Fault detection method and system for computer memory bank

Abstract

The invention relates to the technical field of electronic digital data processing, in particular to a fault detection method and system for a computer memory bank. The method comprises the following steps: after a computer main board inserted into a memory bank is self-checked through a BIOS (Basic Input / Output System), acquiring audio data of a prompt tone played after self-checking from microphone equipment, and acquiring secondary screen image data output after self-checking through USB (Universal Serial Bus) connection; identifying a prompt tone sequence in the audio data; matching the prompt tone sequence with a preset prompt tone sequence, and determining a fault category to which the memory bank belongs; identifying character information in the secondary screen image data; matching each standard fault code under the fault category with character information in the secondary screen image data, and determining a fault code in the secondary screen image data; and determining a fault detection result of the memory bank according to the fault code. By adopting the method, the fault detection efficiency of the memory bank can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic digital data processing, and in particular to a computer memory bar fault detection method and system. Background Art

[0002] Memory sticks are an important part of computers. During the production process, memory sticks need to be strictly monitored for quality. Therefore, it is very important to detect computer memory stick faults. Memory stick failures may cause frequent system crashes, blue screens, or restarts without reason, affecting normal use; or may trigger the motherboard frequency reduction protection, resulting in performance degradation.

[0003] In the traditional method, a fault detection card is generally inserted into the mainboard, and the fault detection card is used to detect the faults of various parts in the mainboard including the memory module and display the fault code. Then the fault type of the memory module is determined according to the fault code on the fault detection card. However, this method has a relatively low detection efficiency. Summary of the invention

[0004] In order to solve the technical problem of low fault detection efficiency, the purpose of the present invention is to provide a computer memory module fault detection method and system, and the technical solution adopted is as follows: The present invention provides a computer memory module fault detection method, the method comprising: After the computer motherboard into which the memory stick is inserted performs a self-test through the BIOS, the audio data of the prompt tone played after the self-test is obtained from the microphone device, and the secondary screen image data output after the self-test is obtained through the USB connection; Identifying a tone sequence in the audio data; Matching the prompt tone sequence with a preset prompt tone sequence to determine the major fault category to which the memory module belongs; Identifying character information in the secondary screen image data; Matching each standard fault code under the fault category with the character information in the secondary screen image data respectively to determine the fault code in the secondary screen image data; A fault detection result of the memory stick is determined according to the fault code.

[0005] According to the computer memory bar fault detection method provided by the present invention, before obtaining the audio data of the prompt sound played after the self-test from the microphone device and obtaining the secondary screen image data output after the self-test through the USB connection, the method also includes: The control robot inserts the memory stick into the computer mainboard to trigger the computer mainboard to perform a self-check through BIOS.

[0006] According to the method for detecting faults of a computer memory module provided by the present invention, the identifying of the beep sequence in the audio data includes: Performing binarization processing on the amplitude values of the audio data to obtain binarized audio data; Determining audio intervals corresponding to multiple beeps respectively according to the values in the binarized audio data; Performing binarization processing on the duration of each audio interval to obtain a beep sequence.

[0007] According to the method for detecting faults of a computer memory module provided by the present invention, the identifying of the character information in the secondary screen image data includes: Processing the secondary screen image data to obtain modulo data of the fault information; Reconstructing the modulo data into a two-dimensional matrix; Identifying the minimum bounding rectangles of each foreground element in the two-dimensional matrix; Matching the data within each minimum bounding rectangle with a standard character template to determine the character information in the secondary screen image data.

[0008] According to the method for detecting faults of a computer memory module provided by the present invention, the matching of each standard fault code under the fault category with the character information in the secondary screen image data to determine the fault code in the secondary screen image data includes: Sequentially selecting each standard fault code under the fault category in descending order of length; In each round of matching, sliding a sliding window with the same length as the currently selected standard fault code on the remaining unmatched character information in the secondary screen image data; Calculating the similarity degree between the characters in the sliding window and the currently selected standard fault code respectively at each position where the sliding window slides to; Determining whether the currently selected standard fault code exists in the secondary screen image data according to the similarity degrees at each position where it slides to.

[0009] According to the method for detecting faults of a computer memory module provided by the present invention, the fault detection result includes the fault category to which the memory module belongs; The determining of the fault detection result of the memory module according to the fault code includes: For each memory module respectively, forming a fault code sequence by arranging the various fault codes in the secondary screen image data corresponding to the memory module in the order of appearance; Performing clustering according to the fault code sequences corresponding to each memory module, and determining the fault category to which each memory module belongs according to the clustering result.

[0010] According to the method for detecting faults of a computer memory module provided by the present invention, the fault detection result includes the main fault cause of the memory module; Determining the fault detection result of the memory module according to the fault code includes: Determining the fault information weight corresponding to each fault code in the secondary screen image data according to the occurrence probability and context association degree of each fault code in the secondary screen image data at each occurrence position; Selecting the main fault code from various fault codes in the secondary screen image data according to the fault information weight, and determining the main fault cause of the memory module according to the main fault code.

[0011] According to the method for detecting faults of a computer memory module provided by the present invention, the method further includes: Determining the window corresponding to each fault code in the secondary screen image data at each occurrence position; the window is obtained by expanding a preset number of strings before and after the occurrence position of the fault code; Determining the context association degree of the fault code at each occurrence position according to the matching result between each string in the window corresponding to the fault code at each occurrence position and the standard fault code; Determining the occurrence probability of the fault code at the occurrence position according to the occurrence position of the fault code each time and the number of times the fault code appears at the occurrence position in history.

