A method for evaluating the life of a motor controller

Abstract

The invention discloses a method for evaluating the life of a motor controller, which includes the following steps: S1, estimating the junction temperature of an IGBT power module; S2, uploading the estimated junction temperature to a new energy vehicle big data platform; S3, obtaining the number of power cycles through classification and statistical processing; S4, substituting the Arrhenius life model formula to obtain the number of failure power cycles; S5, calculating the current life of the motor controller D based on the Miner fatigue damage accumulation method; S6, judging whether the difference between the current life of the motor controller D and the life of the last calculation cycle D' is greater than the update frequency D0, if D-D'≥D0, updating and correcting the junction temperature value, and returning to S1 for a new round of motor controller life evaluation; if D-D'<D0, using the current life of the motor controller obtained in S5 as the motor controller life evaluation value. By introducing the life attenuation factor, the IGBT junction temperature can be estimated more accurately, iterated with the life evaluation, and better fault warning and safety monitoring can be implemented for new energy vehicles.

Description

Technical Field

[0001] The present invention relates to a motor controller of a new energy vehicle, and in particular to a life assessment method for a motor controller. Background Art

[0002] New energy vehicles have achieved rapid development in recent years. As the core component of the electric drive system, the reliability of the motor controller is crucial to the safe operation of new energy vehicles. Therefore, it is necessary to accurately evaluate the life of the motor controller and replace or maintain the motor controller that is approaching the end of its life to ensure the safe operation and power output of new energy vehicles. The IGBT power module is the key core of the motor controller, see Figure 1 The main components of the IGBT power module are: IGBT chip, solder layer, copper-clad ceramic substrate, and base plate. The main component of the IGBT chip is SiO2. The copper-clad ceramic substrate is usually made of Al2O3 or AIN, which has the characteristics of high thermal conductivity and high electrical insulation. The base plate is generally made of copper plate. When the IGBT power module is working, the heat emitted by the semiconductor chip is transferred downward through the layers inside the module. Due to the different expansion coefficients of the materials of each layer, the degree of thermal expansion of each layer will be different, which will generate thermomechanical stress between the layers, causing deformation of the materials of each layer of the package. If the temperature continues to fluctuate, deformation will continue to occur between the layers, and the accumulated thermal fatigue damage will eventually lead to the failure of the IGBT module. In short, the life assessment of the motor controller is the life assessment of the IGBT power module.

[0003] IGBT junction temperature refers to the highest temperature of the IGBT chip during operation. It is a key parameter that affects the life of the IGBT power module. In order to study the influence of IGBT junction temperature on the life of the IGBT power module, analyze the quantitative relationship between junction temperature and the life of the IGBT power module, and develop an IGBT power module life assessment method based on an analytical model. The analytical model method first conducts an accelerated aging test under high stress level, fits the aging parameters in the Arrhenius life model according to the number of failures, and then uses the Arrhenius life model to extrapolate the IGBT life under normal stress level. Then, the rain flow counting method is used to statistically analyze the junction temperature data over a period of time, and the irregular power fluctuations in actual applications are classified according to the average junction temperature and junction temperature fluctuations. The classified data is brought into the life model to calculate the life, and finally the Miner fatigue damage accumulation method is used to calculate the life corresponding to the junction temperature change over a period of time.

[0004] The life assessment of IGBT power modules depends on accurate junction temperature detection. In actual application scenarios, IGBT power modules are encapsulated in motor controllers, and the junction temperature cannot be directly obtained through sensors. Therefore, the IGBT junction temperature is generally obtained by estimation, and the thermal network method is a commonly used junction temperature estimation method. However, the thermal network method does not take into account the increase in thermal resistance caused by the accumulated thermal fatigue damage of the IGBT power module, that is, the thermal network parameters are not updated with the change of the IGBT state. As the IGBT works for a long time, the more thermal fatigue is accumulated, the greater the deviation between the estimated junction temperature of the IGBT and the true value, which ultimately leads to the less accurate life assessment of the IGBT power module. Different thermal fatigue damages correspond to different IGBT lifespans, which means that the IGBT junction temperature and the IGBT lifespan affect each other. The present invention uses the IGBT lifespan as a reference dimension, updates the junction temperature estimation parameters online, and then uses the new parameters for the IGBT junction temperature estimation, and then carries out a new round of life assessment. The frequency of online updating of the IGBT junction temperature estimation parameters can be adjusted according to the requirements of the accuracy of the life assessment.

