CN115398251A - Power semiconductor device's test system, cloud server and test machine - Google Patents

Abstract

A test system, a cloud server and a tester of a power semiconductor device are provided, the system comprises: a cloud server (100) and at least one tester (200); the cloud server (100) is used for training by utilizing historical data of a plurality of power semiconductor devices in advance to obtain an artificial intelligence model; a tester (200) for obtaining data of a power semiconductor device to be tested; the cloud server (100) or the tester (200) is used for inputting data of the power semiconductor device to be tested into the artificial intelligence model, and the output of the artificial intelligence model is the diagnosis result of the power semiconductor device to be tested. The cloud server (100) or the testing machine (200) diagnoses the power semiconductor device to be tested by using the trained artificial intelligence model, so that an accurate diagnosis result can be obtained, and abnormal power semiconductor devices are removed, so that the whole circuit is prevented from being failed due to the fact that the failure occurs in the using process of an actual product.

Description

A test system, cloud server and test machine for power semiconductor devices

technical field

The present application relates to the technical field of power semiconductors, in particular to a testing system, cloud server and testing machine for power semiconductor devices.

Background technique

A power semiconductor device generally refers to a semiconductor device that realizes a circuit switching function, for example, it can be applied to a power conversion circuit of an electric vehicle, a photovoltaic system, or a data center. For example, in electric vehicles, power semiconductor devices can be used in the powertrain control circuit of electric vehicles. In practical applications, in order to ensure the safety of circuits used by power semiconductor devices, performance testing of power semiconductor devices is required after production to reduce the failure risk of power semiconductor devices.

At present, the quality testing of power semiconductor devices mainly sets the upper and lower limit values for the performance parameters of power semiconductor devices through the testing machine, and eliminates the power semiconductor devices whose performance parameters exceed the upper and lower limit values.

However, the current method of testing with the upper and lower limit values needs to be confirmed by a large number of historical products, and the screening method through the upper and lower limit values is highly dependent on experience, and cannot establish a complete correlation with the failure principle of the device. It can only identify Some power semiconductor devices with problems.

Contents of the invention

The present application provides a test system for power semiconductor devices, a cloud server and a test machine, which can more comprehensively screen defective power semiconductor devices.

The test system of the power semiconductor device provided by the embodiment of the present application includes a cloud server and at least one test machine; the cloud server is used to pre-train the historical data of a plurality of power semiconductor devices to obtain an artificial intelligence (AI, Artificial Intelligence) model; The testing machine is used to obtain the data of the power semiconductor device to be tested; the cloud server or the testing machine can use the trained artificial intelligence model to diagnose the power semiconductor device to be tested, that is, the cloud server or the testing machine inputs the data of the power semiconductor device to be tested The artificial intelligence model, the output of the artificial intelligence model is the diagnosis result of the power semiconductor device to be tested.

In the solution provided by the embodiment of this application, a relatively accurate artificial intelligence model can be trained by using the powerful computing and storage capabilities of the cloud server, and then the pre-trained artificial intelligence model can be used to more accurately diagnose the power semiconductor device to be tested. High-risk power semiconductor devices, so as to make up for the limitations of traditional test systems that only use upper-limit data to screen high-risk power semiconductor devices, that is, it is possible to more accurately and comprehensively diagnose whether there are risks in power semiconductor devices.

In a possible implementation, in the embodiment of the present application, the cloud server is used to input the data of the power semiconductor device to be tested into the artificial intelligence model, and the output of the artificial intelligence model is the diagnostic result of the power semiconductor device to be tested, and the diagnostic result sent to the test machine. It should be understood that in the embodiment of the present application, the cloud server inputs the data into the artificial intelligence model, obtains the diagnosis result, and sends the diagnosis result to the test machine, so the test machine can directly obtain the test result, saving the computing resources of the test machine.

In a possible implementation, the cloud server in the embodiment of the present application is used to send the pre-trained artificial intelligence model to the testing machine; the testing machine inputs the data of the power semiconductor device to be tested into the artificial intelligence model, and the output of the artificial intelligence model is the diagnostic result of the power semiconductor device to be tested. It should be understood that the cloud server sends the trained artificial intelligence model to the test machine, and then the test machine inputs data and obtains a diagnosis result. Therefore, the testing machine alone completes the testing work of the power semiconductor device to be tested, and does not need to obtain diagnostic results from the cloud server, which reduces the dependence on the network between the cloud server and the testing machine, and improves the testing of the power semiconductor device in the embodiment of the present application. The system obtains real-time diagnosis results.

In a possible implementation, the testing machine in the embodiment of the present application is also used to collect historical data of multiple power semiconductor devices, and send the collected historical data of multiple power semiconductor devices to the cloud server; historical data includes At least one of the chip test data of multiple power semiconductor devices, the packaged module function test data, or the relevant data when the actual application is abnormal; the cloud server is also used to utilize the historical data of multiple power semiconductor devices to utilize supervised learning At least one of the model or the unsupervised learning model is trained to obtain an artificial intelligence model.

In a possible implementation, the cloud server uses unsupervised learning model training to obtain an artificial intelligence model, which is specifically used to obtain the failure types of multiple power semiconductor devices, and uses expert knowledge to extract each failure type in the test items of power semiconductor devices. All test items corresponding to each type of failure form a subset of test items, and use an unsupervised learning model to distinguish between abnormal subsets and normal subsets for each test item subset, and obtain abnormal subsets and normal subsets. Subsets record the first score and the second score respectively, and the sum of the first score and the second score of all test item subsets of each power semiconductor device is obtained as the total score of the abnormal level; the total score of the abnormal level is greater than or equal to the preset score threshold, it is judged that the power semiconductor device is an abnormal device. It should be understood that the embodiment of the present application adopts the unsupervised learning model to train the AI model, which can make full use of expert knowledge, and jointly perform abnormal point detection on the test item subsets containing multiple test items, and can use the same test item subset. The connection of the same failure type, or the physical basis corresponding to the same failure type, detects the risk of power semiconductor device failure from multiple dimensions, so that the risk of device failure can be more accurately assessed.

In a possible implementation manner, the cloud server in the embodiment of the present application is specifically configured to use expert knowledge to record different first scores for abnormal devices in different test item subsets. It should be understood that different test item subsets may have different weights for judging abnormal devices, and different first scores can be recorded for abnormal devices in different test item subsets, so that more accurate diagnosis results can be obtained.

In a possible implementation, in the embodiment of the present application, the cloud server uses a supervised learning model to train to obtain an artificial intelligence model, which is specifically used to obtain relevant data when power semiconductor devices are abnormal as data labels, and use a supervised learning model to extract data The data characteristics of the label, using the data characteristics to obtain the total score of the abnormal level of each power semiconductor device, when the total score of the abnormal level is greater than the preset score threshold, it is judged that the power semiconductor device is an abnormal device.

In a possible implementation, the testing machine in the embodiment of the present application is also used to compare the data of the power semiconductor device to be tested with the preset upper and lower limit values, when the data of the power semiconductor device to be tested exceeds the upper and lower limit values , judging that the power semiconductor device to be tested is an abnormal device. It should be understood that the test system for power semiconductor devices provided by the embodiment of the present application can combine the traditional test method with the artificial intelligence model, and use the artificial intelligence model in the embodiment of the present application to carry out secondary testing on the basis of the traditional test method. Detection, improving the accuracy of detection, realizing the complementarity between traditional methods and AI model testing methods, can more comprehensively diagnose high-risk power semiconductor devices.

The embodiment of the present application does not limit the number of testing machines included in the testing system, which may include one testing machine, or may include two or more testing machines. When the test system includes multiple test machines, the cloud server can first train the global AI model, and then obtain a local AI model suitable for each test machine based on the global AI model based on the differences between the test machines. That is, in a possible implementation manner, the test system provided by the embodiment of the present application includes at least the following two test machines: the first test machine and the second test machine; the cloud server is specifically used to The historical data and the second historical data sent by the second testing machine are trained to obtain the global artificial intelligence model, the first historical data is used to adjust the global artificial intelligence model to obtain the first artificial intelligence model, and the second historical data is used to adjust the global artificial intelligence model Adjust to obtain the second artificial intelligence model; use the first artificial intelligence model to test the power semiconductor device to be tested corresponding to the first testing machine to obtain the first diagnosis result, and use the second artificial intelligence model to test the power semiconductor device to be tested corresponding to the second testing machine The power semiconductor device is tested to obtain a second diagnosis result, the first diagnosis result is sent to the first testing machine, and the second diagnosis result is sent to the second testing machine. In the embodiment of the present application, the first historical data and the second historical data are used to train the global artificial intelligence model, which can make full use of all historical data and improve the test accuracy of the global artificial intelligence model. On this basis, considering that there may be differences in test performance between different testing machines, the implementation of this application also uses the first historical data to adjust the global artificial intelligence model to obtain the first artificial intelligence model, and uses the second historical data to adjust the global artificial intelligence model. The artificial intelligence model is adjusted to obtain the second artificial intelligence model, so that the partial artificial intelligence model (the first artificial intelligence model) for the characteristics of the first test machine and the partial artificial intelligence model (the second artificial intelligence model) for the characteristics of the second test machine can be obtained. model), which further improves the test accuracy of the artificial intelligence model used in actual detection.