[0012] According to the method for detecting faults of a computer memory module provided by the present invention, after matching the prompt tone sequence with a preset prompt tone sequence to determine the major fault category to which the memory module belongs, the method further includes: If the major fault category to which the memory module belongs is that the memory module is not connected, controlling the robotic arm to insert the contact pins of the memory module into a cleaning device; the cleaning device is used to clean the oxide layer on the contact pins by friction; Controlling the robotic arm to take out the memory module from the cleaning device and re-insert it into the computer motherboard to trigger the computer motherboard to perform self-checking through the BIOS; Returning to execute the step of obtaining the audio data of the prompt tone played after self-checking from the microphone device and obtaining the secondary screen image data output after self-checking through the USB connection and subsequent steps after the computer motherboard with the memory module inserted performs self-checking through the BIOS.

[0013] The present invention provides a fault detection system for a computer memory module. The system includes a memory and a processor. The memory is used to store executable program codes. The processor is used to call and run the executable program codes from the memory to implement the fault detection method for the computer memory module provided by the present invention.

[0014] The present invention has the following beneficial effects: After the computer motherboard with the memory module inserted performs a self-check through the BIOS, it obtains the audio data of the prompt sound played after the self-check from the microphone device, and obtains the secondary screen image data output after the self-check through a USB connection. Then, it identifies the prompt sound sequence in the audio data, matches the prompt sound sequence with a preset prompt sound sequence to determine the major fault category to which the memory module belongs, identifies the character information in the secondary screen image data, matches each standard fault code under the major fault category with the character information in the secondary screen image data respectively to determine the fault code in the secondary screen image data, and finally determines the fault detection result of the memory module according to the fault code, realizing automatic fault detection of the memory module, avoiding the problem of low efficiency caused by manually determining the fault type of the memory module according to the fault codes on the fault detection card, and improving the fault detection efficiency of the memory module. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0016] Figure 1 It is a schematic flowchart of a fault detection method for a computer memory module provided by an embodiment of the present invention; Figure 2 It is a schematic diagram of the connection method of the USB-to-serial port module in a fault detection method for a computer memory module provided by an embodiment of the present invention; Figure 3 and Figure 4 It is a simulation effect diagram of a robotic arm grasping and placing a product at a specific position on a production line in a fault detection method for a computer memory module provided by an embodiment of the present invention; Figure 5 It is a schematic structural diagram of a cleaning device in a fault detection method for a computer memory module provided by an embodiment of the present invention; Figure 6 It is a schematic structural diagram of a fault detection system for a computer memory module provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a computer memory bar fault detection method and system proposed by the present invention, its specific implementation, structure, features and effects, in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0018] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0019] The specific scheme of a computer memory bar fault detection method and system provided by the present invention is described in detail below with reference to the accompanying drawings.

[0020] See also Figure 1 , which shows a flow chart of a computer memory bar fault detection method provided by an embodiment of the present invention, comprising the following steps: Step 102, after the computer motherboard into which the memory stick is inserted performs a self-test through the BIOS, the audio data of the prompt tone played after the self-test is obtained from the microphone device, and the secondary screen image data output after the self-test is obtained through the USB connection.

[0021] Among them, BIOS (Basic Input Output System) is a set of programs solidified on a ROM (Read-Only Memory) chip on the computer motherboard, which stores the most important basic input and output programs of the computer, the self-test program after startup and the system self-starting program, and can read and write specific information of system settings from CMOS (Complementary Metal Oxide Semiconductor). Every time the computer is turned on, BIOS provides a self-test program to self-test various components such as the system's circuit, memory, keyboard, video part, hard disk and floppy drive, and analyzes the hard disk system configuration, initializes the configured basic I / O (Input / Output) settings, and boots the operating system after everything is normal. After the memory stick is inserted into the computer motherboard, BIOS starts power-on self-test. The self-test results of BIOS can be displayed on the secondary screen through fault codes (the present invention does not use the secondary screen for display, but obtains the secondary screen image data through USB connection for character recognition).

[0022] In one embodiment, the fault detection method of the computer memory module in each embodiment of the present invention can be executed by a microcontroller unit (MCU, Microcontroller Unit). In one embodiment, the fault detection method of the computer memory module in each embodiment of the present invention can be executed by the microcontroller unit in a Raspberry Pi.

[0023] Specifically, the microphone device acquires the audio data of the prompt sound played after self-check, and transmits the audio data to the microcontroller unit. The microcontroller unit is connected to the main board through USB (Universal Serial Bus, Universal Serial Bus) to obtain the sub-screen image data output after self-check from the main board.

[0024] In one embodiment, since there may be certain noise effects in the factory assembly line, a soundproof cover can be used to cover the sound module when acquiring audio data through the microphone device, thereby improving the detection accuracy of the audio data of the prompt sound.

[0025] In one embodiment, through a USB to serial port module such as Figure 2 As shown, connect the pins 3V3, TXD, RXD, and GND to the 3V3 power, RXD, TXD, and Ground on the Raspberry Pi respectively. In another embodiment, the Raspberry Pi is directly connected to the USB interface of the main board, and a standard USB communication protocol, such as USB 2.0 or USB 3.0, needs to be used.

