CN110807516A - Junction temperature prediction method of IGBT module for driver - Google Patents
Junction temperature prediction method of IGBT module for driver Download PDFInfo
- Publication number
- CN110807516A CN110807516A CN201911051577.4A CN201911051577A CN110807516A CN 110807516 A CN110807516 A CN 110807516A CN 201911051577 A CN201911051577 A CN 201911051577A CN 110807516 A CN110807516 A CN 110807516A
- Authority
- CN
- China
- Prior art keywords
- igbt module
- junction temperature
- neural network
- network model
- driver
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Abstract
The invention discloses a junction temperature prediction method of an IGBT module for a driver, which comprises the following specific steps: step 1, sampling an IGBT module to be tested, and taking sampling data as training group data; step 2, establishing a BP neural network model; step 3, inputting training group data into a BP neural network model for training; and 4, predicting junction temperature of the IGBT module in real time by adopting the trained BP neural network model, and outputting the predicted junction temperature. The invention solves the problem that the junction temperature of the IGBT module for the driver in the prior art can not be directly measured.
Description
Technical Field
The invention belongs to the technical field of IGBT module drivers, and relates to a junction temperature prediction method of an IGBT module for a driver.
Background
The GBT insulated gate bipolar transistor is a composite fully-controlled voltage-driven power semiconductor device consisting of BJT (bipolar junction transistor) and MOS (insulated gate field effect transistor). The method is applied to the fields of current transformation systems with direct-current voltage of 600V or more, such as alternating-current motors, frequency converters, switching power supplies, lighting circuits, traction transmission and the like. The operating temperature of the IGBT module is a very important parameter, which is related to the normal operation of the device. Because the IGBT chip has the characteristics of difficult direct observation, difficult direct contact and the like in the module, no device capable of realizing the on-line monitoring of the junction temperature of the IGBT module is available in the market at present. Chip failure and packaging failure and highest junction temperature and junction temperature
The amplitude of the fluctuation is related to factors such as the rate of change, the average junction temperature, and the like. Therefore, the real-time junction temperature of the on-line monitoring device is the key for reliable operation of the monitoring device and the system.
Disclosure of Invention
The invention aims to provide a junction temperature prediction method of an IGBT module for a driver. The problem that the junction temperature of the IGBT module for the driver cannot be directly measured in the prior art is solved.
The technical scheme adopted by the invention is that,
a junction temperature prediction method of an IGBT module for a driver comprises the following specific steps:
step 2, establishing a BP neural network model;
step 3, inputting training group data into a BP neural network model for training;
and 4, predicting junction temperature of the IGBT module in real time by adopting the trained BP neural network model, and outputting the predicted junction temperature.
The present invention is also characterized in that,
the sampling data includes a saturation voltage drop Vce, a current Ic, and a resistance Rds.
The BP neural network model comprises an input layer, a hidden layer and an output layerThe number of neurons, the number of neurons in the hidden layer, and the number of neurons in the output layer are niN and n0N is a number of niAnd n0By decision, the formula is:
wherein a is a constant of 0 to 10.
And updating the parameters by the BP neural network model by adopting a back propagation algorithm.
The step 4 is specifically as follows: and inputting test group data to the trained BP neural network model and transplanting the test group data to the CPLD control unit, wherein the CPLD control unit is connected with the IGBT module and inputs the collecting current Ic when the IGBT module works, and the corresponding junction temperature Tc is output in real time through the driving module.
The invention has the advantages that
In an actual working circuit, a chip of the IGBT is arranged inside the IGBT, and the junction temperature is difficult to monitor in real time. According to the invention, experimental data are analyzed by using a BP network algorithm, and junction temperature is predicted by using the algorithm, so that the junction temperature prediction of the IGBT is finally obtained. And then, the algorithm is transplanted into a CPLD control unit, so that the on-line prediction of the junction temperature of the IGBT module is realized.
Drawings
Fig. 1 is a topological structure diagram of a BP neural network model in a junction temperature prediction method of an IGBT module for a driver according to the present invention.
Fig. 2 is a diagram of a test set of data fitting in the method for predicting junction temperature of an IGBT module for a driver in embodiment 1 of the present invention.
Detailed Description
The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
The invention discloses a junction temperature prediction method of an IGBT module for a driver, which comprises the following specific steps of:
step 2, establishing a BP neural network model;
step 3, inputting training group data into a BP neural network model for training;
and 4, predicting junction temperature of the IGBT module in real time by adopting the trained BP neural network model, and outputting the predicted junction temperature.
The sampling data includes a saturation voltage drop Vce, a current Ic, and a resistance Rds.
The BP neural network model comprises an input layer, a hidden layer and an output layer, wherein the number of neurons in the input layer, the number of neurons in the hidden layer and the number of neurons in the output layer are n respectivelyiN and n0N is a number of niAnd n0By decision, the formula is:
wherein a is a constant of 0 to 10.
In the formula n, niAnd n0The number of hidden, input and output layer neurons, respectively, a is a constant from 0 to 10. The number of hidden layer neurons of the BP neural network designed herein is 10, according to the formula and considerations of prediction accuracy.
And updating the parameters by the BP neural network model by adopting a back propagation algorithm.
The step 4 is specifically as follows: and inputting test group data to the trained BP neural network model and transplanting the test group data to the CPLD control unit, wherein the CPLD control unit is connected with the IGBT module and inputs the collecting current Ic when the IGBT module works, and the corresponding junction temperature Tc is output in real time through the driving module.
Example 1
Taking an IGBT module driver which is manufactured by Infineon and has model number FF200R33KF2C and 3300V/200A as an example, the junction temperature prediction method of the IGBT module for the driver is adopted to predict the junction temperature. The method comprises the following specific steps:
the specific sampling method comprises the following steps: setting a current value Ic by a direct current power supply, taking the direct current value as a value in a range of [1,75], changing the value by adopting an interval sampling method, then electrifying an IGBT module driver, conducting the IGBT module driver according to a trigger pulse signal, measuring a saturation voltage drop Vce, and simultaneously recording a current collection current Ic and a resistance Rds corresponding to the saturation voltage drop Vce; the saturation voltage drop Vce, the collecting current Ic, and the resistance Rds are a set of data, and this embodiment 1 collects 360 sets of data.
Step 2, establishing a BP neural network model; the BP neural network model comprises an input layer, a hidden layer and an output layer, the number of neurons in the input layer is set to be 3, for three data of Vce, Ic and Rds, the number of neurons in the hidden layer is set to be 10, and the number of neurons in the output layer is set to be 1.
Step 3, inputting training group data into a BP neural network model for training;
and 4, predicting junction temperature of the IGBT module in real time by adopting the trained BP neural network model, and outputting the predicted junction temperature.
The method specifically comprises the following steps: inputting test group data into a trained BP neural network model, transplanting the test group data into a CPLD control unit, connecting the CPLD control unit with an IGBT module in the IGBT work, inputting a current collection Ic equal to 49, outputting Vce equal to 1.6008 through the IGBT module, and outputting a corresponding junction temperature Tc equal to 42.5 ℃ in real time through prediction.
In this embodiment, test group data is input into a trained BP neural network model to verify a prediction result, and a fitting graph of the test group data is shown in fig. 2, where R is a determination coefficient and a value range is [0,1], a larger number indicates a higher degree of fitting between predicted data and actual data, and the model has better performance, and the expression is:
wherein y isiRepresenting the output junction temperature predicted value;showing the actual junction temperatureThe value of the actual value is the value,representing the average value of the actual junction temperature, wherein m represents the experiment times, and m is more than or equal to 0 and less than or equal to 360;representing the sum of squares of the differences between the predicted values and the average;representing the sum of the squares of the actual and mean values. The figure shows that the decision coefficient R of the prediction results of all samples of the BP neural network is 0.99973, so that the method can realize the on-line prediction of the junction temperature of the IGBT module and has high accuracy.
Claims (5)
1. A junction temperature prediction method of an IGBT module for a driver is characterized by comprising the following specific steps:
step 1, sampling an IGBT module to be tested, and taking sampling data as training group data;
step 2, establishing a BP neural network model;
step 3, inputting the training group data into a BP neural network model for training;
and 4, predicting junction temperature of the IGBT module in real time by adopting the trained BP neural network model, and outputting the predicted junction temperature.
2. The method for predicting junction temperature of the IGBT module for a driver according to claim 1, wherein the sampled data includes a saturation voltage drop Vce, a collector current Ic, and a resistance Rds.
3. The method as claimed in claim 1, wherein the BP neural network model includes an input layer, a hidden layer, and an output layer, and the number of neurons in the input layer, the number of neurons in the hidden layer, and the number of neurons in the output layer are n, respectivelyiN and n0N is a number of niAnd n0By determining the formula:
Wherein a is a constant of 0 to 10.
4. The method for predicting the junction temperature of the IGBT module for a driver according to claim 1, wherein the BP neural network model updates the parameters by using a back propagation algorithm.
5. The method for predicting the junction temperature of the IGBT module for a driver according to claim 1, wherein the step 4 specifically comprises: and inputting test group data to the trained BP neural network model and transplanting the test group data to the CPLD control unit, wherein the CPLD control unit is connected with the IGBT module and inputs the collecting current Ic when the IGBT module works, and the corresponding junction temperature Tc is output in real time through the driving module.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911051577.4A CN110807516A (en) | 2019-10-31 | 2019-10-31 | Junction temperature prediction method of IGBT module for driver |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911051577.4A CN110807516A (en) | 2019-10-31 | 2019-10-31 | Junction temperature prediction method of IGBT module for driver |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110807516A true CN110807516A (en) | 2020-02-18 |
Family
ID=69489820
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911051577.4A Pending CN110807516A (en) | 2019-10-31 | 2019-10-31 | Junction temperature prediction method of IGBT module for driver |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110807516A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112098798A (en) * | 2020-09-18 | 2020-12-18 | 北京航空航天大学 | Neural network-based junction temperature on-line measurement method for silicon carbide MOS (Metal oxide semiconductor) device |
CN112437419A (en) * | 2020-10-30 | 2021-03-02 | 浙江佳乐科仪股份有限公司 | Frequency converter system based on data sharing |
WO2021174907A1 (en) * | 2020-03-03 | 2021-09-10 | 华中科技大学 | Neural network-based igbt junction temperature prediction method |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107545101A (en) * | 2017-08-03 | 2018-01-05 | 西南交通大学 | A kind of design object and the Optimization Design that design variable is section |
CN107543854A (en) * | 2016-06-28 | 2018-01-05 | 中国农业大学 | One heavy metal species quantitative forecasting technique, apparatus and system |
CN108763848A (en) * | 2018-02-10 | 2018-11-06 | 江西航天经纬化工有限公司 | A kind of mechanical properties of propellant prediction technique based on BP artificial neural networks |
CN108959833A (en) * | 2018-09-26 | 2018-12-07 | 北京工业大学 | Tool wear prediction technique based on improved BP neural network |
CN109242147A (en) * | 2018-08-07 | 2019-01-18 | 重庆大学 | Signal fused fan condition prediction technique based on Bp neural network |
CN109376950A (en) * | 2018-11-19 | 2019-02-22 | 国网陕西省电力公司电力科学研究院 | A kind of polynary Load Forecasting based on BP neural network |
CN109492287A (en) * | 2018-10-30 | 2019-03-19 | 成都云材智慧数据科技有限公司 | A kind of solid electrolyte ionic conductivity prediction technique based on BP neural network |
CN109658987A (en) * | 2018-12-14 | 2019-04-19 | 中海石油炼化有限责任公司 | A kind of hydrofining catalyst carrier property prediction technique |
CN109738773A (en) * | 2018-06-19 | 2019-05-10 | 北京航空航天大学 | IGBT module life-span prediction method under a kind of non-stationary operating condition |
CN110222832A (en) * | 2019-06-19 | 2019-09-10 | 中国水产科学研究院东海水产研究所 | Entrance of Changjiang River salt marshes macrobenthos habitat simulation prediction technique |
CN110377991A (en) * | 2019-07-09 | 2019-10-25 | 合肥工业大学 | A kind of insulated gate bipolar transistor IGBT junction temperature on-line prediction method |
-
2019
- 2019-10-31 CN CN201911051577.4A patent/CN110807516A/en active Pending
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107543854A (en) * | 2016-06-28 | 2018-01-05 | 中国农业大学 | One heavy metal species quantitative forecasting technique, apparatus and system |
CN107545101A (en) * | 2017-08-03 | 2018-01-05 | 西南交通大学 | A kind of design object and the Optimization Design that design variable is section |
CN108763848A (en) * | 2018-02-10 | 2018-11-06 | 江西航天经纬化工有限公司 | A kind of mechanical properties of propellant prediction technique based on BP artificial neural networks |
CN109738773A (en) * | 2018-06-19 | 2019-05-10 | 北京航空航天大学 | IGBT module life-span prediction method under a kind of non-stationary operating condition |
CN109242147A (en) * | 2018-08-07 | 2019-01-18 | 重庆大学 | Signal fused fan condition prediction technique based on Bp neural network |
CN108959833A (en) * | 2018-09-26 | 2018-12-07 | 北京工业大学 | Tool wear prediction technique based on improved BP neural network |
CN109492287A (en) * | 2018-10-30 | 2019-03-19 | 成都云材智慧数据科技有限公司 | A kind of solid electrolyte ionic conductivity prediction technique based on BP neural network |
CN109376950A (en) * | 2018-11-19 | 2019-02-22 | 国网陕西省电力公司电力科学研究院 | A kind of polynary Load Forecasting based on BP neural network |
CN109658987A (en) * | 2018-12-14 | 2019-04-19 | 中海石油炼化有限责任公司 | A kind of hydrofining catalyst carrier property prediction technique |
CN110222832A (en) * | 2019-06-19 | 2019-09-10 | 中国水产科学研究院东海水产研究所 | Entrance of Changjiang River salt marshes macrobenthos habitat simulation prediction technique |
CN110377991A (en) * | 2019-07-09 | 2019-10-25 | 合肥工业大学 | A kind of insulated gate bipolar transistor IGBT junction temperature on-line prediction method |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2021174907A1 (en) * | 2020-03-03 | 2021-09-10 | 华中科技大学 | Neural network-based igbt junction temperature prediction method |
CN112098798A (en) * | 2020-09-18 | 2020-12-18 | 北京航空航天大学 | Neural network-based junction temperature on-line measurement method for silicon carbide MOS (Metal oxide semiconductor) device |
CN112437419A (en) * | 2020-10-30 | 2021-03-02 | 浙江佳乐科仪股份有限公司 | Frequency converter system based on data sharing |
CN112437419B (en) * | 2020-10-30 | 2022-12-02 | 浙江佳乐科仪股份有限公司 | Frequency converter system based on data sharing |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110807516A (en) | Junction temperature prediction method of IGBT module for driver | |
CN110221189B (en) | Method for monitoring on-line state of IGBT module bonding wire | |
CN113759225B (en) | IGBT residual life prediction and state evaluation realization method | |
CN103837731B (en) | For the voltage detecting circuit and method of the characteristic for measuring transistor | |
CN112906333B (en) | Photovoltaic inverter IGBT junction temperature online correction method and system considering aging | |
CN110502777A (en) | IGBT module condition detecting system and method based on neural network prediction | |
CN111368454B (en) | SiC MOSFET SPICE model establishment method based on bare chip packaging structure | |
CN105911446A (en) | IGBT aging state monitoring method and IGBT aging state monitoring device | |
CN105811944B (en) | Driving device and method for the estimation of IGBT junction temperature | |
CN106443405A (en) | Integrated multi-IGBT-module aging characteristic measurement device | |
CN206362890U (en) | Electronic power switch device junction temperature on-Line Monitor Device, detection circuit | |
CN109257120B (en) | Optimal method for predicting radio frequency circuit fault characterization parameters | |
CN110082660A (en) | Current transformer IGBT module junction temperature estimation method based on Kalman filter | |
CN113850154A (en) | Inverter IGBT (insulated Gate Bipolar transistor) micro fault feature extraction method based on multi-modal data | |
CN116736063A (en) | IGBT state evaluation method based on weighted LSTM | |
Yang et al. | Hybrid data-driven modeling methodology for fast and accurate transient simulation of SiC MOSFETs | |
CN113092975B (en) | Source-drain breakdown voltage testing method for power MOS device | |
CN111260113A (en) | SiC MOSFET module full life cycle junction temperature online prediction method | |
Huang et al. | Overview of recent progress in condition monitoring for insulated gate bipolar transistor modules: Detection, estimation, and prediction | |
CN109782158A (en) | A kind of Analog circuit diagnosis method based on multiclass classification | |
CN112114237B (en) | IGBT module internal defect monitoring method and circuit based on gate pole charge change | |
CN111832226B (en) | IGBT residual life estimation method based on auxiliary particle filtering | |
CN107544008B (en) | Vehicle-mounted IGBT state monitoring method and device | |
CN113189513A (en) | Ripple-based redundant power supply current sharing state identification method | |
CN107622167B (en) | Collector current soft measurement method for grid control device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination |