Matrix converter common-mode electromagnetic interference modeling and suppression method

Abstract

The invention discloses a common-mode electromagnetic interference modeling and suppression method for a matrix converter, and the method comprehensively considers high-frequency parameters and topological characteristics, builds a conduction CM EMI equivalent circuit model of the matrix converter, completes the impedance network fitting of key elements and the extraction of high-frequency parasitic parameters, and achieves the modeling and suppression of the common-mode electromagnetic interference of the matrix converter. The precision of the CM EMI model of the matrix converter is improved, and the analysis of the CM EMI characteristics of the matrix converter is simplified. According to the common-mode electromagnetic interference suppression method for the matrix converter, CM EMI characteristics of the matrix converter are analyzed, a low-attenuation loop in the converter is constructed by adopting a method of combining a neutral line leading-back type structure and increasing output filter capacitor neutral-point-to-ground parasitic capacitance, the propagation proportion of noise in a linear impedance stabilization network measurement branch is reduced, and the measurement accuracy of the common-mode electromagnetic interference suppression method for the matrix converter is improved. And effective suppression of low-cost CM EMI noise is realized on the premise of not depending on an input EMI filter.

Description

Technical Field

[0001] The present invention belongs to the field of electromagnetic compatibility technology in the field of power electronics, and specifically discloses a method for modeling and suppressing conducted electromagnetic interference of a power electronic converter, and in particular relates to a method for modeling and suppressing common-mode electromagnetic interference of a matrix converter. Background Art

[0002] Matrix converters are primarily used in industrial applications such as variable-frequency drives and wind power generation. The high-speed dv / dt generated by the high-frequency operation of their power switches generates significant electromagnetic interference (EMI) noise. GB4824 / CISPR11 specifies conducted EMI emission limits for power electronic equipment in the B-band (150kHz–30MHz) for industrial, scientific, and medical applications. Therefore, studying the conducted EMI characteristics of matrix converters and effectively mitigating them is essential for device design and development. For efficient EMI analysis, a realistic EMI equivalent circuit model must be established. Based on this model, EMI predictions can be performed. Targeted EMI mitigation methods can then be investigated and evaluated for effectiveness.

[0003] The inherent bidirectional switch array and strong input-output coupling of traditional matrix converters make their EMI modeling methods significantly different from those of other power electronic converters and highly complex. Traditional models cannot accurately characterize their conducted EMI characteristics, necessitating a more comprehensive study of their modeling methods. Furthermore, when the converter's three-phase symmetry is sufficiently high, differential-mode (DM) EMI is minimal, while common-mode (CM) EMI becomes the primary concern in conducted EMI analysis.

[0004] Technical comparison with patent CN107196523A "A T-type three-level active third harmonic injection matrix converter";

[0005] 1. Patent CN107196523A proposes a T-type three-level active third-harmonic injection matrix converter, comprising a rectifier stage, a third-harmonic injection circuit, an inverter stage, and an input LC filter. The rectifier stage is used to rectify the three-phase AC input voltage provided by a three-phase input power supply into a six-pulse DC voltage; the third-harmonic injection circuit is used to generate a third harmonic from the DC power and inject it into the rectifier stage; and the inverter stage is used to convert the DC power into AC power.

[0006] This patent focuses on the traditional matrix converter and its CM EMI suppression method. Taking into account high-frequency parameters and topological characteristics, it establishes a conducted CM EMI equivalent circuit model of the matrix converter, completes the impedance network fitting and high-frequency parasitic parameter extraction of key components, improves the accuracy of the matrix converter CM EMI model, and simplifies the analysis of the matrix converter CM EMI characteristics.

[0007] There are essential differences between the two in terms of converter topology and research content.

[0008] 2. Patent CN107196523A proposes an input LC filter in a T-type three-level active third harmonic injection matrix converter. The filter includes a filter inductor and a filter capacitor. The input LC filter is used for filtering and providing a neutral point for the inverter stage in the inverter circuit. This matrix converter offers improved control performance, such as expanding the control range of input reactive power and eliminating the need for synchronous modulation. It also provides higher output power quality, an improved output waveform, lower common-mode levels, and higher system efficiency.

[0009] This patent discloses a matrix converter CM EMI suppression method. By analyzing the CM EMI characteristics of the matrix converter, a low-attenuation loop is constructed inside the converter by combining a neutral line return structure with an increased parasitic capacitance between the midpoint of the output filter capacitor and ground. This reduces the proportion of noise propagation in the measurement branch of the linear impedance stabilization network, and effectively suppresses low-cost CM EMI noise without relying on an input EMI filter.

[0010] There are essential differences between the two in terms of CM EMI noise suppression methods and beneficial effects.

[0011] 3. Patent CN104935180A discloses a modulation method for suppressing the common-mode voltage of a matrix converter, comprising the following steps: applying singular value decomposition to the space vector form of all switching state matrices of the matrix converter, and dividing all switching states of the matrix converter into three categories based on this. Then, utilizing the equivalent relationship between the switching states of different categories of the matrix converter, the switching state corresponding to the smaller common-mode voltage value is preferentially selected in the calculation of the modulation matrix and the switching sequence of the matrix converter, and the rate of change of the output common-mode voltage is reduced by optimizing the switching sequence.