[0026] In one embodiment, after the hardware connection, it is necessary to download the Raspberry Pi driver and configure the serial port value. For example: the driver configuration can be CH340 and CH9344. The serial port value can be set to 115200.

[0027] It can be understood that the used Raspberry Pi does not include a screen, and only the microcontroller unit of the Raspberry Pi core is required. The microcontroller unit does not directly obtain the fault code, but obtains the data (i.e., sub-screen image data) of the display image (i.e., blue screen) of the fault code through the sub-screen method. Specifically, the microcontroller unit obtains the data packet that conforms to the USB transmission data type and contains the sub-screen image data output after self-check through the USB connection.

[0028] Step 104, identify the prompt sound sequence in the audio data.

[0029] In one embodiment, according to the amplitude value of the audio data, the audio intervals corresponding to multiple prompt sounds in the audio data can be identified, the duration of each audio interval can be determined, the long sound and the short sound can be identified according to the duration of each audio interval, and the prompt sound sequence can be obtained. For example: set the long sound to 1 and the short sound to 0, so as to obtain the prompt sound sequence (a sequence containing 0 and 1).

[0030] Step 106, match the prompt tone sequence with the preset prompt tone sequence to determine the major fault category to which the memory module belongs.

[0031] In one embodiment, multiple standard preset prompt tone sequences can be pre-set in a database. The prompt tone sequence is respectively matched with each preset prompt tone sequence, and the preset major fault category corresponding to the successfully matched preset prompt tone sequence is determined as the major fault category to which the memory module belongs.

[0032] As shown in Table 1 below, it shows the meanings represented by various forms of prompt tones. Therefore, matching the prompt tone sequence with the preset prompt tone sequence can determine the major fault category to which the memory module belongs.

[0033] Table 1 Meanings of Prompt Tones In one embodiment, if the major fault category to which the memory module belongs is that the memory module is not connected, clean the contact pins of the memory module, and re-insert the cleaned memory module into the computer motherboard to trigger the computer motherboard to perform self-checking through the BIOS. Return to execute after the computer motherboard with the memory module inserted performs self-checking through the BIOS, obtain the audio data of the prompt tone played after self-checking from the microphone device, and obtain the secondary screen image data output after self-checking and subsequent steps through USB connection.

[0034] It can be understood that there may be two situations where the major fault category to which the memory module belongs is that the memory module is not connected: one is a memory module fault, and the other is that the memory module is not inserted properly or the contact pins of the memory module are oxidized. Therefore, by cleaning the contact pins of the memory module and then re-inserting the cleaned memory module into the computer motherboard, the problem of false fault detection caused by the memory module not being inserted properly or the contact pins of the memory module being oxidized can be reduced.

[0035] Step 108, identify the character information in the secondary screen image data.

[0036] In one embodiment, the micro control unit processes the secondary screen image data to obtain the modulo data of the fault information. Then reconstruct the modulo data into a two-dimensional matrix, and determine the character information in the secondary screen image data according to the two-dimensional matrix.

[0037] Step 110, respectively match each standard fault code under the major fault category with the character information in the secondary screen image data to determine the fault code in the secondary screen image data.

[0038] In one embodiment, each standard fault code under the major fault category can be selected in descending order of length. In each round of matching, the currently selected standard fault code is matched with the remaining unmatched character information in the secondary screen image data to determine whether the currently selected standard fault code exists in the secondary screen image data and its location.

[0039] It can be understood that since the lengths of different standard fault codes are different, there is an inclusion relationship between the standard fault codes, which is likely to cause mis-matching. For example, there are two standard fault codes, 0001F and 00. If the standard fault code 00 is used to match the character information in the secondary screen image data, 0001F may be matched, resulting in mis-matching. Therefore, by selecting each standard fault code under the major fault category in descending order of length for matching, and the already matched character information is no longer involved in subsequent matching, the situation of mis-matching caused by the inclusion relationship between the standard fault codes can be avoided.

[0040] Step 112: Determine the fault detection result of the memory module according to the fault code.

[0041] In one embodiment, the fault detection result may include at least one of the fault category to which the memory module belongs and the main fault cause of the memory module.

[0042] In one embodiment, the memory modules belonging to the same fault category can be placed at the same position. Anti-oxidation films can be pasted on the memory modules without faults, and then they can be packaged. For the memory modules placed at the same position, corresponding measures are taken according to their respective main fault causes.

[0043] In one embodiment, the robotic arm can be controlled to place the memory modules belonging to the same fault category at the same position.

[0044] The above-mentioned method for detecting faults in computer memory modules, after the computer motherboard with the memory module inserted performs a self-check through the BIOS, obtains the audio data of the prompt sound played after the self-check from the microphone device, and obtains the secondary screen image data output after the self-check through a USB connection. Then, it identifies the prompt sound sequence in the audio data, matches the prompt sound sequence with a preset prompt sound sequence to determine the major fault category to which the memory module belongs, identifies the character information in the secondary screen image data, and respectively matches each standard fault code under the major fault category with the character information in the secondary screen image data to determine the fault code in the secondary screen image data. Finally, it determines the fault detection result of the memory module according to the fault code, realizing the automatic detection of faults in the memory module, avoiding the problem of low efficiency caused by manually determining the fault type of the memory module according to the fault code on the fault detection card, and improving the fault detection efficiency of the memory module. In addition, since the fault detection card is plugged and unplugged through contact pins (i.e., "gold fingers"), the contact pins are easily damaged. However, in the present invention, by obtaining the audio data of the prompt sound played after the self-check from the microphone device and obtaining the secondary screen image data output after the self-check through a USB connection, there is no need to plug and unplug the contact pins, thus avoiding the problem of damage to the contact pins that easily occurs when using a fault detection card and improving the stability of the memory module fault detection. By identifying the character information in the secondary screen image data and performing the matching of fault codes, the present invention also improves the accuracy of the memory module fault detection.

