A method and system for constructing a nonlinear cumulative model of IGBT lifetime loss

By constructing a nonlinear cumulative model of IGBT lifetime loss, the problem of neglecting the mutual influence of loads in the existing technology is solved, high accuracy of IGBT lifetime prediction is achieved, the influence of load history on damage accumulation is reflected, and the accuracy of lifetime prediction is improved.

CN116992621BActive Publication Date: 2026-11-17华能陕西榆阳电力有限公司 +2
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202310619458.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-05-29
Publication Date
2026-11-17
Estimated Expiration
2043-05-29

AI Technical Summary

Technical Problem

Existing technologies for predicting IGBT module lifespan neglect the interaction between different cyclic loads, resulting in a calculated lifespan that is much larger than the actual remaining lifespan, and thus failing to accurately predict the damage accumulation pattern of IGBTs.

Method used

A nonlinear cumulative model of IGBT lifetime loss is constructed. By establishing the functional relationship between input and output, the junction temperature and junction temperature fluctuation of IGBT throughout its entire life cycle are calculated. Combining the Carten-Dolan fatigue cumulative damage theory and the Coffin-Manson-Arrhenius model, the thermal resistance change gradient is introduced as a rainflow counting parameter. Piecewise linear fitting is used to obtain the change curve of the decay coefficient λ, and the damage accumulation is calculated iteratively.

Benefits of technology

It improves the accuracy of IGBT life loss calculation, enabling more accurate prediction of the remaining life of IGBTs and reflecting the impact of load history on damage accumulation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116992621B_ABST
    Figure CN116992621B_ABST
Patent Text Reader

Abstract

This invention provides a method and system for constructing a nonlinear cumulative model of IGBT lifetime loss, comprising the following steps: Step 1, using IGBT full life cycle aging test parameters as input and IGBT loss as output, establishing a functional relationship between the input and output; Step 2, calculating the IGBT junction temperature and IGBT junction temperature vector during the IGBT full life cycle aging test based on the functional relationship obtained in Step 1; Step 3, calculating the average junction temperature vector and junction temperature fluctuation of the IGBT full life cycle based on the results obtained in Step 2; Step 4, calculating the total number of cycles based on the obtained average junction temperature vector and junction temperature fluctuation of the IGBT full life cycle; Step 5, constructing a nonlinear cumulative model of IGBT lifetime loss based on the obtained total number of cycles. This invention establishes damage accumulation models for different life stages, thereby improving the accuracy of IGBT lifetime loss calculation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of new energy (wind power generation, photovoltaic power generation), and specifically relates to a method and system for constructing a nonlinear cumulative model of IGBT lifetime loss. Background Technology

[0002] With the rapid development of new energy power generation industries, represented by wind power and photovoltaic power generation, in my country, IGBT life management, as a core energy conversion device, has become a hot topic. The prerequisite for achieving life management is to deeply explore methods for identifying critical points and calculating equivalent thermal network models at each life stage, accurately analyze the damage accumulation patterns at each stage of the device's entire life cycle, and ultimately develop a practical and highly accurate method for calculating remaining life. When predicting the life of IGBT modules, existing methods mostly rely on Miner's rule to calculate interconnect structure damage accumulation, assuming that damage caused by different cyclic loads is independent and that damage is a linear accumulation process, neglecting the mutual influence between different cyclic loads. This leads to calculated lifespans far exceeding the actual remaining lifespan. However, it should be noted that interconnect structure damage accumulation is not only related to the stress level of the cyclic loads but also to the current degree of damage. Even cyclic loads with the same stress level cause different damage at different aging stages; that is, the load history has a significant impact on damage accumulation. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for constructing a nonlinear cumulative model of IGBT lifetime loss, thereby overcoming the aforementioned shortcomings in the prior art.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0005] The present invention provides a method for constructing a nonlinear cumulative model of IGBT lifetime loss, comprising the following steps:

[0006] Step 1: Using the IGBT full life cycle aging test parameters as input and the IGBT loss as output, establish the functional relationship between the input and output.

[0007] Step 2: Calculate the IGBT junction temperature and IGBT junction temperature vector during the IGBT full life cycle aging test based on the functional relationship obtained in Step 1.

[0008] Step 3: Based on the results obtained in Step 2, calculate the average junction temperature vector and junction temperature fluctuation over the entire IGBT lifecycle.

[0009] Step 4: Calculate the total number of cycles based on the obtained average junction temperature vector and junction temperature fluctuation throughout the IGBT's lifespan.

[0010] Step 5: Construct a nonlinear cumulative model of IGBT lifetime loss based on the total number of cycles obtained.

[0011] In a preferred embodiment, in step 2, the IGBT junction temperature during the IGBT full lifecycle aging test is calculated based on the functional relationship obtained in step 1, using the following formula:

[0012] T j =T c +P×Z jc

[0013] Among them, T j T is the calculated IGBT junction temperature. c Here, P represents the case temperature monitoring data from the aging test data, and Z represents the IGBT loss calculated using the functional relationship H. jc The junction-shell equivalent thermal resistance is obtained based on the dynamic Foster thermal resistance model;

[0014] The IGBT junction temperature vector during the IGBT's entire lifecycle aging test is calculated using the following formula:

[0015] T j_all =[T j1 T j2 , ..., T ji , ..., T jn ]

[0016] Among them, T j1 T j2 , ..., T ji , ..., T jn The vector represents the elements of the IGBT junction temperature vector over its entire lifecycle, where n is the total number of elements in the vector, and T is the total temperature of the IGBT junction. j_all This refers to the junction temperature vector of the IGBT during the IGBT's full life cycle aging test.

[0017] In a preferred embodiment, in step 3, based on the results obtained in step 2, the average junction temperature vector and junction temperature fluctuation over the entire IGBT lifecycle are calculated. The specific method is as follows:

[0018] S31, the IGBT junction temperature vector obtained in step 2 during the IGBT full life cycle aging test is segmented to obtain multiple sub-vectors;

[0019] S32, based on the obtained multiple sub-vectors, the IGBT junction temperature during the IGBT full life cycle aging test is segmented to obtain the IGBT junction temperature segment corresponding to each sub-vector;

[0020] S33, calculate the average junction temperature and junction temperature fluctuation of the IGBT junction temperature range corresponding to each sub-vector;

[0021] S34 calculates the average junction temperature vector and junction temperature fluctuation over the entire IGBT lifecycle.

[0022] In a preferred embodiment, in S31, the IGBT junction temperature vector is segmented by a set length to obtain multiple sub-vectors.

[0023] In a preferred embodiment, in step 4, the total number of cycles is calculated based on the obtained average junction temperature vector over the entire IGBT lifecycle and the junction temperature fluctuation. The specific method is as follows:

[0024] Based on the Carten-Dolan fatigue cumulative damage theory, and on the basis of the Coffin-Manson-Arrhenius model, the junction temperature difference, average junction temperature, power cycle frequency, thermal resistance, and thermal resistance change gradient are used as rainflow counting parameters to calculate the total number of cycles.

[0025] In a preferred embodiment, in step 5, the expression for the IGBT lifetime stage u-th damage in the IGBT lifetime nonlinear cumulative model is as follows:

[0026]

[0027]

[0028] Among them, D u For the damage of N in the total number of cycles yu during the u-th stage of lifespan, au x is the number of iterations under the au condition. u λ is the damage accumulation factor in the first stage of lifespan. u It is an acceleration factor for the first stage of lifespan.

[0029] This preferred embodiment provides a system for constructing a nonlinear cumulative model of IGBT lifetime loss, including:

[0030] The function relationship establishment unit is used to take the IGBT full life cycle aging test parameters as input and IGBT loss as output to establish the function relationship between the input and output.

[0031] The junction temperature calculation unit is used to calculate the IGBT junction temperature and IGBT junction temperature vector during the IGBT's full life cycle aging test based on the obtained functional relationship.

[0032] The average value calculation unit is used to calculate the average junction temperature vector and junction temperature fluctuation throughout the IGBT's entire life cycle.

[0033] The cycle count calculation unit is used to calculate the total number of cycles based on the obtained average junction temperature vector and junction temperature fluctuation throughout the IGBT's entire life cycle.

[0034] The model building unit is used to construct a nonlinear cumulative model of IGBT lifetime loss based on the obtained total number of cycles.

[0035] In a preferred embodiment, the average value calculation unit includes:

[0036] The vector segmentation module is used to segment the IGBT junction temperature vector obtained during the IGBT full life cycle aging test to obtain multiple sub-vectors.

[0037] The junction temperature segmentation module is used to segment the IGBT junction temperature during the IGBT full life cycle aging test based on multiple sub-vectors, and obtain the IGBT junction temperature segment corresponding to each sub-vector.

[0038] The average value calculation module is used to calculate the average junction temperature and junction temperature fluctuation of the IGBT junction temperature range corresponding to each sub-vector.

[0039] And the average junction temperature vector and junction temperature fluctuation of the IGBT throughout its entire life cycle are calculated.

[0040] Compared with the prior art, the beneficial effects of the present invention are:

[0041] This invention provides a method for constructing a nonlinear cumulative model of IGBT lifetime loss. Addressing the current state of IGBT remaining lifetime prediction, this method utilizes the concept of material memory to apply the degradation of material memory performance to fatigue damage analysis. It introduces a decay coefficient λ (corresponding to changes in thermal resistance) to quantitatively characterize the degree of degradation in the device's thermal stress adaptability. The dynamic variation trend of the decay coefficient λ at each lifetime stage is studied, and a piecewise linear fitting method is used to obtain the variation curve of the decay coefficient λ. Based on this, iterative calculations of the corresponding cycle number and nonlinear damage accumulation results are performed to establish damage accumulation models for different lifetime stages, thereby improving the accuracy of IGBT lifetime loss calculation. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0043] The present invention will now be described in further detail with reference to the accompanying drawings.

[0044] This invention addresses the current state of IGBT remaining lifetime prediction by introducing thermal resistance and thermal resistance gradient as rainflow count parameters, building upon the traditional Comn-Manson-Arrhenius model which uses parameters such as junction temperature difference, average junction temperature, and power cycling frequency as rainflow count parameters.

[0045] Nonlinear damage accumulation models are constructed at different life stages. Based on the concept of material memory, the degradation of material memory performance is used for fatigue damage analysis. The decay coefficient λ (corresponding to the change in thermal resistance) is introduced to quantitatively characterize the degree of degradation of the device's thermal stress adaptability. The dynamic change trend of the decay coefficient λ at each life stage is studied. The change curve of the decay coefficient λ is obtained by piecewise linear fitting. On this basis, the corresponding number of cycles and the nonlinear damage accumulation results are calculated iteratively to establish damage accumulation models at different life stages.

[0046] This invention provides a method for constructing a nonlinear cumulative model of IGBT lifetime loss. The specific process of the embodiment is shown in the appendix. Figure 1 As shown, the specific steps are as follows:

[0047] Step 1: Collect IGBT full lifecycle aging test parameters, including test switching frequency (f), module case temperature (T). c ), collector current (I) c ), saturation pressure drop of the focusing stage V cesat A black-box model is adopted, taking aging test parameters as input and IGBT losses (conduction loss, inherent loss, and turn-off loss) as output, to establish a functional relationship H between the input and output quantities, namely:

[0048] P = H[f, T] c V cesat I c ]

[0049] Where H is the functional relationship obtained through the black box model; P is the IGBT loss calculated using the functional relationship H.