[0005] CN110514979A discloses a method for predicting the life of an IGBT module in a subway vehicle traction drive system, but does not consider the thermal resistance Z in the thermal network model. th The situation changes with the life of the IGBT module. Generally speaking, due to the different expansion coefficients of each layer of the IGBT, the degree of thermal expansion of each layer will be different, which will generate thermomechanical stress between the layers, causing the deformation of the packaging materials, and the final result is the thermal resistance Z th Increase. In the method of using a thermal network to estimate junction temperature, even if the IGBT operates under the same operating conditions, once the thermal resistance changes, the IGBT junction temperature will also change. If the change in thermal resistance as the IGBT life decays is not considered, the junction temperature estimation result will be smaller than the actual result, resulting in an overestimation of the IGBT life assessment result.

[0006] CN114004145A discloses a method for estimating the life of an IGBT based on a radial basis neural network. This solution has three problems. The first is that a neural network junction temperature estimation method is used. The neural network relies on a large amount of sample data and requires obvious differentiation between the sample data, otherwise it will affect the prediction accuracy of the neural network. A large amount of non-repetitive and differentiated test data is difficult to obtain in practical applications. Second, the solution mentions that life prediction is performed under the same working conditions, but under the actual irregular operating conditions, it is difficult to extract standard equal working conditions. In actual operation, it may be necessary to apply standard equal working conditions at intervals to predict life, and this operation mode is not real-time. Third, the number of cycles is represented as the service life of the device, which is meaningless. Because in actual applications, IGBTs operate under irregular working conditions, not at fixed stress levels and cycle times. It is necessary to consider how to correspond irregular working conditions in actual applications with standard cycle conditions. Summary of the invention

[0007] The purpose of the present invention is to provide a motor controller life assessment method, which can more accurately estimate the IGBT junction temperature by introducing the life attenuation factor, and iterate with the life assessment, so as to better implement fault warning and safety monitoring for new energy vehicles.

[0008] The motor controller life assessment method of the present invention comprises the following steps:

[0009] S1, estimate the junction temperature based on the loss and thermal resistance of the IGBT to obtain the junction temperature T of the IGBT power module when the motor controller is working 估算 ;

[0010] S2, the junction temperature T obtained by S1 估算 Upload to the new energy vehicle big data platform;

[0011] S3, using the rain flow counting method, according to the average junction temperature T m The junction temperature fluctuation ΔT is used to classify and statistically process multiple junction temperature data on the new energy vehicle big data platform to obtain the average junction temperature T m The number of power cycles N under the junction temperature fluctuation ΔT cyc (ΔT,T m );

[0012] S4, the average junction temperature T m Substitute the junction temperature fluctuation ΔT into the Arrhenius life model formula to obtain the average junction temperature T m The number of power cycles N under the failure temperature fluctuation ΔT f (ΔT,T m );

[0013] S5, calculate the current life D of the motor controller based on Miner fatigue damage accumulation method;

[0014] S6, determine whether the difference between the current life D of the motor controller and the life D' of the previous calculation cycle is greater than the update frequency D0. If D-D'≥D0, then according to the formula T 修正 =T 估算 h updates and corrects the junction temperature value, where h is the life attenuation factor, and returns to S1 for a new round of motor controller life evaluation. The current life D of the motor controller obtained by S5 in the next calculation cycle is used as the motor controller life evaluation value;

[0015] If D-D'<D0, the current life D of the motor controller obtained in S5 is used as the life evaluation value of the motor controller.

[0016] Furthermore, the junction temperature T 估算 The calculation formula is: 估算 =P loss Z th +T c ;

[0017] Where P loss is the IGBT loss, Z th is the thermal resistance, which can be obtained from the IGBT data sheet. c The IGBT shell temperature is obtained in real time through the built-in thermocouple sensor;

[0018] The calculation formula for IGBT loss is P loss =P cond +P sw ,

[0019] P cond is the conduction loss, V ce is the on-state voltage drop, I c is the collector current, m is the modulation ratio, cosφ is the power factor, R T is the internal resistance of IGBT;

[0020] P sw is the switching loss, f sw is the switching frequency, V dc is the bus voltage, I m is the load current, E on To open the energy, E off is the turn-off energy, V nom is the rated voltage, I nom is the rated current;

[0021] The conduction voltage drop Vce , collector current I c , switching frequency f sw , bus voltage V dc and load current I m As an input parameter, the IGBT junction temperature T 估算 Estimates.

[0022] Further, the processing flow of the rain flow counting method in S3 is as follows: when a maximum value is identified using the rain flow counting method, and when a maximum value already exists in the maximum value stack, it is compared with the new maximum value, otherwise, only the new maximum value is stored in the stack; if the new maximum value is greater than the first value in the maximum value stack, the number of values ​​existing in the minimum value stack is detected;

[0023] If the number of minimum value stacks = 1, an equivalent half cycle is identified, the junction temperature fluctuation ΔT = old maximum value - old minimum value, and the average junction temperature T m = old minimum + 1 / 2ΔT, and the old maximum is replaced by the new maximum;

[0024] If the number of minimum value stacks is ≥ 2, an equivalent full cycle is identified, and the junction temperature fluctuation ΔT = new minimum value - old maximum value. At the same time, the old maximum value is replaced by the new maximum value, and the latest minimum value is removed;

[0025] When the new maximum value is smaller than the maximum value stack, the new maximum value is stored on the left side of the maximum value stack as the first value in the maximum value stack;

[0026] After completing the above operations, if there is still more than one value in the maximum value stack, repeat the recursive operation; finally, the average junction temperature T m The number of power cycles of the equivalent half cycle and the equivalent full cycle under the junction temperature fluctuation ΔT.

[0027] Furthermore, the Arrhenius lifespan model formula in S4 is Where A and α are model parameters, which are obtained by accelerated life test combined with maximum likelihood method or least squares method; E a is the activation energy related to the material, and its value is 9.89×10 -20 J;k B is the Boltzmann constant, which is 1.38×10 -23 J.K. -1 .

[0028] Furthermore, the calculation formula for the current life of the motor controller in S5 is:

[0029] Furthermore, the calculation formula of the life attenuation factor in S6 is h = exp(m(T估算 )·D), D is the current life of the motor controller calculated by S5, m(T 估算 ) is the life adjustment coefficient. The higher the junction temperature of the IGBT power module, the smaller the life adjustment coefficient.

[0030] Furthermore, the life adjustment coefficient is obtained through calibration.

[0031] The present invention first estimates the junction temperature according to the loss and thermal resistance of the IGBT, and then uploads the estimated junction temperature to the new energy vehicle big data platform during the working process of the motor controller, obtains the number of power cycles after classification and statistical processing, and obtains the number of failure power cycles through the Arrhenius life model formula, and then uses the Miner fatigue damage accumulation method to calculate the current life of the motor controller within a certain period of time. Finally, the calculated current life is used as a reference dimension, the estimated parameters are updated online, and the new parameters are used for IGBT junction temperature estimation, and then a new round of life evaluation is carried out. The frequency of online updating of the IGBT junction temperature estimation parameters is adjusted according to the accuracy requirements of the life evaluation. Generally speaking, the life of the IGBT will decay after experiencing long-term working conditions, and the junction temperature will increase under the same working conditions. When evaluating the life of the motor controller, if the junction temperature increase caused by life decay is not considered, a longer motor controller life will be calculated, which deviates from the actual situation. The present invention introduces a life decay factor in the junction temperature estimation method, which can more accurately estimate the IGBT junction temperature, and iterate with the life evaluation at the same time, so as to better implement fault warning and safety monitoring for new energy vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 is the thermal expansion coefficient of each layer material of the IGBT power module;

[0033] Figure 2 A flowchart of a motor controller life assessment method;

[0034] Figure 3 Flow chart for calculating junction temperature data using the rain flow counting method;

[0035] Figure 4 This is the flow chart of life assessment using the Miner fatigue damage accumulation method;

[0036] Figure 5 Updated flowchart for life decay factor. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0038] See also Figure 2 , the motor controller life assessment method shown includes the following steps:

[0039] S1, estimate the junction temperature based on the loss and thermal resistance of the IGBT to obtain the junction temperature T of the IGBT power module when the motor controller is working 估算 . Junction temperature T 估算 The calculation formula is: 估算 =P loss Z th +T c .