[0045] In one embodiment, before obtaining the audio data of the prompt sound played after the self-check from the microphone device and obtaining the secondary screen image data output after the self-check through a USB connection, the method further includes: controlling the robotic arm to insert the memory module into the computer motherboard to trigger the computer motherboard to perform a self-check through the BIOS.

[0046] It can be understood that since the position of the motherboard is fixed in the production line for memory module fault detection and the action of plugging and unplugging the memory module on the motherboard is relatively repetitive, the robotic arm can be used to replace manual labor.

[0047] Refer to Figure 3 and Figure 4 , which are the simulation effect diagrams of the robotic arm grasping and placing products at specific positions on the production line. The robotic arm picks up the memory module that needs to be fault-detected on the production line and inserts the memory module into the memory position of the motherboard for subsequent fault detection.

[0048] In one embodiment, the servo motor in the robotic arm can be controlled by a microcontroller unit to achieve the clamping effect.

[0049] In the above embodiment, controlling the robotic arm to insert the memory module into the computer motherboard to trigger the computer motherboard to perform a self-check through the BIOS can improve the automation level of the production line for memory module fault detection and improve the efficiency of memory module fault detection.

[0050] In one embodiment, identifying a beep sequence in audio data includes: binarizing the amplitude values of the audio data to obtain binarized audio data; determining audio intervals corresponding to multiple beeps respectively according to the values in the binarized audio data; and binarizing the duration of each audio interval to obtain the beep sequence.

[0051] In one embodiment, the amplitude values less than the first threshold can be set to 0, and the amplitude values greater than or equal to the first threshold can be set to 1 to obtain the binarized audio data.

[0052] In one embodiment, a first histogram of the amplitude values can be counted, and the segmentation point with the largest variance between two classes in the first histogram can be calculated through the Otsu algorithm to obtain the first threshold.

[0053] In one embodiment, according to the binarized amplitude values in the binarized audio data, an interval with consecutive binarized amplitude values being non-zero and the number of which meets a preset condition is set as an audio interval. Wherein, the preset condition is that the number is greater than or equal to a preset quantity. The preset quantity can be set according to the actual situation. For example: the preset quantity can be 3. That is, an interval with more than 3 consecutive binarized amplitude values being non-zero is set as an audio interval.

[0054] In one embodiment, determining the duration of each audio interval { }, setting the duration less than the second threshold to 0 (corresponding to a short beep), and setting the duration greater than or equal to the second threshold to 1 (corresponding to a long beep) to obtain the beep sequence.

[0055] In one embodiment, a second histogram of the durations of each historical audio interval in the historical data can be counted, and the segmentation point with the largest variance between two classes in the second histogram can be calculated through the Otsu algorithm to obtain the second threshold.

[0056] In the above embodiments, the amplitude values of the audio data are binarized to obtain binarized audio data, and according to the values in the binarized audio data, the audio intervals corresponding to multiple beeps are determined, so that the intervals with beeps in the audio data can be accurately segmented. Then, the duration of each audio interval is binarized to obtain the beep sequence, so that the long beeps and short beeps in the beeps can be accurately distinguished.

[0057] In one embodiment, identifying character information in the secondary screen image data includes: processing the secondary screen image data to obtain modulo data of the fault information; reconstructing the modulo data into a two-dimensional matrix; identifying the minimum bounding rectangles of each foreground element in the two-dimensional matrix; and matching the data in each minimum bounding rectangle with a standard character template to determine the character information in the secondary screen image data.

[0058] Among them, the modulo data is one-dimensional data. The two-dimensional matrix is composed of 0s and 1s, where 0 corresponds to the background and 1 corresponds to the foreground (i.e., character information). The foreground elements refer to the parts formed by the positions where 1s continuously exist in the two-dimensional matrix. The standard character template is also a matrix composed of 0s and 1s obtained after modulo operation.

[0059] In one embodiment, the modulo data can be reconstructed into a two-dimensional matrix according to the resolution of the secondary screen image (i.e., the number of pixel points in each row).

[0060] In one embodiment, the minimum bounding rectangle corresponding to the positions where 1s continuously exist (i.e., foreground elements) in the two-dimensional matrix is determined. Each minimum bounding rectangle is equivalent to separately segmenting each character in the secondary screen image data.

[0061] In one embodiment, in the process of matching the data within each minimum bounding rectangle with the standard character template, the minimum bounding rectangle and the standard character template can be uniformly scaled to a preset size, and binary hash codes corresponding to the data within the minimum bounding rectangle and the data of the standard character template are generated respectively. For each minimum bounding rectangle, count the number of different bits (i.e., Hamming distance) between the binary hash code of this minimum bounding rectangle and each standard character template respectively, perform normalization processing on the Hamming distance, and select the standard character template corresponding to the minimum value in the normalized Hamming distance as the character represented by the data within this minimum bounding rectangle. It can be understood that since the text sizes of the minimum bounding rectangle and the standard character template are different, the direct matching effect is not ideal. Therefore, in this embodiment, the hash algorithm is used for matching, which can more accurately achieve character matching.

