CN110543665A - converter multi-scale modeling method based on micro and macro description - Google Patents
converter multi-scale modeling method based on micro and macro description Download PDFInfo
- Publication number
- CN110543665A CN110543665A CN201910664814.8A CN201910664814A CN110543665A CN 110543665 A CN110543665 A CN 110543665A CN 201910664814 A CN201910664814 A CN 201910664814A CN 110543665 A CN110543665 A CN 110543665A
- Authority
- CN
- China
- Prior art keywords
- scale
- layer
- component
- micro
- macro
- 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.)
- Granted
Links
Landscapes
- Semiconductor Integrated Circuits (AREA)
Abstract
The invention discloses a converter multi-scale modeling method based on microscopic and macroscopic descriptions, which comprises the following steps: 1) selecting a topological structure and an auxiliary circuit of the converter, and analyzing the scale of the converter to be observed; 2) determining the level of research to be developed, and selecting a multi-scale component according to the level; 3) selecting physical field factors according to the operation environment of the converter; 4) determining a research mode by using the information; 5) selecting components and nodes for multi-scale observation, and respectively setting consistency observation points, reliability observation points and corresponding error rates delta; 6) and judging the error between the macro scale and the set reference value through the observation point, switching the micro scale to calculate when the error of the macro scale is larger than the set error rate delta, and correcting the parameter of the macro scale by using the obtained result to complete the data conversion between scales. The method introduces the coupling relation between the transient dynamic characteristics of all components and circuits into the modeling of the power electronic converter, and realizes the accurate simulation, control and reliability analysis of the system.
Description
Technical Field
The invention relates to the technical field of design and reliability analysis of electric energy converters, in particular to a converter multi-scale modeling method based on micro and macro description.
Background
the existing power system continuously has various dynamic problems of unknown mechanisms, and the safety, stability and operation of the power system are seriously threatened. The power electronic converter with the highest fault rate is taken as a complex system, many physical phenomena of the power electronic converter have multi-scale characteristics, such as power electronic equipment in power generation, power transmission and power distribution loads, the nonlinear coupling relation is strong, and the interaction process is complex. The observation of the working process in engineering practice is often performed on different scales. The scale is from nanometer scale to kilometer scale and from picosecond scale to annual scale, and the traditional calculation method is mainly based on classical electrodynamic force and is difficult to uniformly describe.
Modeling methods for power electronic converters are mostly performed on a single scale. In many cases, power electronics transient analysis involves different temporal, spatial scales. Therefore, phenomena such as switching transients, PN junction breakdown, over-current damage, fatigue failure, physical factor effects, etc. cannot be observed. In order to analyze and calculate electromagnetic transients, a non-ideal switching model needs to be established. However, mathematical models describing power electronic systems have high order non-linearities and tend to be very rigid. The transient process of the nonlinear system can only be solved by a numerical solution method of a conventional differential equation, the simulation time is long, and the numerical stability is poor. Although quantum methods can achieve the above calculations, due to the limitations of computational resources, all scales cannot be considered simultaneously.
Therefore, there is a need for a generalized concept and methodology to explore modeling, analysis, and control of the problem of dynamic stabilization of systems from the perspective of physical mechanisms. The research on how to combine multi-scale modeling to realize accurate simulation, efficient calculation, architecture optimization, margin reduction, reliability improvement and the like is a problem to be solved urgently in the field of power electronics.
Disclosure of Invention
The invention aims to overcome the defects and shortcomings of the prior art, and provides a converter multi-scale modeling method based on micro and macro descriptions.
in order to achieve the purpose, the technical scheme provided by the invention is as follows: a converter multi-scale modeling method based on micro and macro description is suitable for a power electronic converter, a component model library with multi-scale and multi-physical field interfaces is deduced according to a micro-scale component layer model, and the component model library is used for reflecting micro information of the power electronic converter under the action of physical field factors and comprises the following steps:
1) Selecting a topological structure and an auxiliary circuit of the converter, analyzing the scale of the converter to be observed, and dividing the scale category into a macro scale, a control strategy scale and a micro scale;
2) Determining the level of research to be developed, and selecting a multi-scale component model from a component library according to the type;
3) Selecting a physical field according to the operation environment of the converter, wherein a single physical field is singly added for environmental factors, and a plurality of physical fields are added for coupling analysis in a complex observation environment;
4) Determining a research mode, specifically dividing the research mode into a long-time steady state and a short-time transient state, and using the long-time steady state and the short-time transient state as a reference simulation environment with a micro scale and a macro scale which are independent from each other;
5) Selecting components and nodes for multi-scale observation, respectively setting consistency observation points and reliability observation points, and setting an error adjustment rate delta of a converter according to requirements;
6) And (3) judging the calculation error between the macro scale and the set reference value through the consistency observation point and the reliability observation point, switching to the micro scale for calculation when the error of the macro scale is greater than the error adjustment rate delta set in the step 5), and correcting the parameter of the macro scale by using the obtained result to complete the data conversion between scales.
in the step 1), the macro scale refers to a power electronic technology for converting and controlling electric energy through power electronic components; the control strategy scale is the response of the macro-scale measurement signal and is used for generating a trigger signal of the micro-scale component layer; the micro-scale is divided into three layers, including a component layer, a material property layer and a multi-physical field layer.