[0040] Where P loss is the IGBT loss, Z th is the thermal resistance, which can be obtained from the IGBT data sheet. c It is the IGBT case temperature, which is obtained in real time through the built-in thermocouple sensor.

[0041] The calculation formula for IGBT loss is P loss =P cond +P sw ,

[0042] P cond is the conduction loss, V ce is the on-state voltage drop, I c is the collector current, m is the modulation ratio, cosφ is the power factor, R T is the internal resistance of IGBT. ce and collector current I c Extracted through the acquisition circuit. Modulation ratio m, power factor cosφ, IGBT internal resistance R T It is a relatively fixed parameter that does not change much in practical applications and is obtained through calibration tests.

[0043] P sw is the switching loss, f sw is the switching frequency, V dc is the bus voltage, I m is the load current, E on To open the energy, E off is the turn-off energy, Vnom is the rated voltage, I nom is the rated current. Switching frequency f sw , bus voltage V dc and load current I m During the operation of the motor controller, the energy is collected in real time by sensors and calculated in combination with the control algorithm. on , Shutoff energy E off It is a characteristic parameter related to the IGBT device itself, measured through a double pulse test, and obtained by looking up the test data in a table in actual application. Rated voltage V related to IGBT operation nom and rated current I nom This can be determined after the motor controller selection and design is completed.

[0044] The conduction voltage drop V ce , collector current I c , switching frequency f sw , bus voltage V dc and load current I m As an input parameter, the IGBT junction temperature T 估算 Estimates.

[0045] S2, the junction temperature T obtained by S1 估算 Upload to the new energy vehicle big data platform.

[0046] S3 uploads irregular multiple junction temperature data to the new energy vehicle big data platform, uses the rain flow counting method, and calculates the average junction temperature T m The junction temperature fluctuation ΔT is used to classify and statistically process multiple junction temperature data on the new energy vehicle big data platform to obtain the average junction temperature T m The number of power cycles N under the junction temperature fluctuation ΔT cyc (ΔT,T m ).

[0047] See also Figure 3 , the specific method of statistical junction temperature data by rain flow counting method shown is as follows: when a maximum value is identified by rain flow counting method, and when a maximum value already exists in the maximum value stack, it is compared with the new maximum value, otherwise, only the new maximum value is stored in the stack; if the new maximum value is greater than the first value in the maximum value stack, the number of values ​​in the minimum value stack is detected; if the number of minimum value stacks = 1, an equivalent half cycle is identified, the junction temperature fluctuation ΔT = old maximum value - old minimum value, and the average junction temperature T m= old minimum value + 1 / 2ΔT, and the old maximum value is replaced by the new maximum value; if the number of minimum value stacks is ≥ 2, an equivalent full cycle is identified, the junction temperature fluctuation ΔT = new minimum value - old maximum value, and the old maximum value is replaced by the new maximum value, and the latest minimum value will be removed; when the new maximum value is smaller than the maximum value stack, the new maximum value is stored on the left side of the maximum value stack as the first value in the maximum value stack; after completing the above operations, if there is still more than one value in the maximum value stack, repeat the recursive operation; finally, the average junction temperature T of each m The number of power cycles of the equivalent half cycle and the equivalent full cycle under the junction temperature fluctuation ΔT.

[0048] S4, the average junction temperature T m Substitute the junction temperature fluctuation ΔT into the Arrhenius life model formula to obtain the average junction temperature T m The number of power cycles N under the failure temperature fluctuation ΔT f (ΔT,T m ).

[0049] The Arrhenius lifespan model formula is Where A and α are model parameters, which are obtained by accelerated life test combined with maximum likelihood method or least squares method; E a is the activation energy related to the material, and its value is 9.89×10 -20 J;k B is the Boltzmann constant, which is 1.38×10 -23 J.K. -1 .

[0050] S5, the current life D of the motor controller is calculated based on the Miner fatigue damage accumulation method.

[0051] See also Figure 4 When calculating the lifetime of irregular junction temperature variation data, first calculate the average junction temperature T m The junction temperature fluctuation ΔT is appropriately grouped and divided to obtain a two-dimensional matrix distribution. Then, the Arrhenius life model is used to calculate the average junction temperature T of each group. m The number of failure cycles corresponding to the junction temperature fluctuation ΔT is summarized to obtain a three-dimensional bar chart, and the Z axis represents the number of failure cycles N f (ΔT,T m ). At the same time, the irregular junction temperature data is calculated by rain flow counting method, and the average junction temperature T in S3 is calculated. m Grouped with junction temperature fluctuation ΔT, and count the corresponding power cycle times N cyc (ΔT,T m). Finally, the Miner fatigue damage accumulation method is used to calculate the corresponding average junction temperature T m The ratio of the power cycle number of the rain flow counting method to the failure cycle number of the Arrhenius life model under the junction temperature fluctuation ΔT is then added to obtain the cumulative value of fatigue damage D of the power module under this junction temperature data, that is, the IGBT life. The specific calculation formula is:

[0052] S6, see Figure 5 , determine whether the difference between the current life D of the motor controller and the life D' of the previous calculation cycle is greater than the update frequency D0. If D-D'≥D0, then according to the formula T 修正 =T 估算 h is used to update and correct the junction temperature value, where h is the life attenuation factor. Return to S1 for a new round of motor controller life evaluation. The current life D of the motor controller obtained by S5 in the next calculation cycle is used as the motor controller life evaluation value. If D-D'<D0, the current life D of the motor controller obtained by S5 is used as the motor controller life evaluation value.

[0053] The calculation formula of life attenuation factor is h = exp(m(T 估算 )·D), D is the current life of the motor controller calculated by S5, m(T 估算 ) is the life adjustment coefficient. The higher the junction temperature of the IGBT power module, the smaller the life adjustment coefficient. The life adjustment coefficient is obtained by calibration. For specific values, see Table 1.

[0054] Table 1 Life adjustment coefficient value table

[0055] <![CDATA[Estimate the junction temperature T 估算 (°C)]]> …… 50 100 150 …… Life adjustment factor m …… 1 0.4 0.2 ……

[0056] The present invention first estimates the junction temperature according to the loss and thermal resistance of the IGBT, and then uploads the estimated junction temperature to the new energy vehicle big data platform during the working process of the motor controller, obtains the number of power cycles after classification and statistical processing, and obtains the number of failure power cycles through the Arrhenius life model formula, and then uses the Miner fatigue damage accumulation method to calculate the current life of the motor controller within a certain period of time. Finally, the calculated current life is used as a reference dimension, the estimated parameters are updated online, and the new parameters are used for IGBT junction temperature estimation, and then a new round of life evaluation is carried out. The frequency of online updating of the IGBT junction temperature estimation parameters is adjusted according to the accuracy requirements of the life evaluation. Generally speaking, the life of the IGBT will decay after experiencing long-term working conditions, and the junction temperature will increase under the same working conditions. When evaluating the life of the motor controller, if the junction temperature increase caused by life decay is not considered, a longer motor controller life will be calculated, which deviates from the actual situation. The present invention introduces a life decay factor in the junction temperature estimation method, which can more accurately estimate the IGBT junction temperature, and iterate with the life evaluation at the same time, so as to better implement fault warning and safety monitoring for new energy vehicles.

[0057] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for evaluating the life of a motor controller, characterized in that: The steps include: S1, estimate the junction temperature based on the loss and thermal resistance of the IGBT to obtain the junction temperature T of the IGBT power module when the motor controller is working 估算 ; S2, the junction temperature T obtained by S1 估算 Upload to the new energy vehicle big data platform; S3, using the rain flow counting method, according to the average junction temperature T m The junction temperature fluctuation ΔT is used to classify and statistically process multiple junction temperature data on the new energy vehicle big data platform to obtain the average junction temperature T m The number of power cycles N under the junction temperature fluctuation ΔT cyc (ΔT,T m ); S4, the average junction temperature T m Substitute the junction temperature fluctuation ΔT into the Arrhenius life model formula to obtain the average junction temperature T m The number of power cycles N under the failure temperature fluctuation ΔT f (ΔT,T m ); S5, calculate the current life D of the motor controller based on Miner fatigue damage accumulation method; S6, determine whether the difference between the current life D of the motor controller and the life D' of the previous calculation cycle is greater than the update frequency D0. If D-D'≥D0, then according to the formula T 修正 =T 估算 h updates and corrects the junction temperature value, where h is the life attenuation factor, and returns to S1 for a new round of motor controller life evaluation. The current life D of the motor controller obtained by S5 in the next calculation cycle is used as the motor controller life evaluation value; If D-D'<D0, the current life D of the motor controller obtained in S5 is used as the life evaluation value of the motor controller.

2. The motor controller life assessment method according to claim 1, characterized in that: Junction temperature in S1 T 估算 The calculation formula is: 估算 =P loss Z th +T c ; Where P loss is the IGBT loss, Z th is the thermal resistance, which can be obtained from the IGBT data sheet. c The IGBT shell temperature is obtained in real time through the built-in thermocouple sensor; The calculation formula for IGBT loss is P loss =P cond +P sw , P cond is the conduction loss, V ce is the on-state voltage drop, I c is the collector current, m is the modulation ratio, cosφ is the power factor, R T is the internal resistance of IGBT; P sw is the switching loss, f sw is the switching frequency, V dc is the bus voltage, I m is the load current, E on To open the energy, E off is the turn-off energy, V nom is the rated voltage, I nom is the rated current; The conduction voltage drop V ce , collector current I c , switching frequency f sw , bus voltage V dc and load current I m As an input parameter, the IGBT junction temperature T 估算 Estimates.

3. The motor controller life assessment method according to claim 1 or 2, characterized in that: The processing flow of the rain flow counting method in S3 is as follows: when a maximum value is identified using the rain flow counting method, and if a maximum value already exists in the maximum value stack, it is compared with the new maximum value, otherwise, only the new maximum value is stored in the stack; if the new maximum value is greater than the first value in the maximum value stack, the number of values ​​in the minimum value stack is detected; If the number of minimum value stacks = 1, an equivalent half cycle is identified, the junction temperature fluctuation ΔT = old maximum value - old minimum value, and the average junction temperature T m = old minimum + 1 / 2ΔT, and the old maximum is replaced by the new maximum; If the number of minimum value stacks is ≥ 2, an equivalent full cycle is identified, and the junction temperature fluctuation ΔT = new minimum value - old maximum value. At the same time, the old maximum value is replaced by the new maximum value, and the latest minimum value is removed; When the new maximum value is smaller than the maximum value stack, the new maximum value is stored on the left side of the maximum value stack as the first value in the maximum value stack; After completing the above operations, if there is still more than one value in the maximum value stack, repeat the recursive operation; finally, the average junction temperature T m The number of power cycles of the equivalent half cycle and the equivalent full cycle under the junction temperature fluctuation ΔT.

4. The motor controller life assessment method according to claim 1 or 2, characterized in that: The Arrhenius life model formula in S4 is: Where A and α are model parameters, which are obtained through accelerated life testing combined with maximum likelihood method or least squares method; E a is the activation energy related to the material, and its value is 9.89×10 -20 J; k B is the Boltzmann constant, which is 1.38×10 -23 J.K. -1 .

5. The motor controller life assessment method according to claim 1 or 2, characterized in that: The calculation formula for the current life of the motor controller in S5 is:

6. The motor controller life assessment method according to claim 1 or 2, characterized in that: The calculation formula of the life attenuation factor in S6 is h = exp(m(T 估算 )·D), D is the current life of the motor controller calculated by S5, m(T 估算 ) is the life adjustment coefficient. The higher the junction temperature of the IGBT power module, the smaller the life adjustment coefficient.

7. The motor controller life assessment method according to claim 1 or 2, characterized in that: The life adjustment coefficient is obtained through calibration.

Images