[0062] In the above embodiment, the modulo data of the fault information is obtained by processing the secondary screen image data, the modulo data is reconstructed into a two-dimensional matrix, the minimum bounding rectangle of each foreground element in the two-dimensional matrix is identified, and the data within each minimum bounding rectangle is matched with the standard character template, so that each individual character in the secondary screen image data can be accurately determined.

[0063] In one embodiment, matching each standard fault code under the fault category with the character information in the secondary screen image data to determine the fault code in the secondary screen image data includes: sequentially selecting each standard fault code under the fault category in descending order of length; in each round of matching, sliding a sliding window with the same length as the currently selected standard fault code on the currently remaining unmatched character information in the secondary screen image data; respectively calculating the similarity degree between the characters in the sliding window and the currently selected standard fault code at each position where the sliding window slides; and determining whether the currently selected standard fault code exists in the secondary screen image data according to the similarity degrees at each position where the sliding window slides.

[0064] In one embodiment, at each position where the sliding window slides to, each character in the sliding window is matched bit by bit with each character in the currently selected standard fault code, and the similarity degree between the characters in the sliding window and the currently selected standard fault code is determined according to the matching results of each bit of characters.

[0065] In one embodiment, when matching each character in the sliding window bit by bit with each character in the currently selected standard fault code, the difference between the value (American Standard Code for Information Interchange) of each character in the sliding window and the character at the corresponding position in the currently selected standard fault code can be determined. According to the difference of the values corresponding to each bit of characters, the similarity degree between the characters in the sliding window and the currently selected standard fault code is determined. In one embodiment, the similarity degree between the characters in the sliding window and the currently selected standard fault code is negatively correlated with the sum of the differences of the values corresponding to each bit of characters.

[0066] In one embodiment, when the sum of the differences of the values corresponding to each bit of characters is 0, the similarity degree between the characters in the sliding window and the currently selected standard fault code is equal to 1. As the sum of the differences of the values corresponding to each bit of characters increases, the similarity degree between the characters in the sliding window and the currently selected standard fault code decays exponentially. In one embodiment, the similarity degree between the characters in the sliding window and the currently selected standard fault code can be calculated according to the following formula:

[0067] In one embodiment, where represents the similarity degree between the character at position a in the sliding window and the j-th standard fault code. represents the difference between the value of the z-th character at position a in the sliding window and the value of the z-th character in the j-th standard fault code.

[0068] In one embodiment, represents the number of characters in the j-th standard fault code. where represents the similarity degree between the character at position a in the sliding window and the j-th standard fault code. represents the value of the z-th character at position a in the sliding window and the value of the z-th character in the j-th standard fault code. represents the number of characters in the j-th standard fault code. represents the exponential function with base e.

[0069] In one embodiment, the similarity degree between the characters within the sliding window and the currently selected standard fault code can be normalized to obtain a normalized similarity degree. If the normalized similarity degree is equal to 1, the currently selected standard fault code exists at the corresponding position of the sliding window. If the normalized similarity degree is not equal to 1, the currently selected standard fault code does not exist at the corresponding position of the sliding window.

[0070] When matching the standard fault code, the distance between characters needs to be considered to avoid incorrect matching. For example: When matching the standard fault code "001F", it may happen that "00,1F" is matched. Therefore, to solve this problem, the distance between the windows enclosing each character in the above-mentioned sliding window should be less than a preset distance threshold, and then the characters in the sliding window can be considered to belong to the same character string. For example: The preset distance threshold can be set to 0.6.

[0071] In the above embodiment, the standard fault codes under the fault category are sequentially selected for matching in the order from longest to shortest, and the character information that has been successfully matched no longer participates in subsequent matching, so as to avoid incorrect matching caused by the inclusion relationship between standard fault codes. In each round of matching, a sliding window with the same length as the currently selected standard fault code is slid on the character information that has not been successfully matched in the current remaining sub-screen image data. At each position where the sliding window slides, the similarity degree between the characters in the sliding window and the currently selected standard fault code is calculated. According to the similarity degrees at each sliding position, it can be accurately determined whether the currently selected standard fault code exists in the sub-screen image data and its existence position.

[0072] In one embodiment, the fault detection result includes the fault category to which the memory module belongs; determining the fault detection result of the memory module according to the fault code includes: for each memory module, forming a fault code sequence from the respective fault codes in the sub-screen image data corresponding to the memory module in the order of appearance; clustering according to the fault code sequences corresponding to each memory module, and determining the fault category to which each memory module belongs according to the clustering result.

[0073] In one embodiment, clustering is performed according to the fault code sequences corresponding to each memory module to obtain a plurality of clustering clusters, and each clustering cluster corresponds to a fault category. The memory modules in the same clustering cluster are classified into the same fault category.

[0074] In one embodiment, clustering can be performed according to the DTW (Dynamic Time Warping) distance between the fault code sequences of each memory module.

[0075] In the above embodiments, for each memory module, the various fault codes in the secondary screen image data corresponding to the memory module are combined into a fault code sequence in the order of appearance, and clustering is performed according to the fault code sequences corresponding to the respective memory modules. According to the clustering results, it is possible to accurately determine the fault classes to which the respective memory modules belong, so that the memory modules can be accurately sorted, and the memory modules of the same fault class are placed at the same position.

[0076] In one embodiment, the fault detection result includes the main fault cause of the memory module; determining the fault detection result of the memory module according to the fault code includes: determining the fault information weight corresponding to each of the various fault codes in the secondary screen image data according to the appearance probability and context correlation degree of each of the various fault codes in the secondary screen image data at each occurrence position; selecting the main fault code from the various fault codes in the secondary screen image data according to the fault information weight, and determining the main fault cause of the memory module according to the main fault code.

[0077] In one embodiment, the fault information weight of the fault code is positively correlated with the appearance probability of the fault code at each occurrence position. The fault information weight of the fault code is positively correlated with the context correlation degree of the fault code at each occurrence position.

[0078] In one embodiment, for each type of fault code in the secondary screen image data, calculate the product of the appearance probability and the context correlation degree of the type of fault code at each occurrence position, and then sum the products corresponding to each occurrence position of the type of fault code to obtain the fault information weight of the type of fault code.

[0079] In one embodiment, the fault information weight corresponding to the j-th type of fault code in the secondary screen image data can be determined according to the following formula: where, represents the fault information weight corresponding to the j-th type of fault code in the secondary screen image data. represents the appearance probability of the j-th type of fault code in the secondary screen image data at the c-th occurrence position. represents the context correlation degree of the j-th type of fault code in the secondary screen image data at the c-th occurrence position. represents the total number of occurrences of the j-th type of fault code in the secondary screen image data.

[0080] In one embodiment, the fault code with the largest fault information weight can be determined as the main fault code, and the fault cause represented by the main fault code is determined as the main fault cause of the memory module.

[0081] In the above embodiments, since fault codes usually occur concentratedly, the fault information weight of a fault code can be more accurately measured according to the degree of context association at each occurrence position of the fault code. Combining with the occurrence probability of the fault code at each occurrence position, the fault information weights corresponding to various fault codes in the secondary screen image data can be accurately determined. According to the fault information weights, the main fault codes can be accurately selected from various fault codes in the secondary screen image data, so as to accurately determine the main fault causes of the memory module.

[0082] In one embodiment, the method further includes: determining a window corresponding to each fault code in the secondary screen image data at each occurrence position; the window is obtained by expanding a preset number of strings before and after the occurrence position of the fault code; determining the degree of context association of the fault code at each occurrence position according to the matching results between each string in the window corresponding to the fault code at each occurrence position and the standard fault code; determining the occurrence probability of the fault code at the occurrence position according to each occurrence position of the fault code and the number of times the fault code appears at the historical occurrence position.

[0083] For example: Taking the current fault code as the center, a window with a size of 5x1 is constructed, and the window contains 5 strings. That is, the window corresponding to the current fault code includes the current fault code, the two strings in front of the current fault code, and the two strings behind it.

[0084] In one embodiment, the matching result between the string and the standard fault code can be determined according to the normalized similarity degree between the string and the standard fault code. If the normalized similarity degree is equal to 1, the matching result is set to 1. If the normalized similarity degree is not equal to 1, the matching result is set to 0.

[0085] In one embodiment, the matching results between each string in the window corresponding to the fault code at each occurrence position and the standard fault code are summed to obtain the degree of context association of the fault code at each occurrence position. The formula is as follows: Among them, represents the degree of context association of the j-th fault code in the secondary screen image data at the c-th occurrence position. represents the -th string in the window corresponding to the j-th fault code in the secondary screen image data at the c-th occurrence position and the matching result with the standard fault code. represents the number of strings in the window corresponding to the j-th fault code in the secondary screen image data at the c-th occurrence position. represents the result of summing the matching results between each string in the window corresponding to the j-th fault code in the secondary screen image data at the c-th occurrence position and the standard fault code.

[0086] In one embodiment, the relative position of the fault code is determined according to the ratio of the line where the fault code appears to the total number of lines. According to the number of times the fault code appears on this line historically and the historical statistical quantity, the historical occurrence probability of the fault code at this occurrence position is determined. The occurrence probability of the fault code at this occurrence position is positively correlated with the relative position of the fault code. The occurrence probability of the fault code at this occurrence position is positively correlated with the historical occurrence probability.

[0087] In one embodiment, the occurrence probability of the fault code at this occurrence position is determined according to the product of the relative position of the fault code and the historical occurrence probability. The formula is as follows: where represents the occurrence probability of the fault code on the th line. represents the line where the occurrence position of the fault code is located. represents the total number of lines. represents the relative position of the fault code. represents the number of times the fault code appears on the th line historically. represents the historical statistical quantity. represents the occurrence probability of the fault code on the th line.

[0088] In the above embodiment, according to the matching results of each string in the window corresponding to each occurrence position of the fault code and the standard fault code, the context association degree of the fault code at each occurrence position can be accurately determined. According to each occurrence position of the fault code and the number of times the fault code appears at the historical occurrence position, the occurrence probability of the fault code at the occurrence position can be accurately determined.