[0012] This patent discloses a matrix converter CM EMI suppression method. By analyzing the CM EMI characteristics of the matrix converter, a low-attenuation loop is constructed inside the converter by combining a neutral line return structure with an increase in the parasitic capacitance between the midpoint of the output filter capacitor and the ground. This reduces the proportion of noise propagation in the measurement branch of the linear impedance stabilization network, and achieves effective suppression of low-cost CM EMI noise without relying on an input EMI filter. Summary of the Invention

[0013] To effectively suppress CM EMI noise in a matrix converter, the present invention discloses a matrix converter CM EMI modeling and suppression method. This CM EMI modeling method comprehensively considers high-frequency parameters and topological characteristics to establish a conducted CMEMI equivalent circuit model of the matrix converter, complete impedance network fitting of key components, and extract high-frequency parasitic parameters. Based on this equivalent model, the conducted CM EMI characteristics are analyzed, achieving effective and low-cost CM EMI noise suppression without relying on input EMI filters.

[0014] To achieve the above object, the technical solution adopted by the present invention is:

[0015] A matrix converter common-mode electromagnetic interference modeling method, characterized by comprising the following steps:

[0016] S1: Parasitic parameters and propagation path analysis;

[0017] S2: Noise source equivalent analysis;

[0018] S3: Establish CM EMI equivalent circuit model;

[0019] S4: High-frequency parameter extraction and feature fitting.

[0020] As a preferred technical solution of the present invention: in step S1, the parasitic parameters and propagation path analysis are specifically as follows:

[0021] The parasitic effects of components in the matrix converter are reflected in the following two aspects:

[0022] Parasitic capacitance of various components and wiring to ground

[0023] Parallel parasitic capacitance of inductors and self-resonance

[0024] The parasitic capacitance of each component and wiring to ground is represented by lumped parameters, as follows:

[0025] C xy1 with C xy2 are the parasitic capacitances of the collector pins of the power switches on the output and input sides to ground, respectively, where x = a, b, c, y = r, s, t,

[0026] C ym with C xm are the parasitic capacitances of the wiring from the output and input power switch pins to the filter terminals to ground,

[0027] C Lo is the output filter inductor L fo Output terminal to ground parasitic capacitance,

[0028] C pfi with C pfo are the midpoint-to-ground capacitances of the input and output filter capacitors,

[0029] C NG is the capacitance from the load midpoint to ground,

[0030] In view of the isolation characteristics of LISN, only the isolation capacitance CLS and the internal resistance RLS of the LISN measurement branch are considered.

[0031] The parasitic capacitance to ground, the inductance and resistance equivalent to CM, and the LISN measurement branch together constitute the CM noise propagation path.

[0032] As a preferred technical solution of the present invention: in step S2, the noise source equivalent analysis is specifically as follows:

[0033] S21. Equivalent alternative methods:

[0034] According to the substitution theorem, Kirchhoff's law, and Ohm's law, a voltage source or current source with the same waveform is used to replace the power switch. After the replacement, the voltage drop and current of each branch are the same as before the replacement. Therefore, the voltage source or current source is used to equivalently replace the nonlinear switch, and the superposition theorem is used to analyze the impact of different noise sources on conducted CM EMI. When using the above equivalent substitution method, the following principles should be followed:

[0035] To ensure that there is a unique current solution, a circuit cannot only have an equivalent voltage source

[0036] To ensure that there is a unique voltage solution, a node cannot be connected only to an equivalent current source.

[0037] Under the premise of meeting the above constraints, priority is given to alternative methods that facilitate circuit CM EMI analysis;

[0038] S21. Noise source equivalent analysis:

[0039] According to the working characteristics of the matrix converter with current source input and voltage source output, the midpoint of the input filter capacitor is taken as the reference point of the noise source to form the noise source equivalent circuit, where V ro , V so , V to are the high-frequency PWM voltages from the three-phase output points of the bidirectional switch array to the midpoint of the input filter capacitor, C sa 、C sb 、C sc 、C sr 、C ss 、C st are the sum of the parasitic capacitances to ground of the power switch pins and related wiring connected to the corresponding phase, that is, C sa = Car2 + C as2 + C at2 + C am , and so on.

[0040] As a preferred technical solution of the present invention: in step S3, the CM EMI equivalent circuit model is established as follows:

[0041] S31. According to the superposition theorem, the current source is open-circuited, and the microfarad input filter capacitor C is used in the 150kHz~30MHz frequency band. fi and the output filter capacitor C fo The impedance is less than the picofarad level C pfi with C pfo Impedance, short circuit treatment;

[0042] S32. Based on the CM EMI propagation path, the noise source, inductance, resistance, and parasitic capacitance are combined according to the CM definition to obtain the CM equivalent circuit model A, where: V cm = (V ro +V so + V to ) / 3,

[0043] In the CM propagation path, the input side parasitic capacitance C sa 、C sb 、C sc are connected in parallel and can be combined into one capacitor, namely C pm = C sa +C sb + C sc , and so on, C ps = C sr +C ss + C st , three-phase C LS Merged into 3C LS Similarly, the inductance and resistance in the CM circuit are treated in parallel, with R LS / 3、(Z Lfi / / R fi ) / 3, Z Lfo / 3 and Z L / 3 means, where Z L = Z(L L + R L ), by further combining the parasitic capacitance, we can get its simplified equivalent circuit B, where: C si = C pm +C pfi , C so = 3C Lo +C pfo ,

[0044] For the convenience of description, let Z X = Z L / 3 + Z(C NG ), then the output side impedance is combined to obtain a simplified equivalent circuit C, where: Z o = (Z Lfo / 3 + (Z(C so ) / / Z X )) / / Z(C ps ),

[0045] According to Thevenin's theorem, the above active two-terminal network is equivalent to the simplest circuit D, which is the conducted CM EMI equivalent circuit model of the matrix converter, where:

[0046] (1)

[0047] (2)

[0048] The final simplification process retains the measured internal resistance of the LISN, and i cm1 The point where the CM current is measured.

[0049] As a preferred technical solution of the present invention: in step S4, high-frequency parameter extraction and characteristic fitting are specifically as follows:

[0050] S41. High-frequency parameter extraction

[0051] First, install the power module on the heat sink and PCB board, and complete the installation of the experimental prototype according to the actual working conditions. Then use the R&S vector network analyzer to directly measure the CM parasitic capacitance of the module's three-phase input to ground. The measured value includes the parasitic capacitance of the input side power switch to ground and the capacitance of the input side PCB wiring and the midpoint of the input filter capacitor to ground, that is, C si ,

[0052] Use a network analyzer to measure the capacitance of each phase output point of the power module to ground. The measured value includes the parasitic capacitance of the output side power switch and the connected wiring to ground. The sum of the three phases is C ps ,

[0053] In addition, separate measurements include 3C Lo The parasitic capacitance C between the midpoint of the output filter capacitor and the ground so Parasitic capacitance C between the load midpoint and ground NG ;

[0054] S42. Measure the impedance characteristics of the inductor used and complete the equivalent impedance network fitting

[0055] First, the possible impedance network structure should be derived. The CM impedance curve of the relevant filter inductor is measured by an impedance analyzer. According to the number of poles and zeros of the filter inductor in the corresponding frequency band shown by the impedance curve, the number of inductors and capacitors required in the impedance network in addition to the nominal inductor is selected. In this way, the simplest impedance network form of multiple RLC series and parallel connections can be obtained.

[0056] On the basis of the necessary network, an RLC parallel network and an RC parallel circuit are connected in series to compensate for the nonlinear characteristics of the impedance curve. On the basis of determining the equivalent impedance network structure, the random system + simplex double iterative algorithm combined with the Levenberg-Marquardt estimation method are used to obtain the equivalent networks of various filter inductors and the corresponding fitting curves.

[0057] A method for suppressing common-mode electromagnetic interference of a matrix converter, characterized by comprising the following steps:

[0058] S1: Connect the midpoint O of the input filter capacitor and the midpoint M of the output filter capacitor with a short high-frequency wire to form a neutral-line return matrix converter;

[0059] S2: Increase the parasitic capacitance between the midpoint of the output filter capacitor and the ground to construct an internal low-attenuation loop.

[0060] As a preferred technical solution of the present invention: in step S1, the neutral line return type matrix converter is established as follows:

[0061] A neutral-line-return matrix converter includes a power grid AC source, a linear stabilizing impedance network, an input LC filter circuit, a bidirectional matrix switch network, and an output LCL filter circuit. The power grid AC source is used to provide power, the linear stabilizing impedance network is used to provide stable line impedance and isolate conducted EMI interference from external systems, and acts as a sensor to couple EMI noise to a measurement port and transmit it to a receiver. The input LC filter circuit is used to filter out high-frequency harmonics and noise, and the output LCL filter circuit is used to suppress output high-frequency switching harmonics. The power grid AC source is connected to one end of the input LC filter, the other end of the input LC filter is connected to one end of the bidirectional matrix switch network, the other end of the bidirectional matrix switch network is connected to one end of the output LCL filter, and the other end of the output LCL filter is connected. The midpoint O of the input filter capacitor of the input LC filter is connected to the midpoint M of the output filter capacitor of the output LCL filter, forming a neutral-line-return structure.

[0062] As a preferred technical solution of the present invention: in step S2, the internal low attenuation loop is constructed as follows:

[0063] The CM EMI model of the neutral-line-return matrix converter is established using the matrix converter common-mode electromagnetic interference modeling method. The midpoint O of the input filter capacitor of the input LC filter and the midpoint M of the output filter capacitor of the output LCL filter are short-circuited to obtain an equivalent circuit. The parasitic capacitance is combined to obtain a simplified equivalent circuit, which is then equivalent to the simplest equivalent circuit according to the Thevenin theorem.

[0064] (3)

[0065] (4)

[0066] Then, according to the EMI test standard, we define the value in dB by taking the logarithm of equation (4), and substituting equations (3) and (4) into the equation, we can get the conducted CMEMI attenuation of the neutral-lead MC:

[0067] (5)

[0068] According to formula (5), reduce Z(C P ) can construct an internal low attenuation loop, C P It includes the input side power switch, the midpoint of the input filter capacitor, and the parasitic capacitance of the midpoint of the output filter capacitor.

[0069] Compared with the prior art, the present invention has the following beneficial effects:

[0070] 1. The present invention discloses a matrix converter CM EMI modeling method, which comprehensively considers high-frequency parameters and topological characteristics, establishes a conducted CM EMI equivalent circuit model of the matrix converter, completes impedance network fitting and high-frequency parasitic parameter extraction of key components, improves the accuracy of the matrix converter CM EMI model, and simplifies the analysis of the matrix converter CM EMI characteristics.

[0071] 2. The present invention discloses a matrix converter CM EMI suppression method. By analyzing the CM EMI characteristics of the matrix converter, a low-attenuation loop is constructed inside the converter by combining a neutral line return structure with an increased parasitic capacitance between the midpoint of the output filter capacitor and ground. This reduces the proportion of noise propagation in the measurement branch of the linear impedance stabilization network, achieving effective and low-cost CM EMI noise suppression without relying on an input EMI filter. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 Schematic diagram of the matrix converter topology considering parasitic capacitance;

[0073] Figure 2 Schematic diagram of the equivalent circuit structure of the noise source of the matrix converter according to an embodiment of the present invention;

[0074] Figure 3 This is a schematic diagram of a simplified equivalent circuit structure of a noise source of a matrix converter according to an embodiment of the present invention;

[0075] Figure 4 This is a schematic diagram of the structure of a simplified equivalent circuit A of a noise source of a matrix converter according to an embodiment of the present invention;

[0076] Figure 5 This is a schematic diagram of the structure of a simplified equivalent circuit B of a noise source of a matrix converter according to an embodiment of the present invention;

[0077] Figure 6 Schematic diagram of the structure of a simplified equivalent circuit C of a noise source of a matrix converter according to an embodiment of the present invention;

[0078] Figure 7 Schematic diagram of the structure of a simplified equivalent circuit D of a noise source of a matrix converter according to an embodiment of the present invention;

[0079] Figure 8 This is a schematic diagram of the internal structure of a power switch array power module used in a matrix converter according to an embodiment of the present invention;

[0080] Figure 9 CM impedance measurement curve a of the filter inductor related to the matrix converter according to an embodiment of the present invention;

[0081] Figure 10 CM impedance measurement curve b of the filter inductor related to the matrix converter according to an embodiment of the present invention;

[0082] Figure 11 CM impedance measurement curve c of the filter inductor related to the matrix converter according to an embodiment of the present invention;

[0083] Figure 12 d is a CM impedance measurement curve of the filter inductor related to the matrix converter according to an embodiment of the present invention;

[0084] Figure 13 The equivalent impedance network I of the filter inductor related to the matrix converter according to the embodiment of the present invention;

[0085] Figure 14 The equivalent impedance network II of the filter inductor related to the matrix converter according to the embodiment of the present invention;

[0086] Figure 15 The equivalent impedance network III of the filter inductor related to the matrix converter according to the embodiment of the present invention;

[0087] Figure 16 The equivalent impedance network IV of the filter inductor related to the matrix converter according to the embodiment of the present invention;

[0088] Figure 17 The topology diagram of the neutral line return type matrix converter considering parasitic capacitance is shown in the figure.

[0089] Figure 18 Schematic diagram of the structure of a simplified equivalent circuit A of a noise source of a line-return matrix converter according to an embodiment of the present invention;

[0090] Figure 19 Schematic diagram of the structure of a simplified equivalent circuit B of a line-return matrix converter noise source in an embodiment of the present invention;

[0091] Figure 20 They are schematic structural diagrams of an equivalent simplified circuit C of a line-return matrix converter noise source in an embodiment of the present invention. DETAILED DESCRIPTION

[0092] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0093] The present invention proposes a matrix converter common-mode electromagnetic interference modeling method, comprising the following steps:

[0094] S1: Parasitic parameters and propagation path analysis;

[0095] S2: Noise source equivalent analysis;

[0096] S3: Establish CM EMI equivalent circuit model;

[0097] S4: High-frequency parameter extraction and feature fitting.

[0098] Figure 1 The topology of the matrix converter considering parasitic capacitance is shown in Figure 2. The bidirectional switches Q1-Q9 in the switch array are in the form of reverse resistance IGBTs, which can be equivalent to IGBT common emitter bidirectional switches. The input end of the switch array is connected to a damped LC filter, which consists of an input filter inductor L. fi and input impedance R fi The output of the switch array is connected to the LCL filter, which consists of an output filter inductor L fo and L L , output filter capacitor C fo The other end is connected to the resistor load R L At the same time, to meet the conducted emission measurement requirements of GB 4824 / CISPR11, a Line Impedance Stabilization Network (LISN) was introduced at the grid port to provide stable line impedance and isolate conducted EMI interference from external systems.

