An electronic device life evaluation method based on absorbing markov chain

Abstract

The application relates to the technical field of electronic device detection, and specifically discloses an electronic device life evaluation method based on an absorbing Markov chain, which comprises the following steps: acquiring the electronic device failure rate of all test collection points in a test period; processing a Weibull distribution model according to the electronic device failure rate of all test collection points and the corresponding test collection point time, so as to obtain a processed Weibull distribution model; inputting a target time point in the test period into the processed Weibull distribution model to calculate the failure rate, so as to output the electronic device failure rate of the target time point; establishing a transition probability matrix of an absorbing Markov chain model according to the electronic device failure rate of the target time point; and evaluating the electronic device life of the target time point according to the transition probability matrix of the absorbing Markov chain model. The application can reduce the number of collection points in the test process, and the output result of the model is consistent with the actual result through the fitting of the test data.

Description

Technical Field

[0001] This invention relates to the field of electronic device testing technology, and in particular to a method for evaluating the lifetime of electronic devices based on an absorption Markov chain. Background Technology

[0002] With the development of the semiconductor industry, electronic devices are becoming increasingly integrated, making reliability verification more challenging. The ability to assess device lifespan using limited experimental data is crucial for reliability verification. An electronic device lifespan assessment model based on an absorption Markov chain offers a solution to this problem.

[0003] Currently, there are exponential distribution models, log-normal distribution models, and Weibull distribution models for evaluating the lifespan of electronic devices. These models construct the failure distribution of the device from a probabilistic perspective based on the performance degradation law of electronic materials. Then, the expected test lifespan of the device can be calculated through the failure distribution. Finally, the expected test lifespan and the acceleration factor are multiplied to obtain the expected lifespan of the device.

[0004] Exponential distribution model lifetime assessment equation

[0005]

[0006] Where, E(t) ed λ represents the expected lifetime of an exponential distribution model experiment. ed Indicates the failure rate of the device; e represents the natural constant; A f It is the acceleration coefficient.

[0007] Log-normal distribution model lifetime assessment equation

[0008]

[0009] Where, E(t) ld σ represents the expected lifetime of an exponentially distributed model experiment. lb Represents the scale parameter; μ lb Represents the position parameter; e represents the natural constant; A f It is the acceleration coefficient.

[0010] Weibull distribution model lifetime assessment equation

[0011]

[0012] Where, m wb It is a proportional parameter, k wb It is a shape parameter. Γ represents the gamma function.

[0013] The Weibull distribution conforms to the material degradation law and is simpler to construct, making it the mainstream basic model for device lifetime assessment.

[0014] Pass rate equation based on Weibull distribution model

[0015]

[0016] A set of m can be obtained by recording the device pass rates at two different time points. wb and k wb Then, the lifetime assessment equation of the Weibull distribution model can be obtained.

[0017] During lifetime testing, it is necessary to select test data collection points to record the device failure rate. Even when using the Weibull distribution model to assess lifetime, the failure rate still needs to be recorded in real time, and the device lifetime is estimated based on the Weibull distribution model using the test data. This lifetime assessment model has high requirements for data real-time performance, and the calculated results may have errors compared to the actual device lifetime. Summary of the Invention

[0018] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a method for evaluating the lifetime of electronic devices based on an absorbing Markov chain. This method can reduce the number of data collection points during the experiment and, by fitting the experimental data, make the model's output results consistent with the actual results.

[0019] As a first aspect of the present invention, a method for assessing the lifetime of electronic devices based on an absorption Markov chain is provided, the method comprising:

[0020] Step S1: Obtain the failure rate of electronic devices at all test collection points during the test period. The failure rate of electronic devices at each test collection point is the number of failed electronic devices at the current test collection point divided by the total number of electronic devices.

[0021] Step S2: Process the Weibull distribution model based on the failure rate of electronic devices at all test acquisition points and the corresponding test acquisition point time to obtain the processed Weibull distribution model;

[0022] Step S3: Input the target time point within the test period into the processed Weibull distribution model to calculate the failure rate, and output the electronic device failure rate at the target time point;

[0023] Step S4: Establish the transition probability matrix of the absorption Markov chain model based on the failure rate of the electronic devices at the target time point;

[0024] Step S5: Evaluate the lifetime of the electronic device at the target time point based on the transition probability matrix of the absorbed Markov chain model.

[0025] Furthermore, the step of processing the Weibull distribution model based on the electronic device failure rate and the corresponding test acquisition time at all test acquisition points to obtain the processed Weibull distribution model also includes:

[0026] The failure rate of electronic devices at all the test acquisition points and the corresponding test acquisition point time are input into the Weibull distribution model to obtain multiple calculated values ​​of proportional parameters and multiple calculated values ​​of shape parameters;

[0027] The proportional parameter fitting value of the target time point within the test period is calculated based on the multiple proportional parameter calculation values, and the shape parameter fitting value of the target time point within the test period is calculated based on the multiple shape parameter calculation values.

[0028] The processed Weibull distribution model is established based on the fitted values ​​of the proportional parameter and the shape parameter at the target time points within the experimental period.

[0029] The pass rate equation Pa(t) based on the processed Weibull distribution model is as follows:

[0030]

[0031] The failure rate equation based on the processed Weibull distribution model is then 1-Pa(t), where t represents the target time point within the experimental time period, and m... wb k represents the fitted value of the scaling parameter corresponding to the target time point. wb This represents the fitted value of the shape parameter corresponding to the target time point.

[0032] Further, the step of calculating the proportional parameter fitting value of the target time point within the test period based on the multiple proportional parameter calculation values, and calculating the shape parameter fitting value of the target time point within the test period based on the multiple shape parameter calculation values, further includes:

[0033] Based on the calculated values ​​of the multiple proportional parameters and their corresponding experimental sampling time, the proportional parameter m is fitted. wb The polynomial function model equation, and the shape parameter k fitted based on the calculated values ​​of the multiple shape parameters and their corresponding experimental acquisition point times. wb The polynomial function model equation;

[0034] The proportional parameter m wb The polynomial function model equation is as follows:

[0035] m wb =a0+a1t+a2t 2 +…+a r t r

[0036] The shape parameter k wb The polynomial function model equation is as follows:

[0037] k wb =b0+b1t+b2t 2 +…+b r t r

[0038] Where a0, a1, a2…a r and b0, b1, b2…b r All represent the fitting coefficients, and t represents the time of the experimental data collection point;

[0039] According to the proportional parameter m wb The polynomial function model equation is used to calculate the fitted value of the proportional parameter corresponding to the target time point t within the experimental time period, and the shape parameter k is used as the basis for the calculation. wb The shape parameter fitting value corresponding to the target time point t within the experimental time period is obtained by calculating the polynomial function model equation.

[0040] Furthermore, the step of inputting the target time point within the test period into the processed Weibull distribution model for failure rate calculation, so as to output the electronic device failure rate at the target time point, further includes:

[0041] The target time point t and the fitted value m of the proportional parameter corresponding to the target time point t are used. wb The shape parameter fitting value k corresponding to the target time point t wb Substitute the values ​​into the failure rate equation based on the processed Weibull distribution model to calculate the failure rate of the electronic device at the target time point t.

[0042] Further, the step of calculating the proportional parameter fitting value of the target time point within the test period based on the multiple proportional parameter calculation values, and calculating the shape parameter fitting value of the target time point within the test period based on the multiple shape parameter calculation values, further includes:

[0043] The optimization equations for the fitted values ​​of the proportional parameter and the shape parameter at the target time points within the experimental period are as follows:

[0044]

[0045] Where L is the optimized scale parameter fitting value or the optimized shape parameter fitting value; f is the input value, i.e., the calculated scale parameter value or the calculated shape parameter value; y iR is the fitted value of the proportional parameter or the fitted value of the shape parameter of the i-th test collection point within the test time period; R is the total number of test collection points within the test time period.

[0046] Furthermore, the step of establishing the transition probability matrix of the absorbing Markov chain model based on the electronic device failure rate at the target time point also includes:

[0047] The transition probability matrix C of the absorbing Markov chain model is in the following form:

[0048]

[0049] Wherein, each element P in the transition probability matrix C ij The calculation formula is as follows:

[0050]

[0051] Among them, P ij This represents the probability that j more failure devices will be added after i failure devices exist at the target time point, where i represents the number of existing failure devices, j represents the number of added failure devices, n represents the total number of electronic devices, and F represents the electronic device failure rate at the target time point calculated based on the processed Weibull distribution model.

[0052] Furthermore, the step of evaluating the electronic device lifetime at the target time point based on the transition probability matrix of the absorbed Markov chain model further includes:

[0053] The evaluation equation for the electronic device lifetime at the target time point is E(t). amc as follows:

[0054]

[0055] Where, E(t) amc This represents the lifetime assessment time of electronic devices based on the absorption Markov chain model; A f Indicates the acceleration coefficient; t int The time interval between adjacent data acquisition points within the experimental period is represented by z; z is a constant; n represents the total number of electronic devices; Q represents the probability matrix for transitioning from a non-absorbing state to a non-absorbing state, and the probability matrix Q has the following form:

[0056]

[0057] Furthermore, the acceleration factor A was calculated using the Arrhenius model. f acceleration coefficient A f The calculation formula is as follows:

[0058]

[0059] Among them, A T A represents the temperature acceleration coefficient. v E represents the voltage acceleration factor. aa T represents the activation energy of the device, k represents the Boltzmann constant, and T represents the activation energy of the device. U The temperature at which the device is used, T A The temperature of the device test is represented by K, the electric field strength determined by the test is represented by X, and the thickness of the stress medium is represented by V. A V represents the accelerating voltage of the device. U This indicates the operating voltage of the device.

[0060] The electronic device lifetime assessment method based on absorbing Markov chains provided by this invention has the following advantages: it can reduce the number of sampling points during the experiment, and at the same time, by fitting the experimental data, the output results of the model can be consistent with the actual results. Attached Figure Description

[0061] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the following detailed description to explain the invention, but do not constitute a limitation thereof.

[0062] Figure 1 A flowchart of the electronic device lifetime assessment method based on absorbing Markov chains provided by the present invention.

[0063] Figure 2 A flowchart illustrating a specific implementation of the electronic device lifetime assessment method based on an absorbing Markov chain provided by this invention. Detailed Implementation

[0064] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0065] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0066] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of the invention described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0067] This embodiment provides a method for evaluating the lifetime of electronic devices based on an absorbing Markov chain. Figure 1 A flowchart of the electronic device lifetime assessment method based on absorbing Markov chains provided by the present invention.

[0068] like Figure 1 As shown, the electronic device lifetime assessment method based on absorbing Markov chains includes:

[0069] Step S1: Obtain the failure rate of electronic devices at all test collection points during the test period. The failure rate of electronic devices at each test collection point is the number of failed electronic devices at the current test collection point divided by the total number of electronic devices.

[0070] Specifically, obtaining the failure rate of electronic devices at all test acquisition points during the test period also includes:

[0071] Determine the normal operating temperature and working voltage of the electronic device, and then determine the test temperature, test voltage and test time of the electronic device according to actual needs;

[0072] Based on the determined test duration, data collection points are taken at 24-hour intervals. For tests shorter than 24 hours, data collection points are taken at 1-hour intervals. The end time of the test must be used as a data collection point. If the number of data collection points is odd, an additional data collection point is added one hour before the end of the test.

[0073] At each sampling point, record the test temperature, voltage, and current of the electronic device. Then, stop the test and remove the thermal and electrical stresses from the electronic device. It is important to note that electrical performance tests must be performed on the electronic device within 24 hours after each test stoppage to determine if the electronic device at the current sampling point has failed. The electrical performance test results for each sampling point should be retained, and the number of failed devices at each sampling point should be recorded. The failure rate of the electronic device at each sampling point should be calculated.

[0074] Step S2: Process the Weibull distribution model based on the failure rate of electronic devices at all test acquisition points and the corresponding test acquisition point time to obtain the processed Weibull distribution model;

[0075] Preferably, the step of processing the Weibull distribution model based on the electronic device failure rate and the corresponding test acquisition time at all test acquisition points to obtain the processed Weibull distribution model further includes:

[0076] The failure rate of electronic devices at all the test acquisition points and the corresponding test acquisition point time are input into the Weibull distribution model to obtain multiple calculated values ​​of proportional parameters and multiple calculated values ​​of shape parameters;

[0077] The proportional parameter fitting value of the target time point within the test period is calculated based on the multiple proportional parameter calculation values, and the shape parameter fitting value of the target time point within the test period is calculated based on the multiple shape parameter calculation values.

[0078] Preferably, the step of calculating the proportional parameter fitting value of the target time point within the test period based on the multiple proportional parameter calculation values, and calculating the shape parameter fitting value of the target time point within the test period based on the multiple shape parameter calculation values, further includes:

[0079] Based on the calculated values ​​of the multiple proportional parameters and their corresponding experimental sampling time, the proportional parameter m is fitted. wb The polynomial function model equation, and the shape parameter k fitted based on the calculated values ​​of the multiple shape parameters and their corresponding experimental acquisition point times. wb The polynomial function model equation;

[0080] The proportional parameter m wb The polynomial function model equation is as follows:

[0081] m wb =a0+a1t+a2t 2 +…+a r t r

[0082] The shape parameter k wb The polynomial function model equation is as follows:

[0083] k wb =b0+b1t+b2t 2 +…+b r t r

[0084] Where a0, a1, a2…a r and b0, b1, b2…b rAll represent the fitting coefficients, and t represents the time of the experimental data collection point;

[0085] According to the proportional parameter m wb The polynomial function model equation is used to calculate the fitted value of the proportional parameter corresponding to the target time point t within the experimental time period, and the shape parameter k is used to calculate the fitted value of the proportional parameter. wb The shape parameter fitting value corresponding to the target time point t within the experimental time period is obtained by calculating the polynomial function model equation.

[0086] The processed Weibull distribution model is established based on the fitted values ​​of the proportional parameter and the shape parameter at the target time points within the experimental period.

[0087] The pass rate equation Pa(t) based on the processed Weibull distribution model is as follows:

[0088]

[0089] The failure rate equation based on the processed Weibull distribution model is then 1-Pa(t), where t represents the target time point within the experimental time period, and m... wb k represents the fitted value of the scaling parameter corresponding to the target time point. wb This represents the fitted value of the shape parameter corresponding to the target time point.

[0090] It should be noted that the instantaneous failure rate of the device is obtained through experimental data, and then the yield rate is calculated. According to the Weibull distribution model equation, given the yield rate Pa(t), we can obtain information about k. wb and m wb The equation. A set of definite k can be obtained at every two fixed intervals of experimental time points. wb and m wb The experiment time is divided into 24-hour intervals as data collection points, and a data collection point is also set at the end of the experiment. If the number of data collection points is odd, then an additional data collection point is added 1 hour before the end of the experiment. After the experiment, k can be obtained respectively. wb and m wb The simple relationship with time is fitted using a polynomial function.

[0091] Specifically, the step of calculating the proportional parameter fitting value of the target time point within the test period based on the multiple proportional parameter calculation values, and calculating the shape parameter fitting value of the target time point within the test period based on the multiple shape parameter calculation values, further includes:

[0092] The optimization equations for the fitted values ​​of the proportional parameter and the shape parameter at the target time points within the experimental period are as follows:

[0093]

[0094] Where L is the optimized fitted value of the scaling parameter or the optimized fitted value of the shape parameter; f is the input value, i.e., the calculated value of the scaling parameter or the calculated value of the shape parameter; y i Let be the fitted value of the proportional parameter or the fitted value of the shape parameter of the i-th test sampling point within the test time period; R is the total number of test sampling points within the test time period. After optimization, k can be obtained. wb and m wb The formula relating to time.

[0095] Step S3: Input the target time point within the test period into the processed Weibull distribution model to calculate the failure rate, and output the electronic device failure rate at the target time point;

[0096] Preferably, such as Figure 2 As shown, the step of inputting the target time point within the test period into the processed Weibull distribution model for failure rate calculation, so as to output the electronic device failure rate at the target time point, further includes:

[0097] The target time point t and the corresponding proportional parameter fitting value m are used to determine the target time point t. wb The shape parameter fitting value k corresponding to the target time point t wb Substitute the values ​​into the failure rate equation based on the processed Weibull distribution model to calculate the failure rate of the electronic device at the target time point t.

[0098] Step S4: Establish the transition probability matrix of the absorption Markov chain model based on the failure rate of the electronic devices at the target time point;

[0099] Preferably, the step of establishing the transition probability matrix of the absorbing Markov chain model based on the electronic device failure rate at the target time point further includes:

[0100] The state change process in lifetime testing exhibits absorption characteristics, and the number of failures at the current sampling point is only related to the number of failures at the previous sampling point. Therefore, the lifetime testing process is established as an Absorbing Markov Chain (AMC) model.

[0101] The transition probability matrix C of the absorbing Markov chain model is in the following form:

[0102]

[0103] Wherein, each element P in the transition probability matrix C ij The calculation formula is as follows:

[0104]

[0105] Among them, P ij This represents the probability that j more failure devices will be added after i failure devices exist at the target time point, where i represents the number of existing failure devices, j represents the number of added failure devices, n represents the total number of electronic devices, and F represents the electronic device failure rate at the target time point calculated based on the processed Weibull distribution model.

[0106] It should be noted that when the number of qualified devices is 0, it is called the absorption state, and all others are called the non-absorption state.

[0107] Step S5: Evaluate the lifetime of the electronic device at the target time point based on the transition probability matrix of the absorbed Markov chain model.

[0108] Preferably, the step of evaluating the electronic device lifetime at the target time point based on the transition probability matrix of the absorbed Markov chain model further includes:

[0109] The evaluation equation for the electronic device lifetime at the target time point is E(t). amc as follows:

[0110]

[0111] Where, E(t) amc This represents the lifetime assessment time of electronic devices based on the absorption Markov chain model; A f Indicates the acceleration coefficient; t int The time interval between adjacent data acquisition points within the experimental period is represented by z; z is a constant; n represents the total number of electronic devices; Q represents the probability matrix for transitioning from a non-absorbing state to a non-absorbing state, and the probability matrix Q has the following form:

[0112]

[0113] It should be noted that, given the limited number of data collection points, some probability elements in Q cannot be obtained through experimentation and require the use of the pass / fail equation of the Weibull distribution model, k... wb and m wb The relationships between time and state transition probabilities are derived sequentially by calculating the relationships between each element.

[0114] Specifically, the acceleration coefficient A is calculated using the Arrhenius model. f acceleration coefficient A f The calculation formula is as follows:

[0115]

[0116] Among them, AT A represents the temperature acceleration coefficient. v E represents the voltage acceleration factor. aa T represents the activation energy of the device, k represents the Boltzmann constant, and T represents the activation energy of the device. U The temperature at which the device is used, T A The temperature of the device test is represented by K, the electric field strength determined by the test is represented by X, and the thickness of the stress medium is represented by V. A V represents the accelerating voltage of the device. U This indicates the operating voltage of the device.