[0089] In one embodiment, after matching the prompt tone sequence with the preset prompt tone sequence to determine the major fault category to which the memory module belongs, the method further includes: if the major fault category to which the memory module belongs is that the memory module is not connected, controlling the robotic arm to insert the contact pins of the memory module into the cleaning device; the cleaning device is used to clean the oxide layer on the contact pins by friction; controlling the robotic arm to take out the memory module from the cleaning device and re-insert it into the computer motherboard to trigger the computer motherboard to perform a self-check through the BIOS; returning to execute after the computer motherboard with the memory module inserted performs a self-check through the BIOS, obtaining the audio data of the prompt tone played after the self-check from the microphone device, and obtaining the secondary screen image data output after the self-check through the USB connection and subsequent steps.

[0090] Refer to Figure 5, which is a schematic structural diagram of a cleaning device. The cleaning device can be made of rubber material. There is a gap in the rubber material for inserting contact pins, and the oxide layer on the contact pins (i.e., "gold fingers") is cleaned by the friction of the rubber.

[0091] In the above embodiment, since the major category of faults to which the memory module belongs is that the memory module is not connected, there may be two situations: one is a memory module fault, and the other is that the memory module is not inserted properly or the contact pins of the memory module are oxidized. Therefore, by cleaning the contact pins of the memory module and then reinserting the cleaned memory module into the computer motherboard, the problem of false fault detection caused by the memory module not being inserted properly or the contact pins of the memory module being oxidized can be reduced. By using a robotic arm to pick up and insert the memory module, the detection efficiency can be improved. In addition, the cleaning device cleans the oxide layer of the contact pins through friction, with high efficiency and good cleaning effect.

[0092] Please refer to Figure 6 , which shows a schematic structural diagram of a fault detection system for a computer memory module provided by an embodiment of the present invention, including a memory and a processor; the memory is used to store executable program code; the processor is used to call and run the executable program code from the memory to implement the following steps: after the computer motherboard with the memory module inserted performs a self-check through the BIOS, obtain the audio data of the prompt sound played after the self-check from the microphone device, and obtain the secondary screen image data output after the self-check through a USB connection; identify the prompt sound sequence in the audio data; match the prompt sound sequence with a preset prompt sound sequence to determine the major category of faults to which the memory module belongs; identify the character information in the secondary screen image data; match each standard fault code under the major category of faults with the character information in the secondary screen image data respectively to determine the fault code in the secondary screen image data; determine the fault detection result of the memory module according to the fault code.

[0093] In one embodiment, the processor also implements the following steps: controlling the robotic arm to insert the memory module into the computer motherboard to trigger the computer motherboard to perform a self-check through the BIOS.

[0094] In one embodiment, the processor also implements the following steps: performing binarization processing on the amplitude values of the audio data to obtain binarized audio data; determining the audio intervals corresponding to multiple prompt sounds according to the values in the binarized audio data; performing binarization processing on the duration of each audio interval to obtain the prompt sound sequence.

[0095] In one embodiment, the processor also implements the following steps: processing the secondary screen image data to obtain modulo data of the fault information; reconstructing the modulo data into a two-dimensional matrix; identifying the minimum circumscribed rectangles of each foreground element in the two-dimensional matrix; matching the data in each minimum circumscribed rectangle with a standard character template to determine the character information in the secondary screen image data.

[0096] In one embodiment, the processor further implements the following steps: successively select each standard fault code under the major fault category in descending order of length; in each round of matching, slide a sliding window with the same length as the currently selected standard fault code over the remaining unmatched character information in the secondary screen image data; at each position where the sliding window slides to, calculate the similarity degree between the characters in the sliding window and the currently selected standard fault code; based on the similarity degrees at each sliding position, determine whether the currently selected standard fault code exists in the secondary screen image data.

[0097] In one embodiment, the fault detection result includes the fault category to which the memory module belongs; the processor further implements the following steps: for each memory module respectively, form a fault code sequence by arranging the various fault codes in the secondary screen image data corresponding to the memory module in the order of appearance; perform clustering based on the fault code sequences corresponding to each memory module, and determine the fault category to which each memory module belongs according to the clustering result.

[0098] In one embodiment, the fault detection result includes the main fault cause of the memory module; the processor further implements the following steps: determine the fault information weight corresponding to each of the various fault codes in the secondary screen image data according to the appearance probability and context correlation degree of each fault code at each appearance position; select the main fault code from the various fault codes in the secondary screen image data according to the fault information weight, and determine the main fault cause of the memory module according to the main fault code.

[0099] In one embodiment, the processor further implements the following steps: determine the window corresponding to each appearance position of each fault code in the secondary screen image data; the window is obtained by expanding a preset number of strings before and after the appearance position of the fault code; determine the context correlation degree of the fault code at each appearance position according to the matching result between each string in the window corresponding to each appearance position of the fault code and the standard fault code; determine the appearance probability of the fault code at the appearance position according to the appearance position of the fault code and the number of times the fault code appears at the historical appearance position.

[0100] In one embodiment, the processor further implements the following steps: If the major fault category to which the memory module belongs is that the memory module is not connected, control the robotic arm to insert the contact pins of the memory module into the cleaning device; the cleaning device is used to clean the oxide layer on the contact pins by friction; control the robotic arm to take out the memory module from the cleaning device and re-insert it into the computer motherboard to trigger the computer motherboard to perform a self-check through the BIOS; return to execute after the computer motherboard with the memory module inserted performs a self-check through the BIOS, obtain the audio data of the prompt sound played after the self-check from the microphone device, and obtain the sub-screen image data output after the self-check and subsequent steps through a USB connection.