In this embodiment, each test machine corresponds to the global AI model, and the diagnosis result is obtained by the cloud server. That is, in a possible implementation manner, the test system provided by the embodiment of the present application includes at least the following two test machines: the first test machine and the second test machine; The first historical data and the second historical data sent by the second testing machine are trained to obtain a global artificial intelligence model, and the global artificial intelligence intelligent model is used to test the power semiconductor device to be tested corresponding to the first testing machine to obtain a first diagnosis result. The artificial intelligence model tests the power semiconductor device to be tested corresponding to the second testing machine to obtain a second diagnostic result, sends the first diagnostic result to the first testing machine, and sends the second diagnostic result to the second testing machine. The cloud server provided by the embodiment of the present application uses the first historical data and the second historical data to train the global artificial intelligence model, and can make full use of all the historical data, thereby improving the generalization ability of the global artificial intelligence model and utilizing the more accurate global artificial intelligence model. The artificial intelligence model can diagnose the power semiconductor device to be tested, which can improve the universality of the artificial intelligence model and increase the accuracy of new device diagnosis.

In this embodiment, the cloud server obtains the local AI models corresponding to each testing machine and sends them to each testing machine, and each testing machine completes the diagnosis. That is, in a possible implementation, the power semiconductor device testing system in the embodiment of the present application includes at least the following two testing machines: a first testing machine and a second testing machine; the cloud server is specifically used to The first historical data sent and the second historical data sent by the second testing machine are trained to obtain the global artificial intelligence model, and the first historical data is used to adjust the global artificial intelligence model to obtain the first artificial intelligence model and send it to the first testing machine. Using the second historical data to adjust the global artificial intelligence model to obtain the second artificial intelligence model and send it to the second testing machine; the first testing machine is specifically used to use the first artificial intelligence model to diagnose the corresponding power semiconductor device to be tested; The second testing machine is specifically used for diagnosing the corresponding power semiconductor device to be tested by using the second artificial intelligence model. The testing machine of the embodiment of the present application can directly use the artificial intelligence model to complete the task of detecting the power semiconductor device to be tested by a single machine, and does not need to communicate with the cloud server in the cloud, so the network between the testing machine at the board end and the cloud server in the cloud can be avoided When interrupted, the power semiconductor device testing system in the embodiment of the present application cannot obtain the diagnosis result of the power semiconductor device to be tested. In this way, the power semiconductor device test system provided in the embodiment of the present application can reduce the network dependence between the cloud server and the test machine, and improve the real-time performance of the diagnostic results obtained by the power semiconductor device test system in the embodiment of the present application.

In this embodiment, the cloud server sends the global AI model to each test machine, and each test machine uses the global AI model to complete the diagnosis. That is, in a possible implementation, the power semiconductor device testing system in the embodiment of the present application includes at least the following two testing machines: the first testing machine and the second testing machine; The first historical data sent by the computer and the second historical data sent by the second testing machine are trained to obtain the global artificial intelligence model, and the global artificial intelligence model is sent to the first testing machine and the second testing machine; the first testing machine is used for Using the global artificial intelligence intelligent model to diagnose the corresponding power semiconductor device to be tested; the second testing machine is used to use the global artificial intelligence model to diagnose the corresponding power semiconductor device to be tested. According to the trained global artificial intelligence model, the testing machine in the embodiment of the present application can complete the testing work of the power semiconductor device to be tested by a single machine. When the test machine obtains the diagnosis result, the cloud server does not need to send the diagnosis result to the test machine. Therefore, the diagnosis result obtained by the test machine is not affected by network faults and has low dependence on the network.

In a possible implementation manner, the testing machine in the embodiment of the present application is also used to send update data to the cloud server; the cloud server is also used to update the artificial intelligence model according to the update data. In addition, the testing machine is also used to fine-tune the pre-trained AI model in combination with the data on the testing machine side, use the fine-tuned model to diagnose the power semiconductor device to be tested, and output the diagnosis result. It should be understood that in the embodiment of the present application, the testing machine can fine-tune the obtained AI model based on its own data or characteristics of the data, which can further improve the accuracy of the AI model in the embodiment of the present application. The test machine can send the fine-tuned AI model to the cloud server for unified management.

Based on the test system for power semiconductor devices provided in the above embodiments, the embodiments of the present application also provide a cloud server. The advantages of the above test system embodiments are also applicable to the following servers, and will not be repeated here. The server includes: a first transceiver device and a first controller; the first transceiver device is used to receive the test data of the power semiconductor device to be tested sent by the testing machine; the first controller is used to pre-use the power semiconductor device The historical data is trained to obtain the artificial intelligence model; it is also used to input the data of the power semiconductor device to be tested into the artificial intelligence model, and the output of the artificial intelligence model is the diagnosis result of the power semiconductor device to be tested; The diagnosis result is sent to the testing machine; or, the first controller is configured to send the artificial intelligence model to the testing machine, so that the testing machine uses the artificial intelligence model to diagnose the power semiconductor device to be tested.

In a possible way, the first controller in the embodiment of the present application is specifically used to obtain the failure types of multiple power semiconductor devices when using unsupervised learning model training to obtain the artificial intelligence model, and uses expert knowledge All test items corresponding to each failure type are extracted from the test items of power semiconductor devices, and all test items corresponding to each failure type form a subset of test items, and an unsupervised learning model is used for abnormal level detection for each subset of test items , get the first score and the second score for the abnormal device and the normal device respectively, and obtain the sum of the first score and the second score of all test item subsets of each power semiconductor device as the total score of the abnormal level; the total score of the abnormal level is greater than When the score threshold is preset, it is judged that the power semiconductor device is an abnormal device.

In a possible manner, the cloud server in the embodiment of the present application corresponds to at least two test machines as follows: the first test machine and the second test machine; the first controller is specifically configured to The first historical data and the second historical data sent by the second testing machine are trained to obtain the global artificial intelligence model, and the first historical data is used to adjust the global artificial intelligence model to obtain the first artificial intelligence model, and the second historical data is used to adjust the global artificial intelligence model. Adjust the model to obtain the second artificial intelligence model; use the first artificial intelligence model to test the power semiconductor device to be tested corresponding to the first testing machine to obtain the first diagnosis result, and use the second artificial intelligence model to test the power semiconductor device to be tested corresponding to the second testing machine The power semiconductor device is tested to obtain a second diagnosis result; the first transceiver device is specifically used to send the first diagnosis result to the first tester, and is also used to send the second diagnosis result to the second tester.

In a possible manner, the cloud server in the embodiment of the present application corresponds to at least two test machines as follows: the first test machine and the second test machine; the first controller is specifically configured to The first historical data and the second historical data sent by the second testing machine are trained to obtain a global artificial intelligence model, and the global artificial intelligence intelligent model is used to test the power semiconductor device to be tested corresponding to the first testing machine to obtain a first diagnosis result. The artificial intelligence model tests the power semiconductor device to be tested corresponding to the second testing machine to obtain a second diagnostic result; the first transceiver device is specifically used to send the first diagnostic result to the first testing machine, and is also used to send the second diagnostic result to the first testing machine. The results are sent to the second testing machine.

In a possible manner, the first transceiver device in the embodiment of the present application is further configured to receive update data sent by the testing machine; the first controller is also configured to update the artificial intelligence model according to the update data.

Based on the test system and server for power semiconductor devices provided in the above embodiments, the embodiment of the present application also provides a test machine. The advantages of the above test system embodiments are also applicable to the following test machines, and will not be repeated here. The testing machine includes: a second transceiver device and a second controller; the second transceiver device is used to receive the artificial intelligence model sent by the cloud server, and the artificial intelligence model is obtained by the cloud server in advance using the historical data of multiple power semiconductor devices for training The second controller is used to obtain the data of the power semiconductor device to be tested; input the data of the power semiconductor device to be tested into the artificial intelligence model, and the output of the artificial intelligence model is the diagnosis result of the power semiconductor device to be tested.

In a possible manner, the second controller in the embodiment of the present application is also used to collect historical data of multiple power semiconductor devices, and send the collected historical data of multiple power semiconductor devices to the cloud server; The data includes at least one of chip test data or packaged module function test data of multiple power semiconductor devices.

In a possible manner, the second controller in the embodiment of the present application is also used to compare the data of the power semiconductor device to be tested with the preset upper and lower limit values, and the data of the power semiconductor device to be tested exceeds the upper and lower limits value, it is judged that the power semiconductor device under test is an abnormal device.

It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

The test system provided in the embodiment of the present application includes a cloud server and a test machine. The computing power and storage capacity of the cloud server are higher than that of the test machine. Therefore, the artificial intelligence model is obtained through pre-training using the powerful computing power and storage capacity of the cloud server. Since the cloud server uses a large amount of historical data of power semiconductor devices to train the artificial intelligence model in advance, the parameters of the artificial intelligence model can be accurately obtained. Therefore, the cloud server or test machine can accurately diagnose the power semiconductor device to be tested by using the trained artificial intelligence model , so as to obtain accurate diagnosis results, and remove abnormal power semiconductor devices, so as to avoid the failure of the entire circuit due to failure during the actual product use. This test system can make up for the disadvantages of the traditional method, that is, traditionally only use the upper and lower limit values to screen high-risk power semiconductor devices, and some cannot be screened out. The test system provided by the embodiment of this application can further screen out high-risk power semiconductor devices, namely Power semiconductor devices can be more comprehensively diagnosed at risk. This scheme can make up for the inaccuracy of simply screening defective products of power semiconductor devices by the testing machine using upper and lower limit values.

