CN109558681A - A kind of preparation method and device of the loss power of insulated gate bipolar transistor - Google Patents
A kind of preparation method and device of the loss power of insulated gate bipolar transistor Download PDFInfo
- Publication number
- CN109558681A CN109558681A CN201811454690.2A CN201811454690A CN109558681A CN 109558681 A CN109558681 A CN 109558681A CN 201811454690 A CN201811454690 A CN 201811454690A CN 109558681 A CN109558681 A CN 109558681A
- Authority
- CN
- China
- Prior art keywords
- basis function
- neural network
- radial basis
- insulated gate
- gate bipolar
- 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
- 238000002360 preparation method Methods 0.000 title claims abstract description 19
- 238000013528 artificial neural network Methods 0.000 claims abstract description 109
- 230000006870 function Effects 0.000 claims description 123
- 238000012549 training Methods 0.000 claims description 44
- 239000000758 substrate Substances 0.000 claims description 27
- 210000002569 neuron Anatomy 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 9
- 238000004088 simulation Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 2
- 238000000034 method Methods 0.000 description 16
- 230000008569 process Effects 0.000 description 6
- 230000006399 behavior Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000005855 radiation Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 239000002826 coolant Substances 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 230000017525 heat dissipation Effects 0.000 description 2
- 238000009413 insulation Methods 0.000 description 2
- 239000007788 liquid Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000003137 locomotive effect Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/39—Circuit design at the physical level
-
- 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 present invention provides a kind of preparation method of the loss power of insulated gate bipolar transistor and device, a kind of preparation method of the loss power of insulated gate bipolar transistor include: obtain insulated gate bipolar transistor it is in running order when running parameter；By the running parameter, inputs a preset radial basis function neural network and calculated, obtain the output valve of the radial basis function neural network；By the output valve, it is determined as the loss power of the insulated gate bipolar transistor.The present invention utilizes the non-linear behavior of radial basis function neural network, can accurately calculate the loss power of insulated gate bipolar transistor, and have preferable Project Realization.
Description
Technical field
The present invention relates to state-of-charges to correct field, in particular to a kind of loss power of insulated gate bipolar transistor
Preparation method and device.
Background technique
IGBT (Insulated Gate Bipolar Transistor, insulated gate bipolar transistor) is a kind of compound
Made of novel power semiconductor, the advantages that possessing the processing capacity of wider operating voltage range and high current because of it,
It is widely used in pure electric automobile drive control, express locomotive (motor-car) and generation of electricity by new energy (such as wind-power electricity generation) neck
Domain.
As the core power conversion equipment of pure electric automobile drive system, IGBT module can generate greatly during the work time
The heat of amount, and IGBT module temperature it is excessively high, especially junction temperature it is excessively high will to its service life generate seriously affect into, and then influence
The reliability of system, pure electric automobile is generally required during the work time to be limited in junction temperature with IGBT module both at home and abroad at present
150 DEG C hereinafter, the temperature is the limiting temperature of IGBT module reliably working.In practical applications the junction temperature of IGBT be can not be direct
Measurement, be the circuit structure diagram of IGBT module, wherein thermistor R can measure temperature, but thermistor is surveyed as shown in Figure 1
The temperature measured is the substrate temperature of IGBT module, rather than junction temperature, and the temperature will be lower than the junction temperature of IGBT under normal circumstances.Therefore
Junction temperature is extrapolated typically only by the design feature of some external parameters and IGBT, wherein obtaining the loss function of IGBT module
Rate is to realize the important prerequisite of estimation IGBT junction temperature, and the true journey of the accuracy of IGBT module loss power and IGBT junction temperature
Spend it is closely related, be based on the reason, the power loss of IGBT module is all the heat in pure electric automobile drive control all the time
Point studies a question.
Research achievement at present about the calculating of IGBT loss power is numerous, wherein with the side based on IGBT loss power model
Method be it is leading, such method has strict theoretical integrality, but the parameter involved in calculating process is numerous, it is some of compared with
Hardly possible obtains, and there are also some parameters to change with the variation of IGBT oneself state, due to above, such method
Project Realization is poor.
Summary of the invention
The present invention provides a kind of preparation method of the loss power of insulated gate bipolar transistor and devices, to solve
The poor problem of the Project Realization of the loss power calculating process of insulated gate bipolar transistor in the prior art.
