Hybrid modeling joint simulation method for modular multilevel converter

Abstract

The invention provides a hybrid modeling joint simulation method of a modular multilevel converter, which fully considers a complex topological structure in an IGBT device-level model in an MMC sub-module, and can accurately simulate the IGBT device-level model under the condition of correctly reflecting IGBT peak voltage and current, Miller platform, tail current and reverse recovery current. A Thevenin equivalent model and an average value model which are commonly used in MMC modeling simulation and a self-defined model meeting device-level nanosecond-level ultra-small step fine simulation modeling requirements are subjected to hybrid modeling joint simulation. The method is specifically implemented as follows: IGBT (Insulated Gate Bipolar Translator) in corresponding sub-modules of an upper bridge arm and a lower bridge arm of the MMC adopt the same device-level modeling simulation method, and a hybrid modeling three-phase six-bridge-arm system is built. According to the method provided by the invention, the internal advantages of different model modeling methods are exerted, the precision, efficiency and flexibility of the MMC joint simulation model are improved, and a new thought is provided for MMC simulation modeling in the field of flexible direct current power transmission.

Description

Technical Field

[0001] The present invention relates to the field of power electronics technology simulation, and particularly to a hybrid modeling and co-simulation method for a modular multilevel converter. Background Art

[0002] Flexible DC power transmission is a key technology for realizing large-scale development and utilization of renewable energy and ensuring reliable power supply. However, it is difficult to simulate the internal characteristics of flexible DC equipment, and the operating mechanism of devices is complex. It is necessary to separately conduct modeling research at the equipment level and the device level. Existing electromagnetic transient equivalent models of MMC mainly include three categories: microsecond-level simulation models, nanosecond-level simulation models, and millisecond-level simulation models. Among them, most nanosecond-level simulation models are used for performance and loss analysis of devices; the representative of millisecond-level simulation models is the electromechanical transient model, which is mainly used for stability analysis of large-scale AC-DC hybrid systems. These two are not commonly used. The models used in mainstream electromagnetic transient simulation software are mostly microsecond-level simulation models. Several common microsecond-level models include the general model of MMC based on controlled sources, the Thevenin model, and the average value model. It is challenging to conduct refined simulation modeling with nanosecond-level ultra-small step sizes for the internal dynamic characteristics of power electronic devices in flexible DC power transmission.

[0003] Currently, in the modeling research of power electronic devices, mainly two categories of mechanism models and behavior models are adopted. When conducting refined simulation modeling at the device level, it is necessary to consider the detailed processes of the on-off voltage and current of switching devices (spikes, Miller platforms), study the electromagnetic transient modeling method at the device level (hundred-nanosecond level), study the composition method of the IGBT transient model, and study and establish a semi-behavioral and semi-mechanistic model of the IGBT device for the physical structure of the IGBT device and the semiconductor element behavior characteristic equation; analyze the change characteristics of parasitic capacitance in the IGBT dynamic process; at the same time, it is also necessary to consider the detailed processes of the on-off voltage and current of switching devices, and propose a parameter extraction method for IGBTs; considering the reverse recovery phenomenon of diodes, establish an equivalent transient model of the freewheeling diode.

[0004] At the same time, in order to accurately describe the operating process of the devices inside the sub-module and give full play to the advantages of the existing microsecond-level models in terms of both accuracy and efficiency in large-scale power grids, after improving each modeling method, a multi-time-scale MMC three-phase six-arm system is hybridly and jointly built, which broadens the thinking for the subsequent development of the flexible DC power transmission field. Summary of the Invention

[0005] In order to meet the needs of the existing technology and address the deficiencies of the microsecond-level modeling method described in the background art, the present invention proposes a hybrid modeling and co-simulation method for a modular multilevel converter.

[0006] A hybrid modeling and co-simulation method for a modular multilevel converter includes the following steps:

[0007] Step 1: Establish the IGBT Thevenin equivalent model;

[0008] Step 2: Establish the IGBT average value model;

[0009] Step 3: Establish a custom component model that conforms to the IGBT switching transient characteristics;

[0010] Step 4: According to the IGBT transient models obtained in Step 1, Step 2, and Step 3, connect the three models according to the MMC sub-module IGBT structure, modify and add corresponding parameters and modules, so as to establish a multi-time-scale MMC sub-module IGBT hybrid transient model.

[0011] In Step 1, using the Thevenin equivalent principle, regard the IGBT switch group in each sub-module as a variable resistor that can switch between high and low resistance values. When the switch group is on, the resistance value is very small (R on ), when the switch group is off, the resistance value is very large (R off ), and its resistance value switching is determined by the switch state, and the switch state is determined by the MMC controller. Specifically, the modeling of the above equivalent module is as follows:

[0012] Perform Thevenin equivalent on the IGBT switch group circuit structure and discretize the capacitor in the sub-module. The Thevenin equivalent resistance R C and the equivalent voltage U CEQ overall equivalent sub-module capacitor C, their values are all functions of time, as shown in equations (1) and (2). The capacitor current I C (t) is calculated by equation (3), and it is used to update the sub-module capacitor voltage in each simulation step. The superscript T in the formula indicates that the trapezoidal integration method is used in the construction of the adjoint circuit.

[0013]

[0014] The discretized circuit model is overall equivalent to the sub-module Thevenin equivalent circuit, and the equivalent parameters R SMEQ and U SMEQ are obtained from equations (4) and (5), and R SMEQ is also a time-varying quantity.

[0015]

[0016] In Step 2, the IGBT switching actions inside the sub-module are simplified, and the internal electrical connection relationships are ignored. A mathematical model of the device ports is established based on the averaging concept, that is, within a time period much longer than a single switching cycle, the switching changes within a single cycle are ignored, and the change in state is equivalent to the average value of the states in consecutive switching cycles. The only equivalent DC-side capacitor in the average value model of the MMC (used to equivalent the capacitors of the three-phase six-arm in the detailed model of the MMC) C AVM It can be calculated from (6).

[0017]

[0018] In the formula, C DM is the capacitance value of a single sub-module in the detailed model of the MMC.

[0019] In Step 3, a custom IGBT switching transient simulation model combines the mechanism model and the behavior model. For the part involving the underlying parameters and physical mechanisms of the device, methods such as mechanism derivation, electrical equivalence, and curve fitting are used to avoid solving complex semiconductor physics equations; in order to reflect the output characteristics of the IGBT, a MOSFET-BJT module and a MILLER capacitance module based on the device manual parameters are built; at the same time, the established model starts from the steady-state and transient characteristics of the IGBT and the diode, fully considers the reverse recovery characteristics of the diode and its influence on the IGBT, and studies the turn-on and turn-off processes of the IGBT module in detail in stages. The specific modeling method is as follows:

[0020] (1) MOSFET-BJT equivalent module:

[0021] The external output characteristics of the IGBT are mainly determined by the MOSFET and the BJT. When the IGBT is conducting, two current paths are generated inside it: 1) The current path In generated by the internal electron flow, corresponding to the MOSFET structure. 2) The current path Ip generated by the hole flow, corresponding to the BJT structure.

[0022] When the IGBT is working in the cut-off, saturation, and amplification states respectively, the current expressions flowing through the MOSFET are as follows:

[0023]

[0024] Due to the complex extraction of the MOSFET transconductance K p and the BJT current gain β parameters, electrical equivalent simplification can be carried out based on the requirements of circuit simulation to obtain the following relationship:

[0025]

[0026] Therefore, the MOSFET-BJT equivalent module can use a voltage-controlled current source to simulate the on-state current Ic of the IGBT module, and its analytical expression is as follows:

[0027]

[0028] Wherein, the equivalent transconductance K = (1 + β)K p ; V ge is the gate-emitter voltage; V T is the IGBT conduction threshold voltage; V ce is the IGBT collector-emitter voltage; K p is the MOSFET transconductance; β is the BJT current gain; I mos is the current flowing through the MOSFET; I c is the current flowing through the IGBT, i.e., the collector current; among which, the equivalent transconductance K and the threshold voltage V T parameters are extracted from the device manual.

[0029] (2) Tail current equivalent module:

[0030] During the IGBT turn-off transient process, due to the existence of BJT in the IGBT, it takes time for a large number of excess carriers in the base region to recombine, resulting in a relatively long tail time for the turn-off current.

[0031]

[0032] Where τ is the minority carrier lifetime, i.e., the tail time constant, which is jointly determined by the steady-state characteristics and the tail current during the turn-off process; t0 is the starting time of the tail current; during the turn-off process, when V ge is less than the threshold voltage, the tailing starts, and at this time, the collector current is the starting current I tail0 of the tail current. Adding the said formula (10) to the said formula (9), the complete MOSFET-BJT equivalent module can be obtained.

