Method and device for generating fault rate estimation model, electronic equipment and medium

Abstract

The invention relates to a method and device for generating a failure rate estimation model, electronic equipment and a medium, and aims to improve the accuracy of failure rate estimation of a display screen cover plate. The method comprises the following steps: obtaining respective training samples of a plurality of historical terminal models, wherein the training sample of each historical terminal model comprises influence factor values of the historical terminal models under a plurality of preset dimensions and actual failure rates of display screen cover plates of the historical terminal models; determining a target dimension of a to-be-estimated terminal model according to the training sample and the fault rate estimation model framework; and according to the target dimension, the influence factor values of the plurality of historical terminal models under the target dimension, the actual fault rates of the display screen cover plates of the historical terminal models and a fault rate estimation model framework, obtaining a target fault rate estimation model, the target fault rate estimation model being used for estimating the fault rate of the display screen cover plate of the terminal model to be estimated.

Description

Technical Field

[0001] The present disclosure relates to the field of terminal technology, and in particular to a method, device, electronic device and medium for generating a failure rate prediction model. Background Art

[0002] Display cover failure, such as cover breakage, is one of the main terminal failures. In the terminal design stage, accurate prediction of cover failure rate facilitates the design of terminals with low failure rates. However, due to the wide range of terminal users and complex user habits, cover breakage is highly random, resulting in accurate prediction of cover failure rate has always been a difficult problem in the industry.

[0003] At present, the main method is to refer to the data of historical terminal models, rely on business experience to determine key factors, assign weights to each key factor, and make estimates through simple formulas. For example, the method of determining weights based on expert experience determines a weight coefficient for each factor through the subjective experience of experts, and then multiplies the coefficient with the factor and sums them to obtain the estimated value of the failure rate. Such estimates are not based on scientific statistical methods, and are often less accurate and less reliable. Summary of the invention

[0004] In order to overcome the problems existing in the related art, the present disclosure provides a method, device, electronic device and medium for generating a failure rate prediction model.

[0005] According to a first aspect of an embodiment of the present disclosure, a method for generating a failure rate prediction model is provided, the method comprising:

[0006] Acquire training samples of each of a plurality of historical terminal models, wherein the training samples of each of the historical terminal models include an influence factor value of the historical terminal model under a plurality of preset dimensions and an actual failure rate of a display screen cover of the historical terminal model, wherein the plurality of preset dimensions refer to dimensions that affect the failure rate of the display screen cover of the historical terminal model;

[0007] Determining the target dimension of the terminal model to be estimated based on the training samples and the failure rate estimation model framework;

[0008] According to the target dimension, the influence factor values ​​of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal models and the failure rate prediction model framework, a target failure rate prediction model is obtained. The target failure rate prediction model is used to estimate the failure rate of the display screen cover of the terminal model to be estimated.

[0009] Optionally, determining the target dimension of the terminal model to be estimated according to the training samples and the failure rate estimation model framework includes:

[0010] Determining each historical terminal model as a target historical terminal model in turn;

[0011] For each of the target historical terminal models, a failure rate prediction model of the target historical terminal model is generated based on training samples of other historical terminal models except the target historical terminal model and a failure rate prediction model framework; a prediction error of the target historical terminal model is determined based on the failure rate prediction model of the target historical terminal model and the training samples of the target historical terminal model;

[0012] If it is determined according to the prediction error of each of the historical terminal models that the preset termination condition is not satisfied, the multiple preset dimensions are updated and the step of determining the prediction error of the target historical terminal model is re-executed, and the preset dimensions when it is determined according to the prediction error of each of the historical terminal models that the termination condition is satisfied are determined as the target dimensions of the terminal model to be estimated.

[0013] Optionally, generating the failure rate prediction model of the target historical terminal model according to the training samples and failure rate prediction model framework of other historical terminal models except the target historical terminal model includes:

[0014] Determine a first target dimension from the multiple preset dimensions according to the impact factor values ​​of other historical terminal models except the target historical terminal model in multiple preset dimensions and the actual failure rates of the display screen covers of the other historical terminal models;

[0015] Generate an initial failure rate prediction model corresponding to the target historical terminal model according to the first target dimension and the failure rate prediction model framework;

[0016] Generate a failure rate estimation model for the target historical terminal model according to the impact factor values ​​of other historical terminal models except the target historical terminal model under the first target dimension, the actual failure rates of the display screen covers of the other historical terminal models, and the initial failure rate estimation model corresponding to the target historical terminal model;

[0017] The step of determining the prediction error of the target historical terminal model according to the failure rate prediction model of the target historical terminal model and the training sample of the target historical terminal model includes:

[0018] Determining an estimated failure rate of a display screen cover of a target historical terminal model according to a failure rate estimation model of the target historical terminal model and an impact factor value of the target historical terminal model under the first target dimension;

[0019] The prediction error of the target historical terminal model is determined based on the estimated failure rate and the actual failure rate.

[0020] Optionally, determining the first target dimension from the multiple preset dimensions according to the impact factor values ​​of other historical terminal models except the target historical terminal model in multiple preset dimensions and the actual failure rates of the display screen covers of the other historical terminal models includes:

[0021] Determine a first parameter group of an initial failure rate prediction model corresponding to the target historical terminal model according to the influence factor values ​​of other historical terminal models other than the target historical terminal model under multiple preset dimensions, the actual failure rates of the display screen covers of the other historical terminal models, and the failure rate prediction model framework, wherein the first parameter group includes first coefficients corresponding to the influence factor values ​​under the multiple preset dimensions;

[0022] The dimension of the impact factor value corresponding to the first coefficient whose absolute value is greater than the first threshold is determined as the first target dimension.

[0023] Optionally, the determining of the first parameter group of the initial failure rate prediction model corresponding to the target historical terminal model according to the influence factor values ​​of each of the other historical terminal models except the target historical terminal model under multiple preset dimensions, the actual failure rate of the display screen cover of each of the other historical terminal models, and the failure rate prediction model framework includes:

[0024] Repeat the step of obtaining the second parameter set a preset number of times, where the preset number is an integer greater than 1:

[0025] Randomly select N-1 historical terminal models from other historical terminal models except the target historical terminal model to form a historical terminal model sequence, where N is the total number of the historical terminal models;

[0026] Determine a second parameter group of the initial failure rate prediction model corresponding to the target historical terminal model according to the influence factor values ​​of each historical terminal model in the historical terminal model sequence under multiple preset dimensions, the actual failure rate of the respective display screen cover plates, and the failure rate prediction model framework;

[0027] According to the preset number of second parameter groups and the preset confidence interval of the target historical terminal model, a first parameter group of the initial failure rate prediction model corresponding to the target historical terminal model is determined.

[0028] Optionally, obtaining the target failure rate prediction model according to the target dimension, the impact factor values ​​of each of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal model and the failure rate prediction model framework includes:

[0029] Determining an initial failure rate prediction model corresponding to the terminal model to be estimated according to the target dimension and the failure rate prediction model framework;

[0030] Determine a third parameter group of the initial failure rate estimation model corresponding to the terminal model to be estimated according to the influence factor values ​​of each of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal model, and the initial failure rate estimation model;

[0031] A target failure rate prediction model is obtained according to the third parameter group and the initial failure rate prediction model.

[0032] Optionally, the determining the target dimension of the terminal model to be estimated according to the training samples and the failure rate estimation model framework further includes:

[0033] Determining a maximum value among the prediction errors of a plurality of the historical terminal models;

[0034] If the maximum value is less than a preset threshold, it is determined that a preset termination condition is met;

[0035] If the maximum value is greater than or equal to the preset threshold, it is determined that the termination condition is not met.

[0036] Optionally, the determining the target dimension of the terminal model to be estimated according to the training samples and the failure rate estimation model framework further includes:

[0037] If the number of updates to the multiple preset dimensions reaches the preset number and does not meet the preset termination condition, the maximum value of the prediction errors of multiple historical terminal models after each update of the preset dimensions is obtained, and the target update number corresponding to the minimum value of the multiple maximum values ​​and the preset dimension corresponding to the target update number are determined as the target dimension of the terminal model to be estimated.

[0038] Optionally, the method further comprises:

[0039] If the preset termination condition is not met, the preset number and / or the confidence interval are adjusted.

[0040] According to a second aspect of an embodiment of the present disclosure, there is provided a device for generating a failure rate prediction model, the device comprising:

[0041] An acquisition module is configured to acquire training samples of each of a plurality of historical terminal models, wherein the training samples of each of the historical terminal models include an influence factor value of the historical terminal model under a plurality of preset dimensions and an actual failure rate of a display screen cover of the historical terminal model, wherein the plurality of preset dimensions refer to dimensions that affect the display screen cover failure of the historical terminal model;

[0042] A first determination module is configured to determine the target dimension of the terminal model to be estimated according to the training sample and the failure rate estimation model framework;

[0043] The second determination module is configured to obtain a target failure rate estimation model based on the target dimension, the influence factor values ​​of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal models and the failure rate estimation model framework. The target failure rate estimation model is used to estimate the failure rate of the display screen cover of the terminal model to be estimated.

[0044] According to a third aspect of an embodiment of the present disclosure, there is provided an electronic device, including:

[0045] processor;

[0046] a memory for storing processor-executable instructions;

[0047] Wherein, the processor is configured to execute the steps of the method described in the first aspect of the present disclosure.

[0048] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, on which computer program instructions are stored. When the program instructions are executed by a processor, the steps of the method for generating a failure rate prediction model provided in the first aspect of the present disclosure are implemented.

[0049] A technical solution is adopted to determine the target dimension of the terminal model to be estimated according to the influencing factor values ​​of multiple historical terminal models under multiple preset dimensions, their actual failure rates and failure rate estimation model frameworks, and then obtain the target failure rate estimation model based on the target dimension, the influencing factor values ​​of multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal model and the failure rate estimation model framework. In this way, during the design stage of the terminal model to be estimated, the failure rate of the display screen cover of the terminal model to be estimated can be accurately evaluated, the quality of the designed terminal model to be estimated can be improved, the service life of the terminal model to be estimated can be increased, and the after-sales rate of the terminal model to be estimated after it is put on the market can be effectively reduced.

[0050] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0052] Figure 1 The present invention is a flowchart of a method for generating a failure rate prediction model according to an exemplary embodiment.

[0053] Figure 2 The invention is a block diagram of a device for generating a failure rate prediction model according to an exemplary embodiment.

[0054] Figure 3 It is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0055] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0056] It should be noted that all actions of acquiring signals, information or data in the present disclosure are carried out in compliance with the relevant data protection laws and policies of the country where the device is located and with the authorization given by the owner of the corresponding device.

[0057] Due to material reasons, according to the after-sales performance of listed terminal models, the probability of straight screen display cover plate breaking is lower than that of curved screen display cover plate. In addition, due to the wide range of straight screen terminal users and complex user habits, the cover plate breakage is relatively random. At present, the estimation of the breakage of straight screen mobile phone screen cover plates mainly refers to the data of historical terminal models, relies on business experience to determine the key factors, assigns weights to each key factor, and estimates through simple formulas. Such estimates are not based on scientific statistical methods, and are often less accurate and less reliable.

[0058] In the related art, the weights of various factors are determined based on expert experience, which has the following disadvantages: (1) the weight determination lacks scientific and effective data support; (2) the prediction effect of the empirical prediction formula on new projects is biased; (3) there is no unified method for adjusting the weights. Therefore, in the related art, it is impossible to accurately predict the failure rate of the display cover.

[0059] In view of this, the present disclosure provides a method, device, electronic device and medium for generating a failure rate prediction model to improve the accuracy of estimating the failure rate of a display screen cover.

[0060] Figure 1 FIG. 1 is a flow chart showing a method for generating a failure rate prediction model according to an exemplary embodiment. Figure 1 As shown, the method may include:

[0061] In step S11, training samples of respective multiple historical terminal models are obtained.

[0062] Among them, the training samples of each historical terminal model include the impact factor value of the historical terminal model under multiple preset dimensions and the failure rate of the display screen cover of the historical terminal model, and the multiple preset dimensions refer to the dimensions that affect the failure rate of the display screen cover of the historical terminal model.

[0063] For example, taking the display screen as a straight screen, according to expert experience, the factors that affect the breakage of the cover of the straight screen display screen include but are not limited to the screen type, cover plate cutting thickness, cover plate material, single ROR data, single 4PB data, single sandpaper drop height, the weight of the entire terminal, the screen size of the display screen, and the simulated stiffness of the middle frame. Therefore, in the present disclosure, multiple preset dimensions include but are not limited to the screen type dimension, the cover plate cutting thickness dimension, the cover plate material dimension, the single ROR data dimension, the single 4PB data dimension, the single sandpaper drop height dimension, the weight of the entire terminal, the screen size of the display screen, and the simulated stiffness of the middle frame. Among them, the single ROR data refers to the data of the hardness and strength of the glass monomer measured by the ring-on-ring test method, the single 4PB data refers to the data of the hardness and strength of the glass monomer measured by the test method, and the single sandpaper drop height is also the data characterizing the hardness and strength of the glass monomer.

[0064] Table 1 shows the above-mentioned multiple preset dimensions and the order relationship between the impact factor value and the failure rate FFR under each preset dimension.

[0065] Table 1

[0066] Preset Dimensions The order relationship between the influencing factor value and the failure rate FFR <![CDATA[Order relation identifier S i > Screen type GOLED FFR>POLED FFR>LCD FFR +1 Cover plate cutting thickness The thicker the thickness, the lower the FFR -1 Cover material GG7 FFR<GG5 FFR -1 Monomer ROR data The larger the value, the lower the FFR -1 Single 4PB data The larger the value, the lower the FFR -1 Single sandpaper drop height The higher the altitude, the lower the FFR -1 Weight of the entire terminal The greater the weight, the higher the FFR +1 Display screen size uncertain 0 Middle frame simulation stiffness The greater the stiffness, the lower the FFR -1

[0067] Among them, in Table 1, the screen types include GOLED type, POLED type and LCD type, wherein GOLED refers to an organic light-emitting semiconductor OLED with a glass substrate, and POLED refers to an organic light-emitting semiconductor OLED with a polyethylene substrate. The failure rate FFR of the GOLE type cover is greater than the failure rate FFR of the POLED type cover, and the failure rate FFR of the POLED type cover is greater than the failure rate FFR of the LCD type cover. The cover material refers to the material of the glass used to form the cover, and the material of the glass may include GG7 and GG5 materials. It should be understood that for the influencing factor values ​​under the two dimensions of screen type and cover material, each screen type can be quantified, for example, the GOLE type is quantified as a first value, the POLED type is quantified as a second value, and the LCD type is quantified as a third value, wherein the first value, the second value and the third value are all different. For another example, the GG7 material is quantified as a fourth value, and the GG5 material is quantified as a fifth value, wherein the fourth value and the fifth value are different.

[0068] In addition, in Table 1, Si The symbol of the order relation, S i = +1 indicates that the impact factor value under this dimension is proportional to the failure rate FFR, S i =-1 indicates that the impact factor value under this dimension is inversely proportional to the failure rate FFR, S i =0 indicates that the order relationship between the impact factor value and the failure rate FFR under this dimension is uncertain.

[0069] It should be understood that, in the case where the display screen is a curved screen, the preset dimension may also be other dimensions that affect the failure rate of the curved screen cover, and the present disclosure does not make any specific restrictions on this.

[0070] In the present disclosure, historical terminal models refer to terminal models that have been on the market and for which after-sales data has been collected. The actual failure rate of the terminal model can be determined based on the collected after-sales data.

[0071] In step S12, the target dimension of the terminal model to be estimated is determined based on the training samples and the failure rate estimation model framework.

[0072] In one embodiment, the formula of the failure rate prediction model framework is as follows:

[0073] log(FFR)=b0+b1X1+b2X2+…+b n X n +∈ (1)

[0074] Where FFR represents the failure rate, b0 to b n Characterize the parameter group corresponding to the failure rate prediction model framework, X1 to X n Respectively represent the impact factor values ​​under n dimensions, and ∩ represents the residual. It should be understood that n is usually greater than or equal to the number of preset dimensions.

[0075] For example, taking the preset dimensions shown in Table 1 as an example, X1 to X9 respectively represent the impact factor values ​​under the preset dimensions, wherein the impact factor value under the preset dimension refers to the numerical value of the impact factor under the dimension.

[0076] In one embodiment, the preset dimension may be determined as the target dimension of the terminal model to be estimated, wherein the terminal model to be estimated refers to a terminal model that has not yet been launched, or a terminal model that has been launched but is different from the historical terminal model.

[0077] In another embodiment, the impact factor value that has a greater impact on the failure rate can be determined based on the training sample and the failure rate prediction model framework, and then the dimension of the impact factor value that has a greater impact can be determined as the target dimension. The specific method of determining the target dimension will be described below and will not be repeated here.

[0078] In step S13, a target failure rate prediction model is obtained according to the target dimension, the impact factor values ​​of each of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal models, and the failure rate prediction model framework.

[0079] Among them, the target failure rate prediction model is used to estimate the failure rate of the display screen cover of the terminal model to be estimated.

[0080] For example, the number of dimensions included in the failure rate prediction model framework is generally greater than or equal to the number of target dimensions. Therefore, in the present disclosure, after determining the target dimension, only the model part corresponding to the target dimension in the failure rate prediction model framework can be trained to obtain the target failure rate prediction model.

[0081] For example, based on the target dimension and the failure rate prediction model framework, an initial failure rate prediction model that only includes the target dimension can be obtained. Then, the influencing factor values ​​of multiple historical terminal models under the target dimension are used as model input parameters, and the actual failure rates of the display screen covers of the historical terminal models are used as model output parameters. The initial failure rate prediction model that only includes the target dimension is trained to obtain the target failure rate prediction model.

[0082] A technical solution is adopted to determine the target dimension of the terminal model to be estimated according to the influencing factor values ​​of multiple historical terminal models under multiple preset dimensions, their actual failure rates and failure rate estimation model frameworks, and then obtain the target failure rate estimation model based on the target dimension, the influencing factor values ​​of multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal model and the failure rate estimation model framework. In this way, during the design stage of the terminal model to be estimated, the failure rate of the display screen cover of the terminal model to be estimated can be accurately evaluated, the quality of the designed terminal model to be estimated can be improved, the service life of the terminal model to be estimated can be increased, and the after-sales rate of the terminal model to be estimated after it is put on the market can be effectively reduced.

[0083] In order to facilitate those skilled in the art to better understand the method for generating a failure rate prediction model provided by the present disclosure, the method for generating a failure rate prediction model is described below with a complete embodiment.

[0084] Figure 1Step S12 of determining the target dimension of the terminal model to be estimated according to the training samples and the failure rate estimation model framework may include: step a), determining each historical terminal model as the target historical terminal model in turn; step b), for each target historical terminal model, generating a failure rate estimation model of the target historical terminal model according to the training samples and the failure rate estimation model framework of other historical terminal models except the target historical terminal model; determining the prediction error of the target historical terminal model according to the failure rate estimation model of the target historical terminal model and the training samples of the target historical terminal model; step c), if it is determined according to the prediction error of each historical terminal model that the preset termination condition is not satisfied, then updating multiple preset dimensions and re-executing the step of determining the prediction error of the target historical terminal model, and determining the preset dimension when it is determined according to the prediction error of each of the historical terminal models that the termination condition is satisfied as the target dimension of the terminal model to be estimated.

[0085] In one embodiment, generating a failure rate prediction model for a target historical terminal model may include: using the influencing factor values ​​under preset dimensions in training samples of other historical terminal models except the target historical terminal model as model input parameters, using the actual failure rates of display screen covers of other historical terminal models as model output parameters, training the failure rate prediction model framework to obtain a model, and the model is recorded as the failure rate prediction model for the target historical terminal model.

[0086] In another embodiment, generating a failure rate prediction model for a target historical terminal model may include: step 1), determining a first target dimension from multiple preset dimensions based on the impact factor values ​​of other historical terminal models other than the target historical terminal model in multiple preset dimensions and the actual failure rates of the display screen covers of the other historical terminal models; step 2), generating an initial failure rate prediction model corresponding to the target historical terminal model based on the first target dimension and the failure rate prediction model framework; step 3), generating a failure rate prediction model for the target historical terminal model based on the impact factor values ​​of other historical terminal models other than the target historical terminal model in the first target dimension, the actual failure rates of the display screen covers of the other historical terminal models and the initial failure rate prediction model corresponding to the target historical terminal model.

[0087] Among them, step 1) specifically includes: according to the influence factor values ​​of other historical terminal models except the target historical terminal model in multiple preset dimensions, the actual failure rate of the display screen cover of other historical terminal models and the failure rate prediction model framework, determine the first parameter group of the initial failure rate prediction model corresponding to the target historical terminal model, the first parameter group includes the first coefficients corresponding to the influence factor values ​​in multiple preset dimensions; determine the dimension of the influence factor value corresponding to the first coefficient whose absolute value is greater than the first threshold as the first target dimension.

[0088] In this embodiment, the optimization algorithm library of Python, R and other software can be used to solve the optimization problem of the following formula (2) to obtain the parameter group [b0, b1, ..., b9] of the initial failure rate prediction model corresponding to the target historical terminal model:

[0089]

[0090] In the above formula, it is assumed that the number of the preset dimensions is 9, N represents the number of historical terminal models, and K sample represents the historical terminal model sequence composed of other historical terminal models except the target historical terminal model, and k represents the sequence K sample The kth historical terminal model in .

[0091] In one possible way of this embodiment, other historical terminal models except the target historical terminal model are combined into a historical terminal model sequence. For example, assuming that the multiple historical terminal models include model 1, model 2, model 3, model 4 and model 5, and the target historical terminal model is model 1, then the sequence K sample For [Model 2, Model 3, Model 4, Model 5], based on the above optimization algorithm, according to the training samples of Model 2, Model 3, Model 4 and Model 5, a group of parameter groups are obtained, and the parameter group is determined as the first parameter group of the initial failure rate estimation model corresponding to the target historical terminal model. Afterwards, the dimension of the impact factor value corresponding to the first coefficient whose absolute value is greater than the first threshold is determined as the first target dimension. Among them, the larger the absolute value of the coefficient, the greater the influence of the dimension of the impact factor value on the failure rate of the display cover. Therefore, the dimension of the impact factor value corresponding to the first coefficient whose absolute value is greater than the first threshold is determined as the first target dimension to ensure that the impact factor value under the determined target dimension is the impact factor value that has a greater impact on the failure rate of the display cover. Among them, the first threshold can be 0.

[0092] It should be understood that the present disclosure uses the least square method with constraints to optimize the above formula (2) to obtain a parameter set. The constraint condition is b i S i ≥0. b i The coefficient corresponding to the impact factor value under the i-th dimension, S i An identifier that represents the ordinal relationship between the impact factor value and the failure rate FFR in the i-th dimension.

[0093] In another possible manner in this embodiment, in order to ensure the accuracy of the determined first target dimension, the above formula (2) can be optimized multiple times, and multiple first coefficients are obtained for the impact factor value under each preset dimension, and then based on the multiple first coefficients, it is determined whether the impact factor value under the preset dimension has a greater impact on the failure rate of the display cover. For example, determining the first parameter group of the initial failure rate prediction model corresponding to the target historical terminal model includes:

[0094] Repeat the steps of obtaining the second parameter set for a preset number of times, where the preset number is an integer greater than 1:

[0095] Randomly select N-1 historical terminal models from other historical terminal models except the target historical terminal model to form a historical terminal model sequence, where N is the total number of the historical terminal models;

[0096] Determine a second parameter group of an initial failure rate prediction model corresponding to a target historical terminal model according to the influencing factor values ​​of each historical terminal model in the historical terminal model sequence under multiple preset dimensions, the actual failure rate of the respective display screen cover plates, and the failure rate prediction model framework;

[0097] According to a preset number of second parameter groups and a preset confidence interval of the target historical terminal model, a first parameter group of the initial failure rate prediction model corresponding to the target historical terminal model is determined.

[0098] For example, assuming that the preset number is 100, multiple historical terminal models include model 1, model 2, model 3, model 4 and model 5, and the target historical terminal model is model 1, then 4 models are randomly selected from model 2, model 3, model 4 and model 5 to form a historical terminal model sequence. The extraction here can be repeated extraction, for example, the 4 models extracted are model 2, model 2, model 3 and model 3. Afterwards, according to the impact factor values ​​of each historical terminal model in the historical terminal model sequence formed by this extraction under multiple preset dimensions, the actual failure rate of each display screen cover and the failure rate prediction model framework, the second parameter group of the initial failure rate prediction model corresponding to the target historical terminal model is determined. Among them, the second parameter group of the initial failure rate prediction model corresponding to the target historical terminal model can be obtained by optimizing the above formula (2). Repeating the above method 100 times can obtain 100 second parameter groups.

[0099] Assume that the confidence interval is the confidence interval corresponding to the upper and lower 2.5% quantiles For each impact factor value under the preset dimension, the 100 second coefficients and confidence intervals corresponding to the impact factor value are used. Determine the first parameter group of the initial failure rate prediction model corresponding to the target historical terminal model. Then, determine the dimension of the impact factor value corresponding to the first coefficient whose absolute value is greater than the first threshold as the first target dimension. For example, the confidence interval of the coefficient The dimension whose impact factor value does not include 0 is determined as the first target dimension.

[0100] At this point, according to the above method, the first target dimension can be determined from multiple preset dimensions.

[0101] It should be understood that, assuming that the failure rate prediction model framework includes the influencing factor values ​​of the 9 dimensions in Table 1, which are respectively recorded as X1, ..., X9, the number of the first target dimensions is 5, and the first target dimensions are the first five dimensions in Table 1, then the initial failure rate prediction model corresponding to the generated target historical terminal model is shown in formula (3): log(FFR)=b0+b1X1+b2X2+…+b5X5+∈ (3).

[0102] The above step 3) can be specifically performed in the following manner: according to the impact factor values ​​of other historical terminal models except the target historical terminal model in the first target dimension, the actual failure rates of the display screen covers of other historical terminal models and the initial failure rate estimation model corresponding to the target historical terminal model, a fourth parameter group of the initial failure rate estimation model corresponding to the target historical terminal model is obtained, and the fourth group of parameters is substituted into the initial failure rate estimation model corresponding to the target historical terminal model to obtain the failure rate estimation model of the target historical terminal model.

[0103] Among them, the method of obtaining the fourth parameter group is similar to the method of directly obtaining the first parameter group according to the above method, except that the initial failure rate prediction model corresponding to the target historical terminal model is optimized, which will not be repeated here.

[0104] The specific method of determining the prediction error of the target historical terminal model is as follows: first, according to the failure rate prediction model of the target historical terminal model and the impact factor value of the target historical terminal model under the first target dimension, the estimated failure rate of the display screen cover of the target historical terminal model is determined. For example, the impact factor value under the first target dimension of the target historical terminal model is substituted into the failure rate prediction model of the target historical terminal model to obtain the logarithm of the estimated failure rate of the display screen cover of the target historical terminal model, and the estimated failure rate is obtained after taking the exponent.

[0105] Then, the prediction error of the target historical terminal model is determined according to the estimated failure rate and the actual failure rate. For example, the ratio of the absolute value of the difference between the estimated failure rate and the actual failure rate to the actual failure rate is determined as the prediction error of the target historical terminal model.

[0106] Thus, the prediction error of each target historical terminal model can be determined according to the above method, that is, the prediction error of each historical terminal model is determined. Then, the target dimension of the terminal model to be estimated is determined according to the prediction error of each historical terminal model.

[0107] For example, Figure 1 Step S12 determines the target dimension of the terminal model to be estimated based on the training samples and the failure rate prediction model framework, and can also include: step d), determining the maximum value among the prediction errors of multiple historical terminal models; step e), if the maximum value is less than a preset threshold, determining that the preset termination condition is met; step f), if the maximum value is greater than or equal to the preset threshold, determining that the termination condition is not met.

[0108] In the present disclosure, if the maximum value is less than the preset threshold, it indicates that the multiple preset dimensions are accurate, and then the current preset dimension is determined as the target dimension of the terminal model to be estimated. If the maximum value is greater than or equal to the preset threshold, it indicates that the multiple preset dimensions are inaccurate, and the preset dimensions need to be updated at this time, for example, some dimensions in the preset dimensions shown in the first column of Table 1 are deleted and / or other dimensions are added, etc.

[0109] After the prediction dimension is updated, the prediction error of each historical terminal model and the maximum value of the prediction errors of multiple historical terminal models are determined in the above manner. If the maximum value is less than the preset threshold, the preset dimension after this update is determined as the target dimension of the terminal model to be estimated. If it is still greater than or equal to the preset threshold, the preset dimension is continuously updated until the maximum value of the prediction errors of multiple historical terminal models is less than the preset threshold, and the most recently updated preset dimension is determined as the target dimension of the terminal model to be estimated.

[0110] In addition, in order to avoid too many updates, the number of updates can also be limited. For example, if the number of updates to multiple preset dimensions reaches the preset number and still does not meet the preset termination condition, the maximum value of the prediction errors of multiple historical terminal models after each update of the preset dimension is obtained, and the target update number corresponding to the minimum value of the multiple maximum values ​​and the preset dimension corresponding to the target update number are determined as the target dimension of the terminal model to be estimated.

[0111] Assuming that the preset number of times is 10 times, after each update of the preset dimension, the maximum values ​​of the prediction errors of multiple historical terminal models after this update are obtained, that is, 10 maximum values ​​are obtained, and then the minimum value is determined from the 10 maximum values. Assuming that the minimum value is the prediction error of a historical terminal model after the fourth update, the target update number is determined to be the fourth time, and then the preset dimension after the fourth update is determined as the target dimension of the terminal model to be estimated.

[0112] It should be understood that when repeatedly determining a preset number of second parameter groups, when the preset training termination condition is not met, in addition to updating the preset dimensions, the preset number and / or confidence interval can also be adjusted to make the prediction error of each historical terminal model less than a preset threshold on average by adjusting the number and / or confidence interval.

[0113] Figure 1 In step S13, according to the target dimension, the impact factor values ​​of each of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal model and the failure rate prediction model framework, the target failure rate prediction model can be obtained, which may include:

[0114] First, according to the target dimension and the failure rate prediction model framework, the initial failure rate prediction model corresponding to the terminal model to be estimated is determined. The method of determining the initial failure rate model corresponding to the terminal model to be estimated is similar to the method of generating the initial failure rate prediction model corresponding to the target historical terminal model, which will not be repeated here.

[0115] Next, based on the impact factor values ​​of multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal models and the initial failure rate prediction model, the third parameter group of the initial failure rate prediction model corresponding to the terminal model to be estimated is obtained.

[0116] For example, the optimization algorithm library of Python, R and other software can be used to solve the optimization problem of the following formula (4) to obtain the third parameter group of the initial failure rate prediction model corresponding to the terminal model to be estimated:

[0117]

[0118] Wherein, Y represents the number of target dimensions, and K represents a historical terminal model sequence including all historical terminal models used in the present disclosure.

[0119] Finally, the target failure rate prediction model is obtained according to the third parameter group and the initial failure rate prediction model. That is, the target failure rate prediction model can be obtained by substituting the third parameter group into the initial failure rate prediction model formula.

[0120] After obtaining the target failure rate estimation model, the impact factor value of the terminal model to be estimated under the target dimension is obtained, and the impact factor value is input into the target failure rate estimation model to obtain the failure rate of the display screen cover of the terminal model to be estimated.

[0121] In the present disclosure, according to the above method, the target failure rate estimation model corresponding to the terminal model to be estimated can be obtained. After testing, the target failure rate estimation model generated according to the method provided by the present disclosure can achieve an accuracy of more than 80% in failure rate estimation. In this way, the target failure rate estimation model generated by the method provided by the present disclosure effectively improves the accuracy of the display cover failure rate estimation.

[0122] Based on the same inventive concept, the present disclosure also provides a device for generating a failure rate prediction model. Figure 2 is a block diagram of a device for generating a failure rate prediction model according to an exemplary embodiment. Figure 2 , the device 200 for generating a failure rate prediction model may include:

[0123] The acquisition module 201 is configured to acquire training samples of each of a plurality of historical terminal models, wherein the training samples of each of the historical terminal models include an impact factor value of the historical terminal model under a plurality of preset dimensions and an actual failure rate of a display screen cover of the historical terminal model, wherein the plurality of preset dimensions refer to dimensions that affect the display screen cover failure of the historical terminal model;

[0124] A first determination module 202 is configured to determine the target dimension of the terminal model to be estimated according to the training sample and the failure rate estimation model framework;

[0125] The second determination module 203 is configured to obtain a target failure rate estimation model based on the target dimension, the influence factor values ​​of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal models and the failure rate estimation model framework. The target failure rate estimation model is used to estimate the failure rate of the display screen cover of the terminal model to be estimated.

[0126] Optionally, the first determining module 202 may include:

[0127] A first determination submodule is configured to sequentially determine each historical terminal model as a target historical terminal model;

[0128] The second determination submodule is configured to generate, for each of the target historical terminal models, a failure rate prediction model of the target historical terminal model based on training samples of other historical terminal models except the target historical terminal model and a failure rate prediction model framework; determine a prediction error of the target historical terminal model based on the failure rate prediction model of the target historical terminal model and the training samples of the target historical terminal model;

[0129] The updating submodule is configured to update the multiple preset dimensions and re-execute the step of determining the prediction error of the target historical terminal model if it is determined that the preset termination condition is not satisfied according to the prediction error of each of the historical terminal models, and determine the preset dimensions when the termination condition is satisfied according to the prediction error of each of the historical terminal models as the target dimensions of the terminal model to be estimated.

[0130] Optionally, the second determination submodule is configured to: determine the first target dimension from the multiple preset dimensions according to the impact factor values ​​of other historical terminal models except the target historical terminal model in the multiple preset dimensions and the actual failure rates of the display screen covers of the other historical terminal models;

[0131] Generate an initial failure rate prediction model corresponding to the target historical terminal model according to the first target dimension and the failure rate prediction model framework;

[0132] Generate a failure rate estimation model for the target historical terminal model according to the impact factor values ​​of other historical terminal models except the target historical terminal model under the first target dimension, the actual failure rates of the display screen covers of the other historical terminal models, and the initial failure rate estimation model corresponding to the target historical terminal model;

[0133] The second determination submodule is further configured to: determine the estimated failure rate of the display screen cover of the target historical terminal model according to the failure rate estimation model of the target historical terminal model and the impact factor value of the target historical terminal model under the first target dimension;

[0134] The prediction error of the target historical terminal model is determined based on the estimated failure rate and the actual failure rate.

[0135] Optionally, the second determining submodule may include:

[0136] A third determination submodule is configured to determine a first parameter group of an initial failure rate prediction model corresponding to the target historical terminal model according to the influence factor values ​​of other historical terminal models other than the target historical terminal model under multiple preset dimensions, the actual failure rates of the display screen covers of the other historical terminal models, and the failure rate prediction model framework, wherein the first parameter group includes first coefficients corresponding to the influence factor values ​​under the multiple preset dimensions;

[0137] The fourth determination submodule is configured to determine the dimension of the impact factor value corresponding to the first coefficient whose absolute value is greater than the first threshold as the first target dimension.

[0138] Optionally, the third determination submodule is configured to: repeat the step of obtaining the second parameter group a preset number of times, where the preset number is an integer greater than 1:

[0139] Randomly select N-1 historical terminal models from other historical terminal models except the target historical terminal model to form a historical terminal model sequence, where N is the total number of the historical terminal models;

[0140] Determine a second parameter group of the initial failure rate prediction model corresponding to the target historical terminal model according to the influence factor values ​​of each historical terminal model in the historical terminal model sequence under multiple preset dimensions, the actual failure rate of the respective display screen cover plates, and the failure rate prediction model framework;

[0141] According to the preset number of second parameter groups and the preset confidence interval of the target historical terminal model, a first parameter group of the initial failure rate prediction model corresponding to the target historical terminal model is determined.

[0142] Optionally, the second determination module 203 is configured to: determine an initial failure rate estimation model corresponding to the terminal model to be estimated according to the target dimension and the failure rate estimation model framework;

[0143] Determine a third parameter group of the initial failure rate estimation model corresponding to the terminal model to be estimated according to the influence factor values ​​of each of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal model, and the initial failure rate estimation model;

[0144] A target failure rate prediction model is obtained according to the third parameter group and the initial failure rate prediction model.

[0145] Optionally, the first determining module 202 may further include:

[0146] a fourth determination submodule, configured to determine a maximum value among the prediction errors of the plurality of historical terminal models;

[0147] a fifth determination submodule, configured to determine that a preset termination condition is satisfied if the maximum value is less than a preset threshold;

[0148] The sixth determination submodule is configured to determine that the termination condition is not satisfied if the maximum value is greater than or equal to the preset threshold.

[0149] Optionally, the first determining module 202 may further include:

[0150] The seventh determination submodule is configured to obtain the maximum value of the prediction errors of multiple historical terminal models after each update of the preset dimensions if the number of updates of the multiple preset dimensions reaches the preset number and does not meet the preset termination condition, and determine the target update number corresponding to the minimum value of the multiple maximum values, and the preset dimension corresponding to the target update number as the target dimension of the terminal model to be estimated.

[0151] Optionally, the device 200 for generating a failure rate prediction model may further include:

[0152] The adjustment module is configured to adjust the preset number and / or the confidence interval if a preset termination condition is not met.

[0153] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0154] The present disclosure also provides a computer-readable storage medium having computer program instructions stored thereon. When the program instructions are executed by a processor, the steps of the method for generating a failure rate prediction model provided by the present disclosure are implemented.

[0155] Figure 3 8 is a block diagram of an electronic device according to an exemplary embodiment. For example, the electronic device 800 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, etc.

[0156] Reference Figure 3 , the electronic device 800 may include one or more of the following components: a processing component 802 , a memory 804 , a power component 806 , a multimedia component 808 , an audio component 810 , an input / output interface 812 , a sensor component 814 , and a communication component 816 .

[0157] The processing component 802 generally controls the overall operation of the electronic device 800, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the method for generating a failure rate prediction model. In addition, the processing component 802 may include one or more modules to facilitate the interaction between the processing component 802 and other components. For example, the processing component 802 may include a multimedia module to facilitate the interaction between the multimedia component 808 and the processing component 802.

[0158] The memory 804 is configured to store various types of data to support operations on the electronic device 800. Examples of such data include instructions for any application or method operating on the electronic device 800, contact data, phone book data, messages, pictures, videos, etc. The memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0159] The power supply component 806 provides power to the various components of the electronic device 800. The power supply component 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the electronic device 800.

[0160] The multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 808 includes a front camera and / or a rear camera. When the electronic device 800 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

[0161] The audio component 810 is configured to output and / or input audio signals. For example, the audio component 810 includes a microphone (MIC), and when the electronic device 800 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 804 or sent via the communication component 816. In some embodiments, the audio component 810 also includes a speaker for outputting audio signals.

[0162] The input / output interface 812 provides an interface between the processing component 802 and the peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: a home button, a volume button, a start button, and a lock button.

[0163] The sensor assembly 814 includes one or more sensors for providing various aspects of status assessment for the electronic device 800. For example, the sensor assembly 814 can detect the open / closed state of the electronic device 800, the relative positioning of the components, such as the display and keypad of the electronic device 800, and the sensor assembly 814 can also detect the position change of the electronic device 800 or a component of the electronic device 800, the presence or absence of contact between the user and the electronic device 800, the orientation or acceleration / deceleration of the electronic device 800, and the temperature change of the electronic device 800. The sensor assembly 814 may include a proximity sensor configured to detect the presence of a nearby object without any physical contact. The sensor assembly 814 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0164] The communication component 816 is configured to facilitate wired or wireless communication between the electronic device 800 and other devices. The electronic device 800 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 816 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0165] In an exemplary embodiment, the electronic device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to execute the method for generating a failure rate prediction model.

[0166] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 804 including instructions, and the instructions can be executed by a processor 820 of an electronic device 800 to complete the method of generating a failure rate prediction model. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

[0167] In another exemplary embodiment, a computer program product is further provided. The computer program product includes a computer program executable by a programmable device. The computer program has a code portion for executing the above-mentioned method for generating a failure rate prediction model when executed by the programmable device.

[0168] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the present disclosure. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0169] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for generating a failure rate prediction model, characterized in that: The method comprises: Acquire training samples of each of a plurality of historical terminal models, wherein the training samples of each of the historical terminal models include an influence factor value of the historical terminal model under a plurality of preset dimensions and an actual failure rate of a display screen cover of the historical terminal model, wherein the plurality of preset dimensions refer to dimensions that affect the failure rate of the display screen cover of the historical terminal model; Determining the target dimension of the terminal model to be estimated based on the training samples and the failure rate estimation model framework; According to the target dimension, the influence factor values ​​of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal models and the failure rate prediction model framework, a target failure rate prediction model is obtained. The target failure rate prediction model is used to estimate the failure rate of the display screen cover of the terminal model to be estimated.

2. The method according to claim 1, characterized in that: The step of determining the target dimension of the terminal model to be estimated according to the training samples and the failure rate estimation model framework includes: Determining each historical terminal model as a target historical terminal model in turn; For each of the target historical terminal models, a failure rate prediction model of the target historical terminal model is generated based on training samples of other historical terminal models except the target historical terminal model and a failure rate prediction model framework; a prediction error of the target historical terminal model is determined based on the failure rate prediction model of the target historical terminal model and the training samples of the target historical terminal model; If it is determined according to the prediction error of each of the historical terminal models that the preset termination condition is not satisfied, the multiple preset dimensions are updated and the step of determining the prediction error of the target historical terminal model is re-executed, and the preset dimensions when it is determined according to the prediction error of each of the historical terminal models that the termination condition is satisfied are determined as the target dimensions of the terminal model to be estimated.

3. The method according to claim 2, characterized in that The generating the failure rate prediction model of the target historical terminal model according to the training samples and failure rate prediction model framework of other historical terminal models except the target historical terminal model includes: Determine a first target dimension from the multiple preset dimensions according to the impact factor values ​​of other historical terminal models except the target historical terminal model in multiple preset dimensions and the actual failure rates of the display screen covers of the other historical terminal models; Generate an initial failure rate prediction model corresponding to the target historical terminal model according to the first target dimension and the failure rate prediction model framework; Generate a failure rate estimation model for the target historical terminal model according to the impact factor values ​​of other historical terminal models except the target historical terminal model under the first target dimension, the actual failure rates of the display screen covers of the other historical terminal models, and the initial failure rate estimation model corresponding to the target historical terminal model; The step of determining the prediction error of the target historical terminal model according to the failure rate prediction model of the target historical terminal model and the training sample of the target historical terminal model includes: Determining an estimated failure rate of a display screen cover of a target historical terminal model according to a failure rate estimation model of the target historical terminal model and an impact factor value of the target historical terminal model under the first target dimension; The prediction error of the target historical terminal model is determined based on the estimated failure rate and the actual failure rate.

4. The method according to claim 3, characterized in that The determining of the first target dimension from the multiple preset dimensions according to the influence factor values ​​of the other historical terminal models except the target historical terminal model in the multiple preset dimensions and the actual failure rates of the display screen covers of the other historical terminal models includes: Determine a first parameter group of an initial failure rate prediction model corresponding to the target historical terminal model according to the influence factor values ​​of other historical terminal models other than the target historical terminal model under multiple preset dimensions, the actual failure rates of the display screen covers of the other historical terminal models, and the failure rate prediction model framework, wherein the first parameter group includes first coefficients corresponding to the influence factor values ​​under the multiple preset dimensions; The dimension of the impact factor value corresponding to the first coefficient whose absolute value is greater than the first threshold is determined as the first target dimension.

5. The method according to claim 4, characterized in that The determining of a first parameter group of an initial failure rate prediction model corresponding to the target historical terminal model according to the influence factor values ​​of other historical terminal models except the target historical terminal model under multiple preset dimensions, the actual failure rates of display screen covers of the other historical terminal models, and the failure rate prediction model framework includes: Repeat the step of obtaining the second parameter set a preset number of times, where the preset number is an integer greater than 1: Randomly select N-1 historical terminal models from other historical terminal models except the target historical terminal model to form a historical terminal model sequence, where N is the total number of the historical terminal models; Determine a second parameter group of the initial failure rate prediction model corresponding to the target historical terminal model according to the influence factor values ​​of each historical terminal model in the historical terminal model sequence under multiple preset dimensions, the actual failure rate of the respective display screen cover plates, and the failure rate prediction model framework; According to the preset number of second parameter groups and the preset confidence interval of the target historical terminal model, a first parameter group of the initial failure rate prediction model corresponding to the target historical terminal model is determined.

6. The method according to any one of claims 1 to 5, characterized in that The target failure rate prediction model is obtained according to the target dimension, the impact factor values ​​of each of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal model and the failure rate prediction model framework, including: Determining an initial failure rate prediction model corresponding to the terminal model to be estimated according to the target dimension and the failure rate prediction model framework; Determine a third parameter group of the initial failure rate estimation model corresponding to the terminal model to be estimated according to the influence factor values ​​of each of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal model, and the initial failure rate estimation model; A target failure rate prediction model is obtained according to the third parameter group and the initial failure rate prediction model.

7. The method according to claim 2, characterized in that: The step of determining the target dimension of the terminal model to be estimated according to the training samples and the failure rate estimation model framework further includes: Determining a maximum value among the prediction errors of a plurality of the historical terminal models; If the maximum value is less than a preset threshold, it is determined that a preset termination condition is met; If the maximum value is greater than or equal to the preset threshold, it is determined that the termination condition is not met.

8. The method according to claim 2, characterized in that: The step of determining the target dimension of the terminal model to be estimated according to the training samples and the failure rate estimation model framework further includes: If the number of updates to the multiple preset dimensions reaches the preset number and does not meet the preset termination condition, the maximum value of the prediction errors of multiple historical terminal models after each update of the preset dimensions is obtained, and the target update number corresponding to the minimum value of the multiple maximum values ​​and the preset dimension corresponding to the target update number are determined as the target dimension of the terminal model to be estimated.

9. The method according to claim 5, characterized in that The method further comprises: If the preset termination condition is not met, the preset number and / or the confidence interval are adjusted.

10. A device for generating a failure rate prediction model, characterized in that: The device comprises: An acquisition module is configured to acquire training samples of each of a plurality of historical terminal models, wherein the training samples of each of the historical terminal models include an influence factor value of the historical terminal model under a plurality of preset dimensions and an actual failure rate of a display screen cover of the historical terminal model, wherein the plurality of preset dimensions refer to dimensions that affect the display screen cover failure of the historical terminal model; A first determination module is configured to determine the target dimension of the terminal model to be estimated according to the training sample and the failure rate estimation model framework; The second determination module is configured to obtain a target failure rate estimation model based on the target dimension, the influence factor values ​​of the multiple historical terminal models under the target dimension, the actual failure rate of the display screen cover of the historical terminal models and the failure rate estimation model framework. The target failure rate estimation model is used to estimate the failure rate of the display screen cover of the terminal model to be estimated.

11. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to execute the steps of the method according to any one of claims 1 to 9.

12. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the program instructions are executed by a processor, the steps of the method described in any one of claims 1 to 9 are implemented.

Images