Junction temperature prediction method based on SiC MOSFET turn-off loss

Abstract

The invention discloses a junction temperature prediction method based on SiC MOSFET turn-off loss, and belongs to the technical field of power electronics and electricians. The junction temperature prediction method combines a turn-off loss model with a junction temperature prediction model based on turn-off loss, and comprises the following steps: firstly, establishing a SiC MOSFET turn-off loss model considering parasitic parameters, secondly, establishing the junction temperature prediction model based on the SiC MOSFET turn-off loss in an off-line manner, and then, calculating the turn-off loss on line according to the loss model. And finally substituting the loss data into the junction temperature prediction model to obtain a junction temperature prediction value. According to the method, the turn-off loss and the junction temperature of the SiC MOSFET under different working conditions can be predicted, and the method has the advantages of being high in real-time performance, easy to implement, high in model precision and the like; and an efficient and accurate solution is provided for thermal evaluation and junction temperature prediction of the silicon carbide power device.

Description

Technical Field

[0001] The present invention relates to the fields of power electronics technology and electrical engineering technology, and particularly to a method for predicting the junction temperature of a SiC MOSFET. Background Art

[0002] Among the faults in power electronic systems, about 31% are caused by power device faults, and about 60% of the power device faults are related to temperature; although SiC materials have better thermal properties, the limitations of traditional package heat dissipation and the relatively high thermal conductivity of SiC materials will force SiC MOSFETs to bear greater thermal stress during operation. Excessive thermal stress may cause the bonding wires inside the SiC MOSFET to fall off and the solder layer to rupture, resulting in the failure and damage of the entire module. The junction temperature is the main characterization quantity of the thermal performance of SiC MOSFETs, and junction temperature prediction is a prerequisite for the thermal management, condition detection, and life prediction of SiC MOSFETs. Therefore, it is very important to study an accurate estimation method for the junction temperature of SiC MOSFETs. Precise prediction and extraction of the junction temperature are technical problems that need to be overcome in the application of SiC MOSFET devices.

[0003] Junction temperature measurement methods can generally be divided into four categories: physical contact method, optical method, thermal network method, and temperature-sensitive electrical parameter method. Among them, although the physical contact method and the optical method are simple and easy to implement, they both require destroying or changing the package structure to leave a measurement channel, and the invasiveness is very strong. Moreover, in terms of the response speed of junction temperature measurement, the physical contact method is limited by the heat capacity of the contact layer, the optical method requires corresponding optical processing procedures, and the thermal network method requires relatively complex model calculations. In contrast, the temperature-sensitive electrical parameter method uses the device itself as a temperature sensor to establish the relationship between the internal electrical parameters and the chip junction temperature, and obtains the junction temperature by measuring the electrical parameters. It has the advantages of simple principle, rapid response, on-line measurement, and non-destructive device packaging, and is considered the most promising junction temperature measurement method.

[0004] When using the temperature-sensitive electrical parameter method to predict the junction temperature of SiC MOSFETs, there are mainly problems such as some electrical parameter values being small, the temperature sensitivity being low, and the electrical parameters being easily affected by circuit parasitic parameters, resulting in errors in junction temperature estimation. For SiC MOSFETs, it is crucial to select appropriate temperature-sensitive electrical parameters for accurate junction temperature detection.

[0005] In view of the problems that the current junction temperature prediction method based on turn-off loss is difficult to achieve on-line detection and the model accuracy is low, the present patent establishes a turn-off loss model of SiC MOSFET considering non-linear parasitic parameters, and proposes a junction temperature prediction method based on the turn-off loss of SiC MOSFET. Summary of the Invention

[0006] In view of the deficiencies of the prior art, the present invention proposes a method that combines a turn-off loss model with a junction temperature prediction model based on turn-off loss, effectively solving the problems that it is difficult to achieve online detection and the model accuracy is low in the current junction temperature prediction method based on turn-off loss.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for predicting the junction temperature based on the turn-off loss of a SiC MOSFET, comprising the following steps:

