CN115342938A - Method and device for determining junction temperature of power semiconductor device - Google Patents
Method and device for determining junction temperature of power semiconductor device Download PDFInfo
- Publication number
- CN115342938A CN115342938A CN202110463113.5A CN202110463113A CN115342938A CN 115342938 A CN115342938 A CN 115342938A CN 202110463113 A CN202110463113 A CN 202110463113A CN 115342938 A CN115342938 A CN 115342938A
- Authority
- CN
- China
- Prior art keywords
- temperature
- condition information
- power semiconductor
- semiconductor device
- junction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000004065 semiconductor Substances 0.000 title claims abstract description 125
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000013528 artificial neural network Methods 0.000 claims abstract description 75
- 238000012360 testing method Methods 0.000 claims description 54
- 230000032683 aging Effects 0.000 claims description 18
- 238000012549 training Methods 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 12
- 238000002474 experimental method Methods 0.000 claims description 7
- 230000005669 field effect Effects 0.000 claims description 2
- 229910044991 metal oxide Inorganic materials 0.000 claims description 2
- 150000004706 metal oxides Chemical class 0.000 claims description 2
- 238000004806 packaging method and process Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 4
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000009529 body temperature measurement Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01K—MEASURING TEMPERATURE; MEASURING QUANTITY OF HEAT; THERMALLY-SENSITIVE ELEMENTS NOT OTHERWISE PROVIDED FOR
- G01K7/00—Measuring temperature based on the use of electric or magnetic elements directly sensitive to heat ; Power supply therefor, e.g. using thermoelectric elements
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/26—Testing of individual semiconductor devices
- G01R31/2601—Apparatus or methods therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Testing Of Individual Semiconductor Devices (AREA)
Abstract
The invention discloses a method and a device for determining junction temperature of a power semiconductor device. The method comprises the following steps: acquiring working condition information of an electric driving device comprising a power semiconductor device; acquiring a device temperature of the electric drive device corresponding to the working condition information; and determining the junction temperature of the power semiconductor device according to the acquired working condition information and the device temperature through a pre-trained neural network. According to the method and the device for determining the junction temperature of the power semiconductor device, the junction temperature of the power semiconductor device can be determined through the pre-trained neural network according to the related information of the electric driving device comprising the power semiconductor device, and the convenience and the accuracy for determining the junction temperature of the power semiconductor device are improved.
Description
Technical Field
The present invention relates to the field of junction temperature measurement, and in particular, to a method and apparatus for determining the junction temperature of a power semiconductor device.
Background
In an electric drive, power semiconductors are typically used as fast switches to convert direct current to alternating current to power the motor of the electric drive.
The power semiconductor device usually has switching loss and conduction loss during operation, and the switching loss and the conduction loss can increase the temperature of the power semiconductor device and even exceed the tolerance threshold of the power semiconductor device, so that the power semiconductor device can be damaged and even a system comprising the power semiconductor device can be in a potential safety hazard. Therefore, it is important to accurately determine the junction temperature of the power semiconductor device.
In general, a model of the correspondence between the junction temperature of the power semiconductor device and the ambient temperature of the power semiconductor device needs to be established through a large amount of experimental data, such as power loss of the power semiconductor device, temperature rise, cooling conditions, and the ambient temperature of the power semiconductor device, so that the junction temperature of the power semiconductor device is inversely deduced through the model by measuring the ambient temperature of the power semiconductor device. However, the establishment of the model requires a large number of experimental measurements and complicated calibration and debugging processes, and the establishment of the model is very complicated. Furthermore, the junction temperature of the power semiconductor device determined in this way is not accurate enough, since the actual state of the power semiconductor device is not taken into account when determining the junction temperature of the power semiconductor device using such a model.
Therefore, there is a need for a way to more accurately and conveniently determine the junction temperature of a power semiconductor device.
Disclosure of Invention
The invention aims to provide a method and a device for determining junction temperature of a power semiconductor device.