[0050] Step 2: Based on the dynamic Foster thermal resistance model, and according to the functional relationship H obtained in Step 1, calculate the IGBT junction temperature during the aging test. The calculation method is as follows:

[0051] T j =T c +P×Z jc

[0052] Among them, T j T is the calculated IGBT junction temperature. c Here, P represents the case temperature monitoring data from the aging test data, and Z represents the IGBT loss calculated using the functional relationship H. jc This is the junction-shell equivalent thermal resistance obtained based on the dynamic Foster thermal resistance model.

[0053] By calculation, the junction temperature vector T of the IGBT over its entire lifecycle can be obtained. j_all Then we have:

[0054] T j_all=[T j1 T j2 , ..., T ji , ..., T jn ]

[0055] Among them, T j1 T j2 , ..., T ji , ..., T jn The vector represents the junction temperature of the IGBT throughout its entire lifecycle, where n is the total number of elements in the vector.

[0056] Step 3: First, the IGBT junction temperature vector is segmented according to the sampling time t. In this example, the sampling frequency of each aging monitoring data is f. a Then we have:

[0057] f a =k×f

[0058] n = t × f a =t×k×f

[0059] Where t is the full life cycle sampling time (s), and the total number of elements in the IGBT junction temperature vector is the product of the sampling time and the sampling frequency. It can be seen that the sampling frequency of the monitoring data is k times the test switching frequency f.

[0060]

[0061] Where t is the sampling time, l is the number of segments, and l represents t / 600 rounded up, then:

[0062] T j_all =[T1, T2,...,Tq,...,Tl]

[0063]

[0064]

[0065]

[0066] 1 < q < l

[0067] Where T1, T2, ..., Tq, ..., Tl are the sub-vectors after segmentation, and except for Tl, the length of the other sub-vectors is 60×f. a That is, the sum of data collected in 1 minute.

[0068] Step 4, based on the segmentation results in Step 3 and the junction temperature T in Step 2 j The calculation results are used to calculate the segmented average junction temperature T. j-a Junction temperature fluctuation ΔT j-w Taking the calculation of the self-vector of T1 as an example, we have:

[0069]

[0070] v = 60 × f a

[0071] ΔT1 j-w =T1 jmax -T1 jmin

[0072] Among them, T1 jmax T1 jmin Let T1 be the maximum and minimum values ​​among the subvectors.

[0073] Step 5, repeat step 4, calculate the average junction temperature T2, ..., Tq, ..., Tl. j-a T3 j-a , ..., Tq j-a , ..., Tl j-a Junction temperature fluctuation value ΔT2 j-w ΔT3 j-w ,...,ΔTq j-w , ..., ΔTl j-w This step allows us to obtain the average junction temperature vector T over the entire IGBT lifecycle. j-a With junction temperature fluctuation ΔT j-w vector.

[0074] T j-a =[T1 j-a T2 j-a , ..., Tq j-a , ..., Tl j-a ]

[0075] ΔT j-w =[ΔT1 j-w ΔT2 j-w ,...,ΔTq j-w , ..., ΔTl j-w ]

[0076] Step 6: Calculate the average junction temperature and other parameters obtained in steps 1 to 5 to determine the remaining total number of cycles N for the TGBT module. f Based on the Carten-Dolan fatigue cumulative damage theory, and building upon the traditional Coffin-Manson-Arrhenius model which uses parameters such as junction temperature difference, average junction temperature, and power cycle frequency as rainflow counting parameters, this paper further introduces thermal resistance and thermal resistance gradient as rainflow counting parameters. Therefore,

[0077]

[0078] Where, N f Let A be the total number of iterations, and k be the pre-electron coefficient. b T is the Boltzmann constant. j-a E represents the average junction temperature. a As the activation energy, Z jc The junction-shell equivalent thermal resistance ΔZ is obtained based on the dynamic Foster thermal resistance model. jc For the thermal resistance gradient, ΔT j-w For junction temperature fluctuations, ΔT j The junction temperature gradient is represented by , and all other parameters are system parameters.

[0079] Step 7: Based on the total number of cycles obtained in Step 6, calculate the IGBT damage D using the dynamic Foster thermal resistance model for different life stages of the module. The IGBT device has no initial damage at the beginning of the first life stage, i.e., damage D0 = 0. The initial values ​​of the thermal network model are all the original network settings. Initial damage exists from the beginning of the first life stage. Based on the concept of material memory, a decay coefficient (aging acceleration factor) λ is introduced. That is, when accumulating damage in this life stage, the damage already generated in the previous stage and the degradation caused by previous cycles to the device are considered. Taking the first life stage as an example, we have:

[0080]

[0081]

[0082] Where D1 represents the damage over the first stage of lifespan, a total of y1 cycles, and N... a1 λ is the number of cycles under condition a1, x1 is the damage accumulation factor in the first stage of life, and λ1 is the acceleration factor in the first stage of life, which are determined based on the dynamic thermal network parameters.

[0083] The method for calculating the damage in stage u of the IGBT lifetime is as follows:

[0084]

[0085]

[0086] Among them, D u For the damage of N in the total number of cycles yu during the u-th stage of lifespan, au x is the number of iterations under the au condition. u λ is the damage accumulation factor in the first stage of lifespan. u The acceleration factor for the first stage of lifespan is determined based on dynamic thermal network parameters.

[0087] This embodiment also provides a system for constructing a nonlinear cumulative model of IGBT lifetime loss, including:

[0088] The function relationship establishment unit is used to take the IGBT full life cycle aging test parameters as input and IGBT loss as output to establish the function relationship between the input and output.

[0089] The junction temperature calculation unit is used to calculate the IGBT junction temperature and IGBT junction temperature vector during the IGBT's full life cycle aging test based on the obtained functional relationship.

[0090] The average value calculation unit is used to calculate the average junction temperature vector and junction temperature fluctuation throughout the IGBT's entire life cycle.

[0091] The cycle count calculation unit is used to calculate the total number of cycles based on the obtained average junction temperature vector and junction temperature fluctuation throughout the IGBT's entire life cycle.

[0092] The model building unit is used to construct a nonlinear cumulative model of IGBT lifetime loss based on the obtained total number of cycles.

[0093] In a preferred embodiment, the average value calculation unit includes:

[0094] The vector segmentation module is used to segment the IGBT junction temperature vector obtained during the IGBT full life cycle aging test to obtain multiple sub-vectors.

[0095] The junction temperature segmentation module is used to segment the IGBT junction temperature during the IGBT full life cycle aging test based on multiple sub-vectors, and obtain the IGBT junction temperature segment corresponding to each sub-vector.

[0096] The average value calculation module is used to calculate the average junction temperature and junction temperature fluctuation of the IGBT junction temperature range corresponding to each sub-vector.

[0097] And the average junction temperature vector and junction temperature fluctuation of the IGBT throughout its entire life cycle are calculated.

Claims

1. A method for constructing a nonlinear cumulative model of IGBT lifetime loss, characterized in that, Includes the following steps: Step 1: Using the IGBT full life cycle aging test parameters as input and the IGBT loss as output, establish the functional relationship between the input and output. Step 2: Calculate the IGBT junction temperature and IGBT junction temperature vector during the IGBT full life cycle aging test based on the functional relationship obtained in Step 1. Step 3: Based on the results obtained in Step 2, calculate the average junction temperature vector and junction temperature fluctuation over the entire IGBT lifecycle. Step 4: Calculate the total number of cycles based on the obtained average junction temperature vector and junction temperature fluctuation over the entire IGBT lifecycle; Step 5: Construct a nonlinear cumulative model of IGBT lifetime loss based on the obtained total number of cycles; In step 5, the IGBT lifetime loss nonlinear cumulative model is used to calculate the IGBT lifetime loss of the first... The expression for stage damage is as follows: in, For lifespan Phase Overall Cycle Secondary injury. In order to be in Number of iterations under condition u For lifespan Stage damage accumulation factor, For lifespan Stage acceleration factor; This represents the total number of loops.

2. The method for constructing a nonlinear cumulative model of IGBT lifetime loss according to claim 1, characterized in that, In step 2, based on the functional relationship obtained in step 1, the IGBT junction temperature during the IGBT's full life-cycle aging test is calculated using the following formula: in, To calculate the IGBT junction temperature, This refers to the shell temperature monitoring data from the aging test data. To utilize functional relationships The calculated IGBT loss, The junction-shell equivalent thermal resistance is obtained based on the dynamic Foster thermal resistance model; The IGBT junction temperature vector during the IGBT's entire lifecycle aging test is calculated using the following formula: in, For each element of the IGBT junction temperature vector throughout its entire lifecycle, The total number of elements in the vector. This refers to the junction temperature vector of the IGBT during the IGBT's full life cycle aging test.

3. The method for constructing a nonlinear cumulative model of IGBT lifetime loss according to claim 1, characterized in that, In step 3, based on the results obtained in step 2, the average junction temperature vector and junction temperature fluctuation over the entire IGBT lifecycle are calculated. The specific method is as follows: S31, the IGBT junction temperature vector obtained in step 2 during the IGBT full life cycle aging test is segmented to obtain multiple sub-vectors; S32, based on the obtained multiple sub-vectors, the IGBT junction temperature during the IGBT full life cycle aging test is segmented to obtain the IGBT junction temperature segment corresponding to each sub-vector; S33, calculate the average junction temperature and junction temperature fluctuation of the IGBT junction temperature range corresponding to each sub-vector; S34 calculates the average junction temperature vector and junction temperature fluctuation over the entire IGBT lifecycle.

4. The method for constructing a nonlinear cumulative model of IGBT lifetime loss according to claim 3, characterized in that, In S31, the IGBT junction temperature vector is segmented into multiple sub-vectors with a set length.

5. The method for constructing a nonlinear cumulative model of IGBT lifetime loss according to claim 1, characterized in that, In step 4, the total number of cycles is calculated based on the obtained average junction temperature vector over the entire IGBT lifecycle and the junction temperature fluctuation. The specific method is as follows: Based on the Carten-Dolan fatigue cumulative damage theory, and on the basis of the Coffin-Manson-Arrhenius model, the junction temperature difference, average junction temperature, power cycle frequency, thermal resistance, and thermal resistance change gradient are used as rainflow counting parameters to calculate the total number of cycles.

6. A system for constructing a nonlinear cumulative model of IGBT lifetime loss, characterized in that, Based on the construction method according to claim 1, it includes: The function relationship establishment unit is used to take the IGBT full life cycle aging test parameters as input and IGBT loss as output to establish the function relationship between the input and output. The junction temperature calculation unit is used to calculate the IGBT junction temperature and IGBT junction temperature vector during the IGBT full life cycle aging test based on the obtained functional relationship. The average value calculation unit is used to calculate the average junction temperature vector and junction temperature fluctuation throughout the IGBT's entire life cycle. The cycle count calculation unit is used to calculate the total number of cycles based on the obtained average junction temperature vector and junction temperature fluctuation throughout the IGBT's entire life cycle. The model building unit is used to construct a nonlinear cumulative model of IGBT lifetime loss based on the obtained total number of cycles.

7. The system for constructing a nonlinear cumulative model of IGBT lifetime loss according to claim 6, characterized in that, The average calculation unit includes: The vector segmentation module is used to segment the IGBT junction temperature vector obtained during the IGBT full life cycle aging test to obtain multiple sub-vectors. The junction temperature segmentation module is used to segment the IGBT junction temperature during the IGBT full life cycle aging test based on multiple sub-vectors, and obtain the IGBT junction temperature segment corresponding to each sub-vector. The average value calculation module is used to calculate the average junction temperature and junction temperature fluctuation of the IGBT junction temperature range corresponding to each sub-vector. And the average junction temperature vector and junction temperature fluctuation of the IGBT throughout its entire life cycle are calculated.

Citation Information

Patent Citations

  • Method and system for predicting service life of IGBT module of power supply under passenger car

    CN113591336A