steps. The abnormal analysis of computer motherboard processing steps is limited to the results directly shown by the test results, and it is difficult to explore the deep abnormal steps in the computer motherboard processing steps. Summary of the invention

[0004] In order to solve the above technical problems, a tool is provided. The technical solution solves the problem that the above existing solutions lack correlation analysis between various computer motherboard processing steps, and the abnormal analysis of computer motherboard processing steps is limited to the results directly displayed by the test results, and it is difficult to explore the deep abnormal steps in the computer motherboard processing steps.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A method for testing the factory quality of computer motherboards based on artificial intelligence, comprising:

[0007] Obtain factory inspection data of a number of consecutive computer motherboards closest to the current time, and record them as sample inspection data;

[0008] Determine whether the unqualified rate of the sample test data is greater than a preset value. If so, analyze the sample test data based on the detection analysis algorithm. If not, do not respond.

[0009] The detection and analysis algorithm is specifically as follows:

[0010] Determine all types of unqualified defects that appear in sample test data and build a library of unqualified defect types;

[0011] Based on the number of occurrences of each type of unqualified defect in the sample inspection data, a defect index is added to each type of unqualified defect;

[0012] Determine at least one computer motherboard processing step that causes each unqualified defect type, and record it as the relevant processing step of the unqualified defect type;

[0013] Based on the bad indicators of all unqualified defect types and the related processing steps of all unqualified defect types, initial abnormal indicators are added to each computer motherboard processing step;

[0014] Based on the computer motherboard processing flow, determine the correlation between multiple computer motherboard processing steps;

[0015] Based on the initial abnormal indicators of all computer motherboard processing steps and the correlation between multiple computer motherboard processing steps, the actual abnormal indicators of each computer motherboard processing step are calculated through AI mining algorithm;

[0016] At least one computer motherboard processing step whose actual abnormality is greater than a preset value is screened out and recorded as an abnormal step, and the equipment corresponding to the abnormal step is repaired.

[0017] Preferably, the adding of a defect indicator to each unqualified defect type based on the number of occurrences of each unqualified type in the sample detection data specifically includes:

[0018] Based on the comparison between the number of occurrences of each type of unqualified defect in the sample test data and the total number of sample test data, the occurrence frequency of each type of unqualified defect is obtained;

[0019] The occurrence frequency of each unqualified defect type is used as the bad indicator of the unqualified defect type, and a bad indicator set A of the unqualified defect type is constructed, A = {a1…a i …a n}, where a i is the defect index of the i-th unqualified defect type, and n is the total number of unqualified defect types.

[0020] Preferably, the method of adding an initial abnormal indicator to each computer motherboard processing step based on the bad indicators of all unqualified defect types and the related processing steps of all unqualified defect types specifically includes:

[0021] Based on historical processing experience, the direct processing steps that cause the types of unqualified defects are determined as directly related steps, and the related processing steps other than the directly related steps are recorded as indirectly related steps;

[0022] Count all the unqualified defect types for each computer motherboard processing step as a directly related step and all the unqualified defect types for each computer motherboard processing step as an indirectly related step;

[0023] Calculate the directly related abnormal indicators of the computer motherboard processing steps based on the directly related formula;

[0024] Calculate the indirect related abnormal indicators of computer motherboard processing steps based on the indirect related formula;

[0025] Adding the direct related abnormality index and the indirect related abnormality index of the computer motherboard processing step to obtain the initial abnormality index of the computer motherboard processing step;

[0026] Among them, the directly related formula is specifically:

[0027]

[0028] Where, X dir is the directly related abnormal index of the computer motherboard processing step, m1 is the total number of unqualified defect types of the computer motherboard processing step as a directly related step, and a j is the bad indicator of the jth unqualified defect type of the computer motherboard processing step as a directly related step, The total number of related processing steps for the jth unqualified defect type where the computer motherboard processing step is a directly related step;

[0029] The indirect correlation formula is specifically:

[0030]

[0031] Where, X ind is the indirect related abnormal index of the computer motherboard processing step, m2 is the total number of unqualified defect types of the computer motherboard processing step as an indirect related step, a k is the bad indicator of the kth unqualified defect type of the computer motherboard processing step as an indirectly related step, The total number of related processing steps for the kth unqualified defect type of the computer motherboard processing step as an indirect related step.

