Scenario Adaptive Aging Test Method and System for UHV MOS

Abstract

The present invention discloses a scenario-adaptive aging test method and system for ultra-high voltage MOS, which relates to the technical field of MOS testing. The method includes: establishing a device aging test platform; parsing test conditions for the application scenario information of the target ultra-high voltage MOS device; performing pre-test adjustment on the target ultra-high voltage MOS device to determine an aging test update parameter table; monitoring the aging test of the target ultra-high voltage MOS device; invoking an ultra-high voltage MOS aging self-adaptive evaluator to process and evaluate the device aging scenario test data stream set to obtain the device scenario aging performance test result. It solves the technical problems in the prior art that the MOS aging test lacks scenario adaptability and it is difficult to accurately simulate the aging process of MOS devices in the actual application environment, resulting in insufficient test pertinence and poor result accuracy, and achieves the technical effect of improving the pertinence and result accuracy of the MOS device aging test.

Description

Technical Field

[0001] This application relates to the technical field of MOS testing, and specifically relates to a scenario-adaptive aging test method and system for ultra-high voltage MOS. Background Art

[0002] With the continuous increase in the demand for high performance and high reliability in modern electronic devices, metal-oxide-semiconductor field-effect transistors (MOSFETs) are increasingly widely used in the fields of power electronics, automotive electronics, communication devices, etc. Especially in ultra-high voltage application scenarios, the reliability and stability of MOS devices are particularly important. As an important semiconductor device, the aging problem of MOS will gradually emerge during long-term operation. The aging of MOS is mainly manifested as the degradation of device parameters, including changes in threshold voltage (Vth), increase in leakage current, reduction in switching speed, etc. The aging phenomenon is usually caused by the combined action of factors such as high electric fields, high-temperature environments, and long-term high-voltage operation of the device. These factors will cause thinning of the oxide layer, migration phenomenon, interface defects, etc. inside the device. Especially in ultra-high voltage application environments, MOS faces more severe challenges. When operating at high voltages, the electric field strength of the device increases, resulting in an accelerated aging speed and a greatly shortened device life. However, the current MOS device aging test methods usually rely on standard accelerated life tests. By operating the MOSFET under specific accelerated conditions such as temperature, voltage, frequency, etc., to simulate the aging process of the device in the actual working environment, although it can help engineers detect the life and reliability of MOS devices, it often lacks targeted evaluation of specific working scenarios, and the test process is relatively simplified, unable to accurately and comprehensively reflect the aging phenomenon in the actual use environment.

[0003] Therefore, in the current related technologies, there are technical problems that the MOS aging test lacks scenario adaptability and it is difficult to accurately simulate the aging process of MOS devices in the actual application environment, resulting in insufficient test pertinence and poor result accuracy. Summary of the Invention

[0004] This application provides a scenario-adaptive aging test method and system for ultra-high voltage MOS, solving the technical problems in the prior art that the MOS aging test lacks scenario adaptability and it is difficult to accurately simulate the aging process of MOS devices in the actual application environment, resulting in insufficient test pertinence and poor result accuracy, and achieving the technical effect of improving the pertinence and result accuracy of the MOS device aging test.

[0005] The present application provides a scenario - adaptive aging test method for extra - high - voltage MOS. The method includes: establishing a device aging test platform, where the device aging test platform includes an environmental test chamber, a host computer, a test fixture, and a test evaluation module; parsing test conditions for the application scenario information of the target extra - high - voltage MOS device through the host computer to generate an aging test condition parameter table; fixedly installing the target extra - high - voltage MOS device on the test fixture, and performing pre - test adjustment on the target extra - high - voltage MOS device by the environmental test chamber based on the aging test condition parameter table to determine an aging test update parameter table; monitoring the aging test of the target extra - high - voltage MOS device based on the aging test update parameter table to obtain a device aging scenario test data stream set; calling an extra - high - voltage MOS aging self - adaptive evaluator through the test evaluation module, and processing and evaluating the device aging scenario test data stream set based on the extra - high - voltage MOS aging self - adaptive evaluator to obtain a device scenario aging performance test result.

[0006] In a possible implementation manner, when generating the aging test condition parameter table, the following processing is further performed: extracting elements from the application scenario information of the target extra - high - voltage MOS device to obtain device application scenario element information; sequentially parsing the test conditions for each application scenario in the application scenario information according to the device application scenario element information to obtain a device scenario test element parameter set; determining a device scenario aging test parameter set according to the device aging test target; combining the parameter designs of the device scenario test element parameter set and the device scenario aging test parameter set to generate the aging test condition parameter table.

[0007] In a possible implementation manner, when determining the aging test update parameter table, the following processing is further performed: monitoring the pre - test of the target extra - high - voltage MOS device based on the aging test condition parameter table to obtain a device aging pre - test data stream set; performing performance evaluation on the target extra - high - voltage MOS device based on the device aging pre - test data stream set to obtain a device aging performance pre - test result; performing an abnormal threshold analysis on the device aging performance pre - test result to determine a device performance abnormal test parameter threshold; adjusting and optimizing the aging test condition parameter table based on the device performance abnormal test parameter threshold to determine the aging test update parameter table.

[0008] In a possible implementation, the call to the UHV MOS aging adaptive evaluator further performs the following processing: Collect the UHV MOS aging test data set, analyze the key indicators of the UHV MOS aging test data set, and determine the key evaluation index set for aging performance; Shunt and evaluate the identification of the UHV MOS aging test data set according to the key evaluation index set for aging performance to obtain the key index aging test sample set; Select the network structure and evaluate the training for the key index aging test sample set respectively to generate the index aging performance evaluation branch network set; Integrate and fuse the index aging performance evaluation branch network set to obtain the UHV MOS aging adaptive evaluator and store it in the test evaluation module.

