Field effect transistor reliability prediction method, device and system

Abstract

The invention belongs to the technical field of field effect transistor performance prediction, and particularly relates to a reliability prediction method, device and system for a field effect transistor, and the method comprises the steps: determining the correlation between the time sequence data of environment parameters, electrical parameters and performance parameters in historical operation records and the change of the performance parameters through a correlation analysis method when the time sequence data changes; key performance parameters are screened out, then a virtual operation model of the field effect transistor is constructed before life prediction is conducted through the screened key performance parameters, and historical data are used for conducting validity verification on the virtual operation model; finally, dynamic changes of environment parameters and electrical parameters are simulated through historical data when the successfully verified virtual operation model is utilized, the degradation trend of key performance parameters is recorded, residual life prediction is carried out based on the degradation trend, prediction efficiency is improved, life prediction can more accurately reflect actual working conditions, and the service life prediction efficiency is improved. And the accuracy of residual life prediction is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of field effect transistor performance prediction, and in particular relates to a reliability prediction method, device and system for a field effect transistor. Background Art

[0002] As a key component in modern electronics, field effect transistors have significant advantages such as high input impedance, low noise figure and fast switching speed. These characteristics make them widely used in fields such as analog circuits and digital circuits. In many applications, especially those industries with strict requirements on safety and reliability (such as aerospace, medical equipment, and automotive industries), the failure of field effect transistors may lead to serious consequences. Therefore, it is crucial to evaluate their service life through reliability prediction.

[0003] A variety of reliability prediction schemes for field effect transistors have been proposed in the prior art. For example, the Chinese invention patent with publication number CN118095182B proposes a reliability prediction method for power field effect transistors, and the specific steps are as follows: S1: collect initial data through sensors and preprocess them; S2: use genetic algorithms to optimize the hyperparameters of the Transformer model and build a GA-Transformer model; S3: input the data obtained in step S1 into the GA-Transformer model built in step S2, evaluate the accuracy of the prediction based on performance indicators, predict the remaining service life, and finally output the reliability prediction results.

[0004] Although the above schemes have shown certain advantages in using advanced models for life prediction, they mainly focus on improving prediction accuracy through model optimization and data processing, but lack the screening and differentiation of key performance indicators. Specifically, field effect transistors have a variety of performance indicators, such as threshold voltage (Vth), on-resistance (Rds(on)), transconductance (gm), etc. However, not all of these performance indicators have a significant impact on the life of the device. If key performance indicators are not screened for prediction, the model will contain a lot of redundant information, resulting in increased model complexity. This will not only consume more computing resources, but may also reduce the operating efficiency and prediction accuracy of the model.

[0005] In addition, even if some performance indicators are considered important in general situations, the importance of these indicators may change under specific actual operating conditions. If the key performance indicators are not selected according to the actual operating conditions for prediction, it is easy to cause deviations in the prediction results and fail to accurately reflect the true health status of the device. Summary of the invention

[0006] The purpose of the present invention is to improve the deficiencies in the prior art and to provide a reliability prediction method, device and system for field effect transistors, which can achieve targeted prediction of the life of field effect transistors by introducing screening of key performance indicators under actual operating conditions.

[0007] The purpose of the present invention can be achieved through the following technical solutions: The first aspect of the present invention provides a reliability prediction method for a field effect transistor, comprising the following steps: Step 1, retrieving the historical operation records of the field effect transistor from the operation database, including the timing data of environmental parameters, power parameters and performance parameters.

[0008] Step 2: Use cross-correlation analysis on historical operation records to determine the correlation between changes in environmental parameters or power parameters and changes in performance parameters, and screen out key performance parameters.

[0009] Step 3: Construct a virtual operation model based on the material characteristics and device structure of the field effect transistor, wherein the input of the virtual operation model is environmental parameters and power parameters, and the output is performance parameters.

[0010] Step 4: Input the environmental parameters and electrical parameter timing data in the historical operation records into the virtual operation model, and verify the effectiveness of the model by comparing the performance parameters output by the model with the corresponding performance parameters in the historical operation records.

[0011] Step 5. When the verification model is valid, use historical data in the virtual operation model to simulate the dynamic changes of environmental parameters and electrical parameters, and record the degradation trend of key performance parameters.

[0012] Step 6: predict the remaining life and accelerated degradation time of the field effect transistor according to the degradation trend, and generate a life assessment report.

[0013] A second aspect of the present invention provides a reliability prediction device for a field effect transistor, which includes a reliability prediction method for a field effect transistor described in the present invention.

[0014] The third aspect of the present invention proposes a reliability prediction system for field effect transistors, including the following modules: a historical operation data retrieval module: retrieves the historical operation records of the field effect transistor from the operation database, including time series data of environmental parameters, power parameters and performance parameters.

[0015] Key performance parameter identification module: Use cross-correlation analysis to determine the correlation between changes in environmental parameters or power parameters and changes in performance parameters in historical operation records, and screen out key performance parameters.

[0016] Virtual operation model construction module: constructs a virtual operation model based on the material characteristics and device structure of the field effect transistor, the input of the virtual operation model is environmental parameters and power parameters, and the output is performance parameters.

[0017] Model validity verification module: input the environmental parameter and electrical parameter timing data in the historical operation record into the virtual operation model, and verify the validity of the model by comparing the performance parameters output by the model with the corresponding performance parameters in the historical operation record.

[0018] Degradation dynamic simulation module: When verifying the validity of the model, historical data is used in the virtual operation model to simulate the dynamic changes of environmental parameters and electrical parameters, and record the degradation trend of key performance parameters.

[0019] Remaining life prediction module: predicts the remaining life and accelerated degradation time of the field effect transistor according to the degradation trend, and generates a life assessment report.

[0020] In combination with all the above-mentioned technical solutions, the positive effects of the present invention are as follows: 1. The present invention analyzes the historical operation records of field effect transistors and uses the correlation analysis method to determine the correlation between changes in environmental parameters, electrical parameters and performance parameters and changes in performance parameters from the timing data of the parameters, screens out key performance parameters, ensures the relevance and effectiveness of the subsequent life prediction model input data, and significantly improves the prediction efficiency and its matching with actual working conditions.

[0021] 2. The present invention constructs a virtual operation model of the field effect transistor before using the screened key performance parameters for life prediction, and uses historical data to verify the effectiveness of the virtual operation model. The model is used for life prediction only when the verification is effective, thereby avoiding prediction deviations caused by inaccurate models and improving the accuracy and credibility of the prediction to a certain extent.

[0022] 3. When using a verified effective virtual operation model, the present invention simulates the dynamic changes of environmental parameters and electrical parameters through historical data, records the degradation trend of key performance parameters, and performs remaining life prediction based on this, so that the life prediction can more accurately reflect the actual working conditions and improve the accuracy of the remaining life prediction.

[0023] 4. The present invention not only predicts the remaining life by using a verified and effective virtual operation model, but also adds a prediction of the accelerated degradation time based on the degradation trend curve. This can detect in advance that the device is about to enter a stage of rapid degradation, provide an early warning signal, and help operation and maintenance personnel take preventive measures in time. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The present invention is further described using the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0025] Figure 1 It is a diagram of the steps for implementing the method of the present invention.

[0026] Figure 2 The present invention is a flowchart for implementing the screening of key performance parameters.

[0027] Figure 3 It is a schematic diagram of system module connection of the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] Example 1

[0030] See also Figure 1 As shown, the present invention provides a reliability prediction method for a field effect transistor, comprising the following steps: Step 1, retrieving the historical operation records of the field effect transistor from an operation database, including time series data of environmental parameters, power parameters and performance parameters.

[0031] It should be noted that the present invention is aimed at predicting the life of field effect transistors in actual use, rather than in the factory state. Manufacturers usually predict the life of field effect transistors in the factory state based on specific ideal conditions, but the actual use conditions may be quite different from these ideal conditions, which makes it difficult to apply the life prediction in the factory state to various complex practical application scenarios. In contrast, the life prediction in the use state can be more practical.

[0032] It should be further explained that the life prediction of the field effect transistor under use status needs to rely on the data of its actual operation, which is obtained from the actual operating environment of the field effect transistor through the data acquisition terminal, and generates detailed operation records, which are stored in the operation database. These records contain the time series data of the environmental parameters, power parameters and performance parameters of the device during the actual operation process.

[0033] The above environmental parameters include but are not limited to temperature, humidity, etc., which reflect the external environmental conditions of the field effect transistor.

[0034] Electrical parameters include but are not limited to voltage, current and load, etc., which describe the electrical conditions of the field effect transistor in actual operation.

[0035] Performance parameters include but are not limited to the following: Threshold voltage (Vth): The voltage that determines the turn-on of the transistor.

[0036] On-resistance (Rds(on)): The resistance of a transistor when it is in the on state.

[0037] Transconductance (gm): An important parameter to measure the amplification capability of a transistor.

[0038] Drain-source breakdown voltage (Vds(max)): The maximum voltage that a transistor can withstand.

[0039] Switching delay time (td(on), td(off)): The time required for a transistor to switch from off to on or from on to off.

[0040] In the actual operation of field effect transistors, environmental parameters and power parameters will have a significant impact on performance parameters. Specifically, the impact of environmental parameters: temperature changes will affect the threshold voltage and on-resistance of the transistor, causing changes in its performance. Other environmental factors such as humidity may also affect the long-term reliability of the device.

[0041] Influence of power parameters: Changes in voltage and current directly affect the working state of the transistor. Excessive voltage may cause breakdown, and excessive current may cause overheating and damage. Load changes will also affect the working efficiency and stability of the transistor.

[0042] It should be added that the application circuit of the field effect transistor is also recorded in the historical operation record. The application circuit information reflects the specific topological structure of the application circuit, including the connection method of the device with other components (such as resistors, capacitors, diodes, etc.) and its position in the network.

[0043] Step 2: Use cross-correlation analysis on historical operation records to determine the correlation between changes in environmental parameters or power parameters and changes in performance parameters, and screen out key performance parameters.

[0044] See also Figure 2 In the preferred implementation of the above scheme, the correlation between the change of environmental parameters or power parameters and the change of performance parameters is determined by correlation analysis. See the following process: the time series data of environmental parameters, power parameters and performance parameters in each historical operation record are respectively plotted in a coordinate system with time as the horizontal axis and environmental parameters, power parameters and performance parameters as the vertical axis to construct an environmental parameter change curve, a power parameter change curve, and a performance parameter change curve.

[0045] It should be emphasized that when using historical operation records to screen key performance parameters, the historical operation records of the same application circuit must be used for screening. This is because different application circuits have unique topologies, operating modes, and electrical stress conditions, which affect the performance and degradation mode of field effect transistors. By using data from the same application circuit, it can be ensured that the screened performance parameters match the operating conditions of the specific circuit and can more accurately reflect the degradation of the device in the actual application scenario.

[0046] The inflection points of the environmental parameter change curve, power parameter change curve and performance parameter change curve in each historical operation record are marked, and the timestamps corresponding to the marked inflection points form an environmental parameter inflection point set, a power parameter inflection point set and a performance parameter inflection point set.

[0047] It should be pointed out that the inflection point in a curve refers to the point where the curve changes from one trend to another, which is usually manifested as a change in slope, reflecting the change in the fluctuation characteristics of the curve at that point.

[0048] The environmental parameter inflection point set and the power parameter inflection point set corresponding to each historical operation record are respectively compared with the performance parameter inflection point set using a cross-correlation function to calculate the cross-correlation coefficient between each performance parameter in the historical operation record and the environmental parameter and power parameter.

[0049] It should be added that when using the inflection point set for cross-correlation analysis, it is necessary to align the inflection points in the inflection point set. This is because aligning the inflection points can ensure that the changes in environmental parameters, power parameters and performance parameters are synchronized in time, avoiding analysis errors caused by time offset.

[0050] As an example of inflection point alignment, missing timestamps are filled using linear interpolation.

[0051] It is further necessary to add that the cross-correlation function is used to calculate the cross-correlation coefficient between two parameter time series to determine whether there is a lagged correlation between them. Specifically, the cross-correlation function can identify whether the change of one parameter will cause the follow-up change of another parameter, thereby revealing the potential temporal relationship between the two. The absolute value of the obtained cross-correlation coefficient reflects the strength of the cross-correlation. When the cross-correlation coefficient is large, it indicates that the cross-correlation strength between the two time series is higher, which means that the change of one parameter has a more significant impact on the other parameter.

[0052] It should be pointed out that the use of a cross-correlation function to calculate the cross-correlation coefficient belongs to the prior art, and its specific calculation steps have been described in detail in relevant literature. In view of this, this article will not repeat the specific details of the calculation process.

[0053] Applied to the present invention, using a cross-correlation function to calculate the cross-correlation coefficient between each performance parameter in the historical operation record and the environmental parameters and power parameters can help identify whether changes in environmental parameters (such as temperature, humidity) or power parameters (such as voltage, current) will cause subsequent changes in performance parameters (such as threshold voltage, on-resistance), and quantify the strength of this correlation.

[0054] In a further preferred implementation of the above scheme, the key performance parameters are screened out as follows: the cross-correlation coefficient between each performance parameter and the environmental parameter and the power parameter in each historical operation record is compared with the set correlation coefficient threshold, and the performance parameters whose cross-correlation coefficients reach the correlation coefficient threshold are screened out and marked as associated performance parameters, and the environmental parameters or power parameters related to them are recorded as sensitive factors.

[0055] As an illustration of the above preferred implementation, comparing the cross-correlation coefficient between each performance parameter and the environmental parameter and the power parameter in each historical operation record with the set correlation coefficient threshold is actually comparing the calculated absolute value of the cross-correlation coefficient with the correlation coefficient threshold.

[0056] In one example, since the cross-correlation coefficient ranges from The purpose of setting the correlation coefficient threshold is to effectively distinguish between strongly correlated and weakly correlated performance parameters. The specific threshold can be set according to actual needs. Generally, a higher correlation coefficient threshold may screen out fewer related performance parameters, but the relationship between these parameters is more significant and reliable. On the contrary, a lower correlation coefficient threshold may screen out more related performance parameters, which can provide more prediction dimensions for subsequent life prediction. For example, it can be 0.6, which can ensure a certain number of related performance parameters while ensuring that the relationship between these parameters has sufficient strength and reliability, thereby providing a solid foundation for subsequent in-depth analysis.

[0057] In a specific embodiment, the cross-correlation coefficients of some performance parameters, some environmental parameters, and some power parameters in a certain historical operation record are shown in Table 1.

[0058] Table 1: Cross-correlation coefficients of some performance parameters and some environmental parameters and power parameters

[0059]

[0060] As can be seen from Table 1, the cross-correlation coefficients of threshold voltage and voltage, threshold voltage and temperature, on-resistance and voltage, transconductance and temperature, and transconductance and voltage all reach the correlation coefficient threshold of 0.6. Therefore, threshold voltage, on-resistance, and transconductance are all used as related performance parameters of the corresponding historical operation records, and the sensitive factors of threshold voltage are voltage and temperature, the sensitive factors of on-resistance are voltage, and the sensitive factors of transconductance are voltage and temperature.

[0061] The associated performance parameters of each historical operation record are classified, and the occurrence ratio of the associated performance parameters is counted and compared with the configured effective occurrence ratio. For example, the effective occurrence ratio is 50%, and the associated performance parameters that reach the effective occurrence ratio are selected as key performance parameters.

[0062] By classifying and aggregating the associated performance parameters in each historical operation record and counting the proportion of each associated performance parameter in all records, performance parameters that appear frequently and have significant correlation can be screened out as key performance parameters. This method avoids the problems of insufficient data representativeness, randomness, noise interference, and statistical bias caused by relying solely on the associated performance parameters in a single historical operation record to determine the key performance parameters, thereby improving the reliability and accuracy of the selection of key performance parameters.

[0063] It should be understood that the present invention identifies the associated performance parameters by analyzing the cross-correlation between the performance parameters and the environmental parameters and electrical parameters in the historical operation records, and determines the key performance parameters accordingly. This is because when a certain performance parameter does not change significantly when the operating state (including environmental parameters and electrical parameters) changes, it means that the performance parameter is not sensitive to these changes, its fluctuation is small, and the corresponding degradation rate is also low. In this case, it is of limited significance to study it as a key performance parameter because its degradation information is not sufficient to provide valuable prediction data; on the contrary, if a certain performance parameter shows a significant change when the operating state changes, it indicates that the performance parameter is more sensitive to changes in the environment and electrical parameters, and its degradation rate is higher and more significant. In this case, studying its degradation trend has a high practical significance and can provide an important reference for subsequent life prediction and maintenance strategies.

[0064] The present invention is a data-driven method for determining key performance parameters by using historical operating data through time series analysis and statistical methods to identify inflection points and calculate cross-correlation coefficients. Compared with pure theoretical derivation or assumption, the key performance parameters determined by this method can more accurately reflect the actual operating conditions.

[0065] Step 3: Construct a virtual operation model based on the material characteristics and device structure of the field effect transistor, wherein the input of the virtual operation model is environmental parameters and power parameters, and the output is performance parameters.

[0066] It should be pointed out that the virtual operation model is a simulation tool based on physical principles and mathematical models, which can simulate the behavior of actual equipment under different working conditions. The virtual operation model is built based on the material properties and device structure of the field effect transistor to ensure that the model can be built based on the real physical principles of the field effect transistor, so as to ensure that the performance of the field effect transistor under various environmental and power conditions can be effectively simulated, and provide a prediction model that is highly close to the actual working conditions for subsequent life prediction.

[0067] It is important to note that the application circuit of the field effect transistor needs to be taken into account when building the virtual operation model to ensure that the model can accurately simulate the behavior and performance of the device in the actual application circuit.

[0068] Step 4: Input the environmental parameters and electrical parameter timing data in the historical operation records into the virtual operation model, and verify the effectiveness of the model by comparing the performance parameters output by the model with the corresponding performance parameters in the historical operation records.

[0069] In the manner in which the above scheme can be implemented, the model validity is verified by the following process: the output performance parameters obtained when the time series data of the environmental parameters and power parameters in each historical operation record are used as the input of the virtual operation model are compared with the performance parameters of the corresponding historical operation record, and the simulation error amount is calculated.

[0070] As an example of the above implementation, the simulation error amount may use a mean square error.

[0071] As another example, the simulation error amount may use the mean absolute error.

[0072] The simulation error of each historical operation record is statistically analyzed to obtain the simulation error mean and simulation error standard deviation.

[0073] It can be understood that the mean of the simulation error reflects the overall error level of the model, and the standard deviation of the simulation error reflects the degree of dispersion of the error, which can be used to evaluate the stability of the model.

[0074] The simulation error mean and simulation error standard deviation are compared with the set acceptable critical values ​​respectively. If the simulation error mean and simulation error standard deviation do not exceed the set acceptable critical values, and if the model error mean and standard deviation are within the acceptable range, the model is considered to have high validity and reliability. At this time, the verification model is valid. Otherwise, the verification model is invalid and it is necessary to consider adjusting the model parameters to improve the accuracy of the model.

[0075] The acceptable critical values ​​corresponding to the simulation error mean and simulation error standard deviation mentioned above are initially set according to the requirements. For example, the acceptable critical value of the simulation error mean is 0.01, and the acceptable critical value of the simulation error standard deviation is 0.05, which are used to determine whether the model meets the accuracy requirements.

[0076] In the improved implementation of the above scheme, Step 4 also includes adjusting the accuracy of the virtual operation model when the verification model is invalid. For specific adjustments, please refer to the following process: when comparing the simulation error mean and the simulation error standard deviation with the set acceptable critical values, if only the simulation error mean exceeds the acceptable critical value, it indicates that there are systematic deviations in the model. At this time, the complexity of the virtual operation model can be increased to improve its fitting ability.

[0077] The specific implementation measures to increase complexity are: 1. Increase the number of layers or nodes: appropriately increase the depth and width of the model to enhance the expressiveness of the model.

[0078] 2. Introduce more features: optimize input features, add new features or improve existing feature construction methods.

[0079] If only the standard deviation of the simulation error exceeds the acceptable critical value, indicating that the model has poor stability, regularization is introduced in the virtual operation model to prevent overfitting and improve the generalization ability of the model.

[0080] The specific implementation measures for introducing regularization are: L1 / L2 regularization: adding L1 or L2 regularization terms to the loss function to constrain the size of model parameters.

[0081] If both the simulation error mean and the simulation error standard deviation exceed the acceptable critical value, indicating that the model has both systematic deviations and lacks stability, the complexity of the virtual operation model is increased and regularization is introduced.

[0082] It should be added that after the accuracy of the virtual operation model is adjusted, the validity verification continues until the adjustment is stopped when the model is verified to be valid.

[0083] The present invention constructs a virtual operation model of the field effect transistor before using the selected key performance parameters for life prediction, and uses historical data to verify the effectiveness of the virtual operation model. Only when the verification is effective, the model is used for life prediction, thereby avoiding the prediction deviation caused by inaccurate model and significantly improving the accuracy and credibility of the prediction. When the verification is invalid, by identifying the specific reasons why the model error does not meet the acceptable critical value, the model accuracy is adjusted in a targeted manner to optimize the model performance and provide reliable support for subsequent life prediction.

[0084] Step 5. When the verification model is valid, use historical data in the virtual operation model to simulate the dynamic changes of environmental parameters and electrical parameters, and record the degradation trend of key performance parameters.

[0085] The specific implementation process of the above steps is as follows: extract all environmental parameters and electrical parameters from the historical operation records, and calculate the maximum and minimum values ​​of each parameter respectively as its fluctuation range.

[0086] According to the sensitive factors corresponding to each key performance parameter, the corresponding sensitive factors are obtained from the environmental fluctuation range and the electrical fluctuation range, and these sensitive factors are used as independent variables, other non-sensitive factors are used as fixed quantities, and the key performance parameters are used as dependent variables to form a simulation group.

[0087] In the above implementation example, assuming that the key performance parameters are threshold voltage, on-resistance, and transconductance, and the sensitive factors of threshold voltage are voltage and temperature, the sensitive factors of on-resistance are voltage, and the sensitive factors of transconductance are voltage and temperature, the simulation group formed at this time is shown in Table 2.

[0088] Table 2: Examples of simulation groups

[0089]

[0090] The minimum and maximum values ​​within the fluctuation range of the sensitive factors in the simulation group are combined with the safety margin to define the boundaries of the simulation interval. For the fixed quantities in the simulation group, the normal values ​​within their fluctuation range are used as the simulation values.

[0091] For the example applied to Table 2, assume that the voltage fluctuation range in simulation group 1 is ,in , They represent the minimum and maximum values ​​of the voltage within the corresponding fluctuation range. The safety margin is , then the boundary of the simulation interval is , .

[0092] It should be understood that the main purpose of setting the safety margin is to ensure the safety and reliability of the simulation results and to cover potential extreme conditions. Specifically, the safety margin can provide additional buffer space within the simulation interval to prevent model failure or prediction deviation caused by actual operating conditions exceeding the expected range. By increasing or decreasing a certain percentage of the safety margin based on the minimum and maximum values ​​in the historical data, it can cover those extreme conditions that may occur but have not been recorded, thereby improving the robustness of the model. In short, by introducing a safety margin, the model's adaptability to uncertainty and variability factors can be enhanced to ensure that it can provide reliable prediction results under various conditions. For example, a 5% safety margin can be set.

[0093] It should also be understood that the normal value is selected as the simulation value for the fixed quantity because the fixed quantity has little impact on the key performance parameters. On the one hand, the use of the normal value can better represent the central trend of the data, avoid deviations caused by individual extreme data points, and help improve the stability of the model; on the other hand, it can reflect the normal performance under actual working conditions, so that the model prediction is based on actual normal data.

[0094] In a specific implementation, the normal value can be obtained by drawing a probability distribution diagram (such as a histogram) based on a fixed amount of historical data. If the data is normally distributed, the mean can be selected as the normal value; if the data has skewness or outliers, the median can be selected as the normal value.

[0095] The step size is set within the simulation interval to conduct multiple simulation experiments, so that the sensitive factors are gradually increased according to the set step size within the simulation interval. The choice of the step size should take into account the computing resources and accuracy requirements. Too small a step size may lead to excessively high computing costs, while too large a step size may miss important degradation trends.

[0096] It should be added that the premise of using historical data to simulate the dynamic changes of environmental parameters and electrical parameters is that the historical data should cover a variety of typical working conditions and extreme conditions. If the data only contains records under certain specific conditions, the simulation results may be incomplete. In addition, the time span of the data should be long enough to cover the performance of the equipment under different seasons and operating conditions.

[0097] During the dynamic simulation of each simulation group, the output results of key performance parameters are recorded, and degradation simulation curves are drawn based on these results.

[0098] The degradation simulation curves mentioned above show the trends of key performance parameters changing with sensitive factors.

[0099] A mathematical model is selected to fit the degradation simulation curve of the key performance parameter corresponding to each simulation group, and the key performance parameter value at a future time point is extrapolated using the fitting model as the future trend value of the key performance parameter value.

[0100] Specifically, the selection of mathematical model fitting requires data analysis and preliminary observation of the degradation simulation curve. Common degradation modes include linear, exponential, power function, polynomial, etc.

[0101] Linear degradation: If the key performance parameters show a linear relationship with time or sensitive factors, you can consider using a linear regression model (such as ).

[0102] Exponential degradation: If the rate of change of a key performance parameter increases or decreases over time, an exponential model (such as ).

[0103] Power function degradation: In some cases, the variation of key performance parameters follows a power law relationship, and a power function model (such as ).

[0104] Polynomial regression: When the data exhibits complex nonlinear changes, you can consider using a polynomial regression model ( ).in Indicates sensitive factors, Indicates key performance parameters, , represents the parameter to be estimated.

[0105] Once a suitable mathematical model is selected and fitted, the model can be used to extrapolate the key performance parameter values ​​at future time points. The basic principle of extrapolation is to substitute future sensitive factors into the equations of the fitted model and calculate the corresponding key performance parameter values.

[0106] The specific extrapolation steps are: determine the specific values ​​of sensitive factors at future time points based on demand.

[0107] Substitute the values ​​of sensitive factors at different future time points into the equation of the fitting model to calculate the corresponding predicted values ​​of key performance parameters.

[0108] Assume that an exponential degradation model is fitted through the degradation simulation curve to describe the change of the threshold voltage of the field effect transistor with temperature. The specific fitting model is: , For temperature.

[0109] The threshold voltage prediction under this exponential degradation model is shown in Table 3.

[0110] Table 3: Threshold voltage predictions for some future time periods

[0111]

[0112] It can be seen from the table that as the temperature rises, the threshold voltage of the field effect transistor shows a downward trend, reflecting the performance degradation of the field effect transistor at higher temperatures.

[0113] It should be added that in order to use the fitting model to accurately extrapolate the key performance parameter values ​​at future time points, the values ​​of sensitive factors at future time points must be obtained first. The accuracy of these sensitive factor values ​​directly determines the reliability of the extrapolation results. An effective way to obtain the values ​​of sensitive factors at future time points is to analyze historical data, identify trends, cyclical and seasonal changes, and establish an appropriate regression model to describe the relationship between sensitive factors and time.

[0114] Step 6: Predict the remaining life of the field effect transistor according to the degradation trend and generate a reliability assessment report.

[0115] Specifically, the remaining life prediction in the above solution includes the following process: setting the failure threshold of the key performance parameter according to the technical specifications of the field effect transistor.

[0116] The fitting model is used to solve the time point when the key performance parameter reaches the failure threshold as the life aging time point corresponding to the parameter.

[0117] The above-mentioned life aging time point refers to the critical time point when the key performance parameter reaches failure, reflecting the expiration of the aging of the key performance parameter.

[0118] It is further supplemented that the life aging time point of the key performance parameter is combined with the current time to obtain the remaining life of the key performance parameter.

[0119] The confidence interval of the remaining life is calculated based on the statistical distribution model for each life aging time point corresponding to the key performance parameter.

[0120] In the preferred implementation of the above scheme, the confidence region of the remaining life is calculated in the following steps: a. Select a suitable statistical distribution model (such as normal distribution, Weibull distribution) to describe the probability distribution of the remaining life.

[0121] b. Estimate the parameters of the selected distribution model (such as mean and standard deviation) based on historical data.

[0122] c. Use the estimated parameters to calculate the confidence interval of the remaining life.

[0123] In the above implementation example, it is assumed that the Weibull distribution is selected to describe the probability distribution of the remaining life, and its probability density function is ,in Indicates time, is the scale parameter, is the shape parameter, Represents a natural constant.

[0124] Estimate using maximum likelihood estimation method using historical data and , based on which the confidence interval for the remaining life is calculated.

[0125] It should be pointed out that the calculation method of the confidence interval is well known in the industry and is part of the prior art, so the specific details thereof will not be repeated in the present invention.

[0126] It is important to understand that identifying the life aging time point of a key performance parameter does give a specific prediction value, but there are certain limitations in relying solely on this single time point. Calculating the confidence interval of the remaining life has the following important functions: 1. Quantifying uncertainty: There is a certain degree of uncertainty in the data and models in the actual system. These uncertainties may come from measurement errors, inaccurate model assumptions, environmental changes and other factors. By calculating the confidence interval, this uncertainty can be quantified. Understanding the range of uncertainty can help better conduct risk assessment and management. For example, if the confidence interval is wide, it means that the uncertainty of the prediction result is large, and the maintenance plan needs to be formulated more carefully.

[0127] 2. Improve prediction reliability: A single life span prediction may be affected by outliers or extreme situations, while the confidence interval provides a range that makes the prediction more robust. Even if some data points are biased, the overall prediction result is still highly credible.

[0128] In further implementation of the solution, the degradation time is accelerated as follows: a degradation trend curve is drawn based on future trend values ​​of key performance parameter values.

[0129] The degradation rates at different time points are calculated on the degradation trend curves of the key performance parameters drawn.

[0130] By comparing the degradation rates at different time points, the time point at which the degradation rate increases significantly is identified as the accelerated degradation time point.

[0131] It is important to know that the time point when the degradation rate increases significantly is the time point when the equipment enters the accelerated degradation stage from the stable degradation stage, indicating that the performance of the equipment begins to decline rapidly. Identifying this time point can serve as an early warning signal to remind the operation and maintenance personnel that maintenance is about to be performed to avoid sudden failure.

[0132] In further implementation of the scheme, the life assessment report includes the following contents: S1, the life aging time points of the field effect transistor under each key performance parameter, and the corresponding remaining life confidence interval and accelerated degradation time points.

[0133] S2. The arrangement results of key performance parameters are specifically ranked as follows: S21. Weights are assigned to key performance parameters based on the width of the confidence interval of the remaining life of the field effect transistor under each key performance parameter.

[0134] The specific weight assignment formula is: , where Indicates The weight values ​​of key performance parameters, Indicates the number of the key performance parameter, , Indicates that the field effect transistor is The width of the remaining life confidence interval under key performance parameters, Indicates the longest confidence interval width among the remaining life confidence interval widths of all key performance parameters.

[0135] The principle of the above weight assignment formula is that the wider the confidence interval width is, the greater the uncertainty of the prediction result is, that is, our confidence in the authenticity of the prediction result is lower. It reflects the proportion of the width of the confidence interval of the remaining life of each key performance parameter. The larger the proportion, the greater the uncertainty of the prediction result, and the smaller the corresponding weight should be. Therefore, the weight value obtained by subtracting this proportion from the value 1 can better reflect the true importance and priority of each key performance parameter.

[0136] S22. Calculate the importance of the life aging time points of each key performance parameter in combination with its weight, and arrange them in descending order of importance.

[0137] In the implementation of the above scheme, since the closer the life aging time point is, the greater the importance is, the life aging time point needs to be properly processed in setting the importance calculation, such as processing the life aging time point as a negative exponent. For example, , Indicates the lifespan aging time point, which is scored for importance with the assigned weight value after processing.

[0138] By sorting the key performance parameters, it is possible to quickly identify which key performance parameters will reach the failure threshold in a short period of time. This helps to prioritize those parameters that are about to fail, so as to take timely maintenance or replacement measures.

[0139] It should be emphasized that the present invention mainly performs single-factor prediction on the life of field effect transistors under different key performance parameters, and can independently analyze the impact of each key performance parameter on the life, reducing the complexity of the model, making the calculation easier, and at the same time being able to obtain the life prediction results of each key performance parameter in a relatively short time, facilitating quick decision-making, and more importantly, can formulate targeted maintenance plans, give priority to those key performance parameters that are close to failure, and reduce unnecessary maintenance costs. Of course, multi-factor prediction can be introduced to provide a more comprehensive evaluation perspective, and the prediction model can be flexibly adjusted according to different application scenarios and needs, providing customized solutions to meet diverse needs.

[0140] Example 2

[0141] The present invention provides a reliability prediction device for a field effect transistor, and the device comprises a reliability prediction method for a field effect transistor described in the present invention.

[0142] Example 3

[0143] Reference Figure 3 As shown, the present invention proposes a reliability prediction system for field effect transistors, including the following modules: a historical operation data retrieval module: retrieves historical operation records of field effect transistors from an operation database, including time series data of environmental parameters, power parameters and performance parameters.

[0144] The key performance parameter identification module is connected to the historical operation data retrieval module and is used to determine the correlation between changes in environmental parameters or power parameters and changes in performance parameters through cross-correlation analysis of historical operation records, and to screen out key performance parameters.

[0145] Virtual operation model building module: building a virtual operation model based on the material characteristics and device structure of the field effect transistor, wherein the input of the virtual operation model is environmental parameters and power parameters, and the output is performance parameters;

[0146] The model validity verification module is connected to the virtual operation model construction module, and is used to input the environmental parameter and electrical parameter timing data in the historical operation record into the virtual operation model, and verify the validity of the model by comparing the performance parameters output by the model with the corresponding performance parameters in the historical operation record.

[0147] The degradation dynamic simulation module is connected to the model validity verification module and the key performance parameter identification module respectively, and is used to simulate the dynamic changes of environmental parameters and electrical parameters in the virtual operation model using historical data when verifying the validity of the model, and record the degradation trend of the key performance parameters;

[0148] The remaining life prediction module is connected to the degradation dynamic simulation module and is used to predict the remaining life and accelerated degradation time of the field effect transistor according to the degradation trend and generate a life assessment report.

[0149] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A reliability prediction method for a field effect transistor, characterized in that: The following steps are involved: Step 1, retrieve the historical operation records of the field effect transistor from the operation database, including the time series data of environmental parameters, power parameters and performance parameters; Step 2: Use cross-correlation analysis to determine the correlation between changes in environmental parameters or power parameters and changes in performance parameters, and select key performance parameters; Step 3, constructing a virtual operation model based on the material characteristics and device structure of the field effect transistor, wherein the input of the virtual operation model is environmental parameters and power parameters, and the output is performance parameters; Step 4, inputting the environmental parameters and electrical parameter time series data in the historical operation records into the virtual operation model, and verifying the effectiveness of the model by comparing the performance parameters output by the model with the corresponding performance parameters in the historical operation records; Step 5. When the verification model is effective, use historical data to simulate the dynamic changes of environmental parameters and electrical parameters in the virtual operation model, and record the degradation trend of key performance parameters; Step 6: predict the remaining life and accelerated degradation time of the field effect transistor according to the degradation trend, and generate a life assessment report.

2. A reliability prediction method for a field effect transistor as claimed in claim 1, characterized in that: The content of Step 2 is as follows: The time series data of environmental parameters, power parameters and performance parameters in each historical operation record are respectively plotted into an environmental parameter change curve, a power parameter change curve and a performance parameter change curve in a coordinate system with time as the horizontal axis and environmental parameters, power parameters and performance parameters as the vertical axis; Mark inflection points of the environmental parameter change curve, power parameter change curve, and performance parameter change curve in each historical operation record, and form an environmental parameter inflection point set, a power parameter inflection point set, and a performance parameter inflection point set with the timestamps corresponding to the marked inflection points; The environmental parameter inflection point set and the power parameter inflection point set corresponding to each historical operation record are respectively combined with the performance parameter inflection point set using a cross-correlation function to calculate the cross-correlation coefficient between each performance parameter in the historical operation record and the environmental parameter and the power parameter; Compare the cross-correlation coefficient between each performance parameter and the environmental parameter and the power parameter in each historical operation record with the set correlation coefficient threshold, select the performance parameters whose cross-correlation coefficient reaches the correlation coefficient threshold and mark them as associated performance parameters, and record the environmental parameters or power parameters related to them as sensitive factors; The associated performance parameters of each historical operation record are classified, the occurrence ratio of the associated performance parameters is counted, and compared with the configured effective occurrence ratio, and the associated performance parameters that reach the effective occurrence ratio are selected as key performance parameters.

3. A reliability prediction method for a field effect transistor according to claim 1, characterized in that: The validation of the model by comparing the performance parameters output by the model with the corresponding performance parameters in the historical operation records is as follows: The output performance parameters obtained when the time series data of the environmental parameters and power parameters in each historical operation record are used as the input of the virtual operation model are compared with the performance parameters of the corresponding historical operation record to calculate the simulation error; The simulation error amount of each historical operation record is statistically analyzed to obtain the simulation error mean and simulation error standard deviation; The simulation error mean and simulation error standard deviation are compared with the set acceptable critical value respectively. If the simulation error mean and simulation error standard deviation do not exceed the set acceptable critical value, the verification model is valid, otherwise the verification model is invalid.

4. A reliability prediction method for a field effect transistor as claimed in claim 3, characterized in that: The Step 4 also includes adjusting the accuracy of the virtual operation model when the verification model is invalid. For specific adjustments, refer to the following process: When comparing the simulation error mean and simulation error standard deviation with the set acceptable critical value, if only the simulation error mean exceeds the acceptable critical value, the complexity of the virtual operation model is increased; If only the standard deviation of the simulation error exceeds the acceptable critical value, regularization is introduced in the virtual operation model; If both the simulation error mean and the simulation error standard deviation exceed the acceptable critical value, the complexity of the virtual operation model is increased and regularization is introduced.

5. The reliability prediction method of a field effect transistor according to claim 1, characterized in that: The specific implementation of Step 5 is as follows: Extract all environmental parameters and electrical parameters from historical operation records, and calculate the maximum and minimum values ​​of each parameter as its fluctuation range; According to the sensitive factors corresponding to each key performance parameter, the corresponding sensitive factors are obtained from the environmental fluctuation range and the electrical fluctuation range, and these sensitive factors are used as independent variables, other non-sensitive factors are used as fixed quantities, and the key performance parameters are used as dependent variables to form a simulation group; The minimum and maximum values ​​of the fluctuation range of the sensitive factors in the simulation group combined with the safety margin are defined as the boundaries of the simulation interval. For the fixed quantity of the simulation group, the normal value within its fluctuation range is used as the simulation value. The step length is set within the simulation interval so that the sensitive factors are gradually increased according to the set step length within the simulation interval; During the dynamic simulation of each simulation group, the output results of key performance parameters are recorded, and degradation simulation curves are drawn based on these results; A mathematical model is selected to fit the degradation simulation curve of the key performance parameter corresponding to each simulation group, and the key performance parameter value at a future time point is extrapolated using the fitting model as the future trend value of the key performance parameter value.

6. A reliability prediction method for a field effect transistor as claimed in claim 5, characterized in that: The predicting of the remaining life of the field effect transistor according to the degradation trend comprises the following process: Setting failure thresholds of key performance parameters based on technical specifications of field effect transistors; The fitting model is used to solve the time point when the key performance parameter reaches the failure threshold as the life aging time point corresponding to the parameter; The confidence interval of the remaining life is calculated based on the statistical distribution model for each life aging time point corresponding to the key performance parameter.

7. A reliability prediction method for a field effect transistor as claimed in claim 6, characterized in that: The accelerated degradation time is predicted as follows: Draw a degradation trend curve based on the future trend values ​​of key performance parameter values; Calculate the degradation rate at different time points on the degradation trend curve of the key performance parameters drawn; By comparing the degradation rates at different time points, the time point at which the degradation rate increases significantly is identified as the accelerated degradation time point.

8. A reliability prediction method for a field effect transistor as claimed in claim 7, characterized in that: The life assessment report includes the following: S1. Life aging time points of field effect transistors under various key performance parameters, as well as the corresponding remaining life confidence intervals and accelerated degradation time points; S2. The arrangement results of key performance parameters are as follows: S21, assigning weights to the key performance parameters based on the width of the confidence interval of the remaining life of the field effect transistor under each key performance parameter; S22. Calculate the importance of the life aging time points of each key performance parameter in combination with its weight, and arrange them in descending order of importance.

9. A reliability prediction device for a field effect transistor, characterized in that: The device comprises a reliability prediction method for a field effect transistor as described in any one of claims 1-8.

10. A reliability prediction system for a field effect transistor, characterized in that: Includes the following modules: Historical operation data retrieval module: retrieves the historical operation records of the field effect transistor from the operation database, including the time series data of environmental parameters, power parameters and performance parameters; Key performance parameter identification module: Use cross-correlation analysis to determine the correlation between changes in environmental parameters or power parameters and changes in performance parameters, and select key performance parameters; Virtual operation model building module: building a virtual operation model based on the material characteristics and device structure of the field effect transistor, wherein the input of the virtual operation model is environmental parameters and power parameters, and the output is performance parameters; Model validity verification module: inputting the environmental parameters and electrical parameter time series data in the historical operation records into the virtual operation model, and verifying the validity of the model by comparing the performance parameters output by the model with the corresponding performance parameters in the historical operation records; Degradation dynamic simulation module: When verifying the model is valid, historical data is used in the virtual operation model to simulate the dynamic changes of environmental parameters and electrical parameters, and record the degradation trend of key performance parameters; Remaining life prediction module: predicts the remaining life and accelerated degradation time of the field effect transistor according to the degradation trend, and generates a life assessment report.

Images