According to an aspect of the invention, there is provided a method for determining the junction temperature of a power semiconductor device, the method comprising: acquiring working condition information of an electric driving device comprising a power semiconductor device; acquiring a device temperature of the electric drive device corresponding to the working condition information; and determining the junction temperature of the power semiconductor device through a pre-trained neural network according to the acquired working condition information and the device temperature.
According to another aspect of the present invention, there is provided an apparatus for determining the junction temperature of a power semiconductor device, the apparatus comprising: a condition acquisition unit configured to be able to acquire condition information of an electric drive apparatus including a power semiconductor device; a temperature acquisition unit configured to be able to acquire a device temperature of the electric drive device corresponding to the condition information; and the junction temperature determining unit is configured to determine the junction temperature of the power semiconductor device according to the acquired working condition information and the device temperature through a pre-trained neural network.
According to another aspect of the invention, a computer program product is provided, wherein the computer program product comprises a computer program which, when executed by a processor, causes the processor to carry out the method for determining the junction temperature of a power semiconductor device according to the invention.
According to the method and the device for determining the junction temperature of the power semiconductor device, the junction temperature of the power semiconductor device can be determined through the pre-trained neural network according to the relevant information of the electric driving device comprising the power semiconductor device, and the convenience and the accuracy for determining the junction temperature of the power semiconductor device are improved.
Drawings
The foregoing and other aspects of the invention will be more fully understood from the following detailed description taken in conjunction with the accompanying drawings, in which:
fig. 1 shows a flow chart of a method for determining a junction temperature of a power semiconductor device according to an exemplary embodiment of the present invention.
Fig. 2 shows a detailed flow chart of a method for determining the junction temperature of a power semiconductor device according to an exemplary embodiment of the present invention.
Fig. 3 shows a schematic diagram for determining the junction temperature of a power semiconductor device according to an exemplary embodiment of the present invention.
Fig. 4 shows a block diagram of an apparatus for determining the junction temperature of a power semiconductor device according to an exemplary embodiment of the present invention.
Detailed Description
In the following, some exemplary embodiments of the invention will be described in more detail with reference to the drawings in order to better understand the basic ideas and advantages of the invention.
Fig. 1 shows a flow chart of a method for determining a junction temperature of a power semiconductor device according to an exemplary embodiment of the present invention.
Referring to fig. 1, in step S1, condition information of an electric drive apparatus including a power semiconductor device is acquired.
Here, the electric driving apparatus may include a power semiconductor device and a motor coupled with the power semiconductor device, and may further include other related functional devices. The power semiconductor device may function as a fast switch to convert direct current input to the power semiconductor device into alternating current, thereby allowing the motor to receive the converted alternating current to operate. For example, the electric drive may be provided in a vehicle or other industrial system. As an example, the power semiconductor device may include an Insulated Gate Bipolar Transistor (IGBT) and/or a metal-oxide semiconductor field effect transistor (MOSFET).
The acquired condition information may include information capable of reflecting an aging state of the power semiconductor device. As an example, the operating condition information may include at least one of: the current flowing through the power semiconductor device, the voltage across the power semiconductor device, the switching frequency of the power semiconductor device, the torque of the motor, the rotational speed of the motor, the power of the motor, and the accumulated operating time of the electric drive. Here, the accumulated operating time of the electric drive apparatus may reflect the aging state of the power semiconductor device.
In step S2, a device temperature of the electric drive device corresponding to the operating condition information is acquired.
For example, while the condition information of the electric drive device is acquired, the corresponding device temperature of the electric drive device may be acquired.
As examples, the device temperature may include: the temperature inside the module of the package module in which the power semiconductor device is packaged and/or the temperature of the circuit board at the connection position of the printed circuit board to which the power semiconductor device is connected.
For example, the module internal temperature may be a temperature measured by a temperature sensor, such as an NTC temperature sensor, provided in a module, such as a power module, in which the power semiconductor device is packaged, and the circuit board temperature may be a temperature measured by a temperature sensor provided at a connection position on the printed circuit board to which the power module, in which the power semiconductor device is provided, is connected. These temperature sensors may be temperature sensors already present in the electrically driven device, and therefore the method for determining the junction temperature of the power semiconductor device according to the present invention may obtain the device temperature at a low cost without additionally providing a temperature sensor.