In order to solve the above-mentioned technical problem, the present invention adopts the following technical scheme:
According to one aspect of the present invention, a kind of acquisition side of the loss power of insulated gate bipolar transistor is provided
Method, comprising:
Running parameter when acquisition insulated gate bipolar transistor is in running order；
It by the running parameter, inputs a preset radial basis function neural network and is calculated, obtain the radial base
The output valve of Function Neural Network；
By the output valve, it is determined as the loss power of the insulated gate bipolar transistor.
Further, the preset radial basis function neural network is carried out according to initial radial basis function neural network
Training, obtains；
The initial radial basis function neural network are as follows:
Wherein, wherein y (x, w) indicate initial radial basis function neural network output valve；X indicates input vector；wiTable
Show weight；L indicates hidden neuron quantity；ciIndicate center vector；‖x-ci‖ indicates input vector to the distance of center vector；
φ is radial basis function.
Further, it is trained according to initial radial basis function neural network, obtains preset Radial Basis Function neural
Network, comprising:
Obtain the DC terminal voltage V when insulated gate bipolar transistor simulation working conditionDC0, phase current virtual value
IS0And switching frequency F0；
The substrate temperature T of the insulated gate bipolar transistor within a preset time0Fluctuation range be less than fluctuation threshold
When, obtain loss power P0；
By DC terminal voltage VDC0, phase current virtual value IS0, switching frequency F0, substrate temperature T0And loss power P0, really
It is set to training data；
The initial radial basis function neural network is trained by the training data, the power after being trained
Weight；
The weight of the initial radial basis function neural network is replaced with into the weight after the training, obtains preset diameter
To basis function neural network.
Further, the running parameter includes: DC terminal voltage VDC, phase current virtual value IS, switching frequency F and base
Plate temperature T1。
Further, it by the running parameter, inputs a preset radial basis function neural network and is calculated, obtain institute
State the output valve of radial basis function neural network, comprising:
According to the DC terminal voltage VDC, phase current virtual value IS, switching frequency F and substrate temperature T1, establish input
Vector [VDC IS F T1]T；
By the input vector [VDC IS F T1]TIt is input to preset radial basis function neural network:
Wherein, y (x, w) indicates the output valve of radial basis function neural network；X indicates input vector；wiIndicate the instruction
Weight after white silk；L indicates hidden neuron quantity；ciIndicate center vector；‖x-ci‖ indicate input vector to center vector away from
From；φ is radial basis function；
Obtain the output valve of the radial basis function neural network.
Another aspect according to the present invention provides a kind of being filled for the loss power of insulated gate bipolar transistor
It sets, comprising:
First obtain module, for obtain insulated gate bipolar transistor it is in running order when running parameter；
First computing module, for inputting the running parameter a preset radial basis function neural network and being counted
It calculates, obtains the output valve of the radial basis function neural network；
First determining module, for being determined as the loss power of the insulated gate bipolar transistor for the output valve.
Further, the preset radial basis function neural network is carried out according to initial radial basis function neural network
Training, obtains；
The initial radial basis function neural network are as follows:
Wherein, wherein y (x, w) indicate initial radial basis function neural network output valve；X indicates input vector；wiTable
Show weight；L indicates hidden neuron quantity；ciIndicate center vector；‖x-ci‖ indicates input vector to the distance of center vector；
φ is radial basis function.
Further, further includes:
Second obtains module, for obtaining the DC terminal voltage when insulated gate bipolar transistor simulation working condition
VDC0, phase current virtual value IS0And switching frequency F0；
Third obtains module, the substrate temperature T for the insulated gate bipolar transistor within a preset time0Fluctuation
When range is less than fluctuation threshold, loss power P is obtained0；
Second determining module is used for DC terminal voltage VDC0, phase current virtual value IS0, switching frequency F0, substrate temperature T0
And loss power P0, it is determined as training data；
Training module is obtained for being trained by the training data to the initial radial basis function neural network
Weight after to training；
Second computing module, after the weight of the initial radial basis function neural network is replaced with the training
Weight obtains preset radial basis function neural network.
Further, the running parameter includes: DC terminal voltage VDC, phase current virtual value IS, switching frequency F and base
Plate temperature T1。
Further, the first computing module, comprising:
First computing unit, for according to the DC terminal voltage VDC, phase current virtual value IS, switching frequency F and base
Plate temperature T1, establish input vector [VCC IS F T1]D；
Second computing unit is used for the input vector [VDC IS F T1]TIt is input to preset Radial Basis Function neural
Network:
Wherein, y (x, w) indicates the output valve of radial basis function neural network；X indicates input vector；wiIndicate the instruction
Weight after white silk；L indicates hidden neuron quantity；ciIndicate center vector；‖x-ci‖ indicate input vector to center vector away from
From；φ is radial basis function；
Third computing unit, for obtaining the output valve of the radial basis function neural network.
Another aspect according to the present invention provides a kind of being set for the loss power of insulated gate bipolar transistor
It is standby, comprising: memory, processor and storage on a memory and the computer program that can run on a processor, the calculating
Machine program realizes the preparation method of the loss power of insulated gate bipolar transistor as described above when being executed by the processor
The step of.
Another aspect according to the present invention, provides a kind of computer readable storage medium, described computer-readable to deposit
Computer program is stored on storage media, the computer program realizes insulated gate bipolar as described above when being executed by processor
The step of preparation method of the loss power of transistor npn npn.
The beneficial effects of the present invention are:
Above-mentioned technical proposal, by running parameter of insulated gate bipolar transistor when in running order, as radial base
The input vector of Function Neural Network calculates the loss function of insulated gate bipolar transistor by radial basis function neural network
Rate can accurately calculate the loss of insulated gate bipolar transistor using the non-linear behavior of radial basis function neural network
Power, and there is preferable Project Realization.
Detailed description of the invention
Fig. 1 shows IGBT module circuit structure diagrams in the prior art；
Fig. 2 indicates that a kind of preparation method of the loss power of insulated gate bipolar transistor provided in an embodiment of the present invention shows
One of be intended to；
Fig. 3 indicates that a kind of preparation method of the loss power of insulated gate bipolar transistor provided in an embodiment of the present invention shows
The two of intention；
Fig. 4 indicates that training data provided in an embodiment of the present invention obtains system structure diagram；
Fig. 5 indicates that training data provided in an embodiment of the present invention obtains flow chart；
Fig. 6 indicates that loss power provided in an embodiment of the present invention calculates schematic diagram；
Fig. 7 indicates that a kind of acquisition device of the loss power of insulated gate bipolar transistor provided in an embodiment of the present invention shows
It is intended to；
Fig. 8 shows the first computing module schematic diagrames provided in an embodiment of the present invention.
Description of symbols:
41, radiator；411, radiating module；412, thermal insulation layer；42, host computer；43, IGBT module；44, engine；
45, coolant liquid；71, first module is obtained；72, the first computing module；721, the first computing unit；722, the second computing unit；
723, third computing unit；73, the first determining module；74, second module is obtained；75, third obtains module；76, it second determines
Module；77, training module；78, the second computing module.
Specific embodiment
The exemplary embodiment that the present invention will be described in more detail below with reference to accompanying drawings.Although showing the present invention in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the present invention without should be by embodiments set forth here
It is limited.It is to be able to thoroughly understand the present invention on the contrary, providing these embodiments, and can be by the scope of the present invention
It is fully disclosed to those skilled in the art.
As shown in Fig. 2, the embodiment of the invention provides a kind of acquisition sides of the loss power of insulated gate bipolar transistor
The preparation method of method, the loss power of the insulated gate bipolar transistor includes:
S21: running parameter when acquisition insulated gate bipolar transistor is in running order；
It should be noted that the running parameter includes: DC terminal voltage, phase current virtual value, switching frequency and substrate
Temperature, but not limited to this.
S22: it by running parameter, inputs a preset radial basis function neural network and is calculated, obtain radial basis function
The output valve of neural network；
It should be noted that input vector can be established according to running parameter, using the input vector as preset radial direction
The input layer of basis function neural network, to calculate the output valve of radial basis function neural network, output valve is referred to as
The numerical value of output layer.
S23: by output valve, it is determined as the loss power of insulated gate bipolar transistor.
In the embodiment of the present invention, by running parameter of insulated gate bipolar transistor when in running order, as radial direction
The input vector of basis function neural network calculates the loss of insulated gate bipolar transistor by radial basis function neural network
Power can accurately calculate the damage of insulated gate bipolar transistor using the non-linear behavior of radial basis function neural network
Wasted work rate, and there is preferable Project Realization.
In order to guarantee the accuracy of radial basis function neural network calculated result, on the basis of foregoing invention embodiment,
In the embodiment of the present invention, preset radial basis function neural network is trained according to initial radial basis function neural network,
It obtains；
Initial radial basis function neural network are as follows:
Wherein, wherein y (x, w) indicate initial radial basis function neural network output valve；X indicates input vector；wiTable
Show weight；L indicates hidden neuron quantity；ciIndicate center vector；‖x-ci‖ indicates input vector to the distance of center vector；
φ is radial basis function.
It should be noted that being trained by training data to initial radial basis function neural network, after being trained
Radial basis function neural network, the radial basis function neural network after the training can accurately calculate loss power, wherein
Training data refers to DC terminal voltage, phase current virtual value, switch when insulated gate bipolar transistor is in running order
Frequency, substrate temperature and corresponding loss power；Preferably, available a large amount of training datas are trained, to be promoted
The accuracy of the calculated result of radial basis function neural network after training.Training data can use existing experimental data,
Experimental provision can certainly be built to be obtained.
As shown in figure 3, on the basis of above-mentioned each inventive embodiments, in the embodiment of the present invention, according to initial radial base letter
Number neural network is trained, and obtains preset radial basis function neural network, comprising:
S31: DC terminal voltage V when insulated gate bipolar transistor simulation working condition is obtainedDCO, phase current virtual value
ISOAnd switching frequency FO；
S32: the substrate temperature T of insulated gate bipolar transistor within a preset time0Fluctuation range be less than fluctuation threshold
When, obtain loss power P0；
It should be noted that the preset time can be 600 seconds, but not limited to this；Fluctuation threshold can be ± 0.5 DEG C.
S33: by DC terminal voltage VDC0, phase current virtual value IS0, switching frequency F0, substrate temperature T0And loss power
P0, it is determined as training data；
S34: initial radial basis function neural network is trained by training data, the weight after being trained；
S35: replacing with the weight after training for the weight of initial radial basis function neural network, obtains preset radial base
Function Neural Network.
Training data, referring to fig. 4, the training data are obtained it should be noted that training data can be built and obtain system
Acquisition system includes: radiator 41, host computer 42, IGBT module 43 and engine 44；Wherein inside the radiator 41
It is provided with accommodating space, which is located in the accommodating space, and the accommodating space is filled with coolant liquid 45；This is dissipated
41 bottom of thermal is radiating module 411, for radiating to the IGBT module 43, the side wall of the radiator 41 and top
Portion is thermal insulation layer 412 made of heat-barrier material；The radiating module 411 is electrically connected with host computer 42, IGBT module 43 respectively with
Host computer 42 and engine 44 are electrically connected, and can simulate working condition.
Referring to Fig. 5, system is obtained to training data and carries out control to obtain the process of training data, including；
S51: start；
S52: PC control IGBT module working condition；It is sent and is instructed by host computer, by the DC terminal of IGBT module
Voltage, phase current virtual value, switching frequency are respectively set to the first preset value, the second preset value and third preset value；Control
IGBT module simulates working condition；
S53: host computer adjusts radiating module working condition and reaches thermal balance；Specific PC control radiating module carries out
Heat dissipation, and substrate temperature is detected, reach thermal balance if variation in the substrate temperature continuous predetermined time is no more than 0.5 DEG C；Its
The middle predetermined time can be with sets itself, such as can be 600 seconds；
S54: the heat radiation power of host computer calculating radiator；The heat that IGBT module generates when thermal balance is equal to heat dissipation dress
The heat distributed is set, the heat radiation power of radiator is equal to the loss power of IGBT module；
S55: record current system conditions obtain radial basis function neural network training data；By the first preset value, second
Substrate temperature and heat radiation power when preset value, third preset value, thermal balance, as one group of training data；
S56: terminate；
Wherein during S52-S55, S57 may be performed simultaneously: the state of monitoring IGBT module and radiating module, if
It breaks down and then directly executes S56, S57 is continued to execute if not breaking down.
Such as Fig. 6, running parameter includes: DC terminal voltage VDC, phase current virtual value IS, switching frequency F and substrate temperature
T1.By running parameter, inputs a preset radial basis function neural network and calculated, obtain radial basis function neural network
Output valve, comprising:
According to DC terminal voltage VDC, phase current virtual value IS, switching frequency F and substrate temperature T1, establish input vector
[VDC IS F T1]T；
By input vector [VDC IS F T1]TIt is input to preset radial basis function neural network:
Wherein, y (x, w) indicates the output valve of radial basis function neural network；X indicates input vector；wiAfter indicating training
Weight；L indicates hidden neuron quantity；ciIndicate center vector；‖x-ci‖ indicates input vector to the distance of center vector；
φ is radial basis function；
Obtain the output valve of radial basis function neural network.
It should be noted that hidden neuron quantity can be 9, radial basis function can be Gaussian radial basis function, diameter
It is loss power P to the output valve of basis function neural network.
As Figure 7-8, another aspect according to the present invention provides a kind of loss of insulated gate bipolar transistor
The acquisition device of power, comprising:
First obtain module 71, for obtain insulated gate bipolar transistor it is in running order when running parameter；
First computing module 72, for inputting running parameter a preset radial basis function neural network and being calculated,
Obtain the output valve of radial basis function neural network；
First determining module 73, for being determined as the loss power of insulated gate bipolar transistor for output valve.
Second obtains module 74, for obtaining DC terminal voltage when insulated gate bipolar transistor simulation working condition
VDC0, phase current virtual value IS0And switching frequency F0；
Third obtains module 75, the substrate temperature T for insulated gate bipolar transistor within a preset time0Fluctuation model
When enclosing less than fluctuation threshold, loss power P is obtained0；
Second determining module 76 is used for DC terminal voltage VDC0, phase current virtual value IS0, switching frequency F0, substrate temperature
T0And loss power P0, it is determined as training data；
Training module 77 is trained for being trained by training data to initial radial basis function neural network
Weight afterwards；
Second computing module 78, for the weight of initial radial basis function neural network to be replaced with to the weight after training,
Obtain preset radial basis function neural network.
Wherein, preset radial basis function neural network is trained according to initial radial basis function neural network, is obtained
It arrives；
Initial radial basis function neural network are as follows:
Wherein, wherein y (x, w) indicate initial radial basis function neural network output valve；X indicates input vector；wiTable
Show weight；L indicates hidden neuron quantity；ciIndicate center vector；‖x-ci‖ indicates input vector to the distance of center vector；
φ is radial basis function.Running parameter includes: DC terminal voltage VDC, phase current virtual value IS, switching frequency F and substrate temperature
T1。
First computing module 72, comprising:
First computing unit 721, for according to DC terminal voltage VDC, phase current virtual value IS, switching frequency F and substrate
Temperature T1, establish input vector [VDC IS F T1]T；
Second computing unit 722 is used for input vector [VDC IS F T1]TIt is input to preset Radial Basis Function neural
Network:
Wherein, y (x, w) indicates the output valve of radial basis function neural network；X indicates input vector；wiAfter indicating training
Weight；L indicates hidden neuron quantity；ciIndicate center vector；‖x-ci‖ indicates input vector to the distance of center vector；
φ is radial basis function；
Third computing unit 723, for obtaining the output valve of radial basis function neural network.
In the embodiment of the present invention, by running parameter of insulated gate bipolar transistor when in running order, as radial direction
The input vector of basis function neural network calculates the loss of insulated gate bipolar transistor by radial basis function neural network
Power can accurately calculate the damage of insulated gate bipolar transistor using the non-linear behavior of radial basis function neural network
Wasted work rate, and there is preferable Project Realization.
Another aspect according to the present invention provides a kind of being set for the loss power of insulated gate bipolar transistor
It is standby, comprising: memory, processor and storage on a memory and the computer program that can run on a processor, computer journey
The acquisition of the loss power for the insulated gate bipolar transistor that above-mentioned each inventive embodiments provide is realized when sequence is executed by processor
The step of method.
In the embodiment of the present invention, by running parameter of insulated gate bipolar transistor when in running order, as radial direction
The input vector of basis function neural network calculates the loss of insulated gate bipolar transistor by radial basis function neural network
Power can accurately calculate the damage of insulated gate bipolar transistor using the non-linear behavior of radial basis function neural network
Wasted work rate, and there is preferable Project Realization.
Another aspect according to the present invention provides a kind of computer readable storage medium, computer-readable storage medium
It is stored with computer program in matter, the insulated gate that above-mentioned each inventive embodiments provide is realized when computer program is executed by processor
The step of preparation method of the loss power of bipolar junction transistor.
In the embodiment of the present invention, by running parameter of insulated gate bipolar transistor when in running order, as radial direction
The input vector of basis function neural network calculates the loss of insulated gate bipolar transistor by radial basis function neural network
Power can accurately calculate the damage of insulated gate bipolar transistor using the non-linear behavior of radial basis function neural network
Wasted work rate, and there is preferable Project Realization.
Although the preferred embodiment of the embodiment of the present invention has been described, once a person skilled in the art knows bases
This creative concept, then additional changes and modifications can be made to these embodiments.So the following claims are intended to be interpreted as
Including preferred embodiment and fall into all change and modification of range of embodiment of the invention.
Finally, it is to be noted that, herein, relational terms such as first and second and the like be used merely to by
One entity or operation are distinguished with another entity or operation, without necessarily requiring or implying these entities or operation
Between there are any actual relationship or orders.Moreover, the terms "include", "comprise" or its any other variant meaning
Covering non-exclusive inclusion, so that process, method, article or terminal device including a series of elements not only wrap
Those elements are included, but also including other elements that are not explicitly listed, or further includes for this process, method, article
Or the element that terminal device is intrinsic.In the absence of more restrictions, being wanted by what sentence "including a ..." limited
Element, it is not excluded that there is also other identical elements in process, method, article or the terminal device for including the element.
Claims (12)
1. a kind of preparation method of the loss power of insulated gate bipolar transistor characterized by comprising
Running parameter when acquisition insulated gate bipolar transistor is in running order；
By the running parameter, inputs a preset radial basis function neural network and calculated, obtain the radial basis function
The output valve of neural network；
By the output valve, it is determined as the loss power of the insulated gate bipolar transistor.
2. the preparation method of the loss power of insulated gate bipolar transistor according to claim 1, which is characterized in that institute
Stating preset radial basis function neural network is trained according to initial radial basis function neural network, is obtained；
The initial radial basis function neural network are as follows:
Wherein, wherein y (x, w) indicate initial radial basis function neural network output valve；X indicates input vector；wiIndicate power
Weight；L indicates hidden neuron quantity；ciIndicate center vector；||x-ci| | the distance of expression input vector to center vector；φ
For radial basis function.
3. the preparation method of the loss power of insulated gate bipolar transistor according to claim 2, which is characterized in that root
It is trained according to initial radial basis function neural network, obtains preset radial basis function neural network, comprising:
Obtain the DC terminal voltage V when insulated gate bipolar transistor simulation working conditionDC0, phase current virtual value IS0With
And switching frequency F0；
The substrate temperature T of the insulated gate bipolar transistor within a preset time0Fluctuation range be less than fluctuation threshold when, obtain
Take loss power P0；
By DC terminal voltage VDC0, phase current virtual value IS0, switching frequency F0, substrate temperature T0And loss power P0, it is determined as
Training data；
The initial radial basis function neural network is trained by the training data, the weight after being trained；
The weight of the initial radial basis function neural network is replaced with into the weight after the training, obtains preset radial base
Function Neural Network.
4. the preparation method of the loss power of insulated gate bipolar transistor according to claim 3, which is characterized in that institute
Stating running parameter includes: DC terminal voltage VDC, phase current virtual value Is, switching frequency F and substrate temperature T1。
5. the preparation method of the loss power of insulated gate bipolar transistor according to claim 4, which is characterized in that will
The running parameter, one preset radial basis function neural network of input are calculated, and the radial ba-sis function network is obtained
The output valve of network, comprising:
According to the DC terminal voltage VDC, phase current virtual value Is, switching frequency F and substrate temperature T1, establish input vector
[VDC Is F T1]T；
By the input vector [VDC Is F T1]TIt is input to preset radial basis function neural network:
Wherein, y (x, w) indicates the output valve of radial basis function neural network；X indicates input vector；wiAfter indicating the training
Weight；L indicates hidden neuron quantity；ciIndicate center vector；||x-ci| | the distance of expression input vector to center vector；
In be radial basis function；
Obtain the output valve of the radial basis function neural network.
6. a kind of acquisition device of the loss power of insulated gate bipolar transistor characterized by comprising
First obtain module, for obtain insulated gate bipolar transistor it is in running order when running parameter；
First computing module, for inputting the running parameter a preset radial basis function neural network and being calculated, obtained
Obtain the output valve of the radial basis function neural network；
First determining module, for being determined as the loss power of the insulated gate bipolar transistor for the output valve.
7. the acquisition device of the loss power of insulated gate bipolar transistor according to claim 6, which is characterized in that institute
Stating preset radial basis function neural network is trained according to initial radial basis function neural network, is obtained；
The initial radial basis function neural network are as follows:
Wherein, wherein y (x, w) indicate initial radial basis function neural network output valve；X indicates input vector；wiIndicate power
Weight；L indicates hidden neuron quantity；ciIndicate center vector；||x-ci| | the distance of expression input vector to center vector；φ
For radial basis function.
8. the acquisition device of the loss power of insulated gate bipolar transistor according to claim 7, which is characterized in that also
Include:
Second obtains module, for obtaining the DC terminal voltage V when insulated gate bipolar transistor simulation working conditionDC0、
Phase current virtual value IS0And switching frequency F0；
Third obtains module, the substrate temperature T for the insulated gate bipolar transistor within a preset time0Fluctuation range
When less than fluctuation threshold, loss power P is obtained0；
Second determining module is used for DC terminal voltage VDC0, phase current virtual value IS0, switching frequency F0, substrate temperature T0And
Loss power P0, it is determined as training data；
Training module is instructed for being trained by the training data to the initial radial basis function neural network
Weight after white silk；
Second computing module, for the weight of the initial radial basis function neural network to be replaced with to the power after the training
Weight, obtains preset radial basis function neural network.
9. the acquisition device of the loss power of insulated gate bipolar transistor according to claim 8, which is characterized in that institute
Stating running parameter includes: DC terminal voltage VDC, phase current virtual value Is, switching frequency F and substrate temperature T1。
10. the acquisition device of the loss power of insulated gate bipolar transistor according to claim 9, which is characterized in that
First computing module, comprising:
First computing unit, for according to the DC terminal voltage VDC, phase current virtual value Is, switching frequency F and substrate temperature
Spend T1, establish input vector [VDC Is F T1]T；
Second computing unit is used for the input vector [VDC Is F T1]TIt is input to preset radial ba-sis function network
Network:
Wherein, y (x, w) indicates the output valve of radial basis function neural network；X indicates input vector；wiAfter indicating the training
Weight；L indicates hidden neuron quantity；ciIndicate center vector；||x-ci| | the distance of expression input vector to center vector；
φ is radial basis function；
Third computing unit, for obtaining the output valve of the radial basis function neural network.
11. a kind of acquisition equipment of the loss power of insulated gate bipolar transistor characterized by comprising memory, processing
Device and storage are on a memory and the computer program that can run on a processor, the computer program are held by the processor
The preparation method for realizing the loss power of the insulated gate bipolar transistor as described in any one of claims 1 to 5 when row
Step.
12. a kind of computer readable storage medium, which is characterized in that be stored with computer on the computer readable storage medium
Program realizes the insulated gate bipolar as described in any one of claims 1 to 5 when the computer program is executed by processor
The step of preparation method of the loss power of transistor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811454690.2A CN109558681A (en) | 2018-11-30 | 2018-11-30 | A kind of preparation method and device of the loss power of insulated gate bipolar transistor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811454690.2A CN109558681A (en) | 2018-11-30 | 2018-11-30 | A kind of preparation method and device of the loss power of insulated gate bipolar transistor |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109558681A true CN109558681A (en) | 2019-04-02 |
Family
ID=65868259
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811454690.2A Pending CN109558681A (en) | 2018-11-30 | 2018-11-30 | A kind of preparation method and device of the loss power of insulated gate bipolar transistor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109558681A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112297771A (en) * | 2019-07-31 | 2021-02-02 | 北京新能源汽车股份有限公司 | Permanent magnet synchronous motor heat management control method and device and automobile |
| WO2022248532A1 (en) * | 2021-05-25 | 2022-12-01 | Danmarks Tekniske Universitet | Data-driven and temperature-cycles based remaining useful life estimation of an electronic device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103715935A (en) * | 2013-11-27 | 2014-04-09 | 国家电网公司 | Modularized multi-level voltage source type converter-based loss determination method |
| CN104199536A (en) * | 2014-07-23 | 2014-12-10 | 西安空间无线电技术研究所 | FPGA dynamic power consumption estimation method based on BP neural network |
| CN107153735A (en) * | 2017-04-28 | 2017-09-12 | 无锡开放大学 | Motor driven systems PWM inverter method for diagnosing faults |
| CN107357997A (en) * | 2017-07-18 | 2017-11-17 | 合肥工业大学 | A kind of analogy method and system of hydraulic pressure process for machining |
| US20180284181A1 (en) * | 2015-11-05 | 2018-10-04 | Crrc Zhuzhou Institute Co., Ltd. | On-line health management device and method for insulated gate bipolar transistor |
-
2018
- 2018-11-30 CN CN201811454690.2A patent/CN109558681A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103715935A (en) * | 2013-11-27 | 2014-04-09 | 国家电网公司 | Modularized multi-level voltage source type converter-based loss determination method |
| CN104199536A (en) * | 2014-07-23 | 2014-12-10 | 西安空间无线电技术研究所 | FPGA dynamic power consumption estimation method based on BP neural network |
| US20180284181A1 (en) * | 2015-11-05 | 2018-10-04 | Crrc Zhuzhou Institute Co., Ltd. | On-line health management device and method for insulated gate bipolar transistor |
| CN107153735A (en) * | 2017-04-28 | 2017-09-12 | 无锡开放大学 | Motor driven systems PWM inverter method for diagnosing faults |
| CN107357997A (en) * | 2017-07-18 | 2017-11-17 | 合肥工业大学 | A kind of analogy method and system of hydraulic pressure process for machining |
Non-Patent Citations (3)
| Title |
|---|
| 唐莹莹等: "基于径向基函数神经网络的功率器件集总参数热模型", 《科学技术与工程》 * |
| 李志刚 等: ""IGBT 模块开关损耗计算方法综述"", 《电子技术应用》 * |
| 米林等: "基于径向基函数神经网络的电动汽车动力电池SOC模型", 《重庆理工大学学报(自然科学)》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112297771A (en) * | 2019-07-31 | 2021-02-02 | 北京新能源汽车股份有限公司 | Permanent magnet synchronous motor heat management control method and device and automobile |
| WO2022248532A1 (en) * | 2021-05-25 | 2022-12-01 | Danmarks Tekniske Universitet | Data-driven and temperature-cycles based remaining useful life estimation of an electronic device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107219016B (en) | Calculate the method and system of IGBT module transient state junction temperature | |
| CN110765601B (en) | IGBT junction temperature estimation method based on IGBT thermoelectric coupling model | |
| CN105825019B (en) | A kind of insulated gate bipolar transistor IGBT module temperature derivation algorithm | |
| CN111781480B (en) | IGBT junction temperature monitoring method, device and system | |
| CN110161398A (en) | A method of IGBT power module ageing state is assessed using shell temperature | |
| CN104091203B (en) | A kind of reliability in time appraisal procedure of current transformer used for wind power generation | |
| CN108896200B (en) | Method, device, equipment and medium for detecting temperature of power module of converter | |
| CN108647447B (en) | Reliability analysis method and device for MMC converter valve | |
| CN109406004A (en) | A kind of switchgear temperature rise analytical method | |
| CN108802590A (en) | A kind of the power circulation test method and test system of semiconductor devices | |
| CN106255392B (en) | The cooling control method of electrical cabinet, device and system | |
| CN107192934A (en) | A kind of measuring method of crust transient thermal impedance for high-power IGBT | |
| CN103956887A (en) | Wind power converter IGBT module junction temperature online computing method | |
| CN104217130A (en) | Method for calculating loss of MMC (Modular Multilevel Converter) | |
| CN109558681A (en) | A kind of preparation method and device of the loss power of insulated gate bipolar transistor | |
| CN111505475B (en) | Method and device for calibrating parameters of electric heating model of power semiconductor module | |
| CN103970947A (en) | Real-time temperature rise measurement method of electromagnetic system of low-voltage apparatus based on finite element analysis | |
| CN109613352A (en) | Method, device and system for testing service life of power module in converter | |
| CN103870612B (en) | A kind of system and method obtaining IGBT device thermal resistance | |
| CN108072819A (en) | The abatement detecting method and device of IGBT | |
| CN111060797A (en) | IGBT module health state monitoring method based on natural frequency of heat network | |
| Shen et al. | Mathematical modeling and simulation for external electrothermal characteristics of an alkaline water electrolyzer | |
| Inui et al. | Detailed estimation method of heat generation during charge/discharge in lithium‐ion battery using equivalent circuit | |
| JP5042396B2 (en) | Power supply and controller | |
| CN109556759A (en) | A kind of experimental method for probing into switchgear the Temperature Rising Law |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190402 |
|
| RJ01 | Rejection of invention patent application after publication |