[0032] Preferably, the interrelationships among the multiple computer motherboard processing steps are determined based on the computer motherboard processing flow as follows:

[0033] Based on the computer motherboard processing flow, several related processing chains are identified;

[0034] All computer motherboard processing steps before the computer motherboard processing step in the processing chain are recorded as preceding steps, and all computer motherboard processing steps after the computer motherboard processing step are recorded as subsequent steps.

[0035] Preferably, the AI ​​mining algorithm is specifically:

[0036] S1. Set the AI ​​learning index, recorded as Z, and assign the Z value to 1;

[0037] S2. Set the anomaly mining index, denoted as P u (Z), P u (Z) is the Zth generation abnormal mining indicator of the uth computer motherboard processing step, and let is the initial abnormality index of the u-th computer motherboard processing step;

[0038] S3, the value of Z is increased by one, and P is calculated through the AI ​​mining formula u (Z);

[0039] S4, judging whether the termination condition is met, if so, outputting the actual abnormal index as the computer motherboard processing step, if not, returning to S3;

[0040] The AI ​​mining formula is specifically:

[0041]

[0042] Where Yu is the total number of subsequent steps corresponding to the u-th computer motherboard processing step, and W yu is the total number of preceding steps of the yuth subsequent step corresponding to the uth computer motherboard processing step, P yu (Z-1) is the Z-1 generation anomaly mining index of the yu-th subsequent step corresponding to the u-th computer motherboard processing step, and α is the mining coefficient.

[0043] Preferably, the termination condition is specifically:

[0044] Determine whether the average change rate of the Zth generation anomaly mining index and the Z-1th generation anomaly mining index of all computer motherboard processing steps is less than the change rate threshold. If so, it is determined that the termination condition is met; if not, it is determined that the termination condition is not met.

[0045] Furthermore, a computer motherboard factory qualified detection system based on artificial intelligence is proposed, including:

[0046] A storage module, the storage module is used to store a computer motherboard factory inspection database;

[0047] A factory inspection module, the factory inspection module is electrically connected to the storage module, and the factory inspection module is used to detect whether the computer motherboard has unqualified defects when it leaves the factory, and store the detection results in the computer motherboard factory inspection database;

[0048] An initial analysis module, the initial analysis module is electrically connected to the storage module, and is used to determine all types of unqualified defects appearing in the sample detection data, build an unqualified defect type library, and add a bad indicator to each unqualified defect type based on the number of occurrences of each unqualified defect type in the sample detection data;

[0049] a step abnormal initial identification module, the step abnormal initial identification module is electrically connected to the initial analysis module, the step abnormal initial identification module is used to determine at least one computer motherboard processing step that causes each unqualified defect type, recorded as the relevant processing step of the unqualified defect type, and based on the bad indicators of all unqualified defect types and the relevant processing steps of all unqualified defect types, add an initial abnormal indicator to each computer motherboard processing step;

[0050] An AI analysis module is electrically connected to the step abnormality initial identification module, and is used to determine the correlation between multiple computer motherboard processing steps based on the computer motherboard processing flow, and based on the initial abnormality indicators of all computer motherboard processing steps and the correlation between multiple computer motherboard processing steps, calculate the actual abnormality indicators of each computer motherboard processing step through an AI mining algorithm, screen out at least one computer motherboard processing step whose actual abnormality indicator is greater than a preset value, record it as an abnormal step, and repair the equipment corresponding to the abnormal step.

[0051] Optionally, the AI ​​analysis module includes:

[0052] A correlation analysis unit, the correlation analysis unit is used to determine a plurality of processing chains that are related to each other based on the computer motherboard processing flow, and record all computer motherboard processing steps that are before the computer motherboard processing step in the processing chain as preceding steps, and record all computer motherboard processing steps that are after the computer motherboard processing step as subsequent steps;

[0053] A deep mining unit, the deep mining unit is used to calculate the actual abnormality index of each computer motherboard processing step through an AI mining algorithm based on the initial abnormality index of all computer motherboard processing steps and the mutual correlation relationship between multiple computer motherboard processing steps;

[0054] The maintenance planning unit is used to screen out at least one computer motherboard processing step whose actual abnormality index is greater than a preset value, record it as an abnormal step, and perform maintenance on the equipment corresponding to the abnormal step.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] The present invention proposes a computer motherboard factory qualified inspection scheme based on artificial intelligence. By analyzing the factory inspection data of computer motherboards, the initial abnormal indicators of each computer motherboard processing step are preliminarily identified based on the correlation between each unqualified defect and the computer motherboard processing step. The computer motherboard processing steps that are deeply related to the current unqualified performance are mined through an AI mining algorithm combined with the mutual correlation between the computer motherboard processing steps, which facilitates the rapid and accurate identification of hidden danger steps in the computer motherboard processing plant, thereby realizing efficient management of the computer motherboard processing plant and improving the efficiency of computer motherboard processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a flow chart of the method for testing the factory quality of computer motherboards based on artificial intelligence proposed by the present invention;

[0058] Figure 2 A flow chart of a method for adding a bad indicator to each unqualified defect type in the present invention;

[0059] Figure 3 A flow chart of the method for adding an initial abnormality indicator to each computer motherboard processing step in the present invention;

[0060] Figure 4 A flow chart of a method for determining the interrelationships between multiple computer motherboard processing steps in the present invention;

[0061] Figure 5 It is a schematic diagram of the structure of the electronic device of the present invention;

[0062] Figure 6 It is a schematic diagram of the computer-readable storage medium structure of the present invention. DETAILED DESCRIPTION

[0063] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0064] Reference Figure 1 As shown, a method for detecting factory qualified computer motherboards based on artificial intelligence includes:

[0065] Obtain factory inspection data of a number of consecutive computer motherboards closest to the current time, and record them as sample inspection data;

[0066] Determine whether the unqualified rate of the sample test data is greater than a preset value. If so, analyze the sample test data based on the detection analysis algorithm. If not, do not respond.

[0067] The specific detection and analysis algorithm is as follows:

[0068] Determine all types of unqualified defects that appear in sample test data and build a library of unqualified defect types;

[0069] Based on the number of occurrences of each type of unqualified defect in the sample inspection data, a defect index is added to each type of unqualified defect;

[0070] Determine at least one computer motherboard processing step that causes each unqualified defect type, and record it as the relevant processing step of the unqualified defect type;

[0071] Based on the bad indicators of all unqualified defect types and the related processing steps of all unqualified defect types, initial abnormal indicators are added to each computer motherboard processing step;

[0072] Based on the computer motherboard processing flow, determine the correlation between multiple computer motherboard processing steps;

[0073] Based on the initial abnormal indicators of all computer motherboard processing steps and the correlation between multiple computer motherboard processing steps, the actual abnormal indicators of each computer motherboard processing step are calculated through AI mining algorithm;

[0074] At least one computer motherboard processing step whose actual abnormality is greater than a preset value is screened out and recorded as an abnormal step, and the equipment corresponding to the abnormal step is repaired.

[0075] This solution analyzes the factory inspection data of computer motherboards, and preliminarily identifies the initial abnormal indicators of each computer motherboard processing step based on the correlation between each unqualified defect and the computer motherboard processing step. Through the AI ​​mining algorithm, combined with the mutual correlation between the computer motherboard processing steps, the computer motherboard processing steps that are deeply related to the current unqualified performance are mined, which facilitates the rapid and accurate identification of hidden danger steps in computer motherboard processing plants.

[0076] Reference Figure 2 As shown, based on the number of occurrences of each unqualified type in the sample test data, additional defect indicators are added to each unqualified defect type, including:

[0077] Based on the comparison between the number of occurrences of each type of unqualified defect in the sample test data and the total number of sample test data, the occurrence frequency of each type of unqualified defect is obtained;

[0078] The occurrence frequency of each unqualified defect type is used as the bad indicator of the unqualified defect type, and a bad indicator set A of the unqualified defect type is constructed, A = {a1…a i …a n}, where a i is the defect index of the i-th unqualified defect type, and n is the total number of unqualified defect types.

[0079] Reference Figure 3 As shown, based on the bad indicators of all unqualified defect types and the related processing steps of all unqualified defect types, the initial abnormal indicators added to each computer motherboard processing step specifically include:

[0080] Based on historical processing experience, the direct processing steps that cause the types of unqualified defects are determined as directly related steps, and the related processing steps other than the directly related steps are recorded as indirectly related steps;

[0081] Count all the unqualified defect types for each computer motherboard processing step as a directly related step and all the unqualified defect types for each computer motherboard processing step as an indirectly related step;

[0082] Calculate the directly related abnormal indicators of the computer motherboard processing steps based on the directly related formula;

[0083] Calculate the indirect related abnormal indicators of computer motherboard processing steps based on the indirect related formula;

[0084] Adding the direct related abnormality index and the indirect related abnormality index of the computer motherboard processing step to obtain the initial abnormality index of the computer motherboard processing step;

[0085] Among them, the directly related formula is specifically:

[0086]

[0087] Where, X dir is the directly related abnormal index of the computer motherboard processing step, m1 is the total number of unqualified defect types of the computer motherboard processing step as a directly related step, and a j is the bad indicator of the jth unqualified defect type of the computer motherboard processing step as a directly related step, The total number of related processing steps for the jth unqualified defect type where the computer motherboard processing step is a directly related step;

[0088] The indirect correlation formula is as follows:

[0089]

[0090] Where, X ind is the indirect related abnormal index of the computer motherboard processing step, m2 is the total number of unqualified defect types of the computer motherboard processing step as an indirect related step, a k is the bad indicator of the kth unqualified defect type of the computer motherboard processing step as an indirectly related step, The total number of related processing steps for the kth unqualified defect type of the computer motherboard processing step as an indirect related step.

[0091] It can be understood that among the causes of unqualified processing, the factors that directly lead to unqualified defects are the decisive factors. Based on this, when evaluating the relationship between the computer motherboard processing steps and the types of unqualified defects in this scheme, a clear-cut approach is adopted to make judgments, and the indicator ratio between the directly related computer motherboard processing steps and the indirectly related computer motherboard processing steps is set to 2:1. Through the relationship between each computer motherboard processing step and multiple types of unqualified defects, the additional initial abnormal indicators of the computer motherboard processing steps are comprehensively evaluated. The larger the initial abnormal indicator, the more directly related the computer motherboard processing step is to the current unqualified detection status.

[0092] Reference Figure 4 As shown, based on the computer motherboard processing flow, the interrelationships between multiple computer motherboard processing steps are determined as follows:

[0093] Based on the computer motherboard processing flow, several related processing chains are identified;

[0094] All computer motherboard processing steps before the computer motherboard processing step in the processing chain are recorded as preceding steps, and all computer motherboard processing steps after the computer motherboard processing step are recorded as subsequent steps.

[0095] The specific AI mining algorithm is:

[0096] S1. Set the AI ​​learning index, recorded as Z, and assign the Z value to 1;

[0097] S2. Set the anomaly mining index, denoted as P u (Z), P u (Z) is the Zth generation abnormal mining indicator of the uth computer motherboard processing step, and let is the initial abnormality index of the u-th computer motherboard processing step;

[0098] S3, the value of Z is increased by one, and P is calculated through the AI ​​mining formula u (Z);

[0099] S4, judging whether the termination condition is met, if so, outputting the actual abnormal index as the computer motherboard processing step, if not, returning to S3;

[0100] Among them, the AI ​​mining formula is specifically:

[0101]

[0102] Where Yu is the total number of subsequent steps corresponding to the u-th computer motherboard processing step, and W yuis the total number of preceding steps of the yuth subsequent step corresponding to the uth computer motherboard processing step, P yu (Z-1) is the Z-1 generation anomaly mining index of the yu-th subsequent step corresponding to the u-th computer motherboard processing step, and α is the mining coefficient.

[0103] The termination conditions are as follows:

[0104] Determine whether the average change rate of the Zth generation anomaly mining index and the Z-1th generation anomaly mining index of all computer motherboard processing steps is less than the change rate threshold. If so, it is determined that the termination condition is met; if not, it is determined that the termination condition is not met.

[0105] After several mining calculations, the mining formula in this scheme will tend to stabilize the abnormal mining indicators of each computer motherboard processing step. In this scheme, the mining termination condition is set based on the mining change degree of the abnormal mining indicator. The larger the value of the mining change rate threshold, the fewer mining times and the smaller the mining required computing power, but the less accurate the abnormal step identification. Conversely, the smaller the value of the mining change rate threshold, the more mining times, the more mining required computing power, and the deeper the abnormal step identification mining.

[0106] It is understandable that there are multiple correlations between the various steps of computer motherboard processing. One computer motherboard processing step may trigger a chain reaction and affect the execution of other computer motherboard processing steps. Therefore, there are multiple correlations between the various computer motherboard processing steps that are fully explored in this scheme. Based on the fact that a computer motherboard processing step may cause other computer motherboard processing steps to be caused, multiple mining calculations are performed to fully explore the root causes of the unqualified defects in the computer host. For example, step A will affect step B, and step B is repeatedly used as the direct cause of the unqualified defects, resulting in a very high initial abnormal index of step B, while step A did not appear in any of the related steps of the multiple unqualified defects. At this time, since step A is the preceding step of step B, through multiple mining, it can be calculated that the actual abnormal index of step A is also very high, thereby realizing the identification and mining of the abnormal step of step A.

[0107] Further, based on the same inventive concept as the above-mentioned artificial intelligence-based computer motherboard factory qualified detection method, this solution proposes an artificial intelligence-based computer motherboard factory qualified detection system, including:

[0108] A storage module, the storage module is used to store a factory inspection database of computer motherboards;

[0109] A factory inspection module, the factory inspection module is electrically connected to the storage module, and the factory inspection module is used to detect whether the computer motherboard has unqualified defects when it leaves the factory, and store the detection results in the computer motherboard factory inspection database;

[0110] An initial analysis module, the initial analysis module is electrically connected to the storage module, and is used to determine all types of unqualified defects appearing in the sample detection data, build an unqualified defect type library, and add a bad indicator to each unqualified defect type based on the number of occurrences of each unqualified defect type in the sample detection data;

[0111] A step abnormal initial identification module, the step abnormal initial identification module is electrically connected to the initial analysis module, and the step abnormal initial identification module is used to determine at least one computer motherboard processing step that causes each unqualified defect type, recorded as the relevant processing step of the unqualified defect type, and based on the bad indicators of all unqualified defect types and the relevant processing steps of all unqualified defect types, an initial abnormal indicator is added to each computer motherboard processing step;

[0112] AI analysis module, the AI ​​analysis module is electrically connected to the step abnormality initial identification module, the AI ​​analysis module is used to determine the correlation between multiple computer motherboard processing steps based on the computer motherboard processing flow, based on the initial abnormal indicators of all computer motherboard processing steps and the correlation between multiple computer motherboard processing steps, through the AI ​​mining algorithm, calculate the actual abnormal indicators of each computer motherboard processing step, screen out at least one computer motherboard processing step with an actual abnormal indicator greater than a preset value, record it as an abnormal step, and repair the equipment corresponding to the abnormal step.

[0113] AI analysis modules include:

[0114] A correlation analysis unit, the correlation analysis unit is used to determine a number of related processing chains based on the computer motherboard processing flow, and record all computer motherboard processing steps before the computer motherboard processing step in the processing chain as preceding steps, and record all computer motherboard processing steps after the computer motherboard processing step as subsequent steps;

[0115] A deep mining unit is used to calculate the actual abnormality index of each computer motherboard processing step through an AI mining algorithm based on the initial abnormality index of all computer motherboard processing steps and the correlation between multiple computer motherboard processing steps;

[0116] The maintenance planning unit is used to screen out at least one computer motherboard processing step whose actual abnormality index is greater than a preset value, record it as an abnormal step, and repair the equipment corresponding to the abnormal step.

[0117] The working process of the above system is:

[0118] Step 1: The factory inspection module is used to detect whether the computer motherboard has unqualified defects when it leaves the factory, and store the inspection results in the computer motherboard factory inspection database;

[0119] Step 2: The initial analysis module obtains the factory inspection data of several consecutive computer motherboards closest to the current moment from the storage module, records them as sample inspection data, and determines all the unqualified defect types appearing in the sample inspection data, builds an unqualified defect type library, and adds a bad indicator to each unqualified defect type based on the number of occurrences of each unqualified defect type in the sample inspection data;

[0120] Step 3: The step abnormal initial identification module determines at least one computer motherboard processing step that causes each unqualified defect type, which is recorded as the relevant processing step of the unqualified defect type, and based on the bad indicators of all unqualified defect types and the relevant processing steps of all unqualified defect types, an initial abnormal indicator is added to each computer motherboard processing step;

[0121] Step 4: The correlation analysis unit determines a number of related processing chains based on the computer motherboard processing flow, and records all computer motherboard processing steps before the computer motherboard processing step in the processing chain as preceding steps, and records all computer motherboard processing steps after the computer motherboard processing step as subsequent steps;

[0122] Step 5: The deep mining unit calculates the actual abnormality index of each computer motherboard processing step through an AI mining algorithm based on the initial abnormality index of all computer motherboard processing steps and the correlation between multiple computer motherboard processing steps;

[0123] Step 6: The maintenance planning unit selects at least one computer motherboard processing step whose actual abnormality index is greater than a preset value, records it as an abnormal step, and repairs the equipment corresponding to the abnormal step.

[0124] Furthermore, the method according to the embodiment of the present application can also be performed by Figure 5 The electronic device architecture shown in FIG. Figure 5 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as ROM 503 or hard disk 507, may store a computer motherboard factory qualified detection method based on artificial intelligence provided by the present application. The electronic device 500 may also include a user interface 508. Of course, Figure 5 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 5 One or more components of an electronic device are shown.

[0125] Figure 6 Schematic diagram of a computer-readable storage medium structure provided by an embodiment of the present application. Figure 6 As shown, a computer-readable storage medium 600 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, a method for detecting the factory-qualified computer motherboard based on artificial intelligence according to an embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0126] In summary, the advantages of the present invention are that it can quickly and accurately identify the hidden danger steps in the computer motherboard processing plant, thereby achieving efficient management of the computer motherboard processing plant and improving the efficiency of computer motherboard processing.

[0127] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for testing the factory quality of computer motherboards based on artificial intelligence, characterized in that: include: Obtain factory inspection data of a number of consecutive computer motherboards closest to the current time, and record them as sample inspection data; Determine whether the unqualified rate of the sample test data is greater than a preset value. If so, analyze the sample test data based on the detection analysis algorithm. If not, do not respond. The detection and analysis algorithm is specifically as follows: Determine all types of unqualified defects that appear in sample test data and build a library of unqualified defect types; Based on the number of occurrences of each type of unqualified defect in the sample inspection data, a defect index is added to each type of unqualified defect; Determine at least one computer motherboard processing step that causes each unqualified defect type, and record it as the relevant processing step of the unqualified defect type; Based on the bad indicators of all unqualified defect types and the related processing steps of all unqualified defect types, initial abnormal indicators are added to each computer motherboard processing step; Based on the computer motherboard processing flow, determine the correlation between multiple computer motherboard processing steps; Based on the initial abnormal indicators of all computer motherboard processing steps and the correlation between multiple computer motherboard processing steps, the actual abnormal indicators of each computer motherboard processing step are calculated through AI mining algorithm; At least one computer motherboard processing step whose actual abnormality is greater than a preset value is screened out and recorded as an abnormal step, and the equipment corresponding to the abnormal step is repaired.

2. According to the artificial intelligence-based computer motherboard factory qualified detection method of claim 1, it is characterized in that: The method of adding a defect indicator to each unqualified defect type based on the number of occurrences of each unqualified type in the sample detection data specifically includes: Based on the comparison between the number of occurrences of each type of unqualified defect in the sample test data and the total number of sample test data, the occurrence frequency of each type of unqualified defect is obtained; The occurrence frequency of each unqualified defect type is used as the bad indicator of the unqualified defect type, and a bad indicator set A of the unqualified defect type is constructed, A = {a1…a i …a n }, where a i is the defect index of the i-th unqualified defect type, and n is the total number of unqualified defect types.

3. According to claim 2, a method for detecting the factory quality of computer motherboards based on artificial intelligence is characterized in that: The above-mentioned method of adding initial abnormal indicators to each computer motherboard processing step based on the bad indicators of all unqualified defect types and the related processing steps of all unqualified defect types specifically includes: Based on historical processing experience, the direct processing steps that cause the types of unqualified defects are determined as directly related steps, and the related processing steps other than the directly related steps are recorded as indirectly related steps; Count all the unqualified defect types for each computer motherboard processing step as a directly related step and all the unqualified defect types for each computer motherboard processing step as an indirectly related step; Calculate the directly related abnormal indicators of the computer motherboard processing steps based on the directly related formula; Calculate the indirect related abnormal indicators of computer motherboard processing steps based on the indirect related formula; Adding the direct related abnormality index and the indirect related abnormality index of the computer motherboard processing step to obtain the initial abnormality index of the computer motherboard processing step; The directly related formula is specifically: Where, X dir is the directly related abnormal index of the computer motherboard processing step, m1 is the total number of unqualified defect types of the computer motherboard processing step as a directly related step, and a j is the bad indicator of the jth unqualified defect type of the computer motherboard processing step as a directly related step, The total number of related processing steps for the jth unqualified defect type where the computer motherboard processing step is a directly related step; The indirect correlation formula is specifically: Where, X ind is the indirect related abnormal index of the computer motherboard processing step, m2 is the total number of unqualified defect types of the computer motherboard processing step as an indirect related step, a k is the bad indicator of the kth unqualified defect type of the computer motherboard processing step as an indirectly related step, The total number of related processing steps for the kth unqualified defect type of the computer motherboard processing step as an indirect related step.

4. According to claim 3, a method for detecting the factory quality of computer motherboards based on artificial intelligence is characterized in that: The interrelationships between the multiple computer motherboard processing steps are determined based on the computer motherboard processing flow as follows: Based on the computer motherboard processing flow, several related processing chains are identified; All computer motherboard processing steps before the computer motherboard processing step in the processing chain are recorded as preceding steps, and all computer motherboard processing steps after the computer motherboard processing step are recorded as subsequent steps.