And S3, determining the junction temperature of the power semiconductor device through a pre-trained neural network according to the acquired working condition information and the device temperature.
In other words, the operating condition information obtained in step S1 and the corresponding device temperature obtained in step S2 may be used as input data of a pre-trained neural network, and the output data of the neural network may be used as the junction temperature of the power semiconductor device.
Here, in order to determine the junction temperature of the power semiconductor device in real time, the operating condition information and the corresponding device temperature may be periodically acquired, and the junction temperature of the power semiconductor device may be periodically determined. As an example, the condition information and the device temperature may be periodically acquired at a first frequency, and the junction temperature of the power semiconductor device may be periodically determined at a second frequency by a pre-trained neural network based on the condition information and the device temperature that have been acquired during the current period at the second frequency. Also, the first frequency may be equal to or greater than the second frequency.
For example, the junction temperature of the power semiconductor device may be determined each time after the condition information and the corresponding device temperature are acquired in steps S1 and S2, i.e., in step S3. I.e. the first frequency is equal to the second frequency. In addition, the working condition information and the corresponding device temperature can be cached after being acquired, and then the junction temperature of the power semiconductor device can be determined or respectively determined through the neural network according to the plurality of working condition information and the corresponding device temperatures. I.e. the first frequency is greater than the second frequency.
The neural network may be any type of neural network, such as RNN, CNN, deep learning neural network, or the like. Further, the neural network may be trained using, as training data, a set of condition information and corresponding device temperatures obtained in advance from the electric drive device in an arbitrary state.
For example, the electric drive may be in any one of the following states: the test system comprises a working state, a normal aging test state and an accelerated aging test state.
The working state may represent a working state of the electric drive device in at least one working environment, and the working condition information obtained in the working state may be referred to as working condition information, and the device temperature obtained in the working state may be referred to as a working device temperature.
The normal aging test state and the accelerated aging test state can represent the aging test state of the electric drive device in a laboratory environment, the working condition information acquired in the aging test state can be called test working condition information, and the device temperature acquired in the aging test state can be called test device temperature.
As an example, the neural network may be trained by: and the pre-acquired test working condition information, the corresponding set of the temperature of the test device and the junction temperature measured by the aging experiment as the corresponding labeled junction temperature.
For example, in a laboratory, the junction temperature of the power semiconductor device may be directly measured by an infrared camera, a temperature sensor, or other means as the noted junction temperature.
Further, as another example, the neural network may be trained by: the method comprises the steps of obtaining working condition information in advance, a set of corresponding working device temperatures and estimated junction temperatures serving as corresponding labeled junction temperatures, wherein the estimated junction temperatures are junction temperatures corresponding to the internal temperatures of the modules estimated according to the corresponding relation between the internal temperatures of the modules and the junction temperatures in factory information of packaging modules packaged with power semiconductor devices.
For example, the corresponding relationship between the internal temperature of the module and the junction temperature may be obtained from factory information (data sheet) of the module in which the power semiconductor device is packaged, for example, the power module, so that the junction temperature is determined by the internal temperature of the module included in the obtained device temperature as the labeled junction temperature.
It should be appreciated that the amount of training data used to train the neural network may be determined based on actual requirements. The electric drive device used to acquire the training data in training the neural network may be a different but the same type (e.g., model) of electric drive device as that used to acquire the condition information and the device temperature in steps S1 and S2 above.
In order to improve the training efficiency, the neural network can be trained in a cloud server or an edge server, and the trained neural network is stored in the cloud server or the edge server, so that resources of a vehicle or other industrial systems provided with the electric driving device are saved. It should be understood that the trained neural network may also be stored in a vehicle or other industrial system in which the electric drive is provided.
In addition, in order to make the trained neural network more adaptive to various working situations of the electric power working system, the neural network can be further trained by using the working condition information and the corresponding device temperature obtained after the training of the neural network is completed, so as to update the neural network (S7).
As an example, the method for determining the junction temperature of a power semiconductor device according to the present invention may further comprise: and when a preset number of pieces of working condition information and corresponding device temperatures are obtained, training the neural network by using the preset number of pieces of working condition information and the corresponding device temperatures so as to update the neural network. Here, the predetermined number may correspond to one set of condition information and device temperature or two or more sets of condition information and device temperature.
In this case, the junction temperature of the power semiconductor device may be determined in step S3 above by the newly updated neural network from the acquired operating condition information and the device temperature.
Furthermore, after the training of the neural network is completed, the neural network may be used to determine the junction temperature of the power semiconductors in all electrically driven devices of the respective type (e.g. in case the neural network is stored in a cloud server or edge server). Therefore, although the working condition information and the working device temperature are generally obtained from the electric driving device in the working state to determine the junction temperature, it is also possible to obtain the test condition information and the test device temperature from the electric driving device in the aging test state in the laboratory.
Thus, as an example, after obtaining the operating condition information and the corresponding device temperature, the method for determining the junction temperature of the power semiconductor device according to the present invention may further include: it is determined whether the acquired condition information and the device temperature are the operating condition information and the operating device temperature (S4). For example, the condition information and the device temperature may have associated indicia that may be used to distinguish between the condition information and the device temperature from the electrically powered device in an operational running state, in a normal burn-in test state, or in an accelerated burn-in test state.
And under the condition that the obtained working condition information and the device temperature are determined to be the working condition information and the working device temperature, the junction temperature of the power semiconductor device can be determined through a pre-trained neural network according to the obtained working condition information and the working device temperature (S3).
Under the condition that the obtained working condition information and the device temperature are determined not to be the working condition information and the working device temperature (but to be the test working condition information and the test device temperature), the neural network can be trained through the obtained test working condition information and the test device temperature so as to update the neural network (S7).
As an example, the neural network may be trained to update the neural network using the first predetermined number of operating condition information and corresponding operating device temperatures, and the corresponding determined junction temperatures as the labeled junction temperatures whenever the first predetermined number of operating condition information and corresponding operating device temperatures are obtained; and/or when a second preset number of pieces of test condition information and corresponding test device temperatures are obtained, the second preset number of pieces of test condition information and corresponding test device temperatures and corresponding junction temperatures measured by the aging experiment are used as marked junction temperatures, and the neural network is trained to be updated (S7). Here, the first predetermined number may be the same as or different from the second predetermined number, and may correspond to, for example, several tens of sets of condition information and device temperature, respectively.
Furthermore, in order to ensure the safety of the electric drive apparatus, the method for determining the junction temperature of the power semiconductor device according to the present invention may further include, as an example: comparing the determined junction temperature with a predetermined junction temperature threshold to determine whether the determined junction temperature is greater than the predetermined junction temperature threshold; in case the determined junction temperature is larger than said predetermined junction temperature threshold, the electric drive is shut down.
By the mode, the junction temperature of the semiconductor device can be determined in real time, and in the process of determining the junction temperature, the aging condition of the power semiconductor device is considered, so that the accuracy of the determined junction temperature is higher. And because the junction temperature of the power semiconductor device is determined through the neural network trained in advance, the determination of the junction temperature is very convenient. In addition, the neural network used for determining the junction temperature of the power semiconductor device can be updated in real time, so that the neural network can adapt to more working conditions, and the accuracy of determining the junction temperature through the neural network can be continuously improved.
Fig. 2 shows a detailed flow chart of a method for determining the junction temperature of a power semiconductor device according to an exemplary embodiment of the present invention.
Referring to fig. 2, in step S1, condition information of the electric driving apparatus including the power semiconductor device may be acquired.
In step S2, a device temperature of the electric drive device corresponding to the operating condition information may be acquired.
In step S4, it may be determined whether the acquired condition information and device temperature are the operating condition information and the operating device temperature.
In the case where it is determined in step S4 that the acquired operating condition information and device temperature are the operating condition information and the operating device temperature ("Y" in fig. 2), step S3 may be performed to determine the junction temperature of the power semiconductor device based on the acquired operating condition information and operating device temperature through the newly updated neural network.
Next, step S5 may be performed to determine whether a first predetermined amount of operating condition information and corresponding operating device temperatures have been obtained. It should be appreciated that this step may be omitted where the first predetermined amount corresponds to a set of operating condition information and operating device temperature.
In case it is determined in step S5 that the first predetermined amount of operating condition information and the corresponding operating device temperatures have been obtained ("Y" in fig. 2), the neural network may be trained in a following step S7 using the first predetermined amount of operating condition information and the corresponding operating device temperatures, and the respective determined junction temperatures as the annotated junction temperatures, to update the neural network.
Further, in the case where it is determined in step S4 that the acquired condition information and device temperature are not the operating condition information and the operating device temperature but the test condition information and the corresponding test device temperature ("N" in fig. 2), step S6 may be performed to determine whether the second predetermined number of test condition information and the corresponding test device temperature have been acquired. It should be appreciated that this step may be omitted where the second predetermined amount corresponds to a set of test condition information and test device temperature.
In the case where it is determined in step S6 that the second predetermined number of pieces of test condition information and the corresponding test apparatus temperatures have been acquired ("Y" in fig. 2), the neural network may be trained to update the neural network in the next step S7, using the second predetermined number of pieces of test condition information and the corresponding test apparatus temperatures, and the corresponding junction temperatures measured in the aging experiment as the labeled junction temperatures.
In addition, after the junction temperature is determined in step S3, step S8 may also be performed to compare the determined junction temperature with a predetermined junction temperature threshold value, so as to determine whether the determined junction temperature is greater than the predetermined junction temperature threshold value.
In case it is determined in step S8 that the determined junction temperature is greater than the predetermined junction temperature threshold, step S9 may be performed to stop the electric drive apparatus.
It should be understood that fig. 2 is only an example of a method for determining the junction temperature of a power semiconductor device according to the present invention, and that the method for determining the junction temperature of a power semiconductor device according to the present invention may comprise more or less steps than fig. 2 according to actual requirements.
Fig. 3 shows a schematic diagram of determining the junction temperature of a power semiconductor device according to an exemplary embodiment of the present invention.
With reference to fig. 3, the electric drive comprising the power semiconductor device of which the junction temperature is to be determined may be located in a vehicle, shown in the figure of an automobile, at the lowermost part of fig. 3, or in another industrial system or laboratory, schematically shown as a circuit board.
The pre-trained neural network for determining the junction temperature of the power semiconductor devices is schematically shown in a cylindrical form, which may be stored in an edge server graphically shown in the middle rectangular box of fig. 3, or may be stored in a cloud server graphically shown in the server at the top of fig. 3. In the case where the neural network is stored in the cloud server in the upper part of FIG. 3, the edge server may be omitted.
The condition information and device temperature may be obtained from the electric drive system in the vehicle, other industrial system, or laboratory shown below in fig. 3. The obtained working condition information can be sent to the edge service or the cloud server through transmission equipment schematically shown by a radar, so that the junction temperature of the power semiconductor device is determined or the neural network is updated through the neural network stored in the edge service or the cloud server, in addition, the determined junction temperature can be compared with the junction temperature with the preset threshold value at the edge service or the cloud server or corresponding devices in vehicles, other industrial systems or laboratories, and whether the electric driving device needs to be stopped or not is determined according to the comparison result.
For example, where the neural network is stored on an edge server or a cloud server, the frequency of obtaining the condition information and device temperature from the vehicle, other industrial system, or laboratory may be equal to or less than the frequency of transmitting the condition information and device temperature to the edge server or cloud server.
According to the method for determining the junction temperature of the power semiconductor device, the junction temperature of the power semiconductor device can be determined through the pre-trained neural network according to the related information of the power driving device comprising the power semiconductor device, and convenience and accuracy for determining the junction temperature of the power semiconductor device are improved.
Fig. 4 shows a block diagram of an apparatus for determining the junction temperature of a power semiconductor device according to an exemplary embodiment of the present invention.
Referring to fig. 4, an apparatus for determining the junction temperature of a power semiconductor device according to the present invention includes: the device comprises a working condition acquisition unit 1, a temperature acquisition unit 2 and a junction temperature determination unit 3.
The condition acquisition unit 1 is configured to be able to acquire condition information of an electric drive apparatus including a power semiconductor device.
The temperature acquisition unit 2 is configured to be able to acquire a device temperature of the electric drive device corresponding to the condition information.
The junction temperature determination unit 3 is configured to determine the junction temperature of the power semiconductor device according to the acquired operating condition information and the device temperature through a pre-trained neural network.
The determination of the condition information, the device temperature, and the training and junction temperatures of the neural network has been described in detail above with reference to fig. 1 to 3, and will not be described herein again.
According to the device for determining the junction temperature of the power semiconductor device, the junction temperature of the power semiconductor device can be determined through the pre-trained neural network according to the related information of the electric driving device comprising the power semiconductor device, and convenience and accuracy for determining the junction temperature of the power semiconductor device are improved.
There is further provided in accordance with an exemplary embodiment of the invention a computer program product, wherein the computer program product comprises a computer program which, when executed by a processor, causes the processor to carry out the method for determining the junction temperature of a power semiconductor device according to the invention. The computer program product may include a computer program, program code, instructions, or some combination thereof for instructing or configuring hardware devices, individually or collectively, to operate as desired. The computer program and/or program code can include a program or computer-readable instructions, software components, software modules, data files, data structures, etc., that can be implemented by one or more hardware devices. Examples of program code may include machine code, such as produced by a compiler, and higher level program code, such as executed using an interpreter.
Furthermore, each unit in the above-described apparatuses and devices according to exemplary embodiments of the present invention may be implemented as a hardware component or a software module. Further, the respective units may be implemented by using, for example, a Field Programmable Gate Array (FPGA), an Application Specific Integrated Circuit (ASIC), or a processor according to the processing performed by the respective units defined by those skilled in the art.
Although the invention is illustrated and described herein with reference to specific embodiments, the invention is not intended to be limited to the details shown. Rather, various modifications may be made in the details within the scope of the invention.
List of reference numerals
S1, acquiring working condition information of electric driving device comprising power semiconductor device
S2, acquiring device temperature corresponding to working condition information of electric drive device
S3, determining junction temperature of the power semiconductor device according to the acquired working condition information and the device temperature through a pre-trained neural network
S4, determining whether the obtained working condition information and the device temperature are the working condition information and the working device temperature
S5, whether the first preset amount of working condition information and the corresponding working device temperature are acquired
S6, whether a second preset amount of test working condition information and corresponding test device temperature are acquired or not
S7, training the neural network to update the neural network
S8 determines whether the junction temperature is greater than a predetermined junction temperature threshold
S9, stopping the electric driving device
1 operating mode acquisition unit
2 temperature acquisition Unit
3 junction temperature determining unit
Claims (10)
1. A method for determining junction temperature of a power semiconductor device, the method comprising:
acquiring working condition information of an electric driving device comprising a power semiconductor device;
acquiring a device temperature of the electric drive device corresponding to the working condition information;
and determining the junction temperature of the power semiconductor device according to the acquired working condition information and the device temperature through a pre-trained neural network.
2. The method of claim 1, wherein the method further comprises:
training the neural network to update the neural network using a predetermined number of pieces of condition information and corresponding device temperatures whenever the predetermined number of pieces of condition information and corresponding device temperatures are acquired,
the step of determining the junction temperature of the power semiconductor device through a pre-trained neural network according to the obtained working condition information and the device temperature comprises the following steps:
and determining the junction temperature of the power semiconductor device according to the acquired working condition information and the device temperature through the newly updated neural network.
3. The method according to claim 2, wherein the operating condition information is operating condition information, the device temperature is an operating device temperature corresponding to the operating condition information, wherein the operating condition information and the operating device temperature respectively represent the operating condition information and the device temperature obtained from the electric drive device in an operating state, and/or
The operating mode information is test operating mode information, and the device temperature is the testing arrangement temperature that corresponds with test operating mode information, and test operating mode information and testing arrangement temperature show respectively from the operating mode information and the device temperature that are in the electric drive device of ageing experiment test state and acquire, and wherein, ageing experiment test state includes: a normal burn-in test condition and/or an accelerated burn-in test condition,
wherein the method further comprises:
determining whether the acquired condition information and the device temperature are the operating condition information and the operating device temperature,
wherein, under the condition that the obtained working condition information and the device temperature are determined to be the working condition information and the working device temperature, the junction temperature of the power semiconductor device is determined through a pre-trained neural network according to the obtained working condition information and the working device temperature,
and training the neural network through the acquired test working condition information and the acquired test device temperature to update the neural network under the condition that the acquired working condition information and the acquired device temperature are determined not to be the working condition information and the working device temperature.
4. The method of claim 3, wherein training the neural network to update the neural network using a predetermined number of operating condition information and corresponding device temperatures each time the predetermined number of operating condition information and corresponding device temperatures are obtained comprises:
when a first preset amount of working condition information and corresponding working device temperatures are obtained, the first preset amount of working condition information, the corresponding working device temperatures and the corresponding determined junction temperatures are used as labeled junction temperatures, and the neural network is trained to update the neural network; and/or
And when a second preset number of test working condition information and corresponding test device temperatures are obtained, the second preset number of test working condition information and corresponding test device temperatures and corresponding junction temperatures measured by the aging experiment are used as labeled junction temperatures to train the neural network so as to update the neural network.
5. The method of claim 3, wherein,
the electric drive apparatus further comprises a motor, wherein the power semiconductor device is used for converting direct current into alternating current, the motor is coupled to the power semiconductor device to receive the converted alternating current,
wherein the operating condition information comprises at least one of: the current flowing through the power semiconductor device, the voltage across the power semiconductor device, the switching frequency of the power semiconductor device, the torque of the motor, the rotational speed of the motor, the power of the motor, and the accumulated operating time of the electric drive,
wherein the device temperature comprises: the module internal temperature of a package module in which the power semiconductor device is packaged and/or the circuit board temperature at the connection position of the printed circuit board to which the power semiconductor device is connected.
6. The method of claim 5, wherein the neural network is trained by: pre-acquiring test working condition information, a corresponding set of test device temperatures and junction temperatures measured by an aging experiment as corresponding labeled junction temperatures; and/or
The method comprises the steps of obtaining working condition information in advance, a set of corresponding working device temperatures and estimated junction temperatures serving as corresponding labeled junction temperatures, wherein the estimated junction temperatures are junction temperatures corresponding to the internal temperatures of the modules estimated according to the corresponding relation between the internal temperatures of the modules and the junction temperatures in factory information of packaging modules packaged with power semiconductor devices.
7. The method of any one of claims 1 to 6,
periodically acquiring working condition information and device temperature at a first frequency; periodically determining the junction temperature of the power semiconductor device through a pre-trained neural network at a second frequency according to the working condition information and the device temperature which are acquired in the current period at the second frequency,
wherein the first frequency is equal to or greater than the second frequency.
8. The method according to any of claims 1 to 6, wherein the power semiconductor device comprises an insulated gate bipolar transistor and/or a metal-oxide semiconductor field effect transistor,
wherein the method further comprises:
comparing the determined junction temperature with a predetermined junction temperature threshold to determine whether the determined junction temperature is greater than the predetermined junction temperature threshold;
in case the determined junction temperature is larger than said predetermined junction temperature threshold, the electric drive is shut down.
