WO2020108172A1 - 一种功率模块温度估算方法 - Google Patents

一种功率模块温度估算方法 Download PDF

Info

Publication number
WO2020108172A1
WO2020108172A1 PCT/CN2019/112503 CN2019112503W WO2020108172A1 WO 2020108172 A1 WO2020108172 A1 WO 2020108172A1 CN 2019112503 W CN2019112503 W CN 2019112503W WO 2020108172 A1 WO2020108172 A1 WO 2020108172A1
Authority
WO
WIPO (PCT)
Prior art keywords
loss
specified
temperature
turn
igbt
Prior art date
Application number
PCT/CN2019/112503
Other languages
English (en)
French (fr)
Inventor
蔡磊
彭再武
杨洪波
陈慧民
石高峰
凌岳伦
Original Assignee
中车时代电动汽车股份有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 中车时代电动汽车股份有限公司 filed Critical 中车时代电动汽车股份有限公司
Publication of WO2020108172A1 publication Critical patent/WO2020108172A1/zh

Links

Images

Classifications

    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/15—Correlation function computation including computation of convolution operations

Definitions

  • the invention relates to the technical field of power electronics, in particular to a power module temperature estimation method.
  • NTC NTC temperature sensor inside the IGBT module for over-temperature protection
  • NTC has a slow response time and cannot reflect the transient fluctuations in junction temperature in a timely manner, and under extreme operating conditions such as locked rotor, the heat energy is unevenly distributed, which can easily cause the IGBT to overheat and permanently fail.
  • the technical solution has limitations.
  • the present invention provides a power module temperature estimation method.
  • the method includes:
  • Step 1 Obtain the loss of the power module to be analyzed, and determine the input heat of the first layer in the network model of the power module to be analyzed according to the loss;
  • Step 2 Determine the single layer temperature difference of each layer in the network model according to the input heat of the first layer
  • Step 3 Determine the junction temperature of the power module to be analyzed according to the temperature difference between the single layer of each layer and the ambient temperature.
  • the power module to be analyzed includes an IGBT module.
  • current data of the IGBT module is obtained, and the current data is based on a preset loss power-current relationship function Determine the loss of the IGBT module.
  • the power module to be analyzed includes an IGBT module.
  • the IGBT on-state loss, the IGBT on-state loss, the IGBT off-state loss, and the FWD on-state loss of the IGBT module are respectively obtained , FWD turn-off loss and IGBT module line loss, and determine the loss of the IGBT module according to the above loss.
  • a loss curve corresponding to each loss component is fitted by fitting, and according to each loss curve Determine each loss component.
  • the step of determining the loss curve corresponding to the IGBT on-state loss power of the IGBT module to be analyzed includes:
  • the loss curve corresponding to the on-state loss power of the IGBT is determined according to the functional relationship between the on-state loss power and the on-state current at the specified junction temperature and the specified GE voltage and a preset temperature-on-state loss coefficient.
  • the loss curve corresponding to the on-state loss power includes:
  • P I represents the IGBT on-state power loss
  • temp represents the junction temperature obtained during the previous round of estimation
  • temp′ represents the specified junction temperature
  • I c and I max represent the on-state current and the limit current, respectively.
  • P I_temp′ represents the IGBT on-state loss power corresponding to the specified junction temperature
  • k represents a constant.
  • the step of determining the loss curve corresponding to the IGBT turn-on loss of the IGBT module includes:
  • the function of the specified junction temperature, the specified gate resistance, the specified GE voltage and the specified CE voltage and the turn-on loss and the on-state current are fitted, and the specified temperature, The functional relationship between the turn-on loss and the gate resistance at the specified GE voltage, the specified CE voltage, and the specified on-state current;
  • the loss curve corresponding to the turn-on loss of the IGBT is determined according to the functional relationship between the turn-on loss and the on-state current and the gate resistance-turn-on loss coefficient.
  • the loss curve corresponding to the IGBT turn-on loss includes:
  • E on represents the IGBT turn- on loss
  • temp represents the junction temperature obtained during the previous round of estimation
  • temp′ represents the specified junction temperature
  • I c and I max represent the on-state current and the limit current
  • E on_temp′ represents the specified junction IGBT turn-on loss corresponding to temperature
  • R g represents the gate resistance.
  • the step of determining the loss curve corresponding to the IGBT turn-off loss power of the IGBT module includes:
  • the loss curve corresponding to the IGBT turn-off loss is determined according to the functional relationship between the turn-off loss and the gate resistance and the gate resistance-turn-off loss coefficient.
  • the loss curve corresponding to the IGBT turn-off loss includes:
  • E off represents the IGBT turn-off loss
  • temp represents the junction temperature obtained during the previous round of estimation
  • temp′ represents the specified junction temperature
  • I c and I max represent the on-state current and the limit current respectively
  • E off_temp′ represents the reference IGBT turn-off loss corresponding to temperature
  • R g represents the gate resistance.
  • the step of determining the loss curve corresponding to the FWD on-state loss of the IGBT module includes:
  • the function relationship between the forward current and the forward voltage of the freewheeling diode at the first specified temperature, the second specified temperature and the third specified temperature are fitted respectively.
  • the third specified temperature, the first specified temperature and the second specified temperature increase sequentially;
  • the functional relationship between the forward current and the forward voltage the functional relationship between the on-state loss of the freewheeling diode and the forward current at the first specified temperature, the second specified temperature and the third specified temperature are determined respectively ;
  • the loss curve corresponding to the FWD on-state loss is determined.
  • the loss curve corresponding to the FWD on-state loss is determined according to the following expression:
  • the loss curve corresponding to the FWD on-state loss is determined according to the following expression:
  • P D represents the FWD on-state loss
  • temp represents the junction temperature obtained during the previous round of estimation
  • P D_temp1 , P D_temp2 and P D_temp3 represent the first specified temperature temp1, the second specified temperature temp2 and the third specified temperature respectively The FWD on-state loss corresponding to temp3.
  • the step of determining the loss curve corresponding to the FWD turn-off loss of the IGBT module includes:
  • the function relationship between the turn-off loss and the gate resistance at the first specified temperature, the specified CE voltage and the specified forward current is fitted to obtain the gate resistance-reverse recovery coefficient
  • the loss curve corresponding to the FWD turn-off loss is determined according to the following expression:
  • E D represents FWD turn-off loss
  • temp represents the junction temperature obtained during the previous round of estimation
  • temp1 and temp2 represent the first specified temperature and second specified temperature
  • I f represents the forward current
  • E D_temp1 and E D_temp2 represents the turn-off loss of the freewheeling diode corresponding to the first specified temperature and the second specified temperature
  • R g represents the gate resistance.
  • the gate resistance-reverse recovery coefficient is determined according to the following expression
  • f(R g ) represents the functional relationship between the turn-off loss and the gate resistance at the first specified temperature, the specified CE voltage, and the specified forward current.
  • the current single-layer temperature difference of each layer is determined according to the acquired temperature difference of each layer in the network model and the input heat of the first layer.
  • the current single layer temperature difference of the layer is determined according to the following expression:
  • r m represents the thermal resistance of the mth layer
  • ⁇ m represents the thermal time constant of the mth layer
  • T represents the preset time interval
  • the output heat of any layer in the network model is equal to the input heat of the next layer, wherein the output heat of any layer is determined according to the following expression:
  • step three the junction temperature of the power module to be analyzed is determined according to the following expression:
  • T j represents the junction temperature of the power module to be analyzed
  • T a represents the ambient temperature
  • N represents the number of layers included in the network model
  • ⁇ temp m represents the temperature difference of the single layer of the mth layer in the network model.
  • the power module temperature estimation method provided by the present invention estimates the junction temperature of the power module through an iterative calculation, which is convenient to be implemented in a microprocessor.
  • the method introduces parameters such as temperature and gate resistance when calculating the loss of the power module, which can effectively improve the accuracy of the loss calculation, thereby helping to improve the accuracy of the junction temperature estimation, and thus allowing the motor controller to be protected On the premise of maximizing the output capacity.
  • FIG. 1 is a schematic diagram of an implementation process of a power module temperature estimation method according to an embodiment of the present invention
  • FIG. 2 is a schematic diagram of a Foster network model according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of an implementation process of determining a single layer temperature difference according to an embodiment of the present invention
  • FIG. 4 is a schematic diagram of a single layer temperature difference calculation principle according to an embodiment of the present invention.
  • FIG. 5 is a schematic diagram of an implementation process of determining the IGBT on-state loss power with the corresponding loss curve according to an embodiment of the present invention
  • FIG. 8 is a schematic diagram of an implementation process of determining a loss curve corresponding to an IGBT turn-on loss according to an embodiment of the present invention
  • FIG. 10 is a schematic diagram of an implementation process of determining a loss curve corresponding to an IGBT turn-off loss according to an embodiment of the present invention
  • 11 is a turn-off loss error curve with a gate resistance R g of 0.51 ⁇ and 150°C according to an embodiment of the present invention
  • FIG. 12 is a schematic diagram of an implementation process of determining the FWD on-state loss of an IGBT module according to an embodiment of the present invention
  • 13 and 14 are the curves of the FWD on-state power loss of the IGBT module when the junction temperature obtained during the previous round of estimation according to an embodiment of the present invention is 65°C
  • FIG. 15 is a schematic diagram of an implementation process of determining the FWD turn-off loss of an IGBT module according to an embodiment of the present invention.
  • 16 is an FWD turn-off loss curve when the gate resistance R g is 0.51 ⁇ and the temperature is 135°C according to an embodiment of the present invention.
  • the instantaneous junction temperature of the IGBT module can be monitored in real time, and overtemperature protection can be performed even under the extreme conditions of locked rotor;
  • the margin of overtemperature protection can be reduced and the output capacity of the IGBT module can be effectively improved , So as to make the vehicle more dynamic;
  • junction temperature calculation scheme for junction temperature.
  • This method discusses the calculation method of the junction temperature of the IGBT module in one switching period based on the electrothermal analogy theory, and calculates the power frequency period junction temperature of the IGBT module through an iterative algorithm.
  • the junction temperature calculation scheme provided by this method has some shortcomings. For example, this method does not introduce junction temperature parameters during the calculation of the IGBT heat loss, which also affects the accuracy of the junction temperature calculation.
  • junction temperature of the IGBT is directly related to its loss, and the junction temperature in turn affects the loss, that is, the loss of the IGBT under the same electrical parameters and different junction temperatures is different, so the loss must be related to the current junction temperature of the IGBT.
  • complex matrix operations in the calculation process of this method which is not suitable for general motor controller processing chips.
  • this method ignores the consideration of heat sink thermal parameters when calculating the junction temperature of the IGBT, and also affects the accuracy of the junction temperature calculation.
  • IGBT junction temperature detection device and method there is also an IGBT junction temperature detection device and method in the prior art.
  • This method first detects the switching frequency of the IGBT and the current when the IGBT is turned on, then calculates the temperature rise of the IGBT junction temperature according to a predetermined model, and finally the junction of the IGBT The temperature rise plus the temperature of the IGBT heat sink is taken as the junction temperature of the IGBT.
  • the current detection unit, frequency detection unit, isolation unit, etc. bring additional hardware cost overhead and may introduce detection errors.
  • the loss calculation model used in this method is a simple linear calculation formula, which is difficult to match the IGBT loss nonlinear curve, resulting in inaccurate loss calculation.
  • this method does not introduce junction temperature parameters for the calculation of IGBT heat loss, which affects the accuracy of junction temperature calculation.
  • this method simplifies the calculation of the junction temperature into a two-stage calculation, that is, the junction temperature is equal to the temperature of the heat sink plus the temperature rise of the IGBT.
  • the device selects the voltage VCE of the collector and the emitter under a small current as the temperature sensitive parameter, which is tested by a pre-heating box Determine the functional relationship between the saturation voltage drop V CE and the junction temperature, and then control the large current turn-off by switching the IGBT module under power cycle conditions, monitor the change of the voltage drop V CE at the moment of IGBT turn-off, and finally substitute The functional relationship gets the junction temperature. Since this method requires the use of a small current detection unit, etc., it brings additional hardware cost, and additional isolation circuits need to be added to the unit. At the same time, this method is more time-consuming due to the need for a pre-warm chamber test.
  • the present invention provides a new power module temperature estimation method, which is particularly suitable for estimating the junction temperature of the motor controller power module.
  • FIG. 1 shows a schematic diagram of an implementation process of the power module temperature estimation method provided in this embodiment.
  • the power module temperature estimation method provided in this embodiment first obtains the power loss of the power module to be analyzed in step S101, and obtains the network model of the power module to be analyzed according to the determined power loss Input heat in the first floor.
  • this method equates the power module to the Foster network model shown in FIG. 2.
  • the heat is generated by the IGBT chip and the FWD chip, and then flows through multiple material layers and is taken away by the heat sink. Since the IGBT chip and the FWD chip share the same substrate, in this embodiment, the IGBT chip and the FWD chip are considered to have the same case temperature T c .
  • the Foster network model there is preferably a solder layer, a copper foil layer, a ceramic layer, etc. between the chip and the substrate, so in the Foster network model, the heat generated by the chip is transferred to the heat resistance through 4 pairs of heat capacity to The substrate simulates these physical layers, but the heat capacity, thermal resistance, and temperature of these four layers are not related to the heat capacity and thermal resistance of the actual structural layer.
  • the heat resistance and thermal resistance parameters of the first 4 layers are generally obtained from the IGBT module data sheet, and the heat resistance and thermal resistance parameters of the last 3 layers can be obtained by measuring the heat dissipation curve of the controller radiator.
  • the input heat of the first layer in the network model can be obtained by acquiring the power loss of the power module to be analyzed. That is, the input heat of the first layer in the network model is equal to the power loss of the power module to be analyzed.
  • the method determines the temperature difference of each layer in the network model according to the input heat of the first layer in step S102. Subsequently in step S103 to be analyzed to estimate the power module according to the junction temperature T j of each layer and the temperature difference monolayer acquired ambient temperature T a.
  • the method estimates the junction temperature T j of the power module to be analyzed according to the following expression:
  • N represents the number of layers included in the network model
  • ⁇ temp m represents the single layer temperature difference of the mth layer in the network model. It should be noted that in different embodiments of the present invention, the number N of layers included in the network model may be different reasonable values according to actual conditions, and the present invention does not limit the specific value of N.
  • the method preferably adopts an iterative manner to determine and determine the single-layer temperature difference of each layer in the network model.
  • FIG. 3 shows a schematic diagram of an implementation process of determining a single layer temperature difference in this embodiment.
  • the method will obtain the input heat of the i-th time (that is, the current estimation process) of the layer in step S301, and according to step S302
  • the previous temperature difference of the layer and the input component acquired in step S301 determine the current temperature difference of the layer.
  • the single-layer temperature difference can be characterized as shown in FIG. 4.
  • ⁇ temp represents the single layer temperature difference of this layer
  • E in represents the input heat (energy)
  • E out represents the output heat (energy)
  • the thermal model of this layer can be expressed as:
  • P in represents the input heat flow (power)
  • P r represents the heat flow on the thermal resistance r
  • P c represents the heat flow on the heat capacity c.
  • r m represents the thermal resistance of the mth layer
  • ⁇ m represents the thermal time constant of the mth layer
  • T represents the preset time interval
  • the thermal time constant ⁇ m of the m- th layer can be determined by the following expression:
  • c m represents the heat capacity of the m-th layer.
  • T represents the preset time interval.
  • the preset time interval T is preferably configured as the switching cycle of the controller.
  • the input heat of the first layer It is preferably equal to the loss of the power module.
  • the power module to be analyzed preferably includes an IGBT module
  • the loss of the IGBT module preferably includes: IGBT on-state loss, IGBT on-loss, IGBT off loss, FWD on-state loss, FWD off loss, and IGBT module Line loss is 6 parts.
  • the loss of the IGBT module depends not only on the amount of current flowing, but also on the chip junction temperature, gate resistance, driving voltage, etc., and these are non-linear, which makes the calculation of the loss of the IGBT module more complicated .
  • the method provided in this embodiment preferably obtains various loss components (i.e., IGBT on-state loss, IGBT on-loss, IGBT off-loss, FWD on-state loss, etc.) by fitting according to the given performance data of the IGBT module.
  • FWD turn-off loss and IGBT module line loss (6 parts) corresponding to the loss curve, and determine each loss component according to each loss curve. This not only preserves input parameters such as junction temperature and gate resistance, but also improves calculation accuracy.
  • the method may also use other reasonable methods to determine the loss of the IGBT module to be analyzed, and the present invention is not limited to this.
  • the method can also directly ignore the relatively small impact parameters such as junction temperature, gate resistance, and driving voltage, and only introduce an input parameter of current in the loss calculation, and the loss The relationship with the current is linearized, which simplifies the calculation process.
  • the loss calculation needs to be maximized during linearization to leave sufficient protection margin, so that the loss calculation value is too large in most cases.
  • FIG. 5 shows a schematic diagram of an implementation process of determining the IGBT on-state loss power with the corresponding loss curve in this embodiment.
  • the inventor found that the IGBT on-state loss is related to the on-state current I c , the GE voltage V GE and the junction temperature T j . Therefore, as shown in FIG. 5, in this embodiment, in step S501, according to the given performance data of the IGBT module to be analyzed, the method obtains the on-state power loss at the specified junction temperature T j and the specified GE voltage V GE Functional relationship with on-state current.
  • step S501 the method will first fit the CE voltage at the specified junction temperature T j and the specified GE voltage V GE according to the performance data provided in the data manual of the IGBT module to be analyzed
  • the functional relationship between V CE and the on-state current I c namely exists:
  • V CE f(I c ) (6)
  • the method can further obtain the functional relationship between the IGBT on-state power loss and the on-state current I c at the specified junction temperature T j , that is, there is:
  • step S502 the function relationship between the IGBT on-state power loss and the on-state current I c at the specified junction temperature T j.
  • the loss curve corresponding to the IGBT on-state loss power determined by this method can be expressed as:
  • P I represents the IGBT on-state power loss
  • temp represents the junction temperature obtained during the previous round of estimation
  • temp′ represents the reference temperature
  • I c and I max represent the on-state current and the limit current, respectively.
  • P I_temp′ represents the IGBT on-state loss power corresponding to the reference temperature
  • k represents a constant.
  • the value range of the junction temperature temp obtained in the previous round of estimation process is preferably [25,150]. Among them, if the value of the junction temperature temp exceeds 150, then its value will be calculated according to 150; and if the value of the junction temperature temp is less than 25, then its value will be calculated according to 25.
  • the value range of the on-state current I c is preferably [0.5, I max ]. Among them, if the value of the on-state current I c exceeds the limit current I max , then its value will be calculated according to I max ; and if the value of the on-state current I c is less than the limit current 0.5, then its value will be according to 0.5 calculation.
  • the value of the limit current I max is preferably 1200.
  • the value of the limit current I max can also be configured to other reasonable values, and the present invention does not limit the specific value of the limit current I max .
  • Fig. 6 shows the on-state power loss at different temperatures fitted by Infineon FF600R12ME4A_B11.
  • Fig. 7 shows the power error curves corresponding to 25°C and 150°C after the introduction of the temperature coefficient of Infineon FF600R12ME4A_B11.
  • FIG. 8 shows a schematic diagram of an implementation process of determining the loss curve corresponding to the IGBT turn-on loss in this embodiment.
  • the inventor found that the IGBT on-state loss is related to the on-state current I c , the gate resistance R g and the junction temperature T j .
  • the specified temperature, the specified gate resistance, the specified GE voltage, and the specified CE voltage at the turn-on loss and pass are fitted. The functional relationship between the state current.
  • step S802 according to the given performance data of the IGBT module, the function relationship between the turn-on loss and the gate resistance at the specified temperature, the specified GE voltage, the specified CE voltage, and the specified on-state current is fitted.
  • IGBT turn-on loss and gate resistance At a specified temperature of 125°C, a specified GE voltage of 15V, a specified on-state current of 600A, and a specified CE voltage of 600V, the functional relationship between IGBT turn-on loss and gate resistance can be expressed as:
  • step S803 Get the gate resistance-turn-on loss coefficient
  • the method preferably determines the gate resistance-on loss coefficient according to the following expression
  • the method may also use other reasonable methods to determine the gate resistance-on loss coefficient
  • the present invention is not limited to this.
  • step S804 the method can be based on the functional relationship between the turn-on loss and the on-state current and the gate resistance-turn-on loss coefficient To determine the loss curve corresponding to the IGBT turn-on loss.
  • the method preferably determines the loss curve corresponding to the IGBT turn-on loss according to the following expression:
  • E on represents the IGBT turn- on loss
  • temp represents the junction temperature obtained during the previous round of estimation
  • temp′ represents the reference temperature
  • I c and I max represent the on-state current and the limit current
  • E on_temp′ represents the reference temperature.
  • Corresponding IGBT turn-on loss Represents the gate resistance-turn-on loss coefficient.
  • Fig. 9 shows the turn-on loss error at a gate resistance R g of 0.51 ⁇ at 150°C.
  • the value range of the junction temperature temp obtained in the previous round of estimation is preferably [125,150]. Among them, if the value of the junction temperature temp exceeds 150, then its value will be calculated according to 150; and if the value of the junction temperature temp is less than 125, then its value will be calculated according to 125.
  • the value range of the gate resistance R g is preferably [0.51, 4.95].
  • the value range of the on-state current I c is preferably [62, I max ]. Among them, if the value of the on-state current I c exceeds the limit current I max , then its value will be calculated according to I max ; and if the value of the on-state current I c is less than the limit current 62, then its value will be according to 62 calculations.
  • the value of the limit current I max is preferably 1180.
  • the value of the limit current I max can also be configured to other reasonable values, and the present invention does not limit the specific value of the limit current I max .
  • FIG. 10 shows a schematic diagram of an implementation process of determining a loss curve corresponding to an IGBT turn-off loss in this embodiment.
  • the inventor found that the IGBT on-state loss is related to the on-state current I c , the gate resistance R g and the junction temperature T j .
  • the method obtains the specified temperature, the specified gate resistance, the specified GE voltage, and the specified CE voltage. The functional relationship between the on-state current.
  • step S1002 according to the given performance data of the IGBT module, the fitting is performed to obtain the specified temperature, the specified GE voltage, the specified CE voltage, and the specified turn-off loss and gate resistance under specified on-state current. Functional relationship between.
  • IGBT turn-on loss and gate resistance At a specified temperature of 125°C, a specified GE voltage of 15V, a specified on-state current of 600A, and a specified CE voltage of 600V, the functional relationship between IGBT turn-on loss and gate resistance can be expressed as:
  • step S1003 will determine Determine the gate resistance-turn-off loss coefficient as a function of the on-state current and the turn-off loss and gate resistance
  • the method preferably determines the gate resistance-turn-off loss coefficient according to the following expression
  • the method may also use other reasonable methods to determine the gate resistance-off loss coefficient
  • the present invention is not limited to this.
  • step S1004 the method according to the functional relationship between the turn-off loss and the on-state current and the gate resistance-turn-off loss coefficient Determine the loss curve corresponding to IGBT turn-off losses.
  • the method preferably determines the loss curve corresponding to the IGBT turn-off loss according to the following expression:
  • E off represents the IGBT turn-off loss
  • temp represents the junction temperature obtained during the previous round of estimation
  • temp′ represents the reference temperature
  • I c and I max represent the on-state current and the limit current
  • E off_temp′ represents the reference temperature The corresponding IGBT turn-off losses.
  • Fig. 11 shows the turn-off loss error curve at a gate resistance R g of 0.51 ⁇ at 150°C.
  • the value range of the junction temperature temp obtained in the previous round of estimation process is preferably [125,150]. Among them, if the value of the junction temperature temp exceeds 150, then its value will be calculated according to 150; and if the value of the junction temperature temp is less than 125, then its value will be calculated according to 125.
  • the value range of the gate resistance R g is preferably [0.51, 4.95].
  • the value range of the on-state current I c is preferably [62, I max ]. Among them, if the value of the on-state current I c exceeds the limit current I max , then its value will be calculated according to I max ; and if the value of the on-state current I c is less than the limit current 62, then its value will be according to 62 calculations.
  • the value of the limit current I max is preferably 1800.
  • the value of the limit current I max can also be configured to other reasonable values, and the present invention does not limit the specific value of the limit current I max .
  • the power loss of the IGBT module also includes FWD on-state loss and FWD off loss.
  • FIG. 12 shows a schematic diagram of an implementation process of determining the FWD on-state loss of the IGBT module in this embodiment.
  • the inventor found out from the analysis of the IGBT module that, from the forward characteristics of the FWD of the IGBT module, the FWD on-state loss is related to the forward current I f flowing through the FWD and the temperature T j .
  • the method when determining the FWD on-state loss of the IGBT module, the method preferably in step S1201 fits the first specified temperature and The functional relationship between the forward current and forward voltage of the freewheeling diode at the second specified temperature and the third specified temperature. Among them, the third specified temperature, the first specified temperature, and the second specified temperature increase sequentially.
  • step S1202 the method determines the on-state loss and forward direction of the freewheeling diode at the first specified temperature, the second specified temperature, and the third specified temperature according to the functional relationship between the forward current and the forward voltage.
  • the third specified temperature may be set to 25°C
  • the first specified temperature may be set to 125°C
  • the second specified temperature may be set to 150°C.
  • the first specified temperature, the second specified temperature, and the third specified temperature may also be configured to other reasonable values.
  • the present invention does not treat the first specified temperature, the second specified temperature, and the third specified temperature. 3.
  • the specific value of the specified temperature is limited.
  • the functional relationship between the on-state loss of the freewheeling diode and the forward current can also be obtained when the second specified temperature temp2 is 150°C and when the third specified temperature temp3 is 25°C. :
  • step S1203 the method can determine the FWD on-state loss corresponding to the on-state loss of the freewheeling diode and the forward current at the first specified temperature, the second specified temperature, and the third specified temperature. Loss curve.
  • the method uses temperature as one of the parameters to determine the FWD on-state loss of the IGBT. Specifically, in step S1203, the method determines whether the junction temperature obtained during the previous round of estimation is greater than or equal to the first specified temperature. Among them, if the junction temperature obtained during the previous round of estimation is greater than or equal to the first specified temperature, the method will determine the FWD on-state loss according to the following expression:
  • the method will determine the FWD on-state loss according to the following expression:
  • P D represents the FWD on-state loss
  • temp represents the junction temperature obtained during the previous round of estimation
  • P D_temp1 , P D_temp2 and P D_temp3 represent the first specified temperature temp1, the second specified temperature temp2 and the third specified temperature respectively The FWD on-state loss corresponding to temp3.
  • FIG. 13 and FIG. 14 show curves of FWD on-state power loss of the IGBT module when the junction temperature obtained in the previous round of estimation process in this embodiment is 65°C.
  • the value interval of the junction temperature obtained in the previous round of estimation process is preferably [25, 125], wherein, if the temperature exceeds 125°C, it is calculated according to 125°C, and if the temperature is lower than 25°C, it is calculated according to Calculation at 25°C.
  • the method may also use other reasonable methods to determine the FWD on-state loss of the IGBT module, and the present invention is not limited to this.
  • FIG. 15 shows a schematic diagram of an implementation process of determining the FWD turn-off loss of the IGBT module in this embodiment.
  • step S1501 according to the given performance data of the IGBT module, the forward current and off current of the freewheeling diode at the first specified temperature and the second specified temperature are fitted respectively.
  • the functional relationship between the breaking loss is shown in FIG. 15, in this embodiment, according to the given performance data of the IGBT module, the forward current and off current of the freewheeling diode at the first specified temperature and the second specified temperature.
  • step S1502 according to the given performance data of the IGBT module, the function relationship between the first specified temperature, the specified CE voltage, and the specified forward current at the turn-off loss and the gate resistance is fitted.
  • step S1503 the method according to the functional relationship between the turn-off loss of the freewheeling diode and the forward current at the first specified temperature and the second specified temperature, and the above relationship between the turn-off loss and the gate resistance , To determine the loss curve corresponding to FWD turn-off loss.
  • step S1503 the method firstly according to the functional relationship between the turn-off loss obtained in step S1502 and the gate resistance and between the forward current and the turn-off loss at the first specified temperature Function relationship to determine the gate resistance-reverse recovery coefficient
  • the gate resistance-reverse recovery coefficient Preferably, it can be determined according to the following expression:
  • I 1 designates the forward current.
  • the specified forward current I 1 is preferably configured to 600A.
  • the specified forward current I 1 can also be configured to other reasonable values according to actual needs.
  • the present invention does not limit the specific value of the specified forward current I 1 .
  • the FWD turn-off loss of the freewheeling diode and the gate resistance R g can be fitted.
  • the functional relationship between E D_125 f(R g ), so that the gate resistance-reverse recovery coefficient can also be obtained
  • the gate resistance-reverse recovery coefficient can be obtained There is the following relationship with the gate resistance R g :
  • the turn-off loss of the freewheeling diode is also related to temperature, and the FWD turn-off loss E D introduced into the temperature can preferably be determined according to the expression
  • E D represents FWD turn-off loss
  • temp represents the junction temperature obtained during the previous round of estimation
  • temp1 and temp2 represent the first specified temperature and second specified temperature
  • I f represents the forward current
  • E D_temp1 and E D_temp2 represents the turn-off loss of the freewheeling diode corresponding to the first specified temperature and the second specified temperature
  • R g represents the gate resistance.
  • the value interval of the junction temperature obtained in the previous round of estimation process is preferably [25, 125], wherein, if the temperature exceeds 125°C, it is calculated according to 125°C, and if the temperature is lower than 25°C, then Calculation at 25°C.
  • the value of the gate resistance R g is preferably in the range [0.51, 4.95].
  • the value range of the forward current I f is preferably [62, 1200], where if the actual forward current exceeds 1200, it is calculated according to 1200, and if the actual forward current is lower than 62, it is calculated according to 62.
  • FIG. 16 shows the FWD turn-off loss curve when the gate resistance R g is 0.51 ⁇ and the temperature is 135°C in this embodiment.
  • the method may also use other reasonable methods to determine the FWD turn-off loss of the IGBT module, and the present invention is not limited to this.
  • the method determines the line loss of the GIBT module according to the lead resistance of the IGBT module (ie, the resistance between the terminal and the chip).
  • the method preferably determines the line loss of the IGBT module according to the following expression:
  • P L represents the line loss of the IGBT module
  • R CCEE represents the lead resistance
  • I L represents the module current of the IGBT module.
  • the power module temperature estimation method estimates the junction temperature of the power module by iterative calculation, which is convenient to be implemented in a microprocessor.
  • the method introduces parameters such as temperature and gate resistance when calculating the loss of the power module, which can effectively improve the accuracy of the loss calculation, thereby helping to improve the accuracy of the junction temperature estimation, and thus allowing the motor controller to be protected On the premise of maximizing the output capacity.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Computational Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Power Conversion In General (AREA)
  • Inverter Devices (AREA)

Abstract

一种功率模块温度估算方法，包括：步骤一、获取待分析功率模块的损耗，并根据损耗确定待分析功率模块的网络模型中首层的输入热量；步骤二、根据首层的输入热量确定网络模型中各层的单层温差；步骤三、根据各层的单层温差和环境温度确定待分析功率模块的结温。本方法通过迭代计算的方式来对功率模块的结温进行估算，其方便在微处理器中实现。同时，本方法在计算功率模块的损耗时引入了温度、栅极电阻等参数，其能够有效提高损耗计算的精度，从而有助于提高结温估算的精度，进而使得电机控制器能够在得到保护的前提下最大限度的提升输出能力。

Description

一种功率模块温度估算方法
相关技术的交叉引用
本申请要求享有2018年11月30日提交的名称为：“一种功率模块温度估算方法”的中国专利申请CN201811455463.1的优先权，其全部内容通过引用并入本文中。
技术领域
本发明涉及一种电力电子技术领域，具体地说，涉及一种功率模块温度估算方法。
背景技术
近年来，大功率电力电子变流系统广泛应用于柔性直流输电、新能源发电和新能源汽车等领域。大容量IGBT模块作为大功率变流系统中的重要组成部分，其运行状态和可靠性越来越受到行业的高度重视。
根据调查表明，变流器失效中约34％的案例是由器件失效所导致，且55％的系统故障主要是由温度引发。据研究表明，结温每升高10℃，功率器件失效概率将增加一倍。由此可见，过温是IGBT模块可靠运行的主要制约因素之一。
利用IGBT模块内部的NTC温度传感器进行过温保护是功率器件常用的保护策略。但是NTC响应时间慢，不能及时反映出结温的瞬态波动，且在堵转等极端工况下，热能分布不均匀，容易致使IGBT过热而永久失效，因此通过NTC对IGBT进行过温保护的技术方案存在局限性。
发明内容
为解决上述问题，本发明提供了一种功率模块温度估算方法，所述方法包括：
步骤一、获取待分析功率模块的损耗，并根据所述损耗确定所述待分析功率模块的网络模型中首层的输入热量；
步骤二、根据所述首层的输入热量确定所述网络模型中各层的单层温差；
步骤三、根据所述各层的单层温差和环境温度确定所述待分析功率模块的结温。
根据本发明的一个实施例，所述待分析功率模块包括IGBT模块，在所述步骤一 中，获取所述IGBT模块的电流数据，并基于预设损耗功率-电流关系函数来根据所述电流数据确定所述IGBT模块的损耗。
根据本发明的一个实施例，所述待分析功率模块包括IGBT模块，在所述步骤一中，分别获取所述IGBT模块的IGBT通态损耗、IGBT开通损耗、IGBT关断损耗、FWD通态损耗、FWD关断损耗以及IGBT模块线损耗，并根据上述损耗确定所述IGBT模块的损耗。
根据本发明的一个实施例，在所述步骤一中，根据所述待分析IGBT模块的给定性能数据，通过拟合的方式拟合得到各个损耗成分所对应的损耗曲线，并根据各个损耗曲线确定各个损耗成分。
根据本发明的一个实施例，确定所述待分析IGBT模块的IGBT通态损耗功率所对应的损耗曲线的步骤包括：
根据所述待分析IGBT模块的给定性能数据，拟合得到指定结温和指定GE电压下通态损耗功率与通态电流之间的函数关系；
根据所述指定结温和指定GE电压下通态损耗功率与通态电流之间的函数关系以及预设温度-通态损耗系数确定所述IGBT通态损耗功率所对应的损耗曲线。
根据本发明的一个实施例，所述通态损耗功率所对应的损耗曲线包括：
Figure PCTCN2019112503-appb-000001
其中，P I表示IGBT通态损耗功率，temp表示前一轮估算过程中所得到的结温，temp′表示指定结温，I c和I max分别表示通态电流和极限电流，
Figure PCTCN2019112503-appb-000002
表示预设温度-通态损耗系数，P I_temp′表示指定结温所对应的IGBT通态损耗功率，k表示常数。
根据本发明的一个实施例，确定所述IGBT模块的IGBT开通损耗所对应的损耗曲线的步骤包括：
根据所述IGBT模块的给定性能数据，拟合得到指定结温、指定栅极电阻、指定GE电压以及指定CE电压下开通损耗与通态电流之间的函数关系，并拟合得到指定温度、指定GE电压、指定CE电压以及指定通态电流下开通损耗与栅极电阻之间的函数关系；
根据所述开通损耗与通态电流之间的函数关系以及开通损耗与栅极电阻之间的函数关系，确定栅极电阻-开通损耗系数；
根据所述开通损耗与通态电流之间的函数关系以及栅极电阻-开通损耗系数确定所述IGBT开通损耗所对应的损耗曲线。
根据本发明的一个实施例，所述IGBT开通损耗所对应的损耗曲线包括：
Figure PCTCN2019112503-appb-000003
其中，E on表示IGBT开通损耗，temp表示前一轮估算过程中所得到的结温，temp′表示指定结温，I c和I max分别表示通态电流和极限电流，E on_temp′表示指定结温所对应的IGBT开通损耗，
Figure PCTCN2019112503-appb-000004
表示栅极电阻-开通损耗系数，R g表示栅极电阻。
根据本发明的一个实施例，确定所述IGBT模块的IGBT关断损耗功率所对应的损耗曲线的步骤包括：
根据所述IGBT模块的给定性能数据，拟合得到指定温度、指定栅极电阻、指定GE电压以及指定CE电压下开通损耗与通态电流之间的函数关系，并拟合得到指定温度、指定GE电压、指定CE电压以及指定通态电流下关断损耗与栅极电阻之间的函数关系；
根据所述关断损耗与通态电流之间的函数关系以及关断损耗与栅极电阻之间的函数关系，确定栅极电阻-关断损耗系数；
根据所述关断损耗与栅极电阻之间的函数关系以及栅极电阻-关断损耗系数确定所述IGBT关断损耗所对应的损耗曲线。
根据本发明的一个实施例，所述IGBT关断损耗所对应的损耗曲线包括：
Figure PCTCN2019112503-appb-000005
其中，E off表示IGBT关断损耗，temp表示前一轮估算过程中所得到的结温，temp′表示指定结温，I c和I max分别表示通态电流和极限电流，E off_temp′表示参考温度所对应的IGBT关断损耗，
Figure PCTCN2019112503-appb-000006
表示栅极电阻-开通损耗系数，R g表示栅极电阻。
根据本发明的一个实施例，确定所述IGBT模块的FWD通态损耗所对应的损耗曲线的步骤包括：
根据所述IGBT模块的给定性能数据，分别拟合得到第一指定温度、第二指定温度和第三指定温度下续流二极管的正向电流与正向电压之间的函数关系，其中，所述第三指定温度、第一指定温度和第二指定温度依次增大；
根据所述正向电流与正向电压之间的函数关系分别确定第一指定温度、第二指定温度和第三指定温度下所述续流二极管的通态损耗与正向电流之间的函数关系；
根据所述第一指定温度、第二指定温度和第三指定温度下所述续流二极管的通态损耗与正向电流之间的函数关系，确定FWD通态损耗所对应的损耗曲线。
根据本发明的一个实施例，如果前一轮估算过程中所得到的结温大于或等于所述第一指定温度，则根据如下表达式确定FWD通态损耗所对应的损耗曲线：
P D(temp,I f)＝(temp-temp1)/(temp2-temp1)*(P D_temp2-P D_temp1)+P D_temp1
如果前一轮估算过程中所得到的结温小于所述第一指定温度，则根据如下表达式确定FWD通态损耗所对应的损耗曲线：
P D(temp,I f)＝(temp1-temp)/(temp1-temp3)*(P D_temp3-P D_temp1)+P D_temp1
其中，P D表示FWD通态损耗，temp表示前一轮估算过程中所得到的结温，P D_temp1、P D_temp2和P D_temp3分别表示第一指定温度temp1、第二指定温度temp2和第三指定温度temp3所对应的FWD通态损耗。
根据本发明的一个实施例，确定所述IGBT模块的FWD关断损耗所对应的损耗曲线的步骤包括：
根据所述IGBT模块的给定性能数据，分别拟合得到第一指定温度和第二指定温度下续流二极管的正向电流与关断损耗之间的函数关系；
根据所述IGBT模块的给定性能数据，拟合得到第一指定温度、指定CE电压和指定正向电流下关断损耗与栅极电阻之间的函数关系，进而得到栅极电阻-反向恢复系数；
根据所述第一指定温度和第二指定温度下所述续流二极管的关断损耗与正向电流之间的函数关系，以及所述栅极电阻-反向恢复系数，确定FWD关断损耗所对应的损耗曲线。
根据本发明的一个实施例，根据如下表达式确定FWD关断损耗所对应的损耗曲线：
Figure PCTCN2019112503-appb-000007
其中，E D表示FWD关断损耗，temp表示前一轮估算过程中所得到的结温，temp1和temp2分别表示第一指定温度和第二指定温度，I f表示正向电流，E D_temp1和E D_temp2分别表示第一指定温度和第二指定温度所对应的续流二极管的关断损耗，
Figure PCTCN2019112503-appb-000008
表示栅极电阻-反向恢复系数，R g表示栅极电阻。
根据本发明的一个实施例，根据如下表达式确定栅极电阻-反向恢复系数
Figure PCTCN2019112503-appb-000009
Figure PCTCN2019112503-appb-000010
其中，f(R g)表示第一指定温度、指定CE电压和指定正向电流下关断损耗与栅极电阻之间的函数关系。
根据本发明的一个实施例，在所述步骤二中，根据获取到的所述网络模型中各层前一次的温差以及所述首层的输入热量，确定各层当前的单层温差。
根据本发明的一个实施例，对于所述网络模型中的任一层，根据如下表达式确定该层当前的单层温差：
Figure PCTCN2019112503-appb-000011
其中，
Figure PCTCN2019112503-appb-000012
表示网络模型中第m层第i次的单层温差，r m表示第m层的热阻，τ m表示第m层的热时间常数，T表示预设时间间隔，
Figure PCTCN2019112503-appb-000013
表示第m层第i次的输入热量，
Figure PCTCN2019112503-appb-000014
表示第m层第i-1次的单层温差。
根据本发明的一个实施例，对于所述网络模型中的任一层的输出热量等于其下一层的输入热量，其中，根据如下表达式确定任一层的输出热量：
Figure PCTCN2019112503-appb-000015
其中，
Figure PCTCN2019112503-appb-000016
表示第m层第i次的输出热量。
根据本发明的一个实施例，在所述步骤三中，根据如下表达式确定所述待分析功率模块的结温：
Figure PCTCN2019112503-appb-000017
其中，T j表示待分析功率模块的结温，T a表示环境温度，N表示网络模型所包含的层数，△temp m表示网络模型中第m层的单层温差。
本发明所提供的功率模块温度估算方法通过迭代计算的方式来对功率模块的结温进行估算，其方便在微处理器中实现。
同时，本方法在计算功率模块的损耗时引入了温度、栅极电阻等参数，其能够有效提高损耗计算的精度，从而有助于提高结温估算的精度，进而使得电机控制器能够在得到保护的前提下最大限度的提升输出能力。
本发明的其它特征和优点将在随后的说明书中阐述，并且，部分地从说明书中变得显而易见，或者通过实施本发明而了解。本发明的目的和其他优点可通过在说明书、权利要求书以及附图中所特别指出的结构来实现和获得。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要的附图做简单的介绍：
图1是根据本发明一个实施例的功率模块温度估算方法的实现流程示意图；
图2是根据本发明一个实施例的Foster网络模型示意图；
图3是根据本发明一个实施例的确定单层温差的实现流程示意图；
图4是根据本发明一个实施例的单层温差计算原理示意图；
图5是根据本发明一个实施例的确定IGBT通态损耗功率随对应的损耗曲线的实现流程示意图；
图6是根据本发明一个实施例的英飞凌FF600R12ME4A_B11所拟合得到的不同温度下的通态损耗功率；
图7是根据本发明一个实施例的英飞凌FF600R12ME4A_B11引入温度系数后25℃和150摄氏度所对应的功率误差曲线；
图8是根据本发明一个实施例的确定IGBT开通损耗所对应的损耗曲线的实现流程示意图；
图9是根据本发明一个实施例的栅极电阻R g为0.51Ω、150℃下的开通损耗误差；
图10是根据本发明一个实施例的确定IGBT关断损耗所对应的损耗曲线的实现流程示意图；
图11是根据本发明一个实施例的栅极电阻R g为0.51Ω、150℃下的关断损耗误差曲线；
图12是根据本发明一个实施例的确定IGBT模块的FWD通态损耗的实现流程示意图；
图13和图14是根据本发明一个实施例的前一轮估算过程中所得到的结温为65℃时IGBT模块的FWD通态损耗功率的曲线
图15是根据本发明一个实施例的确定IGBT模块的FWD关断损耗的实现流程示意图；
图16是根据本发明一个实施例的栅极电阻R g为0.51Ω、温度为135℃时的FWD关 断损耗曲线。
具体实施方式
以下将结合附图及实施例来详细说明本发明的实施方式，借此对本发明如何应用技术手段来解决技术问题，并达成技术效果的实现过程能充分理解并据以实施。需要说明的是，只要不构成冲突，本发明中的各个实施例以及各实施例中的各个特征可以相互结合，所形成的技术方案均在本发明的保护范围之内。
同时，在以下说明中，出于解释的目的而阐述了许多具体细节，以提供对本发明实施例的彻底理解。然而，对本领域的技术人员来说显而易见的是，本发明可以不用这里的具体细节或者所描述的特定方式来实施。
另外，在附图的流程图示出的步骤可以在诸如一组计算机可执行指令的计算机系统中执行，并且，虽然在流程图中示出了逻辑顺序，但是在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤。
发明人通过研究发现，根据实时采集的一些控制器参数，以热网络模型实现对IGBT的结温估算具有三个优势。其一，能实时监测IGBT模块的瞬时结温，即便是在堵转的极限工况下也能进行过温保护；其二，能够减小过温保护的裕量，有效提高IGBT模块的输出能力，从而使整车动力性更强；其三，能检测功率模块的损耗功率，为整车电耗优化提供参数支持。
现有技术中存在一种用于风电变流器可靠性评估的结温数值计算方法，该方法提出了一种结温迭代数值计算方案。该方法基于电热比拟理论讨论了IGBT模块在一个开关周期内的结温计算方法，并通过迭代算法计算了IGBT模块的工频周期结温。但是该方法提供的结温计算方案存在一些缺点。例如，该方法在进行IGBT的热损耗的计算过程中并未引入结温参数，这样也就影响了结温计算精度。IGBT的结温与其损耗直接相关，而结温反过来影响损耗，即IGBT在同等电参数不同结温下，其损耗是不同的，因此损耗必须与IGBT当前的结温相关。另外，该方法的计算过程存在复杂的矩阵运算，不适应于一般的电机控制器处理芯片。此外，该方法在进行IGBT的结温计算时忽略了散热器热参数的考虑，同样也影响结温计算精度。
现有技术中还存在一种IGBT结温检测装置及其方法，该方法首先检测IGBT的开关频率和IGBT导通时的电流，然后根据预定模型计算IGBT的结温温升，最后将IGBT的结温温升加上IGBT散热器的温度作为IGBT的结温。而该方法同样存在诸多缺点。例 如，该方法在实施过程中，电流检测单位、频率检测单元和隔离单元等带来了额外硬件成本开销，并且可能引入检测误差。另外，该方法所使用的损耗计算模型为简单的线性计算公式，难以匹配IGBT损耗非线性曲线，从而导致损耗计算不准确。同时，该方法对于IGBT的热损耗的计算没有引入结温参数，从而影响结温计算精度。另外，该方法将结温计算简化为两阶计算，即结温等于散热器温度加IGBT温升，其难以模拟IGBT与散热器构成的多层热模型。
此外，现有技术中还存在一种基于饱和压降测量IGBT功率模块结温的在线检测装置，该装置选取小电流下集电极和发射极的电压VCE作为温度敏感参数，通过预先的温箱实验确定饱和压降V CE与结温之间的函数关系，然后在功率循环条件下通过开关IGBT模块控制大电流关断，监测IGBT关断瞬间的压降V CE的变化，最终代入预先求出的函数关系得到结温。该方法在实施中由于需要使用小电流检测单元等，因而带来了额外的硬件成本，且需要对该单元增加其他隔离电路。同时该方法由于需要进行预先的温箱试验，因此较为耗时。
针对现有技术中所存在的上述问题，本发明提供了一种新的功率模块温度估算方法，该方法特别适用于对于电机控制器功率模块的结温的估算。
图1示出了本实施例所提供的功率模块温度估算方法的实现流程示意图。
如图1所示，本实施例所所提供的功率模块温度估算方法首先会在步骤S101中获取待分析功率模块的损耗功率，并根据所确定出的损耗功率来得到待分析功率模块的网络模型中首层的输入热量。
具体地，如图2所示，本方法将功率模块等效为如图2所示的Foster网络模型。热量由IGBT芯片和FWD芯片产生，然后流经多个材料层后由散热器带走。由于IGBT芯片和FWD芯片共基板，因此本实施例中将IGBT芯片和FWD芯片视为具有相同的壳温T c。
本实施例中，在Foster网络模型中，芯片与基板之间优选地存在焊接层、铜箔层、陶瓷层等，因此在Foster网络模型中，芯片产生的热量通过4对热容热阻传至基板以模拟这些物理层，但是这4层的热容、热阻和温度与实际结构层的热容热阻没有关系。
在Foster网络模型中，基板与散热介质之间存在导热胶、散热器两个物理层，为了更加精细的对结温进行估算，本实施例中，这两层用到了3对热容热阻来表示。
前4层的热容热阻参数一般从IGBT模块的数据手册得到，而后3层的热容热阻参数则可以通过测量控制器散热器的散热曲线来得到。
本实施例中，通过获取待分析功率模块的损耗功率，也就可以得到网络模型中首层的输入热量。即网络模型中首层的输入热量等于待分析功率模块的损耗功率。
如图1所示，本实施例中，在得到网络模型中首层的输入热量后，该方法会在步骤S102中根据首层的输入热量确定网络模型中各层的单层温差。随后再在步骤S103中根据各层的单层温差以及获取到的环境温度T a来估算出待分析功率模块的结温T j。
具体地，本实施例中，该方法根据如下表达式估算待分析功率模块的结温T j：
Figure PCTCN2019112503-appb-000018
其中，N表示网络模型所包含的层数，△temp m表示网络模型中第m层的单层温差。需要指出的是，在本发明的不同该实施例中，根据实际情况，网络模型的所包含的层数N可以为不同的合理值，本发明并不对N的具体取值进行限定。
本实施例中，该方法优选地采用迭代的方式来确定确定网络模型中各层的单层温差。图3示出了本实施例中确定单层温差的实现流程示意图。
如图3所示，本实施例中，对于网络模型中的任一层，该方法会在步骤S301中获取该层第i次(即本次估算过程)的输入热量，并在步骤S302中根据该层前一次的温差以及步骤S301中所获取到的输入分量来确定该层的当前温差。
如图4所示，对于网络模型中的任一层，其单层温差可以表征为图4所示。在图4中，△temp表示该层的单层温差，E in表示输入热量(能量)，E out表示输出热量(能量)，那么该层的热路模型可以表示为：
Figure PCTCN2019112503-appb-000019
其中，P in表示输入热流量(功率)，P r表示热阻r上的热流量，P c表示热容c上的热流量。
简化表达式(2)也就可以得到单层温差的计算表达式：
Figure PCTCN2019112503-appb-000020
其中，
Figure PCTCN2019112503-appb-000021
表示网络模型中第m层第i次的单层温差，r m表示第m层的热阻，τ m表示第m层的热时间常数，T表示预设时间间隔，
Figure PCTCN2019112503-appb-000022
表示第m层第i次的输入热 量，
Figure PCTCN2019112503-appb-000023
表示第m层第i-1次的单层温差。
对于表达式(3)来说，如果网络模型所包含的层数为N，那么m的取值也就为[1,N]。
其中，第m层的热时间常数τ m可以采用如下表达式确定：
τ m＝r mc m     (4)
其中，c m表示第m层的热容。
而第m层的输出热量
Figure PCTCN2019112503-appb-000024
可以通过如下表达式确定：
Figure PCTCN2019112503-appb-000025
其中，T表示预设时间间隔。本实施例中，预设时间间隔T优选的配置为控制器的开关周期。
如果需要确定网络模型中各层的单层温差，也就需要确定出本次估算过程中网络模型中首层的输入热量
Figure PCTCN2019112503-appb-000026
而本实施例中，首层的输入热量
Figure PCTCN2019112503-appb-000027
优选地等于功率模块的损耗。
本实施例中，待分析功率模块优选地包括IGBT模块，而IGBT模块的损耗优选地包括：IGBT通态损耗、IGBT开通损耗、IGBT关断损耗、FWD通态损耗、FWD关断损耗以及IGBT模块线损耗6部分。
由于IGBT模块的损耗不仅取决于流经的电流大小，还与芯片结温、栅极电阻、驱动电压等有关系，而且这些都是非线性的，这样也就使得对于IGBT模块的损耗的计算较为复杂。
本实施例所提供的方法优选地根据IGBT模块的给定性能数据，通过拟合的方式拟合得到各个损耗成分(即IGBT通态损耗、IGBT开通损耗、IGBT关断损耗、FWD通态损耗、FWD关断损耗以及IGBT模块线损耗6部分)所对应的损耗曲线，并根据各个损耗曲线确定各个损耗成分。这样也就不仅保留了结温、栅极电阻等输入参数，还提高了计算精度。
当然，在本发明的其他实施例中，根据实际需要，该方法还可以采用其他合理方式来确定待分析IGBT模块的损耗，本发明不限于此。例如，在本发明的一个实施例中，该方法还可以将结温、栅极电阻、驱动电压这些影响相对较小的参数直接忽略掉，在损耗计算时只引入电流一个输入参数，并且将损耗与电流关系线性化，从而简化了计算过 程。但是这种方式在线性化时需要将损耗计算最大化来留足保护裕量，使得大部分情况下损耗计算值偏大。
为了更加清楚地阐述本发明所提供的功率模块温度估算方法确定各个损耗成分的原理以及过程，以下以英飞凌FF600R12ME4A_B11为例进行说明。需要指出的是，由于拟合的多项式计算量较大导致程序开销较大，因此可以根据实际需要可以降低拟合多项式阶数来减少计算时间。
图5示出了本实施例中确定IGBT通态损耗功率随对应的损耗曲线的实现流程示意图。
通过对英飞凌FF600R12ME4A_B11的数据手册中的传输特性进行分析，发明人发现IGBT通态损耗与通态电流I c、GE电压V GE和结温T j有关。因此，如图5所示，本实施例中，该方法会在步骤S501中根据待分析IGBT模块的给定性能数据，拟合得到指定结温T j和指定GE电压V GE下通态损耗功率与通态电流之间的函数关系。
具体地，本实施例中，该方法在步骤S501中首先会根据待分析IGBT模块的数据手册中所提供的性能数据来在指定结温T j和指定GE电压V GE下，拟合得到CE电压V CE与通态电流I c的函数关系，即存在：
V CE＝f(I c)      (6)
根据上述函数关系，该方法可以进一步得到该指定结温T j下IGBT通态损耗功率与通态电流I c之间的函数关系，即存在：
Figure PCTCN2019112503-appb-000028
其中，
Figure PCTCN2019112503-appb-000029
表示指定结温T j以及通态电流I c下IGBT通态损耗功率。
如图5所示，本实施例中，在得到指定结温T j下IGBT通态损耗功率与通态电流I c之间的函数关系
Figure PCTCN2019112503-appb-000030
后，该方法会在步骤S502中根据指定结温T j和指定GE电压V GE下通态损耗功率与通态电流之间的函数关系
Figure PCTCN2019112503-appb-000031
以及预设温度-通态损耗系数确定IGBT通态损耗功率所对应的损耗曲线。
具体地，本实施例中，该方法所确定出的IGBT通态损耗功率所对应的损耗曲线可以表示为：
Figure PCTCN2019112503-appb-000032
其中，P I表示IGBT通态损耗功率，temp表示前一轮估算过程中所得到的结温，temp′表示参考温度，I c和I max分别表示通态电流和极限电流，
Figure PCTCN2019112503-appb-000033
表示预设温度- 通态损耗系数，P I_temp′表示参考温度所对应的IGBT通态损耗功率，k表示常数。
本实施例中，前一轮估算过程中所得到的结温temp的取值范围优选为[25,150]。其中，如果结温temp的取值超过150，那么其取值将会按照150计算；而如果结温temp的取值小于25，那么其取值则会按照25计算。
通态电流I c的取值范围优选地为[0.5,I max]。其中，如果通态电流I c的取值超过极限电流I max，那么其取值将会按照I max计算；而如果通态电流I c的取值小于极限电流0.5，那么其取值将会按照0.5计算。本实施例中，极限电流I max的取值优选地为1200。当然，在本发明的其他实施例中，根据实际情况，极限电流I max的取值还可以配置为其他合理值，本发明并不对极限电流I max的具体取值进行限定。
例如，对于英飞凌FF600R12ME4A_B11来说，在GE电压V GE为15V、结温T j为125℃的情况下，可以拟合得到如下IGBT通态损耗功率与通态电流I c之间的函数关系：
Figure PCTCN2019112503-appb-000034
通过调整温度-功率系数
Figure PCTCN2019112503-appb-000035
的取值，使得P I_25＝f(25,P I_125)并且P I_150＝f(150,P I_125)。经过计算，调整温度-功率系数
Figure PCTCN2019112503-appb-000036
优选地取值为2.14279E-3。这样也就存在：
Figure PCTCN2019112503-appb-000037
图6示出了英飞凌FF600R12ME4A_B11所拟合得到的不同温度下的通态损耗功率，图7则示出了英飞凌FF600R12ME4A_B11引入温度系数后25℃和150摄氏度所对应的功率误差曲线。
图8示出了本实施例中确定IGBT开通损耗所对应的损耗曲线的实现流程示意图。
通过对英飞凌FF600R12ME4A_B11的数据手册中的传输特性进行分析，发明人发现IGBT通态损耗与通态电流I c、栅极电阻R g和结温T j有关。如图8所示，本实施例中，该方法会在步骤S801中根据IGBT模块的给定性能数据，拟合得到指定温度、指定栅极电阻、指定GE电压以及指定CE电压下开通损耗与通态电流之间的函数关系。
同时，该方法还会在步骤S802中根据IGBT模块的给定性能数据，拟合得到指定温度、指定GE电压、指定CE电压以及指定通态电流下开通损耗与栅极电阻之间的函数关系。
例如，对于英飞凌FF600R12ME4A_B11来说，在指定温度125℃、指定栅极电阻0.51Ω、指定GE电压15V以及指定CE电压600V下，IGBT开通损耗与通态电流之间的 函数关系可以表示为：
Figure PCTCN2019112503-appb-000038
而在指定温度125℃、指定GE电压15V、指定通态电流600A以及指定CE电压600V下，IGBT开通损耗与栅极电阻之间的函数关系可以表示为：
E on_125＝f(R g)      (12)
如图8所示，本实施例中，在得到IGBT开通损耗与通态电流之间的函数关系以及IGBT开通损耗与栅极电阻之间的函数关系后，该方法会在步骤S803中根据上述关系得到栅极电阻-开通损耗系数
Figure PCTCN2019112503-appb-000039
具体地，本实施例中，该方法优选地根据如下表达式确定栅极电阻-开通损耗系数
Figure PCTCN2019112503-appb-000040
Figure PCTCN2019112503-appb-000041
例如，对于英飞凌FF600R12ME4A_B11来说，其栅极电阻-开通损耗系数
Figure PCTCN2019112503-appb-000042
可以根据如下表达式确定得到：
Figure PCTCN2019112503-appb-000043
当然，在本发明的其他实施例中，该方法还可以采用其他合理方式来确定栅极电阻-开通损耗系数
Figure PCTCN2019112503-appb-000044
本发明不限于此。
如图8所示，在得到栅极电阻-开通损耗系数
Figure PCTCN2019112503-appb-000045
后，该方法则可以在步骤S804中根据开通损耗与通态电流之间的函数关系以及栅极电阻-开通损耗系数
Figure PCTCN2019112503-appb-000046
来确定IGBT开通损耗所对应的损耗曲线。
具体地，本实施例中，该方法优选地根据如下表达式确定IGBT开通损耗所对应的损耗曲线：
Figure PCTCN2019112503-appb-000047
其中，E on表示IGBT开通损耗，temp表示前一轮估算过程中所得到的结温，temp′表示参考温度，I c和I max分别表示通态电流和极限电流，E on_temp′表示参考温度所对应的IGBT开通损耗，
Figure PCTCN2019112503-appb-000048
表示栅极电阻-开通损耗系数。
对于英飞凌FF600R12ME4A_B11来说，当栅极电阻-开通损耗系数
Figure PCTCN2019112503-appb-000049
的取值为8.37949E-03时，存在：
Figure PCTCN2019112503-appb-000050
图9示出了栅极电阻R g为0.51Ω、150℃下的开通损耗误差。本实施例中，在确定IGBT开通损耗所对应的损耗曲线的过程中，前一轮估算过程中所得到的结温temp的取值范围优选为[125,150]。其中，如果结温temp的取值超过150，那么其取值将会按照150计算；而如果结温temp的取值小于125，那么其取值则会按照125计算。栅极电阻R g的取值范围优选地为[0.51,4.95]。
通态电流I c的取值范围优选地为[62,I max]。其中，如果通态电流I c的取值超过极限电流I max，那么其取值将会按照I max计算；而如果通态电流I c的取值小于极限电流62，那么其取值将会按照62计算。本实施例中，极限电流I max的取值优选地为1180。当然，在本发明的其他实施例中，根据实际情况，极限电流I max的取值还可以配置为其他合理值，本发明并不对极限电流I max的具体取值进行限定。
图10示出了本实施例中确定IGBT关断损耗所对应的损耗曲线的实现流程示意图。
通过对英飞凌FF600R12ME4A_B11的数据手册中的传输特性进行分析，发明人发现IGBT通态损耗与通态电流I c、栅极电阻R g和结温T j有关。如图10所示，本实施例中，该方法会在步骤S1001中根据IGBT模块的给定性能数据，拟合得到指定温度、指定栅极电阻、指定GE电压以及指定CE电压下关断损耗与通态电流之间的函数关系。
同样的，该方法还会在步骤S1002中根据IGBT模块的给定性能数据，拟合得到拟合得到指定温度、指定GE电压、指定CE电压以及指定通态电流下关断损耗与栅极电阻之间的函数关系。
例如，对于英飞凌FF600R12ME4A_B11来说，在指定温度125℃、指定栅极电阻0.51Ω、指定GE电压15V以及指定CE电压600V下，IGBT关断损耗与通态电流之间的函数关系可以表示为：
Figure PCTCN2019112503-appb-000051
而在指定温度125℃、指定GE电压15V、指定通态电流600A以及指定CE电压600V下，IGBT开通损耗与栅极电阻之间的函数关系可以表示为：
E off_125＝f(R g)    (18)
如图10所示，在得到关断损耗与通态电流之间的函数关系以及关断损耗与栅极电阻之间的函数关系后，该方法会在步骤S1003中会根据所得到的关断损耗与通态电流之间的函数关系以及关断损耗与栅极电阻之间的函数关系，确定栅极电阻-关断损耗系数
Figure PCTCN2019112503-appb-000052
具体地，本实施例中，该方法优选地根据如下表达式确定栅极电阻-关断损耗系数
Figure PCTCN2019112503-appb-000053
Figure PCTCN2019112503-appb-000054
例如，对于英飞凌FF600R12ME4A_B11来说，其栅极电阻-关断损耗系数
Figure PCTCN2019112503-appb-000055
可以根据如下表达式确定得到：
Figure PCTCN2019112503-appb-000056
当然，在本发明的其他实施例中，该方法还可以采用其他合理方式来确定栅极电阻-关断损耗系数
Figure PCTCN2019112503-appb-000057
本发明不限于此。
如图10所示，在得到栅极电阻-关断损耗系数
Figure PCTCN2019112503-appb-000058
后，该方法会在步骤S1004中根据关断损耗与通态电流之间的函数关系以及栅极电阻-关断损耗系数
Figure PCTCN2019112503-appb-000059
确定IGBT关断损耗所对应的损耗曲线。
具体地，本实施例中，该方法优选地根据如下表达式确定IGBT关断损耗所对应的损耗曲线：
Figure PCTCN2019112503-appb-000060
其中，E off表示IGBT关断损耗，temp表示前一轮估算过程中所得到的结温，temp′表示参考温度，I c和I max分别表示通态电流和极限电流，E off_temp′表示参考温度所对应的IGBT关断损耗。
对于英飞凌FF600R12ME4A_B11来说，栅极电阻-开通损耗系数
Figure PCTCN2019112503-appb-000061
的取值为4.99710E-03时，存在：
Figure PCTCN2019112503-appb-000062
图11示出了栅极电阻R g为0.51Ω、150℃下的关断损耗误差曲线。本实施例中，在确定IGBT关断损耗所对应的损耗曲线的过程中，前一轮估算过程中所得到的结温temp的取值范围优选为[125,150]。其中，如果结温temp的取值超过150，那么其取值将会按照150计算；而如果结温temp的取值小于125，那么其取值则会按照125计算。栅极电阻R g的取值范围优选地为[0.51,4.95]。
通态电流I c的取值范围优选地为[62,I max]。其中，如果通态电流I c的取值超过极限 电流I max，那么其取值将会按照I max计算；而如果通态电流I c的取值小于极限电流62，那么其取值将会按照62计算。本实施例中，极限电流I max的取值优选地为1800。当然，在本发明的其他实施例中，根据实际情况，极限电流I max的取值还可以配置为其他合理值，本发明并不对极限电流I max的具体取值进行限定。
正如前述内容所分析的那样，IGBT模块的损耗功率还包括FWD通态损耗和FWD关断损耗。图12示出了本实施例中确定IGBT模块的FWD通态损耗的实现流程示意图。
发明人通过对IGBT模块进行分析发现，从IGBT模块的FWD的正向特性来看，FWD通态损耗与FWD流过的正向电流I f以及温度T j有关。如图12所示，本实施例中，该方法在确定IGBT模块的FWD通态损耗时，优选地会在步骤S1201中根据IGBT模块的给定性能数据，分别拟合得到第一指定温度、第二指定温度和第三指定温度下续流二极管的正向电流与正向电压之间的函数关系。其中，第三指定温度、第一指定温度和第二指定温度依次增大。
随后，该方法会在步骤S1202中根据上述正向电流与正向电压之间的函数关系分别确定第一指定温度、第二指定温度和第三指定温度下续流二极管的通态损耗与正向电流之间的函数关系。
具体地，本实施例中，对于一指定温度，其续流二极管的正向电流与正向电压之间的函数关系可以表征为V f＝f(I f)，因此该温度下的通态损耗也就可以表示为：
P D(temp,I f)＝V f*I f＝f(I f)*I f     (23)
例如，本实施例中，第三指定温度可以设定为25℃，第一指定温度可以设定为125℃，第二指定温度可以设定为150℃。
当然，在本发明的其他实施例中，上述第一指定温度、第二指定温度和第三指定温度还可以配置为其他合理值，本发明并不对上述第一指定温度、第二指定温度和第三指定温度的具体取值进行限定。
例如，对于英飞凌FF600R12ME4A_B11来说，在第一指定温度temp1为125℃的情况下，可以拟合得到该温度下续流二极管的通态损耗与正向电流之间的函数关系，即存在：
Figure PCTCN2019112503-appb-000063
同样地，也可以分别得到第二指定温度temp2为150℃的情况下以及第三指定温度temp3为25℃的情况下，续流二极管的通态损耗与正向电流之间的函数关系，即存在：
Figure PCTCN2019112503-appb-000064
Figure PCTCN2019112503-appb-000065
该方法在步骤S1203中则可以根据上述第一指定温度、第二指定温度和第三指定温度下续流二极管的通态损耗与正向电流之间的函数关系，确定FWD通态损耗所对应的损耗曲线。
本实施例中，该方法会将温度作为参数之一来确定IGBT的FWD通态损耗。具体地，该方法在步骤S1203中会判断前一轮估算过程中所得到的结温是否大于或等于上述第一指定温度。其中，如果前一轮估算过程中所得到的结温大于或等于上述第一指定温度，该方法则会根据如下表达式确定FWD通态损耗：
P D(temp,I f)＝(temp-temp1)/(temp2-temp1)*(P D_temp2-P D_temp1)+P D_temp1  (27)
而如果前一轮估算过程中所得到的结温小于第一指定温度，该方法则会根据如下表达式确定FWD通态损耗：
P D(temp,I f)＝(temp1-temp)/(temp1-temp3)*(P D_temp3-P D_temp1)+P D_temp1  (28)
其中，P D表示FWD通态损耗，temp表示前一轮估算过程中所得到的结温，P D_temp1、P D_temp2和P D_temp3分别表示第一指定温度temp1、第二指定温度temp2和第三指定温度temp3所对应的FWD通态损耗。
其中，图13和图14示出了本实施例中前一轮估算过程中所得到的结温为65℃时IGBT模块的FWD通态损耗功率的曲线。
本实施例中，前一轮估算过程中所得到的结温的取值区间优选地为[25，125]，其中，温度超过125℃则按照125℃计算，而温度如果低于25℃则按照25℃计算。
当然，在本发明的其他实施例中，该方法还可以采用其他合理方式来确定IGBT模块的FWD通态损耗，本发明不限于此。
发明人通过对IGBT模块进行分析发现，从IGBT模块的传输特性来看，FWD关断损耗ED与流过的正向电流I f、温度T j以及栅极电阻R g有关。图15示出了本实施例中确定IGBT模块的FWD关断损耗的实现流程示意图。
如图15所示，本实施例中，该方法会在步骤S1501中根据IGBT模块的给定性能数据，分别拟合得到第一指定温度和第二指定温度下续流二极管的正向电流与关断损耗之间的函数关系。
同时，该方法还会在步骤S1502中根据IGBT模块的给定性能数据，拟合得到第一 指定温度、指定CE电压和指定正向电流下关断损耗与栅极电阻之间的函数关系。
随后，该方法会在步骤S1503中根据第一指定温度和第二指定温度下续流二极管的关断损耗与正向电流之间的函数关系，以及关断损耗与栅极电阻之间的上述关系，来确定FWD关断损耗所对应的损耗曲线。
具体地，本实施例中，该方法在步骤S1503中首先会根据步骤S1502中所得到的关断损耗与栅极电阻之间的函数关系以及第一指定温度下正向电流与关断损耗之间的函数关系来确定出栅极电阻-反向恢复系数
Figure PCTCN2019112503-appb-000066
其中，栅极电阻-反向恢复系数
Figure PCTCN2019112503-appb-000067
优选地可以根据如下表达式确定：
Figure PCTCN2019112503-appb-000068
其中，I 1表示指定正向电流。
本实施例中，指定正向电流I 1优选地配置为600A。当然，在本发明的其他实施例中，指定正向电流I 1还可以根据实际需要配置为其他合理值，本发明并不对指定正向电流I 1的具体取值进行限定。
例如，对于英飞凌FF600R12ME4A_B11来说，在第一指定温度为125℃、指定CE电压为600V、指定栅极电阻为0.51Ω的情况下，可以拟合得到续流二极管的正向电流与关断损耗之间的函数关系为：
Figure PCTCN2019112503-appb-000069
同样可以得到在第二指定温度为150℃、指定CE电压为600V、指定栅极电阻为0.51Ω的情况下，可以拟合得到续流二极管的正向电流与关断损耗之间的函数关系为：
Figure PCTCN2019112503-appb-000070
本实施例中，在第一指定温度为125℃、指定CE电压为600V以及正向电流I f为600A的情况下，可以拟合得到续流二极管的FWD关断损耗与栅极电阻R g之间的函数关系E D_125＝f(R g)，这样也就可以得到栅极电阻-反向恢复系数
Figure PCTCN2019112503-appb-000071
本实施例中，经过尝试，可以得到栅极电阻-反向恢复系数
Figure PCTCN2019112503-appb-000072
与栅极电阻R g存在如下关系：
Figure PCTCN2019112503-appb-000073
续流二极管的关断损耗同样与温度有关，引入温度的FWD关断损耗E D优选地可以 根据如下表达式确定：
Figure PCTCN2019112503-appb-000074
其中，E D表示FWD关断损耗，temp表示前一轮估算过程中所得到的结温，temp1和temp2分别表示第一指定温度和第二指定温度，I f表示正向电流，E D_temp1和E D_temp2分别表示第一指定温度和第二指定温度所对应的续流二极管的关断损耗，
Figure PCTCN2019112503-appb-000075
表示栅极电阻-反向恢复系数，R g表示栅极电阻。
本实施例中，前一轮估算过程中所得到的结温的取值区间优选地为[25，125]，其中，温度超过125℃则按照125℃计算，而温度如果低于25℃则按照25℃计算。栅极电阻R g的取值为范围优选地为[0.51,4.95]。正向电流I f的取值范围优选的为[62，1200]，其中，如果实际正向电流超过1200，则按照1200进行计算，而如果实际正向电流低于62，则按照62进行计算。
图16示出了本实施例中栅极电阻R g为0.51Ω、温度为135℃时的FWD关断损耗曲线。
当然，在本发明的其他实施例中，该方法还可以采用其他合理方式来确定IGBT模块的FWD关断损耗，本发明不限于此。
IGBT模块在运行的过程中，传输线路同样存在损耗。本实施例中，该方法还会根据IGBT模块的引线电阻(即端子到芯片之间的电阻)来确定GIBT模块线损耗。
具体地，本实施例中，该方法优选地根据如下表达式来确定IGBT模块线损耗：
Figure PCTCN2019112503-appb-000076
其中，P L表示IGBT模块线损耗，R CCEE表示引线电阻，I L表示IGBT模块的模块电流。
从上述描述中可以看出，本实施例所提供的功率模块温度估算方法通过迭代计算的方式来对功率模块的结温进行估算，其方便在微处理器中实现。
同时，本方法在计算功率模块的损耗时引入了温度、栅极电阻等参数，其能够有效提高损耗计算的精度，从而有助于提高结温估算的精度，进而使得电机控制器能够在得到保护的前提下最大限度的提升输出能力。
应该理解的是，本发明所公开的实施例不限于这里所公开的特定结构或处理步骤，而应当延伸到相关领域的普通技术人员所理解的这些特征的等同替代。还应当理解的是，在此使用的术语仅用于描述特定实施例的目的，而并不意味着限制。
说明书中提到的“一个实施例”或“实施例”意指结合实施例描述的特定特征、结构或特性包括在本发明的至少一个实施例中。因此，说明书通篇各个地方出现的短语“一个实施例”或“实施例”并不一定均指同一个实施例。
虽然上述示例用于说明本发明在一个或多个应用中的原理，但对于本领域的技术人员来说，在不背离本发明的原理和思想的情况下，明显可以在形式上、用法及实施的细节上作各种修改而不用付出创造性劳动。因此，本发明由所附的权利要求书来限定。

Claims (19)

  1. 一种功率模块温度估算方法，其中，所述方法包括：
    步骤一、获取待分析功率模块的损耗，并根据所述损耗确定所述待分析功率模块的网络模型中首层的输入热量；
    步骤二、根据所述首层的输入热量确定所述网络模型中各层的单层温差；
    步骤三、根据所述各层的单层温差和环境温度确定所述待分析功率模块的结温。
  2. 如权利要求1所述的方法，其中，所述待分析功率模块包括IGBT模块，在所述步骤一中，获取所述IGBT模块的电流数据，并基于预设损耗功率-电流关系函数来根据所述电流数据确定所述IGBT模块的损耗。
  3. 如权利要求1所述的方法，其中，所述待分析功率模块包括IGBT模块，在所述步骤一中，分别获取所述IGBT模块的IGBT通态损耗、IGBT开通损耗、IGBT关断损耗、FWD通态损耗、FWD关断损耗以及IGBT模块线损耗，并根据上述损耗确定所述IGBT模块的损耗。
  4. 如权利要求3所述的方法，其中，在所述步骤一中，根据所述待分析IGBT模块的给定性能数据，通过拟合的方式拟合得到各个损耗成分所对应的损耗曲线，并根据各个损耗曲线确定各个损耗成分。
  5. 如权利要求4所述的方法，其中，确定所述待分析IGBT模块的IGBT通态损耗功率所对应的损耗曲线的步骤包括：
    根据所述待分析IGBT模块的给定性能数据，拟合得到指定结温和指定GE电压下通态损耗功率与通态电流之间的函数关系；
    根据所述指定结温和指定GE电压下通态损耗功率与通态电流之间的函数关系以及预设温度-通态损耗系数确定所述IGBT通态损耗功率所对应的损耗曲线。
  6. 如权利要求5所述的方法，其中，所述通态损耗功率所对应的损耗曲线包括：
    Figure PCTCN2019112503-appb-100001
    其中，P I表示IGBT通态损耗功率，temp表示前一轮估算过程中所得到的结温，temp′表示指定结温，I c和I max分别表示通态电流和极限电流，
    Figure PCTCN2019112503-appb-100002
    表示预设温度-通态损耗系数，P I_temp′表示指定结温所对应的IGBT通态损耗功率，k表示常数。
  7. 如权利要求4～6中任一项所述的方法，其中，确定所述IGBT模块的IGBT开通损耗所对应的损耗曲线的步骤包括：
    根据所述IGBT模块的给定性能数据，拟合得到指定结温、指定栅极电阻、指定GE电压以及指定CE电压下开通损耗与通态电流之间的函数关系，并拟合得到指定温度、指定GE电压、指定CE电压以及指定通态电流下开通损耗与栅极电阻之间的函数关系；
    根据所述开通损耗与通态电流之间的函数关系以及开通损耗与栅极电阻之间的函数关系，确定栅极电阻-开通损耗系数；
    根据所述开通损耗与通态电流之间的函数关系以及栅极电阻-开通损耗系数确定所述IGBT开通损耗所对应的损耗曲线。
  8. 如权利要求7所述的方法，其中，所述IGBT开通损耗所对应的损耗曲线包括：
    Figure PCTCN2019112503-appb-100003
    其中，E on表示IGBT开通损耗，temp表示前一轮估算过程中所得到的结温，temp′表示指定结温，I c和I max分别表示通态电流和极限电流，E on_temp′表示指定结温所对应的IGBT开通损耗，
    Figure PCTCN2019112503-appb-100004
    表示栅极电阻-开通损耗系数，R g表示栅极电阻。
  9. 如权利要求4～8中任一项所述的方法，其中，确定所述IGBT模块的IGBT关断损耗功率所对应的损耗曲线的步骤包括：
    根据所述IGBT模块的给定性能数据，拟合得到指定温度、指定栅极电阻、指定GE电压以及指定CE电压下开通损耗与通态电流之间的函数关系，并拟合得到指定温度、指定GE电压、指定CE电压以及指定通态电流下关断损耗与栅极电阻之间的函数关系；
    根据所述关断损耗与通态电流之间的函数关系以及关断损耗与栅极电阻之间的函数关系，确定栅极电阻-关断损耗系数；
    根据所述关断损耗与栅极电阻之间的函数关系以及栅极电阻-关断损耗系数确定所述IGBT关断损耗所对应的损耗曲线。
  10. 如权利要求9所述的方法，其中，所述IGBT关断损耗所对应的损耗曲线包括：
    Figure PCTCN2019112503-appb-100005
    其中，E off表示IGBT关断损耗，temp表示前一轮估算过程中所得到的结温，temp′表示指定结温，I c和I max分别表示通态电流和极限电流，E off_temp′表示参考温度所对应的IGBT关断损耗，
    Figure PCTCN2019112503-appb-100006
    表示栅极电阻-开通损耗系数，R g表示栅极电阻。
  11. 如权利要求4～10中任一项所述的方法，其中，确定所述IGBT模块的FWD通态损耗所对应的损耗曲线的步骤包括：
    根据所述IGBT模块的给定性能数据，分别拟合得到第一指定温度、第二指定温度和第三指定温度下续流二极管的正向电流与正向电压之间的函数关系，其中，所述第三指定温度、第一指定温度和第二指定温度依次增大；
    根据所述正向电流与正向电压之间的函数关系分别确定第一指定温度、第二指定温度和第三指定温度下所述续流二极管的通态损耗与正向电流之间的函数关系；
    根据所述第一指定温度、第二指定温度和第三指定温度下所述续流二极管的通态损耗与正向电流之间的函数关系，确定FWD通态损耗所对应的损耗曲线。
  12. 如权利要求11所述的方法，其中，如果前一轮估算过程中所得到的结温大于或等于所述第一指定温度，则根据如下表达式确定FWD通态损耗所对应的损耗曲线：
    P D(temp,I f)＝(temp-temp1)/(temp2-temp1)*(P D_temp2-P D_temp1)+P D_temp1
    如果前一轮估算过程中所得到的结温小于所述第一指定温度，则根据如下表达式确定FWD通态损耗所对应的损耗曲线：
    P D(temp,I f)＝(temp1-temp)/(temp1-temp3)*(P D_temp3-P D_temp1)+P D_temp1
    其中，P D表示FWD通态损耗，temp表示前一轮估算过程中所得到的结温，P D_temp1、P D_temp2和P D_temp3分别表示第一指定温度temp1、第二指定温度temp2和第三指定温度temp3所对应的FWD通态损耗，I f表示正向电流。
  13. 如权利要求4～12中任一项所述的方法，其中，确定所述IGBT模块的FWD关断损耗所对应的损耗曲线的步骤包括：
    根据所述IGBT模块的给定性能数据，分别拟合得到第一指定温度和第二指定温度下续流二极管的正向电流与关断损耗之间的函数关系；
    根据所述IGBT模块的给定性能数据，拟合得到第一指定温度、指定CE电压和指定正向电流下关断损耗与栅极电阻之间的函数关系，进而得到栅极电阻-反向恢复系数；
    根据所述第一指定温度和第二指定温度下所述续流二极管的关断损耗与正向电流之间的函数关系，以及所述栅极电阻-反向恢复系数，确定FWD关断损耗所对应的损耗曲线。
  14. 如权利要求13所述的方法，其中，根据如下表达式确定FWD关断损耗所对应的损耗曲线：
    Figure PCTCN2019112503-appb-100007
    其中，E D表示FWD关断损耗，temp表示前一轮估算过程中所得到的结温，temp1 和temp2分别表示第一指定温度和第二指定温度，I f表示正向电流，E D_temp1和E D_temp2分别表示第一指定温度和第二指定温度所对应的续流二极管的关断损耗，
    Figure PCTCN2019112503-appb-100008
    表示栅极电阻-反向恢复系数，R g表示栅极电阻。
  15. 如权利要求14所述的方法，其中，根据如下表达式确定栅极电阻-反向恢复系数
    Figure PCTCN2019112503-appb-100009
    Figure PCTCN2019112503-appb-100010
    其中，f(R g)表示第一指定温度、指定CE电压和指定正向电流下关断损耗与栅极电阻之间的函数关系。
  16. 如权利要求1～15中任一项所述的方法，其中，在所述步骤二中，根据获取到的所述网络模型中各层前一次的温差以及所述首层的输入热量，确定各层当前的单层温差。
  17. 如权利要求16所述的方法，其中，对于所述网络模型中的任一层，根据如下表达式确定该层当前的单层温差：
    Figure PCTCN2019112503-appb-100011
    其中，
    Figure PCTCN2019112503-appb-100012
    表示网络模型中第m层第i次的单层温差，r m表示第m层的热阻，τ m表示第m层的热时间常数，T表示预设时间间隔，
    Figure PCTCN2019112503-appb-100013
    表示第m层第i次的输入热量，
    Figure PCTCN2019112503-appb-100014
    表示第m层第i-1次的单层温差。
  18. 如权利要求17所述的方法，其中，对于所述网络模型中的任一层的输出热量等于其下一层的输入热量，其中，根据如下表达式确定任一层的输出热量：
    Figure PCTCN2019112503-appb-100015
    其中，
    Figure PCTCN2019112503-appb-100016
    表示第m层第i次的输出热量。
  19. 如权利要求1～18中任一项所述的方法，其中，在所述步骤三中，根据如下表达式确定所述待分析功率模块的结温：
    Figure PCTCN2019112503-appb-100017
    其中，T j表示待分析功率模块的结温，T a表示环境温度，N表示网络模型所包含的层数，△temp m表示网络模型中第m层的单层温差。
PCT/CN2019/112503 2018-11-30 2019-10-22 一种功率模块温度估算方法 WO2020108172A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811455463.1A CN111339623B (zh) 2018-11-30 2018-11-30 一种功率模块温度估算方法
CN201811455463.1 2018-11-30

Publications (1)

Publication Number Publication Date
WO2020108172A1 true WO2020108172A1 (zh) 2020-06-04

Family

ID=70852696

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/112503 WO2020108172A1 (zh) 2018-11-30 2019-10-22 一种功率模块温度估算方法

Country Status (2)

Country Link
CN (1) CN111339623B (zh)
WO (1) WO2020108172A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112670962A (zh) * 2020-12-23 2021-04-16 日立楼宇技术（广州）有限公司 一种igbt过温保护方法、系统、装置及存储介质
CN113776693A (zh) * 2021-09-16 2021-12-10 浙江吉利控股集团有限公司 一种车辆电机的控制器温度估算方法、系统及汽车

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112765786B (zh) * 2020-12-31 2024-04-26 联合汽车电子有限公司 功率器件的结温估算方法、功率器件、电机控制器及计算机可读存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20170089179A (ko) * 2016-01-26 2017-08-03 엘에스산전 주식회사 고온 차단 기능이 내장된 igbt 모듈
CN107219016A (zh) * 2017-05-24 2017-09-29 湖南大学 计算igbt模块瞬态结温的方法和系统
CN107525990A (zh) * 2017-09-18 2017-12-29 天津农学院 多电平功率变换器状态监测系统及功率器件损耗计算方法
CN108319763A (zh) * 2018-01-09 2018-07-24 全球能源互联网研究院有限公司 Igbt多层热传导建模方法

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104848961A (zh) * 2015-05-14 2015-08-19 哈尔滨工业大学 基于饱和导通压降测量igbt结温的温度定标平台及实现igbt结温测量的方法
CN105574285B (zh) * 2015-12-31 2019-05-03 杭州士兰集成电路有限公司 功率模块的损耗与结温仿真系统
CN107643777B (zh) * 2016-07-20 2020-12-04 湖南中车时代电动汽车股份有限公司 用于电路板试验的环境温度调整方法及装置
CN106960085A (zh) * 2017-03-07 2017-07-18 株洲中车时代电气股份有限公司 一种在线预估igbt元件结温的方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20170089179A (ko) * 2016-01-26 2017-08-03 엘에스산전 주식회사 고온 차단 기능이 내장된 igbt 모듈
CN107219016A (zh) * 2017-05-24 2017-09-29 湖南大学 计算igbt模块瞬态结温的方法和系统
CN107525990A (zh) * 2017-09-18 2017-12-29 天津农学院 多电平功率变换器状态监测系统及功率器件损耗计算方法
CN108319763A (zh) * 2018-01-09 2018-07-24 全球能源互联网研究院有限公司 Igbt多层热传导建模方法

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112670962A (zh) * 2020-12-23 2021-04-16 日立楼宇技术（广州）有限公司 一种igbt过温保护方法、系统、装置及存储介质
CN112670962B (zh) * 2020-12-23 2023-03-24 日立楼宇技术（广州）有限公司 一种igbt过温保护方法、系统、装置及存储介质
CN113776693A (zh) * 2021-09-16 2021-12-10 浙江吉利控股集团有限公司 一种车辆电机的控制器温度估算方法、系统及汽车

Also Published As

Publication number Publication date
CN111339623A (zh) 2020-06-26
CN111339623B (zh) 2022-09-23

Similar Documents

Publication Publication Date Title
WO2020108172A1 (zh) 一种功率模块温度估算方法
US9601404B2 (en) Thermal resistance measuring method and thermal resistance measuring device
WO2017071365A1 (zh) 数字化驱动的igbt电流检测系统及其检测方法
JP6557517B2 (ja) 半導体集積回路装置および電子装置
WO2019063000A1 (zh) 半导体功率器件动态结温的实时在线预测方法
CN107944209A (zh) 一种计算光伏逆变器元器件igbt工作温度的方法
EP2998715B1 (en) Temperature sensing system for switching device
CN106257251B (zh) 估算车辆的转换器的结温的方法
US9929073B2 (en) Semiconductor device
US20140225659A1 (en) Thermal controller
US10966291B2 (en) Power conversion apparatus and power conversion method
CN111355440B (zh) 控制电路、电驱动系统、逆变器系统及其控制方法
CN109450269B (zh) 风电变流器功率模块的结温检测方法和装置、存储介质
CN108072819B (zh) Igbt的失效检测方法和装置
Ma et al. Method of junction temperature estimation and over temperature protection used for electric vehicle's IGBT power modules
JP2012252887A (ja) 外部環境からの熱量を考慮した温調制御システム
CN109120138A (zh) 用于监测半导体开关的方法和装置
Liu et al. BP neural network for non-invasive IGBT junction temperature online detection
CN104596655A (zh) 智能功率模块、功率器件及其温度检测电路和方法
BR112020001767B1 (pt) Aparelho de proteção de dispositivo e método de proteção de dispositivo
JP6511220B2 (ja) 電力変換装置及び方法
CN112765786B (zh) 功率器件的结温估算方法、功率器件、电机控制器及计算机可读存储介质
JP2020088884A (ja) 電力変換装置
CN115343589B (zh) 一种igbt结温波动快速测量方法
JP2016116411A (ja) 半導体装置の異常検出方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19889393

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19889393

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 19889393

Country of ref document: EP

Kind code of ref document: A1