Intelligent automatic test system of IGBT device

Abstract

The invention relates to the technical field of IGBT (Insulated Gate Bipolar Translator) device performance testing, in particular to an intelligent automatic testing system for IGBT devices, which comprises a data acquisition module for acquiring various data of each IGBT device at each moment at the same temperature; the data analysis module is used for analyzing differences between all crest values in all short-circuit currents of each IGBT device and corresponding adjacent trough values, determining an oscillation difference value of each IGBT device, and determining an oscillation abnormal coefficient of each IGBT device by combining the crest values, distribution conditions of all short-circuit currents in neighborhoods of the crest values and the abrupt change offset coefficients; and the data test module is used for detecting the IGBT device. According to the invention, by analyzing the change trend characteristics of the short-circuit current of the IGBT device at different temperatures and the consistency of the oscillation characteristics of the short-circuit current and the temperature change, the accuracy of the performance test of the IGBT device is improved.

Description

Technical Field

[0001] The invention relates to the technical field of IGBT device performance testing, and in particular to an intelligent automatic testing system for IGBT devices. Background Art

[0002] IGBT devices have the advantages of high blocking voltage, low conduction voltage, and high current density, and are widely used in rail transportation, new energy power generation, electric vehicles, etc. In order to improve the efficiency of the converter system and reduce production and maintenance costs, such IGBT devices generally adopt a multi-chip integrated package power module structure. This module has become the main power electronic switching device in the converter system. Fully exerting the performance of IGBT and ensuring the reliable operation of IGBT are the prerequisites for the normal operation of the converter system.

[0003] In the production process, the performance of IGBT devices needs to be automatically tested. The existing methods mainly collect electrical performance parameters such as saturation voltage and short-circuit current of IGBT devices, and compare whether the electrical performance parameters obtained by the test meet the design specifications to judge the quality of IGBT devices. However, the performance of IGBT devices is affected by the change of junction temperature. Under different junction temperature conditions, the short-circuit current will show different degrees of change characteristics. Therefore, the existing methods have the defect of insufficient accuracy of performance test results. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide an intelligent automatic testing system for IGBT devices, and the technical solutions adopted are as follows: The present invention provides an intelligent automatic testing system for IGBT devices, the system comprising: The performance data acquisition module is used to obtain various types of data of each IGBT device at different temperatures at different times, where the short-circuit current is one type of data among all types of data, and all types of data of each IGBT device at different temperatures at different times form a triplet; A performance data analysis module is used to cluster all triplets of each IGBT device at each temperature to obtain two clusters, and determine the mutation difference coefficient of each IGBT device at each temperature based on the difference of all triplets between the two clusters in each IGBT device; Based on the difference between various data of each IGBT device at all times and the corresponding similar data of other IGBT devices at all times, combined with the mutation difference coefficient, determine the mutation offset coefficient of each IGBT device at each temperature; Based on the differences between all peak values ​​and their respective adjacent trough values ​​in the fitted curves of all short-circuit currents of each IGBT device, the oscillation difference value of each IGBT device at each temperature is determined, and combined with the distribution of all short-circuit currents in each peak value and its neighborhood, as well as the mutation offset coefficient, the oscillation abnormality coefficient of each IGBT device at each temperature is determined; The performance data test module is used to record the average value of all short-circuit currents of each IGBT device at each temperature as the average short-circuit current of each IGBT device at each temperature; Based on the changing trend and distribution of the average short-circuit current of each IGBT device at all temperatures, the trend significance of each IGBT device is determined, and combined with the correlation between the oscillation abnormality coefficient of each IGBT device at all temperatures and the oscillation abnormality coefficient of other IGBT devices at all temperatures, the quality assessment coefficient of each IGBT device is determined, and the IGBT devices are tested.

[0005] Preferably, the mutation difference coefficient of each IGBT device at each temperature is the average of the differences between all triplets in one cluster of each IGBT device and all triplets in another cluster.

[0006] Preferably, the method for determining the sudden change offset coefficient of each IGBT device at each temperature is: Based on the differences between various data of each IGBT device at all times and the corresponding similar data of other IGBT devices at all times, the comprehensive difference coefficient of each IGBT device at each temperature is determined; The expression of the sudden shift coefficient of each IGBT device at each temperature is: ; In the formula, represents the sudden shift coefficient of the ith IGBT device at temperature k; represents the mutation difference coefficient of the ith IGBT device at temperature k; Represents the comprehensive difference coefficient of the i-th IGBT device at temperature k.

[0007] Preferably, the method for determining the comprehensive difference coefficient of each IGBT device at each temperature is: The mean values ​​of the differences between different IGBT devices for each type of data are calculated, and the average value of the mean values ​​for all types of data is taken as the comprehensive difference coefficient.

[0008] Preferably, the method for determining the oscillation difference value of each IGBT device at each temperature is: Fitting all short-circuit currents of each IGBT device at each temperature to obtain a current fitting curve, analyzing the average of the differences between each peak value and its adjacent valley value on the current fitting curve of each IGBT device at each temperature, and recording it as the first average of each peak value on the current fitting curve of each IGBT device at each temperature; The oscillation difference value of each IGBT device at each temperature is the average of the first average values ​​of all wave peak values ​​on the current fitting curve of each IGBT device at each temperature.

[0009] Preferably, the method for determining the oscillation abnormality coefficient of each IGBT device at each temperature is: Determine the kurtosis of each peak value and all fitting values ​​in its neighborhood on the current fitting curve of each IGBT device at each temperature; Determine the cumulative sum of the peaks of all the peaks of each IGBT device at each temperature, and record it as the peak steepness index of each IGBT device; The expression of the oscillation anomaly coefficient of each IGBT device at each temperature is: ; In the formula, represents the oscillation anomaly coefficient of the ith IGBT device at temperature k; represents the sudden shift coefficient of the ith IGBT device at temperature k; It represents the peak steepness index of the ith IGBT device at temperature k; Represents the oscillation difference value of the i-th IGBT device at temperature k.

[0010] Preferably, the method for determining the trend significance of each IGBT device is: The average short-circuit current of each IGBT device at all temperatures is used as the input of the trend test algorithm, and the absolute value of the output trend test result is used as the trend significance feature of each IGBT device; The trend significance of each IGBT device is the product of the mean of the average short-circuit current of each IGBT device at all temperatures and the trend significance feature.

[0011] Preferably, the method for determining the quality assessment coefficient of each IGBT device is: Determine the oscillation consistency of each IGBT device based on the correlation between the oscillation anomaly coefficient of each IGBT device at all temperatures and the oscillation anomaly coefficient of other IGBT devices at all temperatures; The expression of the quality assessment coefficient of each IGBT device is: ; In the formula, represents the quality assessment index of the i-th IGBT device; Indicates the trend significance of the i-th IGBT device; Represents the oscillation consistency of the i-th IGBT device.

[0012] Preferably, the method for determining the oscillation consistency of each IGBT device is: The oscillation abnormality coefficients of each IGBT device at all temperatures are combined into an oscillation abnormality sequence of each IGBT device; The oscillation consistency of each IGBT device is the average value of the correlation between the oscillation abnormality sequence of each IGBT device and the oscillation abnormality sequences of all other IGBT devices.

[0013] Preferably, the detecting of the IGBT device includes: If the normalized value of the quality assessment index of the IGBT device is less than a preset value, the corresponding IGBT device has a production fault; otherwise, the IGBT device is produced normally.

[0014] The present invention has the following beneficial effects: The present application determines the mutation offset coefficient of each IGBT device at each temperature by analyzing the differences between various types of data of each IGBT device at all times and the corresponding similar data of other IGBT devices at all times, combined with the mutation difference coefficient; based on the differences between all peak values ​​and their respective adjacent trough values ​​in the fitted curve of all short-circuit currents of each IGBT device, the oscillation difference value of each IGBT device at each temperature is determined, and combined with the distribution of all short-circuit currents in each peak value and its neighborhood, as well as the mutation offset coefficient, the oscillation abnormality coefficient of each IGBT device at each temperature is determined; the beneficial effect is that it can reflect the stability of the performance of the IGBT device.

[0015] This application determines the trend significance of each IGBT device based on the changing trend and distribution of the average short-circuit current of each IGBT device at all temperatures, and determines the quality assessment coefficient of each IGBT device based on the correlation between the oscillation abnormality coefficient of each IGBT device at all temperatures and the oscillation abnormality coefficient of other IGBT devices at all temperatures; the beneficial effect is that the quality and reliability of IGBT devices can be comprehensively evaluated.

[0016] This application determines the quality evaluation index of IGBT devices by analyzing the changing trend characteristics of IGBT short-circuit current at different temperatures, as well as the consistency of the oscillation characteristics of the short-circuit current and temperature changes, comprehensively evaluates the quality of IGBT devices, and improves the accuracy of IGBT device performance testing. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A block diagram of an intelligent automated testing system for an IGBT device provided in one embodiment of the present application; Figure 2 A schematic diagram of a quality assessment coefficient extraction process is provided for one embodiment of the present application; Figure 3 A schematic diagram of an automated testing process for an IGBT device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation, structure, features and effects of an intelligent automatic test system for IGBT devices proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0020] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0021] A specific solution of an intelligent automatic testing system for IGBT devices provided by the present invention is described in detail below with reference to the accompanying drawings.

[0022] See also Figure 1 , which shows a block diagram of an intelligent automatic testing system for an IGBT device provided by an embodiment of the present invention. The system includes: a performance data acquisition module 101, a performance data analysis module 102, and a performance data testing module 103.

[0023] The performance data acquisition module is used to obtain various types of data of each IGBT device at different temperatures at different times, among which the short-circuit current is one of all types of data, and all types of data of each IGBT device at each temperature at each time form a triplet.

[0024] Multiple IGBT devices to be tested are connected in parallel in a test circuit for testing as the same batch, and an oscilloscope is used to collect various data of each IGBT device at various times within a preset time period at different temperatures. The various data include three performance parameters: short-circuit current, saturation voltage, and gate current. The duration of the preset time period is set to t, and the sampling time interval is T.

[0025] It should be noted that the values ​​of the collection time t and the time interval T are set manually. In this embodiment, the value of the collection time t is 50 , the time interval T is 1 The implementer can also set it according to the specific situation, and this embodiment does not impose any special restrictions.

[0026] Furthermore, in order to obtain the performance of the IGBT under different temperature conditions, measurements are performed in different temperature environments, a temperature range is set during the test, and detection and testing are performed every preset interval within the temperature range.

[0027] It should be noted that the values ​​of the temperature range and the detection interval are set manually. In this embodiment, the temperature range is 25°C-145°C, and the preset interval value is 10°C. The implementer can also set it according to the specific situation. This embodiment does not impose any special restrictions.

[0028] The performance data analysis module is used to cluster all triplets of each IGBT device at each temperature to obtain two clustering clusters, and determine the mutation difference coefficient of each IGBT device at each temperature based on the difference between all triplets between the two clustering clusters in each IGBT device; determine the mutation offset coefficient of each IGBT device at each temperature based on the difference between various types of data at all times of each IGBT device and the corresponding similar data at all times of other IGBT devices, combined with the mutation difference coefficient; determine the oscillation difference value of each IGBT device at each temperature based on the difference between all peak values ​​and their respective adjacent trough values ​​in the fitted curve of all short-circuit currents of each IGBT device, and determine the oscillation abnormality coefficient of each IGBT device at each temperature based on the distribution of each peak value and all short-circuit currents in its neighborhood, as well as the mutation offset coefficient.

[0029] S1: Based on the differences between various data of each IGBT device at all times and the corresponding similar data of other IGBT devices at all times, combined with the mutation difference coefficient, determine the mutation offset coefficient of each IGBT device at each temperature.

[0030] The measurement of short-circuit current can help evaluate the performance of IGBT devices under abnormal conditions. The saturation voltage is a key parameter to measure the voltage drop of IGBT in the on state. The gate current directly affects the switching speed and driving power of IGBT. During the production process of IGBT devices, lax process control may cause abnormal amplitude or overall deviation of some of the above performance parameters in IGBT devices.

[0031] When the quality of the IGBT device is good, all short-circuit currents, all saturation voltages and all gate currents in the corresponding period are relatively stable as a whole. When the IGBT device is abnormal, all short-circuit currents, all saturation voltages and all gate currents in the corresponding period will change dramatically in some positions.

[0032] Therefore, the following analysis is performed based on the abnormal amplitudes or overall deviation characteristics of all short-circuit currents, all saturation voltages, and all gate currents in each time period, specifically: (1) First, the short-circuit current, saturation voltage and gate current of each IGBT device at each time and temperature are combined into a triplet.

[0033] Furthermore, all triples of each IGBT device at each temperature are used as input of the clustering algorithm, and two clusters are output.

[0034] It should be noted that there are many commonly used clustering algorithms. In this embodiment, the fuzzy C-means clustering algorithm is used to cluster all triples of each IGBT device at each temperature to obtain two cluster clusters. In actual application, the implementer can also use the k-means clustering algorithm to cluster the triples. This embodiment does not impose any special restrictions on the choice of clustering method.