[0009] Step S1: Establish an equivalent circuit model of a SiC MOSFET considering parasitic parameters, and obtain the state equations of each stage of the turn-off process through the analysis and modeling of the turn-off transient process, and then establish a turn-off loss model of the SiC MOSFET considering parasitic parameters;

[0010] Step S2: Perform an offline calibration program to obtain the mapping relationship between the turn-off loss of the selected SiC MOSFET and the known junction temperature in an offline manner, and establish a junction temperature prediction model based on the turn-off loss of the SiC MOSFET;

[0011] Step S3: Perform numerical solution. Input the initial values of each state variable in the calculation tool, and then obtain the values of each state variable after completing the solution within the solution period. Update the parameters in the state equation according to these calculated values, and repeat this process until the relevant boundary conditions are met, then enter the next state equation, and save the data calculated each time, so that the turn-off loss can be obtained through numerical calculation;

[0012] Step S4: Simulate different operating conditions of the power device by adjusting the initial values of the state variables in Step S3, substitute the calculation results of the turn-off loss of the SiC MOSFET into the junction temperature prediction model based on the turn-off loss of the SiC MOSFET established in Step S2, and calculate the junction temperature value according to the mapping relationship between the turn-off loss, the junction temperature and the operating conditions.

[0013] Further, in Step S1, when establishing the equivalent circuit model of the SiC MOSFET considering parasitic parameters, the state equations of each stage of the turn-off process obtained through the analysis and modeling of the turn-off transient process are respectively: the state equation set A1 in the turn-off delay stage, the state equation set A2 in the voltage rise stage, the state equation set A3 in the current decline stage, and the state equation set A4 in the turn-off oscillation stage;

[0014] Among them, the state equation set A1 in the turn-off delay stage is as follows:

[0015]

[0016] The state equation set A2 in the voltage rising stage is as follows:

[0017]

[0018] The state equation set A3 in the current falling stage is as follows:

[0019]

[0020] The state equation set A4 in the turn-off oscillation stage is as follows:

[0021]

[0022] Among them, v GS is the gate-source voltage of the SiC MOSFET, v DS is the drain-source voltage, i D is the drain current, V DRV is the driving positive voltage, V DC is the ideal voltage source equivalent to the DC bus, V TH is the threshold voltage, g fs is the transconductance; I L is the ideal current source equivalent to the load inductance, C J is the equivalent junction capacitance of the SiC SBD; C ISS , C OSS , C RSS are the input capacitance, output capacitance, and transfer capacitance of the SiC MOSFET respectively; R G is the gate driving resistance of the SiC MOSFET, L G , L D , L S are the gate, drain, and source parasitic inductances of the SiC MOSFET respectively; L loop is the stray inductance, R loop is the stray resistance;

[0023] Through the turn-off delay stage state equation set A1, voltage rising stage state equation set A2, current falling stage state equation set A3, and turn-off oscillation stage state equation set A4, a turn-off loss model of the SiC MOSFET considering parasitic parameters is established:

[0024]

[0025] Among them, E off is the turn-off loss, R G is the gate driving resistance of the SiC MOSFET, C GD is the gate-drain parasitic capacitance, I L is the inductor current, V DC is the bus voltage, V DS(ON) is the conduction voltage drop, VDRV For the driving positive voltage, V TH For the threshold voltage, C ISS For the input capacitance of the SiCMOSFET, g fs For the transconductance, L S For the source parasitic inductance of the SiC MOSFET. Further, the process of establishing the junction temperature prediction model based on the turn-off loss of the SiC MOSFET in step S2 is as follows:

[0026] Step S21: By building a double-pulse test platform, set the junction temperature T of the SiC MOSFET j and the operating conditions: the bus voltage V DC , the inductor current I L ;

[0027] Step S22: Test the turn-off loss values of the SiC MOSFET under different junction temperatures and operating conditions, and draw a relationship diagram between the turn-off loss and the junction temperature T j and the operating conditions;

[0028] Step S23: Decouple and fit the relationship between the turn-off loss and the junction temperature T j and the operating conditions to obtain the mapping relationship between the turn-off loss and the junction temperature T j and the operating conditions, which is the junction temperature prediction model based on the turn-off loss of the SiC MOSFET:

[0029] T j =αE off +βI L +δ

[0030] where α, β, δ are the constant values obtained by fitting, E off is the turn-off loss, and I L is the inductor current value.

[0031] Further, in step S3, according to the state equations of each stage of the turn-off process, input the initial values of each state variable into the calculation tool and set the time step:

[0032] Step S31: First, calculate sub-period 1, solve the state equation set A1 of the turn-off delay stage, and judge whether the boundary condition v GS <V p is satisfied after the end of a single time step, where v GS is the gate-source voltage, V p is the Miller plateau voltage, and define the duration of this stage as t 1 -t 0 , if this boundary condition is satisfied, store the drain-source voltage, gate-source voltage, drain current, and duration parameters calculated during this time period, and start calculating sub-period 2, if not satisfied, run a single time step again;

[0033] Step S32: After satisfying the boundary conditions of sub-cycle 1, start calculating sub-cycle 2, solve the state equation set A2 in the voltage rising stage, and define the duration of this stage as t 2 -t 1 , after satisfying the boundary condition v D <-V F , store the solution data of this stage and start calculating sub-cycle 3;

[0034] Step S33: After satisfying the boundary conditions of sub-cycle 2, start calculating sub-cycle 3, solve the state equation set A3 in the current falling stage, and define the duration of this stage as t 3 -t 2 , after satisfying the boundary condition v GS <V TH , store the solution data of this stage and start calculating sub-cycle 4;

[0035] Step S34: After satisfying the boundary conditions of sub-cycle 3, start calculating sub-cycle 4, solve the state equation set A4 in the turn-off oscillation stage, and define the duration of sub-cycle 4 as t 4 -t 3 , store the solution data of this stage after running a single time step;

[0036] Step S35: Once the data calculated each time is saved, the turn-off loss can be obtained through numerical calculation. The formula for calculating the turn-off loss by the integration method is:

[0037]

[0038] Further, in step S4, different operating conditions of the power device are simulated by adjusting the initial values of the state variables in step S3, and the data and turn-off loss of each stage are obtained through a calculation tool;

[0039] Substitute the calculation result of the turn-off loss of the SiC MOSFET into the junction temperature prediction model based on the turn-off loss of the SiC MOSFET established in step S2, and calculate the junction temperature value according to the mapping relationship between the turn-off loss and the junction temperature and the operating conditions.

[0040] Further, the state variables in step S3 include drain-source voltage, drain current, and duration.

[0041] Compared with the prior art, the present invention adopts the above technical solutions and has the following technical effects:

[0042] (1) The present invention proposes a method that combines a turn-off loss model with a junction temperature prediction model based on turn-off loss, and accurately predicts the junction temperature of SiC MOSFETs through numerical calculations. An equivalent circuit model of SiC MOSFET considering parasitic parameters is established, and the influence of non-linear parasitic parameters is fully considered when establishing the state equations of each stage during the analysis of the turn-off process, so as to closely approximate the actual circuit and improve the prediction accuracy of the model;

[0043] (2) Aiming at the problem that it is difficult to achieve on-line detection of the turn-off loss of SiC MOSFETs at present, this method does not require complex steps for detecting and extracting the drain-source voltage and drain current. The accurate and rapid calculation of the turn-off loss can be achieved by substituting the circuit operating conditions into the calculation tool. Description of the Drawings

[0044] Figure 1 is a framework diagram of a junction temperature prediction method based on the turn-off loss of SiC MOSFETs in the present invention;

[0045] Figure 2 is a double-pulse equivalent circuit diagram of SiC MOSFET considering non-linear parasitic parameters in the present invention;

[0046] Figure 3 is a flow chart for establishing a junction temperature prediction model based on the turn-off loss of SiC MOSFETs in the present invention;

[0047] Figure 4 is a flow chart for solving the state equation group during the turn-off process in the present invention;

[0048] Figure 5 is an overall framework diagram of the junction temperature prediction method based on the turn-off loss model of SiC MOSFETs in the present invention;

[0049] Figure 6 is a double-pulse simulation result diagram of the turn-off loss varying with the junction temperature in the present invention; Detailed Embodiments

[0050] The following details some key technologies involved in the present invention with reference to the drawings to support the claims; to describe the purpose, technical solutions and advantages of the present invention more clearly, the present invention will be further described in detail below with reference to the drawings and embodiments.

[0051] Figure 1 is a framework diagram of a junction temperature prediction method based on the turn-off loss of SiC MOSFETs, including the following steps:

[0052] Step S1: Establish an equivalent circuit model of SiC MOSFET considering parasitic parameters. By analyzing and modeling the turn-off transient process, obtain the state equations for each stage of the turn-off process, and then establish a turn-off loss model of SiC MOSFET considering parasitic parameters;

[0053] Step S2: Perform an offline calibration procedure to obtain the mapping relationship between the turn-off loss of the selected SiC MOSFET and the known junction temperature in an offline manner, and establish a junction temperature prediction model based on the turn-off loss of SiC MOSFET;

[0054] Step S3: Perform numerical solution. Input the initial values of each state variable into the calculation tool, and then obtain the values of each state variable after completing the solution within the solution period. Update the parameters in the state equation according to these calculated values, and repeat this process until the relevant boundary conditions are met, then enter the next state equation. The data calculated each time is saved, and the turn-off loss can be obtained through numerical calculation;

[0055] Step S4: Simulate different operating conditions of the power device by adjusting the initial values of the state variables in Step S3. Substitute the calculation results of the turn-off loss of SiC MOSFET into the junction temperature prediction model based on the turn-off loss of SiC MOSFET established in Step S2, and calculate the junction temperature value according to the mapping relationship between the turn-off loss, junction temperature, and operating conditions.

[0056] Figure 2 is the double-pulse equivalent circuit diagram of SiC MOSFET considering nonlinear parasitic parameters. Among them, V DC is an ideal voltage source equivalent to the DC bus; I L is an ideal current source equivalent to the load inductor, C L is the parasitic capacitance of the load inductor; D H is an ideal SiC SBD, C J is the equivalent junction capacitance of SiC SBD; C GS 、C GD 、C DS are the gate-source capacitance, gate-drain capacitance, and drain-source capacitance of SiC MOSFET respectively; L D(int) 、L S(int) are the drain and source parasitic inductances introduced by the SiC MOSFET package respectively; R G(int) is the gate internal resistance of SiC MOSFET, R G(ext) is the external drive resistance; L G is the parasitic inductance of the gate circuit; L D(ext) and R loop are the equivalent parasitic inductance and stray resistance of the line between the positive terminal of the DC bus and the drain of SiC MOSFET respectively; L S(ext)is the parasitic inductance of the circuit between the source and ground of the SiC MOSFET. i D is the drain current, i G is the gate current, i GS 、i GD and i DS are the displacement currents of C GS 、C GD and C DS respectively, i CH is the channel current of the SiC MOSFET, I L is the load current.

[0057] Furthermore, in step S1, an equivalent circuit model of the SiC MOSFET considering parasitic parameters is established, and the state equations for each stage of the turn-off transient process are obtained through analysis and modeling: the state equation set A1 for the turn-off delay stage, the state equation set A2 for the voltage rise stage, the state equation set A3 for the current fall stage, and the state equation set A4 for the turn-off oscillation stage;

[0058] State equation set A1 for the turn-off delay stage:

[0059]

[0060] State equation set A2 for the voltage rise stage:

[0061]

[0062] State equation set A3 for the current fall stage:

[0063]

[0064] State equation set A4 for the turn-off oscillation stage:

[0065]

[0066] Through the analysis and modeling of the turn-off transient process, the mathematical expressions and state equations for each stage of the turn-off process are obtained, and then a turn-off loss model of the SiC MOSFET considering parasitic parameters is established:

[0067]

[0068] Among them, E off is the turn-off loss, R G is the gate drive resistance of the SiC MOSFET, C GD is the gate-drain parasitic capacitance, I L is the inductor current, V DC is the bus voltage, V DS(ON) is the conduction voltage drop, V DRV is the driving positive voltage, V THis the threshold voltage, C ISS is the input capacitance of the SiCMOSFET, g fs is the transconductance, L S is the source parasitic inductance of the SiC MOSFET.

[0069] Furthermore, as Figure 3 shown, the process of establishing the junction temperature prediction model based on the turn-off loss of the SiC MOSFET in step S2 is as follows:

[0070] Step S21: By building a double-pulse test platform, set the junction temperature T j of the SiC MOSFET and the working conditions;

[0071] Step S22: Test the turn-off loss values of the SiC MOSFET at different junction temperatures and working conditions, and draw a relationship diagram of the turn-off loss versus the junction temperature T j and the working conditions;

[0072] Step S23: Decouple and fit the relationship between the turn-off loss and the junction temperature T j and the working conditions to obtain the mapping relationship formula between the turn-off loss and the junction temperature T j and the working conditions, which is the junction temperature prediction model based on the turn-off loss of the SiC MOSFET:

[0073] T j = αE off + βI L + δ

[0074] where α, β, and δ are constant values obtained by fitting, E off is the turn-off loss, and I L is the inductor current value.

[0075] Furthermore, Figure 4 is the flow chart for solving the state equations of the turn-off process. In step S3, according to the state equations of each stage of the turn-off process, input the initial values of each state variable in the calculation tool and set the time step:

[0076] Step S31: First, calculate sub-period 1, solve the state equations A1 of the turn-off delay stage, and judge whether the boundary condition v GS < V p is satisfied after a single time step, where v GS is the gate-source voltage and V p is the Miller plateau voltage. Define the duration of this stage as t 1 - t 0, if the boundary condition is satisfied, store the drain-source voltage, gate-source voltage, drain current, and duration parameters calculated for this time period, and start calculating sub-cycle 2; if not, run a single time step again;

[0077] Step S32: After satisfying the boundary condition of sub-cycle 1, start calculating sub-cycle 2, solve the state equation set A2 in the voltage rising stage, and define the duration of this stage as t 2 -t 1 , when the boundary condition v D <-V F is satisfied, store the solution data of this stage and start calculating sub-cycle 3;

[0078] Step S33: After satisfying the boundary condition of sub-cycle 2, start calculating sub-cycle 3, solve the state equation set A3 in the current falling stage, and define the duration of this stage as t 3 -t 2 , when the boundary condition v GS <V TH is satisfied, store the solution data of this stage and start calculating sub-cycle 4;

[0079] Step S34: After satisfying the boundary condition of sub-cycle 3, start calculating sub-cycle 4, solve the state equation set A4 in the turn-off oscillation stage, and define the duration of sub-cycle 4 as t 4 -t 3 , store the solution data of this stage after running a single time step;

[0080] Step S35: Once the data calculated each time is saved, the turn-off loss can be obtained through numerical calculation. The formula for calculating the turn-off loss by the integration method is:

[0081]

[0082] Figure 5 is the overall framework diagram of the junction temperature prediction method based on the turn-off loss of SiC MOSFET. The corresponding relationship between the turn-off loss of the selected SiC MOSFET and the known junction temperature is obtained offline, and this offline curve of the junction temperature vs. SiC MOSFET turn-off loss is used as a reference in the next temperature measurement program; adjust the initial values of the state variables to simulate different operating conditions of the power device, and obtain the data and turn-off loss of each stage through a calculation tool; substitute the calculated result of the SiC MOSFET turn-off loss into the junction temperature prediction model based on the turn-off loss, and calculate the junction temperature value according to the mapping relationship between the turn-off loss, junction temperature, and operating conditions.

[0083] Simulate the SiC MOSFET double-pulse circuit considering non-linear parasitic parameters, Figure 6It is a double-pulse simulation result diagram of the turn-off loss varying with the junction temperature; the simulation results show that as the junction temperature increases, the turn-off loss value of the SiC MOSFET gradually increases and has good linearity. Selecting the turn-off loss as the temperature-sensitive electrical parameter can well reflect the change of the junction temperature. By decoupling and fitting the relationship diagram between the turn-off loss, the junction temperature T j and the working conditions, the mapping relationship formula between the turn-off loss and the junction temperature T j and the working conditions is obtained, which is the junction temperature prediction model based on the turn-off loss of the SiC MOSFET; taking this junction temperature prediction model as a reference in the next temperature measurement program, and then obtaining the data and turn-off loss at each stage by solving the state equations; substituting the calculated result of the turn-off loss of the SiC MOSFET into the junction temperature prediction model based on the turn-off loss, and calculating the junction temperature value according to the mapping relationship between the turn-off loss, the junction temperature and the working conditions.

[0084] The above embodiments are only used to illustrate the technical idea of the present invention, and the protection scope of the present invention cannot be limited thereby. Any changes made on the basis of the technical solution according to the technical idea proposed by the present invention shall fall within the protection scope of the present invention.

Claims

1. A junction temperature prediction method based on SiC MOSFET turn-off loss, characterized in that: The following steps are involved: Step S1: Establishing an equivalent circuit model of SiC MOSFET considering parasitic parameters, obtaining state equations of each stage of the shutdown process by analyzing and modeling the shutdown transient process, and then establishing a SiC MOSFET shutdown loss model considering parasitic parameters; Step S2: performing an offline calibration procedure, obtaining a mapping relationship between the selected SiC MOSFET turn-off loss and the known junction temperature in an offline manner, and establishing a junction temperature prediction model based on the SiC MOSFET turn-off loss; Step S3: perform numerical solution, input the initial value of each state variable in the calculation tool, and then obtain the value of each state variable after completing the solution within the solution cycle, update the parameters in the state equation according to these calculated values, and repeat this cycle until the relevant boundary conditions are met, and then enter the next state equation. The data calculated each time is saved, and the turn-off loss can be obtained by numerical calculation; Step S4: By adjusting the initial value of the state variable in step S3 to simulate different operating conditions of the power device, the calculation result of the SiCMOSFET turn-off loss is substituted into the junction temperature prediction model based on the SiC MOSFET turn-off loss established in step S2, and the junction temperature value is calculated according to the mapping relationship between the turn-off loss, the junction temperature and the operating condition.

2. The method for predicting junction temperature based on SiC MOSFET turn-off loss according to claim 1, characterized in that: The step S1 establishes a SiC MOSFET equivalent circuit model taking into account parasitic parameters, and analyzes and models the shutdown transient process to obtain state equations for each stage of the shutdown process: shutdown delay stage state equation group A1, voltage rise stage state equation group A2, current drop stage state equation group A3, and shutdown oscillation stage state equation group A4; Among them, the state equation group A1 of the shutdown delay stage is as follows: The state equation group A2 in the voltage rising stage is as follows: The state equation group A3 in the current drop stage is as follows: The state equation group A4 in the shutdown oscillation stage is as follows: Among them, v GS is the SiC MOSFET gate-source voltage, v DS is the drain-source voltage, i D is the drain current, V DRV For positive driving pressure, V DC is the ideal voltage source equivalent to the DC bus, V TH is the threshold voltage, g fs is the transconductance; I L C is an ideal current source equivalent to the load inductance. J is the SiC SBD equivalent junction capacitance; C ISS , C OSS , C RSS They are the input capacitance, output capacitance, and transfer capacitance of SiC MOSFET respectively; R G is the gate drive resistance of SiC MOSFET, L G , L D , L S They are the parasitic inductances of the gate, drain, and source of SiC MOSFET respectively; L loop is the stray inductance, R loop is the stray resistance; Through the turn-off delay stage state equation group A1, the voltage rise stage state equation group A2, the current drop stage state equation group A3, and the turn-off oscillation stage state equation group A4, the SiCMOSFET turn-off loss model considering parasitic parameters can be established by the integration method: Among them, E off is the turn-off loss, R G is the gate drive resistance of SiC MOSFET, C GD is the gate-drain parasitic capacitance, I L is the inductor current, V DC is the bus voltage, V DS(ON) is the conduction voltage drop, V DRV For positive driving pressure, V TH is the threshold voltage, C ISS is the input capacitance of SiCMOSFET, g fs is the transconductance, L S is the SiC MOSFET source parasitic inductance.

3. The method for predicting junction temperature based on SiC MOSFET turn-off loss according to claim 1, characterized in that: The process of establishing the junction temperature prediction model based on the SiC MOSFET turn-off loss in step S2 is as follows: Step S21: Setting the junction temperature T of SiC MOSFET by building a double pulse test platform j And working conditions: bus voltage V DC , inductor current I L ; Step S22: Test the turn-off loss value of SiC MOSFET under different junction temperatures and working conditions, and plot the turn-off loss and junction temperature T j and working conditions; Step S23: Compare the turn-off loss and the junction temperature T j The relationship between the turn-off loss and the junction temperature T is decoupled and fitted. j The mapping relationship between the operating conditions and the junction temperature prediction model based on the SiC MOSFET turn-off loss is: T j =αE off +βI L +d Among them, α, β, δ are the constant values ​​obtained by fitting, E off is the turn-off loss, I L is the inductor current value.

4. The method for predicting junction temperature based on SiC MOSFET turn-off loss according to claim 1, characterized in that: In step S3, according to the state equations of each stage of the shutdown process, the initial values ​​of each state variable are input into the calculation tool, and the time step is set: Step S31: First, calculate sub-cycle 1, solve the shutdown delay phase state equation group A1, and determine whether the boundary condition v is satisfied after a single time step ends. GS <V p , where v GS is the gate-source voltage, V p is the Miller platform voltage, and the duration of this stage is defined as t1-t0. If the boundary condition is met, the drain-source voltage, gate-source voltage, drain current, and duration parameters calculated in this time period are stored, and sub-cycle 2 is calculated. If not, a single time step is run again; Step S32: After the boundary conditions of sub-period 1 are met, sub-period 2 is calculated, and the state equation group A2 of the voltage rising stage is solved. The duration of this stage is defined as t2-t1. When the boundary conditions v are met, D <-V F Then store the solution data of this stage and start calculating sub-cycle 3; Step S33: After the boundary conditions of sub-period 2 are met, sub-period 3 is calculated, and the state equation group A3 of the current drop phase is solved. The duration of this phase is defined as t3-t2. When the boundary conditions v are met, GS <V TH Then store the solution data of this stage and start calculating sub-cycle 4; Step S34, after the boundary conditions of sub-period 3 are met, sub-period 4 is calculated, the state equation group A4 of the shutdown oscillation stage is solved, the duration of sub-period 4 is defined as t4-t3, and the solution data of this stage is stored after running a single time step; Step S35, each calculated data is saved, and the turn-off loss can be obtained by numerical calculation. The formula for calculating the turn-off loss by the integral method is:

5. The method for predicting junction temperature based on SiC MOSFET turn-off loss according to claim 1, characterized in that: The step S4 simulates different operating conditions of the power device by adjusting the initial value of the state variable in step S3, and obtains the data and turn-off loss of each stage by using a calculation tool; The calculation result of the SiC MOSFET turn-off loss is substituted into the junction temperature prediction model based on the SiC MOSFET turn-off loss established in step S2, and the junction temperature value is calculated according to the mapping relationship between the turn-off loss, the junction temperature and the operating conditions.

6. The method for predicting junction temperature based on SiC MOSFET turn-off loss according to claim 1, characterized in that: The state variables in step S3 include drain-source voltage, gate-source voltage, drain current, and duration.

Images