CN105825019A - Temperature solving algorithm of insulated gate bipolar transistor (IGBT) module - Google Patents

Abstract

The invention discloses a temperature solving algorithm of an insulated gate bipolar transistor (IGBT) module. The temperature solving algorithm of the IGBT module comprises the following steps of firstly, extracting the fitting parameters of a power loss model, thermal resistance and thermal capacitance parameters of a seven-order equivalent Cauer heat transfer network model and an ambient temperature of the IGBT module, afterwards, judging the states in which an IGBT and a freewheeling diode (FWD) are in a current iteration cycle according to gate triggering signals, which are detected from an electrical network model, of the current iteration cycle and a previous iteration cycle, subsequently, calculating the power losses of the IGBT and the FWD in the corresponding states of the current iteration cycle, and finally, calculating the temperature of the IGBT module by the heat transfer network model of the IGBT module through synthesizing the power loss and the ambient temperature of the current iteration cycle. By using the temperature solving algorithm of the IGBT module, which is provided by the invention, not only can the real-time calculation of the temperature of the IGBT module be realized, but also foundations can be laid for the aspects of the heat radiation design, the performance optimization, the reliability evaluation and the like of the IGBT module.

Description

A kind of insulated gate bipolar transistor IGBT module temperature derivation algorithm

Technical field

The present invention relates to a kind of temperature derivation algorithm, be specifically related to a kind of insulated gate bipolar transistor IGBT module temperature derivation algorithm.

Background technology

Igbt (IGBT) module temperature not only directly affects the heat dissipation design of IGBT module, service behaviour and service life, but also the long-term reliability of converter system can be directly influenced, therefore obtain IGBT module temperature and run significant for the normal safe guaranteeing converter system.

The existing acquisition methods about IGBT module silicon PN junction temperature (junction temperature), skin temperature is more.Chinese patent ZL:201410205679.8 is disclosed " wind electric converter IGBT module junction temperature on-line calculation method ", on the premise of considering IGBT module electro thermal coupling characteristic, is lost in line computation IGBT module junction temperature based on switch periods；Chinese patent ZL:201310442045.X is disclosed " real-time estimating method of IGBT module shell temperature ", simulates the heat transfer process of IGBT module shell, and then estimation IGBT module skin temperature by setting up the thermal resistance heater circuit model in parallel with thermal capacitance；Chinese patent ZL:201410345265.5 is disclosed " on-line detecting system of a kind of IGBT module working junction temperature and detection method ", utilize the temperature sensitive time under fixing shutoff voltage, current conditions with the substantial connection of IGBT module working junction temperature, calculate junction temperature by extracting temperature sensitive time indirect；Chinese patent ZL:201410853960.2 disclosed " steady temperature of a kind of IGBT module calculates method ", by the power attenuation of input IGBT module, crust thermal resistance, thermal conductive contact material thermal resistance, the thermal resistance of radiator to environment, calculate IGBT module silicon PN junction, the steady temperature of shell.But, above-mentioned patent is all not directed to the calculating of IGBT module temperature, because IGBT module not only includes silicon PN junction, shell, also includes the structure sheafs such as solder layer, upper layers of copper, ceramic layer, lower layers of copper, lower solder layer and substrate.In sum, there is no the acquisition methods about IGBT module temperature at present.

Therefore, invent a kind of brand-new IGBT module temperature derivation algorithm, significant with application efficiency for promoting IGBT module work in engineer applied designs.

Summary of the invention

Problem to be solved by this invention is for above-mentioned existing model and the deficiency of method, it is provided that a kind of insulated gate bipolar transistor IGBT module temperature derivation algorithm.

The technical solution adopted in the present invention is:

To achieve these goals, the present invention provides following technical scheme:

Step one: extract power attenuation model fitting parameter a, b, c, d of IGBT module and 7 rank equivalence Cauer heat transfer network model thermal resistance thermal capacitance parameter R, C and ambient temperature T_A；

Step 2: trigger signal V according to the gate pole of the current iteration cycle detected from electrical network model and a upper iteration cycle_G[k]、V_G[k-1], it is judged that IGBT, anti-parallel diodes (FWD), in current iteration cycle state in which, i.e. turn on, are switched on or off state；

Step 3: flow through the current/voltage I of IGBT, FWD based on the current iteration cycle_C[k]、V_CE[k]、I_F[k]、V_D[k] and junction temperature T_Tj[k]、T_Dj[k], calculates IGBT, FWD power attenuation P at current iteration cycle corresponding state_loss[k]；

Step 4: power attenuation P in IGBT module heat transfer network model's comprehensive current iteration cycle_loss[k] and ambient temperature T_A, solve IGBT module temperature T [k+1]；

Step 5: if the temperature iteration variable in adjacent output cycle meets the condition of convergence, then solve calculating and terminate, power attenuation P of output IGBT module_loss[k] and each layer temperature T [k]；If the temperature iteration variable in adjacent output cycle is unsatisfactory for the condition of convergence, then by T [k], V_G[k-1] is updated to T [k+1], V respectively_G[k], repetition step 2 is to four, until the temperature iteration variable in adjacent output cycle meets the condition of convergence.

Further, step 2 judges that IGBT, FWD in the standard of current iteration cycle status are:

Further, the IGBT conduction loss P in step 3_Tcon, IGBT turn-on consumption P_on, IGBT turn-off power loss P_off, FWD conduction loss P_Dcon, FWD turn-off power loss (also known as reverse recovery loss) P_rrIt is respectively as follows:

P_Tcon(T_Tj,I_C)=a_T·T_Tj+b_T

In above formula, a_TAnd b_TIt is respectively as follows:

$a_T=\frac{P_{Tcon}(T_{T\min },I_C)-P_{T\mathrm{co}n}(T_{T\max },I_C)}{T_{T\min }-T_{T\max }}$

b_T=P_Tcon(T_Tmin,I_C)-a_T·T_Tmin

P_Tcon(T_Tmin,I_C)=a₁·I_C ³+b₁·I_C ²+c₁·I_C+d₁

P_Tcon(T_Tmax,I_C)=a₂·I_C ³+b₂·I_C ²+c₂·I_C+d₂

Fitting coefficient a in above formula₁、b₁、c₁、d₁、a₂、b₂、c₂、d₂Can be obtained by MATLAB Fitting Toolbox matching.

FWD conduction loss P_DconComputational methods and above-mentioned IGBT conduction loss P_TconComputational methods are similar to.

IGBT turn-on consumption P_onFor:

P_on=E_on·f_sw·V_CE/V_rated

In above formula, f_swFor IGBT switching frequency；V_CERunning voltage, V is penetrated for IGBT collection_ratedFor IGBT rated operational voltage；Open ENERGY E_onFor:

E_on=a_on·T_Tj+b_on

E_on(T_Tmin,I_C)=a₃·I_C ³+b₃·I_C ²+c₃·I_C+d₃

E_on(T_Tmax,I_C)=a₄·I_C ³+b₄·I_C ²+c₄·I_C+d₄

$a_{on}=\frac{P_{on}(T_{Tmin},I_C)-P_{on}(T_{Tmax},I_C)}{T_{T\min }-T_{Tmax}}$

b_on=P_on(T_Tmin,I_C)-a_on·T_Tmin

Fitting coefficient a in above formula₃、b₃、c₃、d₃、a₄、b₄、c₄、d₄Can be obtained by MATLAB Fitting Toolbox matching.

IGBT turn-off power loss P_off, FWD reverse recovery loss P_rrComputational methods and above-mentioned IGBT conduction loss P_TconComputational methods are similar to.

Further, IGBT module in step 4 heat transfer network model is to be conducted heat network model, through Runge-Kutta method difference processing gained by the IGBT module 7 rank equivalence Cauer corresponding with accompanying drawing 1, it may be assumed that

In above formula, k=0,1,2,3 ...；Matrix T represents the matrix of IGBT module the most each layer temperature, i.e. T=[T₁T₂T₃T₄T₅T₆T₇]^T；U=[P_lossT_A]^T；t_simFor simulation step length；Matrix A, B, K₁、K₂、K₃、K₄It is shown below respectively, wherein, R₈For heat radiator thermal resistance.

$A=\left[\begin{array}{ccccccc}\frac{-1}{R_1{C}_1}& \frac{-1}{R_1{C}_1}& 0& 0& 0& 0& 0\\ \frac{1}{R_1{C}_2}& -(\frac{1}{R_1}+\frac{1}{R_2})g\frac{1}{C_2}& \frac{1}{R_2{C}_2}& 0& 0& 0& 0\\ 0& \frac{1}{R_2{C}_3}& -(\frac{1}{R_2}+\frac{1}{R_3})g\frac{1}{C_3}& \frac{1}{R_3{C}_3}& 0& 0& 0\\ 0& 0& \frac{1}{R_3{C}_4}& -(\frac{1}{R_3}+\frac{1}{R_4})g\frac{1}{C_4}& \frac{1}{R_4{C}_4}& 0& 0\\ 0& 0& 0& \frac{1}{R_4{C}_5}& -(\frac{1}{R_4}+\frac{1}{R_5})g\frac{1}{C_5}& \frac{1}{R_5{C}_5}& 0\\ 0& 0& 0& 0& \frac{1}{R_5{C}_6}& -(\frac{1}{R_5}+\frac{1}{R_6})g\frac{1}{C_6}& \frac{1}{R_6{C}_6}\\ 0& 0& 0& 0& 0& \frac{1}{R_6{C}_7}& -(\frac{1}{R_6}+\frac{1}{R_7+R_8})g\frac{1}{C_7}\end{array}\right]B=\left[\begin{array}{cc}\frac{1}{C_1}& 0\\ 0& 0\\ 0& 0\\ 0& 0\\ 0& 0\\ 0& 0\\ 0& \frac{1}{(R_7+R_8){\mathrm{gC}}_7}\end{array}\right]$

K₁=A T [k]+B U [k]

$K_2=A\·(T\[k\]+\frac{t_{sim}}{2}\·K_1)+B\·U(t_k+\frac{t_{sim}}{2},T\[k\]+\frac{t_{sim}}{2}\·K_1)$

$K_3=A\·(T\[k\]+\frac{t_{sim}}{2}\·K_2)+B\·U(t_k+\frac{t_{sim}}{2},T\[k\]+\frac{t_{sim}}{2}\·K_2)$

K₄=A (T [k]+t_sim·K₃)+B·U(t_k+1,T[k]+t_sim·K₃)

Further, the condition of convergence of step 5 is:

||T[k+n]-T[k]||_∞≤M

In formula, M is predetermined accuracy, and n is the iterative computation number of times in an output cycle.

One insulated gate bipolar transistor IGBT module temperature derivation algorithm of the present invention, can be on the basis of ensureing computational accuracy, avoid the problems such as modeling process is complicated, model parameter extraction difficulty, amount of data storage are excessive, emulation is the most serious, for IGBT module and the efficient design of converter system radiator structure, service behaviour and the raising in service life, and the aspect such as the Topology Structure Design of project of transmitting and converting electricity high voltage direct current converter valve, SVC, THE UPFC, long-term reliability assessment provides service and lays the foundation.

Accompanying drawing explanation

With specific embodiment, the present invention is described in further detail below in conjunction with the accompanying drawings:

Fig. 1 is igbt (IGBT) module equivalence Cauer heat transfer network model schematic.

Fig. 2 is IGBT module temperature derivation algorithm schematic diagram.

Fig. 3 is IGBT module vertical layer structure schematic diagram.

Fig. 4 is IGBT module temperature simulation resolution principle block diagram in two level three-phase voltage source inverter.

Fig. 5 is IGBT module igbt chip and layers below temperature thereof in two level three-phase voltage source inverter.

Fig. 6 is IGBT module FWD chip and layers below temperature thereof in two level three-phase voltage source inverter.

Detailed description of the invention

As in figure 2 it is shown, a kind of insulated gate bipolar transistor IGBT module temperature derivation algorithm, specifically include following steps:

Step one: extract power attenuation model fitting parameter a, b, c, d of IGBT module.

First, IGBT minimum representative temperature T IGBT module product data handbook provided_TminWith maximum representative temperature T_TmaxUnder V_CE-I_COutput characteristic curve, changes into corresponding P_Tcon-I_CCharacteristic curve；Then, IGBT on-state loss P in MATLAB Fitting Toolbox_TconUse collector current I_CFitting of a polynomial, obtain fitting coefficient a₁、b₁、c₁、d₁And a₂、b₂、c₂、d₂.FWD on-state loss P_TconFitting coefficient extracting method be similar to.

The IGBT minimum representative temperature T provided according to IGBT module product data handbook_TminWith maximum representative temperature T_TmaxUnder E_on-I_CCharacteristic curve, opening ENERGY E in MATLAB Fitting Toolbox_onUse collector current I_CFitting of a polynomial, obtain fitting coefficient a₃、b₃、c₃、d₃And a₄、b₄、c₄、d₄；IGBT turns off ENERGY E_on, FWD Reverse recovery ENERGY E_rrFitting coefficient extracting method be similar to.

Step 2: extract heat transfer network model thermal resistance thermal capacitance parameter R, C and ambient temperature T of IGBT module_A。

As shown in Figure 3, inside IGBT module, there are 7 different layers, are respectively as follows: silicon, upper solder layer, upper layers of copper, ceramic layer, lower layers of copper, lower solder layer and substrate from top to bottom.IGBT module 7 rank equivalence Foster heat transfer network model transient thermal impedance expression formula be shown below.

$Z_{th(j-c)}\left(t\right)=\underset{i=1}{\overset{7}{\Σ}}{r}_i(1-e^{-t/{\mathrm{\τ}}_i})$

In formula, r_i、τ_iIt is respectively the equivalence Foster heat transfer thermal resistance of network model i-th layer and time constant.

Owing to IGBT module product data handbook generally provides IGBT, FWD transient thermal impedance curve recorded by experiment, therefore it is carried out 7 rank exponential series matchings, each unknown parameters ' value of above formula can be obtained.Above formula is carried out Laplace transform, obtains:

$Z_{h(j-c)}\left(s\right)=\underset{i=1}{\overset{7}{\Σ}}\frac{r_i}{{\mathrm{\τ}}_{i}s+1}$

According to the definition of complex impedance, the thermal impedance of 7 rank equivalence Cauer heat transfer network modeies is:

$Z_{h(j-c)}^{\′}\left(s\right)=\frac{1}{{\mathrm{sC}}_1+\frac{1}{R_1+\frac{1}{{\mathrm{sC}}_2+...+\frac{1}{R_7}}}}$

According to the principle that IGBT module equivalence Foster, Cauer heat transfer network model thermal impedance is identical, i.e. Z_th(j-c)(s)=Z ＇_th(j-c)(s), each rank thermal resistance R of network model of can conducting heat in the hope of 7 rank equivalence Cauer_iWith thermal capacitance C_i, i.e. the thermal resistance of IGBT module each physical arrangement layer and thermal capacitance value.

Step 3: trigger signal V according to the gate pole of the current iteration cycle detected from electrical network model and a upper iteration cycle_G[k]、V_G[k-1], it is judged that IGBT, FWD, in current iteration cycle state in which, i.e. turn on, are switched on or off state.

Judge that IGBT, FWD in the standard of current iteration cycle status are:

Step 4: flow through IGBT, the current/voltage I of anti-parallel diodes (FWD) based on the current iteration cycle_C[k]、V_CE[k]、I_F[k]、V_D[k] and junction temperature T_Tj[k]、T_Dj[k], calculates IGBT, FWD power attenuation P at current iteration cycle corresponding state_loss[k]。

IGBT conduction loss P_Tcon, IGBT turn-on consumption P_on, IGBT turn-off power loss P_offFWD conduction loss P_Dcon, FWD turn-off power loss (also known as reverse recovery loss) P_rrComputing formula is respectively as follows:

P_Tcon(T_Tj,I_C)=a_T·T_Tj+b_T

In above formula, a_TAnd b_TIt is respectively as follows:

$a_T=\frac{P_{Tcon}(T_{T\min },I_C)-P_{Tcon}(T_{Tmax},I_C)}{T_{T\min }-T_{Tmax}}$

b_T=P_Tcon(T_Tmin,I_C)-a_T·T_Tmin

P_Tcon(T_Tmin,I_C)=a₁·I_C ³+b₁·I_C ²+c₁·I_C+d₁

P_Tcon(T_Tmax,I_C)=a₂·I_C ³+b₂·I_C ²+c₂·I_C+d₂

FWD conduction loss P_DconComputational methods and above-mentioned IGBT conduction loss P_TconComputational methods are similar to.

IGBT turn-on consumption P_onComputing formula is:

P_on=E_on·f_sw·V_CE/V_rated

In above formula, f_swFor IGBT switching frequency；V_CERunning voltage, V is penetrated for IGBT collection_ratedFor IGBT rated operational voltage；Open ENERGY E_onFor:

E_on=a_on·T_Tj+b_on

E_on(T_Tmin,I_C)=a₃·I_C ³+b₃·I_C ²+c₃·I_C+d₃

E_on(T_Tmax,I_C)=a₄·I_C ³+b₄·I_C ²+c₄·I_C+d₄

$a_{on}=\frac{P_{on}(T_{Tmin},I_C)-P_{on}(T_{Tmax},I_C)}{T_{T\min }-T_{T\max }}$

b_on=P_on(T_Tmin,I_C)-a_on·T_Tmin

IGBT turn-off power loss P_off, FWD reverse recovery loss P_rrComputational methods and above-mentioned IGBT conduction loss P_TconComputational methods are similar to.

Step 5: the loss P in IGBT module heat transfer network model's comprehensive current iteration cycle_loss[k] and ambient temperature T_A, solve the temperature matrices T [k+1] of IGBT module.

IGBT module heat transfer network model be shown below, by the loss P in IGBT module current iteration cycle_loss[k] and ambient temperature T_ASubstitute into wherein, can calculate and try to achieve IGBT module temperature matrices T [k+1].

$T\[k+1\]=T\[k\]+\frac{t_{\mathrm{si}m}}{6}(K_1+2K_2+2K_3+K_4)$

In above formula, k=0,1,2,3 ...；Matrix T represents the matrix of IGBT module the most each layer temperature, i.e. T=[T₁T₂T₃T₄T₅T₆T₇]^T；t_simFor simulation step length；K₁、K₂、K₃、K₄It is shown below respectively.

K₁=A T [k]+B U [k]

$K_2=A\·(T\[k\]+\frac{t_{sim}}{2}\·K_1)+B\·U(t_k+\frac{t_{sim}}{2},T\[k\]+\frac{t_{sim}}{2}\·K_1)$

$K_3=A\·(T\[k\]+\frac{t_{sim}}{2}\·K_2)+B\·U(t_k+\frac{t_{sim}}{2},T\[k\]+\frac{t_{sim}}{2}\·K_2)$

K₄=A (T [k]+t_sim·K₃)+B·U(t_k+1,T[k]+t_sim·K₃)

In above formula, U=[P_lossT_A]^T；Matrix A, B are shown below, wherein, and R₈For heat radiator thermal resistance.

$A=\left[\begin{array}{ccccccc}\frac{-1}{R_1{C}_1}& \frac{-1}{R_1{C}_1}& 0& 0& 0& 0& 0\\ \frac{1}{R_1{C}_2}& -(\frac{1}{R_1}+\frac{1}{R_2})g\frac{1}{C_2}& \frac{1}{R_2{C}_2}& 0& 0& 0& 0\\ 0& \frac{1}{R_2{C}_3}& -(\frac{1}{R_2}+\frac{1}{R_3})g\frac{1}{C_3}& \frac{1}{R_3{C}_3}& 0& 0& 0\\ 0& 0& \frac{1}{R_3{C}_4}& -(\frac{1}{R_3}+\frac{1}{R_4})g\frac{1}{C_4}& \frac{1}{R_4{C}_4}& 0& 0\\ 0& 0& 0& \frac{1}{R_4{C}_5}& -(\frac{1}{R_4}+\frac{1}{R_5})g\frac{1}{C_5}& \frac{1}{R_5{C}_5}& 0\\ 0& 0& 0& 0& \frac{1}{R_5{C}_6}& -(\frac{1}{R_5}+\frac{1}{R_6})g\frac{1}{C_6}& \frac{1}{R_6{C}_6}\\ 0& 0& 0& 0& 0& \frac{1}{R_6{C}_7}& -(\frac{1}{R_6}+\frac{1}{R_7+R_8})g\frac{1}{C_7}\end{array}\right]B=\left[\begin{array}{cc}\frac{1}{C_1}& 0\\ 0& 0\\ 0& 0\\ 0& 0\\ 0& 0\\ 0& 0\\ 0& \frac{1}{(R_7+R_8){\mathrm{gC}}_7}\end{array}\right]$

Step 6: if the temperature iteration variable in adjacent output cycle meets the condition of convergence, then simulation calculation terminates, power attenuation P of output IGBT module_loss[k] and IGBT module each layer temperature T [k]；If the temperature iteration variable in adjacent output cycle is unsatisfactory for the condition of convergence, then by T [k], V_G[k-1] is updated to T [k+1], V respectively_G[k], repetition step 2 is to four, until the temperature iteration variable in adjacent output cycle meets the condition of convergence.

Wherein, the condition of convergence is:

||T[k+n]-T[k]||_∞≤M

In formula, M is predetermined accuracy, and n is the iterative computation number of times in an output cycle.

The IGBT module temperature derivation algorithm that the invention described above proposes is compiled into dynamic link library file (DLL), and imports in Saber software, set up the temperature calculation models of IGBT module in two level three-phase voltage source inverter.Accompanying drawing 4 is IGBT module temperature simulation resolution principle block diagram in two level three-phase voltage source inverter, and the result of calculation of IGBT module temperature is as shown in accompanying drawing 5,6.Fig. 5 is respectively igbt chip and layers below temperature thereof from top to bottom；Fig. 6 is respectively FWD chip and layers below temperature thereof from top to bottom.

Claims (4)

1. an insulated gate bipolar transistor IGBT module temperature derivation algorithm, it is characterised in that the rapid solving realizing IGBT module temperature comprises the steps of

Step one: extract power attenuation model fitting parameter a, b, c, d of IGBT module and 7 rank equivalence Cauer heat transfer network model thermal resistance thermal capacitance parameter R, C and ambient temperature T_A；

Step 2: trigger signal V according to the gate pole of the current iteration cycle detected from electrical network model and a upper iteration cycle_G[k]、V_G[k-1], it is judged that IGBT, anti-parallel diodes FWD, in current iteration cycle state in which, i.e. turn on, are switched on or off state；

Step 3: flow through the current/voltage I of IGBT, FWD based on the current iteration cycle_C[k]、V_CE[k]、I_F[k]、V_D[k] and junction temperature T_Tj[k]、T_Dj[k], calculates IGBT, FWD power attenuation P at current iteration cycle corresponding state_loss[k]；

Step 4: power attenuation P in IGBT module heat transfer network model's comprehensive current iteration cycle_loss[k] and ambient temperature T_A, solve IGBT module temperature T [k+1]；

Step 5: if the temperature iteration variable in adjacent output cycle meets the condition of convergence, then solve calculating and terminate, power attenuation P of output IGBT module_loss[k] and each layer temperature T [k]；If the temperature iteration variable in adjacent output cycle is unsatisfactory for the condition of convergence, then by T [k], V_G[k-1] is updated to T [k+1], V respectively_G[k], repetition step 2 is to four, until the temperature iteration variable in adjacent output cycle meets the condition of convergence.

A kind of insulated gate bipolar transistor IGBT module temperature derivation algorithm the most according to claim 1, it is characterised in that step 2 judges that IGBT, FWD in the standard of current iteration cycle status are:

A kind of insulated gate bipolar transistor IGBT module temperature derivation algorithm the most according to claim 1, it is characterized in that, in step 4 IGBT module heat transfer network model be by IGBT module 7 rank equivalence Cauer conduct heat network model, through Runge-Kutta method difference processing gained.

A kind of insulated gate bipolar transistor IGBT module temperature derivation algorithm the most according to claim 1, it is characterised in that the condition of convergence of step 5 is:

||T[k+n]-T[k]||_∞≤M

In formula, M is predetermined accuracy, and n is the iterative computation number of times in an output cycle.