A dynamic evaluation method and system for aging performance of high-voltage silicon carbide MOS

Abstract

The present invention discloses a method and system for dynamically evaluating the aging performance of a high-voltage silicon carbide MOS, and relates to the technical field of semiconductor performance evaluation. The method comprises: performing dynamic performance testing on a high-voltage silicon carbide MOS to obtain a plurality of performance parameter sets; obtaining a leakage current matching information set, a breakdown voltage matching information set, and a switch loss matching information set, and constructing a leakage current special degree set, a breakdown voltage special degree set, and a switch special degree set; obtaining a plurality of parameter utility information, and configuring a plurality of aging analysis computing powers; obtaining a plurality of aging performance evaluation results and aging performance variation coefficients as aging performance dynamic evaluation results. The method solves the technical problem that the existing silicon carbide MOS aging performance evaluation cannot effectively integrate the dynamic changes of a plurality of performance parameters, resulting in difficulty in comprehensively and accurately evaluating the actual aging state of the device and predicting the aging risk of the device, and achieves the technical effect of improving the comprehensiveness and accuracy of the aging performance evaluation.

Description

Technical Field

[0001] The present application relates to the technical field related to semiconductor performance evaluation, and specifically to a method and system for dynamically evaluating the aging performance of a high-voltage silicon carbide MOS. Background Art

[0002] With the rapid development of power electronics technology, silicon carbide (SiC) power devices, especially high-voltage silicon carbide MOSFETs (metal oxide semiconductor field effect transistors), have been widely used in high-efficiency power conversion, electric vehicles, photovoltaic power generation, and smart grids due to their excellent electrical properties and advantages such as high voltage and high temperature resistance. Although silicon carbide MOS has excellent performance, its reliability and aging problems are still urgent problems to be solved in actual applications. Power devices will age with the increase of usage time under long-term high voltage, high temperature and high frequency working environment, resulting in gradual performance degradation and even failure. Traditional aging performance evaluation methods usually rely on offline testing or accelerated life testing in laboratory environments. Although they can evaluate the aging of devices to a certain extent, due to the large difference between test conditions and actual application scenarios, it is difficult to accurately reflect the aging performance of devices in real working environments. In addition, with the complexity of the working environment of power devices and the diversification of application scenarios, traditional analysis based on a single parameter cannot effectively integrate the dynamic changes of multiple performance parameters, and it is difficult to comprehensively and systematically evaluate the overall aging status of devices.

[0003] Therefore, at this stage, the relevant technologies for evaluating the aging performance of high-voltage silicon carbide MOS have the technical problem of being unable to effectively integrate the dynamic changes of multiple performance parameters, resulting in difficulty in comprehensively and accurately evaluating the actual aging status of the device and predicting the aging risk of the device. Summary of the invention

[0004] The present application provides a method and system for dynamically evaluating the aging performance of a high-voltage silicon carbide MOS, thereby solving the technical problem that the existing silicon carbide MOS aging performance evaluation cannot effectively integrate the dynamic changes of multiple performance parameters, resulting in difficulty in comprehensively and accurately evaluating the actual aging state of the device and predicting the aging risk of the device, thereby achieving the technical effect of improving the comprehensiveness and accuracy of the aging performance evaluation.

[0005] The present application provides a method for dynamically evaluating the aging performance of a high-voltage silicon carbide MOS, the method comprising: performing dynamic performance tests on the high-voltage silicon carbide MOS in multiple working stages to obtain multiple performance parameter sets, dividing and obtaining an on-resistance set, a gate leakage current set, a breakdown voltage set, and a switching loss set, wherein each performance parameter set includes an on-resistance, a gate leakage current, a breakdown voltage, and a switching loss; performing a test matching analysis on the data in the gate leakage current set, the breakdown voltage set, and the switching loss set according to the on-resistance set, obtaining a leakage current matching information set, a breakdown voltage set, and a switching loss set; The voltage matching information set and the switch loss matching information set are obtained, and matching special degree analysis is performed to construct a leakage current special degree set, a breakdown voltage special degree set and a switch special degree set; according to the leakage current special degree set, the breakdown voltage special degree set and the switch special degree set, parameter utility fusion is performed on the multiple performance parameter sets to obtain multiple parameter utility information, and multiple aging analysis computing powers are configured; according to the multiple aging analysis computing powers, aging performance analysis is performed on the multiple performance parameter sets to obtain multiple aging performance evaluation results and aging performance variation coefficients as aging performance dynamic evaluation results.

[0006] In a possible implementation, dynamic performance tests are performed on the high-voltage silicon carbide MOS in multiple working stages to obtain multiple performance parameter sets, which are divided into an on-resistance set, a gate leakage current set, a breakdown voltage set, and a switching loss set. The following processing is also performed: multiple accumulated working times of the high-voltage silicon carbide MOS are obtained as multiple working stages; in the multiple working stages, dynamic performance tests are performed on the high-voltage silicon carbide MOS to obtain multiple performance parameter sets, which are divided into an on-resistance set, a gate leakage current set, a breakdown voltage set, and a switching loss set.

[0007] In a possible implementation, a test matching analysis of the data in the gate leakage current set, the breakdown voltage set and the switching loss set is performed according to the on-resistance set, and the following processing is also performed: a pre-training test matching analyzer is performed, wherein the test matching analyzer includes a leakage current matching analysis branch, a breakdown voltage matching analysis branch and a switching loss matching analysis branch; the first on-resistance in the first working stage in the on-resistance set is combined with the first gate leakage current in the gate leakage current set, and the combination is input into the leakage current matching analysis branch to perform a test matching analysis, and the first leakage current matching information is obtained as an output; and the test matching analysis of the data in the gate leakage current set, the breakdown voltage set and the switching loss set is continued to be performed according to the on-resistance set to obtain a leakage current matching information set, a breakdown voltage matching information set and a switching loss matching information set.

[0008] In a possible implementation, the pre-trained test matching analyzer also performs the following processing: based on the historical performance test data of the high-voltage silicon carbide MOS, a sample on-resistance set, a sample gate leakage current set, a sample breakdown voltage set and a sample switching loss set are collected; based on the standard aging test data of the high-voltage silicon carbide MOS, a standard gate leakage current set, a standard breakdown voltage set and a standard switching loss set corresponding to multiple sample on-resistances in the sample on-resistance set are obtained, and the deviations of the sample gate leakage current set, the sample breakdown voltage set and the sample switching loss set from the standard values ​​are calculated, and the sample leakage current matching information set, the sample breakdown voltage matching information set and the sample switching loss matching information set are obtained by calculation and annotation; the sample on-resistance set and the sample gate leakage current set are used as training inputs, and the sample leakage current matching information set is used as training outputs, and the leakage current matching analysis branch is trained, and the breakdown voltage matching analysis branch and the switching loss matching analysis branch are trained; the leakage current matching analysis branch, the breakdown voltage matching analysis branch and the switching loss matching analysis branch are combined to obtain a test matching analyzer.

[0009] In a possible implementation, a matching specificity analysis is performed to construct a leakage current specificity set, a breakdown voltage specificity set and a switch specificity set, and the following processing is also performed: the first leakage current matching information of the first working stage is selected in the leakage current matching information set, and the mean of the leakage current matching information set is calculated to obtain the average leakage current matching information; the ratio of the average leakage current matching information to the first leakage current matching information is calculated as the first leakage current specificity; the leakage current specificity set is continued to be calculated and constructed based on the leakage current matching information set; matching specificity analysis is performed based on the breakdown voltage matching information set and the switching loss matching information set respectively to construct and obtain the breakdown voltage specificity set and the switch specificity set.

[0010] In a possible implementation, according to the leakage current special degree set, the breakdown voltage special degree set and the switch special degree set, the parameter utility fusion is performed on the multiple performance parameter sets to obtain multiple parameter utility information, and multiple aging analysis computing powers are configured, and the following processing is also performed: according to the multiple performance parameter sets, the leakage current special degree set, the breakdown voltage special degree set and the switch special degree set are divided to obtain multiple performance parameter special degree sets, wherein each performance parameter special degree set includes the leakage current special degree, the breakdown voltage special degree and the switch special degree; the multiple performance parameter special degree sets are divided into The leakage current specificity, breakdown voltage specificity and switch specificity are weightedly fused and calculated to obtain multiple fused specificities; multiple parameter utility information are generated according to the multiple fused specificities, wherein the size of the fused specificity is negatively correlated with the size of the parameter utility information; a preset concurrent processing quantity is obtained, and multiple aging analysis computing powers are calculated and configured based on the multiple parameter utility information, wherein the product of the preset concurrent processing quantity and the ratio of each parameter utility information to the maximum value of the multiple parameter utility information is calculated to obtain multiple concurrent processing quantities as multiple aging analysis computing powers.

[0011] In a possible implementation, aging performance analysis is performed on the multiple performance parameter sets according to the multiple aging analysis computing powers, and the following processing is also performed: based on the aging performance detection data of the high-voltage silicon carbide MOS, multiple sample performance parameter sets are collected, and based on the aging process mark of the high-voltage silicon carbide MOS, a sample aging performance evaluation result set is obtained, and an aging evaluation training data set is obtained by combining; the aging evaluation training data set is divided to obtain the sub-training data sets of the preset concurrent processing number, and multiple aging performance analysis branches are trained respectively to obtain a plurality of aging performance analysis branches, and an aging performance analyzer is obtained by integration; the multiple performance parameter sets are respectively input into the aging performance analysis branches of the randomly selected multiple concurrent processing numbers, and multiple aging performance evaluation result sets are obtained by analysis, and the averages are respectively calculated to obtain multiple aging performance evaluation results; based on the multiple aging performance evaluation results, a variation coefficient of the aging performance evaluation results is calculated to obtain an aging performance variation coefficient.

[0012] The present application also provides a dynamic evaluation system for aging performance of a high-voltage silicon carbide MOS, including: a performance parameter set acquisition module, used to perform dynamic performance tests on the high-voltage silicon carbide MOS in multiple working stages, obtain multiple performance parameter sets, and divide them into an on-resistance set, a gate leakage current set, a breakdown voltage set, and a switching loss set, wherein each performance parameter set includes an on-resistance, a gate leakage current, a breakdown voltage, and a switching loss; a matching specificity analysis module, used to perform test matching analysis of the data in the gate leakage current set, the breakdown voltage set, and the switching loss set according to the on-resistance set, and obtain a leakage current matching information set, a breakdown voltage set, and a switching loss set. A matching information set and a switching loss matching information set are prepared, and matching specificity analysis is performed to construct a leakage current specificity set, a breakdown voltage specificity set and a switch specificity set; a parameter utility information acquisition module is used to perform parameter utility fusion on the multiple performance parameter sets according to the leakage current specificity set, the breakdown voltage specificity set and the switch specificity set, obtain multiple parameter utility information, and configure multiple aging analysis computing powers; an aging performance analysis module is used to perform aging performance analysis on the multiple performance parameter sets according to the multiple aging analysis computing powers, obtain multiple aging performance evaluation results and aging performance variation coefficients as aging performance dynamic evaluation results.

[0013] Through the proposed method and system for dynamic evaluation of aging performance of high-voltage silicon carbide MOS in this application, dynamic performance test of high-voltage silicon carbide MOS is carried out to obtain multiple performance parameter sets; leakage current matching information set, breakdown voltage matching information set and switch loss matching information set are obtained, leakage current special degree set, breakdown voltage special degree set and switch special degree set are constructed; multiple parameter utility information is obtained, and multiple aging analysis computing power is configured; multiple aging performance evaluation results and aging performance change coefficients are obtained as aging performance dynamic evaluation results. The technical problem that the existing silicon carbide MOS aging performance evaluation cannot effectively integrate the dynamic changes of multiple performance parameters, resulting in difficulty in comprehensively and accurately evaluating the actual aging state of the device and predicting the aging risk of the device is solved, and the technical effect of improving the comprehensiveness and accuracy of aging performance evaluation is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0015] Figure 1A schematic flow chart of a method for dynamically evaluating aging performance of a high-voltage silicon carbide MOS provided in an embodiment of the present application;

[0016] Figure 2 A schematic diagram of the structure of a dynamic evaluation system for aging performance of a high-voltage silicon carbide MOS provided in an embodiment of the present application.

[0017] Explanation of reference numerals: performance parameter set acquisition module 10 , matching specificity analysis module 20 , parameter utility information acquisition module 30 , aging performance analysis module 40 . DETAILED DESCRIPTION

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0019] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0020] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0021] The present application embodiment provides a method for dynamically evaluating the aging performance of a high-voltage silicon carbide MOS. Figure 1 As shown, the method includes:

[0022] Step S100, performing dynamic performance tests on the high-voltage silicon carbide MOS in multiple working stages to obtain multiple performance parameter sets, which are divided into an on-resistance set, a gate leakage current set, a breakdown voltage set and a switching loss set, wherein each performance parameter set includes an on-resistance, a gate leakage current, a breakdown voltage and a switching loss.

[0023] Preferably, the high-voltage silicon carbide MOS refers to MOSFETs (Metal-Oxide-Semiconductor Field-Effect Transistors), and dynamic performance tests are performed on the high-voltage silicon carbide MOS in multiple working stages, and its performance parameters are divided into four key indicator sets. Specifically, the key performance parameters of the high-voltage silicon carbide MOS are continuously monitored and evaluated in working stages under multiple operating conditions and states. The multiple working stages may include different voltage, current, temperature and load conditions. The diversity of these working stages can more comprehensively reflect the performance of the device in various working environments. The dynamic performance test can capture the changes of the device at different time points and under different working conditions, and provide more comprehensive data support for aging performance analysis, thereby obtaining multiple performance parameter sets, each of which contains multiple key performance parameters of the device in a certain working stage (including on-resistance, gate leakage current, breakdown voltage and switching loss), which are important indicators for evaluating MOS performance and aging. Each indicator reflects the specific performance of the high-voltage silicon carbide MOS in this working stage.

[0024] Preferably, the on-resistance reflects the resistance of the MOS in the on state. The lower the on-resistance, the higher the efficiency of the device. With aging, the on-resistance may increase; the gate leakage current refers to the leakage current between the gate and other ports. An increase in the gate leakage current may be a sign of aging and may cause device failure; the breakdown voltage indicates the maximum voltage that the device can withstand. When the breakdown voltage decreases, it means that the voltage resistance of the device decreases, which is one of the typical manifestations of aging; the switching loss refers to the energy consumed by the device during the switching process. An increase in the switching loss will affect the efficiency and heat generation of the system, and is also an important indicator of device aging; then, multiple performance parameter sets are divided according to these performance indicators, that is, the on-resistance data set is classified as an on-resistance set, the gate leakage current data set is classified as a gate leakage current set, the breakdown voltage data set is classified as a breakdown voltage set, and the switching loss data set is classified as a switching loss set, and the obtained multiple performance parameter sets are divided into an on-resistance set, a gate leakage current set, a breakdown voltage set, and a switching loss set.

[0025] Step S100 further includes step S110, obtaining multiple accumulated working times of the high-voltage silicon carbide MOS as multiple working stages; step S120, performing dynamic performance tests on the high-voltage silicon carbide MOS under the multiple working stages to obtain multiple performance parameter sets, which are divided into on-resistance set, gate leakage current set, breakdown voltage set and switching loss set.

[0026] Preferably, multiple cumulative working hours of the high-voltage silicon carbide MOS are obtained. The cumulative working time refers to multiple working hours accumulated over time during the actual operation of the high-voltage silicon carbide MOS. For example, the MOS may record the cumulative working time at different time points in different stages of use (such as 10 hours, 100 hours, and 500 hours). These time periods can reflect the performance changes of the device in long-term operation. These time periods are respectively used as multiple working stages, and each stage represents the working state and performance of the MOS in a specific time period. Then, in each working stage, the performance of the high-voltage silicon carbide MOS is monitored and tested in real time. As the cumulative working time increases, the performance of the MOS will gradually change, and a A series of performance parameters are obtained to form multiple performance parameter sets. Each performance parameter set represents the performance of MOS in a specific working stage and contains multiple key performance indicators, such as on-resistance, gate leakage current, breakdown voltage and switching loss. Finally, the obtained performance parameter sets are divided into different sets according to specific performance indicators to obtain on-resistance set, gate leakage current set, breakdown voltage set and switching loss set. Among them, the on-resistance set records the on-resistance changes in different working stages; the gate leakage current set records the leakage current change trend between the gate and other ports; the breakdown voltage set records the breakdown voltage changes of MOS in different working stages; the switching loss set records the energy loss of MOS in different working stages.

[0027] Step S200, based on the on-resistance set, perform test matching analysis on the data in the gate leakage current set, breakdown voltage set and switching loss set to obtain a leakage current matching information set, a breakdown voltage matching information set and a switching loss matching information set, and perform matching specificity analysis to construct a leakage current specificity set, a breakdown voltage specificity set and a switching specificity set.

[0028] Preferably, based on the matching analysis and specificity analysis process of the on-resistance set, the relationship between the changing trend of other performance parameters (gate leakage current, breakdown voltage, switching loss) and the on-resistance is evaluated, so as to conduct a more in-depth analysis of the aging state of the device. Specifically, the test matching analysis refers to analyzing the relationship between the on-resistance set and other performance parameters (such as gate leakage current, breakdown voltage and switching loss) to determine whether these performance parameters show similar changing trends with the on-resistance in different working stages, that is, checking the correlation between these parameters and the on-resistance to see whether they change together during the aging process. By comparing the on-resistance set and the gate leakage current set, the matching degree between them is evaluated. For example, if the gate leakage current is different in multiple working stages, the gate leakage current is different from the on-resistance set. If the changing trend of the current is consistent with the changing trend of the on-resistance (such as both increase with time), it means that the two have a high matching degree, and the leakage current matching information set is used to record the correlation between the leakage current and the on-resistance; similarly, by comparing the on-resistance set and the breakdown voltage set, the matching between them is evaluated. If the breakdown voltage shows the same or similar trend with the change of the on-resistance (such as the on-resistance increases when the breakdown voltage decreases), the matching between the two is high, and the breakdown voltage matching information set records the matching between the two; then the matching analysis is performed by comparing the on-resistance set and the switching loss set. If the changing trend of the switching loss is consistent with the on-resistance, it indicates that there is a strong correlation between them, and the switching loss matching information set is used to record this correlation.

[0029] Preferably, the matching specificity analysis is a further in-depth analysis of the matching information to evaluate the matching specificity between each performance parameter set (gate leakage current, breakdown voltage, switching loss) and the on-resistance set, that is, whether there are abnormal matching situations. For example, if the gate leakage current is abnormally high or abnormally low in matching with the on-resistance in certain working stages, this may indicate that there are some special aging phenomena or abnormal conditions in this stage, thereby identifying the stage of abnormal changes in device performance and performing fault prediction. Specifically, the leakage current specificity set, the breakdown voltage specificity set and the switching specificity set are constructed through the matching specificity analysis. The leakage current specificity set reflects the special matching degree between the gate leakage current and the on-resistance in different working stages. For example, if the leakage current exhibits abnormal Increases, and the relationship with the change of on-resistance is abnormal (such as too high or too low matching), the leakage current in these stages is recorded as a special degree set; similarly, the breakdown voltage special degree set records the special matching of the breakdown voltage in certain working stages, for example, the breakdown voltage suddenly drops in certain stages, and the matching with the on-resistance is abnormal, then the breakdown voltage in these stages is classified into the special degree set; the switch special degree set records the matching special degree of the switching loss with the on-resistance in different working stages. If the matching of the switching loss with the on-resistance in certain stages is very abnormal, for example, the switching loss suddenly increases or decreases, and is inconsistent with the change of the on-resistance, it is included in the switch special degree set to identify abnormal phenomena or aging trends in different working stages, and help to more accurately identify the aging status of the device.

[0030] Step S200 further includes step S210, pre-training a test matching analyzer, wherein the test matching analyzer includes a leakage current matching analysis branch, a breakdown voltage matching analysis branch, and a switching loss matching analysis branch; step S220, combining the first on-resistance in the first working stage in the on-resistance set with the first gate leakage current in the gate leakage current set, inputting the combination into the leakage current matching analysis branch, performing a test matching analysis, and outputting the first leakage current matching information; step S230, continuing to perform a test matching analysis on the data in the gate leakage current set, the breakdown voltage set, and the switching loss set according to the on-resistance set, and obtaining a leakage current matching information set, a breakdown voltage matching information set, and a switching loss matching information set.

[0031] Preferably, the test matching analysis of the data in the gate leakage current set, breakdown voltage set and switching loss set is continued based on the on-resistance set, that is, not only the matching of the leakage current and the on-resistance is paid attention to, but also other key electrical characteristics such as the breakdown voltage and switching loss are considered at the same time. Specifically, the test matching analyzer traverses the on-resistance set, and combines each on-resistance value with the corresponding gate leakage current, breakdown voltage and switching loss value, and then performs test matching analysis on these parameter combinations in the corresponding input leakage current matching analysis branch, breakdown voltage matching analysis branch and switching loss matching analysis branch. The output results of each analysis branch are respectively a leakage current matching information set, a breakdown voltage matching information set and a switching loss matching information set. Each information set contains the matching evaluation results between the corresponding electrical characteristics and the on-resistance, which together constitute a comprehensive evaluation of the aging performance of silicon carbide MOSFET in multiple working stages.

[0032] Step S210 further includes step S211, collecting a sample on-resistance set, a sample gate leakage current set, a sample breakdown voltage set and a sample switching loss set according to historical performance test data of the high-voltage silicon carbide MOS; step S212, based on the standard aging test data of the high-voltage silicon carbide MOS, obtaining a standard gate leakage current set, a standard breakdown voltage set and a standard switching loss set corresponding to multiple sample on-resistances in the sample on-resistance set, and calculating the deviations of the sample gate leakage current set, the sample breakdown voltage set and the sample switching loss set from the standard values, calculating and marking to obtain a sample leakage current matching information set, a sample breakdown voltage matching information set and a sample switching loss matching information set; step S213, using the sample on-resistance set and the sample gate leakage current set as training inputs, using the sample leakage current matching information set as training outputs, training a leakage current matching analysis branch, and training a breakdown voltage matching analysis branch and a switching loss matching analysis branch; step S214, combining the leakage current matching analysis branch, the breakdown voltage matching analysis branch and the switching loss matching analysis branch to obtain a test matching analyzer.

[0033] Preferably, a sample on-resistance set (representing the resistance characteristics of the device under different working conditions), a sample gate leakage current set (multiple gate leakage currents), a sample breakdown voltage set (multiple breakdown voltages) and a sample switching loss set (multiple switching loss values) are collected from historical performance test data of high-voltage silicon carbide MOS, and then based on the standard aging test data of high-voltage silicon carbide MOS, a standard gate leakage current set, a standard breakdown voltage set and a standard switching loss set corresponding to the sample on-resistance set are obtained, and these parameter standard values ​​represent the electrical characteristics of the device under ideal or standard aging conditions, and then the deviations between the sample gate leakage current set, the sample breakdown voltage set and the sample switching loss set and the corresponding parameter standard values ​​are calculated to reflect the difference between the actual performance of the device and the standard performance, and based on the deviation calculation results, a sample leakage current matching information set, a sample breakdown voltage matching information set and a sample switching loss matching information set are annotated and generated, which include the matching results of the aging performance evaluation of the device under different working conditions.

[0034] Preferably, a sample on-resistance set and a sample gate leakage current set are used as training inputs, a sample leakage current matching information set is used as training output, and a leakage current matching analysis branch is trained, the leakage current matching analysis branch is intended to learn the matching relationship between the on-resistance and the gate leakage current, and to predict whether the gate leakage current under a given on-resistance meets expectations. Similarly, a sample on-resistance set and a sample breakdown voltage set are used as training inputs, a sample breakdown voltage matching information set is used as training output, and a breakdown voltage matching analysis branch is trained, the breakdown voltage matching analysis branch is intended to learn the matching relationship between the on-resistance and the breakdown voltage, and to predict whether the breakdown voltage under a given on-resistance meets expectations. The sample on-resistance set and the sample switching loss set are used as training inputs, and the sample switching loss matching information set is used as training output. The switching loss matching analysis branch is trained. The switching loss matching analysis branch aims to learn the matching relationship between on-resistance and switching loss, and predict whether the breakdown voltage under a given on-resistance meets expectations. Finally, the trained leakage current matching analysis branch, breakdown voltage matching analysis branch and switching loss matching analysis branch are combined to form a complete test matching analyzer. The test matching analyzer can receive the on-resistance and other electrical characteristic parameters of the silicon carbide MOSFET as input, and output the aging performance evaluation results of the device through the evaluation of each embedded analysis branch.

[0035] Step S200 further includes step S240, selecting the first leakage current matching information of the first working stage in the leakage current matching information set, and calculating the mean of the leakage current matching information set to obtain the average leakage current matching information; step S250, calculating the ratio of the average leakage current matching information to the first leakage current matching information as the first leakage current particularity; step S260, continuing to calculate and construct a leakage current particularity set based on the leakage current matching information set; step S270, performing matching particularity analysis based on the breakdown voltage matching information set and the switching loss matching information set, respectively, to construct a breakdown voltage particularity set and a switching particularity set.

[0036] Preferably, matching specificity analysis is performed on the matching information of leakage current, breakdown voltage and switching loss, and the specificity (or deviation, abnormality, etc.) in a specific working stage is calculated, so as to further evaluate the aging or abnormal state of high-voltage silicon carbide MOSFET in different electrical characteristics. Specifically, leakage current matching information corresponding to a specific working stage (such as an initial working stage, a working stage under specific temperature / voltage conditions, etc.) is selected from the leakage current matching information set, and the mean of the entire leakage current matching information set is calculated to obtain the average leakage current matching information. The ratio of the first leakage current matching information in the first working stage to the average leakage current matching information is taken as the first leakage current specificity, which measures the degree of deviation of the leakage current matching in the first working stage from the overall average level. If the ratio is much greater than or much less than 1, it indicates that the leakage current matching in the working stage has significant specificity or abnormality; if the ratio is close to 1, it indicates that the smaller the specificity of the leakage current matching in the working stage, the better the data quality.

[0037] Preferably, for each element in the leakage current matching information set (i.e., the leakage current matching information in each working stage), the leakage current matching information is repeatedly calculated to obtain the leakage current special degree in each working stage, and then the leakage current special degrees in all working stages are combined to form a leakage current special degree set, which provides a comprehensive set of leakage current special properties of the device in different working stages; similarly, for the matching information set of breakdown voltage and switching loss, the same matching special degree analysis is performed, specifically, the special degree of breakdown voltage and switching loss in each working stage is calculated respectively, and the average values ​​of the breakdown voltage matching information set and the switching loss matching information set are calculated, and then the corresponding ratios are calculated, which respectively reflect the specialness or abnormality of the breakdown voltage and switching loss in different working stages, and finally the breakdown voltage special degree and the switching loss special degree in all working stages are combined to form a breakdown voltage special degree set and a switching special degree set. By calculating the special degree of electrical characteristics in different working stages (i.e., the degree of deviation from the average value), the aging or abnormal state of high-voltage silicon carbide MOSFET in different electrical characteristics can be evaluated.

[0038] Step S300, performing parameter utility fusion on the multiple performance parameter sets according to the leakage current specificity set, the breakdown voltage specificity set and the switch specificity set, obtaining multiple parameter utility information, and configuring multiple aging analysis computing powers.

[0039] Preferably, parameter utility fusion is performed on multiple performance parameter sets according to the leakage current special degree set, the breakdown voltage special degree set and the switch special degree set, which means that based on the leakage current special degree, the breakdown voltage special degree and the switch loss special degree, the data validity of multiple performance parameter sets is analyzed and calculated, for example, the inverse of the special degree is fused as the corresponding parameter utility information, the larger the special degree is, the greater the difference between the matching degree of the performance parameter and the matching degree of other parameters is, the smaller the data validity is, and the smaller the corresponding parameter utility information is, and then multiple aging analysis computing powers are configured according to the obtained multiple parameter utility information, and configuring multiple aging analysis computing powers means allocating appropriate computing resources for the aging analysis process according to the evaluation requirements and the data processing volume, that is, configuring the number of branches for actual performance analysis, if the parameter utility information obtained is more and more accurate, it is necessary to configure more analysis branches, that is, configure the corresponding aging analysis computing power, which is used to perform multi-parameter comprehensive evaluation and resource optimization allocation on the aging performance of high-voltage silicon carbide MOS, to ensure the smooth progress and efficient completion of the aging analysis process.

[0040] Step S300 further includes step S310, dividing the leakage current special degree set, breakdown voltage special degree set and switch special degree set according to the multiple performance parameter sets to obtain multiple performance parameter special degree sets, wherein each performance parameter special degree set includes leakage current special degree, breakdown voltage special degree and switch special degree; step S320, performing weighted fusion calculation on the leakage current special degree, breakdown voltage special degree and switch special degree in the multiple performance parameter special degree sets to obtain multiple fused special degrees; step S330, generating multiple parameter utility information according to the multiple fused special degrees, wherein the size of the fused special degree is negatively correlated with the size of the parameter utility information; step S440, obtaining a preset concurrent processing number, and calculating and configuring multiple aging analysis computing powers based on the multiple parameter utility information, wherein based on the ratio of each parameter utility information to the maximum value of the multiple parameter utility information, the product with the preset concurrent processing number is calculated to obtain multiple concurrent processing numbers as multiple aging analysis computing powers.

[0041] Preferably, according to multiple performance parameter sets (which may represent different working conditions, test cycles or device batches, etc.), the leakage current specificity set, the breakdown voltage specificity set and the switch specificity set are divided to obtain multiple performance parameter specificity sets, so that each performance parameter specificity set contains the leakage current specificity, breakdown voltage specificity and switch specificity under corresponding conditions, and each performance parameter specificity set reflects the specificity information of the three electrical characteristics under specific conditions; then, according to the importance of each electrical characteristic in aging evaluation, different weights are assigned to the leakage current specificity, breakdown voltage specificity and switch specificity in each performance parameter specificity set, and weighted fusion calculation is performed to obtain the fused specificity of each performance parameter specificity set, and each fused specificity represents the comprehensive evaluation result of the device aging state under the corresponding conditions.

[0042] Preferably, the fusion specificity represents the degree to which the performance parameters in the performance parameter set deviate from the average value. The size of the fusion specificity is negatively correlated with the size of the parameter utility information. The larger the fusion specificity, the greater the deviation, the lower the test accuracy of the performance parameter, and thus the lower the data value, that is, the smaller the parameter utility information. Specifically, based on multiple fusion specificities, multiple parameter utility information are generated. For example, the inverse of the fusion specificity is used as the parameter utility information, and each information corresponds to a fusion specificity conversion result of a performance parameter specificity set; then, multiple aging analysis computing powers are configured based on the parameter utility information to optimize resource allocation and improve analysis efficiency. Specifically, the number of analysis tasks that can be processed simultaneously is obtained. The upper limit of the amount is calculated. For each parameter utility information, its ratio to the maximum value of all parameter utility information is calculated, which reflects the relative importance of the parameter utility information relative to the maximum utility information. Finally, this ratio is multiplied by the preset concurrent processing number to obtain the concurrent processing number corresponding to the parameter utility information, which indicates the size of the aging analysis computing power that should be allocated to the performance parameter particularity set. Finally, multiple aging analysis computing powers are configured, which will be dynamically allocated based on the importance and urgency of each performance parameter particularity set, so as to realize a comprehensive evaluation of the aging status of the high-voltage silicon carbide MOSFET under different performance parameters, and dynamically configure the aging analysis computing power according to the evaluation results to improve the analysis efficiency and accuracy.

[0043] Step S400: performing aging performance analysis on the multiple performance parameter sets according to the multiple aging analysis computing powers to obtain multiple aging performance evaluation results and aging performance variation coefficients as aging performance dynamic evaluation results.

[0044] Preferably, an aging performance analysis is performed on multiple performance parameter sets, and the analysis is performed based on multiple aging analysis computing powers to obtain evaluation results on the aging performance of high-voltage silicon carbide MOSFET and the corresponding aging performance variation coefficient, that is, by applying multiple aging analysis computing powers (that is, computing resources dynamically allocated according to parameter utility information), an in-depth aging performance analysis is performed on multiple performance parameter sets (each set contains electrical characteristic data under different conditions, such as leakage current, breakdown voltage, switching loss, etc.). Specifically, the data in the performance parameter set is preprocessed, including cleaning, denoising, standardization, etc., to extract key features related to aging performance, such as specificity value, trend change, fluctuation range, etc., and the allocated aging analysis computing power is used to run an aging performance evaluation model to predict and evaluate the aging performance of the device, wherein the aging performance evaluation model It is a prediction model built based on machine learning models (such as neural networks, random forests, etc.) according to the historical data and performance indicators of the device. It learns the relationship between performance parameters and device aging performance. For each performance parameter set, the aging performance analysis model outputs an aging performance evaluation result, including a quantitative value of the degree of aging, a prediction of the remaining service life, and a trend of performance degradation; and calculates a corresponding aging performance variation coefficient, which is used to quantify the rate or degree of change of device performance over time or working conditions. For example, the growth rate of leakage current, the rate of decrease of breakdown voltage, or the rate of increase of switching loss can be calculated as the aging performance variation coefficient to more accurately evaluate the aging trend and speed of the device; finally, multiple aging performance evaluation results and aging performance variation coefficients are combined to form a comprehensive dynamic evaluation result of the aging performance of high-voltage silicon carbide MOS.

[0045] Step S400 further includes step S410, collecting multiple sample performance parameter sets according to the aging performance detection data of the high-voltage silicon carbide MOS, and obtaining a sample aging performance evaluation result set according to the aging process mark of the high-voltage silicon carbide MOS, and combining them to obtain an aging evaluation training data set; step S420, dividing the aging evaluation training data set to obtain the sub-training data sets of the preset concurrent processing number, training them respectively to obtain multiple aging performance analysis branches, and integrating them to obtain an aging performance analyzer; step S430, respectively inputting the multiple performance parameter sets into the aging performance analysis branches of the randomly selected multiple concurrent processing numbers, analyzing to obtain multiple aging performance evaluation result sets, and respectively calculating the mean to obtain multiple aging performance evaluation results; step S440, calculating the variation coefficient of the aging performance evaluation results according to the multiple aging performance evaluation results to obtain the aging performance variation coefficient.

[0046] Preferably, based on the aging performance test data of the high-voltage silicon carbide MOSFET, performance parameter sets of multiple samples under different conditions are collected, each performance parameter set includes key performance indicators such as leakage current, breakdown voltage, switching loss, etc., and according to the aging process of the high-voltage silicon carbide MOSFET (such as through time, working conditions, failure criteria, etc.), the collected samples are evaluated for aging performance to obtain a set of sample aging performance evaluation results, and then the collected sample performance parameter sets and the corresponding aging performance evaluation results are combined to form an aging evaluation training data set; then, according to a preset number of concurrent processing (i.e., the expected parallel processing capability), the aging evaluation training data set is divided into a corresponding number of sub-training data sets, and each sub-training data set is used to train an independent aging performance analysis branch based on machine learning (such as random forest, neural network, etc.), and then the trained multiple aging performance analysis branches are integrated to form a comprehensive aging performance analyzer.

[0047] Preferably, for each performance parameter set, multiple (matching the preset concurrent processing number) aging performance analysis branches are randomly selected for analysis, that is, multiple performance parameter sets are respectively input into the randomly selected aging performance analysis branches for parallel analysis, and for each performance parameter set, the aging performance evaluation results given by all branches participating in the analysis are collected, and their mean is calculated as the final evaluation result, so as to reduce the error and bias that may be caused by a single branch; then for each performance parameter set, the change of multiple aging performance evaluation results is analyzed, that is, the change amount or difference between the aging performance evaluation results obtained by multiple analyses is calculated, and based on the calculated result of the change amount, the variation coefficient of the aging performance evaluation result is further calculated, such as the ratio of the standard deviation of the aging performance evaluation result to its mean, so as to measure the discrete degree or change rate of the aging performance evaluation result.

[0048] In the above, refer to Figure 1 A method for dynamically evaluating the aging performance of a high-voltage silicon carbide MOS according to an embodiment of the present invention is described in detail. Figure 2 A dynamic evaluation system for aging performance of a high-voltage silicon carbide MOS according to an embodiment of the present invention is described.

[0049] A dynamic evaluation system for aging performance of a high-voltage silicon carbide MOS according to an embodiment of the present invention is used to solve the technical problem that the existing silicon carbide MOS aging performance evaluation cannot effectively integrate the dynamic changes of multiple performance parameters, resulting in difficulty in comprehensively and accurately evaluating the actual aging state of the device and predicting the aging risk of the device, thereby achieving the technical effect of improving the comprehensiveness and accuracy of aging performance evaluation. A dynamic evaluation system for aging performance of a high-voltage silicon carbide MOS includes: a performance parameter set acquisition module 10, a matching specificity analysis module 20, a parameter utility information acquisition module 30, and an aging performance analysis module 40.

[0050] The performance parameter set acquisition module 10 is used to perform dynamic performance testing on the high-voltage silicon carbide MOS in multiple working stages to obtain multiple performance parameter sets, which are divided into an on-resistance set, a gate leakage current set, a breakdown voltage set and a switching loss set, wherein each performance parameter set includes an on-resistance, a gate leakage current, a breakdown voltage and a switching loss; the matching specificity analysis module 20 is used to perform test matching analysis on the data in the gate leakage current set, the breakdown voltage set and the switching loss set according to the on-resistance set, and obtain a leakage current matching information set, a breakdown voltage matching information set and a switching loss matching information set. set, and perform matching special degree analysis to construct a leakage current special degree set, a breakdown voltage special degree set and a switch special degree set; a parameter utility information acquisition module 30, used to perform parameter utility fusion on the multiple performance parameter sets according to the leakage current special degree set, the breakdown voltage special degree set and the switch special degree set, obtain multiple parameter utility information, and configure multiple aging analysis computing powers; an aging performance analysis module 40, used to perform aging performance analysis on the multiple performance parameter sets according to the multiple aging analysis computing powers, obtain multiple aging performance evaluation results and aging performance variation coefficients as aging performance dynamic evaluation results.

[0051] The specific configuration of the performance parameter set acquisition module 10 will be described in detail below. The performance parameter set acquisition module 10 further includes: obtaining multiple accumulated working times of the high-voltage silicon carbide MOS as multiple working stages; performing dynamic performance tests on the high-voltage silicon carbide MOS in the multiple working stages to obtain multiple performance parameter sets, and dividing and obtaining an on-resistance set, a gate leakage current set, a breakdown voltage set, and a switching loss set.

[0052] The specific configuration of the matching specificity analysis module 20 will be described in detail below. The matching specificity analysis module 20 may further include: a pre-trained test matching analyzer, wherein the test matching analyzer includes a leakage current matching analysis branch, a breakdown voltage matching analysis branch, and a switch loss matching analysis branch; combining the first on-resistance in the first working stage in the on-resistance set with the first gate leakage current in the gate leakage current set, inputting the first gate leakage current matching analysis branch, performing a test matching analysis, and outputting the first leakage current matching information; continuing to perform a test matching analysis of the data in the gate leakage current set, the breakdown voltage set, and the switch loss set according to the on-resistance set, and obtaining a leakage current matching information set, a breakdown voltage matching information set, and a switch loss matching information set.

[0053] The specific configuration of the matching specificity analysis module 20 will be described in detail below. The matching specificity analysis module 20 may further include: collecting a sample on-resistance set, a sample gate leakage current set, a sample breakdown voltage set, and a sample switch loss set according to the historical performance test data of the high-voltage silicon carbide MOS; obtaining a standard gate leakage current set, a standard breakdown voltage set, and a standard switch loss set corresponding to multiple sample on-resistances in the sample on-resistance set based on the standard aging test data of the high-voltage silicon carbide MOS, and calculating the deviations of the sample gate leakage current set, the sample breakdown voltage set, and the sample switch loss set from the standard values, calculating and marking to obtain a sample leakage current matching information set, a sample breakdown voltage matching information set, and a sample switch loss matching information set; using the sample on-resistance set and the sample gate leakage current set as training inputs, using the sample leakage current matching information set as training outputs, training a leakage current matching analysis branch, and training a breakdown voltage matching analysis branch and a switch loss matching analysis branch; combining the leakage current matching analysis branch, the breakdown voltage matching analysis branch, and the switch loss matching analysis branch to obtain a test matching analyzer.

[0054] The specific configuration of the matching specificity analysis module 20 will be described in detail below. The matching specificity analysis module 20 may further include: selecting the first leakage current matching information of the first working stage in the leakage current matching information set, and calculating the mean of the leakage current matching information set to obtain the average leakage current matching information; calculating the ratio of the average leakage current matching information to the first leakage current matching information as the first leakage current specificity; continuing to calculate and construct the leakage current specificity set according to the leakage current matching information set; performing matching specificity analysis according to the breakdown voltage matching information set and the switching loss matching information set, respectively, to construct and obtain the breakdown voltage specificity set and the switching specificity set.

[0055] The specific configuration of the parameter utility information acquisition module 30 will be described in detail below. The parameter utility information acquisition module 30 may further include: dividing the leakage current special degree set, the breakdown voltage special degree set and the switch special degree set according to the multiple performance parameter sets to obtain multiple performance parameter special degree sets, wherein each performance parameter special degree set includes leakage current special degree, breakdown voltage special degree and switch special degree; performing weighted fusion calculation on the leakage current special degree, breakdown voltage special degree and switch special degree in the multiple performance parameter special degree sets to obtain multiple fusion special degrees; generating multiple parameter utility information according to the multiple fusion special degrees, wherein the size of the fusion special degree is negatively correlated with the size of the parameter utility information; obtaining a preset concurrent processing number, and calculating and configuring multiple aging analysis computing powers based on the multiple parameter utility information, wherein the product of the preset concurrent processing number is calculated based on the ratio of each parameter utility information to the maximum value of the multiple parameter utility information, and obtaining multiple concurrent processing numbers as multiple aging analysis computing powers.

[0056] The specific configuration of the aging performance analysis module 40 will be described in detail below. The aging performance analysis module 40 may further include: collecting multiple sample performance parameter sets according to the aging performance detection data of the high-voltage silicon carbide MOS, and obtaining a sample aging performance evaluation result set according to the aging process mark of the high-voltage silicon carbide MOS, and combining to obtain an aging evaluation training data set; dividing the aging evaluation training data set to obtain the sub-training data sets of the preset concurrent processing number, respectively training to obtain multiple aging performance analysis branches, and integrating to obtain an aging performance analyzer; respectively inputting the multiple performance parameter sets into the aging performance analysis branches of the randomly selected multiple concurrent processing numbers, analyzing to obtain multiple aging performance evaluation result sets, and respectively calculating the mean to obtain multiple aging performance evaluation results; according to the multiple aging performance evaluation results, calculating the variation coefficient of the aging performance evaluation results to obtain the aging performance variation coefficient.

[0057] A system for dynamically evaluating aging performance of a high-voltage silicon carbide MOS provided in an embodiment of the present invention can execute a method for dynamically evaluating aging performance of a high-voltage silicon carbide MOS provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0058] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0059] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for dynamically evaluating the aging performance of a high-voltage silicon carbide MOS, characterized in that: The method comprises: Performing dynamic performance tests on high-voltage silicon carbide MOS in multiple working stages to obtain multiple performance parameter sets, which are divided into an on-resistance set, a gate leakage current set, a breakdown voltage set, and a switching loss set, wherein each performance parameter set includes an on-resistance, a gate leakage current, a breakdown voltage, and a switching loss; According to the on-resistance set, test matching analysis of the data in the gate leakage current set, breakdown voltage set and switching loss set is performed to obtain a leakage current matching information set, a breakdown voltage matching information set and a switching loss matching information set, and matching specificity analysis is performed to construct a leakage current specificity set, a breakdown voltage specificity set and a switching specificity set; According to the leakage current special degree set, the breakdown voltage special degree set and the switch special degree set, parameter utility fusion is performed on the multiple performance parameter sets to obtain multiple parameter utility information, and multiple aging analysis computing powers are configured; Performing aging performance analysis on the multiple performance parameter sets according to the multiple aging analysis computing powers to obtain multiple aging performance evaluation results and aging performance variation coefficients as aging performance dynamic evaluation results; According to the on-resistance set, the test matching analysis of the data in the gate leakage current set, the breakdown voltage set and the switching loss set is performed, including: A pre-trained test match analyzer, wherein the test match analyzer includes a leakage current match analysis branch, a breakdown voltage match analysis branch, and a switching loss match analysis branch; Combine the first on-resistance in the first working stage in the on-resistance set and the first gate leakage current in the gate leakage current set, input the combination into the leakage current matching analysis branch, perform a test matching analysis, and output first leakage current matching information; Continue to perform test matching analysis on the data in the gate leakage current set, the breakdown voltage set and the switching loss set according to the on-resistance set to obtain a leakage current matching information set, a breakdown voltage matching information set and a switching loss matching information set; Among them, the pre-trained test matching analyzer includes: Based on the historical performance test data of high-voltage silicon carbide MOS, a sample on-resistance set, a sample gate leakage current set, a sample breakdown voltage set, and a sample switching loss set are collected; Based on the standard aging test data of high-voltage silicon carbide MOS, a standard gate leakage current set, a standard breakdown voltage set and a standard switching loss set corresponding to a plurality of sample on-resistances in the sample on-resistance set are obtained, and deviations of the sample gate leakage current set, the sample breakdown voltage set and the sample switching loss set from the standard values ​​are calculated, and a sample leakage current matching information set, a sample breakdown voltage matching information set and a sample switching loss matching information set are obtained by calculation and annotation; Using the sample on-resistance set and the sample gate leakage current set as training inputs, using the sample leakage current matching information set as training outputs, training a leakage current matching analysis branch, and training a breakdown voltage matching analysis branch and a switching loss matching analysis branch; Combining the leakage current matching analysis branch, the breakdown voltage matching analysis branch and the switching loss matching analysis branch to obtain a test matching analyzer; Among them, matching special degree analysis is performed to construct leakage current special degree set, breakdown voltage special degree set and switch special degree set, including: Selecting first leakage current matching information of a first working stage from the leakage current matching information set, and calculating a mean value of the leakage current matching information set to obtain average leakage current matching information; Calculating a ratio of the average leakage current matching information to the first leakage current matching information as a first leakage current specificity; Continuing to calculate and construct a leakage current specificity set according to the leakage current matching information set; According to the breakdown voltage matching information set and the switching loss matching information set, matching specificity analysis is performed to construct a breakdown voltage specificity set and a switching specificity set; According to the leakage current special degree set, the breakdown voltage special degree set and the switch special degree set, the parameter utility fusion is performed on the multiple performance parameter sets to obtain multiple parameter utility information, and multiple aging analysis computing powers are configured, including: According to the multiple performance parameter sets, the leakage current specificity set, the breakdown voltage specificity set and the switch specificity set are divided to obtain multiple performance parameter specificity sets, wherein each performance parameter specificity set includes a leakage current specificity, a breakdown voltage specificity and a switch specificity; Performing weighted fusion calculation on the leakage current specificity, the breakdown voltage specificity and the switch specificity in the plurality of performance parameter specificity sets to obtain a plurality of fused specificities; Generate multiple parameter utility information according to the multiple fusion specificities, wherein the magnitude of the fusion specificity is negatively correlated with the magnitude of the parameter utility information; Obtaining a preset concurrent processing quantity, and calculating and configuring a plurality of aging analysis computing powers based on the plurality of parameter utility information, wherein a product of a ratio of each parameter utility information to a maximum value among the plurality of parameter utility information and the preset concurrent processing quantity is calculated to obtain a plurality of concurrent processing quantities as a plurality of aging analysis computing powers; The step of performing aging performance analysis on the multiple performance parameter sets according to the multiple aging analysis computing powers includes: According to the aging performance test data of the high-voltage silicon carbide MOS, multiple sample performance parameter sets are collected, and according to the aging process mark of the high-voltage silicon carbide MOS, a sample aging performance evaluation result set is obtained, and an aging evaluation training data set is obtained by combining; Dividing the aging evaluation training data set to obtain the preset concurrent processing number of sub-training data sets, respectively training to obtain multiple aging performance analysis branches, and integrating to obtain an aging performance analyzer; Input the multiple performance parameter sets into a randomly selected multiple concurrent processing number of aging performance analysis branches respectively, analyze to obtain multiple aging performance evaluation result sets, and calculate the mean values ​​respectively to obtain multiple aging performance evaluation results; According to the plurality of aging performance evaluation results, a variation coefficient of the aging performance evaluation result is calculated to obtain an aging performance variation coefficient.

2. The method for dynamically evaluating aging performance of a high-voltage silicon carbide MOS according to claim 1, characterized in that: Dynamic performance tests are performed on high-voltage silicon carbide MOS in multiple working stages to obtain multiple performance parameter sets, which are divided into on-resistance set, gate leakage current set, breakdown voltage set and switching loss set, including: Acquire multiple accumulated working times of the high-voltage silicon carbide MOS as multiple working stages; In the multiple working stages, dynamic performance tests are performed on the high-voltage silicon carbide MOS to obtain multiple performance parameter sets, which are divided into an on-resistance set, a gate leakage current set, a breakdown voltage set and a switching loss set.

3. A dynamic evaluation system for aging performance of high-voltage silicon carbide MOS, characterized in that: The system is used to implement a method for dynamically evaluating aging performance of a high-voltage silicon carbide MOS according to any one of claims 1 to 2, and the system comprises: A performance parameter set acquisition module is used to perform dynamic performance tests on high-voltage silicon carbide MOS in multiple working stages to obtain multiple performance parameter sets, which are divided into an on-resistance set, a gate leakage current set, a breakdown voltage set and a switching loss set, wherein each performance parameter set includes an on-resistance, a gate leakage current, a breakdown voltage and a switching loss; A matching specificity analysis module is used to perform test matching analysis of the data in the gate leakage current set, the breakdown voltage set and the switching loss set according to the on-resistance set, obtain a leakage current matching information set, a breakdown voltage matching information set and a switching loss matching information set, and perform matching specificity analysis to construct a leakage current specificity set, a breakdown voltage specificity set and a switching specificity set; A parameter utility information acquisition module, configured to perform parameter utility fusion on the plurality of performance parameter sets according to the leakage current special degree set, the breakdown voltage special degree set and the switch special degree set, obtain a plurality of parameter utility information, and configure a plurality of aging analysis computing powers; The aging performance analysis module is used to perform aging performance analysis on the multiple performance parameter sets according to the multiple aging analysis computing powers, and obtain multiple aging performance evaluation results and aging performance variation coefficients as aging performance dynamic evaluation results.

Images