In step 2), the multi-scale component model includes: the system comprises a multi-scale resistance model, a multi-scale inductance model, a multi-scale capacitance model and a multi-scale semiconductor switch component model; the multi-scale resistance model comprises a temperature interface, a stress interface, an electric field interface, a temperature interface, a stress interface, a magnetic field:
V＝V+I+Δ
+(V+I+I+Δ+Δ)
+(V+I+I+Δ+Δ)
+(V+I+Δ)
In the formula, V represents an electrical parameter, corner marks represent levels, I represents the influence of adjacent scales, and delta represents the coupling factor of the adjacent scales; v1_ cir is a circuit layer parameter, V2_ con is a micro-scale component layer parameter, V3_ mat is a material property layer parameter, and V4_ phy is a multi-physical field layer parameter; i2_ cir is an influence factor of the circuit layer on the micro-scale component layer, I1_ con and I3_ con are influence factors of the micro-scale component layer on the circuit layer and the material attribute layer respectively, I2_ mat and I4_ mat are influence factors of the multi-physical-field layer on the micro-scale component layer and the material attribute layer respectively, and I3_ phy is an influence factor of the multi-physical-field layer on the material attribute layer; Δ 1_ cir/con is a coupling item between the circuit layer and the micro-scale component layer, Δ 2_ con/mat is a coupling item between the component layer and the material property layer, Δ 2_ con/cir is a coupling item between the micro-scale component layer and the circuit layer, Δ 3_ mat/phy is a coupling item between the material property layer and the multi-physical field layer, Δ 3_ mat/con is a coupling item between the material property layer and the micro-scale component layer, and Δ 4_ phy/mat is a coupling item between the multi-physical field layer and the material property layer.
in step 3), one or more physical fields of an additive temperature field, an electric field, a magnetic field and a mechanical stress field are selected according to the application environment to be combined, and the internal parameters of the microscale component layer are changed through the change of material properties, so that the connection between the physical fields and the mechanism model layer is realized.
in step 5), the same power electronic converter contains a multi-scale component model and has multi-level zooming observation characteristics in the range from micro-scale to macro-scale, and accordingly consistency observation points and reliability observation points are set.
In the step 6), the multi-scale information transfer direction is progressive from a micro-scale to a macro-scale hierarchy, and the micro-scale is divided into three layers: the device layer, the material attribute layer and the multi-physical field layer are specifically explained as follows: the multi-physical field layer transmits physical field factors to the material attribute layer to cause material performance change, the material attribute layer acts the material characteristic change and constitutive relation on the micro-scale component layer to cause component electrical characteristic change, and the component layer transmits accurate component working characteristics to the macro-scale circuit layer through an operator to cause circuit electrical characteristic change; the different scales carry out data interaction at the interface of the adjacent scales.
In step 6), the data interaction and conversion between scales are realized by adopting a compression operator and a reconstruction operator, which are specifically embodied as follows: the micro-scale is averaged through a compression operator, information is transmitted to the macro-scale, the working state information of the macro-scale is used as a boundary condition, and the information is transmitted to the micro-scale through a continuation operator.
Furthermore, unidirectional connection or bidirectional coupling is adopted among different scales, and decoupling is realized only by interaction between the levels and the adjacent levels, which is specifically embodied as follows:
a. the circuit layer is connected with the component layer in a bidirectional mode, so that decoupling of the circuit layer and the multiple physical field layers is achieved;
b. the component layer is connected with the material attribute layer in a one-way mode, so that decoupling of the circuit layer and the material attribute layer is achieved;
c. the material property layer is connected with the multiple physical field layers in a one-way mode, and decoupling of the multiple physical field layers and the component layer is achieved.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. By using the multi-scale modeling method, multi-scale modeling can be performed by adopting one or a combination of several scales in the scale category according to the requirement of researching the attention level information.
2. only the meaning of a component model is considered to be very little, information such as the state, the aging, the reliability and the like of the component must be reflected through a physical field, research contents can be customized according to research needs, and a single field or a plurality of physical fields are added for research.
3. the scaling observation of the components can be carried out by using the micro scale according to the existing state of the circuit layer and the external environment, and the micro detail information which is difficult to reflect by a macro model can not be lost.
4. The macro-scale serial and micro-scale parallel computing method is combined with a data processing method of a compression operator and a continuation operator, so that the contradiction between the computing precision and the computing efficiency can be effectively relieved. Under the condition of the same calculation precision and calculation resource occupation, the method can replace the obviously improved calculation precision with less resource occupation, and has the characteristics of physical field interface and multi-scale observation.
5. The method is a universal modeling method for the power electronic converter, does not need to be reconstructed according to different converter types, can be applied to the existing power electronic converter, is developed from single scale to multi-scale steps, and is flexible and easy to use.
In conclusion, the method has advantages in the aspects of calculation efficiency, calculation accuracy, consistency, multiple observation characteristics and the like, is used for component type selection and determining system performance and safe working boundary, and has wide application prospect.