Description of drawings

FIG. 1 is a structural diagram of a test system for a power semiconductor device provided in an embodiment of the present application;

FIG. 2 is a schematic diagram of a data source of a power semiconductor device provided by the implementation of the present application;

FIG. 3A is a flow chart of a testing method for a power semiconductor device provided in an embodiment of the present application;

FIG. 3B is a flow chart of another power semiconductor device testing method provided by the embodiment of the present application;

FIG. 4 is a schematic diagram of a test system for another power semiconductor device provided in an embodiment of the present application;

FIG. 5 is a schematic diagram of another test system for a power semiconductor device provided in an embodiment of the present application;

FIG. 6 is a schematic diagram of another test system for a power semiconductor device provided in an embodiment of the present application;

FIG. 7 is a schematic diagram of another test system for a power semiconductor device provided in an embodiment of the present application;

FIG. 8 is a schematic diagram of a testing system for a power semiconductor device including multiple testing machines provided in an embodiment of the present application;

FIG. 9 is a schematic diagram of another testing system for a power semiconductor device including multiple testing machines provided in an embodiment of the present application;

FIG. 10 is a schematic diagram of another testing system for a power semiconductor device including multiple testing machines provided by the embodiment of the present application;

FIG. 11 is a schematic diagram of another test system for a power semiconductor device including multiple test machines provided in the embodiment of the present application;

FIG. 12 is a schematic structural diagram of a cloud server provided by an embodiment of the present application;

FIG. 13 is a schematic structural diagram of a testing machine provided in an embodiment of the present application.

Detailed ways

Words such as "first" and "second" in the following descriptions are used for description purposes only, and should not be understood as indicating or implying relative importance or implicitly indicating the quantity of indicated technical features. Thus, a feature defined as "first", "second", etc. may expressly or implicitly include one or more of that feature. In the description of the present application, unless otherwise specified, "plurality" means two or more.

In addition, in this application, orientation terms such as "upper" and "lower" may include, but are not limited to, definitions relative to the schematic placement orientations of components in the drawings. It should be understood that these directional terms may be relative concepts, They are used for description and clarification relative to, which may change accordingly according to changes in the orientation in which parts of the figures are placed in the figures.

In this application, unless otherwise specified and limited, the term "connection" should be understood in a broad sense, for example, "connection" can be a fixed connection, a detachable connection, or an integral body; it can be a direct connection, or It can be connected indirectly through an intermediary. In addition, the term "coupled" may be an electrical connection for signal transmission. "Coupling" can be a direct electrical connection, or an indirect electrical connection through an intermediary.

An embodiment of the present application relates to a testing system for a power semiconductor device, wherein the testing system for a power semiconductor device includes: a cloud server and at least one testing machine. Cloud servers have powerful computing energy and storage space. Therefore, cloud servers can use a large amount of historical data of power semiconductor devices to train artificial intelligence models. Since the AI module is trained through a large amount of historical data, the cloud server or testing machine can accurately diagnose the power semiconductor device to be tested by using the trained AI model, and remove the abnormal power semiconductor device to prevent it from being used in the real process. A fault occurs that causes the entire circuit to fail. This solution can solve the inaccuracy of simply screening defective products of power semiconductor devices to be tested by using the preset upper and lower limit values of the testing machine in the traditional solution.

In order for those skilled in the art to better understand the technical solutions provided by the embodiments of the present application, the test system provided by the embodiments of the present application will be described in detail below with reference to the accompanying drawings. The test system includes a cloud server and at least one test machine.

The embodiment of the present application does not specifically limit the number of testing machines, which can be set according to the actual number of product lines. One testing machine can correspond to one product line, and one product line can also correspond to multiple testing machines. The types of power semiconductor devices corresponding to multiple product lines may be the same or different. In the following, three testing machines are taken as an example in conjunction with accompanying drawing 1. It should be understood that there may be one testing machine, two testing machines, or more testing machines.

System embodiment

Referring to FIG. 1 , this figure is a structural diagram of a test system for a power semiconductor device provided by an embodiment of the present application.

As shown in FIG. 1 , the power semiconductor device testing system provided by the embodiment of the present application includes a cloud server 100 , a first testing machine 201 , a second testing machine 202 and a third testing machine 203 . Wherein, the cloud service 100 is located on a cloud platform, and the cloud platform may include multiple cloud servers, and only one cloud server is shown in the figure as an example. The first testing machine 201 , the second testing machine 202 and the third testing machine 203 at the board end can respectively communicate with the cloud server 100 in the cloud for data exchange or transmission.

Wherein, the cloud server 100 is configured to use historical data of a plurality of power semiconductor devices for training in advance to obtain an AI model.

It should be understood that the embodiment of the present application does not specifically limit the specific model used by the cloud server 100 for AI model training, for example, a supervised learning model or an unsupervised learning model may be used. The historical data used by the cloud server 100 to train the AI model can be the data uploaded by the first testing machine 201, the second testing machine 202, and the third testing machine 203, or it can be the data uploaded by more testing machines. Not specifically limited. It should be understood that, when the cloud server 100 performs AI model training, the more historical data is used, the more accurate the parameters of the AI model obtained after training will be.

The first testing machine 201 , the second testing machine 202 and the third testing machine 203 are used to obtain data corresponding to power semiconductor devices to be tested.

During the specific test, the cloud server 100 may perform the test, or the cloud server 100 may send the trained AI model to the test machine, and the test machine may perform the test.

In the first type, the cloud server 100 performs a test.

The cloud server 100 is used to input the data of the power semiconductor device to be tested into the pre-trained artificial intelligence model, and the output of the artificial intelligence model is the diagnosis result of the power semiconductor device to be tested. Send the diagnostic results to the tester.

It should be understood that the power semiconductor device to be tested here generally refers to any power semiconductor device that needs to be tested; the testing machine here generally refers to any testing machine.

The second type: test machine for testing.

The first testing machine 201, the second testing machine 202 and the third testing machine 203 are respectively used to input the data corresponding to the power semiconductor device to be tested into the pre-trained artificial intelligence model respectively, and obtain the output of the pre-trained model as respective The diagnosis result of the corresponding power semiconductor device to be tested.

The embodiment of the present application does not limit whether the first testing machine 201, the second testing machine 202, and the third testing machine 203 test the corresponding power semiconductor devices to be tested at the same time, or test them sequentially, because the testing action of each testing machine can be Completed independently without interfering with each other.

It can be understood that, in order to improve the accuracy of intercepting defective power semiconductor devices in the embodiments of the present application, an artificial intelligence model is used to test the performance of power semiconductor devices. However, due to the limited computing power and storage capacity of the tester on the board, training the artificial intelligence model using the tester on the board takes a long time and takes up too many computing resources, affecting other functions of the tester. Therefore, the power semiconductor device test system provided in the embodiment of the present application uses the powerful computing power and storage capacity of the cloud server to obtain an artificial intelligence model through the training of historical data of multiple power semiconductor devices, and then uses the artificial intelligence model to treat the test power Semiconductor devices are tested. The AI model trained by the cloud server can be sent to the test machine, and the test machine uses the AI model trained by the cloud server to test the power semiconductor device to be tested, and finally outputs the diagnosis result.

It can be seen that the power semiconductor device testing system provided in the embodiment of the present application can use the artificial intelligence model to diagnose the performance of the power semiconductor and screen defective products in the power semiconductor device. And considering the limited computing power and storage capacity of the testing machine on the board side, this application uses a cloud server with powerful computing power and storage capacity to train the artificial intelligence model, so as to obtain an artificial intelligence model with high diagnostic accuracy.

The above content mainly introduces the architecture of the test system for power semiconductor devices provided by the embodiment of the present application. The following describes the specific process of AI model training performed by the cloud server with reference to the accompanying drawings.

Referring to FIG. 2 , this figure is a schematic diagram of a data source of a power semiconductor device provided by the implementation of the present application.

In the embodiment of this application, the cloud server uses a large amount of historical data for training to obtain the AI model. Historical data can include production line data and application-side data.

Wherein, the source of the production line data D1 may include data obtained from the following types of tests: chip (CP, ChipProbe) test, functional test (FT, Functional Test), single-board test and complete machine test. Wherein, the CP test may be a test of a single chip, and the FT test may be a test of a packaged module with certain circuit functions. The single board test can be the upper board test, and the whole machine test can be the upper machine test. It can be understood that the chip test data and the power test data can be used separately for AI model training, or can be used jointly for AI model training.

It should be noted that the CP test is mainly aimed at a single chip, and a single chip can be a single insulated gate bipolar transistor (IGBT, Insulated Gate Bipolar Transistor,), a single metal-oxide semiconductor field effect transistor (MOSFET, Metal-Oxide -Semiconductor Field-Effect Transistor) chip, diode (Doide), triode (Bipolar Junction Transistor), thyristor (Thyristor), integrated gate commutation thyristor (IGCT, Integrated Gate-Commutated Thyristor) and other power semiconductor devices, can also be used for A chip formed of a plurality of IGBTs may also be a chip formed of IGBTs and diodes. The FT test is mainly aimed at a packaged module,

The application-side data D2 mainly includes relevant data corresponding to failures in the actual use of power semiconductor devices after production, that is, the actual working condition data of power semiconductor devices fed back by the client side, which can include test data sources or verification data source

As an example, the test data of the power semiconductor device to be tested in the embodiment of the present application may include but not limited to the following breakdown voltage, leakage current, conduction voltage drop, parasitic capacitance, gate charge, turn-on loss and turn-off loss.

The principle of using test data to obtain diagnosis results by the test system for power semiconductor devices provided by the present application is introduced below.

The embodiment of the present application provides an AI model obtained by cloud server training, which can be obtained by using an unsupervised integrated learning model, or can be obtained by using a supervised learning model; it can also be obtained by using a supervised learning model and an unsupervised learning model Jointly obtained, not specifically limited in the embodiments of this application

The following first introduces the embodiment of the present application to provide an AI model obtained by training with an unsupervised ensemble learning model.

As an example, the unsupervised learning model in the embodiment of the present application may be an unsupervised ensemble learning model. It can be understood that the unsupervised model in the embodiment of the present application uses an unsupervised learning algorithm and expert knowledge to construct an unsupervised integrated learning model, which can make full use of expert experience and reduce the impact of too few failure samples on the accuracy of the artificial intelligence model.

In the embodiment of the present application, when the cloud server uses unsupervised learning model training to obtain the artificial intelligence model, as a possible implementation, the cloud server is specifically used to obtain the failure types of multiple power semiconductor devices, and use expert knowledge to determine the failure types of power semiconductor devices. All test items corresponding to each failure type are extracted from the test items, and all test items corresponding to each failure type form a subset of test items, and an unsupervised learning model is used to distinguish between abnormal subsets and normal subsets for each test item subset , get the first score and the second score for the abnormal subset and the normal subset respectively, and obtain the sum of the first score and the second score of all test item subsets of each power semiconductor device as the total score of the abnormal level; the total score of the abnormal level When the score is greater than or equal to the preset score threshold, it is determined that the power semiconductor device is an abnormal device.

Wherein, the failure type may specifically be various failure types that cannot be intercepted by the card control during the individual test of the power semiconductor device. For example, early failures occurred in power semiconductor devices in actual use, but they were not screened out during controllable testing.

In addition, one failure type may have problems corresponding to many test items, such as gate leakage, which may be caused by problems in the physical layer, and many test items involved are caused by problems in the physical layer. Therefore, all test items associated with a failure type are extracted to form a subset of test items corresponding to the test item.

As an example, the scheme for obtaining an artificial intelligence model through unsupervised ensemble learning model training provided by the embodiment of the present application is specifically introduced below in conjunction with Table 1 and Table 2.

Table 1

Table 2

Subset1 Subset2 Subset3 ... Abnormal Level Total Score Device A 0 0 0 0 Device B 0 1 0 1 Device C 0 0 0 0 ... ...

Among them, Gate failure, PN junction failure or withstand voltage ring failure are different failure types; IGES (GE leakage current), VTH (threshold voltage), VCESAT (conduction voltage drop between C pole and E pole), ICES (CE leakage current) and Eon (turn-on loss) etc. are different test items, as can be seen from Table 1, different failure types include different test items, and each failure type includes multiple test items, therefore, this application In the embodiment, when the AI model is trained, the data used are all multi-dimensional data, not traditional one-dimensional data. Therefore, the AI model obtained through training is more accurate and comprehensive. When the trained AI model is finally used for diagnosis, the output Diagnosis is also more accurate.

Wherein, device A, device B and device C are different power semiconductor devices respectively. Subset1, Subset2 and Subset3 are different subsets of test items; that is, each device includes at least three subsets of test items. The content and quantity of test items included in each test item subset are different. In the table, the second score of the abnormal subset is recorded as 1, and the first score of the normal subset is recorded as 0, and the sum of the first score and the second score of all test item subsets of each power semiconductor device is obtained as the total score of the abnormal level ; That is, the total score of abnormal level of device A is 0, the total score of abnormal level of device B is 1, and the total score of abnormal level of device C is 0. If the total score of the abnormal level is greater than or equal to the preset score threshold, it is determined that the power semiconductor device is an abnormal device. For example, if the preset score threshold is 1, then the test device B is an abnormal device.

Referring to FIG. 3A , this figure is a flow chart of a testing method for a power semiconductor device provided by an embodiment of the present application.

A method for testing a power semiconductor device provided in an embodiment of the present application includes:

S301: Obtain failure types of multiple power semiconductor devices.

As an example, failure types such as Gate failure, PN junction failure, or withstand voltage ring failure can be obtained. It can be understood that when there are more historical data of the same power semiconductor device, the more types of failures are collected and the more comprehensive it is.

S302: Using expert knowledge to extract all test items corresponding to each failure type from the test items of the power semiconductor device, and all test items corresponding to each failure type form a test item subset.

In this example, Gate failure corresponds to multiple test items such as IGES, VTH, VCESAT, and Eon, and these test items form a test item subset subset1. PN junction failure corresponds to test items such as VCESAT and ICES, and these test items form a subset2 of test items. The failure of the pressure ring corresponds to test items such as ICES, and these test items form a subset3 of test items.

S303: For each subset of test items, use an unsupervised learning model to distinguish between an abnormal subset and a normal subset, and obtain a first score and a second score for the abnormal subset and the normal subset respectively.

For example, the unsupervised learning model is used to detect outliers for the subset of test items corresponding to Gate failure, and the obtained normal device A, normal device B, and normal device C are respectively recorded with the second score. For the subset of test items corresponding to PN junction failure, the unsupervised learning model is used to detect abnormal points, and the obtained normal device A and normal device C are respectively recorded as the second score, and the abnormal device B is recorded as the first score. In the example of Table 2, the first score is 1 and the second score is 0.

In practical applications, in order to more accurately assess the failure risk of each device, as a possible implementation, the cloud server also uses expert knowledge to record different first scores for abnormal devices in different test item subsets. That is, weights corresponding to the subset of test items are added on the basis of the first score and the second score. For example, there are three failure types in total, and the three failure types correspond to three test item subsets a, b, and c respectively. The weight of a test subset is 0.2, the weight of b test subset is 0.3, and the weight of c test subset The weight is 0.5. Correspondingly, if the device A and the device B in the b-test subset are normal, then the second score of 0 can be recorded for the device A and the device B respectively. Device C is abnormal, and device C can record initial score 1*b test subset weight 0.3=first score 0.3.

S304: Obtain the sum of the first score and the second score of all test item subsets of each power semiconductor device as the total abnormal level score.

In the example corresponding to Table 1 and Table 2, the abnormal level of device A is divided into the second score 0 of the test item subset subset1, the second score 0 of the test item subset subset2 and the second score of the test item subset subset3 The sum of 0 is 0. The total score of the abnormal level of device B is the sum of the second score 0 of the test item subset subset1 and the first score 1 of the test item subset subset2 and the second score 0 of the test item subset subset3, which is 1. The total score of abnormal level of device C is 0, which is the sum of the second score 0 of the test item subset subset1, the second score 0 of the test item subset2 and the second score 0 of the test item subset subset3.

S305: When the total score of the abnormal level is greater than or equal to the preset score threshold, it is determined that the power semiconductor device is an abnormal device.

In this example, the preset score threshold may be 1, and of course, other values may also be obtained according to actual conditions, which are not limited in the implementation of this application. When the preset score threshold is 1, since the abnormal level total score of device B is 1 greater than or equal to the preset score threshold, it is determined that device B is an abnormal device.

It can be seen from this that, in the embodiment of the present application, each failure type is corresponding to multiple test items, and the test item subsets containing multiple test items are jointly performed for outlier detection, and the connection between devices in the same test item subset can be used. Contact, or the physical basis corresponding to the same failure type, detects the risk of power semiconductor device failure from multiple dimensions, so that the risk of device failure can be more accurately assessed.

The solution introduced in the above embodiment is that the cloud server obtains the artificial intelligence model through unsupervised learning model training. As another possible implementation manner, the cloud server obtains the artificial intelligence model through supervised learning model training.

Referring to FIG. 3B , this figure is a flow chart of another power semiconductor device testing method provided by the embodiment of the present application.

The test method of the power semiconductor device provided in the embodiment of the present application includes:

S311: Obtain relevant data when the power semiconductor device is abnormal as a data label. For example, a very small number of failure samples can be selected as data labels.

S312: Using a supervised learning model to extract data features of data labels.

S313: Obtain the total abnormality level score of each power semiconductor device by using the data feature.

S314: When the total score of the abnormal level is greater than the preset score threshold, it is determined that the power semiconductor device is an abnormal device. It can be seen that, the solutions provided in the embodiments of the present application can obtain an artificial intelligence model for diagnosing power semiconductor devices through a supervised learning model or an unsupervised learning model.

It should be understood that the method for screening defective power semiconductor devices using an artificial intelligence model in the embodiment of the present application is an improved solution based on the traditionally set upper and lower limit numerical diagnosis methods. That is, the power semiconductor device testing method in the embodiment of the present application is that the testing machine uses the preset upper and lower limit values to first diagnose the power semiconductor device to be tested, that is, the testing machine is also used to compare the data of the power semiconductor device to be tested with The preset upper and lower limit values are compared, and when the data of the power semiconductor device to be tested exceeds the upper and lower limit values, it is judged that the power semiconductor device to be tested is an abnormal device. Then the testing machine or cloud server uses the AI model to diagnose the power semiconductor device to be tested, and perform logical OR operation on the diagnosis results of the two schemes. Therefore, the testing system provided by the embodiment of the present application can complement the two schemes, so as to more comprehensively diagnose the power semiconductor device to be tested.

It can be understood that the power semiconductor device testing system provided by the embodiment of the present application can use the artificial intelligence model to diagnose the power semiconductor device, and can also use the traditional method to compare the data of the power semiconductor device to be tested with the preset upper and lower limits Values are compared to diagnose power semiconductor devices. As a possible implementation, when at least one of the schemes of the artificial intelligence model and the scheme of the upper and lower limit values diagnoses the failure of the power semiconductor device, it is judged that the power semiconductor device is failure, that is, a logical OR operation is performed on the two diagnosis results. In this way, the test system for power semiconductor devices provided by the embodiments of the present application can combine traditional test methods with artificial intelligence models to further improve the accuracy of detection.

For example, the data of the power semiconductor device X does not exceed the preset upper and lower limit data, but after the data of the power semiconductor device X is input into the artificial intelligence model, the diagnosis result output by the artificial intelligence model is invalid, and at this time it is judged that the power semiconductor device X is invalid . The data of the power semiconductor device Y exceeds the preset upper and lower limit data, but after the data of the power semiconductor device Y is input into the artificial intelligence model, the diagnostic result output by the artificial intelligence model is not invalid, and at this time, the power semiconductor device Y is judged to be invalid. The data of the power semiconductor device Z exceeds the preset upper and lower limit data, and after the data of the power semiconductor device Z is input into the artificial intelligence model, the diagnosis result output by the artificial intelligence model is invalid, and at this time, the power semiconductor device Z is judged to be invalid.

In the power semiconductor device testing system provided in the embodiment of the present application, in order to make the diagnosis of the artificial intelligence model more accurate, as a possible implementation, the testing machine is also used to send update data to the cloud server; the cloud server is also used to update the artificial intelligence model based on updated data.

It should be noted that in the embodiment of the present application, the step of inputting the data of the power semiconductor device to be tested into the pre-trained artificial intelligence model and obtaining the diagnosis result can be completed by the cloud server in the cloud, or by the test machine on the board end. The embodiments of the present application are not limited here. The following will specifically introduce two different implementation manners through embodiments.

Referring to FIG. 4 , this figure is a schematic diagram of another test system for a power semiconductor device provided by an embodiment of the present application.

The testing machine 200 is also used to collect historical data of multiple power semiconductor devices, and send the collected historical data of multiple power semiconductor devices to the cloud server 100, and the cloud server 100 can train artificial intelligence through the historical data of multiple power semiconductor devices Model.

The cloud server 100 is further configured to use historical data of a plurality of power semiconductor devices to perform training using at least one of a supervised learning model or an unsupervised learning model to obtain an artificial intelligence model.

In the embodiment of the present application, the historical data may include at least one of chip test data of multiple power semiconductor devices, function test data of packaged modules, or relevant data when the actual application is abnormal. Specifically, as a possible implementation manner, the historical data may include one or more items of the training data in FIG. 2 .

It should be noted that the artificial intelligence model obtained through training in the embodiment of the present application is stored in the cloud server. When it is necessary to use the artificial intelligence model to detect the power semiconductor device to be tested, the cloud server uses the artificial intelligence model to detect the power semiconductor device to be tested. The power semiconductor device is tested and the result of the test is obtained.

The following will specifically introduce the specific solution of the cloud server using the artificial intelligence model to detect the power semiconductor device to be tested in this example with reference to the accompanying drawings.

Referring to FIG. 5 , this figure is a schematic diagram of another test system for a power semiconductor device provided by an embodiment of the present application.

The cloud server 100 is used to input the data of the power semiconductor device to be tested obtained from the testing machine into the pre-trained artificial intelligence model, and the output of the pre-trained artificial intelligence model is the diagnosis result of the power semiconductor device to be tested. Then, the cloud server 100 sends the diagnosis result to the testing machine 200, and finally the testing machine 200 outputs the diagnosis result.

It can be understood that, in the solution provided by the embodiment of the present application, the cloud server 100 will store its trained artificial intelligence model. When it is necessary to use the artificial intelligence model to detect the power semiconductor device to be tested, the cloud server 100 can obtain the data of the power semiconductor device to be tested from the testing machine 200, and input these data into the pre-trained artificial intelligence model to obtain the diagnosis result . Then, the cloud server 100 sends the obtained diagnosis result to the testing machine 200, so that the testing machine 200 on the end board outputs the diagnosis result, that is, the testing machine 200 only receives the diagnosis result, and does not test the power semiconductor device itself.

As another possible implementation, the step of inputting the data of the power semiconductor to be tested into the pre-selected trained artificial intelligence model and obtaining the diagnosis result in the embodiment of the present application can also be completed by the tester on the board side. The following describes the training process of the AI model.

Referring to FIG. 6 , this figure is a schematic diagram of another test system for a power semiconductor device provided by an embodiment of the present application.

The testing machine 200 is also used to collect historical data of a plurality of power semiconductor devices, and send the collected historical data of a plurality of power semiconductor devices to the cloud server 100, so that the cloud server 100 can be trained through the historical data of a plurality of power semiconductor devices artificial intelligence model. The historical data includes at least one of the chip test data or packaged module function test data of the plurality of power semiconductor devices in the above embodiments. It should be noted that, after training the artificial intelligence model obtained in the embodiment of the present application, the cloud server 100 sends the obtained artificial intelligence model to the testing machine 200 on the board end, and the testing machine 200 stores the artificial intelligence model.

The following will specifically introduce the specific solution of the testing machine using the artificial intelligence model to detect the power semiconductor device to be tested in this example with reference to the accompanying drawings.

Referring to FIG. 7 , this figure is a schematic diagram of another test system for a power semiconductor device provided by an embodiment of the present application.

The cloud server 100 is used to send the pre-trained artificial intelligence model to the testing machine 200 . The testing machine 200 is used to input the data of the power semiconductor device to be tested into the pre-trained artificial intelligence model, and the output of the pre-trained artificial intelligence model is the diagnosis result of the power semiconductor device to be tested.

It can be understood that, in the solution provided by the embodiment of the present application, the testing machine 200 will receive the artificial intelligence model sent by the cloud server 100 and store the artificial intelligence model. When the artificial intelligence model needs to be used to detect the power semiconductor device to be tested, the testing machine 200 inputs the data of the power semiconductor device to be tested into the artificial intelligence model, and obtains and outputs a diagnosis result.

It should be noted that, in the embodiment of the present application, since the artificial intelligence model is stored in the testing machine, the testing machine can directly use the artificial intelligence model to complete the task of detecting the power semiconductor device to be tested on a single machine without communicating with the cloud server in the cloud. Therefore, it can be avoided that when the network between the testing machine at the board end and the cloud server in the cloud is interrupted, the power semiconductor device testing system in the embodiment of the present application cannot obtain the diagnosis result of the power semiconductor device to be tested. In this way, the power semiconductor device test system provided in the embodiment of the present application can reduce the network dependence between the cloud server and the test machine, and improve the real-time performance of the diagnostic results obtained by the power semiconductor device test system in the embodiment of the present application.

To sum up, the test system provided by the embodiment of the present application can use the cloud server in the cloud to input the data of the power semiconductor to be tested into the pre-selected training artificial intelligence model and obtain the diagnosis result, or the test machine on the board side can input the data of the power semiconductor to be tested Data from power semiconductors is fed into preselected trained artificial intelligence models and diagnostic results are obtained. When the diagnosis result is obtained from the cloud server, the cloud server can directly store the trained artificial intelligence model without sending it to the test machine on the board. When the diagnostic result is obtained from the tester on the board, the cloud server needs to send the pre-trained artificial intelligence model to the tester, and the tester can complete the test of the power semiconductor device to be tested on a single machine. When the test machine obtains the diagnosis result, the cloud server does not need to send the diagnosis result to the test machine. Therefore, the diagnosis result obtained by the test machine is not affected by network faults and has low dependence on the network.

The testing system provided in the embodiment of the present application may include one testing machine, or may include multiple testing machines. The working principle of the testing system provided by the embodiment of the present application including multiple testing machines will be described below with reference to the accompanying drawings.

Referring to FIG. 8 , this figure is a schematic diagram of a testing system for a power semiconductor device including multiple testing machines provided in an embodiment of the present application.

The power semiconductor device testing system provided in the embodiment of the present application includes at least the following two testing machines: a first testing machine 201 and a second testing machine 202;

The cloud server 100 is specifically used to perform training according to the first historical data sent by the first testing machine 201 and the second historical data sent by the second testing machine 202 to obtain a global artificial intelligence model, and use the global artificial intelligence intelligent model to train the first testing machine The power semiconductor device to be tested corresponding to 201 is tested to obtain a first diagnosis result, and the global artificial intelligence model is used to test the power semiconductor device to be tested corresponding to the second testing machine 202 to obtain a second diagnosis result, and the first diagnosis result is sent to the second diagnosis result. A testing machine 201 sends the second diagnosis result to the second testing machine 202 .

It can be understood that the cloud server provided by the embodiment of the present application uses the first historical data and the second historical data to train the global artificial intelligence model, and can make full use of all the historical data, thereby improving the generalization ability of the global artificial intelligence model and using accurate The more accurate global artificial intelligence model can diagnose the power semiconductor device to be tested, which can improve the universality of the artificial intelligence model and increase the accuracy of new device diagnosis.

The cloud server provided by the embodiment of the present application can not only directly send the diagnosis result to the test machine at the board end, but also can send the trained global artificial intelligence model to the test machine to complete the test by each test machine, which is not limited in the embodiment of the present application. .

Referring to FIG. 9 , this figure is a schematic diagram of another testing system for a power semiconductor device including multiple testing machines provided by an embodiment of the present application.

The power semiconductor device testing system provided in the embodiment of the present application includes at least the following two testing machines: a first testing machine 201 and a second testing machine 202 .

Wherein, the cloud server 100 is specifically used to perform training according to the first historical data sent by the first testing machine 201 and the second historical data sent by the second testing machine 202 to obtain a global artificial intelligence model, and send the global artificial intelligence model to the first A testing machine 201 and a second testing machine 202 .

The first testing machine 201 is used to test the corresponding power semiconductor device to be tested by using the global artificial intelligence model; the second testing machine 202 is used to test the corresponding power semiconductor device to be tested by using the global artificial intelligence model.

In the testing system provided by the embodiment of the present application, since the cloud server directly sends the global artificial intelligence model, the testing machine can directly use the stored global artificial intelligence model to complete the task of detecting the power semiconductor device to be tested on a single machine, and does not need to communicate with the cloud server in the cloud. to communicate. It is avoided that when the network is interrupted between the testing machine at the board end and the cloud server in the cloud, the power semiconductor device testing system in the embodiment of the present application cannot obtain the diagnosis result of the power semiconductor device to be tested. In this way, the power semiconductor device test system provided in the embodiment of the present application can reduce the dependence on network stability between the cloud server and the test machine, and improve the real-time diagnosis results obtained by the power semiconductor device test system in the embodiment of the present application. sex.

It can be seen from the above that the power semiconductor device testing system provided by the embodiment of the present application can obtain a global artificial intelligence model through all historical data, so as to obtain a detection result by using the global artificial intelligence model. However, considering that the testing machine may have certain systematic errors, the systematic errors in the data measured by different testing machines are different, and this error may be related to the testing machine itself. Therefore, the embodiment of the present application also provides a test system for power semiconductor devices. After obtaining the global artificial intelligence model, the system fine-tunes the global artificial intelligence model according to the historical data obtained by different testing machines to obtain different The local artificial intelligence model corresponding to the test machine.

Referring to FIG. 10 , this figure is a schematic diagram of another testing system for a power semiconductor device including multiple testing machines provided by an embodiment of the present application.

As shown in FIG. 10 , the power semiconductor device testing system provided by the present application includes at least the following two testing machines: a first testing machine 201 and a second testing machine 202 .

Among them, the cloud server 100 is specifically used to perform training according to the first historical data sent by the first testing machine 201 and the second historical data sent by the second testing machine 202 to obtain a global artificial intelligence model, and use the first historical data to train the global artificial intelligence model. Adjust the model to obtain the first artificial intelligence model, use the second historical data to adjust the global artificial intelligence model to obtain the second artificial intelligence model; use the first artificial intelligence model to test the power semiconductor device to be tested corresponding to the first testing machine 201 Obtain the first diagnosis result, use the second artificial intelligence model to test the power semiconductor device to be tested corresponding to the second tester 202 to obtain the second diagnosis result, send the first diagnosis result to the first tester 201, and send the second diagnosis The results are sent to the second testing machine 202 .

It should be noted that, in the embodiment of the present application, the first historical data sent by the first testing machine 201 is historical data corresponding to the first testing machine 201 . That is, the first historical data includes data measured directly or indirectly by the first testing machine 201 . The second historical data sent by the first testing machine 202 is historical data corresponding to the second testing machine 202 .

In the embodiment of the present application, the first historical data and the second historical data are used to train the global artificial intelligence model, which can make full use of all historical data and improve the test accuracy of the global artificial intelligence model. On this basis, considering that there may be differences in test performance between different testing machines, the implementation of this application also uses the first historical data to adjust the global artificial intelligence model to obtain the first artificial intelligence model, and uses the second historical data to adjust the global artificial intelligence model. The artificial intelligence model is adjusted to obtain the second artificial intelligence model, so that the partial artificial intelligence model (the first artificial intelligence model) for the characteristics of the first test machine and the partial artificial intelligence model (the second artificial intelligence model) for the characteristics of the second test machine can be obtained. model), which further improves the test accuracy of the artificial intelligence model used in actual detection.

For example, the testing system for power semiconductor devices provided in the embodiment of the present application includes: testing machine Q and testing machine W, testing machine Q sends historical data of testing machine Q to the cloud server, testing machine W sends historical data of testing machine W to Cloud Server. The cloud server uses the historical data of the test machine Q and the historical data of the test machine W to train the global artificial intelligence model. Then, the cloud server uses the historical data of the testing machine Q to adjust the global artificial intelligence model, obtains the local artificial intelligence model corresponding to the testing machine Q, and uses the local artificial intelligence model to obtain the power semiconductor device to be tested corresponding to the testing machine Q. test. The cloud server uses the historical data of the testing machine W to adjust the global artificial intelligence model, obtains the local artificial intelligence model corresponding to the testing machine W, and uses the local artificial intelligence model to obtain the power semiconductor device to be tested corresponding to the testing machine W for testing.

The cloud server provided by the embodiment of the present application can not only directly send the diagnosis result to the test machine on the board end, but also send a trained artificial intelligence model to the test machine on the board end, and the test machine on the board end uses the trained artificial intelligence model to analyze the The device is tested and the diagnosis result is obtained, which will be described in detail below with reference to the accompanying drawings.

Referring to FIG. 11 , this figure is a schematic diagram of another testing system for a power semiconductor device including multiple testing machines provided by an embodiment of the present application.

The testing system for power semiconductor devices provided by the application includes at least the following two testing machines: a first testing machine 201 and a second testing machine 202;

The cloud server 100 is specifically used to perform training according to the first historical data sent by the first testing machine 201 and the second historical data sent by the second testing machine 202 to obtain a global artificial intelligence model, and use the first historical data to perform training on the global artificial intelligence model. The adjusted first artificial intelligence model is sent to the first testing machine 201 , and the second historical data is used to adjust the global artificial intelligence model to obtain the second artificial intelligence model and sent to the second testing machine 202 .

The first testing machine 201 is specifically used to test the corresponding power semiconductor device to be tested by using the first artificial intelligence model; the second testing machine 202 is specifically used to test the corresponding power semiconductor device to be tested by using the second artificial intelligence model test.

In the embodiment of the present application, since the cloud server directly sends the first artificial intelligence model and the second artificial intelligence model to the first testing machine and the second testing machine respectively, the testing machine can directly use the corresponding artificial intelligence model, and a single machine completes the detection. The task of testing power semiconductor devices does not need to communicate with the cloud server in the cloud, which prevents the power semiconductor device test system in the embodiment of the present application from being unable to be tested when the network is interrupted between the test machine at the board end and the cloud server in the cloud. Diagnostic results of power semiconductor devices. In this way, the power semiconductor device test system provided in the embodiment of the present application can reduce the dependence on network stability between the cloud server and the test machine, and improve the real-time diagnosis results obtained by the power semiconductor device test system in the embodiment of the present application. sex.

It can be seen from the above that the power semiconductor device testing system provided by the embodiment of the present application can directly use all historical data to obtain the global artificial intelligence model and use it directly, or use different testing machines to correspond to the global artificial intelligence model after obtaining the global artificial intelligence model. According to the historical data, adjust the global artificial intelligence model to obtain the local artificial intelligence model, and use the partial artificial intelligence model to detect the power semiconductor device of the corresponding testing machine.

In order to obtain a higher-precision artificial intelligence model in the embodiment of the present application, as a possible implementation mode, a follow-up test can also be performed on the power semiconductor device diagnosed as failure to verify whether the power semiconductor device is invalid, and then verify the AI model Whether it is accurate, if it is verified that the power semiconductor device does not fail, it means that the AI model is not accurate, and the AI model needs to be adjusted. For power semiconductor devices that are judged not to have failed, continue to collect their operating data on the application side, and then upload the data of the above two power semiconductor devices to the cloud server, and the cloud server uses these two data to adjust the AI model.

In order to make the accuracy of the artificial intelligence model in the embodiment of the present application higher, as a possible implementation, the testing machine in the embodiment of the present application is also used to fine-tune the pre-trained model in combination with data from the testing machine side , using the fine-tuned model to diagnose the power semiconductor device to be tested, and output a diagnosis result. It should be noted that the fine-tuning in this embodiment of the present application may include optimizing parameters of a pre-trained model. As an example, transfer learning can be used to fine-tune a pre-trained AI model. For example, the fine-tuned AI model can be used to test whether the wafer fails.

Considering that the testing machine in the embodiment of the present application stores a small amount of historical data collected by it, but these data have not been uploaded to the cloud server, as a possible implementation, the data on the testing machine side in the embodiment of the application is also Data not uploaded to the cloud server may be included. It can be understood that the test machine in the embodiment of the present application usually uploads the collected data to the cloud server regularly, and the test machine in the embodiment of the present application usually stores a certain amount of historical data that has not been uploaded. Therefore, the test machine can use these The AI model is fine-tuned on the unuploaded data, and then the power semiconductor device is tested based on the fine-tuned AI model. After the testing machine in the embodiment of the present application fine-tunes the AI model, it can also upload the fine-tuned AI model to the cloud server, so that the cloud server can uniformly manage the AI model.

In summary, the test system for power semiconductor devices provided by the embodiment of the present application, considering the limited computing power and storage capacity of the testing machine, uses a cloud server with relatively large computing power and storage capacity to train the artificial intelligence model. The artificial intelligence model can be used in the cloud server to directly obtain the diagnosis result, or the artificial intelligence model can be sent to the test machine on the board, and the test machine can use the artificial intelligence model to obtain the diagnosis result, or only the artificial intelligence model that has not been fully trained can be sent To the testing machine on the board side, the testing machine on the board side completes the training of the artificial intelligence model according to the stored data, and uses the trained artificial intelligence model to obtain the diagnosis result.

server embodiment

According to the power semiconductor device testing system provided in the foregoing embodiments, an embodiment of the present application further provides a cloud server.

Referring to FIG. 12 , this figure is a schematic structural diagram of a cloud server provided by an embodiment of the present application.

As shown in FIG. 12 , the cloud server provided by the embodiment of the present application includes: a first transceiver device 1201 and a first controller 1202 .

Wherein, the first transceiver device 1201 is configured to receive the test data of the power semiconductor device to be tested sent by the testing machine.

The first controller 1202 is used to pre-train the historical data of multiple power semiconductor devices to obtain the artificial intelligence model; it is also used to input the data of the power semiconductor devices to be tested into the artificial intelligence model, and the output of the pre-trained AI model is The diagnosis result of the power semiconductor device to be tested; the first transceiver device is also used to send the diagnosis result to the testing machine;

Or, the first controller 1202 is configured to send the pre-trained artificial intelligence model to the testing machine, so that the testing machine uses the pre-trained artificial intelligence model to diagnose the power semiconductor device to be tested.

For the specific implementation manner of the cloud server provided by the embodiment of the present application, reference may be made to the introduction of the cloud server in the above test system embodiment, and only a brief introduction is given here.

In a possible implementation of the embodiment of the present application, the first controller is specifically used to obtain the failure types of multiple power semiconductor devices when using the unsupervised learning model training to obtain the artificial intelligence model, and uses expert knowledge in the power semiconductor All test items corresponding to each failure type are extracted from the test items of the device, and all test items corresponding to each failure type form a subset of test items. For each subset of test items, an unsupervised learning model is used to detect the abnormal level, and the abnormal The first score and the second score are respectively recorded for the device and the normal device, and the sum of the first score and the second score of all test item subsets of each power semiconductor device is obtained as the total score of the abnormal level; the total score of the abnormal level is greater than the preset score threshold, it is judged that the power semiconductor device is an abnormal device.

In a possible implementation manner of the embodiment of the present application, the cloud server corresponds to at least two testing machines as follows: a first testing machine and a second testing machine. Wherein, the first controller is specifically configured to perform training according to the first historical data sent by the first testing machine and the second historical data sent by the second testing machine to obtain the global artificial intelligence model, and use the first historical data to train the global artificial intelligence model. Adjust to obtain the first artificial intelligence model, use the second historical data to adjust the global artificial intelligence model to obtain the second artificial intelligence model; use the first artificial intelligence model to test the power semiconductor device to be tested corresponding to the first testing machine to obtain the second artificial intelligence model A diagnostic result, using the second artificial intelligence model to test the power semiconductor device to be tested corresponding to the second testing machine to obtain a second diagnostic result; the first transceiver device is specifically used to send the first diagnostic result to the first testing machine, It is also used to send the second diagnosis result to the second testing machine.

In a possible implementation manner of the embodiment of the present application, the cloud server corresponds to at least two testing machines as follows: a first testing machine and a second testing machine. Wherein, the first controller is specifically configured to perform training according to the first historical data sent by the first testing machine and the second historical data sent by the second testing machine to obtain a global artificial intelligence model, and use the global artificial intelligence intelligent model to test the first The power semiconductor device to be tested corresponding to the tester is tested to obtain the first diagnosis result, and the global artificial intelligence model is used to test the power semiconductor device to be tested corresponding to the second tester to obtain the second diagnosis result; the first transceiver device is specifically used to The first diagnosis result is sent to the first testing machine, and is also used to send the second diagnosis result to the second testing machine.

In a possible implementation manner of the embodiment of the present application, the first transceiver device is further configured to receive update data sent by the testing machine; the first controller is further configured to update the artificial intelligence model according to the update data.

The cloud server provided by the embodiment of the present application may also include a first storage device (not shown in the figure) in addition to the first transceiver device and the first controller, and the first storage device may be used to store the history uploaded by the testing machine The data and the updated data are convenient for the first controller to perform subsequent processing. Before the first controller uses the historical data to train the AI model, in order to make the parameters of the trained AI model more accurate, some interference data may be eliminated, specifically data cleaning may be performed.

To sum up, the cloud server in the embodiment of the present application can use its powerful computing power and storage capacity to train the artificial intelligence model, so as to obtain an artificial intelligence model with high diagnostic accuracy. After the artificial intelligence model is obtained, the artificial intelligence model can be directly used in the cloud server to obtain the diagnosis result, or only the artificial intelligence model can be provided for the testing machine without direct diagnosis result.

Test machine embodiment

Based on the power semiconductor device testing system and cloud server provided in the above embodiments, the embodiment of the present application also provides a testing machine, which will be described in detail below with reference to the accompanying drawings.

Referring to FIG. 13 , this figure is a schematic structural diagram of a testing machine provided in an embodiment of the present application.

As shown in Figure 13, the testing machine in the embodiment of the present application includes: a second transceiver device 1301 and a second controller 1302;

The second transceiver device 1301 is used to receive the artificial intelligence model sent by the cloud server, and the artificial intelligence model is obtained by the cloud server in advance using the historical data of multiple power semiconductor devices for training;

The second controller 1302 is used to obtain the data of the power semiconductor device to be tested; input the data of the power semiconductor device to be tested into the pre-trained artificial intelligence model, and the output of the pre-trained AI model is the diagnosis result of the power semiconductor device to be tested.

The test machine provided by the embodiment of the present application generally refers to any test machine. One cloud server can correspond to multiple test machines provided by the embodiment of the present application. For the specific working mode of the test machine, please refer to the introduction of the test system embodiment above for the test machine. Only a brief description is given here.

In a possible implementation manner of the embodiment of the present application, the second controller is also used to collect historical data of multiple power semiconductor devices, and send the collected historical data of multiple power semiconductor devices to the cloud server; the historical data includes multiple At least one of the chip test data of a power semiconductor device or the packaged module function test data.

In a possible implementation manner of the embodiment of the present application, the second controller is also used to compare the data of the power semiconductor device to be tested with the preset upper and lower limit values, and when the data of the power semiconductor device to be tested exceeds the upper and lower limit values , judging that the power semiconductor device to be tested is an abnormal device. And in this application, the testing machine can directly obtain the diagnosis result sent from the cloud server, or can obtain the artificial intelligence model sent by the cloud server, and obtain the diagnosis result through the artificial intelligence model in the testing machine.

In summary, the testing machine provided by the embodiment of the present application has limited computing power and storage capacity, so it sends the data to the cloud server, and uses the cloud server with powerful computing power and storage capacity to train the artificial intelligence model. In this way, an artificial intelligence model with high diagnostic accuracy can be obtained.

It should be understood that in this application, "at least one (item)" means one or more, and "multiple" means two or more. "And/or" is used to describe the association relationship of associated objects, indicating that there can be three types of relationships, for example, "A and/or B" can mean: only A exists, only B exists, and A and B exist at the same time , where A and B can be singular or plural. The character "/" generally indicates that the contextual objects are an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one item (piece) of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c ", where a, b, c can be single or multiple.

As mentioned above, the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: it can still understand the foregoing The technical solutions described in each embodiment are modified, or some of the technical features are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the application.

Claims (22)

1. A test system for a power semiconductor device, comprising: a cloud server and at least one test machine; The cloud server is used to pre-train the historical data of multiple power semiconductor devices to obtain an artificial intelligence model; The testing machine is used to obtain the data of the power semiconductor device to be tested; The cloud server or the testing machine is used to input the data of the power semiconductor device to be tested into the artificial intelligence model, and the output of the artificial intelligence model is a diagnosis result of the power semiconductor device to be tested. 2. The system according to claim 1, wherein the cloud server is configured to input the data of the power semiconductor device to be tested into the artificial intelligence model, and the output of the artificial intelligence model is the A diagnostic result of the power semiconductor device is tested, and the diagnostic result is sent to the testing machine. 3. The system according to claim 1, wherein the cloud server is configured to send the pre-trained artificial intelligence model to the testing machine; The testing machine is configured to input the data of the power semiconductor device to be tested into the artificial intelligence model, and the output of the artificial intelligence model is a diagnosis result of the power semiconductor device to be tested. 4. The system according to any one of claims 1-3, wherein the testing machine is also used to collect historical data of the multiple power semiconductor devices, and the collected multiple power semiconductor devices The historical data of the device is sent to the cloud server; the historical data includes at least one of the chip test data of the plurality of power semiconductor devices, the packaged module function test data, or the relevant data when the actual application is abnormal; The cloud server is further configured to use historical data of the plurality of power semiconductor devices to perform training using at least one of a supervised learning model or an unsupervised learning model to obtain the artificial intelligence model. 5. The system according to any one of claims 1-4, wherein the cloud server uses the unsupervised learning model training to obtain the artificial intelligence model, which is specifically used to obtain the plurality of power semiconductor devices failure types, use expert knowledge to extract all test items corresponding to each failure type from the test items of power semiconductor devices, all test items corresponding to each failure type form a subset of test items, and use no test items for each subset of test items The supervised learning model differentiates the abnormal subset and the normal subset, obtains the first score and the second score respectively for the abnormal subset and the normal subset, and obtains the described The sum of the first score and the second score is used as the total score of the abnormal level; when the total score of the abnormal level is greater than or equal to the preset score threshold, it is judged that the power semiconductor device is an abnormal device. 6 . The system according to claim 5 , wherein the cloud server is specifically configured to use expert knowledge to record different first scores for abnormal devices in different test item subsets. 7. The system according to any one of claims 1-4, wherein the cloud server uses the supervised learning model training to obtain the artificial intelligence model, which is specifically used to obtain the power semiconductor device when it is abnormal The relevant data of the data is used as a data label, and the data feature of the data label is extracted by using a supervised learning model, and the abnormal level total score of each power semiconductor device is obtained by using the data feature, and when the abnormal level total score is greater than the preset score threshold , it is judged that the power semiconductor device is an abnormal device. 8. The system according to any one of claims 1-7, wherein the testing machine is also used to compare the data of the power semiconductor device to be tested with preset upper and lower limit values, the When the data of the power semiconductor device to be tested exceeds the upper and lower limit values, it is determined that the power semiconductor device to be tested is an abnormal device. 9. The system according to claim 2, wherein the system comprises at least the following two testing machines: a first testing machine and a second testing machine; The cloud server is specifically configured to perform training according to the first historical data sent by the first testing machine and the second historical data sent by the second testing machine to obtain a global artificial intelligence model, and use the first historical data to train the Adjusting the global artificial intelligence model to obtain a first artificial intelligence model, using the second historical data to adjust the global artificial intelligence model to obtain a second artificial intelligence model; using the first artificial intelligence model to test the first artificial intelligence model Test the power semiconductor device to be tested corresponding to the testing machine to obtain a first diagnosis result, use the second artificial intelligence model to test the power semiconductor device to be tested corresponding to the second testing machine to obtain a second diagnosis result, and use the second artificial intelligence model to test the power semiconductor device to be tested corresponding to the second testing machine. A diagnosis result is sent to the first testing machine, and the second diagnosis result is sent to the second testing machine. 10. The system according to claim 2, characterized in that the system comprises at least the following two testing machines: a first testing machine and a second testing machine; The cloud server is specifically configured to perform training according to the first historical data sent by the first testing machine and the second historical data sent by the second testing machine to obtain a global artificial intelligence model, and use the global artificial intelligence intelligent model to The power semiconductor device to be tested corresponding to the first testing machine is tested to obtain a first diagnosis result, and the global artificial intelligence model is used to test the power semiconductor device to be tested corresponding to the second testing machine to obtain a second diagnosis result, and the The first diagnosis result is sent to the first testing machine, and the second diagnosis result is sent to the second testing machine. 11. The system according to claim 3, wherein the system comprises at least the following two testing machines: a first testing machine and a second testing machine; The cloud server is specifically configured to perform training according to the first historical data sent by the first testing machine and the second historical data sent by the second testing machine to obtain a global artificial intelligence model, and use the first historical data to train the Adjust the global artificial intelligence model to obtain the first artificial intelligence model and send it to the first tester, and use the second historical data to adjust the global artificial intelligence model to obtain the second artificial intelligence model and send it to the second tester machine; The first testing machine is specifically configured to use the first artificial intelligence model to diagnose the corresponding power semiconductor device to be tested; The second testing machine is specifically configured to use the second artificial intelligence model to diagnose the corresponding power semiconductor device to be tested. 12. The system according to claim 3, wherein the system comprises at least the following two testing machines: a first testing machine and a second testing machine; The cloud server is specifically configured to perform training according to the first historical data sent by the first testing machine and the second historical data sent by the second testing machine to obtain a global artificial intelligence model, and send the global artificial intelligence model to the said first testing machine and said second testing machine; The first testing machine is configured to use the global artificial intelligence intelligent model to diagnose the corresponding power semiconductor device to be tested; The second testing machine is configured to use the global artificial intelligence model to diagnose the corresponding power semiconductor device to be tested. 13. The system according to any one of claims 1-12, wherein the testing machine is also used to send update data to the cloud server; The cloud server is further configured to update the artificial intelligence model according to the update data. 14. The system according to any one of claims 3-13, wherein the testing machine is also used to fine-tune the pre-trained AI model in combination with data from the testing machine side, and use the fine-tuned model to The power semiconductor device to be tested is diagnosed, and a diagnosis result is output. 15. A cloud server, comprising: a first transceiver device and a first controller; The first transceiver device is used to receive the test data of the power semiconductor device to be tested sent by the testing machine; The first controller is configured to use the historical data of a plurality of power semiconductor devices for training in advance to obtain an artificial intelligence model; it is also configured to input the data of the power semiconductor devices to be tested into the artificial intelligence model, and the artificial The output of the intelligent model is the diagnosis result of the power semiconductor device to be tested; the first transceiver device is also used to send the diagnosis result to the testing machine; or, The first controller is configured to send the artificial intelligence model to the testing machine, so that the testing machine uses the artificial intelligence model to diagnose the power semiconductor device to be tested. 16. The cloud server according to claim 15, wherein the first controller is specifically used to obtain the artificial intelligence model by using unsupervised learning model training, and is specifically used to obtain the plurality of power semiconductor The failure type of the device, use expert knowledge to extract all the test items corresponding to each failure type in the test items of the power semiconductor device, all the test items corresponding to each failure type form a subset of test items, and use The unsupervised learning model detects the abnormal level, and the abnormal device and the normal device are recorded as the first score and the second score respectively, and the sum of the first score and the second score of all test item subsets of each power semiconductor device is obtained as the abnormal level Total score; when the total score of the abnormal level is greater than the preset score threshold, it is judged that the power semiconductor device is an abnormal device. 17. The cloud server according to claim 15 or 16, wherein the cloud server corresponds to at least two testing machines as follows: a first testing machine and a second testing machine; The first controller is specifically configured to perform training according to the first historical data sent by the first testing machine and the second historical data sent by the second testing machine to obtain a global artificial intelligence model, and use the first historical data to The global artificial intelligence model is adjusted to obtain a first artificial intelligence model, and the second historical data is used to adjust the global artificial intelligence model to obtain a second artificial intelligence model; A power semiconductor device to be tested corresponding to a testing machine is tested to obtain a first diagnosis result, and the second artificial intelligence model is used to test a power semiconductor device to be tested corresponding to the second testing machine to obtain a second diagnosis result; The first transceiver device is specifically configured to send the first diagnosis result to the first testing machine, and is also used to send the second diagnosis result to the second testing machine. 18. The cloud server according to claim 15 or 16, wherein the cloud server corresponds to the following at least two testing machines: a first testing machine and a second testing machine; The first controller is specifically configured to perform training according to the first historical data sent by the first testing machine and the second historical data sent by the second testing machine to obtain a global artificial intelligence model, and use the global artificial intelligence smart model Test the power semiconductor device to be tested corresponding to the first testing machine to obtain a first diagnosis result, and use the global artificial intelligence model to test the power semiconductor device to be tested corresponding to the second testing machine to obtain a second diagnosis result ; The first transceiver device is specifically configured to send the first diagnosis result to the first testing machine, and is also used to send the second diagnosis result to the second testing machine. 19. The cloud server according to any one of claims 14-17, wherein the first transceiver device is further configured to receive update data sent by the testing machine; The first controller is further configured to update the artificial intelligence model according to the update data. 20. A testing machine, comprising: a second transceiver device and a second controller; The second transceiver device is configured to receive the artificial intelligence model sent by the cloud server, and the artificial intelligence model is obtained by the cloud server using historical data of multiple power semiconductor devices for training in advance; The second controller is configured to obtain data of the power semiconductor device to be tested; input the data of the power semiconductor device to be tested into the artificial intelligence model, and the output of the artificial intelligence model is the power semiconductor device to be tested diagnosis results. 21. The testing machine according to claim 20, wherein the second controller is also used to collect historical data of the plurality of power semiconductor devices, and to collect the historical data of the plurality of power semiconductor devices The historical data is sent to the cloud server; the historical data includes at least one of chip test data or packaged module function test data of the plurality of power semiconductor devices. 22. The testing machine according to claim 20 or 21, wherein the second controller is further configured to compare the data of the power semiconductor device to be tested with preset upper and lower limit values, the When the data of the power semiconductor device to be tested exceeds the upper and lower limit values, it is determined that the power semiconductor device to be tested is an abnormal device.

Images