Tablet computer fault detection method and system

Abstract

The invention discloses a tablet computer fault detection method and system, and relates to the field of computer products, and the method comprises the following steps: constructing to obtain sample test data; obtaining at least one sample fault feature of the sample test data under the condition of the sample fault reason; obtaining suspected fault features corresponding to the sample fault reasons; actual fault features are obtained; obtaining suspected sample fault reasons, and taking sample test data corresponding to the suspected sample fault reasons as target sample test data; obtaining at least one actual detection feature; taking at least one sample fault feature of the target sample test data under the suspected sample fault reason condition as a target sample fault feature; and calculating the similarity between the actual detection feature and the fault feature of the at least one target sample. By arranging the data construction module, the feature acquisition module, the matching degree calculation module and the similarity calculation module, accurate determination of faults can be completed, and meanwhile, the detection time can be effectively controlled.

Description

Technical Field

[0001] The present invention relates to the field of computer products, and in particular to a method and system for detecting faults of a tablet computer. Background Art

[0002] As tablets become more popular, users' expectations for their performance and functionality continue to grow. However, these devices may be subject to a variety of technical glitches during daily use, which can sometimes prevent the device from functioning properly and cause inconvenience to users.

[0003] Since tablet computers have many fault conditions, it is difficult to accurately determine the cause of the fault when an abnormality occurs, which poses a great challenge to fault handling and prolongs the time for fault handling. Summary of the invention

[0004] In order to solve the above technical problems, a tablet computer fault detection method and system are provided. The technical solution solves the problem raised in the above background technology that due to the large number of fault conditions existing in tablet computers, it is difficult to accurately determine the cause of the fault when an abnormality occurs, which will pose a great challenge to fault handling and prolong the time of fault handling.

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

[0006] A tablet computer fault detection method, comprising:

[0007] Based on big data, obtain at least one sample failure cause of the tablet computer;

[0008] Based on the sample failure causes, sample test data is constructed, and the sample test data is paired with the sample failure causes;

[0009] Obtain at least one sample fault feature of the sample test data under the sample fault cause condition;

[0010] Based on big data, at least one suspected fault feature corresponding to the sample fault cause is obtained;

[0011] Performing fault detection on the tablet computer waiting to be detected to obtain at least one actual fault feature;

[0012] Calculating a matching degree between a suspected fault feature and at least one actual fault feature, and selecting a suspected fault feature whose matching degree is greater than a first preset value as a target suspected fault feature;

[0013] Screening the sample failure causes based on the target suspected failure characteristics to obtain the suspected sample failure causes, and using the sample test data corresponding to the suspected sample failure causes as the target sample test data;

[0014] Using the target sample test data to test the tablet computer, obtaining at least one actual detection feature, wherein the actual detection feature is the effect of the tablet computer on the target sample test data;

[0015] taking at least one sample fault feature of the target sample test data under the condition of the suspected sample fault cause as the target sample fault feature;

[0016] The similarity between the actual detection feature and at least one target sample fault feature is calculated. When the similarity is greater than a preset ratio, the suspected sample fault cause is taken as the actual fault cause.

[0017] Preferably, constructing sample test data based on the sample failure cause includes the following steps:

[0018] Selecting one of at least one sample failure cause as a calibration sample failure cause;

[0019] Acquire at least one characteristic data based on the big data;

[0020] Obtain the effect of the cause of the calibration sample failure in the characteristic data to obtain the calibration effect;

[0021] Obtain the effects of the remaining sample failure causes in the characteristic data and summarize them to obtain the non-calibrated effects;

[0022] Obtaining the target calibration effect, and satisfying the difference between the target calibration effect and the non-calibration effect;

[0023] When the target calibration effect cannot be obtained, at least one feature data is expanded based on big data until the target calibration effect can be found;

[0024] The difference between the target calibration effect and the non-calibration effect is taken as the target difference part;

[0025] The characteristic data corresponding to the target calibration effect is used as sample test data;

[0026] matching the sample test data to a sample failure cause that is consistent with the calibration sample failure cause among at least one sample failure cause, and pairing the target difference portion with the sample test data;

[0027] When the calibration sample failure causes traverse at least one sample failure cause, the construction is completed.