5. The method for detecting the factory quality of computer motherboards based on artificial intelligence according to claim 4 is characterized in that: The AI ​​mining algorithm is specifically: S1. Set the AI ​​learning index, recorded as Z, and assign the Z value to 1; S2. Set the anomaly mining index, denoted as P u (Z), P u (Z) is the Zth generation abnormal mining indicator of the uth computer motherboard processing step, and let is the initial abnormality index of the u-th computer motherboard processing step; S3, the value of Z is increased by one, and P is calculated through the AI ​​mining formula u (Z); S4, judging whether the termination condition is met, if so, outputting the actual abnormal index as the computer motherboard processing step, if not, returning to S3; The AI ​​mining formula is specifically: Where Yu is the total number of subsequent steps corresponding to the u-th computer motherboard processing step, and W yu is the total number of preceding steps of the yuth subsequent step corresponding to the uth computer motherboard processing step, P yu (Z-1) is the Z-1 generation anomaly mining index of the yu-th subsequent step corresponding to the u-th computer motherboard processing step, and α is the mining coefficient.

6. The method for detecting the factory quality of computer motherboards based on artificial intelligence according to claim 5 is characterized in that: The termination conditions are specifically: Determine whether the average change rate of the Zth generation anomaly mining index and the Z-1th generation anomaly mining index of all computer motherboard processing steps is less than the change rate threshold. If so, it is determined that the termination condition is met; if not, it is determined that the termination condition is not met.

7. A computer motherboard factory qualified inspection system based on artificial intelligence, characterized in that: include: A storage module, the storage module is used to store a computer motherboard factory inspection database; A factory inspection module, the factory inspection module is electrically connected to the storage module, and the factory inspection module is used to detect whether the computer motherboard has unqualified defects when it leaves the factory, and store the detection results in the computer motherboard factory inspection database; An initial analysis module, the initial analysis module is electrically connected to the storage module, and is used to determine all types of unqualified defects appearing in the sample detection data, build an unqualified defect type library, and add a bad indicator to each unqualified defect type based on the number of occurrences of each unqualified defect type in the sample detection data; a step abnormal initial identification module, the step abnormal initial identification module is electrically connected to the initial analysis module, the step abnormal initial identification module is used to determine at least one computer motherboard processing step that causes each unqualified defect type, recorded as the relevant processing step of the unqualified defect type, and based on the bad indicators of all unqualified defect types and the relevant processing steps of all unqualified defect types, add an initial abnormal indicator to each computer motherboard processing step; An AI analysis module is electrically connected to the step abnormality initial identification module, and is used to determine the correlation between multiple computer motherboard processing steps based on the computer motherboard processing flow, and based on the initial abnormality indicators of all computer motherboard processing steps and the correlation between multiple computer motherboard processing steps, calculate the actual abnormality indicators of each computer motherboard processing step through an AI mining algorithm, screen out at least one computer motherboard processing step whose actual abnormality indicator is greater than a preset value, record it as an abnormal step, and repair the equipment corresponding to the abnormal step.

8. The artificial intelligence-based computer motherboard factory qualified detection system according to claim 7 is characterized in that: The AI ​​analysis module includes: A correlation analysis unit, the correlation analysis unit is used to determine a plurality of processing chains that are related to each other based on the computer motherboard processing flow, and record all computer motherboard processing steps that are before the computer motherboard processing step in the processing chain as preceding steps, and record all computer motherboard processing steps that are after the computer motherboard processing step as subsequent steps; A deep mining unit, the deep mining unit is used to calculate the actual abnormality index of each computer motherboard processing step through an AI mining algorithm based on the initial abnormality index of all computer motherboard processing steps and the mutual correlation relationship between multiple computer motherboard processing steps; The maintenance planning unit is used to screen out at least one computer motherboard processing step whose actual abnormality index is greater than a preset value, record it as an abnormal step, and perform maintenance on the equipment corresponding to the abnormal step.

9. An electronic device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor, wherein: The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the artificial intelligence-based computer motherboard factory qualification detection method as described in any one of claims 1-6.

10. A computer-readable storage medium storing a computer-readable program, characterized in that: When the computer-readable program is executed by the processor, the method for detecting factory qualified computer motherboards based on artificial intelligence described in any one of claims 1 to 6 is implemented.

Images