9. An apparatus for determining a junction temperature of a power semiconductor device, the apparatus comprising:
a condition acquisition unit configured to be able to acquire condition information of an electric drive apparatus including a power semiconductor device;
a temperature acquisition unit configured to be able to acquire a device temperature of the electric drive device corresponding to the condition information;
and the junction temperature determining unit is configured to determine the junction temperature of the power semiconductor device according to the acquired working condition information and the device temperature through a pre-trained neural network.
10. A computer program product, wherein the computer program product comprises a computer program which, when executed by a processor, causes the processor to carry out the method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110463113.5A CN115342938A (en) | 2021-04-27 | 2021-04-27 | Method and device for determining junction temperature of power semiconductor device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110463113.5A CN115342938A (en) | 2021-04-27 | 2021-04-27 | Method and device for determining junction temperature of power semiconductor device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115342938A true CN115342938A (en) | 2022-11-15 |
Family
ID=83977683
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110463113.5A Pending CN115342938A (en) | 2021-04-27 | 2021-04-27 | Method and device for determining junction temperature of power semiconductor device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115342938A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115828699A (en) * | 2022-12-19 | 2023-03-21 | 华中科技大学 | Power semiconductor module full-life-cycle junction temperature prediction method, system and terminal |
-
2021
- 2021-04-27 CN CN202110463113.5A patent/CN115342938A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115828699A (en) * | 2022-12-19 | 2023-03-21 | 华中科技大学 | Power semiconductor module full-life-cycle junction temperature prediction method, system and terminal |
| US11976984B1 (en) * | 2022-12-19 | 2024-05-07 | Huazhong University Of Science And Technology | Method and system for predicting junction temperature of power semiconductor module in full life cycle, and terminal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106291303B (en) | The determination of the aging of power semiconductor in power-driven system | |
| Celaya et al. | Accelerated aging system for prognostics of power semiconductor devices | |
| Sonnenfeld et al. | An agile accelerated aging, characterization and scenario simulation system for gate controlled power transistors | |
| EP3203250A1 (en) | Method and device for estimating a level of damage or a lifetime expectation of a power semiconductor module | |
| CN103207362A (en) | System and method for monitoring in real time the operating state of an IGBT device | |
| CN106575106B (en) | Simulation device and method for simulating a peripheral circuit arrangement which can be connected to a control device | |
| US11169201B2 (en) | Diagnostic device and method to establish degradation state of electrical connection in power semiconductor device | |
| US11527780B2 (en) | Battery model and control application calibration systems and methods | |
| US11888321B2 (en) | Power conversion apparatus and method for controlling output impedance of power conversion apparatus | |
| EP3382407B1 (en) | Method for controlling the health of a multi-die power module and a multi-die health monitoring device | |
| US11320321B2 (en) | Determining a characteristic temperature of an electric or electronic system | |
| CN107479367B (en) | Motor control method and system and vehicle | |
| CN115342938A (en) | Method and device for determining junction temperature of power semiconductor device | |
| Roy et al. | Aging Detection and State of Health Estimation of Live Power Semiconductor Devices Using SSTDR Embedded PWM Sequence | |
| Krone et al. | Advanced condition monitoring system based on on-line semiconductor loss measurements | |
| Babel et al. | Condition monitoring and failure prognosis of IGBT inverters based on on-line characterization | |
| EP3933361A1 (en) | Arrangement and method for measuring temperature of cable connection terminal of electric drive | |
| US9859739B2 (en) | Load driver circuit including load model parameter estimation | |
| US20170077862A1 (en) | Calculation of mosfet switch temperature in motor control | |
| WO2022217597A1 (en) | Fault early-warning method and apparatus for motor driver | |
| CN111512170A (en) | Power conversion device, rotating electric machine system using the same, and diagnostic method therefor | |
| CN109945396B (en) | Air conditioner, and detection method and detection device for driving circuit fault | |
| Oettmeier et al. | Test-bench set-up for high-power PMSM test drive for validating novel control schemes | |
| Wang et al. | Reliability evaluation method for robot servo system based on accelerated degradation test | |
| US10879888B2 (en) | Method for actuating at least one semiconductor switch, in particular in a component of a motor vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |