WO2019063000A1 - 半导体功率器件动态结温的实时在线预测方法 - Google Patents

半导体功率器件动态结温的实时在线预测方法 Download PDF

Info

Publication number
WO2019063000A1
WO2019063000A1 PCT/CN2018/108960 CN2018108960W WO2019063000A1 WO 2019063000 A1 WO2019063000 A1 WO 2019063000A1 CN 2018108960 W CN2018108960 W CN 2018108960W WO 2019063000 A1 WO2019063000 A1 WO 2019063000A1
Authority
WO
WIPO (PCT)
Prior art keywords
power device
semiconductor power
junction temperature
loss
real
Prior art date
Application number
PCT/CN2018/108960
Other languages
English (en)
French (fr)
Inventor
姚瑱
余国军
苗茂宇
徐小军
钱巍
Original Assignee
南京埃斯顿自动化股份有限公司
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 南京埃斯顿自动化股份有限公司 filed Critical 南京埃斯顿自动化股份有限公司
Priority to US16/387,310 priority Critical patent/US11215657B2/en
Priority to EP18847181.7A priority patent/EP3492937A4/en
Publication of WO2019063000A1 publication Critical patent/WO2019063000A1/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
    • G01R31/26—Testing of individual semiconductor devices
    • G01R31/2601—Apparatus or methods therefor
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01K—MEASURING TEMPERATURE; MEASURING QUANTITY OF HEAT; THERMALLY-SENSITIVE ELEMENTS NOT OTHERWISE PROVIDED FOR
    • G01K7/00—Measuring temperature based on the use of electric or magnetic elements directly sensitive to heat ; Power supply therefor, e.g. using thermoelectric elements
    • G01K7/42—Circuits effecting compensation of thermal inertia; Circuits for predicting the stationary value of a temperature
    • G01K7/427—Temperature calculation based on spatial modeling, e.g. spatial inter- or extrapolation

Definitions

  • the invention relates to a method for predicting junction temperature prediction of a semiconductor power device, in particular to a real-time online prediction method for dynamic junction temperature of a semiconductor power device.
  • junction temperature of semiconductor power devices is the most critical parameter to determine their lifetime. Research shows that 60% of semiconductor power device failures are caused by excessive internal junction temperature. Almost all mechanical failures of power devices are caused by junction temperature fluctuations. Power cycling and thermal cycling.
  • junction temperature for semiconductor power devices there are three main methods for obtaining junction temperature for semiconductor power devices:
  • Direct measurement method directly measure the junction temperature of the internal chip of the power device by using a temperature sensor, a thermal imager, and the like.
  • a temperature sensor a thermal imager
  • the front surface of the IGBT device is opened, the chip surface of the IGBT device is completely exposed, and the IGBT device is fixed on the heat sink with small holes, in a small A thermocouple is inserted into the hole so that one end of the thermocouple is in contact with the outer casing of the IGBT device, and the other end of the thermocouple is connected to the test device.
  • the thermal resistance values of the junction of the IGBT device and the junction-to-ambient thermal resistance values are calculated.
  • the method uses a thermocouple to directly measure the junction temperature of the power device.
  • Such a method has the following disadvantages: 1
  • the power device has a relatively high integration degree, and the installation of the temperature sensor beside the chip requires relatively high process requirements and high rejection rate, and the internal chip is The inter-voltage and insulation levels also have an impact, greatly reducing the reliability of the power device.
  • Power devices with embedded temperature sensors can only be used for measurement studies and cannot be used as batch products.
  • Conventional temperature sensors have large thermal time constants and cannot be used for dynamic junction temperature acquisition of power devices. Acquisition of junction temperature.
  • the temperature sensor made for dynamic junction temperature acquisition is very small and easily damaged. It can only be used for measurement research and it is difficult to achieve productization. 3 Using a thermal imager for junction temperature measurement requires destruction of the power device and the chip. The surface is blackened. Compared with the temperature sensor, this method can obtain good response and accuracy, but it will cause irreversible damage to the packaging and insulation of the power device. This method cannot be used for productization.
  • junction temperature is obtained by indirect conversion by measuring the thermistor parameters of the power device. As described in the patent CN106199367A "A IGBT junction temperature measuring device", the junction temperature of the power device chip is determined by detecting the turn-off delay of the power switching device in combination with the on-current.
  • junction temperature there are many practical examples of indirect acquisition of junction temperature through the thermal parameters of power devices, such as gate-level turn-on voltage, gate-level Miller platform voltage, gate-level Miller platform duration, and turn-on voltage drop, etc.
  • the thermistor parameters of the power device are usually different from the electrical parameters required for the closed-loop control of the system.
  • the problem to be solved by the invention lies in the defects of the existing technology of the customer service, and a real-time online prediction method for the dynamic junction temperature of the semiconductor power device is proposed.
  • the necessary power is controlled by the closed loop of the multiplexing system.
  • the parameter sample value is discretized to calculate the loss of the internal chip of the power device.
  • the real-time junction temperature of the internal chip of the power device is obtained through online iterative calculation.
  • the technical solution proposed by the present invention for achieving the object of the invention is: a real-time online prediction method for dynamic junction temperature of a semiconductor power device, comprising the following steps:
  • Step 1 Obtain the working state parameters of the semiconductor power device, including the power module output current, the power module bus voltage, and the heat sink temperature of the power module.
  • Step 2 Obtain the semiconductor power device loss parameter, including the conduction voltage drop U ce , the turn-on loss E on and the turn-off loss E off corresponding to the semiconductor power device current I T at a given temperature.
  • Step 3 Calculate the conduction loss of the semiconductor power device, and calculate the conduction loss P DC of the semiconductor power device corresponding to the current calculation period T S according to the obtained conduction voltage drop U CE flowing through the semiconductor power device current I T . .
  • D is the on-duty of the semiconductor power device within a calculation period T S .
  • Step 4 Calculate the switching loss of the semiconductor power device, and calculate the switching loss P sw of the semiconductor power device according to the turn- on loss E on and the turn-off loss E off corresponding to the current I T of the semiconductor power device.
  • f is the switching frequency of the semiconductor power device.
  • Step 5 Calculate the total loss of the semiconductor power device. Calculate the total loss of the power device based on the conduction loss and switching loss obtained in steps 3 and 4.
  • Step 6 Calculate the temperature difference ⁇ T tjh1 between the junction of the semiconductor power device and the surface of the heat sink. According to the obtained total loss P of the semiconductor power device corresponding to the calculation period T S , using the formula for inference and simplification of the thermal resistance model of the power device, the calculation of the sum of the power devices after a calculation period T S can be accurately calculated. The temperature difference ⁇ T tjh1 of the surface of the heat sink.
  • ⁇ T tjh0 is the temperature difference between the power device junction and the surface of the heat sink after the end of the previous period
  • ⁇ is obtained by optimizing the dynamic thermal resistance curve of the power module, and is related to the calculation period T S and the dynamic thermal resistance curve. Discretization factor. To calculate the transient thermal resistance of the power module corresponding to the period Ts.
  • Step 7 Iteratively calculate the junction temperature of the power device at each moment. Steps 1 to 6 are repeated, and then the system platform controller performs an iterative operation on the result of the previous cycle calculation and then adds the real-time temperature of the heat sink to real-time online prediction to obtain the junction temperature of each semiconductor power device in the power module.
  • the specific application of real-time junction temperature T tjh of semiconductor power devices The following two applications are taken as examples to illustrate the application of the junction temperature T tjh of the semiconductor power device (the invention is effective for all applications of the real-time junction temperature T tjh of the power device, not limited to the following two), 1 according to the real-time junction temperature of the power device T Tjh can judge the running status of the power device in real time, estimate the service life of the power device, and promptly remind the customer to replace it. 2 According to the real-time junction temperature T tjh of the power device, the junction temperature protection threshold of the power device is set. When the junction temperature exceeds the threshold, the protection can be timely, so that the power device is more reliable when used.
  • the advantages of the present invention are as follows: (1) The electrical parameter sampling value necessary for the closed-loop control of the multiplexing system is taken as an input, and a software algorithm is added to the original control platform of the system, Add any additional system hardware circuitry and cost. (2) This method adopts the idea of discretized iterative calculation, which can save the processor resources to the maximum, realize online calculation, and guarantee the real-time performance of dynamic junction temperature calculation. (3) The method uses the discrete-time dynamic thermal resistance model of optimized fitting to perform iterative calculation, which ensures the real-time performance of the dynamic junction temperature calculation of the power device, and also ensures the accuracy of calculation, which can meet the protection and life. Requirements for forecasting, reliability design, etc. (4) In summary, the method can well meet the needs of practical engineering applications.
  • the electrical parameter sampling value necessary for the closed-loop control of the multiplexing system of the invention is input as an input, and a software algorithm is added on the original control platform of the system.
  • the multiplexing of the electrical parameter sampling circuit is realized, and no additional hardware development cost is required.
  • the calculation parameters of the software algorithm can be adjusted according to different measurement objects, thereby realizing different series on the basis of no increase in cost, Estimate the junction temperature of different types of power devices.
  • the invention adopts the idea of discretized iterative calculation, can save the processor resources to the maximum extent, realize online calculation, and ensure the real-time performance of dynamic junction temperature calculation while ensuring measurement accuracy.
  • the inventive innovative iterative calculation using the optimized fitting of the discretized dynamic thermal resistance model ensures the real-time performance of the dynamic junction temperature calculation of the power device while ensuring the accuracy of the calculation, and can meet the protection, life prediction, Reliability design and other requirements.
  • the invention adopts a method for directly calculating the junction temperature, does not require special treatment of the module, does not affect the performance of various aspects, and can well meet the requirements of practical engineering applications.
  • FIG. 1 is a block diagram of a real-time online prediction method for dynamic junction temperature of a semiconductor power device according to the present invention.
  • Embodiment A specific embodiment of the method of the present invention is described by taking a power device IGBT module (FF600R12ME4) as an example (the present invention is effective for all power modules, and is not limited to an IGBT module). As shown in FIG. 1, the specific implementation of the method of the present invention mainly includes the following steps:
  • Step 1 Obtain the working state parameters of the IGBT module (FF600R12ME4): power module output current, power module bus voltage and heat sink temperature of the power module.
  • the bus voltage of the power module is 600V
  • the heat sink temperature is 100 °C.
  • Step Two Get semiconductor power device IGBT module (FF600R12ME4) loss parameters, including at a given temperature, the power device flows through the IGBT current I T corresponding to the voltage drop U ce, flowing through the power device IGBT current I T corresponding opening Loss E on (IGBT) and turn-off loss E off (IGBT) .
  • IGBT Loss E on
  • IGBT turn-off loss E off
  • Step 3 Calculate the conduction loss of the semiconductor power device IGBT module (FF600R12ME4 ) , and calculate the calculation cycle according to the obtained conduction voltage drop U CE (IGBT) flowing through the semiconductor power device IGBT current I T (IGBT) .
  • T S corresponding IGBT conduction loss P DC (IGBT) .
  • D is the on-duty of the semiconductor power device within a calculation period T S .
  • Step 4 Calculate the switching loss of the semiconductor power device according to the turn-on loss corresponding to the current I T of the semiconductor power device IGBT
  • E on (IGBT) and turn-off loss E off (IGBT) can calculate the switching loss P sw (IGBT) of the IGBT of the semiconductor power device.
  • f is the switching frequency of the semiconductor power device.
  • the switching frequency of the IGBT module (FF600R12ME4) is 5000Hz.
  • Step 5 Calculate the total loss of the semiconductor power device. According to the conduction loss and switching loss obtained in the third step and the fourth step, the total loss P (IGBT) of the semiconductor power device can be obtained.
  • the total loss P (IGBT) of the semiconductor power device can be obtained.
  • Step 6 Calculate the temperature difference ⁇ T tjh1 between the junction of the semiconductor power device and the surface of the heat sink.
  • the semiconductor power device after a calculation period T S can be accurately calculated.
  • ⁇ T tjh0 IGBT
  • ⁇ is obtained by optimizing the dynamic thermal resistance curve of the power module and calculating the period T S and the dynamic thermal resistance curve.
  • Related discretization coefficients To calculate the transient thermal resistance of the power module corresponding to the period Ts.
  • ⁇ T tjh1(FF600R12ME4) ⁇ T tjh0(FF600R12ME4) +P (FF600R12ME4) *R thjc [@10ms] - ⁇ T tjh0(FF600R12ME4) * ⁇
  • ⁇ T tjh0 (FF600R12ME4) is the difference between the junction temperature of the IGBT module (FF600R12ME4) and the surface temperature of the heat sink after the end of the previous cycle is 25 °C.
  • ⁇ is obtained by optimizing the dynamic thermal resistance curve of the power module and calculating the cycle.
  • T S and the discrete coefficient of the IGBT module FF600R12ME4 dynamic thermal resistance curve, here select 0.366, R thjc [@10ms] is the transient thermal resistance of the IGBT module (FF600R12ME4) corresponding to the calculation period of 10ms is 0.015.
  • Step 7 Iteratively calculate the junction temperature of the power device at each moment. Steps 1 to 6 are repeated, and then the system platform controller performs an iterative operation on the result of the previous cycle calculation and adds it to the real-time temperature of the heat sink, so that the junction temperature of each power device in the power module can be predicted online in real time.
  • the junction temperature of the IGBT module (FF600R12ME4) can be calculated at any time.
  • the real-time online prediction method for the dynamic junction temperature of the power device of the invention adopts the electrical parameter sampling value necessary for the closed-loop control of the multiplexing system as an input, and adds a software algorithm to the original control platform of the system.
  • the multiplexing of the electrical parameter sampling circuit is realized, and no additional hardware development cost is required.
  • the calculation parameters of the software algorithm can be adjusted according to different measurement objects, thereby realizing different series on the basis of no increase in cost, Estimate the junction temperature of different types of power devices.
  • the method of the invention adopts the idea of discretization iterative calculation, can save the processor resources to the maximum extent, realize online calculation, and ensure the real-time performance of the dynamic junction temperature calculation while ensuring the measurement precision.
  • the method of the invention uses the discretized dynamic thermal resistance model of the optimized fitting to perform iterative calculation, and ensures the real-time performance of the dynamic junction temperature calculation of the power device, and also ensures the accuracy of the calculation, and can satisfy the protection, life prediction and reliability design. Wait for the request.
  • the method of the invention adopts a method for directly calculating the junction temperature, does not require special treatment of the module, does not affect the performance of various aspects, and can well meet the requirements of practical engineering applications.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Power Conversion In General (AREA)
  • Testing Of Individual Semiconductor Devices (AREA)

Abstract

一种半导体功率器件动态结温的实时在线预测方法，该方法通过复用原系统闭环控制所必须的电参数采样值，离散化计算功率器件损耗，结合优化拟合的离散化动态热阻模型，通过在线迭代计算获得功率器件的实时结温。具有如下优点：复用系统闭环控制所必须的电参数采样值作为输入，不增加任何额外的系统硬件电路和成本；采用离散化迭代计算的思想，能够最大限度的节约处理器的资源，实现在线计算，并且保证动态结温计算的实时性；使用优化拟合的离散化动态热阻模型进行迭代计算，在保证功率器件动态结温计算的实时性的同时，还保证了计算的准确性，能够满足保护、寿命预测、可靠性设计等要求且很好的契合实际工程应用的需求。

Description

半导体功率器件动态结温的实时在线预测方法 技术领域
本发明涉及一种半导体功率器件结温预测检测方法，具体说是一种半导体功率器件动态结温的实时在线预测方法。
背景技术
近年来，半导体功率器件作为工业自动化、交通运输、智能电网、新能源等领域的核心零部件，其可靠性相关的研究已经成为国际国内高等院校、科研机构及大中型高新企业研究的热点。半导体功率器件的结温是决定其寿命的最关键参数，研究表明60％的半导体功率器件失效都是由于内部结温过高造成的，几乎所有的功率器件机械失效都源于结温波动引起的功率循环和热循环。
目前在针对半导体功率器件结温的获取方法主要有三种：
(1)直接测量法：通过使用温度传感器、热成像仪等设备直接测量功率器件内部芯片的结温。如专利CN103954900A《一种测量IGBT稳态热阻值的方法》所述，将IGBT器件的正面开帽，完全露出IGBT器件的芯片表面，将IGBT器件固定在带有小孔的散热片上，在小孔中插入热偶，使热偶的一端与IGBT器件的背面的外壳接触，热偶的另一端连接测试设备。根据施加电压的电压值，管壳温度、表面温度和环境温度，分别计算得到IGBT器件的结壳热阻值和结到环境的热阻值。该方法使用热电偶对功率器件结温进行直接测量，此类方法存在以下缺点：①功率器件集成度比较高，在芯片旁边安装温度传感器需要比较高的工艺要求且废品率高，对内部芯片之间的耐压和绝缘水平也会造成一定影响，很大程度降低了功率器件的可靠性。内埋温度传感器的功率器件一般仅能够应用于测量研究，不能作为批量产品使用；② 常规温度传感器具有较大的热时间常数，无法用于功率器件动态结温度的获取，只能用于稳态结温的获取。为了动态结温获取而制作的温度传感器直径非常小，很容易损坏，只能用于测量研究，难以实现产品化；③使用热成像仪进行结温测量，需要对功率器件进行破坏，并对芯片表面进行黑化处理，相比于温度传感器，此方法虽然能够获得良好的响应和精度，但会对功率器件的封装、绝缘造成不可逆破坏，该方法同样不能用于产品化。
(2)间接测量法：通过测量功率器件的热敏电参数来间接折算获得结温。如专利CN106199367A《一种IGBT结温的测量装置》所述，通过检测功率开关器件的关断延迟，结合导通电流来判断功率器件芯片的结温。此外通过功率器件的热敏电参数来间接获取结温的实际案例还有很多，如通过门级开启电压、门级米勒平台电压、门级米勒平台持续时间和导通压降等，但此类方法也存在缺陷：①功率器件的热敏电参数通常和系统闭环控制所需的电参数不同，需要在系统硬件架构上额外增加检测电路，势必会增加硬件成本；②功率器件热敏电参数的测量必须实时、精确，对测量电路、采样电路的性能和设计提出了较高的要求，实际工程应用困难会有一定的困难；③功率器件热敏电参数与结温的线性关系通常只在较小的范围内成立，该方法难以覆盖全范围的结温获取。
(3)直接功耗计算法：通过建立功率器件功耗和热阻的数学模型，用数学计算的方法来估算模块的结温。如专利CN103956887A《风电变流器IGBT模块结温在线测试方法》所述，通过功率器件损耗的数学模型计算功率器件的损耗，结合稳态热阻计算出功率器件的稳态结温。以上专利使用稳态热阻(常数)，计算的是稳态结温度。当计算动态结温时，要用到动态热阻模型(4阶Foster模型)，直接采用上述方法计算会耗费很多CPU资源，实时性也难以保证，实际工程应用上鲜有采用此类方法进行动态结温的实时在线计算的案例。
发明内容
本发明所要解决的问题在于，客服现有技术存在的缺陷，提出了一种半导体功率器件动态结温的实时在线预测方法，在系统原有控制平台上，通过复用系统闭环控制所必须的电参数采样值，离散化计算功率器件内部芯片的损耗，结合优化拟合的离散化动态热阻模型，通过在线迭代计算获得功率器件内部芯片的实时结温。
本发明为实现发明目的所提出的技术方案是：一种半导体功率器件动态结温的实时在线预测方法，包括以下步骤：
步骤一：获取半导体功率器件的工作状态参数，包括功率模块输出电流、功率模块母线电压和功率模块所在的散热器温度。
步骤二：获取半导体功率器件损耗参数，包括在给定温度下流过半导体功率器件电流I T对应的导通压降U ce、开通损耗E on和关断损耗E off。
步骤三：计算半导体功率器件的导通损耗，根据获取的流过半导体功率器件电流I T对应的导通压降U CE，计算出这一个计算周期T S对应的半导体功率器件导通损耗P DC。
P DC＝U CE*I T*D
D为一个计算周期T S内半导体功率器件的导通占空比。
步骤四：计算半导体功率器件的开关损耗，根据获取半导体功率器件在电流I T对应的开通损耗E on和关断损耗E off，计算出半导体功率器件的开关损耗P sw。
P sw＝(E on+E off)*f
f为半导体功率器件的开关频率。
步骤五：计算半导体功率器件的总损耗。根据步骤三和步骤四求得的导通损耗和开关损 耗，求出功率器件的总损耗P
P＝P DC+P sw
步骤六：计算半导体功率器件结与散热器表面的温度差ΔT tjh1。根据获得的一个计算周期T S对应的半导体功率器件的总损耗P，使用本专利针对功率器件热阻模型进行推理和简化的公式，可以准确的计算出一个计算周期T S后功率器件的结与散热器表面的温度差ΔT tjh1。
Figure PCTCN2018108960-appb-000001
其中ΔT tjh0为上一个周期结束后功率器件结与散热器表面的温度差，λ是通过对功率模块动态热阻曲线进行优化拟合之后得到的，且与计算周期T S以及动态热阻曲线相关的离散化系数。
Figure PCTCN2018108960-appb-000002
为计算周期Ts对应的功率模块的瞬态热阻。
步骤七：迭代计算出每一时刻功率器件的结温。重复步骤一～步骤六，然后通过系统平台控制器对上一个周期计算的结果进行迭代运算之后与散热器的实时温度相加，实时在线预测得到功率模块内每一个半导体功率器件的结温。
半导体功率器件实时结温T tjh的具体应用。下面以其中两种应用为例阐述半导体功率器件的结温T tjh的应用(本发明对功率器件实时结温T tjh的所有应用有效，不局限以下两种)，①根据功率器件实时结温T tjh可以实时判断功率器件的运行状态，估算功率器件的使用寿命，及时提醒客户进行更换。②根据功率器件实时结温T tjh，设定功率器件的结温保护阀值，当结温超过阀值能够及时保护，使功率器件在使用时更加的可靠。
本发明的技术效果如下：
相比于现有的功率器件结温获取方法，本发明的优势如下：(1)复用系统闭环控制所必须的电参数采样值作为输入，在系统原有控制平台上添加软件算法实现，不增加任何额外的 系统硬件电路和成本。(2)本方法采用离散化迭代计算的思想，能够最大限度的节约处理器的资源，实现在线计算，并且保证动态结温计算的实时性。(3)本方法创新性的使用优化拟合的离散化动态热阻模型进行迭代计算，在保证功率器件动态结温计算的实时性的同时，还保证了计算的准确性，能够满足保护、寿命预测、可靠性设计等要求。(4)综上所述，本方法能够很好的契合实际工程应用的需求。
1本发明复用系统闭环控制所必须的电参数采样值作为输入，在系统原有控制平台上添加软件算法实现。一方面实现电参数采样电路的复用，不需要额外的增加硬件开发成本，另一方面，可以根据测量对象的不同调节软件算法的计算参数，从而实现在不增加成本的基础上对不同系列、不同型号的功率器件的结温进行估算。
2本发明采用离散化迭代计算的思想，在保证测量精度的情况下，能够最大限度的节约处理器的资源，实现在线计算，并且保证动态结温计算的实时性。
3本发明创新性的使用优化拟合的离散化动态热阻模型进行迭代计算，在保证功率器件动态结温计算的实时性的同时，还保证了计算的准确性，能够满足保护、寿命预测、可靠性设计等要求。
4本发明采用直接计算结温的方法，不需要对模块进行特殊处理，不会影响各个方面的性能，能够很好的契合实际工程应用的需求。
附图说明
图1是本发明半导体功率器件动态结温的实时在线预测方法程序框图。
具体实施方式
下面结合实施例对本发明方法做进一步详细说明。
实施例：以功率器件IGBT模块(FF600R12ME4)为例说明本发明方法的具体实施方式(本发明对所有功率模块有效，不局限于IGBT模块)。如图1所示，本发明方法的具体实施主要包括以下步骤：
步骤一：获取IGBT模块(FF600R12ME4)的工作状态参数：功率模块输出电流、功率模块母线电压和功率模块所在的散热器温度。
如：IGBT模块(FF600R12ME4)在某一个时间周期T S＝10ms的平均电流I T＝300A,功率模块的母线电压为600V，散热器温度为100℃。
步骤二：获取半导体功率器件IGBT模块(FF600R12ME4)损耗参数，包括在给定温度下，流过功率器件IGBT电流I T对应的导通压降U ce，流过功率器件IGBT电流I T对应的开通损耗E on(IGBT)和关断损耗E off(IGBT)。
如：在给定温度下，流过IGBT模块(FF600R12ME4)电流I T＝300A对应的IGBT导通压降U CE(FF600R12ME4)＝1.4V，对应的开通损耗E on(FF600R12ME4)＝0.033J,对应的关断损耗E off(FF600R12ME4)＝0.034J。
步骤三：计算半导体功率器件IGBT模块(FF600R12ME4)的导通损耗，根据获取的流过半导体功率器件IGBT电流I T(IGBT)对应的导通压降U CE(IGBT)，计算出这一个计算周期T S对应的IGBT导通损耗P DC(IGBT)。
P DC(IGBT)＝U CE(IGBT)*I T*D
D为一个计算周期T S内半导体功率器件的导通占空比。
如：计算IGBT模块(FF600R12ME4)的导通损耗
P DC(FF600R12ME4)＝U CE(FF600R12ME4)*I T*D＝1.4*300*0.5＝210W
在T S＝10ms这个计算周期内IGBT模块(FF600R12ME4)的导通占空比为0.5。
步骤四：计算半导体功率器件的开关损耗，根据获取半导体功率器件IGBT在电流I T对应的开通损耗
E on(IGBT)和关断损耗E off(IGBT)可以计算出半导体功率器件的IGBT的开关损耗P sw(IGBT)。
P sw(IGBT)＝[E on(IGBT)+E off(IGBT)]*f
f为半导体功率器件的开关频率。
如：计算半导体功率器件IGBT模块(FF600R12ME4)的开关损耗
P sw(FF600R12ME4)＝[E on(FF600R12ME4)+E off(FF600R12ME4)]*f＝(0.034+0.033)*5000
＝335W
IGBT模块(FF600R12ME4)工作的开关频率为5000Hz。
步骤五：计算半导体功率器件的总损耗。根据步骤三和步骤四求得的导通损耗和开关损耗，可以求出，半导体功率器件的总损耗P (IGBT)
P (IGBT)＝P DC(IGBT)+P sw(IGBT)
如：计算半导体功率器件IGBT模块(FF600R12ME4)的总损耗
P (FF600R12ME4)＝P DC(FF600R12ME4)+P sw(FF600R12ME4)＝210+335＝545W
步骤六：计算半导体功率器件结与散热器表面的温度差ΔT tjh1。根据获得的一个计算周期T S对应的IGBT的总损耗P (IGBT)，使用本专利针对半导体功率器件热阻模型进行推理和简化的公式，可以准确的计算出一个计算周期T S后半导体功率器件IGBT的结温与散热器表面温度的差ΔT tjh1(IGBT)。
Figure PCTCN2018108960-appb-000003
其中ΔT tjh0(IGBT)为上一个周期结束后IGBT结温与散热器表面温度的差，λ是通过对功 率模块动态热阻曲线进行优化拟合之后得到的与计算周期T S以及动态热阻曲线相关的离散化系数,
Figure PCTCN2018108960-appb-000004
为计算周期Ts对应的功率模块的瞬态热阻。
如：计算一个计算周期T S＝10ms之后IGBT模块(FF600R12ME4)结与散热器表面的温度差ΔT tjh1。
ΔT tjh1(FF600R12ME4)＝ΔT tjh0(FF600R12ME4)+P (FF600R12ME4)*R thjc[@10ms]-ΔT tjh0(FF600R12ME4)*λ
＝25+545*0.015-25*0.366＝24.025
其中ΔT tjh0(FF600R12ME4)为上一个周期结束后IGBT模块(FF600R12ME4)结温与散热器表面温度的差为25℃，λ是通过对功率模块动态热阻曲线进行优化拟合之后得到的与计算周期T S以及IGBT模块FF600R12ME4动态热阻曲线相关的离散化系数，在这里选择0.366,R thjc[@10ms]为计算周期10ms对应的IGBT模块(FF600R12ME4)的瞬态热阻为0.015。
步骤七：迭代计算出每一时刻功率器件的结温。重复步骤一～步骤六，然后通过系统平台控制器对上一个周期计算的结果进行迭代运算之后与散热器的实时温度相加，可以实时在线预测功率模块内每一个功率器件的结温。
如:计算这个10ms计算周期后IGBT模块(FF600R12ME4)的结温
T tjh1(FF600R12ME4)＝24.025+100℃＝124.025℃
按照此种方法可以计算出任何时刻IGBT模块(FF600R12ME4)的结温。
本发明功率器件动态结温的实时在线预测方法，通过复用系统闭环控制所必须的电参数采样值作为输入，在系统原有控制平台上添加软件算法实现。一方面实现电参数采样电路的复用，不需要额外的增加硬件开发成本，另一方面，可以根据测量对象的不同调节软件算法的计算参数，从而实现在不增加成本的基础上对不同系列、不同型号的功率器件的结温进行估算。
本发明方法采用离散化迭代计算的思想，在保证测量精度的情况下，能够最大限度的节约处理器的资源，实现在线计算，并且保证动态结温计算的实时性。
本发明方法使用优化拟合的离散化动态热阻模型进行迭代计算，在保证功率器件动态结温计算的实时性的同时，还保证了计算的准确性，能够满足保护、寿命预测、可靠性设计等要求。
本发明方法采用直接计算结温的方法，不需要对模块进行特殊处理，不会影响各个方面的性能，能够很好的契合实际工程应用的需求。

Claims (1)

  1. 一种半导体功率器件动态结温的实时在线预测方法，包括以下步骤：
    步骤一：获取半导体功率器件的工作状态参数：功率模块输出电流、功率模块母线电压和功率模块所在的散热器温度；
    步骤二：获取半导体功率器件损耗参数：在给定温度下流过半导体功率器件电流I T对应的导通压降U CE、开通损耗E on和关断损耗E off；
    步骤三：根据获取的流过半导体功率器件电流I T对应的导通压降U CE，计算出这一个计算周期T S对应的半导体功率器件导通损耗P DC：
    P DC＝U CE*I T*D
    D为一个计算周期T S内半导体功率器件的导通占空比；
    步骤四：根据获取半导体功率器件在电流I T对应的开通损耗E on和关断损耗E off，计算出半导体功率器件的开关损耗P sw：
    P sw＝(E on+E off)*f
    f为半导体功率器件的开关频率；
    步骤五：根据步骤三和步骤四求得的导通损耗和开关损耗，求出半导体功率器件的总损耗P：
    P＝P DC+P sw；
    步骤六：计算半导体功率器件结与散热器表面的温度差ΔT tjh1：
    Figure PCTCN2018108960-appb-100001
    其中：ΔT tjh0为上一个周期结束后功率器件结与散热器表面的温度差，λ是通过对功率模块动态热阻曲线进行优化拟合之后得到的，且与计算周期T S以及动态热阻曲线相关的离散 化系数；
    Figure PCTCN2018108960-appb-100002
    为计算周期Ts对应的功率模块的瞬态热阻；
    步骤七：迭代计算出每一时刻功率器件的结温：重复步骤一～步骤六，然后对上一个周期计算的结果进行迭代运算之后与散热器的实时温度相加，实时在线预测得到功率模块内每一个半导体功率器件的结温。
PCT/CN2018/108960 2017-12-06 2018-09-30 半导体功率器件动态结温的实时在线预测方法 WO2019063000A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US16/387,310 US11215657B2 (en) 2017-12-06 2018-09-30 Real-time online prediction method for dynamic junction temperature of semiconductor power device
EP18847181.7A EP3492937A4 (en) 2017-12-06 2018-09-30 METHOD FOR REAL-TIME LINE PREDICTION OF DYNAMIC JUNCTION TEMPERATURE OF SEMICONDUCTOR ELECTRICAL DEVICE

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711273620.2 2017-12-06
CN201711273620.2A CN108072821B (zh) 2017-12-06 2017-12-06 半导体功率器件动态结温的实时在线预测方法

Publications (1)

Publication Number Publication Date
WO2019063000A1 true WO2019063000A1 (zh) 2019-04-04

Family

ID=62157959

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/108960 WO2019063000A1 (zh) 2017-12-06 2018-09-30 半导体功率器件动态结温的实时在线预测方法

Country Status (4)

Country Link
US (1) US11215657B2 (zh)
EP (1) EP3492937A4 (zh)
CN (1) CN108072821B (zh)
WO (1) WO2019063000A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110147578A (zh) * 2019-04-19 2019-08-20 西安中车永电电气有限公司 基于半实物仿真平台的igbt器件的寿命预测方法
CN111260113A (zh) * 2020-01-08 2020-06-09 西安工程大学 SiC MOSFET模块全生命周期结温在线预测方法
CN113030683A (zh) * 2021-03-15 2021-06-25 五羊—本田摩托（广州）有限公司 一种测量功率开关器件温度的方法、介质及计算机设备
CN114441923A (zh) * 2022-04-08 2022-05-06 北京智芯微电子科技有限公司 瞬态热阻的模拟系统与方法

Families Citing this family (26)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108072821B (zh) 2017-12-06 2018-11-16 南京埃斯顿自动控制技术有限公司 半导体功率器件动态结温的实时在线预测方法
EP3627121B1 (en) * 2018-09-21 2022-07-06 Maschinenfabrik Reinhausen GmbH Determining a characteristic temperature of an electric or electronic system
CN111024260B (zh) * 2018-10-09 2021-08-17 上海汽车集团股份有限公司 一种功率半导体器件热点温度估计方法和装置
CN109193572B (zh) * 2018-10-11 2020-06-30 四川长虹电器股份有限公司 Igbt的保护控制方法
EP3696558A1 (de) * 2019-02-15 2020-08-19 Siemens Aktiengesellschaft Vorrichtung und verfahren zur automatischen prüfung eines schaltorgans
CN110011276B (zh) * 2019-04-08 2020-12-04 南京埃斯顿自动化股份有限公司 一种基于寿命衰减速率控制的芯片热保护方法
CN110471012B (zh) * 2019-06-03 2022-05-27 上海联影医疗科技股份有限公司 功率开关器件温度的预测方法和装置、存储介质及医疗设备
CN110502720B (zh) * 2019-08-26 2024-02-09 阳光电源股份有限公司 功率半导体模块的损耗在线计算方法及其应用方法和装置
CN111293671B (zh) * 2020-02-07 2021-01-08 山东大学 基于结温预测的功率器件热保护与预警方法及系统
JP7361624B2 (ja) * 2020-02-12 2023-10-16 東京エレクトロン株式会社 加熱源の寿命推定システム、寿命推定方法、および検査装置
CN111665430A (zh) * 2020-03-27 2020-09-15 厦门市三安集成电路有限公司 一种GaN HEMT器件的热可靠性评估方法
CN111783287B (zh) * 2020-06-19 2022-10-04 合肥工业大学 三相igbt功率模块的在线结温计算方法
CN112014707A (zh) * 2020-07-13 2020-12-01 北京工业大学 一种功率循环实验中测量与控制SiC功率VDMOS器件结温的方法
CN112014708B (zh) * 2020-07-27 2023-02-07 西安中车永电电气有限公司 基于FPGA的SiC功率器件在线结温计算方法
CN111896799B (zh) * 2020-08-05 2023-08-08 合肥零碳技术有限公司 一种功率器件平均损耗的计算方法及装置
CN112001127B (zh) * 2020-08-28 2022-03-25 河北工业大学 一种igbt结温预测的方法
CN112327130A (zh) * 2020-09-07 2021-02-05 珠海格力电器股份有限公司 一种开关管工作结温实时测量装置及方法
CN112162186A (zh) * 2020-09-25 2021-01-01 华电(烟台)功率半导体技术研究院有限公司 一种自我校准的电力电子器件温度系数校准方法
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
CN112858866A (zh) * 2021-01-20 2021-05-28 合肥工业大学 一种基于米勒平台持续时间的igbt结温监测系统和方法
CN113514166B (zh) * 2021-03-03 2024-06-18 中国南方电网有限责任公司超高压输电公司天生桥局 一种hvdc换流阀晶闸管温度监测方法和系统
CN113258912B (zh) * 2021-04-29 2024-05-03 珠海格力电器股份有限公司 一种开关管的控制装置、方法和电器设备
CN113435151B (zh) * 2021-07-19 2023-06-13 西安热工研究院有限公司 一种运行过程中igbt结温的预测系统及方法
CN113971101B (zh) * 2021-10-15 2023-06-02 飞腾信息技术有限公司 一种服务器温度故障诊断方法、装置、存储介质及系统
CN114384391B (zh) * 2022-03-23 2022-07-05 北京宏景智驾科技有限公司 一种域控制器主要芯片结温估算的方法
CN117034786B (zh) * 2023-10-09 2024-01-05 山东芯赛思电子科技有限公司 一种igbt结温预测方法

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040066837A1 (en) * 2002-10-04 2004-04-08 Armour Joshua W. Method and apparatus for providing accurate junction temperature in an integrated circuit
CN103954900A (zh) * 2014-05-08 2014-07-30 江苏中科君芯科技有限公司 一种测量igbt稳态热阻值的方法
CN103956887A (zh) * 2014-05-15 2014-07-30 重庆大学 风电变流器igbt模块结温在线计算方法
US20160328512A1 (en) * 2014-11-17 2016-11-10 Teledyne Scientific & Imaging, Llc Method of modeling the temperature profile of an ic transistor junction
CN106199367A (zh) * 2016-07-06 2016-12-07 中国科学院电工研究所 一种igbt结温测量装置
CN106610445A (zh) * 2015-10-27 2017-05-03 全球能源互联网研究院 一种数字化驱动的igbt电流检测系统及其检测方法
CN108072821A (zh) * 2017-12-06 2018-05-25 南京埃斯顿自动控制技术有限公司 半导体功率器件动态结温的实时在线预测方法

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7356441B2 (en) * 2005-09-28 2008-04-08 Rockwell Automation Technologies, Inc. Junction temperature prediction method and apparatus for use in a power conversion module
US10337930B2 (en) * 2015-05-12 2019-07-02 GM Global Technology Operations LLC Online IGBT junction temperature estimation without the use of a dedicated temperature estimation or measurement device
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
CN106849011B (zh) * 2016-12-30 2019-03-15 南京埃斯顿自动控制技术有限公司 一种伺服电机过热保护方法

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040066837A1 (en) * 2002-10-04 2004-04-08 Armour Joshua W. Method and apparatus for providing accurate junction temperature in an integrated circuit
CN103954900A (zh) * 2014-05-08 2014-07-30 江苏中科君芯科技有限公司 一种测量igbt稳态热阻值的方法
CN103956887A (zh) * 2014-05-15 2014-07-30 重庆大学 风电变流器igbt模块结温在线计算方法
US20160328512A1 (en) * 2014-11-17 2016-11-10 Teledyne Scientific & Imaging, Llc Method of modeling the temperature profile of an ic transistor junction
CN106610445A (zh) * 2015-10-27 2017-05-03 全球能源互联网研究院 一种数字化驱动的igbt电流检测系统及其检测方法
CN106199367A (zh) * 2016-07-06 2016-12-07 中国科学院电工研究所 一种igbt结温测量装置
CN108072821A (zh) * 2017-12-06 2018-05-25 南京埃斯顿自动控制技术有限公司 半导体功率器件动态结温的实时在线预测方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3492937A4 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110147578A (zh) * 2019-04-19 2019-08-20 西安中车永电电气有限公司 基于半实物仿真平台的igbt器件的寿命预测方法
CN110147578B (zh) * 2019-04-19 2022-12-06 西安中车永电电气有限公司 基于半实物仿真平台的igbt器件的寿命预测方法
CN111260113A (zh) * 2020-01-08 2020-06-09 西安工程大学 SiC MOSFET模块全生命周期结温在线预测方法
CN113030683A (zh) * 2021-03-15 2021-06-25 五羊—本田摩托（广州）有限公司 一种测量功率开关器件温度的方法、介质及计算机设备
CN114441923A (zh) * 2022-04-08 2022-05-06 北京智芯微电子科技有限公司 瞬态热阻的模拟系统与方法
CN114441923B (zh) * 2022-04-08 2022-06-28 北京智芯微电子科技有限公司 瞬态热阻的模拟系统与方法

Also Published As

Publication number Publication date
CN108072821A (zh) 2018-05-25
EP3492937A4 (en) 2019-08-28
EP3492937A1 (en) 2019-06-05
US20190377023A1 (en) 2019-12-12
CN108072821B (zh) 2018-11-16
US11215657B2 (en) 2022-01-04

Similar Documents

Publication Publication Date Title
WO2019063000A1 (zh) 半导体功率器件动态结温的实时在线预测方法
CN106443400B (zh) 一种igbt模块的电-热-老化结温计算模型建立方法
Wu et al. A temperature-dependent thermal model of IGBT modules suitable for circuit-level simulations
CN110161398B (zh) 一种利用壳温评估igbt功率模块老化状态的方法
CN104458039B (zh) Igbt模块壳温的实时估算方法
CN107944209A (zh) 一种计算光伏逆变器元器件igbt工作温度的方法
Zhou et al. Dynamic junction temperature estimation via built-in negative thermal coefficient (NTC) thermistor in high power IGBT modules
CN110658435B (zh) 一种igbt结温监测装置及方法
CN112364591B (zh) SiC MOSFET结温实时动态预测方法及电压检测电路
CN108680847A (zh) 基于故障电流下的igbt结温的热计算方法
Yang et al. A temperature-dependent Cauer model simulation of IGBT module with analytical thermal impedance characterization
Chen et al. Electrothermal-based junction temperature estimation model for converter of switched reluctance motor drive system
Lai et al. Condition monitoring in a power module using on-state resistance and case temperature
CN110020447A (zh) 电动汽车控制器中功率元件igbt结温的估算方法
Liu et al. Simplified junction temperature estimation using integrated NTC sensor for SiC modules
CN111060798B (zh) 一种mos管自动功率老化测试系统及测试方法
Xu et al. Full-time junction temperature extraction of IGCT based on electrothermal model and TSEP method for high-power applications
Shao et al. On the junction temperature extraction approach with a hybrid model of voltage-rise time and voltage-rise loss
Li et al. Measurement method of the IGBT chip temperature fluctuation based on electrothermal model derivation
Musallam et al. Real-time compact electronic thermal modelling for health monitoring
Wang et al. Fast and accurate temperature estimation of three-level IGBT converter based on 3-D coupled thermal model
Li et al. Sustainable energy saving: A junction temperature numerical calculation method for power insulated gate bipolar transistor module
Zhang et al. Thermal network parameters identifying during the cooling procedure of IGBT module
Sun et al. Junction temperature estimation in IGBT power modules based on Kalman filter
Peng et al. Method for obtaining junction temperature of power semiconductor devices combining computational fluid dynamics and thermal network

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2018847181

Country of ref document: EP

Effective date: 20190226

ENP Entry into the national phase

Ref document number: 2018847181

Country of ref document: EP

Effective date: 20190226

NENP Non-entry into the national phase

Ref country code: DE