[0028] Preferably, the step of expanding at least one feature data comprises the following steps:

[0029] Acquire at least one piece of data to be supplemented, so that the piece of data to be supplemented changes when the cause of the sample failure changes;

[0030] Evenly divide the data to be supplemented into at least one local data;

[0031] The characteristic data are supplemented by partial data which are different from the at least one characteristic data.

[0032] Preferably, the step of obtaining at least one sample fault feature of the sample test data under the sample fault cause condition comprises the following steps:

[0033] Evenly divide the sample test data into at least one data block;

[0034] Evenly divide the target difference portion corresponding to the sample test data into at least one difference block, wherein the data block and the difference block have the same size;

[0035] Acquire a data block having data overlap with the difference block as a reference data block;

[0036] Aggregating at least one benchmark data block into a comprehensive data block;

[0037] At least one data feature in the target difference portion corresponding to the sample test data that is different from the integrated data block is used as a sample fault feature and matched to the sample test data.

[0038] Preferably, obtaining at least one suspected fault feature corresponding to the sample fault cause based on big data comprises the following steps:

[0039] Obtain at least one operating indicator of the tablet computer;

[0040] Based on the big data, an actual value of an operation indicator of at least one tablet computer having a sample failure cause is obtained;

[0041] Taking an average of at least one actual value of the operating indicator to obtain an actual average value;

[0042] Based on the big data, a baseline value of an operating indicator of at least one normally operating tablet computer is obtained;

[0043] Taking an average of at least one benchmark value of the operating indicator to obtain a benchmark average value;

[0044] The operating indicators whose actual average value differs from the benchmark average value by a larger margin than the preset value are regarded as suspected fault features.

[0045] Preferably, the performing fault detection on the tablet computer to be detected to obtain at least one actual fault feature comprises the following steps:

[0046] Obtaining a median of at least one value of an operating indicator of the tablet computer to be detected during operation as a value to be evaluated;

[0047] The operating indicators whose difference between the value to be evaluated and the benchmark average value is greater than the preset value are regarded as the actual fault characteristics.

[0048] Preferably, the calculating the matching degree between the suspected fault feature and at least one actual fault feature comprises the following steps:

[0049] Based on big data, the value range of suspected fault features is obtained as the feature range;

[0050] Obtain the fluctuation range of the actual fault characteristic corresponding to the same operating indicator as the suspected fault characteristic;

[0051] Obtain the overlapping part of the characteristic range and the fluctuation range as the overlapping range;

[0052] Using a matching degree formula, calculating a matching degree between a suspected fault feature and at least one actual fault feature;

[0053] The matching formula is as follows:

[0054]

[0055] Among them, A is the matching degree, a is the length of the overlapping range, b is the length of the characteristic range, and c is the length of the fluctuation range.

[0056] Preferably, the screening of the sample failure causes based on the target suspected failure characteristics to obtain the suspected sample failure causes comprises the following steps:

[0057] Obtaining the number of times at least one target suspected fault feature appears in at least one suspected fault feature corresponding to the sample fault cause as a feature value;

[0058] The characteristic value is divided by the total number of at least one suspected fault characteristic corresponding to the sample fault cause to obtain a screening value of the sample fault cause;

[0059] When the screening value of the sample fault cause is greater than the sample fault cause of the second preset value, it is regarded as a suspected sample fault cause.

[0060] Preferably, the calculating of the similarity between the actual detection feature and at least one target sample fault feature comprises the following steps:

[0061] Based on big data, the value range of the target sample fault characteristics is obtained as the target range;

[0062] Obtain the value range of the actual detection feature during the test as the actual measurement range;

[0063] Pairing actual detection features and target sample fault features that appear at the same position in the target sample test data;

[0064] The length of the target range is divided by the length of the measured range to obtain the local similarity between the paired actual detection features and the target sample fault features;

[0065] The local similarity of at least one target sample fault feature corresponding to the suspected sample fault cause is superimposed to obtain the similarity between the actual detection feature and the at least one target sample fault feature.

[0066] A tablet computer fault detection system, used to implement the above tablet computer fault detection method, comprises:

[0067] A fault acquisition module, wherein the fault acquisition module acquires at least one sample fault cause of the tablet computer based on big data;

[0068] A data construction module, wherein the data construction module constructs sample test data based on the sample failure cause and pairs the sample test data with the sample failure cause;

[0069] A feature acquisition module, wherein the feature acquisition module acquires at least one sample fault feature of the sample test data under the sample fault cause condition, acquires at least one suspected fault feature corresponding to the sample fault cause based on big data, performs fault detection on the tablet computer to be detected, and obtains at least one actual fault feature;

[0070] A matching degree calculation module, wherein the matching degree calculation module calculates a matching degree between a suspected fault feature and at least one actual fault feature, and selects a suspected fault feature whose matching degree is greater than a first preset value as a target suspected fault feature;

[0071] A data screening module, wherein the data screening module screens the sample fault causes based on the target suspected fault characteristics to obtain the suspected sample fault causes, and uses the sample test data corresponding to the suspected sample fault causes as the target sample test data;

[0072] A data testing module, wherein the data testing module uses the target sample test data to test the tablet computer to obtain actual detection features, and uses at least one sample failure feature of the target sample test data under the condition of the suspected sample failure cause as the target sample failure feature;

[0073] A similarity calculation module calculates the similarity between the actual detection feature and at least one target sample fault feature, and when the similarity is greater than a preset ratio, the suspected sample fault cause is taken as the actual fault cause.

[0074] Compared with the prior art, the present invention has the following beneficial effects:

[0075] By setting up a data construction module, a feature acquisition module, a matching degree calculation module and a similarity calculation module, corresponding sample test data can be generated for each sample failure cause for testing, and different detection features can be formed for each sample failure cause, so that it is easy to determine the actual failure cause. At the same time, in order to reduce the workload of detection, the suspected sample failure cause is determined according to the actual failure characteristics, so that only the target sample test data corresponding to the suspected sample failure cause is used for testing, without using all the sample test data, which can not only complete the accurate determination of the fault, but also effectively control the detection time. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 It is a flow chart of a method for detecting a fault of a tablet computer according to the present invention;

[0077] Figure 2 A schematic diagram of a flow chart of constructing and obtaining sample test data based on sample failure causes of the present invention;

[0078] Figure 3 A schematic diagram of a process of expanding at least one characteristic data according to the present invention;

[0079] Figure 4 A schematic diagram of a flow chart of obtaining at least one sample fault feature of sample test data under a sample fault cause condition according to the present invention;

[0080] Figure 5 A schematic diagram of a process of obtaining at least one suspected fault feature corresponding to a sample fault cause based on big data according to the present invention;

[0081] Figure 6 A schematic diagram of a flow chart of performing fault detection on a tablet computer waiting to be detected and obtaining at least one actual fault feature according to the present invention;

[0082] Figure 7 A schematic diagram of a flow chart of calculating a matching degree between a suspected fault feature and at least one actual fault feature according to the present invention;

[0083] Figure 8 A schematic diagram of a process of screening sample failure causes based on target suspected failure features to obtain suspected sample failure causes in the present invention;

[0084] Fig. 9 The figure is a flow chart of calculating the similarity between the actual detection feature and at least one target sample fault feature according to the present invention. DETAILED DESCRIPTION

[0085] 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.

[0086] Reference Figure 1 As shown, a tablet computer fault detection method includes:

[0087] Based on big data, obtain at least one sample failure cause of the tablet computer;

[0088] Based on the sample failure causes, sample test data is constructed, and the sample test data is paired with the sample failure causes;

[0089] Obtain at least one sample fault feature of the sample test data under the sample fault cause condition;

[0090] Based on big data, at least one suspected fault feature corresponding to the sample fault cause is obtained;

[0091] Performing fault detection on the tablet computer waiting to be detected to obtain at least one actual fault feature;

[0092] Calculating a matching degree between a suspected fault feature and at least one actual fault feature, and selecting a suspected fault feature whose matching degree is greater than a first preset value as a target suspected fault feature;

[0093] Screening the sample failure causes based on the target suspected failure characteristics to obtain the suspected sample failure causes, and using the sample test data corresponding to the suspected sample failure causes as the target sample test data;

[0094] Using the target sample test data to test the tablet computer, obtaining at least one actual detection feature, wherein the actual detection feature is the effect of the tablet computer on the target sample test data;

[0095] taking at least one sample fault feature of the target sample test data under the condition of the suspected sample fault cause as the target sample fault feature;

[0096] The similarity between the actual detection feature and at least one target sample fault feature is calculated. When the similarity is greater than a preset ratio, the suspected sample fault cause is taken as the actual fault cause.

[0097] Since the sample failure causes will produce different effects on different data, based on this, in this solution, sample test data is constructed, and the sample test data is specially set for the sample failure causes. There is a one-to-one correspondence between the sample test data and the sample failure causes. At the same time, the effect of the sample failure cause on the sample test data is different from the effect of the other sample failure causes on the sample test data. This difference is the basis for distinguishing the sample failure causes;

[0098] However, when using this method, it is necessary to use sample test data corresponding to the sample failure cause to test the tablet computer. In order to avoid this situation, the suspected sample failure cause is identified to narrow the scope of detection. The principle is: the tablet computer waiting to be tested only has its own data and architecture, and no sample test data. Therefore, when a failure occurs, there will be no features corresponding to the sample test data. However, the failure of the tablet computer's own data and architecture can be matched with the sample failure cause, that is, at least one suspected failure feature corresponding to the sample failure cause is obtained. Therefore, when at least one actual failure feature of the tablet computer waiting to be tested is obtained, the possible sample failure causes of the tablet computer waiting to be tested can be determined based on the comparison between the suspected failure feature and the actual failure feature, and these sample failure causes can be tested using the corresponding sample test data without testing all the sample test data, thereby achieving the effect of reducing the workload.

[0099] Reference Figure 2 As shown, based on the sample failure cause, constructing sample test data includes the following steps:

[0100] Selecting one of at least one sample failure cause as a calibration sample failure cause;

[0101] Acquire at least one characteristic data based on the big data;

[0102] Obtain the effect of the cause of the calibration sample failure in the characteristic data to obtain the calibration effect;

[0103] Obtain the effects of the remaining sample failure causes in the characteristic data and summarize them to obtain the non-calibrated effects;

[0104] Obtaining the target calibration effect, and satisfying the difference between the target calibration effect and the non-calibration effect;

[0105] When the target calibration effect cannot be obtained, at least one feature data is expanded based on big data until the target calibration effect can be found;

[0106] The difference between the target calibration effect and the non-calibration effect is taken as the target difference part;

[0107] The characteristic data corresponding to the target calibration effect is used as sample test data;

[0108] matching the sample test data to a sample failure cause that is consistent with the calibration sample failure cause among at least one sample failure cause, and pairing the target difference portion with the sample test data;

[0109] When the calibration sample failure causes traverse at least one sample failure cause, the construction is completed.

[0110] When constructing sample test data, the key is to ensure that the effect of the sample test data under the sample failure cause condition is different from the effect of the sample test data under the other sample failure cause conditions. Therefore, the characteristics of the difference can be used as the basis for identifying the sample failure cause. However, it is not certain that data that meets this requirement can be found in at least one characteristic data. Therefore, when such data cannot be found, data supplementation is required. When supplementing, the data to be supplemented is limited, so that the efficiency of data supplementation is guaranteed, and at least one characteristic data that meets the requirement can be supplemented more quickly.

[0111] Reference Figure 3 As shown, expanding at least one feature data includes the following steps:

[0112] Acquire at least one piece of data to be supplemented, so that the piece of data to be supplemented changes when the cause of the sample failure changes;

[0113] Evenly divide the data to be supplemented into at least one local data;

[0114] The characteristic data are supplemented by partial data which are different from the at least one characteristic data.

[0115] Reference Figure 4 As shown, obtaining at least one sample fault feature of sample test data under the sample fault cause condition includes the following steps:

[0116] Evenly divide the sample test data into at least one data block;

[0117] Evenly divide the target difference portion corresponding to the sample test data into at least one difference block, wherein the data block and the difference block have the same size;

[0118] Acquire a data block having data overlap with the difference block as a reference data block;

[0119] Aggregating at least one benchmark data block into a comprehensive data block;

[0120] At least one data feature in the target difference portion corresponding to the sample test data that is different from the integrated data block is used as a sample fault feature and matched to the sample test data.

[0121] The target difference part is the data that has changed due to the fault. Therefore, when comparing, in order to obtain the corresponding features, it is necessary to compare the data at the same position. Therefore, the data is divided into data blocks and difference blocks. When the positions of the data blocks and the difference blocks are the same, they must correspond to the same data. It is just that some data in the fault difference blocks has changed. Therefore, when there is data overlap between the data blocks and the difference blocks, it means that the two are in the same position. Based on this, a comprehensive data block is obtained. The comprehensive data block corresponds to the target difference part. Therefore, by comparing the two, the required sample fault features can be obtained.

[0122] Reference Figure 5 As shown, based on big data, obtaining at least one suspected fault feature corresponding to the sample fault cause includes the following steps:

[0123] Obtain at least one operating indicator of the tablet computer;

[0124] Based on the big data, an actual value of an operation indicator of at least one tablet computer having a sample failure cause is obtained;

[0125] Taking an average of at least one actual value of the operating indicator to obtain an actual average value;

[0126] Based on the big data, a baseline value of an operating indicator of at least one normally operating tablet computer is obtained;

[0127] Taking an average of at least one benchmark value of the operating indicator to obtain a benchmark average value;

[0128] The operating indicators whose actual average value differs from the benchmark average value by a larger margin than the preset value are regarded as suspected fault features.

[0129] The acquisition of suspected fault characteristics cannot be based on sample test data, but can only be based on the existing data or architecture of the tablet computer itself. Therefore, at least one corresponding operating indicator is obtained, and a connection is established between these operating indicators and the causes of sample failures. In actual testing, the causes of sample failures that are ultimately involved in the test are also determined through the operating indicators, thereby narrowing the scope of detection.

[0130] Reference Figure 6 As shown, performing fault detection on a tablet computer waiting to be detected to obtain at least one actual fault feature includes the following steps:

[0131] Obtaining a median of at least one value of an operating indicator of the tablet computer to be detected during operation as a value to be evaluated;

[0132] The operating indicators whose difference between the value to be evaluated and the benchmark average value is greater than the preset value are regarded as the actual fault characteristics.

[0133] During actual detection, the value of the operating indicator is not fixed, therefore, at least one value of the operating indicator is obtained intermittently, and the median thereof is used to screen the operating indicator.

[0134] Reference Figure 7 As shown, calculating the matching degree between the suspected fault feature and at least one actual fault feature includes the following steps:

[0135] Based on big data, the value range of suspected fault features is obtained as the feature range;

[0136] Obtain the fluctuation range of the actual fault characteristic corresponding to the same operating indicator as the suspected fault characteristic;

[0137] Obtain the overlapping part of the characteristic range and the fluctuation range as the overlapping range;

[0138] Using a matching degree formula, calculating a matching degree between a suspected fault feature and at least one actual fault feature;

[0139] The matching formula is as follows:

[0140]

[0141] Among them, A is the matching degree, a is the length of the overlapping range, b is the length of the characteristic range, and c is the length of the fluctuation range.

[0142] Since the values ​​of the suspected fault characteristics and the actual fault characteristics are both range values, there will inevitably be certain differences. In actual measurement, individual extreme values ​​may appear. In order to avoid this situation, the proportion of the length of the overlapping range is calculated and the judgment is made based on the calculated results.

[0143] Reference Figure 8 As shown, screening the sample fault causes based on the target suspected fault characteristics to obtain the suspected sample fault causes includes the following steps:

[0144] Obtaining the number of times at least one target suspected fault feature appears in at least one suspected fault feature corresponding to the sample fault cause as a feature value;

[0145] The characteristic value is divided by the total number of at least one suspected fault characteristic corresponding to the sample fault cause to obtain a screening value of the sample fault cause;

[0146] When the screening value of the sample fault cause is greater than the sample fault cause of the second preset value, it is regarded as a suspected sample fault cause.

[0147] There are multiple suspected fault features corresponding to the sample fault cause, but in actual situations, not all of the suspected fault features corresponding to the sample fault cause may appear. In order to avoid omissions in detection, when the proportion of suspected fault features corresponding to the sample fault cause is large enough, the sample fault cause is likely to appear, and therefore, the sample fault cause is detected.

[0148] Reference Fig. 9 As shown, calculating the similarity between the actual detection feature and at least one target sample fault feature includes the following steps:

[0149] Based on big data, the value range of the target sample fault characteristics is obtained as the target range;

[0150] Obtain the value range of the actual detection feature during the test as the actual measurement range;

[0151] Pairing actual detection features and target sample fault features that appear at the same position in the target sample test data;

[0152] The length of the target range is divided by the length of the measured range to obtain the local similarity between the paired actual detection features and the target sample fault features;

[0153] The local similarity of at least one target sample fault feature corresponding to the suspected sample fault cause is superimposed to obtain the similarity between the actual detection feature and the at least one target sample fault feature.

[0154] When comparing the actual detection features with the target sample fault features, it is necessary to ensure that they are corresponding features, and the basis for judgment is the position where the actual detection features and the target sample fault features appear. When their relative positions are consistent, it means that the two are in a corresponding relationship. Therefore, a comparison can be performed, and when comparing, the comparison results of at least one target sample fault feature of the target sample test data are integrated to avoid the impact of extreme situations during a single comparison.

[0155] A tablet computer fault detection system, used to implement the above tablet computer fault detection method, comprises:

[0156] A fault acquisition module, wherein the fault acquisition module acquires at least one sample fault cause of the tablet computer based on big data;

[0157] A data construction module, wherein the data construction module constructs sample test data based on the sample failure cause and pairs the sample test data with the sample failure cause;

[0158] A feature acquisition module, wherein the feature acquisition module acquires at least one sample fault feature of the sample test data under the sample fault cause condition, acquires at least one suspected fault feature corresponding to the sample fault cause based on big data, performs fault detection on the tablet computer to be detected, and obtains at least one actual fault feature;

[0159] A matching degree calculation module, wherein the matching degree calculation module calculates a matching degree between a suspected fault feature and at least one actual fault feature, and selects a suspected fault feature whose matching degree is greater than a first preset value as a target suspected fault feature;

[0160] A data screening module, wherein the data screening module screens the sample fault causes based on the target suspected fault characteristics to obtain the suspected sample fault causes, and uses the sample test data corresponding to the suspected sample fault causes as the target sample test data;

[0161] A data testing module, wherein the data testing module uses the target sample test data to test the tablet computer to obtain actual detection features, and uses at least one sample failure feature of the target sample test data under the condition of the suspected sample failure cause as the target sample failure feature;

[0162] A similarity calculation module calculates the similarity between the actual detection feature and at least one target sample fault feature, and when the similarity is greater than a preset ratio, the suspected sample fault cause is taken as the actual fault cause.

[0163] Furthermore, the present solution also proposes a storage medium on which a computer-readable program is stored, and when the computer-readable program is called, the above-mentioned tablet computer fault detection method is executed.

[0164] It is understandable that the storage medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid state drive (SSD).