[0099] The common-mode electromagnetic interference modeling method of the matrix converter is as follows:

[0100] S1: Parasitic Parameters and Propagation Path Analysis

[0101] exist Figure 1 The lumped parameters are used to represent the parasitic capacitance of each device and wiring to the ground, as follows: C xy1 (x= a, b, c, y= r, s, t) with C xy2 are the parasitic capacitances of the collector pins of the power switches on the output and input sides to the ground respectively; C ym with C xm are the parasitic capacitances of the wiring from the output side and input side power switch pins to the filter terminals to ground; C Lo is the output filter inductor L fo Output terminal to ground parasitic capacitance; C pfi with C pfo are the midpoint-to-ground capacitances of the input and output filter capacitors respectively; C NG is the capacitance from the midpoint of the load to ground.

[0102] In view of the high isolation characteristics of LISN, it is assumed that there is no mutual propagation of CM EMI between the converter and the grid. Therefore, only the isolation capacitance C of the LISN measurement branch is considered. LS (0.1μF) and measure the internal resistance R LS (50Ω).

[0103] The matrix converter CM EMI noise propagation path includes the aforementioned parasitic capacitance to ground, the CM equivalent model of the filter inductor, resistor, and capacitor, and the LISN measurement branch.

[0104] S2: Noise source equivalent analysis:

[0105] According to the substitution theorem and the limitations of Kirchhoff's law and Ohm's law, if a voltage source or current source with the same waveform is used to replace the power switch, the operating characteristics of the related branches will remain unchanged, that is, the voltage drop and current of each branch after the replacement are the same as before the replacement. According to the operating characteristics of the matrix converter with current source input and voltage source output, the midpoint of the input filter capacitor is taken as the reference point of the noise source, and the noise source equivalent circuit is formed as follows: Figure 2 As shown in the figure, V ro , V so , V to are the high-frequency pulse width modulation voltages from the three-phase output points of the bidirectional switch array to the midpoint of the input filter capacitor, C sa 、C sb 、C sc 、C sr 、C ss 、C st are the sum of the parasitic capacitances to ground of the power switch pins and related wiring connected to the corresponding phase, that is, C sa = C ar2 + C as2 +C at2 + Cam , and so on.

[0106] S3: Establish CM EMI equivalent circuit model:

[0107] According to the superposition theorem, the current source is treated as an open circuit, and the microfarad input filter capacitor C is used in the 150kHz~30MHz frequency band. fi and the output filter capacitor C fo The impedance of the picofarad C pfi with C pfo Obviously smaller, it can be treated as a short circuit, and the equivalent circuit can be simplified, such as Figure 3 shown.

[0108] According to the CM EMI propagation path, the noise source, inductance, resistance and parasitic capacitance can be combined according to the CM definition to obtain the CM equivalent circuit model A, as shown in the following example: Figure 4 As shown, where: V cm = (V ro +V so + V to ) / 3. In the CM propagation path, the input side parasitic capacitance C sa 、C sb 、C sc are connected in parallel and can be combined into one capacitor, namely C pm = C sa +C sb + C sc , and so on, C ps = C sr +C ss + C st , three-phase C LS Merged into 3C LS Similarly, on the basis of high enough three-phase symmetry of the circuit, the inductor and resistor in the CM circuit are treated in parallel, with R LS / 3、(Z Lfi / / R fi ) / 3, Z Lfo / 3 and Z L / 3 means, where Z L = Z(L L +R L ). By further combining the parasitic capacitance, we can get its simplified equivalent circuit B, as shown in Figure 5 As shown, where: C si = C pm +C pfi , C so = 3C Lo +C pfo For the convenience of description, let Z X = ZL / 3 + Z(C NG ), then the output side impedance is combined to obtain a simplified equivalent circuit C, as shown in Figure 6 As shown, where: Z o = (Z Lfo / 3 + (Z(C so ) / / Z X )) / / Z(C ps ). According to Thevenin's theorem, the above active two-terminal network is equivalent to the simplest circuit D, as shown in Figure 7 As shown in Figure 1, this is the conduction CM EMI equivalent circuit model of the matrix converter.

[0109] (1)

[0110] (2)

[0111] The final simplification process retains the measured internal resistance of the LISN, and i cm1 The CM current measurement point is shown in Figure 2. Obviously, this equivalent circuit model can intuitively reflect the CM EMI characteristics of the matrix converter.

[0112] S4: High-frequency parameter extraction and characteristic fitting:

[0113] In this embodiment, a Fuji 18MBI power module with 9 sets of bidirectional reverse-blocking IGBTs is used as the power switch array of the matrix converter. Figure 8 To ensure a close approximation to actual conditions and facilitate measurement, the experimental prototype must be installed according to actual operating conditions (the power module has been installed on the heat sink and PCB board). The R&S vector network analyzer is then used to directly measure the CM parasitic capacitance of the module's three-phase input to ground. This measurement value includes the parasitic capacitance of the input-side power switch to ground and the capacitance of the input-side PCB wiring and the midpoint of the input filter capacitor to ground, namely, C si The power module package and PCB board are highly symmetrical. Use a network analyzer to measure the capacitance of each phase output point of the power module to ground. The measured value includes the parasitic capacitance of the output side power switch and the connected wiring to ground. The sum of the three phases is C ps In addition, a separate measurement includes 3C Lo The parasitic capacitance C between the midpoint of the output filter capacitor and the ground so Parasitic capacitance C between the load midpoint and ground NG .

[0114] In the 150kHz to 30MHz frequency range, conventional filter inductors exhibit self-resonance due to parasitic parameters, resulting in complex impedance characteristics. In this case, simply characterizing the impedance using nominal inductance is inadequate. Impedance measurement of the inductor and equivalent impedance network fitting are required to ensure the accuracy of the CM EMI equivalent model.

[0115] In order to improve the accuracy of fitting, the possible impedance network structure should be derived first. The CM impedance measurement curve of the relevant filter inductor is as follows: Figure 9 As shown. The impedance curve shows that the filter inductor has two obvious poles and one zero point in the corresponding frequency band. In addition to the nominal inductor, at least one inductor and two capacitors are required in the impedance network. Therefore, the simplest impedance network is two RLC parallel circuits in series. In order to achieve the best impedance characteristic fitting effect, this embodiment connects an RLC parallel circuit and an RC parallel circuit in series on the basis of the necessary network to compensate for the nonlinear characteristics of the impedance curve. On the basis of determining the equivalent impedance network structure, the random system + simplex double iterative algorithm is combined with the Levenberg-Marquardt estimation method to obtain the equivalent networks of the three filter inductors as shown below. Figure 9 、 10 , 11, including the output side 2mH inductor, the output measurement 0.47mH inductor, the input measurement 1mH inductor, the corresponding fitting curve is as follows Figure 12 、 13 , 14. In addition, in order to further improve the accuracy of the CMEMI equivalent model, this embodiment also fits the CM impedance characteristics of the parallel branch of the filter inductor and damping resistor of the input filter, and obtains the equivalent network as shown in Figure 15 The corresponding fitting curve is shown by Figure 12 As shown. According to the fitting results, within the required frequency band, the CM impedance characteristics of the input filter inductor parallel branch are dominated by the damping resistor, which can be approximately equivalent to the damping resistor CM value connected in series with a nanohenry-level inductor. Figure 9-12 It can be seen that the fitting results of the impedance characteristics of each inductor match the actual measurement curve well, laying the foundation for improving the accuracy of the CM EMI equivalent model.

[0116] The present invention proposes a common-mode electromagnetic interference modeling method for a matrix converter. By comprehensively considering high-frequency parameters and topological characteristics, a conducted CM EMI equivalent circuit model of the matrix converter is established. The impedance network fitting and high-frequency parasitic parameter extraction of key components are completed, thereby improving the accuracy of the matrix converter CM EMI model and simplifying the analysis of the matrix converter CM EMI characteristics.

[0117] A method for suppressing common-mode electromagnetic interference of a matrix converter comprises the following steps:

[0118] S1: Connect the midpoint O of the input filter capacitor and the midpoint M of the output filter capacitor with a short high-frequency wire to form a neutral-line return matrix converter;

[0119] S2: Increase the parasitic capacitance between the midpoint of the output filter capacitor and the ground to construct an internal low-attenuation loop:

[0120] The key approach to achieving conducted CM EMI suppression is to construct a low-attenuation loop within the matrix converter to reduce the proportion of noise propagating through the LISN measurement branch. Constrained by its circuit structure, the matrix converter lacks the freedom to suppress CM EMI, making it difficult to form an internal low-attenuation loop. Therefore, without compromising operational characteristics, the simplest approach is to connect the input filter capacitor midpoint O and the output filter capacitor midpoint M with the shortest possible high-frequency wire, forming a neutral-line-return matrix converter, as shown in Figure 1. Figure 17 As shown, this provides a basis for constructing a low attenuation loop.

[0121] Step S1 is specifically as follows:

[0122] like Figure 17 As shown, a neutral-line return matrix converter circuit includes a power grid AC source, a linear stabilizing impedance network, an input LC filter circuit, a bidirectional matrix switch network, and an output LCL filter circuit. The power grid AC source is used to provide power. The linear stabilizing impedance network is used to provide stable line impedance and isolate conducted EMI interference from external systems. It also acts as a sensor to couple EMI noise to a measurement port and transmit it to a receiver. The input LC filter circuit is used to filter out high-frequency harmonics and noise. The output LCL filter circuit is used to suppress output high-frequency switching harmonics. The power grid AC source is connected to one end of the input LC filter, the other end of the input LC filter is connected to one end of the bidirectional matrix switch network, the other end of the bidirectional matrix switch network is connected to one end of the output LCL filter, and the other end of the output LCL filter is connected. The midpoint O of the input filter capacitor of the input LC filter is connected to the midpoint M of the output filter capacitor of the output LCL filter, forming a neutral-line return structure for increasing the degree of freedom of conducted CM EMI suppression.

[0123] Step S2 is specifically as follows:

[0124] The midpoint O of the input filter capacitor of the input LC filter and the midpoint M of the output filter capacitor of the output LCL filter are short-circuited to obtain an equivalent circuit E of a neutral line-returned matrix converter with parasitic capacitance, as shown in FIG. Figure 18 As shown. Combining the parasitic capacitances gives the simplified equivalent circuit F, as Figure 19 As shown. Then according to Thevenin's theorem, we can get its simplest equivalent circuit G, as shown in Figure 20 As shown,

[0125] (3)

[0126] (4)

[0127] Then, according to the EMI test standard, we define it in dB and take the logarithm of equation (4). Then, we can substitute equations (3) and (4) to obtain the conducted CM EMI attenuation of the neutral-lead MC:

[0128] (5)

[0129] According to formula (5), reduce Z(C P ) (Increase C P Capacitance) can construct an internal low attenuation circuit, C P This includes the input-side power switch, the parasitic capacitance at the midpoint of the input filter capacitor, and the parasitic capacitance at the midpoint of the output filter capacitor. Therefore, by increasing the parasitic capacitance at the midpoint of the output filter capacitor, this embodiment can easily implement a low-attenuation loop within the matrix converter without relying on an input-side EMI filter, thereby reducing the proportion of noise propagation in the measurement branch of the linear impedance stabilization network and achieving low-cost CM EMI noise suppression.

[0130] The present invention proposes a method for suppressing common-mode electromagnetic interference in a matrix converter. By analyzing the CMEMI characteristics of the matrix converter, a low-attenuation loop is constructed inside the converter by combining a neutral line return structure with increased parasitic capacitance between the midpoint of the output filter capacitor and ground. This reduces the proportion of noise propagation in the measurement branch of the linear impedance stabilization network, achieving effective and low-cost suppression of CM EMI noise without relying on an input EMI filter.

[0131] The above description is merely a preferred embodiment of the present invention and does not constitute any other form of limitation to the present invention. Any modification or equivalent variation based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A matrix converter common mode electromagnetic interference modeling method, characterized in that: The steps include: S1: Parasitic parameters and propagation path analysis; S2: Noise source equivalent analysis; S3: Establish CM EMI equivalent circuit model; S4: High-frequency parameter extraction and feature fitting.

2. A matrix converter common mode electromagnetic interference modeling method according to claim 1, characterized in that: In step S1, the parasitic parameters and propagation path analysis are as follows: The parasitic effects of components in the matrix converter are reflected in the following two aspects: Parasitic capacitance of various components and wiring to ground Parallel parasitic capacitance of inductors and self-resonance The parasitic capacitance of each component and wiring to ground is represented by lumped parameters, as follows: C xy1 with C xy2 are the parasitic capacitances of the collector pins of the power switches on the output and input sides to ground, respectively, where x = a, b, c, y = r, s, t, C ym with C xm are the parasitic capacitances of the wiring from the output and input power switch pins to the filter terminals to ground, C Lo is the output filter inductor L fo Output terminal to ground parasitic capacitance, C pfi with C pfo are the midpoint-to-ground capacitances of the input and output filter capacitors, C NG is the capacitance from the load midpoint to ground, In view of the isolation characteristics of LISN, only the isolation capacitance CLS and the internal resistance RLS of the LISN measurement branch are considered. The parasitic capacitance to ground, the inductance and resistance equivalent to CM, and the LISN measurement branch together constitute the CM noise propagation path.

3. The matrix converter common mode electromagnetic interference modeling method according to claim 1, characterized in that: In step S2, the noise source equivalent analysis is as follows: S21. Equivalent alternative methods: According to the substitution theorem, Kirchhoff's law, and Ohm's law, a voltage source or current source with the same waveform is used to replace the power switch. After the replacement, the voltage drop and current of each branch are the same as before the replacement. Therefore, the voltage source or current source is used to equivalently replace the nonlinear switch, and the superposition theorem is used to analyze the impact of different noise sources on conducted CM EMI. When using the above equivalent substitution method, the following principles should be followed: To ensure that there is a unique current solution, a circuit cannot only have an equivalent voltage source To ensure that there is a unique voltage solution, a node cannot be connected only to an equivalent current source. Under the premise of meeting the above constraints, priority is given to alternative methods that facilitate circuit CM EMI analysis; S21. Noise source equivalent analysis: According to the working characteristics of the matrix converter with current source input and voltage source output, the midpoint of the input filter capacitor is taken as the reference point of the noise source to form the noise source equivalent circuit, where V ro , V so , V to are the high-frequency PWM voltages from the three-phase output points of the bidirectional switch array to the midpoint of the input filter capacitor, C sa 、C sb 、C sc 、C sr 、C ss 、C st are the sum of the parasitic capacitances to ground of the power switch pins and related wiring connected to the corresponding phase, that is, C sa = C ar2 + C as2 + C at2 + C am , and so on.

4. The matrix converter common mode electromagnetic interference modeling method according to claim 1, characterized in that: In step S3, the CM EMI equivalent circuit model is established as follows: S31. According to the superposition theorem, the current source is open-circuited, and the microfarad input filter capacitor C is used in the 150kHz~30MHz frequency band. fi and the output filter capacitor C fo The impedance is less than the picofarad level C pfi with C pfo Impedance, short circuit treatment; S32. Based on the CM EMI propagation path, the noise source, inductance, resistance, and parasitic capacitance are combined according to the CM definition to obtain the CM equivalent circuit model A, where: V cm = (V ro +V so + V to ) / 3, In the CM propagation path, the input side parasitic capacitance C sa 、C sb 、C sc are connected in parallel and can be combined into one capacitor, namely C pm = C sa +C sb + C sc , and so on, C ps = C sr +C ss + C st , three-phase C LS Merged into 3C LS Similarly, the inductance and resistance in the CM circuit are treated in parallel, with R LS / 3、(Z Lfi / / R fi ) / 3, Z Lfo / 3 and Z L / 3 means, where Z L = Z(L L + R L ), by further combining the parasitic capacitance, we can get its simplified equivalent circuit B, where: C si = C pm +C pfi , C so = 3C Lo +C pfo , For the convenience of description, let Z X = Z L / 3 + Z(C NG ), then the output side impedance is combined to obtain a simplified equivalent circuit C, where: Z o = (Z Lfo / 3 + (Z(C so ) / / Z X )) / / Z(C ps ), According to Thevenin's theorem, the above active two-terminal network is equivalent to the simplest circuit D, which is the conduction CMEMI equivalent circuit model of the matrix converter, where: (1) (2) The final simplification process retains the measured internal resistance of the LISN, and i cm1 The point where the CM current is measured.

5. The matrix converter common mode electromagnetic interference modeling method according to claim 1, characterized in that: In step S4, high-frequency parameter extraction and characteristic fitting are specifically as follows: S41. High-frequency parameter extraction First, install the power module on the heat sink and PCB board, and complete the installation of the experimental prototype according to the actual working conditions. Then use the R&S vector network analyzer to directly measure the CM parasitic capacitance of the module's three-phase input to ground. The measured value includes the parasitic capacitance of the input side power switch to ground and the capacitance of the input side PCB wiring and the midpoint of the input filter capacitor to ground, that is, C si , Use a network analyzer to measure the capacitance of each phase output point of the power module to ground. The measured value includes the parasitic capacitance of the output side power switch and the connected wiring to ground. The sum of the three phases is C ps , In addition, separate measurements include 3C Lo The parasitic capacitance C between the midpoint of the output filter capacitor and the ground so Parasitic capacitance C between the load midpoint and ground NG ; S42. Measure the impedance characteristics of the inductor used and complete the equivalent impedance network fitting First, the possible impedance network structure should be derived. The CM impedance curve of the relevant filter inductor is measured by an impedance analyzer. According to the number of poles and zeros of the filter inductor in the corresponding frequency band shown by the impedance curve, the number of inductors and capacitors required in the impedance network in addition to the nominal inductor is selected. In this way, the simplest impedance network form of multiple RLC series and parallel connections can be obtained. On the basis of the necessary network, an RLC parallel network and an RC parallel circuit are connected in series to compensate for the nonlinear characteristics of the impedance curve. On the basis of determining the equivalent impedance network structure, the random system + simplex double iterative algorithm combined with the Levenberg-Marquardt estimation method are used to obtain the equivalent networks of various filter inductors and the corresponding fitting curves.

6. A method for suppressing common-mode electromagnetic interference of a matrix converter, characterized in that: The steps include: S1: Connect the midpoint O of the input filter capacitor and the midpoint M of the output filter capacitor with a short high-frequency wire to form a neutral-line return matrix converter; S2: Increase the parasitic capacitance between the midpoint of the output filter capacitor and the ground to construct an internal low-attenuation loop.

7. The method for suppressing common-mode electromagnetic interference of a matrix converter according to claim 6, characterized in that: In step S1, the neutral line return type matrix converter is established as follows: A neutral-line-return matrix converter includes a power grid AC source, a linear stabilizing impedance network, an input LC filter circuit, a bidirectional matrix switch network, and an output LCL filter circuit. The power grid AC source is used to provide power, the linear stabilizing impedance network is used to provide stable line impedance and isolate conducted EMI interference from external systems, and acts as a sensor to couple EMI noise to a measurement port and transmit it to a receiver. The input LC filter circuit is used to filter out high-frequency harmonics and noise, and the output LCL filter circuit is used to suppress output high-frequency switching harmonics. The power grid AC source is connected to one end of the input LC filter, the other end of the input LC filter is connected to one end of the bidirectional matrix switch network, the other end of the bidirectional matrix switch network is connected to one end of the output LCL filter, and the other end of the output LCL filter is connected. The midpoint O of the input filter capacitor of the input LC filter is connected to the midpoint M of the output filter capacitor of the output LCL filter, forming a neutral-line-return structure.

8. The method for suppressing common-mode electromagnetic interference of a matrix converter according to claim 6, wherein: In step S2, the internal low attenuation loop is constructed as follows: The CM EMI model of the neutral-line-return matrix converter is established using the matrix converter common-mode electromagnetic interference modeling method. The midpoint O of the input filter capacitor of the input LC filter and the midpoint M of the output filter capacitor of the output LCL filter are short-circuited to obtain an equivalent circuit. The parasitic capacitance is combined to obtain a simplified equivalent circuit, which is then equivalent to the simplest equivalent circuit according to the Thevenin theorem. (3) (4) Then, according to the EMI test standard, we define the value in dB by taking the logarithm of equation (4), and substituting equations (3) and (4) into the equation, we can get the conducted CMEMI attenuation of the neutral-lead MC: (5) According to formula (5), reduce Z(C P ) can construct an internal low attenuation loop, C P It includes the input side power switch, the midpoint of the input filter capacitor, and the parasitic capacitance of the midpoint of the output filter capacitor.

Images