[0033] (3) Miller capacitance module:

[0034] There are two types of parasitic capacitances between the IGBT poles. One is related to the MOS structure, and the other is related to the PN junction. In the data manual, the input capacitance C ies , the output capacitance C oes and the feedback capacitance C res are commonly used parameters in applications. Their relationships with the inter-pole capacitances are as follows:

[0035]

[0036] Using the said formula (11) combined with the data in the device manual, the corresponding inter-pole parasitic capacitance values can be obtained.

[0037] (4) Diode reverse recovery current module:

[0038] The transition of a diode from the forward conduction state to the reverse blocking state is called reverse recovery, and the resulting reverse recovery current will affect the switching characteristics of the IGBT. Based on the reverse recovery characteristics of the diode and its influence on the IGBT, an equivalent diode reverse recovery model is established based on the macro-model design idea. This model consists of an ideal diode, an RL parallel circuit, and a voltage-controlled current source. The relevant parameters of the model are shown in Equation (12).

[0039]

[0040] Among them, τ re is the reverse recovery decay time constant; R and L are free variables, and since the calculated values of the device parameters are very small, they are often ignored in the modeling process; I rm is the reverse recovery peak current; dI f / dt is the reverse recovery current slope; t rr is the reverse recovery time; Q rr is the reverse recovery charge amount.

[0041] According to the above formulas of each part, respective defined modules are constructed, connected according to the IGBT module circuit structure, a complete IGBT custom component model is built, and after encapsulation, three electrodes G (gate), C (collector), and E (emitter) are led out and connected to the main circuit.

[0042] In Step 4, the equivalent simulation models in Step 1, Step 2, and Step 3 are subjected to hybrid co-simulation to realize the equivalent modeling of each sub-module and the connection of the corresponding circuit structure, and finally a three-phase six-arm MMC system is built.

[0043] Compared with the closest prior art, the excellent effects of the present invention are as follows:

[0044] 1. The present invention performs nested processing and improvement on the existing microsecond-level model. Specifically, it conducts device-level nanosecond-level ultra-small step-size refined simulation modeling on the internal dynamic characteristics of the power electronic devices required for flexible DC transmission, and the established model meets the needs of current technology.

[0045] 2. The present invention can realize waveform simulation of each working state of the IGBT in circuit simulation, and can simulate the IGBT switching transient characteristics such as IGBT module voltage and current spikes, tail current, Miller plateau, and diode reverse recovery current.

[0046] 3. The present invention can not only simulate the IGBT inside the MMC sub-module, but also meet the needs of different time scales on the three-phase six-arm system, and switch between the microsecond level and the nanosecond level to achieve the development needs of modern engineering technology. Description of the Drawings

[0047] The present invention will be further described below with reference to the accompanying drawings.

[0048] Figure 1 The MMC three-phase six-arm circuit diagram built for the hybrid modeling co-simulation method of a modular multilevel converter provided by the present invention;

[0049] Figure 2 The circuit diagram of the Thevenin equivalent model implemented on the PSCAD / EMTDC platform;

[0050] Figure 3 The circuit diagram of the average value model implemented on the PSCAD / EMTDC platform;

[0051] Figure 4 The circuit diagrams of the (a) MOSFET-BJT module, (b) Miller capacitance module, and (c) diode reverse recovery module in the custom IGBT component implemented on the PSCAD / EMTDC platform;

[0052] Figure 5 The circuit diagram of the custom IGBT component after packaging implemented on the PSCAD / EMTDC platform. Specific embodiments

[0053] The present invention provides a hybrid modeling co-simulation method for a modular multilevel converter.

[0054] Figure 1 It is the MMC three-phase six-arm circuit diagram built for the hybrid modeling co-simulation method of a modular multilevel converter provided by the present invention. Figure 1 In it, the MMC three-phase six-arm system is composed of a hybrid co-simulation sub-module of a microsecond-level model and a nanosecond-level model. The microsecond-level model is composed of an MMC Thevenin equivalent model (such as Figure 2 ) and an average value model (such as Figure 3 ); the nanosecond-level model is composed of a custom IGBT model (such as Figure 4 ).

[0055] Figure 4 (a) is the MOSFET-BJT module, where the module inputs are the turn-on threshold voltage V T , the gate-emitter voltage V ge , the collector-emitter voltage V ce , the collector current I c , the simulation time t, and related control parameters, etc., and the output is the current value I mos1 of the MOSFET-BJT equivalent voltage-controlled current source. Its custom component is built according to the content described in step 3 to simulate the characteristics of the MOSFET and BJT.

[0056] Figure 4(b) is the Miller capacitance module, where the module inputs are the collector-emitter voltage V ce , the simulation time t, and the module output is the input capacitance C ies , the output capacitance C oes , and the feedback capacitance C res . It is implemented by custom programming based on the capacitance characteristic curve in the device manual. According to the description in step 3, the capacitance values C ge , C gc , and C ce are changed under different conditions, and the capacitance value is changed by judging the condition of the magnitude of the voltage V ge to simulate the IGBT entering the Miller plateau.

[0057] Figure 4 (c) is the diode reverse recovery current module, where the module inputs are the diode current I d , the simulation time t, the peak value of the reverse recovery current I rm , the reverse recovery current slope dif and other reverse recovery parameters, and the output is the current value I f of the diode reverse recovery equivalent current source. The collector current I c is input into the diode module to participate in the calculation and judgment of the reverse recovery current to simulate the reverse recovery characteristics of the diode.

[0058] Figure 5 is the encapsulation after connecting the circuit structures of the corresponding modules in Figure 4 (a)-(c). After encapsulation, the IGBT circuit structure module leads out three electrodes G, C, and E to connect to the main circuit. The main circuit module voltage is used to control the conduction and cutoff of the IGBT and assign custom parameters to the model. At the same time, it accepts the output of the custom parameter module as the control source of the voltage-controlled current source, and a drive voltage signal is introduced from the gate G to realize the control of the working state and the voltages and currents of each electrode of the IGBT. The circuit structure module is closely corresponding to the static and dynamic characteristics of the IGBT.

[0059] Finally, it should be noted that the above-mentioned are only the preferred embodiments of the present invention that are easily understood by those skilled in the art, and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A hybrid modeling and co - simulation method for a modular multilevel converter, characterized in that: The method includes the following steps: Step 1: Establish a Thevenin equivalent model of IGBT; Step 2: Establish an average value model of IGBT; Step 3: Establish a custom component model that conforms to the switching transient characteristics of IGBT; Step 4: According to the IGBT transient models obtained in Step 1, Step 2, and Step 3, connect the three models according to the IGBT structure of the MMC sub-module, modify and add corresponding parameters and modules, so as to establish a multi-time scale MMC sub-module IGBT hybrid transient model.

2. The hybrid modeling and co-simulation method of a modular multilevel converter according to claim 1, characterized in that In the above step 1, according to the Thevenin equivalent principle, the IGBT switch group in each sub-module is regarded as a variable resistor that can switch between high and low resistance values. When the switch group is turned on, the resistance value is very small (R on ), and when the switch group is turned off, the resistance value is very large (R off ). The switching of its resistance value is determined by the switch state, and the switch state is determined by the MMC controller. The specific modeling of the above equivalent module is as follows: Perform Thevenin equivalent on the IGBT switch group circuit structure and discretize the capacitors in the sub-module. The Thevenin equivalent resistance R C and the equivalent voltage U CEQ The overall equivalent sub-module capacitance C, whose values are all functions of time, as shown in Eqs. (1) and (2). The capacitor current I C (t) is calculated by Eq. (3), which is used to update the sub-module capacitor voltage in each simulation step. The superscript T in the formula indicates that the trapezoidal integration method is used in the construction of the adjoint circuit. The overall discretized circuit model is equivalently regarded as a sub-module Thevenin equivalent circuit, and the equivalent parameters R SMEQ and U SMEQ According to Equations (4) and (5), R SMEQ is also a time-varying quantity.

3. The hybrid modeling and co - simulation method of a modular multilevel converter according to claim 1, characterized in that, In the said step 2, the IGBT switching actions inside the sub-module are simplified, and the internal electrical connection relationship is ignored. A mathematical model of the device port is established with the idea of averaging. That is, in a period much longer than a single switching period, the switching changes within a single period are ignored, and the change of state is equivalent to the state mean value of continuous switching periods. The only equivalent DC-side capacitor in the average value model of MMC (used to equivalent the capacitors of three-phase six-arm bridges in the detailed model of MMC) C AVM It can be obtained by calculation from (6). Where C DM is the capacitance value of a single sub-module in the MMC detailed model.

4. The hybrid modeling and co - simulation method of a modular multilevel converter according to claim 1, characterized in that, In Step 3, for the custom IGBT switching transient simulation model, the mechanism model and the behavior model are combined. For the part involving the underlying device parameters and physical mechanisms, methods such as mechanism derivation, electrical equivalence, and curve fitting are used to avoid solving complex semiconductor physical equations; in order to reflect the output characteristics of IGBT, a MOSFET-BJT module and a Miller capacitance module based on the device manual parameters are built; at the same time, the established model starts from the steady-state and transient characteristics of IGBT and diode, fully considers the reverse recovery characteristics of the diode and its influence on IGBT, and studies the turn-on and turn-off processes of the IGBT module in detail in stages. The specific modeling method is as follows: (1) MOSFET-BJT equivalent module: The external output characteristics of IGBT are mainly determined by MOSFET and BJT. When IGBT is turned on, two current paths will be generated inside it: 1) The current path In generated by the internal electron flow, corresponding to the MOSFET structure. 2) The current path Ip generated by the hole flow, corresponding to the BJT structure. When IGBT works in the cut-off, saturation, and amplification states respectively, the current expression flowing through MOSFET is: Due to the complex extraction of the MOSFET transconductance K p and the BJT current gain β parameters, electrical equivalent simplification can be adopted based on the requirements of circuit simulation to obtain the following relationships: Therefore, the MOSFET-BJT equivalent module can use a voltage-controlled current source to simulate the on-state current Ic of the IGBT module, and its analytical expression is as follows: where the equivalent transconductance \(K=(1 + \beta)K\) p ; \(V\) ge is the gate-emitter voltage; \(V\) T is the IGBT turn-on threshold voltage; \(V\) ce is the IGBT collector-emitter voltage; \(K\) p is the MOSFET transconductance; \(\beta\) is the BJT current gain; \(I\) mos is the current flowing through the MOSFET; \(I\) c is the current flowing through the IGBT, i.e., the collector current. Among them, the equivalent transconductance \(K\) and the threshold voltage \(V\) T parameters are extracted from the device manual. (2) Tail current equivalent module: During the turn-off transient process of IGBT, due to the existence of BJT inside IGBT, it takes time for a large number of excess carriers in the base region to recombine, resulting in a relatively long tail time for the turn-off current. Where τ is the minority carrier lifetime, that is, the tail time constant, which is jointly determined by the steady-state characteristics and the tail current during the turn-off process; t0 is the starting time of the tail current; turn-off During the process, when V ge is less than the threshold voltage, tailing starts, and at this time the collector current is the tailing start current I tail0 . Adding the said formula (10) to the said formula (9), the complete MOSFET-BJT equivalent module can be obtained. (3) Miller capacitance module: There are two types of IGBT parasitic capacitance between terminals. One is related to the MOS structure, and the other is related to the PN junction. In the data sheet, the input capacitance C ies , the output capacitance C oes and the feedback capacitance C res are commonly used parameters in applications. Their relationships with the parasitic capacitance between terminals are as follows: Using the formula (11) and combining the device manual data, the corresponding inter-electrode parasitic capacitance value is obtained. (4) Diode reverse recovery current module: The process that the diode changes from the forward conduction state to the reverse blocking is called reverse recovery, and the generated reverse recovery current will affect the switching characteristics of IGBT. According to the reverse recovery characteristics of the diode and its influence on IGBT, an equivalent diode reverse recovery model is established based on the macro-model design idea. The model is composed of an ideal diode, an RL parallel circuit, and a voltage-controlled current source. The relevant parameters of the model are shown in formula (12). Among them, τ re is the reverse recovery decay time constant; R and L are free variables, which are often ignored in the modeling process because their calculated values of device parameters are very small; I rm is the reverse recovery peak current; dI f / dt is the reverse recovery current slope; t rr is the reverse recovery time; Q rr is the reverse recovery charge quantity. Construct respective defined modules according to the above partial formulas, connect them according to the IGBT module circuit structure, build a complete IGBT custom component model, and after encapsulating it, lead out three electrodes of G (gate), C (collector), and E (emitter) to connect with the main circuit.

5. The hybrid modeling and co-simulation method of a modular multilevel converter according to claim 1, characterized in that In step 4, perform hybrid co-simulation on the equivalent simulation models in step 1, step 2, and step 3 to achieve equivalent modeling of each sub-module and connection of the corresponding circuit structure, and finally build a three-phase six-arm MMC system.

Images