Drawings
FIG. 1 is a multi-scale hierarchical partitioning diagram of the method of the present invention.
FIG. 2 is a circuit diagram of an embodiment of the method of the present invention.
Fig. 3 is a flow chart of an embodiment of the method of the present invention.
fig. 4 is a comparison graph of the multi-scale IGBT and macro-scale IGBT switching process in the method of the present invention. In the graph, (a) is a comparison graph of gate voltage vge of the IGBT in two scales, (b) is a comparison graph of collector-emitter voltage vce of the IGBT in two scales, and (c) is a comparison graph of on-current ic of the IGBT in two scales. The two curves respectively represent the switching characteristics of the multi-scale IGBT and the macro-scale circuit layer IGBT.
Fig. 5 is a comparison graph of load currents io of the multi-scale buck circuit under the condition that multiple sets of error adjustment rates are set for scale conversion. Wherein, the error adjustment rates from near to far are respectively 1%, 10% and 20%, and scale conversion and data interaction are performed.
figure 6 is a multi-scale result of the emitter voltages at different error adjustment rates delta.
Detailed Description
to further illustrate the content and features of the present invention, the following detailed description of the embodiments of the present invention is provided with reference to the accompanying drawings, but the present invention is not limited thereto.
the invention takes a buck converter as an example, and specifically comprises the following components: inputting a direct-current voltage source, a multi-scale resistance model R1, a multi-scale inductance model L1, a multi-scale capacitance model C1, a multi-scale diode model and a multi-scale Insulated Gate Bipolar Transistor (IGBT) switch component model. Wherein the multi-scale resistance model R1 contains a temperature and stress interface; the multi-scale inductance model L1 contains temperature, stress and electric field interfaces; the multi-scale capacitance model C1 contains temperature, stress and magnetic field interfaces; the multi-scale diode model comprises a temperature field, an electric field, a magnetic field and a stress field interface; a multiscale Insulated Gate Bipolar Transistor (IGBT) switch component model comprises temperature, electric field, magnetic field and stress interfaces. The multi-scale component models jointly form a power electronic converter network.
The embodiment takes fig. 1 as a basis for scale division, and the structure applied to the topology is shown in fig. 2. For the sake of visual understanding, only the macro-scale circuit layers, the micro-scale device layers, and the micro-scale multi-physical field layers associated with this example are drawn in the drawings.
As shown in fig. 2, the multi-scale topological structure adopts a multi-scale resistance model, a multi-scale inductance model, a multi-scale capacitance model, a multi-scale diode model and a multi-scale Insulated Gate Bipolar Transistor (IGBT) switch component model, which are respectively described as follows:
1) multi-scale resistance model: and deducing a resistance constitutive relation expression containing a physical field interface by using the resistance definition formula R0 as L0/(rho 0. S) and the physical field influence factor thereof:
Where ρ 0 represents the electrical conductivity, determined by the material of the conductor and the ambient temperature, and is also the interface of the material properties; t is independent variable temperature; TC1 is a linear temperature coefficient, and TC2 is a quadratic temperature coefficient; tnom is the normal working environment temperature; l0 is the initial length of the resistor, Δ L is the change in length due to mechanical stress; and S is the cross-sectional area of the resistor.
the above equation can be written as:
The above process can separate the dimensions and physical factors: the factor L0/(ρ 0 · S) represents the ideal resistance, i.e. the circuit layer on a macroscopic scale; other items are microscale component layers: the factor [ TC 1- (T-Tnom) + TC 2- (T-Tnom)2 ]. L0/(rho 0. S) represents the resistance value variation under the influence of the temperature field; the factor [ Delta L/(rho 0 & S) ] represents the resistance value variation under the influence of the mechanical stress field; the factor [ TC 1- (T-Tnom) + TC 2- (T-Tnom)2 ]. delta L/(rho 0. S) represents the resistance variation under the mutual coupling action of the temperature field and the mechanical stress field.
2) Multi-scale inductance model: the constitutive relation expression of the inductance containing the physical field interface is derived by the inductance definition formula L0 ═ k μ 0 μ sN2S)/L and the physical field influence factor thereof:
In the formula, mu 0 is vacuum magnetic conductivity and is taken as 4 pi x 10-7; μ s is the relative permeability of the magnetic core inside the coil, determined by the material of the conductor and the ambient temperature, and is also the interface of the material properties; n2 is the square of the number of turns of the coil; s is the sectional area of the coil; l is the length of the coil, and delta l is the expansion variable of the coil; k is a coefficient, depending on the ratio of the radius to the length of the coil; i is the conduction current, IL1 is the linear current coefficient, IL2 is the secondary current coefficient; t is independent variable temperature; TC1 is a linear temperature coefficient, and TC2 is a quadratic temperature coefficient; tnom is the normal operating ambient temperature.
The above equation can be written as:
The above process can separate the dimensions and physical factors: the factor L0 represents the ideal inductance definition, i.e., the macro-scale circuit layer; other items are microscale component layers: the factor L0 [ (IL1 XI) + (IL2 XI 2) ] represents the amount of change in inductance under the influence of a magnetic field; the factor L0 (TC 1 (T-Tnom) + TC2 (T-Tnom) 2) represents the inductance transformation under the influence of the temperature field; the factor L0 [ (IL1 XI) + (IL2 XI 2) ] [ TC1 ] (T-Tnom) + TC2 (T-Tnom)2] represents the inductance value change under the simultaneous action of the temperature field and the magnetic field; the factor [ Δ L · L0/(1+ Δ L) ] represents the amount of change in inductance value under the influence of a mechanical stress field or magnetostriction; the factor Δ L · L0 · [ (IL1 × I) + (IL2 × I2) ]/(1+ Δ L) represents the amount of change in inductance value under the simultaneous action of a mechanical stress field and a magnetic field; the factor L0 (TC 1 (T-Tnom) + TC2 (T-Tnom) 2.delta L/(1+ delta L) represents the inductance value variation under the simultaneous action of the mechanical stress field and the temperature field; the factor L0 [ (IL1 XI) + (IL2 XI 2) ] [ TC1 ] (T-Tnom) + TC 2] (T-Tnom)2 ]. delta L/(1+ delta L) represents the inductance value change under the mutual coupling action of the mechanical stress field and the magnetic field and the temperature field.
3) Multi-scale capacitance model: and deriving a capacitance constitutive relation expression containing a physical field interface by using a capacitance definition formula C0 ═ epsilon S/4 pi kd and a physical field influence factor thereof:
Wherein epsilon is the dielectric constant of the medium between the polar plates, is determined by the material of the conductor and the ambient temperature, and is also the interface of the material property; s is the positive area of the capacitor plate; π is a constant circumference ratio; k represents an electrostatic force constant; d is the distance of the capacitor plate, and delta d is the change of the distance between the capacitor plates caused by mechanical stress; v is the voltage at two ends of the capacitor, VC1 is a linear voltage coefficient, and VC2 is a secondary voltage coefficient; t is independent variable temperature; TC1 is a linear temperature coefficient, and TC2 is a quadratic temperature coefficient; tnom is the normal operating ambient temperature.
The above equation can be written as:
The above process can separate the dimensions and physical factors: the factor C0 represents the ideal capacitance definition, i.e., the circuit layer; other items are microscale component layers: the factor C0[ (VC1 XV) + (VC2 XV 2) ] represents the variation of capacitance under the influence of electric field; the factor C0 (TC 1 (T-Tnom) + TC2 (T-Tnom) 2) represents the capacitance value transformation under the influence of the temperature field; the factor C0[ (VC1 XV) + (VC2 XV 2) ] [ TC1 ] (T-Tnom) + TC2 (T-Tnom)2] represents the capacitance variation under the simultaneous action of the temperature field and the electric field; the factor C0 Δ d/(1+ Δ d) represents the amount of change in capacitance under the influence of the mechanical stress field; the factor C0[ (VC1 XV) + (VC2 XV 2) ]. delta d/(1+ delta d) represents the capacitance variation under the simultaneous action of the mechanical stress field and the electric field; the factor C0 (TC 1 (T-Tnom) + TC2 (T-Tnom) 2. delta d/(1+ delta d) represents the capacitance variation under the simultaneous action of the mechanical stress field and the temperature field; the factor C0[ (VC1 XV) + (VC2 XV 2) ] [ TC1 ] (T-Tnom) + TC 2] (T-Tnom)2 ]. delta d/(1+ delta d) represents the capacitance variation under the mutual coupling action of the mechanical stress field and the electric field and the temperature field.
4) multi-scale diode model: using a mechanistic model, the current expression describing the model is shown below:
i＝i+i+i+i (7)
idhl, idrec, idb, ic are the components that make up the diode current, respectively, as follows:
Wherein the content of the first and second substances,
in the formula, Rs is a series resistor; vd is the diode terminal voltage; VJ is junction potential; VT is the temperature dependent interface, at constant temperature: kT ═ kBTnom/q, taking into account the temperature field: kT ═ kB (Tnom + Δ T)/q; q is an electronic charge; vd is the external terminal voltage of the diode; n is an ideal factor; is saturation current; IKF is a large current roll-off angle; NBV is an ideal factor for failure; BV is reverse breakdown voltage; IBV is the current at breakdown voltage; CJ0 is zero offset junction capacitance; FC is the forward bias capacitance coefficient; m is a grading coefficient.
In the multi-scale diode model, the capacitor CJ0 contains interfaces of temperature fields, electric fields and mechanical stress fields, and the multi-scale capacitor modeling method is adopted.
5) Multiscale Insulated Gate Bipolar Transistor (IGBT) switch component model:
Distribution of remaining holes p on the x-axis:
Wherein x is the moving distance of the cavity on the x axis; t is a time variable; p0 is the initial concentration of holes; d2 DnDp/(Dn + Dp) is bipolar diffusion coefficient; dn and Dp are electrons and hole diffusion coefficients; l ═ D τ HL is the bipolar diffusion length; w is the quasi-neutral base width.
the BJT base current Ib is equal to the electron current of the quasi-neutral base region, i.e.:
In the formula, A is an equipment activity area; b is the base width; d2 DnDp/(Dn + Dp) is bipolar diffusion coefficient; dn and Dp are electrons and hole diffusion coefficients; in + Ip, In, Ip is electron, hole current; n, p are electron and hole carrier concentrations; q is an electronic charge; q is the total excess charge carrier group charge; w is the quasi-neutral base width.
when the IGBT is turned off, the residual carrier charge Q is mainly the consumption of itself and the difference between the electron current flowing out and in:
The electron current input by the BJT emitting end is as follows:
wherein P0 is the initial concentration of holes; a is an equipment active area; in is electron current; isne is the emitter electron saturation current; ni is the electron concentration; q is an electronic charge; w is the quasi-neutral base width; τ HL is the high level supercarrier lifetime.
When the IGBT turns on, the on-current Imos of the MOSFET section is related to the drive voltage Vgs:
In the formula, Kp is the transconductance of the MOSFET channel; vgs is the driving voltage; vds is the positive drain source voltage; VT is the threshold voltage.
since the voltage drop of the BJT mainly lies in the voltage drop ubc between the base and the collector of the BJT, i.e. the voltage across the parasitic capacitor cbbj in the IGBT, and in the switching transient state, the voltage across the emitter and the base of the BJT is constant, i.e.: since the duty/dt is dubc/dt, it can be found that:
In the formula, A is an equipment activity area; cbcj is the parasitic capacitance of the IGBT; d2 DnDp/(Dn + Dp) is bipolar diffusion coefficient; dp is electron, hole diffusion coefficient; ib is base current; imos is the MOSFET on current; iT is IGBT conduction current; NB is base region doping concentration; b is the base width; q is the residual electron charge; q is the amount of residual carrier charge; w is the quasi-neutral base width.
uce is temperature dependent in a multiscale Insulated Gate Bipolar Transistor (IGBT) switch component model; q is related to the temperature field and implies the information of the aging of the components; the capacitor Cbcj comprises a temperature field, an electric field degree and a stress field interface; IT is IB + IC, which includes interfaces related to aging, and reflects the limit value of the on-current of the device, as shown in the following formula.
In the formula, A is an equipment activity area; b is the base width; d2 DnDp/(Dn + Dp) is bipolar diffusion coefficient; dn and Dp are electrons and hole diffusion coefficients; p0 is the initial concentration of holes; isne is the emitter electron saturation current; l ═ D τ HL is the bipolar diffusion length; q is an electronic charge; ni is the electron concentration; w is the quasi-neutral base width.
In order to avoid the influence of the circuit layer parameters on the IGBT, the gate voltage, the collector-emitter voltage and the on-current waveform of the multi-scale IGBT and the macro-scale circuit layer IGBT in a single switching period are compared by a method of driving a resistive load, as shown in (a), (b) and (c) of fig. 4.
fig. 4 (a) shows the VGE triggering process of the IGBT in the circuit layer, which is an ideal switching process that can be instantly completed without delay. After passing through the miller platform, VGE of the component layer IGBT gradually rises to the driving voltage von. The maximum error rate of VGE can reach 150%. For the turn-off process, the IGBT experiences a turn-off time toff ═ td (off) + Δ t + tf ═ 1.11 μ s, including a turn-off delay td (off), a fall time tf, and a time Δ t from 30% of the VCC voltage to the start of the current drop.
Similar to the ice current spike in fig. 4 (b), the maximum error rate between the two scales can reach 50.5% at a supply voltage of 300V to ground.
For VCE in fig. 4 (c), the conductance GOFF of the circuit layer IGBT can be adjusted to a voltage close to that of the component layer with a maximum error rate of 19.5%.
It can be seen that the circuit layer IGBT uses a linear polygonal line model to approximate the voltage and current characteristics of the switching element, and has the best calculation efficiency but the poor accuracy and the approximate dynamic characteristics.
Compared with the circuit layer IGBT, the component layer IGBT is consistent with the real situation, the obvious transient difference of the drop voltage of the switch is considered, accurate information at the current peak value is observed, and the circuit layer IGBT is suitable for simulating detailed switch characteristics and finely adjusting the current-voltage characteristics. Therefore, the IGBT of the component layer has a strong guiding function on parameter selection and heat loss prediction, and expensive cost caused by overdesign is avoided.
As shown in fig. 2, the multi-scale model operating state of the buck converter is simulated by using the data exchange method shown in fig. 3 in consideration of the influence of the temperature and the bidirectional coupling relationship between the circuit layer and the component layer, and the simulation result is shown in fig. 5 and 6.
The conversion condition of the component layer and the updating time of the circuit layer are determined by the intersection of two switching strategies:
1) The component layer runs once every other settable fixed time;
2) When the error adjustment rate delta exceeds a set value.
In the aspect of buck circuit load, each component updates from the component layer to the circuit layer according to the switching strategy. In order to achieve uniform observation effect, simulation results obtained by different error adjustment rates are obtained by the circuit layer, and as shown in fig. 5, the error adjustment rates δ of the component layer update circuit layer are set to δ 1%, 10%, and 20% respectively from near to far. Different error adjustment rates have different observations, and the update node is marked with a vertical dashed line.
The concrete expression is as follows: when the error adjustment rate δ of the load is 1%, the frequency of updating the circuit scale at the component layer is observed to be 28 times, and is hardly seen at a significant difference between the update points. In contrast, when the error adjustment rates are 10% and 20%, respectively, the significant differences of the update points are 9 times and 5 times, respectively.
From the results of the load current, it can be seen that the lower the error adjustment rate, the more consistent the frequency of circuit layer updates and temperature changes to achieve an immediate response, especially when the error adjustment rate δ is 1%. The higher the operating temperature, the smaller the current conducting capacity of the components and circuits. Different temperatures also have an effect on the current: the more accurate the transmission of the temperature influence factors is, the more practical the transmission is, so that the micro-scale is connected with the multiple physical field layers. When the error adjustment rate δ is higher, the lower the circuit layer update frequency is, the smaller the temperature influence is, and the closer the analysis result is to the ideal circuit.
fig. 6 shows the multiscale result of collector-emitter voltages at different error adjustment ratios δ, where the error adjustment ratios δ from the device layer to the circuit layer are set to δ being 1%, 10%, 20%, and 50% respectively from near to far. Different error adjustment rates have different update frequencies.
the concrete expression is as follows: as can be seen from the four enlarged views on the right side of the figure, when the load setting error adjustment rate δ is 1%, the frequency of the component layer refresh circuit layer is 9 times (the obvious refresh point is indicated by the vertical broken line in the figure). When the error adjustment rates are 10%, 20%, and 50%, respectively, the update frequencies are 5 times, 4 times, and 2 times, respectively.
the measurement result of the collector-emitter voltage shows that the lower the error adjustment rate is, the higher the update frequency is, the more accurate the transmission of the temperature effect is, and the more practical the transmission is. On the contrary, the higher the error adjustment rate, the lower the circuit layer update frequency, the smaller the temperature influence, and the more consistent the analysis result with the ideal circuit. The transient process of the components only affects the micro scale, the influence on the overall external characteristics of the converter is not large, and the steady-state waveforms of the two models are the same. The multi-scale model contains more information and higher accuracy than the circuit layer model.
in summary, the multi-scale modeling method of the converter based on the micro and macro descriptions provided by the invention verifies the scaling characteristics of the multi-scale in the modeling process, and compared with the traditional single-scale converter modeling method, the multi-scale model contains more information than the circuit layer model: voltage spike, accurate value of withstand voltage, physical field influence, current oscillation, overshoot, accurate current capacity of components and the like. Meanwhile, the transient process of the micro-scale component layer only affects the micro-scale, the influence on the overall external characteristics of the converter is small, and the result of the macro-scale and the result of the micro-scale can be kept consistent in a stable state. Moreover, under the condition of the same calculation precision and calculation resource occupation, the method can replace the significantly improved calculation precision with less resource occupation, and has the characteristics of physical field interface, multi-scale observation and the like. The modeling method is a universal modeling method for the power electronic converter, does not need to be reconstructed according to different converter types, can be applied to the existing converter, is developed from a single scale to multi-scale steps, and is flexible and easy to use. Therefore, the circuit has very wide application prospect.
the above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be regarded as equivalent replacements within the protection scope of the present invention.
Claims (8)
1. a converter multi-scale modeling method based on micro and macro description is characterized in that: the method is suitable for the power electronic converter, a component model library with multi-scale and multi-physical field interfaces is derived according to a component model, and the component model library is used for reflecting microscopic information of the power electronic converter under the action of physical field factors, and the method comprises the following steps:
1) Selecting a topological structure and an auxiliary circuit of the converter, analyzing the scale of the converter to be observed, and dividing the scale category into a macro scale, a control strategy scale and a micro scale;
2) Determining the level of research to be developed, and selecting a multi-scale component model from a component library according to the type;
3) selecting a physical field according to the operation environment of the converter, wherein a single physical field is singly added for environmental factors, and a plurality of physical fields are added for coupling analysis in a complex observation environment;
4) determining a research mode, specifically dividing the research mode into a long-time steady state and a short-time transient state, and using the long-time steady state and the short-time transient state as a reference simulation environment with a micro scale and a macro scale which are independent from each other;
5) Selecting components and nodes for multi-scale observation, respectively setting consistency observation points and reliability observation points, and setting an error adjustment rate delta of a converter according to requirements;
6) And judging the calculation error between the macro scale and the set reference value through the consistency observation point and the reliability observation point, switching the micro scale for calculation when the error of the macro scale is larger than the error adjustment rate delta set in the step 5), and correcting the parameter of the macro scale by using the obtained result to finish the data conversion between scales.
2. The method of claim 1, wherein the method comprises the following steps: in the step 1), the macro scale refers to a power electronic technology for converting and controlling electric energy through power electronic components; the control strategy scale is the response of a circuit layer measurement signal and is used for generating a trigger signal of a microscale component layer; the micro-scale is divided into three layers, including a component layer, a material property layer and a multi-physical field layer.