[0009] In a possible implementation, the determination of the key evaluation index set for aging performance further performs the following processing: Extract the associated indicators of the UHV MOS aging test data set to obtain the device aging associated evaluation index set; Perform a clustering operation on the device aging associated evaluation index set based on the UHV MOS aging test data set to obtain the associated aging index clustering result; Calculate and evaluate the within-cluster divergence and between-cluster divergence of the associated aging index clustering result to determine the clustering divergence quality index; Optimize and update the associated aging index clustering result and screen the key indicators based on the clustering divergence quality index to determine the key evaluation index set for aging performance.

[0010] In a possible implementation, the determination of the key evaluation index set for aging performance further performs the following processing: Iteratively optimize and update the associated aging index clustering result based on the clustering divergence quality index until a preset termination condition is reached to obtain the aging index clustering update result; Evaluate the relevance of each index cluster in the aging index clustering update result to obtain the aging index cluster relevance set; Determine the aging index cluster screening threshold set according to the aging index cluster relevance set; Screen the key indicators for each index cluster in the aging index clustering update result based on the aging index cluster screening threshold set to determine the key evaluation index set for aging performance.

[0011] In a possible implementation, the obtaining of the UHV MOS aging adaptive evaluator further performs the following processing: Evaluate the criticality of each branch network in the index aging performance evaluation branch network set to obtain the branch network criticality coefficient set; Verify the output performance of each branch network in the index aging performance evaluation branch network set to obtain the branch network accuracy factor set; Construct the branch network decision function: , where represents the th branch network criticality coefficient, identifies the criticality coefficient of the detection factor, identifies the A branch network precision factor; based on the branch network decision function, parameter calculations are performed on the branch network critical coefficient set and the branch network precision factor set to obtain a branch network decision parameter set; based on the branch network decision parameter set, weighted integration fusion is performed on the index aging performance evaluation branch network set to obtain the UHV MOS aging adaptive evaluator.

[0012] This application also provides a scenario adaptive aging test system for UHV MOS, including: a device aging test platform establishment module for establishing a device aging test platform, where the device aging test platform includes an environmental test chamber, a host computer, a test fixture, and a test evaluation module; a test condition analysis module for analyzing the application scenario information of the target UHV MOS device through the host computer to generate an aging test condition parameter table; a pre-test adjustment module for fixedly installing the target UHV MOS device on the test fixture and performing pre-test adjustment on the target UHV MOS device through the environmental test chamber based on the aging test condition parameter table to determine an aging test update parameter table; an aging test monitoring module for monitoring the aging test of the target UHV MOS device based on the aging test update parameter table to obtain a device aging scenario test data stream set; a performance test result acquisition module for calling the UHV MOS aging adaptive evaluator through the test evaluation module and processing and evaluating the device aging scenario test data stream set based on the UHV MOS aging adaptive evaluator to obtain a device scenario aging performance test result.

[0013] It is intended to establish a device aging test platform through the scenario adaptive aging test method and system for UHV MOS proposed in this application; analyze the test conditions of the application scenario information of the target UHV MOS device; perform pre-test adjustment on the target UHV MOS device to determine an aging test update parameter table; monitor the aging test of the target UHV MOS device; call the UHV MOS aging adaptive evaluator to process and evaluate the device aging scenario test data stream set to obtain a device scenario aging performance test result. This solves the technical problems in the prior art that the MOS aging test lacks scenario adaptability and it is difficult to accurately simulate the aging process of MOS devices in the actual application environment, resulting in insufficient test pertinence and poor result accuracy, and achieves the technical effect of improving the pertinence and result accuracy of the MOS device aging test. Description of the Drawings

[0014] To more clearly illustrate the technical solutions of the embodiments of the present disclosure, the accompanying drawings of the embodiments of the present disclosure will be briefly introduced below. Flowcharts are used in this application to illustrate the operations performed by the systems according to the embodiments of the present application. It should be understood that the operations described above or below do not necessarily need to be executed precisely in order. Instead, various steps can be executed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more steps can be removed from these processes.

[0015] Figure 1 Schematic flow diagram of the scenario adaptive aging test method for UHV MOS provided by the embodiment of the present application;

[0016] Figure 2 Schematic structural diagram of the scenario adaptive aging test system for UHV MOS provided by the embodiment of the present application.

[0017] Explanation of reference numerals: Device aging test platform establishment module 10, test condition analysis module 20, pre-test adjustment module 30, aging test monitoring module 40, performance test result acquisition module 50. Detailed implementation manners

[0018] The above description is only an overview of the technical solutions of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the following specifically gives the detailed implementation manners of the present application.

[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 with reference to the accompanying drawings. The described embodiments should not be regarded as limitations of the present application. All other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present application.

[0020] In the following description, "some embodiments" are involved, which describe a subset of all possible embodiments. However, it can be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict. The terms "first / second" involved are only used to distinguish similar objects and do not represent a specific order for the objects. The terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or server comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules 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 commonly understood by those skilled in the technical field to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application.

[0021] The embodiments of this application provide a scenario-adaptive aging test method for ultra-high voltage MOS, as Figure 1 shown. The method includes: Step S100, establishing a device aging test platform, where the device aging test platform includes an environmental test chamber, a host computer, a test fixture, and a test evaluation module.

[0022] Preferably, when establishing a device aging test platform, the device aging test platform refers to an integrated test platform containing multiple components and devices constructed for systematic and precise aging tests of target ultra-high voltage MOS devices. The device aging test platform includes an environmental test chamber, a host computer, a test fixture, and a test evaluation module. Specifically, the environmental test chamber is used to simulate specific environmental conditions (such as high temperature, high humidity, high voltage, etc.) for aging tests of MOS devices. The environmental test chamber can control factors such as the environmental temperature, humidity, and voltage of the device during the test, so as to simulate the aging process of the device in different working environments. For example, for ultra-high voltage MOS devices, the environmental test chamber can simulate their working conditions in high-voltage and high-temperature environments to detect their aging characteristics; the host computer refers to a computer system that connects and controls the entire test system, and is responsible for managing, controlling, and data acquisition during the test. Through the host computer, an operator can input test conditions, obtain test data, adjust test parameters, etc. The host computer is also responsible for parsing the application scenario information of the target MOS device and generating a corresponding aging test condition parameter table, so as to achieve precise control and automation of the test process.

[0023] Preferably, the test fixture is used to fix the target MOS device on the test platform to ensure the stability and accuracy of the device during the test. The fixture needs to be designed according to the specific size and shape of the MOS device to ensure that the fixture can correctly contact the device and ensure that the device will not be displaced or otherwise interfered during the test, guaranteeing the consistency and reliability of the device under different test environments. The test evaluation module refers to the system module used to process and analyze test data. It can evaluate the aging process and performance changes of the device based on the data stream generated by the target MOS device under different test conditions. Usually, the test evaluation module is connected to the host computer and is responsible for calling the relevant aging adaptive evaluator to analyze the test data of the device, thereby obtaining the test results of the device's scenario aging performance. The device aging test platform aims to simulate and evaluate the aging performance of MOS devices under different environmental conditions. It controls the test conditions through the environmental test chamber, manages and processes data with the help of the host computer, uses the test fixture to ensure the stable fixation of the device, and at the same time deeply analyzes the test data through the test evaluation module to comprehensively evaluate the aging performance of the device.

[0024] Step S200, the host computer analyzes the test conditions of the application scenario information of the target UHV MOS device to generate an aging test condition parameter table.

[0025] Preferably, in the device aging test platform, the host computer analyzes and derives the specific conditions suitable for the aging test of the device based on the application scenario information of the target UHV MOS device and generates an aging test condition parameter table. Specifically, through the actual usage scenarios of the device (such as power electronic systems, communication equipment, automotive electronic control systems, etc.), the application scenario information of the target UHV MOS device is obtained, that is, various parameters related to the working environment of the target MOS device, such as working voltage, current, working frequency, temperature range, load conditions, running time, environmental humidity, etc. Then, the host computer analyzes the application scenario information of the target UHV MOS device and converts this information into conditions related to the aging test. Specifically, according to the preset rules and combining the electrical characteristics, structural characteristics and working scenarios of the target device, the host computer converts this information into aging test conditions that can reflect the performance degradation of the device. For example, if the device will work in a high-temperature environment, the test conditions may be set to a certain temperature higher than the actual working environment to accelerate the aging process, and then an aging test condition parameter table is generated, specifically listing various settings during the aging test, which may include but are not limited to temperature range, voltage range, current load, test time and test steps (such as preheating, load adjustment, etc.). By generating a detailed test condition parameter table, the host computer ensures that during the entire test process, the test conditions used can accurately simulate the actual working scenario of the target MOS device and can cover the performance of the MOS device under various working conditions to evaluate its long-term reliability.

[0026] Further, step S200 further includes step S210 of extracting elements from the application scenario information of the target UHV MOS device to obtain device application scenario element information; step S220 of sequentially analyzing test conditions for each application scenario in the application scenario information according to the device application scenario element information to obtain a device scenario test element parameter set; step S230 of determining a device scenario aging test parameter set according to the device aging test target; and step S240 of performing parameter design combination on the device scenario test element parameter set and the device scenario aging test parameter set to generate the aging test condition parameter table.

[0027] Preferably, extracting elements from the application scenario information of the target UHV MOS device means extracting key feature data crucial for device aging test from these application scenario information to obtain specific device application scenario element information, which may include voltage level, temperature range, current load, load change, and environmental factors, etc. Then, sequentially analyze the test conditions for each application scenario in the application scenario information according to the device application scenario element information, that is, by analyzing each application scenario (such as high temperature, high voltage, current load, etc.) one by one, and converting each element information extracted from the application scenario into specific test conditions to obtain a device scenario test element parameter set, including temperature, humidity, voltage level, and test time, etc.; then, according to the device aging test target (the target desired in actual test), for example, by simulating extreme environmental conditions to accelerate the aging process of the MOS device and testing the stability of the device under continuous load, such as the change of drain current and threshold voltage of the device, etc., to determine the device scenario aging test parameter set, including the voltage, current load, and temperature of the device, etc.; finally, combine the scenario test element parameter set with the parameter set of the aging test target, that is, combine the conditions such as voltage, current, and temperature in the test element parameter set with the extreme conditions such as high voltage and high temperature required by the aging test target to form the final aging test condition parameter table, which contains all test conditions and parameters, ensuring that the test conditions can comprehensively cover all working conditions that the MOS device may encounter in actual application and improving the accuracy and comprehensiveness of the aging performance test.

[0028] Step S300 of fixedly installing the target UHV MOS device on the test fixture and performing pre-test adjustment on the target UHV MOS device by the environmental test chamber based on the aging test condition parameter table to determine the aging test update parameter table.

[0029] Preferably, during the device aging test, in combination with the operations of device fixing, environment adjustment, and test condition optimization, it is ensured that the test can more accurately simulate the performance of the MOS device in the actual working environment, and the test conditions are updated to improve the effectiveness of the test. Specifically, the target extra-high voltage MOS device is fixedly installed on the test fixture, that is, the target MOS device is fixed using the test fixture to ensure that the device does not displace or damage during the aging test. The design of the fixture needs to match the size, structure, and contact method of the MOS device to ensure that the device can make good contact with the test conditions (such as temperature, voltage) in the environmental test chamber, avoiding test errors caused by looseness or poor contact of the device during the test, and ensuring the stability and accuracy during the test process; then use the environmental test chamber to simulate specific environmental conditions (such as high temperature, high humidity, high voltage, etc.), that is, the environmental test chamber performs pre-test adjustment on the target MOS device according to the aging test condition parameter table. Specifically, by adjusting parameters such as temperature, voltage, and humidity, the working environment that the target MOS device may face in the actual application scenario is simulated. Through pre-test adjustment, it can be ensured in advance that the test conditions reach the best state and prepare for the formal aging test. After the pre-test adjustment, if it is found that some of the originally set test conditions do not adapt to the actual device response or test requirements, the test conditions need to be optimized, such as adjusting environmental factors such as temperature, voltage, and load, adjusting the test time, and optimizing the test steps. These adjusted conditions are recorded and updated as the new aging test update parameter table to ensure that the subsequent aging test is carried out more accurately and efficiently, and real and reliable aging performance data can be obtained, thereby improving the reliability and accuracy of the test data.

[0030] Further, step S300 further includes step S310, pre-test monitoring the target extra-high voltage MOS device based on the aging test condition parameter table to obtain a device aging pre-test data stream set; step S320, performing performance evaluation on the target extra-high voltage MOS device based on the device aging pre-test data stream set to obtain a device aging performance pre-test result; step S330, performing abnormal threshold analysis on the device aging performance pre-test result to determine a device performance abnormal test parameter threshold; step S340, adjusting and optimizing the aging test condition parameter table based on the device performance abnormal test parameter threshold to determine the aging test update parameter table.

[0031] Preferably, based on the aging test condition parameter table, pre - test monitoring is carried out on the target UHV MOS device to verify whether the set test conditions meet the actual requirements and ensure the stability and accuracy of data during the test. In the pre - test stage, the MOS device will operate for a period of time under the specified environmental conditions, and the performance data (such as current, voltage, leakage current, temperature change, etc.) during this period will be monitored and recorded in real - time to generate an aging pre - test data stream set, which is usually recorded in a time - series manner, reflecting the performance changes of the MOS device during the preliminary test, including the changes of multiple dimensions of information (such as voltage, current, temperature, etc.) over time, and embodying the aging trend of the device under given conditions. Then, various evaluation methods are used to process and analyze the data, including analyzing the changing trends of various performance indicators of the device (such as leakage current, threshold voltage, current - carrying capacity, etc.) over time, judging whether the device has experienced unexpected degradation or abnormal phenomena during the aging process, and the evaluation results are used as the pre - test results of the device aging performance, such as the predicted aging trend (such as whether it conforms to the normal aging curve), whether there is premature performance decline or abnormality.

[0032] Preferably, perform an abnormal threshold analysis on the pre - test results of the device aging performance, that is, analyze the changes in the device performance, especially those performance degradations that exceed the normal range. Among them, the abnormal threshold refers to that during the aging process, some performances of the device (such as leakage current, threshold voltage, switching speed, etc.) exceed the preset safety range, indicating that the device may have experienced abnormal aging or failure, so as to define the abnormal test parameter threshold of the device performance as the standard for monitoring and evaluation in the subsequent formal aging test, helping to judge whether there are sudden changes in the device performance during the test process. Finally, according to the abnormal test parameter threshold of the performance, optimize and adjust the original aging test condition parameter table, that is, when it is found that the device performance has significantly declined, the test parameters need to be optimized to avoid device damage, identify any abnormal or unexpected results, optimize the test element parameters in the test plan, and obtain the updated parameter table for the aging test to ensure that the subsequent formal aging test can more accurately simulate the long - term use state of the MOS device.

[0033] Step S400, based on the updated parameter table for the aging test, carry out aging test monitoring on the target UHV MOS device to obtain a data stream set for the device aging scenario test.

[0034] Preferably, update the aging test conditions in the parameter table according to the aging test, conduct actual test monitoring on the target UHV MOS device, collect real-time data of the device under different working environments, and form a data stream set for the device aging scenario test. Specifically, apply these specific environmental conditions to the target MOS device according to the updated parameter table of the aging test, and conduct an actual aging test. For example, the test may include working at high temperature and high pressure for a long time, or simulating the performance of the device under extreme current and voltage changes. Then, during the entire process of the aging test, obtain various operating data of the target MOS device in real time through a monitoring system (such as various sensors, data acquisition systems, and instrument devices), including current, voltage, temperature, power consumption, leakage current, etc., ensure that the test environment and parameters meet the requirements of the aging test, and at the same time capture the performance changes of the device during operation in a timely manner to help testers discover abnormal situations in a timely manner, so as to obtain a data stream set for the device aging scenario test, including various performance indicators (such as voltage, temperature, leakage current, etc.) of the device at each time point. For example, multi-dimensional information such as voltage, current, temperature, power, temperature inside the device, and leakage current. The data in each dimension changes with time, recording the change trend of the device performance during aging. The data stream set provides the performance data of the device under different aging conditions and is an important basis for evaluating the aging process of the MOS device. By analyzing these data, it can be judged whether the device meets the design requirements, whether its performance degrades during aging, and the service life and reliability of the MOS device can be inferred.

[0035] Step S500, call the UHV MOS aging self-adaptive evaluator through the test evaluation module, and process and evaluate the data stream set of the device aging scenario test based on the UHV MOS aging self-adaptive evaluator to obtain the test result of the device scenario aging performance.

[0036] Preferably, call the UHV MOS aging self-adaptive evaluator through the test evaluation module to analyze and process the test data stream set obtained from the aging test, so as to evaluate the aging performance of the target MOS device and obtain the final aging performance test result, that is, the test result of the device scenario aging performance. Specifically, during the test, the test evaluation module calls the latter to conduct in-depth analysis on the collected data through the connection with the UHV MOS aging self-adaptive evaluator. The UHV MOS aging self-adaptive evaluator can automatically adjust the analysis model according to the working environment and aging test data of the target MOS device to adapt to different aging test scenarios, taking into account not only the conventional electrical performance changes in the test data, but also some environmental changes, device life prediction and other factors for comprehensive evaluation.

[0037] Preferably, processing and evaluating the aging scenario test data flow set refers to screening, cleaning, analyzing, and modeling these original test data to extract valuable information, including dynamically evaluating the parameters in the data flow set according to algorithms (such as regression analysis, machine learning algorithms, etc.) to quantify the aging state and performance degradation of the device. Specifically, the ultra-high voltage MOS aging adaptive evaluator determines whether the device conforms to the expected degradation mode during aging and whether there is abnormal degradation according to the change trend of electrical performance (such as leakage current, threshold voltage, switching speed, etc.) in the data flow. It also evaluates the adaptability of the device in a specific environment according to the temperature and voltage changes in the aging test scenario. For example, under ultra-high voltage conditions, the threshold voltage and leakage current of MOS devices may change significantly, and the stability of the device needs to be evaluated. By analyzing the aging trend of the device, its possible performance in long-term use is evaluated, and the remaining service life of the device is predicted. Furthermore, the aging performance test results of the device scenario are obtained, such as the degree of performance degradation, the reliability evaluation of the device (reliability in specific environments such as high temperature and high voltage and whether it can work stably for a long time), and life prediction data (the expected life or remaining life of the device), which reflect the performance degradation of the target MOS device in a given aging test scenario. Through a data-driven approach, the accuracy and efficiency of the test are improved, and the limitations of manual judgment in traditional tests are avoided.

[0038] Further, step S500 further includes step S510, collecting and obtaining the ultra-high voltage MOS aging test data set, analyzing the key indicators of the ultra-high voltage MOS aging test data set, and determining the key evaluation index set of aging performance; step S520, performing shunt evaluation and identification on the ultra-high voltage MOS aging test data set according to the key evaluation index set of aging performance to obtain the key index aging test sample set; step S530, respectively selecting and evaluating and training the network structure for the key index aging test sample set to generate the index aging performance evaluation branch network set; step S540, integrating and fusing the index aging performance evaluation branch network set to obtain the ultra-high voltage MOS aging adaptive evaluator and storing it in the test evaluation module.

[0039] Preferably, performance data during the aging test of UHV MOS devices are collected through various sensors to obtain an aging test data set of UHV MOS, including various indicators such as voltage, current, leakage current, threshold voltage, switching speed, power loss, temperature, etc., which reflects the aging process of the device under specific test conditions. Then, the collected aging test data are analyzed to identify the core indicators that can truly reflect the aging degree of MOS devices, such as leakage current, threshold voltage change, and switching performance, etc., and then determine the key evaluation index set of aging performance, which contains the indicators that can best reflect the aging characteristics and performance degradation of the device; according to the determined key evaluation index set of aging performance, the collected aging test data are evaluated by shunting according to these key indicators, that is, separate processing and analysis are carried out on the data of different indicators, so that the change trend of each indicator can be independently evaluated, and then data sets of each type of key indicator (such as leakage current, threshold voltage, etc.) are obtained, forming a key indicator aging test sample set.

[0040] Preferably, the network structure selection and evaluation training are respectively carried out on the key indicator aging test sample set, including selecting a suitable neural network architecture (such as convolutional neural network, long short-term memory network, etc.) for modeling according to different aging test sample sets (such as leakage current, threshold voltage, etc.), and then using the test sample set of each key indicator for training. The network will learn the relationship between each key indicator and the performance aging of MOS devices, model the device aging process based on a large amount of historical data, and then generate multiple indicator aging performance evaluation branch networks, forming an indicator aging performance evaluation branch network set. Each branch network is specifically used to process one or more specific key performance indicators and perform aging prediction on these indicators; finally, the indicator aging performance evaluation branch network set is integrated and fused, that is, multiple indicator aging performance evaluation branch networks are integrated through ensemble learning or model fusion techniques such as weighted average and voting mechanism to obtain a UHV MOS aging adaptive evaluator, which can combine all key performance indicators, comprehensively evaluate the aging process of MOS devices, predict the future aging trend of the device, and then store it in the test evaluation module. The test evaluation module can call the evaluator in real time to perform real-time analysis on the newly collected MOS device aging test data to ensure that it can handle situations under different test conditions.

[0041] Further, step S510 further includes step S511 of extracting associated metrics from the UHV MOS aging test dataset to obtain a set of device aging associated evaluation metrics; step S512 of performing a clustering operation on the set of device aging associated evaluation metrics based on the UHV MOS aging test dataset to obtain an associated aging metrics clustering result; step S513 of calculating and evaluating the within-cluster divergence and between-cluster divergence of the associated aging metrics clustering result to determine a clustering divergence quality index; and step S514 of optimizing and updating the associated aging metrics clustering result and screening key metrics based on the clustering divergence quality index to determine the set of key evaluation metrics for aging performance.

[0042] Preferably, extracting associated metrics from the UHV MOS aging test dataset, that is, extracting different metrics that are associated, such as the relationship between voltage change and leakage current, and the change trend of temperature change and threshold voltage, etc., which can jointly reflect the aging process of the MOS device. These associated metrics are usually the result of the mutual influence between multiple different test data. The obtained set of device aging associated evaluation metrics may include voltage-leakage current association, temperature-threshold voltage association, and switching frequency-power consumption association. Then, perform a clustering operation on the set of device aging associated evaluation metrics based on the UHV MOS aging test dataset, and use (K-means clustering, hierarchical clustering, etc.) to classify metrics with similar aging trends or performance change patterns to generate an associated aging metrics clustering result. For example, at different temperatures, the leakage current change and threshold voltage change of some MOS devices may show similar patterns and are clustered into the same group.

[0043] Preferably, calculate and evaluate the within-cluster divergence and between-cluster divergence of the associated aging metrics clustering result. The within-cluster divergence refers to the degree of dispersion between data points within the same cluster, reflecting the tightness within the cluster. If the within-cluster divergence is small, it indicates that the metrics within the cluster have high similarity, otherwise, it indicates a poor clustering effect. The between-cluster divergence refers to the degree of dispersion between different clusters, reflecting the distinguishability between clusters. A larger between-cluster divergence indicates a significant difference between different clusters and a better clustering effect; while a smaller between-cluster divergence indicates a large overlap between different clusters and a poor clustering effect. By comprehensively calculating the within-cluster divergence and between-cluster divergence, a clustering divergence quality index is obtained to measure the quality of the clustering result. The smaller the index, the higher the clustering quality. According to the clustering divergence quality index, select and optimize the clustering result. If the quality of some clustering results is poor (i.e., the within-cluster divergence is large or the between-cluster divergence is small), then the clustering algorithm needs to be adjusted or features need to be reselected to improve the clustering quality, and then screen out the key evaluation metrics that have the greatest impact on the aging process, which can best reflect the aging performance of the MOS device, and form a set of key evaluation metrics for aging performance to ensure that the aging performance of the MOS device can be effectively evaluated.

[0044] Further, step S514 further includes step S514a of iteratively optimizing and updating the clustering result of the associated aging indicators based on the clustering divergence quality index until a preset termination condition is reached to obtain an updated clustering result of the aging indicators; step S514b of performing a relevance evaluation on each indicator cluster in the updated clustering result of the aging indicators to obtain a set of aging indicator cluster relevance degrees; step S514c of determining a set of aging indicator cluster screening thresholds according to the set of aging indicator cluster relevance degrees; and step S514d of respectively performing key indicator screening on each indicator cluster in the updated clustering result of the aging indicators based on the set of aging indicator cluster screening thresholds to determine the set of key evaluation indicators for aging performance.

[0045] Preferably, the clustering result of the associated aging indicators is iteratively optimized and updated according to the clustering divergence quality index, that is, based on the clustering divergence quality index, the preliminary clustering result is repeatedly optimized. By adjusting the clustering method and parameters, the within-cluster divergence is minimized (the compactness of the cluster is improved) and the between-cluster divergence is increased (the difference between clusters is increased), so as to obtain a more accurate clustering result. In each round of optimization, the clustering boundary is readjusted, a different clustering method is selected, or data points are reallocated until an optimal clustering structure is reached. By repeatedly comparing the difference between the current clustering result and the target quality index until a preset termination condition is reached, such as reaching a predetermined clustering divergence quality index (i.e., the clustering quality reaches a certain standard), reaching the maximum number of iterations, or the clustering result no longer changes significantly, an updated clustering result of the aging indicators is obtained; then a relevance evaluation is performed on it, that is, the correlation between different clusters is measured by statistical methods (such as Pearson correlation coefficient, etc.), and the relevance between each indicator cluster is evaluated to determine which indicator clusters have a strong mutual correlation, so as to obtain a set of aging indicator cluster relevance degrees, which describes the degree of correlation between each cluster.

[0046] Preferably, according to the relevance degree data in the set of aging indicator cluster relevance degrees, a suitable threshold is determined for screening out indicator clusters with high relevance. The aging indicator cluster screening threshold is determined by setting a fixed value or dynamic calculation (such as through statistical analysis, average relevance degree of clustering, etc.) to obtain a set of aging indicator cluster screening thresholds. Finally, according to the set of aging indicator cluster screening thresholds, each indicator cluster in the updated clustering result is screened. Based on whether the relevance degree of each cluster reaches the predetermined threshold, it is decided whether to retain the indicators in the cluster as key evaluation indicators. If the relevance degree of a cluster is higher than the threshold, then the indicators in the cluster are considered to have a significant impact on the aging process and are added to the set of key evaluation indicators for aging performance. Finally, by screening out indicator clusters with significant impacts, the set of key evaluation indicators for aging performance is determined, which includes all core indicators that can accurately reflect the aging process of MOS devices, making the evaluation process more scientific and efficient and capable of accurately evaluating the performance changes of UHV MOS devices during the aging process.

[0047] Further, step S540 further includes step S541, performing a criticality assessment on each branch network in the index aging performance evaluation branch network set to obtain a branch network criticality coefficient set; step S542, performing an output performance verification on each branch network in the index aging performance evaluation branch network set to obtain a branch network accuracy factor set; step S543, constructing a branch network decision function: , where represents the th branch network criticality coefficient, identifies the criticality coefficient of the detection factor, identifies the th branch network accuracy factor; step S544, performing parameter calculation on the branch network criticality coefficient set and the branch network accuracy factor set based on the branch network decision function to obtain a branch network decision parameter set; step S545, performing weighted integration fusion on the index aging performance evaluation branch network set based on the branch network decision parameter set to obtain the UHV MOS aging adaptive evaluator.

[0048] Preferably, each branch network represents an evaluation model for different aging performance indicators. For each branch network, its criticality in the aging test is evaluated, that is, the contribution degree of the network to the overall aging evaluation is obtained, and multiple branch network criticality coefficients are obtained, which reflect the importance of the network in the overall evaluation. Then, the output results of each branch network are verified. Usually, the actual test data is compared with the model output results to check the performance of the model in predicting the aging characteristics of MOS devices and evaluate its accuracy. Specifically, through the verification process, the prediction accuracy of each branch network is evaluated to obtain the accuracy factor. The accuracy factor represents the accuracy and reliability of the output result of the branch network. The accuracy factor set is the set of all branch network accuracy factors. If a certain branch network performs more precisely in the prediction process, the corresponding accuracy factor will be higher.

[0049] Preferably, construct the branch network decision function: , where represents the nth branch network criticality coefficient, identifies the criticality coefficient of the detection factor, identifies the nth branch network accuracy factor, identifies the (n - 1)th branch network accuracy factor, represents the total criticality coefficient of all branch networks, It represents the total accuracy factor of all branch networks. The branch network decision function weights the criticality and accuracy of each branch network, comprehensively considers the role of each branch network in the final evaluation, and thus helps to generate the final aging performance evaluation result. Then, based on the decision function, the criticality coefficient, accuracy factor, and criticality coefficient of the detection factor of all branch networks are calculated to obtain the decision parameters of each branch network, and a set of branch network decision parameters is obtained, that is, the influence degree of each branch network in the final evaluation. Finally, the weighted integration and fusion of the index aging performance evaluation branch network set is performed according to the set of branch network decision parameters to ensure that each branch network occupies an appropriate weight in the final evaluation result according to its criticality and accuracy, and a UHV MOS aging adaptive evaluator is generated. It can make a comprehensive evaluation of the aging process of MOS devices based on the input aging test data and the weighted results of each branch network, providing strong evaluation support for the long-term stability and reliability of the equipment.

[0050] In the above text, reference is made to Figure 1 a scenario adaptive aging test method for UHV MOS according to an embodiment of the present invention is described in detail. Next, reference will be made to Figure 2 describe a scenario adaptive aging test system for UHV MOS according to an embodiment of the present invention.

[0051] The scenario adaptive aging test system for UHV MOS according to an embodiment of the present invention is used to solve the technical problems in the prior art that the MOS aging test lacks scenario adaptability and it is difficult to accurately simulate the aging process of MOS devices in the actual application environment, resulting in insufficient test pertinence and poor result accuracy, and achieves the technical effect of improving the pertinence and result accuracy of the MOS device aging test. The scenario adaptive aging test system for UHV MOS includes: a device aging test platform establishment module 10, a test condition analysis module 20, a pre-test adjustment module 30, an aging test monitoring module 40, and a performance test result acquisition module 50.

[0052] The device aging test platform establishment module 10 is used to establish a device aging test platform, and the device aging test platform includes an environmental test chamber, a host computer, a test fixture, and a test evaluation module; the test condition analysis module 20 is used to analyze the test conditions of the application scenario information of the target UHV MOS device through the host computer to generate an aging test condition parameter table; the pre-test adjustment module 30 is used to fixedly install the target UHV MOS device on the test fixture, and perform pre-test adjustment on the target UHV MOS device through the environmental test chamber based on the aging test condition parameter table to determine an aging test update parameter table; the aging test monitoring module 40 is used to monitor the aging test of the target UHV MOS device based on the aging test update parameter table to obtain a device aging scenario test data flow set; the performance test result acquisition module 50 is used to call the UHV MOS aging self-adaptive evaluator through the test evaluation module, and process and evaluate the device aging scenario test data flow set based on the UHV MOS aging self-adaptive evaluator to obtain a device scenario aging performance test result.

[0053] Next, the specific configuration of the test condition analysis module 20 will be described in detail. The test condition analysis module 20 further includes: extracting elements from the application scenario information of the target UHV MOS device to obtain device application scenario element information; sequentially analyzing the test conditions of each application scenario in the application scenario information according to the device application scenario element information to obtain a device scenario test element parameter set; determining a device scenario aging test parameter set according to the device aging test target; and performing parameter design combination on the device scenario test element parameter set and the device scenario aging test parameter set to generate the aging test condition parameter table.

[0054] Next, the specific configuration of the pre-test adjustment module 30 will be further described in detail. The pre-test adjustment module 30 further includes: performing pre-test monitoring on the target UHV MOS device based on the aging test condition parameter table to obtain a device aging pre-test data flow set; performing performance evaluation on the target UHV MOS device based on the device aging pre-test data flow set to obtain a device aging performance pre-test result; performing abnormal threshold analysis on the device aging performance pre-test result to determine a device performance abnormal test parameter threshold; and adjusting and optimizing the aging test condition parameter table based on the device performance abnormal test parameter threshold to determine the aging test update parameter table.

[0055] Next, the specific configuration of the performance test result acquisition module 50 will be described in detail. The performance test result acquisition module 50 further includes: collecting and obtaining a UHV MOS aging test data set, analyzing key indicators of the UHV MOS aging test data set, and determining a key evaluation index set for aging performance; performing shunt evaluation and identification on the UHV MOS aging test data set according to the key evaluation index set for aging performance to obtain a key indicator aging test sample set; respectively performing network structure selection and evaluation training on the key indicator aging test sample set to generate an index aging performance evaluation branch network set; integrating and fusing the index aging performance evaluation branch network set to obtain a UHV MOS aging adaptive evaluator and storing it in the test evaluation module.

[0056] Next, the specific configuration of the performance test result acquisition module 50 will be further described in detail. The performance test result acquisition module 50 further includes: extracting associated indicators from the UHV MOS aging test data set to obtain a device aging associated evaluation index set; performing a clustering operation on the device aging associated evaluation index set based on the UHV MOS aging test data set to obtain an associated aging index clustering result; calculating and evaluating the within-cluster divergence and between-cluster divergence of the associated aging index clustering result to determine a clustering divergence quality index; optimizing and updating the associated aging index clustering result and screening key indicators based on the clustering divergence quality index to determine the key evaluation index set for aging performance.

[0057] Next, the specific configuration of the performance test result acquisition module 50 will be further described in detail. The performance test result acquisition module 50 further includes: iteratively optimizing and updating the associated aging index clustering result based on the clustering divergence quality index until a preset termination condition is reached to obtain an aging index clustering update result; performing a relevance evaluation on each index cluster in the aging index clustering update result to obtain an aging index cluster relevance degree set; determining an aging index cluster screening threshold set according to the aging index cluster relevance degree set; screening key indicators for each index cluster in the aging index clustering update result based on the aging index cluster screening threshold set to determine the key evaluation index set for aging performance.

[0058] Next, the specific configuration of the performance test result acquisition module 50 will be further described in detail. The performance test result acquisition module 50 further includes: performing a criticality evaluation on each branch network in the index aging performance evaluation branch network set to obtain a branch network criticality coefficient set; performing an output performance verification on each branch network in the index aging performance evaluation branch network set to obtain a branch network accuracy factor set; constructing a branch network decision function: , where represents the criticality coefficient of the nth branch network, Identify the criticality coefficient of the detection factor, Identify the precision factor of the nth branch network; perform parameter calculations on the criticality coefficient set and the precision factor set of the branch network based on the branch network decision function to obtain a branch network decision parameter set; perform weighted integration fusion on the index aging performance evaluation branch network set based on the branch network decision parameter set to obtain the UHV MOS aging adaptive evaluator.

[0059] The scenario adaptive aging test system for UHV MOS provided by the embodiments of the present invention can execute the scenario adaptive aging test method for UHV MOS provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the method.

[0060] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, however, any number of different modules can be used and run on the user terminal and / or the server. The included units and modules are only divided according to the 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 mutual distinction and do not limit the protection scope of the present invention.

[0061] The above specific embodiments do not constitute a limitation to the protection scope of the present application. Those skilled in the art should understand 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 principle of the present application shall be included within the protection scope of the present application.

Claims

1. A scenario-adaptive aging test method for UHV MOS, characterized in that: The method comprises: Establishing a device aging test platform, the device aging test platform includes an environmental test chamber, a host computer, a test fixture and a test evaluation module; The host computer performs test condition analysis on the application scenario information of the target UHV MOS device to generate an aging test condition parameter table; The target UHV MOS device is fixedly mounted on the test fixture, and a pre-test adjustment is performed on the target UHV MOS device based on the aging test condition parameter table through the environmental test chamber to determine an aging test update parameter table; Perform aging test monitoring on the target UHV MOS device based on the aging test update parameter table to obtain a device aging scenario test data stream set; Calling the UHV MOS aging adaptive evaluator through the test evaluation module, processing and evaluating the device aging scenario test data stream set based on the UHV MOS aging adaptive evaluator, and obtaining the device scenario aging performance test result; The calling of the UHV MOS aging adaptive evaluator includes: Acquiring a UHV MOS aging test data set, performing key indicator analysis on the UHV MOS aging test data set, and determining a key evaluation indicator set for aging performance; According to the aging performance key evaluation indicator set, the UHV MOS aging test data set is shunted and evaluated to obtain a key indicator aging test sample set; Performing network structure selection and evaluation training on the key indicator aging test sample sets respectively to generate an indicator aging performance evaluation branch network set; Integrate and fuse the index aging performance evaluation branch network set to obtain a UHV MOS aging adaptive evaluator and store it in the test evaluation module; Wherein, the determination of the key evaluation index set of aging performance includes: Extracting related indicators from the UHV MOS aging test data set to obtain a device aging related evaluation indicator set; Based on the UHV MOS aging test data set, a clustering operation is performed on the device aging associated evaluation index set to obtain a clustering result of the associated aging index; Performing intra-cluster divergence and inter-cluster divergence calculation and evaluation on the clustering results of the associated aging indicators to determine a cluster divergence quality index; Based on the cluster divergence quality index, the clustering results of the associated aging indicators are optimized and updated, and key indicators are screened to determine the key evaluation indicator set of aging performance; Wherein, determining the aging performance key evaluation indicator set includes: Iteratively optimizing and updating the associated aging indicator clustering result based on the cluster divergence quality index until a preset termination condition is reached to obtain an aging indicator clustering update result; Performing a correlation evaluation on each indicator cluster in the aging indicator clustering update result to obtain an aging indicator cluster correlation degree set; Determining an aging indicator cluster screening threshold set according to the aging indicator cluster association degree set; Based on the aging indicator cluster screening threshold set, each indicator cluster in the aging indicator cluster update result is screened for key indicators to determine the aging performance key evaluation indicator set; Wherein, obtaining the UHV MOS aging adaptive evaluator comprises: Performing a criticality assessment on each branch network in the indicator aging performance assessment branch network set to obtain a branch network criticality coefficient set; Verifying the output performance of each branch network in the indicator aging performance evaluation branch network set to obtain an accurate factor set of the branch network; Construct a branch network decision function: , where Fn represents the criticality coefficient of the nth branch network, δn identifies the criticality coefficient of the detection factor, and an identifies the precise factor of the nth branch network; Perform parameter calculation on the branch network criticality coefficient set and the branch network precise factor set based on the branch network decision function to obtain a branch network decision parameter set; The indicator aging performance evaluation branch network set is weightedly integrated and fused based on the branch network decision parameter set to obtain the ultra-high voltage MOS aging adaptive evaluator.

2. The scenario-adaptive aging test method for UHV MOS according to claim 1, characterized in that: The step of generating the aging test condition parameter table includes: Extracting elements of the application scenario information of the target UHV MOS device to obtain device application scenario element information; According to the device application scenario element information, test condition analysis is performed on each application scenario in the application scenario information in turn to obtain a device scenario test element parameter set; Determine the device scenario aging test parameter set according to the device aging test objectives; The device scenario test factor parameter set and the device scenario aging test parameter set are combined by parameter design to generate the aging test condition parameter table.

3. The scenario-adaptive aging test method for UHV MOS according to claim 1, characterized in that: The step of determining an aging test update parameter table includes: Performing pre-test monitoring on the target UHV MOS device based on the aging test condition parameter table to obtain a device aging pre-test data stream set; Performing performance evaluation on the target UHV MOS device based on the device aging pre-test data stream set to obtain a device aging performance pre-test result; Performing abnormal threshold analysis on the device aging performance pre-test results to determine device performance abnormality test parameter thresholds; The aging test condition parameter table is adjusted and optimized based on the device performance abnormality test parameter threshold to determine the aging test update parameter table.

4. The scenario-adaptive aging test system for UHV MOS is characterized by: The system is used to implement the scenario-adaptive aging test method for UHV MOS according to any one of claims 1 to 3, and the system includes: A device aging test platform establishment module is used to establish a device aging test platform, wherein the device aging test platform includes an environmental test chamber, a host computer, a test fixture and a test evaluation module; A test condition analysis module, used to analyze the application scenario information of the target UHV MOS device through the host computer to generate an aging test condition parameter table; A pre-test adjustment module, used for fixing the target UHV MOS device on the test fixture, performing pre-test adjustment on the target UHV MOS device based on the aging test condition parameter table through the environmental test chamber, and determining an aging test update parameter table; An aging test monitoring module, used to perform aging test monitoring on the target UHV MOS device based on the aging test update parameter table, and obtain a device aging scenario test data stream set; The performance test result acquisition module is used to call the UHV MOS aging adaptive evaluator through the test evaluation module, process and evaluate the device aging scenario test data stream set based on the UHV MOS aging adaptive evaluator, and obtain the device scenario aging performance test result.

Images