[0035] Among them, fuzzy C-means clustering is a well-known technology, and its specific process of data clustering is not described in detail.

[0036] (2) If there is a quality problem with the IGBT device, there will be an abnormal mutation in one or more parameter values ​​in the triplet, which will cause the differences between clusters to become larger, while the differences between clusters in a good quality state are relatively small.

[0037] Therefore, the mean of the difference between all triplets in one cluster of each IGBT device at each temperature and all triplets in another cluster is further analyzed and recorded as the mutation difference coefficient of each IGBT device at each temperature.

[0038] If the value of the mutation difference coefficient is larger, it indicates that the short-circuit current, saturation voltage or gate current of the corresponding IGBT device is more likely to have an abnormal amplitude.

[0039] It should be noted that there are many methods to measure the differences between triplets. In this embodiment, the Mahalanobis distance between all triplets in each cluster and all triplets in another cluster is calculated to measure the differences between corresponding triplets in different clusters. Implementers can also use other methods such as Euclidean distance to measure the differences between three-dimensional data, and this embodiment does not impose any special restrictions.

[0040] The calculation process of the Mahalanobis distance is a well-known technique, and the specific calculation process will not be described in detail.

[0041] (3) Further, in order to analyze the performance differences between different IGBT devices in the same batch, the mean values ​​of the differences between different IGBT devices of various types of data are calculated, and the average value of the mean values ​​of all types of data is taken as the comprehensive difference coefficient. The specific implementation steps are as follows: Firstly, the short-circuit current of each IGBT device at each temperature at all times is combined into a short-circuit current sequence of each IGBT device at each temperature; Furthermore, the mean of the difference between the short-circuit current sequence of each IGBT device at each temperature and the short-circuit current sequence of all other IGBT devices is analyzed and recorded as the short-circuit current difference of each IGBT device at each temperature; wherein, the greater the short-circuit current difference, the more likely that the short-circuit current of the corresponding IGBT is offset.

[0042] It should be noted that there are many methods for measuring the difference between two sequences. In this embodiment, the difference between the short-circuit current sequences of different devices is measured by calculating the DTW distance between the short-circuit current sequence of each IGBT device at each temperature and the short-circuit current sequence of all other IGBT devices. In actual application, as other implementation methods, the implementer may also use other methods for measuring the difference between sequences, such as Euclidean distance. In this embodiment, the following method for analyzing the difference between sequences also uses DTW distance. Regarding the method for calculating the difference between sequences, this embodiment does not impose any special restrictions.

[0043] The calculation process of the DTW distance is a well-known technology, and the specific calculation process will not be described in detail.

[0044] Further, the saturation voltage of each IGBT device at all times at each temperature is combined into a saturation voltage sequence of each IGBT device at each temperature; Furthermore, the average of the differences between the saturation voltage sequence of each IGBT device at each temperature and the saturation voltage sequences of all other IGBT devices is analyzed and recorded as the saturation voltage difference of each IGBT device at each temperature.

[0045] The gate current of each IGBT device at each temperature at all times is used to form a gate current sequence of each IGBT device at each temperature; Furthermore, the average of the differences between the gate current sequence of each IGBT device at each temperature and the gate current sequences of all other IGBT devices is analyzed and recorded as the gate current difference of each IGBT device at each temperature.

[0046] The greater the difference in saturation voltage and gate current of the IGBT devices, the more likely it is that the saturation voltage and gate current of the corresponding IGBT devices will be offset.

[0047] (4) Further, based on the short-circuit current difference, saturation voltage difference and gate current difference of each IGBT device at each temperature, the comprehensive difference coefficient of each IGBT device at each temperature is determined, specifically: The comprehensive difference coefficient of each IGBT device at each temperature is the average of the short-circuit current difference, saturation voltage difference and gate current difference of each IGBT device at each temperature.

[0048] (5) Based on the comprehensive difference coefficient and mutation difference coefficient of each IGBT device at each temperature, determine the mutation offset coefficient of each IGBT device at each temperature to evaluate the quality of the IGBT device, specifically: The expression of the oscillation anomaly coefficient of each IGBT device at each temperature is: ; In the formula, represents the oscillation anomaly coefficient of the ith IGBT device at temperature k; represents the sudden shift coefficient of the ith IGBT device at temperature k; It represents the peak steepness index of the ith IGBT device at temperature k; Represents the oscillation difference value of the i-th IGBT device at temperature k.

[0049] Furthermore, according to the mutation offset coefficient of each IGBT device at each temperature, it can be understood that if the performance of the IGBT device is more unstable, the mutation offset coefficient of the IGBT device will be larger, the mutation difference coefficient of the corresponding IGBT device will be larger, and the comprehensive difference coefficient of the corresponding IGBT device will be larger; conversely, if the performance of the IGBT device is more stable, the mutation offset coefficient of the IGBT device will be smaller, the mutation difference coefficient of the corresponding IGBT device will be smaller, and the comprehensive difference coefficient of the corresponding IGBT device will be smaller.

[0050] S2: Based on the differences between all peak values ​​and their respective adjacent trough values ​​in the fitted curves of all short-circuit currents of each IGBT device, the oscillation difference value of each IGBT device at each temperature is determined, and combined with the distribution of all short-circuit currents in each peak value and its neighborhood, as well as the mutation offset coefficient, the oscillation abnormality coefficient of each IGBT device at each temperature is determined.

[0051] When an IGBT fails, the short-circuit current changes significantly. For example, if the IGBT has poor bonding wires, it is easy to cause short-circuit current oscillations, and the greater the amplitude of the oscillation, the worse the IGBT quality may be. Therefore, the performance of the IGBT device can be further evaluated by combining the change characteristics of the short-circuit current sequence.

[0052] The current will be in a low-value stable state at the beginning. When the short circuit just occurs, the current will suddenly increase to a very high value. At this time, the current will experience a transient process. After reaching the highest value, it will decrease slightly and then increase slightly. It will continue to oscillate to a new larger value stable state. After maintaining for a period of time, it will quickly decay to a low value stable state. In the above transient process, there are multiple peaks and troughs in the short-circuit current sequence. The oscillation characteristics of the short-circuit current are analyzed based on the peaks and troughs.

[0053] (1) First, all short-circuit currents in the short-circuit current sequence of each IGBT device at each temperature are fitted to obtain the current fitting curve of each IGBT device at each temperature.

[0054] It should be noted that there are many commonly used fitting algorithms. In this embodiment, polynomial fitting technology is used to fit the discrete short-circuit current data in the short-circuit current sequence. The implementer may also use other fitting methods such as least squares method, which is not particularly limited in this embodiment.

[0055] (2) Further, the point where the first-order derivative of the current fitting curve of each IGBT device at each temperature is 0 and the second-order derivative is less than 0 is taken as the peak position; similarly, the point where the first-order derivative of the current fitting curve of each IGBT device at each temperature is 0 and the second-order derivative is greater than 0 is taken as the trough position.

[0056] (3) Further, the average value of the difference between each peak value and its adjacent valley value on the current fitting curve of each IGBT device at each temperature is analyzed and recorded as the first average value of each peak value on the current fitting curve of each IGBT device at each temperature; The average of the first differences of all wave peak values ​​on the current fitting curve of each IGBT device at each temperature is analyzed and recorded as the oscillation difference value of each IGBT device at each temperature; the larger the oscillation difference value, the larger the oscillation amplitude of the short-circuit current of the corresponding IGBT device.

[0057] It should be noted that there are many ways to measure the differences between numerical values. In this embodiment, the absolute value of the difference between each peak value and its adjacent valley value is taken as a way to measure the difference between adjacent peak and valley values. In actual application, implementers can also use other methods to measure the difference, such as ratios, and this embodiment does not impose any special restrictions.

[0058] (4) The sharpness of the peak can reflect the response capability of the short-circuit protection circuit inside the IGBT device. The sharper the peak, the faster the short-circuit current rises, and the performance of the IGBT may be relatively poor.

[0059] Therefore, a neighborhood is set with each peak position of each IGBT device as the center point. There are many ways to divide the neighborhood. In this embodiment, the distance between the nearest trough positions is used as the neighborhood radius of the corresponding peak. The implementer can also choose the neighborhood division method according to the specific situation. This embodiment does not impose any special restrictions.

[0060] Determine the kurtosis of each peak value and all fitting values ​​in its neighborhood on the current fitting curve of each IGBT device at each temperature; The cumulative sum of the peaks of all the peaks of each IGBT device at each temperature is determined, and recorded as the peak steepness index of each IGBT device; wherein, the larger the peak steepness index is, the poorer the response capability of the short-circuit protection circuit inside the corresponding IGBT device may be.

[0061] The calculation process of kurtosis is a well-known technique, and the specific calculation process will not be described in detail.

[0062] (5) Further, based on the peak steepness index, oscillation difference value and mutation offset coefficient of each IGBT device at each temperature, the oscillation abnormality coefficient of each IGBT device at each temperature is determined to evaluate whether there is any fault inside the IGBT device, specifically: The expression of the oscillation anomaly coefficient of each IGBT device at each temperature is: ; In the formula, represents the oscillation anomaly coefficient of the ith IGBT device at temperature k; represents the sudden shift coefficient of the ith IGBT device at temperature k; It represents the peak steepness index of the ith IGBT device at temperature k; Represents the oscillation difference value of the i-th IGBT device at temperature k.

[0063] According to the oscillation abnormality coefficient of each IGBT device at each temperature, it can be understood that if the short-circuit current oscillation of the IGBT device is more significant, the oscillation abnormality coefficient of the IGBT device will be larger, the oscillation abnormality coefficient of the corresponding device will be larger, the mutation offset coefficient will be larger, and the oscillation difference value will be larger; conversely, if the short-circuit current of the IGBT device is more stable, the oscillation abnormality coefficient of the IGBT device will be smaller, the oscillation abnormality coefficient of the corresponding device will be smaller, the mutation offset coefficient will be smaller, and the oscillation difference value will be smaller.

[0064] The performance data testing module 103 is used to record the mean of all short-circuit currents of each IGBT device at each temperature as the average short-circuit current of each IGBT device at each temperature; based on the change trend and distribution of the average short-circuit current of each IGBT device at all temperatures, determine the trend significance of each IGBT device, and combine the correlation between the oscillation abnormality coefficient of each IGBT device at all temperatures and the oscillation abnormality coefficient of other IGBT devices at all temperatures to determine the quality assessment coefficient of each IGBT device and detect the IGBT device.

[0065] S3: Based on the changing trend and distribution of the average short-circuit current of each IGBT device at all temperatures, determine the trend significance of each IGBT device, and combine the correlation between the oscillation abnormality coefficient of each IGBT device at all temperatures and the oscillation abnormality coefficient of other IGBT devices at all temperatures to determine the quality assessment coefficient of each IGBT device.

[0066] IGBT devices are often used in parallel and work in high-power scenarios. They are frequently subjected to electrothermal stress, which causes the junction temperature inside the device to rise, leading to failure. For IGBT devices with good quality, the short-circuit current is larger, and the increasing trend of the short-circuit current is more significant as the temperature rises. For IGBT devices with defects, the short-circuit current is smaller, and the increasing trend of the short-circuit current is slower as the temperature rises.

[0067] Therefore, the magnitude and change trend characteristics of the short-circuit current of each IGBT device at different temperatures are analyzed, specifically: (1) The average of all short-circuit currents of each IGBT device at each temperature is recorded as the average short-circuit current of each IGBT device at each temperature.

[0068] The average short-circuit current of each IGBT device at all temperatures is used as the input of the trend detection algorithm, and the absolute value of the output trend detection result is used as the trend significance feature of each IGBT device.

[0069] It should be noted that there are many commonly used trend test algorithms. In this embodiment, the Mann-Kendall trend test algorithm is used to obtain the trend-significant characteristics of each IGBT device. The implementer can also use the Cox-Stuart trend test algorithm to analyze the changing trend of the average short-circuit current at all temperatures. This embodiment does not impose any special restrictions.

[0070] The process of using the Mann-Kendall trend test algorithm to detect the trend of a set of data changes is a well-known technology, and the specific process will not be described in detail.

[0071] The trend significance of each IGBT device is the product of the mean of the average short-circuit current of each IGBT device at all temperatures and the trend significance feature.

[0072] Among them, the greater the trend significance, the more obvious the short-circuit current of the IGBT device changes with temperature, the better the quality of the IGBT device, the larger the mean of the average short-circuit current of the IGBT device at all temperatures, and the greater the trend significance feature; conversely, the smaller the trend significance, the less obvious the short-circuit current of the IGBT device changes with temperature, the worse the quality of the IGBT device, the smaller the mean of the average short-circuit current of the IGBT device at all temperatures, and the smaller the trend significance feature.

[0073] (2) In parallel applications, due to the differences between devices, the loss distribution of IGBT devices will be uneven, resulting in different junction temperatures of IGBT devices, which in turn affects the degree of change in short-circuit current.

[0074] Thus, the oscillation anomaly coefficients of each IGBT device at all temperatures are combined into an oscillation anomaly sequence of each IGBT device in ascending order of temperature; The average value of the correlation between the abnormal oscillation sequence of each IGBT device and the abnormal oscillation sequence of all other IGBT devices is analyzed and recorded as the oscillation consistency of each IGBT device. The greater the oscillation consistency, the less the corresponding IGBT device is affected by uneven loss in the line.

[0075] It should be noted that there are many methods for measuring the correlation between sequences. In this embodiment, the correlation between the oscillation abnormality sequences of different IGBT devices is measured by calculating the Pearson correlation coefficient between the oscillation abnormality sequences of different IGBT devices. The implementer may also use the Kendall rank correlation coefficient to measure the correlation between different sequences. Regarding the selection of the method for measuring the correlation between sequences, this embodiment does not impose any special restrictions, and the implementer may choose according to the specific circumstances.

[0076] (3) Further, based on the trend significance and oscillation consistency of each IGBT device, the quality assessment coefficient of each IGBT device is determined to comprehensively evaluate the quality and reliability of the IGBT device, specifically: The expression of the quality assessment coefficient of each IGBT device is: ; In the formula, represents the quality assessment index of the i-th IGBT device; Indicates the trend significance of the i-th IGBT device; Represents the oscillation consistency of the i-th IGBT device.

[0077] Furthermore, according to the quality assessment coefficient of each IGBT device, it can be understood that if the quality of the IGBT device is better, the quality assessment index of the IGBT device is larger, the trend significance of the corresponding IGBT device is larger, and the oscillation consistency of the corresponding IGBT device is larger; conversely, if the quality of the IGBT device is worse, the quality assessment index of the IGBT device is smaller, the trend significance of the corresponding IGBT device is smaller, and the oscillation consistency of the corresponding IGBT device is smaller.

[0078] Preferably, this embodiment provides a schematic diagram of the quality assessment coefficient extraction process as shown in Figure 2 shown.

[0079] S4: Detect each IGBT device based on the quality assessment coefficient of each IGBT device.

[0080] This embodiment obtains the quality assessment coefficient of the IGBT device by analyzing the trend change characteristics of the short-circuit current of the IGBT device at different temperatures and the degree of short-circuit current oscillation.

[0081] Further, based on the quality assessment coefficient of each IGBT device, each IGBT device is tested, specifically: If the normalized value of the quality assessment index of the IGBT device is less than a preset value, the corresponding IGBT device has a production failure; conversely, if the normalized value of the quality assessment index of the IGBT device is greater than or equal to the preset value, the corresponding IGBT device is produced normally.

[0082] It should be noted that the value of the preset value is manually set. In this embodiment, the value of the preset value is 0.75. The implementer can also set it by himself according to the specific situation. This embodiment does not impose any special restrictions.

[0083] So far, by analyzing the changing trend characteristics of the IGBT short-circuit current at different temperatures, as well as the consistency between the oscillation characteristics of the short-circuit current and the temperature change, the quality evaluation index of the IGBT device is determined, the quality of the IGBT device is comprehensively evaluated, and the performance test of the IGBT device is completed.

[0084] Preferably, the schematic diagram of the automated testing process of the IGBT device provided in this embodiment is as follows: Figure 3 shown.

[0085] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0086] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0087] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent automatic test system for IGBT devices, characterized in that: The system comprises: The performance data acquisition module is used to obtain various types of data of each IGBT device at different temperatures at different times, where the short-circuit current is one type of data among all types of data, and all types of data of each IGBT device at different temperatures at different times form a triplet; A performance data analysis module is used to cluster all triplets of each IGBT device at each temperature to obtain two clusters, and determine the mutation difference coefficient of each IGBT device at each temperature based on the difference of all triplets between the two clusters in each IGBT device; Based on the difference between various data of each IGBT device at all times and the corresponding similar data of other IGBT devices at all times, combined with the mutation difference coefficient, determine the mutation offset coefficient of each IGBT device at each temperature; Based on the differences between all peak values ​​and their respective adjacent trough values ​​in the fitted curves of all short-circuit currents of each IGBT device, the oscillation difference value of each IGBT device at each temperature is determined, and combined with the distribution of all short-circuit currents in each peak value and its neighborhood, as well as the mutation offset coefficient, the oscillation abnormality coefficient of each IGBT device at each temperature is determined; The performance data test module is used to record the average value of all short-circuit currents of each IGBT device at each temperature as the average short-circuit current of each IGBT device at each temperature; Based on the changing trend and distribution of the average short-circuit current of each IGBT device at all temperatures, the trend significance of each IGBT device is determined, and combined with the correlation between the oscillation abnormality coefficient of each IGBT device at all temperatures and the oscillation abnormality coefficient of other IGBT devices at all temperatures, the quality assessment coefficient of each IGBT device is determined, and the IGBT devices are tested.

2. The intelligent automatic test system for IGBT devices according to claim 1, characterized in that: The mutation difference coefficient of each IGBT device at each temperature is the average of the differences between all triplets in one cluster of each IGBT device and all triplets in another cluster.

3. The intelligent automatic test system for IGBT devices according to claim 1, characterized in that: The method for determining the sudden shift coefficient of each IGBT device at each temperature is as follows: Based on the differences between various data of each IGBT device at all times and the corresponding similar data of other IGBT devices at all times, the comprehensive difference coefficient of each IGBT device at each temperature is determined; The expression of the sudden shift coefficient of each IGBT device at each temperature is: ; In the formula, represents the sudden shift coefficient of the ith IGBT device at temperature k; represents the mutation difference coefficient of the ith IGBT device at temperature k; Represents the comprehensive difference coefficient of the i-th IGBT device at temperature k.

4. The intelligent automatic test system for IGBT devices according to claim 3, characterized in that: The method for determining the comprehensive difference coefficient of each IGBT device at each temperature is as follows: The mean values ​​of the differences between different IGBT devices for each type of data are calculated, and the average value of the mean values ​​for all types of data is taken as the comprehensive difference coefficient.

5. The intelligent automatic test system for IGBT devices according to claim 1, characterized in that: The method for determining the oscillation difference value of each IGBT device at each temperature is as follows: Fitting all short-circuit currents of each IGBT device at each temperature to obtain a current fitting curve, analyzing the average of the differences between each peak value and its adjacent valley value on the current fitting curve of each IGBT device at each temperature, and recording it as the first average of each peak value on the current fitting curve of each IGBT device at each temperature; The oscillation difference value of each IGBT device at each temperature is the average of the first average values ​​of all wave peak values ​​on the current fitting curve of each IGBT device at each temperature.

6. The intelligent automatic test system for IGBT devices according to claim 5, characterized in that: The method for determining the oscillation abnormality coefficient of each IGBT device at each temperature is as follows: Determine the kurtosis of each peak value and all fitting values ​​in its neighborhood on the current fitting curve of each IGBT device at each temperature; Determine the cumulative sum of the peaks of all the peaks of each IGBT device at each temperature, and record it as the peak steepness index of each IGBT device; The expression of the oscillation anomaly coefficient of each IGBT device at each temperature is: ; In the formula, represents the oscillation anomaly coefficient of the ith IGBT device at temperature k; represents the sudden shift coefficient of the ith IGBT device at temperature k; It represents the peak steepness index of the ith IGBT device at temperature k; Represents the oscillation difference value of the i-th IGBT device at temperature k.

7. The intelligent automatic test system for IGBT devices according to claim 1, characterized in that: The method for determining the trend significance of each IGBT device is: The average short-circuit current of each IGBT device at all temperatures is used as the input of the trend test algorithm, and the absolute value of the output trend test result is used as the trend significance feature of each IGBT device; The trend significance of each IGBT device is the product of the mean of the average short-circuit current of each IGBT device at all temperatures and the trend significance feature.

8. The intelligent automatic test system for IGBT devices according to claim 1, characterized in that: The method for determining the quality assessment coefficient of each IGBT device is: Determine the oscillation consistency of each IGBT device based on the correlation between the oscillation anomaly coefficient of each IGBT device at all temperatures and the oscillation anomaly coefficient of other IGBT devices at all temperatures; The expression of the quality assessment coefficient of each IGBT device is: ; In the formula, represents the quality assessment index of the i-th IGBT device; Indicates the trend significance of the i-th IGBT device; Represents the oscillation consistency of the i-th IGBT device.

9. The intelligent automatic test system for IGBT devices according to claim 8, characterized in that: The method for determining the oscillation consistency of each IGBT device is: The oscillation abnormality coefficients of each IGBT device at all temperatures are combined into an oscillation abnormality sequence of each IGBT device; The oscillation consistency of each IGBT device is the average value of the correlation between the oscillation abnormality sequence of each IGBT device and the oscillation abnormality sequences of all other IGBT devices.

10. The intelligent automatic test system for IGBT devices according to claim 1, characterized in that: The detecting of the IGBT device comprises: If the normalized value of the quality assessment index of the IGBT device is less than a preset value, the corresponding IGBT device has a production fault; otherwise, the IGBT device is produced normally.

Images