3. The method of claim 1, wherein the method comprises the following steps: in step 2), the multi-scale component model includes: the system comprises a multi-scale resistance model, a multi-scale inductance model, a multi-scale capacitance model and a multi-scale semiconductor switch component model; the multi-scale resistance model comprises a temperature interface, a stress interface, an electric field interface, a temperature interface, a stress interface, a magnetic field:
V＝V+I+Δ
+(V+I+I+Δ+Δ)
+(V+I+I+Δ+Δ)
+(V+I+Δ)
In the formula, V represents an electrical parameter, corner marks represent levels, I represents the influence of adjacent scales, and delta represents the coupling factor of the adjacent scales; v1_ cir is a circuit layer parameter, V2_ con is a component layer parameter, V3_ mat is a material property layer parameter, and V4_ phy is a multi-physical field layer parameter; i2_ cir is an influence factor of a circuit layer on a component layer, I1_ con and I3_ con are influence factors of the component layer on the circuit layer and a material attribute layer respectively, I2_ mat and I4_ mat are influence factors of a multi-physical field layer on the component layer and the material attribute layer respectively, and I3_ phy is an influence factor of the multi-physical field layer on the material attribute layer; Δ 1_ cir/con is a coupling item between a circuit layer and a component layer, Δ 2_ con/mat is a coupling item between the component layer and a material attribute layer, Δ 2_ con/cir is a coupling item between the component layer and the circuit layer, Δ 3_ mat/phy is a coupling item between the material attribute layer and a multi-physical field layer, Δ 3_ mat/con is a coupling item between the material attribute layer and the component layer, and Δ 4_ phy/mat is a coupling item between the multi-physical field layer and the material attribute layer.
4. the method of claim 1, wherein the method comprises the following steps: in step 3), one or more physical fields of an additive temperature field, an electric field, a magnetic field and a mechanical stress field are selected according to the application environment to be combined, and the internal parameters of the component are changed through the change of material properties, so that the connection between the physical field and the mechanism model layer is realized.
5. The method of claim 1, wherein the method comprises the following steps: in step 5), the same power electronic converter contains a multi-scale component model and has multi-level zooming observation characteristics in the range from micro-scale to macro-scale, and accordingly consistency observation points and reliability observation points are set.
6. the method of claim 1, wherein the method comprises the following steps: in the step 6), the multi-scale information transfer direction is progressive from a micro-scale to a macro-scale hierarchy, and the micro-scale is divided into three layers: the device layer, the material attribute layer and the multi-physical field layer are specifically explained as follows: the multi-physical field layer transmits physical field factors to the material property layer to cause material performance change, the material property layer acts the material characteristic change and the constitutive relation on the micro-scale component layer to cause component electrical characteristic change, and the component layer transmits the accurate component working characteristics to the macro-scale circuit layer through an operator to cause circuit electrical characteristic change; the different scales carry out data interaction at the interface of the adjacent scales.
7. The method of claim 1 or 6, wherein the method comprises the following steps: in step 6), the data interaction and conversion between scales are realized by adopting a compression operator and a reconstruction operator, which are specifically embodied as follows: the micro-scale is averaged through a compression operator, information is transmitted to the macro-scale, the working state information of the macro-scale is used as a boundary condition, and the information is transmitted to the micro-scale through a continuation operator.
8. The method of claim 6, wherein the method comprises the following steps: the different scales are in one-way connection or two-way coupling, and the levels only interact with the adjacent levels to realize decoupling, which is embodied as follows:
a. The circuit layer is connected with the component layer in a bidirectional mode, so that decoupling of the circuit layer and the multiple physical field layers is achieved;
b. the component layer is connected with the material attribute layer in a one-way mode, so that decoupling of the circuit layer and the material attribute layer is achieved;
c. the material property layer is connected with the multiple physical field layers in a one-way mode, and decoupling of the multiple physical field layers and the component layer is achieved.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910664814.8A CN110543665B (en) | 2019-07-23 | 2019-07-23 | Converter multi-scale modeling method based on micro and macro description |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910664814.8A CN110543665B (en) | 2019-07-23 | 2019-07-23 | Converter multi-scale modeling method based on micro and macro description |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110543665A true CN110543665A (en) | 2019-12-06 |
| CN110543665B CN110543665B (en) | 2021-03-30 |
Family
ID=68710285
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910664814.8A Active CN110543665B (en) | 2019-07-23 | 2019-07-23 | Converter multi-scale modeling method based on micro and macro description |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110543665B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111597735A (en) * | 2020-06-19 | 2020-08-28 | 华南理工大学 | Component prediction method combining machine learning and CVD modeling |
| CN113009226A (en) * | 2021-02-03 | 2021-06-22 | 长江存储科技有限责任公司 | Method and device for obtaining contact resistance |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060025989A1 (en) * | 2004-07-28 | 2006-02-02 | Nima Mesgarani | Discrimination of components of audio signals based on multiscale spectro-temporal modulations |
| CN202427452U (en) * | 2011-09-02 | 2012-09-12 | 中国第一汽车股份有限公司 | Patch output converter (POC) postprocessor metal carrier micro trapezoidal structure |
| CN102945548A (en) * | 2012-11-20 | 2013-02-27 | 成都晶石石油科技有限公司 | Directional pyramid filtering-based image processing method and device |
| CN109310933A (en) * | 2016-04-24 | 2019-02-05 | 弗洛设计声能学公司 | For generating electron configuration and the control of sound standing wave |
| CN109661684A (en) * | 2016-09-06 | 2019-04-19 | 苏黎世联合高等工业学校 | Ray-tracing procedure for real interactive ultrasonic simulation |
-
2019
- 2019-07-23 CN CN201910664814.8A patent/CN110543665B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060025989A1 (en) * | 2004-07-28 | 2006-02-02 | Nima Mesgarani | Discrimination of components of audio signals based on multiscale spectro-temporal modulations |
| CN202427452U (en) * | 2011-09-02 | 2012-09-12 | 中国第一汽车股份有限公司 | Patch output converter (POC) postprocessor metal carrier micro trapezoidal structure |
| CN102945548A (en) * | 2012-11-20 | 2013-02-27 | 成都晶石石油科技有限公司 | Directional pyramid filtering-based image processing method and device |
| CN109310933A (en) * | 2016-04-24 | 2019-02-05 | 弗洛设计声能学公司 | For generating electron configuration and the control of sound standing wave |
| CN109661684A (en) * | 2016-09-06 | 2019-04-19 | 苏黎世联合高等工业学校 | Ray-tracing procedure for real interactive ultrasonic simulation |
Non-Patent Citations (2)
| Title |
|---|
| HUAI WANG等: "Transitioning to Physics-of-Failure as a Reliability Driver in Power Electronics", 《IEEE JOURNAL OF EMERGING AND SELECTED TOPICS IN POWER ELECTRONICS》 * |
| 刘更 等: "宏观 —微观多尺度数值计算方法研究进展", 《中国机械工程》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111597735A (en) * | 2020-06-19 | 2020-08-28 | 华南理工大学 | Component prediction method combining machine learning and CVD modeling |
| CN111597735B (en) * | 2020-06-19 | 2022-06-14 | 华南理工大学 | Component prediction method combining machine learning and CVD modeling |
| CN113009226A (en) * | 2021-02-03 | 2021-06-22 | 长江存储科技有限责任公司 | Method and device for obtaining contact resistance |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110543665B (en) | 2021-03-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Analysis and design considerations of integrated 3-level buck converters | |
| Hefner et al. | An experimentally verified IGBT model implemented in the Saber circuit simulator | |
| Locorotondo et al. | Analytical model of power MOSFET switching losses due to parasitic components | |
| Li et al. | SiC/GaN power semiconductor devices: a theoretical comparison and experimental evaluation under different switching conditions | |
| Drofenik et al. | Voltage-dependent capacitors in power electronic multi-domain simulations | |
| CN110543665B (en) | Converter multi-scale modeling method based on micro and macro description | |
| CN103354417B (en) | Constant-current control circuit with peak current compensation and Switching Power Supply | |
| Xiao et al. | A universal power flow model for dual active bridge-based converters with phase shift modulation | |
| Liu et al. | Multiscale modeling and analysis of DC/DC converter based on macro and micro-scale description | |
| Bazzano et al. | A new analog behavioral SPICE macro model with thermal and self-heating effects for silicon carbide power MOSFETs | |
| Simões | Power bipolar transistors | |
| Igic et al. | Dynamic electro-thermal physically based compact models of the power devices for device and circuit simulations | |
| Raffa et al. | A new analog behavioral SPICE macro model with self-heating effects and 3rd quadrant behavior for Silicon Carbide Power MOSFETs | |
| Zenk | Comparison of Electrical Performances of Power Electronics Switches and an Effective Switch Selection Algorithm | |
| Hao et al. | Calculation and analysis of switching losses in IGBT devices based on switching transient processes | |
| Ren et al. | Bridging gaps in paper design considering impacts of switching speed and power-loop layout | |
| Górecki et al. | Influence of MOSFET model form on characteristics of the boost converter | |
| Santi et al. | Variable model levels for power semiconductor devices | |
| Zappulla et al. | Optimization of bidirectional modular dc/dc converter for low and high power operation in aircraft applications | |
| Pekdemir et al. | Stress Analysis Based on Semiconductor Equivalent Circuit Model in Flyback Converters | |
| Khanna et al. | A linear model for characterizing transient behaviour in wide bandgap semiconductor-based switching circuits | |
| Hammer et al. | Modeling the effects of printed-circuit-board parasitics on the switching performance of wide-bandgap applications | |
| Rengarajan | Characterization and Analysis of 10 kV Generation-3 Silicon Carbide Power MOSFET Model | |
| Chen | Power losses of silicon carbide MOSFET in HVDC application | |
| Inamdar et al. | Metastability of pulse power loads with nonlinear coupled magnetics |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |