Power electronic component reliability prediction method based on accelerated life model

Abstract

The invention discloses a power electronic component reliability prediction method based on an accelerated life model, and relates to the technical field of electric data processing. The method comprises the following steps: comprehensively analyzing an influence evaluation coefficient of environmental acceleration stress on life quality prediction of the power electronic component and an influence evaluation coefficient of mechanical acceleration stress on life quality prediction of the power electronic component to obtain an accuracy evaluation coefficient in life quality prediction of the power electronic component. According to the method, the accuracy evaluation coefficient in the power electronic component life quality prediction is compared with the accuracy evaluation threshold in the power electronic component life quality prediction, so that the power electronic component life quality prediction is finally optimized according to the comparison result; the effect of improving the reliability prediction accuracy of the power electronic component based on the accelerated life model is achieved, and the problem of inaccurate reliability prediction of the power electronic component based on the accelerated life model in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical data processing, and particularly to a reliability prediction method for power electronic components based on an accelerated life model. Background Art

[0002] With the wide application of power electronic devices in fields such as industry, transportation, and medical treatment, their reliability issues have received increasing attention. Traditional reliability analysis of power electronic components is mostly based on statistical models, such as exponential distribution and Weibull distribution, to describe the relationship between the failure rate of components and time. Since the lifespan of power electronic components is often very long, it is not practical to directly conduct life tests. Accelerated life tests accelerate the failure process of components by increasing stress levels (such as temperature, voltage, etc.), so as to obtain reliability information in a shorter time.

[0003] Existing Stress-Life (S-N) models achieve reliability prediction by establishing the relationship between stress levels and lifespan. Existing multi-physical field coupling analysis comprehensively considers the influence of multiple factors such as heat, electricity, and magnetism on the reliability of components.

[0004] For example, a method for predicting the lifespan of a leakage signal conditioning circuit disclosed in the invention patent with the publication number of CN116029234A includes: taking the leakage signal conditioning circuit as the research object, first obtaining the key components that affect the overall output performance of the circuit through sensitivity analysis, using the Monte-Carlo method to analyze the influence of the tolerances of the components on the performance of the leakage signal conditioning circuit, and obtaining the conclusion that the smaller the tolerances of the components and the higher the accuracy, the higher the reliability of the circuit operation; secondly, establishing a performance degradation model of the leakage signal conditioning circuit based on the Wiener process, and using the Arrhenius equation to extrapolate the predicted lifespan under accelerated stress to the predicted lifespan under normal stress, so as to obtain the predicted lifespan of the leakage signal conditioning circuit at room temperature.

[0005] For example, a method for evaluating an accelerated life prediction model of components disclosed in the invention patent with the publication number of CN116090190A includes: S1. Establishing a reliability model R*(t) using all data; setting T1, T2, T3,..., TS as the accelerated stresses of the product; S2. Dividing the accelerated stresses in a ratio of S-1:1, and dividing the data into training data and test data; S3. Establishing a reliability model according to the training data in each combination, deriving the reliability function corresponding to the stress of the test data, denoted as R(t); separately conducting reliability modeling using the test data, and setting the obtained reliability function as the standard reliability function, denoted as ; S4. Comparing R(t) of all combinations with If the errors are all within the given threshold range, the model is accurate.

[0006] However, in the process of implementing the inventive technical solution in the embodiments of the present application, it is found that the above technologies have at least the following technical problems: In the prior art, under different stress conditions, the failure mechanisms of the dominant power electronic components may be different. If it is an accelerated stress, it may cause the prediction method to be unable to accurately and reliably reflect the prediction result, resulting in inaccurate reliability prediction of power electronic components based on the accelerated life model. Summary of the Invention

[0007] By providing a reliability prediction method for power electronic components based on the accelerated life model, the embodiments of the present application solve the problem of inaccurate reliability prediction of power electronic components based on the accelerated life model in the prior art, and achieve the effect of improving the accuracy of reliability prediction of power electronic components based on the accelerated life model.

[0008] The embodiments of the present application provide a reliability prediction method for power electronic components based on an accelerated life model, including the following steps: collecting environmental acceleration stress-related data for power electronic component life quality prediction and mechanical acceleration stress-related data for power electronic component life quality prediction through sensors and databases respectively, obtaining accelerated life experiment-related data through experiments, preprocessing the environmental acceleration stress-related data for power electronic component life quality prediction, the mechanical acceleration stress-related data for power electronic component life quality prediction, and the accelerated life experiment-related data to obtain the preprocessed environmental acceleration stress-related data for power electronic component life quality prediction, the mechanical acceleration stress-related data for power electronic component life quality prediction, and the accelerated life experiment-related data; conducting an accelerated life experiment on power electronic component samples, modeling the accelerated life experiment-related data through statistical analysis methods to obtain an accelerated life model; analyzing the environmental acceleration stress-related data for power electronic component life quality prediction to obtain an influence evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction, analyzing the mechanical acceleration stress-related data for power electronic component life quality prediction to obtain an influence evaluation coefficient of mechanical acceleration stress on power electronic component life quality prediction, comprehensively analyzing the influence evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction and the influence evaluation coefficient of mechanical acceleration stress on power electronic component life quality prediction to obtain an accuracy evaluation coefficient in power electronic component life quality prediction; obtaining an influence evaluation threshold of environmental acceleration stress on power electronic component life quality prediction and an accuracy evaluation threshold in power electronic component life quality prediction from the database, comparing the influence evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction with the influence evaluation threshold of environmental acceleration stress on power electronic component life quality prediction, and preliminarily optimizing the power electronic component life quality prediction according to the comparison result, comparing the accuracy evaluation coefficient in power electronic component life quality prediction with the accuracy evaluation threshold in power electronic component life quality prediction, and finally optimizing the power electronic component life quality prediction according to the comparison result.

[0009] Further, the specific process of conducting an accelerated life experiment on power electronic component samples is as follows: Selecting temperature, humidity, and vibration as the acceleration stresses of the experiment according to the failure mechanism of the power electronic component samples, designing the stress application method and stress change law, monitoring the experimental parameters of the samples, and recording the failure time of the samples or the time to reach the predetermined performance degradation standard.

[0010] Further, the specific modeling process of modeling the accelerated life experiment-related data through statistical analysis methods is as follows: Selecting the Arrhenius model as the model for the accelerated life experiment, preprocessing the accelerated life experiment-related data, and using statistical analysis methods to estimate the model parameters to obtain an accelerated life model.

[0011] Furthermore, the specific process for preprocessing the environmental acceleration stress-related data for power electronic component life quality prediction, the mechanical acceleration stress-related data for power electronic component life quality prediction, and the accelerated life test-related data is as follows: The preprocessing includes data cleaning and data transformation; data cleaning includes filling missing values, deleting outliers and duplicate values; data transformation: normalizing the environmental acceleration stress-related data for power electronic component life quality prediction, the mechanical acceleration stress-related data for power electronic component life quality prediction, and the accelerated life test-related data.

[0012] Furthermore, the specific analysis process for analyzing the environmental acceleration stress-related data for power electronic component life quality prediction is as follows: The environmental acceleration stress-related data for power electronic component life quality prediction includes neutron flux, temperature acceleration factor, Boltzmann constant, number of humidity failures under reference stress, material constant, reference stress range, environmental oxygen concentration, activation energy, and stress range. Process the neutron flux, temperature acceleration factor, Boltzmann constant, number of humidity failures under reference stress, material constant, reference stress range, environmental oxygen concentration, activation energy, and stress range to obtain the impact evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction.

[0013] Furthermore, the specific analysis process for analyzing the mechanical acceleration stress-related data for power electronic component life quality prediction is as follows: The mechanical acceleration stress-related data for power electronic component life quality prediction includes vibration frequency, shock duration, and bending moment. Process the vibration frequency, shock duration, and bending moment to obtain the impact evaluation coefficient of mechanical acceleration stress on power electronic component life quality prediction.

[0014] Furthermore, the specific comparison process for comparing the impact evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction with the impact evaluation threshold of environmental acceleration stress on power electronic component life quality prediction is as follows: Compare the impact evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction with the impact evaluation threshold of environmental acceleration stress on power electronic component life quality prediction. If the impact evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction is greater than or equal to the impact evaluation threshold of environmental acceleration stress on power electronic component life quality prediction, preliminarily optimize the power electronic component life quality prediction method. If the impact evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction is less than the impact evaluation threshold of environmental acceleration stress on power electronic component life quality prediction, mark that the life prediction of this power electronic component sample is not affected by environmental acceleration stress.

[0015] Further, the specific comparison process of comparing the accuracy evaluation coefficient in the life quality prediction of power electronic components with the accuracy evaluation threshold in the life quality prediction of power electronic components is as follows: Compare the accuracy evaluation coefficient in the life quality prediction of power electronic components with the accuracy evaluation threshold in the life quality prediction of power electronic components. If the accuracy evaluation coefficient in the life quality prediction of power electronic components is greater than or equal to the accuracy evaluation threshold in the life quality prediction of power electronic components, mark that the life prediction of the power electronic component sample is accurate. If the accuracy evaluation coefficient in the life quality prediction of power electronic components is less than the accuracy evaluation threshold in the life quality prediction of power electronic components, deeply optimize the life quality prediction method of power electronic components.

[0016] Further, the specific optimization process of finally optimizing the life quality prediction of power electronic components according to the comparison result is as follows: The final optimization includes using PID (Proportional-Integral-Derivative) controller software to adjust the heater and cooling device, using an integrated control software platform to simultaneously control multiple environmental stresses and mechanical stresses, allowing engineers to remotely monitor the test process and control hardware devices through remote access software, and performing fault diagnosis and early warning through the software system when the stress control is abnormal.

[0017] Further, the specific method for obtaining the accuracy evaluation coefficient in the life quality prediction of power electronic components is as follows: ; In the formula, is expressed as the accuracy evaluation coefficient in the life quality prediction of power electronic components, represents the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components, represents the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components, represents the material degradation rate of power electronic components, represents the processing cooling rate of power electronic components, represents the weight factor of the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components, represents the weight factor of the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components, represents the weight factor of the material degradation rate of power electronic components, represents the weight factor of the processing cooling rate of power electronic components.

[0018] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. By comparing the accuracy evaluation coefficient in the life quality prediction of power electronic components with the accuracy evaluation threshold in the life quality prediction of power electronic components, the life quality prediction of power electronic components is finally optimized according to the comparison result, thereby achieving the effect of improving the reliability prediction accuracy of power electronic components based on the accelerated life model, and effectively solving the problem of inaccurate reliability prediction of power electronic components based on the accelerated life model in the prior art.

[0019] 2. By conducting an accelerated life experiment on power electronic component samples, relevant data of the accelerated life experiment is modeled through statistical analysis methods to obtain an accelerated life model, thereby achieving the effect of shortening the reliability prediction time of power electronic components.

[0020] 3. By comprehensively analyzing the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components and the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components, the accuracy evaluation coefficient in the life quality prediction of power electronic components is obtained, thereby achieving the effect of improving the comprehensiveness of the reliability prediction of power electronic components. Description of the Drawings

[0021] Figure 1 It is a flowchart of the reliability prediction method for power electronic components based on the accelerated life model provided by the embodiment of the present application; Figure 2 It is an image of the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components in the reliability prediction method for power electronic components based on the accelerated life model provided by the embodiment of the present application. Detailed Embodiment

[0022] The embodiment of the present application provides a reliability prediction method for power electronic components based on the accelerated life model, which solves the problem of inaccurate reliability prediction of power electronic components based on the accelerated life model in the prior art. By comparing the accuracy evaluation coefficient in the life quality prediction of power electronic components with the accuracy evaluation threshold in the life quality prediction of power electronic components, the life quality prediction of power electronic components is finally optimized according to the comparison result, achieving the effect of improving the reliability prediction accuracy of power electronic components based on the accelerated life model.

[0023] The technical solution in the embodiment of the present application for solving the above problem of inaccurate reliability prediction of power electronic components based on the accelerated life model has the following general idea: By preprocessing the environmental acceleration stress-related data for power electronic component life quality prediction, the mechanical acceleration stress-related data for power electronic component life quality prediction, and the accelerated life test-related data, conducting an accelerated life test on power electronic component samples, modeling the accelerated life test-related data through statistical analysis methods to obtain an accelerated life model, comprehensively analyzing the influence evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction and the influence evaluation coefficient of mechanical acceleration stress on power electronic component life quality prediction to obtain the accuracy evaluation coefficient in power electronic component life quality prediction, comparing the accuracy evaluation coefficient in power electronic component life quality prediction with the accuracy evaluation threshold in power electronic component life quality prediction, and finally optimizing the power electronic component life quality prediction according to the comparison result, the effect of improving the reliability prediction accuracy of power electronic components based on the accelerated life model is achieved.

[0024] To better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings of the specification and specific implementation manners.

[0025] As Figure 1As shown in the figure, it is a flowchart of a reliability prediction method for power electronic components based on an accelerated life model provided by an embodiment of the present application. The method includes the following steps: Collect relevant data on environmental acceleration stress for power electronic component life quality prediction and relevant data on mechanical acceleration stress for power electronic component life quality prediction through sensors and databases respectively. Obtain relevant data on accelerated life experiments through experiments. Preprocess the relevant data on environmental acceleration stress for power electronic component life quality prediction, the relevant data on mechanical acceleration stress for power electronic component life quality prediction, and the relevant data on accelerated life experiments to obtain the preprocessed relevant data on environmental acceleration stress for power electronic component life quality prediction, the relevant data on mechanical acceleration stress for power electronic component life quality prediction, and the relevant data on accelerated life experiments; Conduct accelerated life experiments on power electronic component samples, and construct a model for the relevant data on accelerated life experiments through statistical analysis methods to obtain an accelerated life model; Analyze the relevant data on environmental acceleration stress for power electronic component life quality prediction to obtain an influence evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction. Analyze the relevant data on mechanical acceleration stress for power electronic component life quality prediction to obtain an influence evaluation coefficient of mechanical acceleration stress on power electronic component life quality prediction. Conduct a comprehensive analysis of the influence evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction and the influence evaluation coefficient of mechanical acceleration stress on power electronic component life quality prediction to obtain an accuracy evaluation coefficient in power electronic component life quality prediction; Obtain an influence evaluation threshold of environmental acceleration stress on power electronic component life quality prediction and an accuracy evaluation threshold in power electronic component life quality prediction from the database. Compare the influence evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction with the influence evaluation threshold of environmental acceleration stress on power electronic component life quality prediction, and preliminarily optimize the power electronic component life quality prediction according to the comparison result. Compare the accuracy evaluation coefficient in power electronic component life quality prediction with the accuracy evaluation threshold in power electronic component life quality prediction, and finally optimize the power electronic component life quality prediction according to the comparison result.

[0026] Further, the specific process of conducting accelerated life experiments on power electronic component samples is as follows: Select temperature, humidity, and vibration as the acceleration stresses of the experiment according to the failure mechanism of the power electronic component samples, design the stress application method and stress change law, monitor the experimental parameters of the samples, and record the failure time of the samples or the time to reach the predetermined performance degradation standard.

[0027] In this embodiment, the power electronic component sample is placed in a thermostatic and humidistatic test chamber. The temperature range is set from -40°C to 150°C, with 50°C as one step, and it is divided into five temperature levels in total. On the basis of the temperature stress, the relative humidity range is set from 20% to 90%, with 20% as one step, and it is divided into four humidity levels in total. The power electronic component sample is fixed on a vibration table, and the vibration frequency range is set from 10 Hz to 50 Hz, with 10 Hz as one step, and it is divided into five frequency levels in total. During the experiment, parameters such as the voltage, current, power, and temperature of the power electronic component sample are monitored in real time to ensure the accuracy of the experimental data. The first threshold and the second threshold of the voltage are obtained from the database. If the real-time voltage is less than the first threshold and greater than the second threshold, it is determined that the power electronic component reaches the predetermined performance degradation standard. If the real-time voltage is less than the second threshold, it is determined that the power electronic component fails. The first threshold and the second threshold of the current are obtained from the database. If the real-time current is less than the first threshold and greater than the second threshold, it is determined that the power electronic component reaches the predetermined performance degradation standard. If the real-time current is less than the second threshold, it is determined that the power electronic component fails. The first threshold and the second threshold of the power are obtained from the database. If the real-time power is less than the first threshold and greater than the second threshold, it is determined that the power electronic component reaches the predetermined performance degradation standard. If the real-time power is less than the second threshold, it is determined that the power electronic component fails. The first threshold and the second threshold of the temperature are obtained from the database. If the real-time temperature is less than the first threshold and greater than the second threshold, it is determined that the power electronic component reaches the predetermined performance degradation standard. If the real-time temperature is less than the second threshold, it is determined that the power electronic component fails. For example, with 20 Hz as a reference, the time when the sample reaches the predetermined performance degradation standard when the real-time voltage is less than the first threshold and greater than the second threshold is preferably recorded.

[0028] Further, the specific modeling process of modeling the data related to the accelerated life experiment by statistical analysis methods is as follows: Select the Arrhenius model as the model for the accelerated life experiment, preprocess the data related to the accelerated life experiment, and use statistical analysis methods to estimate the model parameters to obtain the accelerated life model.

[0029] In this embodiment, the statistical methods include the least squares method and Bayesian estimation.

[0030] The least squares method estimates the model parameters by minimizing the sum of the squares of the differences between the predicted values and the actual observed values. The specific steps are as follows: Calculate the sum of the squares of the differences between the predicted values and the actual observed values to obtain the residual sum of squares, find the model parameters that minimize the residual sum of squares, and finally record the model parameters as the parameters of the Arrhenius equation.

[0031] Bayesian estimation is a method of estimating parameters by using Bayes' theorem to combine prior information and sample data.

[0032] Taking the natural logarithm of the Arrhenius model gives a linear form. Experimental data is collected and a linear regression model is constructed using the experimental data. The form of the linear regression model is y = kx + b, and the model parameters in the linear regression are calculated using the least squares method.

[0033] A prior distribution is set for the model parameters. Based on the experimental data and the prior distribution, the posterior distribution is calculated through Bayes' theorem, and the estimated values of the parameters are obtained through data analysis of the posterior distribution.

[0034] The prior distribution refers to the mathematical representation of the model parameters before observing any experimental data, which can be obtained from a database. The posterior distribution refers to the mathematical representation of the model parameters after considering the experimental data, and is calculated through Bayes' theorem. The specific formula is: ; and the estimated values of the parameters can be extracted by calculating the mean or median of the posterior distribution as the estimated values of the parameters.

[0035] Furthermore, the specific process for preprocessing the environmental acceleration stress-related data for power electronic component life quality prediction, the mechanical acceleration stress-related data for power electronic component life quality prediction, and the accelerated life test-related data is as follows: The preprocessing includes data cleaning and data transformation; data cleaning includes filling missing values, deleting outliers and duplicate values; data transformation: normalizing the environmental acceleration stress-related data for power electronic component life quality prediction, the mechanical acceleration stress-related data for power electronic component life quality prediction, and the accelerated life test-related data.

[0036] In this embodiment, there are missing vibration frequencies in 5 power electronic component samples, and it is decided to fill these missing values with the average value of this feature. It is found that the environmental oxygen concentration of 10 samples is much higher than that of other samples, and these may be measurement errors, so they are deleted. It is found that the data of 10 samples are completely repeated, and these repeated samples are deleted. Data cleaning ensures the integrity and accuracy of the data set, and reduces the impact of noise and errors on the model.

[0037] Furthermore, the specific analysis process for analyzing the environmental acceleration stress-related data for power electronic component life quality prediction is as follows: The environmental acceleration stress-related data for power electronic component life quality prediction includes neutron flux, temperature acceleration factor, Boltzmann constant, number of humidity failures under reference stress, material constant, reference stress range, environmental oxygen concentration, activation energy, and stress range. The neutron flux, temperature acceleration factor, Boltzmann constant, number of humidity failures under reference stress, material constant, reference stress range, environmental oxygen concentration, activation energy, and stress range are processed to obtain the impact evaluation coefficient of environmental acceleration stress on power electronic component life quality prediction.

[0038] In this embodiment, the specific method for obtaining the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components is as follows: ; ; ; ; In the formula, represents the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components, represents the temperature cycle index, represents the weight factor of the temperature cycle index, represents the humidity cycle index, represents the weight factor of the humidity cycle index, represents the environmental oxygen concentration, represents the weight factor of the environmental oxygen concentration, represents the neutron flux, represents the weight factor of the neutron flux, represents the temperature acceleration factor, represents the activation energy, represents the Boltzmann constant, represents the number of humidity failure times under the reference stress, represents the material constant, represents the stress range, represents the reference stress range.

[0039] In this embodiment, the weight factor of the temperature cycle index, the weight factor of the humidity cycle index, the weight factor of the environmental oxygen concentration, and the weight factor of the neutron flux respectively represent the numerical values of the influence degrees of the temperature cycle index, the humidity cycle index, the environmental oxygen concentration, and the neutron flux on the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components.

[0040] In the process of implementing the reliability prediction of power electronic components, several key indicators are directly extracted from a dedicated database - the weight factors corresponding to the temperature cycle index, humidity cycle index, ambient oxygen concentration, and neutron flux. The determination of these weight factors depends on a complete mapping system. For example, various monitored indicator values in real time, such as the temperature cycle index, humidity cycle index, etc., are converted into corresponding weight factors. Specifically, this conversion process is based on the following mapping logic: after real-time data is input, through the conversion of the mapping set, the corresponding weight factors are output. For example, when inputting the real-time temperature cycle index, humidity cycle index, ambient oxygen concentration, and neutron flux, through the conversion of the mapping set, the weight factors corresponding to the temperature cycle index, humidity cycle index, ambient oxygen concentration, and neutron flux are output. This mapping may be a direct correspondence, or it may be a complex mapping where one indicator value corresponds to multiple weight factors. In short, this is a process of converting real-time data indicators into the weight values of the key factors affecting the reliability prediction of components, and its conversion rules can be either a simple direct mapping or a more complex one-to-many mapping.

[0041] High temperature, high humidity, high oxygen concentration, and high radiation flux may produce a synergistic effect, that is, when these factors act on the component together, the resulting damage and degradation rate may be greater than the sum of the effects of each factor acting alone. The increase in temperature will accelerate the corrosion effects of humidity and oxygen, the increase in humidity will promote the dissolution of oxygen and chemical reactions, and the increase in oxygen concentration will accelerate the oxidation process.

[0042] Neutron flux refers to the number of neutrons passing through a unit area per unit time and is usually used to describe the radiation environment. It can be measured by radiation monitoring equipment.

[0043] The temperature acceleration factor refers to the degree of influence of temperature on the aging rate of components. In this embodiment, it is usually calculated by the Arrhenius equation. The activation energy and Boltzmann constant need to be determined through experiments. The specific process of calculating using the Arrhenius equation is as follows: determine the reaction rate constants at 10°C and 20°C, apply these two temperatures to the Arrhenius equation, calculate the temperature acceleration factor, and substitute the reaction rate constants at 10°C and 20°C. The reaction rate constant refers to the rate at which a component fails (or a specific failure mechanism, such as corrosion, fatigue, electromigration, etc.) at a certain temperature and can be directly obtained from the database.

[0044] The Boltzmann constant is a physical constant used to convert temperature to energy. Its value is fixed, approximately 1.380649*10^-22 J / K, and can be obtained from the database.

[0045] The number of humidity failures under the reference stress refers to the number of times a component fails in an accelerated life test under specific humidity conditions, which can be obtained from a database.

[0046] The material constant refers to the constant of a specific material in an accelerated life test, which can be obtained from a database.

[0047] The reference stress range refers to the range of stress levels for comparing and standardizing test results, which can be obtained from a database.

[0048] The ambient oxygen concentration refers to the concentration of oxygen in the surrounding environment. It can be measured using a gas analysis instrument.

[0049] The activation energy refers to the energy required for a chemical reaction or physical process to occur. It can be obtained from a database.

[0050] The stress range refers to the range of stresses (such as temperature, humidity, mechanical stress, etc.) applied to a component, which can be obtained from a database.

[0051] Furthermore, the specific analysis process for analyzing the mechanical acceleration stress-related data for predicting the life quality of power electronic components is as follows: The mechanical acceleration stress-related data for predicting the life quality of power electronic components includes vibration frequency, shock duration, and bending moment. The vibration frequency, shock duration, and bending moment are processed to obtain the influence evaluation coefficient of mechanical acceleration stress on the prediction of the life quality of power electronic components.

[0052] In this embodiment, the specific method for obtaining the influence evaluation coefficient of mechanical acceleration stress on the prediction of the life quality of power electronic components is as follows: ; In the formula, represents the influence evaluation coefficient of mechanical acceleration stress on the prediction of the life quality of power electronic components, represents the vibration frequency, represents the shock duration, represents the bending moment, represents the weight factor of the vibration frequency, represents the weight factor of the shock duration, represents the weight factor of the bending moment.

[0053] An increase in the vibration frequency will exacerbate the fatigue damage of the component due to resonance. An extension of the shock duration will enhance the destructive effect of the instantaneous force on the component structure, while an increase in the bending moment will directly cause the component to bear greater mechanical stress.

[0054] The vibration frequency refers to the number of vibrations per second of a component under vibration stress. A frequency meter or vibration analyzer is used to measure the vibration frequency of the component in the actual working or test environment.

[0055] The impact duration refers to the length of time during which the impact stress acts on the component and is obtained using an impact testing machine.

[0056] The bending moment refers to the product of the force acting on the component and the distance from its acting point, and a torque sensor or torque measuring instrument is used to directly measure the torque applied to the component.

[0057] In a specific embodiment, the data example of the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components is as follows in the table.

[0058] Table 1 Data Example Table of Influence Evaluation Coefficient of Mechanical Acceleration Stress on Life Quality Prediction of Power Electronic Components

[0059] When the weight factors of vibration frequency, impact duration, and bending moment are 0.01, 0.59, and 0.4 respectively, as Figure 2 shown, it is the image of the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components in the power electronic component reliability prediction method based on the accelerated life model provided by the embodiment of the present application. Through Figure 2 and the data in Table 1, it can be seen that when the vibration frequency and impact duration are fixed, the greater the bending moment, the greater the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components.

[0060] During the process of implementing the reliability prediction of power electronic components, several key indicators - the weight factors of vibration frequency, impact duration, and bending moment - are directly extracted from a dedicated database. The determination of these weight factors depends on a complete mapping system. For example, the values of various indicators monitored in real time, such as vibration frequency, impact duration, and bending moment, are converted into corresponding weight factors. Specifically, this conversion process is based on the following mapping logic: after real-time data is input, through the conversion of the mapping set, the corresponding weight factors are output. For example, when real-time vibration frequency, impact duration, and bending moment are input into the mapping set for conversion, the weight factors of vibration frequency, impact duration, and bending moment are output. This mapping may be a direct correspondence relationship or a complex mapping where one indicator value corresponds to multiple weight factors. In short, this is a process of converting real-time data indicators into weight values of key factors affecting component reliability prediction, and its conversion rules can be either a simple direct mapping or a more complex one-to-many mapping.

[0061] Further, the specific comparison process of comparing the influence evaluation coefficient of environmental accelerated stress on the life quality prediction of power electronic components with the influence evaluation threshold of environmental accelerated stress on the life quality prediction of power electronic components is as follows: Compare the influence evaluation coefficient of environmental accelerated stress on the life quality prediction of power electronic components with the influence evaluation threshold of environmental accelerated stress on the life quality prediction of power electronic components. If the influence evaluation coefficient of environmental accelerated stress on the life quality prediction of power electronic components is greater than or equal to the influence evaluation threshold of environmental accelerated stress on the life quality prediction of power electronic components, preliminarily optimize the life quality prediction method of power electronic components. If the influence evaluation coefficient of environmental accelerated stress on the life quality prediction of power electronic components is less than the influence evaluation threshold of environmental accelerated stress on the life quality prediction of power electronic components, mark that the life prediction of this power electronic component sample is not affected by environmental accelerated stress.

[0062] In this embodiment, the preliminary optimization includes re - analyzing data, optimizing stress levels, and improving test plans.

[0063] Re - analyzing data means re - examining experimental data and statistical analysis results, identifying possible data anomalies or deviations, and ensuring data quality and analysis accuracy.

[0064] Optimizing stress levels means re - evaluating and adjusting the stress levels applied to the components, increasing or decreasing the stress levels, and increasing or decreasing the environmental accelerated stress and mechanical accelerated stress by 10% according to the original values.

[0065] Improving test plans means modifying test plans, including test time, test cycles, and stress application methods, and increasing the number of samples or the number of test repetitions.

[0066] Further, the specific comparison process of comparing the accuracy evaluation coefficient in the life quality prediction of power electronic components with the accuracy evaluation threshold in the life quality prediction of power electronic components is as follows: Compare the accuracy evaluation coefficient in the life quality prediction of power electronic components with the accuracy evaluation threshold in the life quality prediction of power electronic components. If the accuracy evaluation coefficient in the life quality prediction of power electronic components is greater than or equal to the accuracy evaluation threshold in the life quality prediction of power electronic components, mark that the life prediction of this power electronic component sample is accurate. If the accuracy evaluation coefficient in the life quality prediction of power electronic components is less than the accuracy evaluation threshold in the life quality prediction of power electronic components, deeply optimize the life quality prediction method of power electronic components.

[0067] In this embodiment, if the accuracy evaluation coefficient of life prediction is only 0.75, while the accuracy evaluation threshold is 0.85, it indicates that the prediction result is not precise enough, and the life quality prediction method of power electronic components needs to be deeply optimized. If the accuracy evaluation coefficient of life prediction is only 0.9, while the accuracy evaluation threshold is 0.85, it is marked that the life prediction of this power electronic component sample is precise.

[0068] Furthermore, the specific optimization process for the final optimization of the life quality prediction of power electronic components according to the comparison results is as follows: The final optimization includes using PID (Proportional-Integral-Derivative) controller software to adjust the heater and cooling device, using an integrated control software platform to control multiple environmental stresses and mechanical stresses simultaneously, allowing engineers to remotely monitor the test process and control hardware devices through remote access software, and performing fault diagnosis and warning through the software system when stress control is abnormal.

[0069] In this embodiment, the PID controller software is used to precisely control the heater and cooling device to maintain a constant temperature or temperature cycle during the test. The PID controller software ensures that the temperature control accuracy is within ±0.5°C by real-time monitoring of temperature changes and adjusting the heating or cooling output. An integrated control software platform is used to control environmental stresses such as temperature, humidity, and vibration, as well as mechanical stresses applied through a vibration table simultaneously. The software platform allows the operator to set complex stress profiles, such as temperature cycles, humidity changes, and the frequency and amplitude of vibration. During the test, the heater of a power module suddenly fails, resulting in abnormal temperature control. The software system immediately detects this abnormality and notifies the engineer through the warning system.

[0070] Furthermore, the specific method for obtaining the accuracy evaluation coefficient in the life quality prediction of power electronic components is as follows: ; where represents the accuracy evaluation coefficient in the life quality prediction of power electronic components, represents the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components, represents the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components, represents the material degradation rate of power electronic components, represents the processing cooling rate of power electronic components, represents the weight factor of the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components, represents the weight factor of the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components, represents the weight factor of the material degradation rate of power electronic components, represents the weight factor of the processing cooling rate of power electronic components.

[0071] In this embodiment, the weight factors of the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components and the weight factors of the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components respectively represent the numerical values of the influence degrees of the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components and the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components on the accuracy evaluation coefficient in the life quality prediction of power electronic components. The processing cooling rate refers to the rate at which the material of the power electronic component cools from a high temperature state to room temperature during the life prediction process, which can be obtained through sensors in the accelerated life experiment. The material degradation rate refers to the speed at which the material performance of the power electronic component degrades over time under specific working conditions, which can be obtained through the accelerated life experiment.

[0072] During the process of implementing the reliability prediction of power electronic components, several key indicators are directly extracted from a dedicated database - the weight factor of the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components, the weight factor of the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components, the weight factor of the material degradation rate of power electronic components, and the weight factor of the processing cooling rate of power electronic components. The determination of these weight factors depends on a complete mapping system. For example, various monitored index values, such as the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components, the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components, the material degradation rate of power electronic components, and the processing cooling rate of power electronic components, are converted into corresponding weight factors. Specifically, this conversion process is based on the following mapping logic: after real-time data is input, through the conversion of the mapping set, the corresponding weight factors are output. For example, when the real-time influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components, the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components, the material degradation rate of power electronic components, and the processing cooling rate of power electronic components are input into the mapping set for conversion, the weight factor of the influence evaluation coefficient of environmental acceleration stress on the life quality prediction of power electronic components, the weight factor of the influence evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components, the weight factor of the material degradation rate of power electronic components, and the weight factor of the processing cooling rate of power electronic components are output. This mapping may be a direct correspondence relationship, or it may be a complex mapping where one index value corresponds to multiple weight factors. In short, this is a process of converting real-time data indicators into weight values of key factors affecting component reliability prediction, and its conversion rules can be either simple direct mapping or more complex one-to-many mapping.

[0073] Those skilled in the art will understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0074] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0075] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing devices to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0076] These computer program instructions can also be loaded onto a computer or other programmable data processing devices, such that a series of operation steps are executed on the computer or other programmable devices to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable devices provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0077] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they know the basic creative concepts. Therefore, the appended claims are intended to be construed to include the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0078] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims

1. A reliability prediction method for power electronic components based on an accelerated life model, characterized in that: The following steps are involved: Collect environmental accelerated stress related data for life quality prediction of power electronic components and mechanical accelerated stress related data for life quality prediction of power electronic components through sensors and databases, obtain accelerated life experiment related data through experiments, preprocess the environmental accelerated stress related data for life quality prediction of power electronic components, mechanical accelerated stress related data for life quality prediction of power electronic components and accelerated life experiment related data, and obtain the preprocessed environmental accelerated stress related data for life quality prediction of power electronic components, mechanical accelerated stress related data for life quality prediction of power electronic components and accelerated life experiment related data; Carry out accelerated life test on power electronic component samples, construct relevant data of accelerated life test by statistical analysis method, and obtain accelerated life model; Analyze the environmental accelerated stress related data of life quality prediction of power electronic components to obtain the impact assessment coefficient of environmental accelerated stress on life quality prediction of power electronic components. Analyze the mechanical accelerated stress related data of life quality prediction of power electronic components to obtain the impact assessment coefficient of mechanical accelerated stress on life quality prediction of power electronic components. Comprehensively analyze the impact assessment coefficient of environmental accelerated stress on life quality prediction of power electronic components and the impact assessment coefficient of mechanical accelerated stress on life quality prediction of power electronic components to obtain the accuracy assessment coefficient in life quality prediction of power electronic components. The impact assessment threshold of environmental accelerated stress on the life quality prediction of power electronic components and the accuracy assessment threshold of power electronic components life quality prediction are obtained from the database, the impact assessment coefficient of environmental accelerated stress on the life quality prediction of power electronic components is compared with the impact assessment threshold of environmental accelerated stress on the life quality prediction of power electronic components, and the life quality prediction of power electronic components is preliminarily optimized according to the comparison results, the accuracy assessment coefficient in the life quality prediction of power electronic components is compared with the accuracy assessment threshold in the life quality prediction of power electronic components, and the life quality prediction of power electronic components is finally optimized according to the comparison results.

2. The reliability prediction method of power electronic components based on the accelerated life model according to claim 1, characterized in that: The specific process of conducting accelerated life test on power electronic component samples is as follows: According to the failure mechanism of power electronic component samples, temperature, humidity and vibration are selected as the accelerated stress of the experiment, the stress application method and stress change law are designed, the experimental parameters of the samples are monitored, and the failure time of the samples or the time to reach the predetermined performance degradation standard is recorded.

3. The reliability prediction method of power electronic components based on the accelerated life model according to claim 1, characterized in that: The specific modeling process of modeling the accelerated life test related data by the statistical analysis method is as follows: The Arrhenius model was selected as the model of accelerated life test, the relevant data of accelerated life test were preprocessed, and the model parameters were estimated by using statistical analysis method to obtain the accelerated life model.

4. The reliability prediction method of power electronic components based on the accelerated life model according to claim 1, characterized in that: The specific process of preprocessing the environmental accelerated stress related data for life quality prediction of power electronic components, the mechanical accelerated stress related data for life quality prediction of power electronic components, and the accelerated life experiment related data is as follows: Preprocessing includes data cleaning and data transformation; Data cleaning includes filling missing values, removing outliers and duplicate values; Data conversion: Normalize the environmental accelerated stress related data for life quality prediction of power electronic components, the mechanical accelerated stress related data for life quality prediction of power electronic components, and the accelerated life experiment related data.

5. The reliability prediction method of power electronic components based on the accelerated life model according to claim 1, characterized in that: The specific analysis process of analyzing the environmental accelerated stress related data for life quality prediction of power electronic components is as follows: The environmental accelerated stress related data for life quality prediction of power electronic components include neutron flux, temperature acceleration factor, Boltzmann constant, number of humidity failures under reference stress, material constant, reference stress range, ambient oxygen concentration, activation energy and stress range. The neutron flux, temperature acceleration factor, Boltzmann constant, number of humidity failures under reference stress, material constant, reference stress range, ambient oxygen concentration, activation energy and stress range are processed to obtain the impact evaluation coefficient of environmental accelerated stress on life quality prediction of power electronic components.

6. The reliability prediction method of power electronic components based on the accelerated life model according to claim 1, characterized in that: The specific analysis process of analyzing the mechanical acceleration stress related data for life quality prediction of power electronic components is as follows: The mechanical acceleration stress related data for life quality prediction of power electronic components include vibration frequency, impact duration and bending moment. The vibration frequency, impact duration and bending moment are processed to obtain the impact evaluation coefficient of mechanical acceleration stress on life quality prediction of power electronic components.

7. The reliability prediction method of power electronic components based on the accelerated life model according to claim 1, characterized in that: The specific comparison process of comparing the impact assessment coefficient of environmental accelerated stress on the life quality prediction of power electronic components with the impact assessment threshold of environmental accelerated stress on the life quality prediction of power electronic components is as follows: The impact assessment coefficient of environmental accelerated stress on the life quality prediction of power electronic components is compared with the impact assessment threshold of environmental accelerated stress on the life quality prediction of power electronic components. If the impact assessment coefficient of environmental accelerated stress on the life quality prediction of power electronic components is greater than or equal to the impact assessment threshold of environmental accelerated stress on the life quality prediction of power electronic components, the life quality prediction method of power electronic components is preliminarily optimized. If the impact assessment coefficient of environmental accelerated stress on the life quality prediction of power electronic components is less than the impact assessment threshold of environmental accelerated stress on the life quality prediction of power electronic components, it is marked that the life prediction of the power electronic component sample is not affected by environmental accelerated stress.

8. The reliability prediction method of power electronic components based on the accelerated life model as claimed in claim 1, characterized in that: The specific comparison process of comparing the accuracy assessment coefficient in the life quality prediction of power electronic components with the accuracy assessment threshold in the life quality prediction of power electronic components is as follows: The accuracy assessment coefficient in the life quality prediction of power electronic components is compared with the accuracy assessment threshold in the life quality prediction of power electronic components. If the accuracy assessment coefficient in the life quality prediction of power electronic components is greater than or equal to the accuracy assessment threshold in the life quality prediction of power electronic components, the life prediction of the power electronic component sample is marked as accurate. If the accuracy assessment coefficient in the life quality prediction of power electronic components is less than the accuracy assessment threshold in the life quality prediction of power electronic components, the life quality prediction method of power electronic components is deeply optimized.

9. The reliability prediction method of power electronic components based on the accelerated life model according to claim 1, characterized in that: The specific optimization process of finally optimizing the life quality prediction of power electronic components according to the comparison results is as follows: Final optimizations included using PID (proportional-integral-derivative) controller software to adjust heaters and coolers, using an integrated control software platform to simultaneously control multiple environmental and mechanical stresses, remote access software that allows engineers to remotely monitor the test process and control hardware equipment, and software systems to perform fault diagnosis and early warning when stress control is abnormal.

10. The reliability prediction method of power electronic components based on the accelerated life model according to claim 1, characterized in that: The specific method for obtaining the accuracy evaluation coefficient in the life quality prediction of power electronic components is as follows: ; In the formula, Expressed as the accuracy evaluation coefficient in the life quality prediction of power electronic components, It represents the impact assessment coefficient of environmental accelerated stress on the life quality prediction of power electronic components. It is expressed as the evaluation coefficient of the impact of mechanical acceleration stress on the life quality prediction of power electronic components, Expressed as the material degradation rate of power electronic components, Expressed as the processing cooling rate of power electronic components, It is expressed as the weight factor of the impact assessment coefficient of environmental accelerated stress on the life quality prediction of power electronic components, It is expressed as the weight factor of the impact evaluation coefficient of mechanical acceleration stress on the life quality prediction of power electronic components, Expressed as a weighting factor for the material degradation rate of power electronic components, Expressed as a weighting factor for the processing cooling rate of power electronic components.

Images