[0117] In this embodiment of the invention, the device failure rate is obtained through experimental data, and the failure rate is input into the Weibull distribution model to output m. wb and k wb Then m wb and k wb By fitting the results to time, we obtain m. wb and k wb The relationship between time and the device's lifespan is established. Based on this relationship, the probabilities of various failure states during device testing are calculated, and these probabilities are used as elements to construct an absorbing Markov chain transition matrix. By calculating the acceleration factor, the device's lifetime can be evaluated.

[0118] In specific implementation, such as Figure 2 As shown, first, the normal operating temperature and voltage of the device are determined. Then, the test temperature, test voltage, and test time are determined according to actual needs. Based on the determined test time, data collection points are taken at 24-hour intervals. For tests shorter than 24 hours, data collection points are taken at 1-hour intervals. The end time of the test must be used as a data collection point. If the number of data collection points is odd, an additional data collection point is added one hour before the end of the test. After the test preparation is completed, the test begins. At the data collection points, the test temperature, voltage, and current of the device are recorded. Then, the test is stopped, and the thermal and electrical stresses on the device are removed. The electrical performance of the device is tested within 24 hours, the test results are retained, and the number of failed devices is recorded to calculate the failure rate. When all test data from all data collection points have been collected, the failure rates calculated from different data collection points are input into a polynomial model for fitting to obtain k. wb and m wb Regarding the relationship with time, a Weibull distribution model is established to calculate the transition probabilities between various non-absorption states of the device. Then, the transition probability matrix of the AMC model is constructed. The acceleration factor of the device is obtained through the normal operating temperature and voltage of the device and the test temperature and voltage. Finally, the estimated lifetime of the device is obtained through the AMC model and the acceleration factor.

[0119] The electronic device lifetime assessment method based on absorbing Markov chains provided by this invention can reduce the number of data collection points during the experiment, and at the same time, the model output results can be made consistent with the actual results by fitting the experimental data.

[0120] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of the present invention, and the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A method for evaluating the lifetime of electronic devices based on an absorbing Markov chain, characterized in that, The electronic device lifetime assessment method based on absorbing Markov chains includes: Step S1: Obtain the failure rate of electronic devices at all test collection points during the test period. The failure rate of electronic devices at each test collection point is the number of failed electronic devices at the current test collection point divided by the total number of electronic devices. Step S2: Process the Weibull distribution model based on the electronic device failure rate and the corresponding test acquisition time at all test acquisition points to obtain the processed Weibull distribution model; which also includes: The failure rate of electronic devices at all the test acquisition points and the corresponding test acquisition point time are input into the Weibull distribution model to obtain multiple calculated values ​​of proportional parameters and multiple calculated values ​​of shape parameters; The proportional parameter fitting value of the target time point within the test period is calculated based on the multiple proportional parameter calculation values, and the shape parameter fitting value of the target time point within the test period is calculated based on the multiple shape parameter calculation values. The processed Weibull distribution model is established based on the fitted values ​​of the proportional parameter and the shape parameter at the target time points within the experimental period. Among them, the pass rate equation based on the processed Weibull distribution model as follows: ; The failure rate equation based on the processed Weibull distribution model is then: ,in, t This represents the target time point within the stated test period. m wb This represents the fitted value of the scaling parameter corresponding to the target time point. k wb This represents the fitted value of the shape parameter corresponding to the target time point; This also includes: The proportional parameters are fitted based on the calculated values ​​of the multiple proportional parameters and their corresponding experimental sampling time. m wb The polynomial function model equation, and the shape parameters fitted based on the calculated values ​​of the multiple shape parameters and their corresponding experimental acquisition point times. k wb The polynomial function model equation; The proportional parameter m wb The polynomial function model equation is as follows: ; The shape parameters k wb The polynomial function model equation is as follows: ； in, a 0, a 1, a 2… a r and b 0, b 1, b 2… b r All represent the fitting coefficients. t Indicates the time of the test sampling point; According to the ratio parameter m wb The target time point within the experimental time period was obtained by calculating the polynomial function model equation. t The corresponding proportional parameter fitting value, and based on the shape parameters k wb The target time point within the experimental time period was obtained by calculating the polynomial function model equation. t The corresponding shape parameter fitting values; Step S3: Input the target time point within the test period into the processed Weibull distribution model to calculate the failure rate, and output the electronic device failure rate at the target time point; Step S4: Establish the transition probability matrix of the absorbing Markov chain model based on the electronic device failure rate at the target time point; wherein, it also includes: The transition probability matrix of the absorbed Markov chain model The format is as follows: ; Wherein, the transition probability matrix The elements in The calculation formula is as follows: ; in, This indicates that the target time point exists. i Add after each failed device j The probability of a failed device. i This indicates the number of existing failed devices. j This indicates the number of additional failed devices. n Indicates the total number of electronic components. F This represents the failure rate of electronic devices at the target time point calculated based on the processed Weibull distribution model. Step S5: Evaluate the lifetime of the electronic device at the target time point based on the transition probability matrix of the absorbed Markov chain model.

2. The method for evaluating the lifetime of electronic devices based on an absorbing Markov chain according to claim 1, characterized in that, The step of inputting the target time point within the test period into the processed Weibull distribution model for failure rate calculation, and outputting the electronic device failure rate at the target time point, further includes: target time point t The target time point t Corresponding proportional parameter fitted value m wb and the target time point t Corresponding shape parameter fitted values k wb Substitute the values ​​into the failure rate equation based on the processed Weibull distribution model to calculate the failure rate and obtain the target time point. t Failure rate of electronic devices.

3. The method for evaluating the lifetime of electronic devices based on an absorbing Markov chain according to claim 1, characterized in that, The step of calculating the proportional parameter fitting value of the target time point within the test period based on the multiple proportional parameter calculation values, and calculating the shape parameter fitting value of the target time point within the test period based on the multiple shape parameter calculation values, further includes: The optimization equations for the fitted values ​​of the proportional parameter and the shape parameter at the target time points within the experimental period are as follows: ; in, L It is the optimized fitting value of the proportional parameter or the optimized fitting value of the shape parameter; f These are input values, i.e., calculated values ​​for scale parameters or shape parameters; y i The first time during the test period i The fitted value of the proportional parameter of the first test collection point or the first i Fitted values ​​of shape parameters for each experimental data collection point; R This represents the total number of data collection points during the experimental period.

4. The method for evaluating the lifetime of electronic devices based on an absorbing Markov chain according to claim 1, characterized in that, The step of evaluating the electronic device lifetime at the target time point based on the transition probability matrix of the absorbed Markov chain model further includes: The evaluation equation for the lifetime of electronic devices at the target time point as follows: ; in, E ( t ) amc This represents the lifetime assessment time for electronic devices based on the absorbed Markov chain model. A f Indicates the acceleration coefficient; t int This indicates the time interval between adjacent data collection points within the test period; z It is a constant; n Indicates the total number of electronic components; Q The probability matrix represents the probability of transitioning from a non-absorbing state to a non-absorbing state. Q The format is as follows: .

5. The method for evaluating the lifetime of electronic devices based on an absorbing Markov chain according to claim 4, characterized in that, Acceleration coefficient calculated using the Arrhenius model A f acceleration coefficient A f The calculation formula is as follows: ； in, A T Indicates the temperature acceleration coefficient. A v Indicates the voltage acceleration factor. E aa Indicates the activation energy of the device. k Represents Boltzmann's constant. T U Indicates the operating temperature of the device. T A Indicates the temperature at which the device is tested. K This represents the electric field strength determined by the experiment. X Indicates the thickness of the stress medium. V A Indicates the accelerating voltage of the device. V U This indicates the operating voltage of the device.

Images