[0101] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0102] The above-described embodiments merely represent several implementation manners of the present application. Their descriptions are relatively specific and detailed, but they should not be construed as limiting the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application.

[0103] It should be noted that the above sequence of the embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0104] Each embodiment in this specification is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for detecting faults in a computer memory module, characterized in that, The method includes: After the computer motherboard with the memory module inserted performs a self-check through the BIOS, obtain the audio data of the prompt sound played after the self-check from the microphone device, and obtain the secondary screen image data output after the self-check through a USB connection; Identify the prompt sound sequence in the audio data; Match the prompt sound sequence with a preset prompt sound sequence to determine the major fault category to which the memory module belongs; Identify the character information in the secondary screen image data; Match each standard fault code under the major fault category with the character information in the secondary screen image data respectively to determine the fault code in the secondary screen image data; Determine the fault detection result of the memory module according to the fault code.

2. The method for detecting a fault of a computer memory module according to claim 1, wherein Before obtaining the audio data of the prompt sound played after the self-check from the microphone device and obtaining the secondary screen image data output after the self-check through a USB connection, the method further includes: Control the robotic arm to insert the memory module into the computer motherboard to trigger the computer motherboard to perform a self-check through the BIOS.

3. The fault detection method of the computer memory module according to claim 1, wherein, The identifying the prompt sound sequence in the audio data includes: Perform binarization processing on the amplitude values of the audio data to obtain binarized audio data; Determine the audio intervals corresponding to multiple prompt sounds respectively according to the values in the binarized audio data; Perform binarization processing on the duration of each audio interval to obtain a prompt sound sequence.

4. The method for detecting a fault of a computer memory module according to claim 1, wherein, The identifying the character information in the secondary screen image data includes: Process the secondary screen image data to obtain the modulo data of the fault information; Reconstruct the modulo data into a two-dimensional matrix; Identify the minimum bounding rectangles of each foreground element in the two-dimensional matrix; Match the data within each minimum bounding rectangle with a standard character template to determine the character information in the secondary screen image data.

5. The method for detecting a fault of a computer memory module according to claim 1, wherein The matching each standard fault code under the major fault category with the character information in the secondary screen image data respectively to determine the fault code in the secondary screen image data includes: Select each standard fault code under the major fault category in order from the longest to the shortest; In each round of matching, slide a sliding window with the same length as the currently selected standard fault code on the currently remaining unmatched character information in the secondary screen image data; Calculate the similarity degree between the characters in the sliding window and the currently selected standard fault code respectively at each position where the sliding window slides to; Determine whether the currently selected standard fault code exists in the secondary screen image data according to the similarity degrees at each position where it slides to.

6. The method for detecting a fault of a computer memory module according to claim 1, wherein, The fault detection result includes the fault category to which the memory module belongs; The determining the fault detection result of the memory module according to the fault code includes: For each memory module respectively, form a fault code sequence by arranging the various fault codes in the secondary screen image data corresponding to the memory module in the order of appearance; Perform clustering according to the fault code sequences corresponding to each memory module, and determine the fault category to which each memory module belongs according to the clustering result.

7. The method for detecting a fault of a computer memory module according to claim 1, characterized in that, The fault detection result includes the main fault cause of the memory module; Determining the fault detection result of the memory module according to the fault code includes: Determining the fault information weights corresponding to various fault codes in the secondary screen image data respectively according to the occurrence probabilities and context correlation degrees of various fault codes in the secondary screen image data at each occurrence position; Selecting the main fault codes from various fault codes in the secondary screen image data according to the fault information weights, and determining the main fault causes of the memory module according to the main fault codes.

8. The method for detecting a fault of a computer memory module according to claim 7, characterized in that, The method further includes: Determining the window corresponding to each fault code in the secondary screen image data at each occurrence position; the window is obtained by expanding a preset number of strings before and after the occurrence position of the fault code; Determining the context correlation degree of the fault code at each occurrence position according to the matching results of each string in the window corresponding to the occurrence position of the fault code and the standard fault code; Determining the occurrence probability of the fault code at the occurrence position according to each occurrence position of the fault code and the number of times the fault code appears at the occurrence position in history.

9. The method for detecting a fault of a computer memory module according to any one of claims 1 to 8, characterized in that, After matching the prompt tone sequence with the preset prompt tone sequence to determine the major fault category to which the memory module belongs, the method further includes: If the major fault category to which the memory module belongs is that the memory module is not connected, controlling the robotic arm to insert the contact pins of the memory module into the cleaning device; the cleaning device is used to clean the oxide layer on the contact pins by friction; Controlling the robotic arm to take out the memory module from the cleaning device and re-insert it into the computer motherboard to trigger the computer motherboard to perform self-checking through the BIOS; Returning to execute obtaining the audio data of the prompt tone played after self-checking from the microphone device after the computer motherboard with the memory module inserted performs self-checking through the BIOS, and obtaining the secondary screen image data output after self-checking and subsequent steps through USB connection.

10. A fault detection system for a computer memory module, characterized in that, The system includes a memory and a processor; the memory is used to store executable program codes; the processor is used to call and run the executable program codes from the memory to implement the fault detection method of the computer memory module according to any one of claims 1 to 9.

Images