[0165] To sum up, the advantages of the present invention are: by setting up a data construction module, a feature acquisition module, a matching degree calculation module and a similarity calculation module, corresponding sample test data can be generated for each sample fault cause for testing, and different detection features can be formed for each sample fault cause, so that it is easy to determine the actual fault cause. At the same time, in order to reduce the workload of detection, the suspected sample fault cause is determined according to the actual fault characteristics, so that only the target sample test data corresponding to the suspected sample fault cause is used for testing, without using all the sample test data, which can not only complete the accurate determination of the fault, but also effectively control the detection time.

[0166] 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 tablet computer fault detection method, characterized in that: include: Based on big data, obtain at least one sample failure cause of the tablet computer; Based on the sample failure causes, sample test data is constructed, and the sample test data is paired with the sample failure causes; Obtain at least one sample fault feature of the sample test data under the sample fault cause condition; Based on big data, at least one suspected fault feature corresponding to the sample fault cause is obtained; Performing fault detection on the tablet computer waiting to be detected to obtain at least one actual fault feature; Calculating a matching degree between a suspected fault feature and at least one actual fault feature, and selecting a suspected fault feature whose matching degree is greater than a first preset value as a target suspected fault feature; Screening the sample failure causes based on the target suspected failure characteristics to obtain the suspected sample failure causes, and using the sample test data corresponding to the suspected sample failure causes as the target sample test data; Using the target sample test data to test the tablet computer, obtaining at least one actual detection feature, wherein the actual detection feature is the effect of the tablet computer on the target sample test data; taking at least one sample fault feature of the target sample test data under the condition of the suspected sample fault cause as the target sample fault feature; The similarity between the actual detection feature and at least one target sample fault feature is calculated. When the similarity is greater than a preset ratio, the suspected sample fault cause is taken as the actual fault cause.

2. A tablet computer fault detection method according to claim 1, characterized in that: The construction of sample test data based on the sample failure cause includes the following steps: Selecting one of at least one sample failure cause as a calibration sample failure cause; Acquire at least one characteristic data based on the big data; Obtain the effect of the cause of the calibration sample failure in the characteristic data to obtain the calibration effect; Obtain the effects of the remaining sample failure causes in the characteristic data and summarize them to obtain the non-calibrated effects; Obtaining the target calibration effect, and satisfying the difference between the target calibration effect and the non-calibration effect; When the target calibration effect cannot be obtained, at least one feature data is expanded based on big data until the target calibration effect can be found; The difference between the target calibration effect and the non-calibration effect is taken as the target difference part; The characteristic data corresponding to the target calibration effect is used as sample test data; matching the sample test data to a sample failure cause that is consistent with the calibration sample failure cause among at least one sample failure cause, and pairing the target difference portion with the sample test data; When the calibration sample failure causes traverse at least one sample failure cause, the construction is completed.

3. A tablet computer fault detection method according to claim 2, characterized in that: The expanding of at least one feature data comprises the following steps: Acquire at least one piece of data to be supplemented, so that the piece of data to be supplemented changes when the cause of the sample failure changes; Evenly divide the data to be supplemented into at least one local data; The characteristic data are supplemented by partial data which are different from the at least one characteristic data.

4. A tablet computer fault detection method according to claim 3, characterized in that: The step of obtaining at least one sample fault feature of the sample test data under the sample fault cause condition comprises the following steps: Evenly divide the sample test data into at least one data block; Evenly divide the target difference portion corresponding to the sample test data into at least one difference block, wherein the data block and the difference block have the same size; Acquire a data block having data overlap with the difference block as a reference data block; Aggregating at least one benchmark data block into a comprehensive data block; At least one data feature in the target difference portion corresponding to the sample test data that is different from the integrated data block is used as a sample fault feature and matched to the sample test data.

5. A tablet computer fault detection method according to claim 4, characterized in that: The method of obtaining at least one suspected fault feature corresponding to the sample fault cause based on big data includes the following steps: Obtain at least one operating indicator of the tablet computer; Based on the big data, an actual value of an operation indicator of at least one tablet computer having a sample failure cause is obtained; Taking an average of at least one actual value of the operating indicator to obtain an actual average value; Based on the big data, a baseline value of an operating indicator of at least one normally operating tablet computer is obtained; Taking an average of at least one benchmark value of the operating indicator to obtain a benchmark average value; The operating indicators whose actual average value differs from the benchmark average value by a larger margin than the preset value are regarded as suspected fault features.

6. A tablet computer fault detection method according to claim 5, characterized in that: The performing fault detection on the tablet computer to be detected to obtain at least one actual fault feature comprises the following steps: Obtaining a median of at least one value of an operating indicator of the tablet computer to be detected during operation as a value to be evaluated; The operating indicators whose difference between the value to be evaluated and the benchmark average value is greater than the preset value are regarded as the actual fault characteristics.

7. A tablet computer fault detection method according to claim 6, characterized in that: The calculating of the matching degree between the suspected fault feature and at least one actual fault feature comprises the following steps: Based on big data, the value range of suspected fault features is obtained as the feature range; Obtain the fluctuation range of the actual fault characteristic corresponding to the same operating indicator as the suspected fault characteristic; Obtain the overlapping part of the characteristic range and the fluctuation range as the overlapping range; Using a matching degree formula, calculating a matching degree between a suspected fault feature and at least one actual fault feature; The matching formula is as follows: Among them, A is the matching degree, a is the length of the overlapping range, b is the length of the characteristic range, and c is the length of the fluctuation range.

8. A tablet computer fault detection method according to claim 7, characterized in that: The method of screening the sample fault causes based on the target suspected fault characteristics to obtain the suspected sample fault causes includes the following steps: Obtaining the number of times at least one target suspected fault feature appears in at least one suspected fault feature corresponding to the sample fault cause as a feature value; The characteristic value is divided by the total number of at least one suspected fault characteristic corresponding to the sample fault cause to obtain a screening value of the sample fault cause; When the screening value of the sample fault cause is greater than the sample fault cause of the second preset value, it is regarded as a suspected sample fault cause.

9. A tablet computer fault detection method according to claim 8, characterized in that: The calculating of the similarity between the actual detection feature and at least one target sample fault feature comprises the following steps: Based on big data, the value range of the target sample fault characteristics is obtained as the target range; Obtain the value range of the actual detection feature during the test as the actual measurement range; Pairing actual detection features and target sample fault features that appear at the same position in the target sample test data; The length of the target range is divided by the length of the measured range to obtain the local similarity between the paired actual detection features and the target sample fault features; The local similarity of at least one target sample fault feature corresponding to the suspected sample fault cause is superimposed to obtain the similarity between the actual detection feature and the at least one target sample fault feature.

10. A tablet computer fault detection system, used to implement the tablet computer fault detection method according to any one of claims 1 to 9, characterized in that: include: A fault acquisition module, wherein the fault acquisition module acquires at least one sample fault cause of the tablet computer based on big data; A data construction module, wherein the data construction module constructs sample test data based on the sample failure cause and pairs the sample test data with the sample failure cause; A feature acquisition module, wherein the feature acquisition module acquires at least one sample fault feature of the sample test data under the sample fault cause condition, acquires at least one suspected fault feature corresponding to the sample fault cause based on big data, performs fault detection on the tablet computer to be detected, and obtains at least one actual fault feature; A matching degree calculation module, wherein the matching degree calculation module calculates a matching degree between a suspected fault feature and at least one actual fault feature, and selects a suspected fault feature whose matching degree is greater than a first preset value as a target suspected fault feature; A data screening module, wherein the data screening module screens the sample fault causes based on the target suspected fault characteristics to obtain the suspected sample fault causes, and uses the sample test data corresponding to the suspected sample fault causes as the target sample test data; A data testing module, wherein the data testing module uses the target sample test data to test the tablet computer to obtain actual detection features, and uses at least one sample failure feature of the target sample test data under the condition of the suspected sample failure cause as the target sample failure feature; A similarity calculation module calculates the similarity between the actual detection feature and at least one target sample fault feature, and when the similarity is greater than a preset ratio, the suspected sample fault cause is taken as the actual fault cause.

Images