Parameter optimization method and system for spread spectrum modulation of microgrid interface inverter

Abstract

The invention discloses a parameter optimization method and system for spread spectrum modulation of a microgrid interface inverter, and the method comprises the following steps: 1, constructing an electromagnetic interference model of a two-stage grid-connected inverter on simulation software, selecting a plurality of key nodes, and obtaining the voltage and current waveforms of the key nodes; 2, on the basis of the voltage and current waveforms of the key nodes obtained in the step 1, an electromagnetic interference model of the two-stage grid-connected inverter is simplified firstly, a common-mode loop of the two-stage grid-connected inverter is constructed, so that a common-mode noise source of the loop is determined, the common-mode loop is equivalent to an RLC series loop, weighting coefficients of front-stage common-mode noise sources and rear-stage common-mode noise sources are obtained, and the weighting coefficients of the front-stage common-mode noise sources and the rear-stage common-mode noise sources are obtained; therefore, the total common-mode noise source of the circuit is determined, and the noise distribution condition is qualitatively estimated. And step 3, determining a used spread spectrum modulation function, and completing optimization. The method aims at reducing conducted emission noise caused by a large number of high-frequency power electronic devices in an existing micro-grid system, and a model basis is provided for design of a post-stage EMI filter.

Description

Technical Field

[0001] This invention belongs to the field of microgrid interface inverter technology, specifically relating to a parameter optimization method and system for spread spectrum modulation of microgrid interface inverters. Background Technology

[0002] With the scale of new energy power generation growing year by year, inverters, as the power conversion interface of AC microgrids, have a significant impact on the electromagnetic compatibility and power quality of the system. However, as users' requirements for inverter efficiency and size continue to increase, high-performance wide-bandgap materials such as silicon carbide (SiC) and gallium nitride (GaN) are widely used in these devices. The excellent high-temperature, high-frequency, and high-voltage operating characteristics of wide-bandgap switching devices allow them to operate stably at higher switching frequencies than silicon material switching devices, thereby reducing system losses and the requirements for filters. However, faster switching speeds do not all bring benefits; they also bring larger di / dt and du / dt, thus introducing serious electromagnetic interference problems into the system. The starting frequency of most conducted emission electromagnetic compatibility standards is 150kHz. Some high-frequency MOSFETs may operate directly in the conducted emission frequency band, leading to a sharp increase in the system's demand for electromagnetic interference (EMI) filters, resulting in increased system size and reduced power density. Reducing the conducted emission of the system is a key technology for further improving the power density of inverters.

[0003] There are numerous methods to reduce conducted emissions from inverters, such as passive and active filters, soft switching, and spread spectrum modulation. Passive filters are simple to design, effective, and suitable for various power applications, but they are relatively large. Active filters are small, but only suitable for low-power applications, and obviously not suitable for high-power microgrid systems. Soft switching technology reduces EMI and switching losses by bringing the voltage or current of devices to zero before switching through resonant circuits, but this requires additional passive components and switches, increasing system complexity, and is difficult to implement in high-voltage applications. Spread spectrum modulation reduces interference by spreading harmonic noise across a wider frequency band by fluctuating the switching frequency. This method can reduce EMI without additional hardware circuitry and has received widespread attention from the academic community in recent years.

[0004] Spread spectrum modulation can be divided into periodic frequency modulation (PFM) and random frequency modulation (RFM) based on its implementation method. RFM can achieve good spread spectrum effect when the randomness of frequency change is good, but uniformly distributed random numbers require complex algorithms or additional circuits. If random numbers are directly generated by a common controller, the spread spectrum effect is not as good as PFM. PFM determines the change of switching frequency through a certain periodic function, which is simple to implement and has good effect. However, most existing PFM research focuses on the curve shape of the spread spectrum modulation function, mainly to reduce noise spikes by increasing the spread spectrum range. The design process is complicated and the effect is not obvious.

[0005] Therefore, designing a parameter optimization method for spread spectrum modulation that can further reduce noise spikes and effectively reduce the conducted emissions of inverters without changing the range of the spread spectrum modulation function is urgently needed in this field. Summary of the Invention

[0006] To address the above technical problems, this invention provides a parameter optimization method and system for spread spectrum modulation of microgrid interface inverters, aiming to reduce conducted emission noise caused by a large number of high-frequency power electronic devices in today's microgrid systems, and to provide a model basis for the design of subsequent EMI filters.

[0007] The technical solution of this invention is: a parameter optimization method for spread spectrum modulation of a microgrid interface inverter, comprising the following steps:

[0008] Step 1: Construct an electromagnetic interference model of a two-stage grid-connected inverter in simulation software. The front stage is a Boost converter circuit, and the back stage is a full-bridge inverter circuit. Select several key nodes and obtain the voltage and current waveforms of the key nodes.

[0009] Step 2: Based on the voltage and current waveforms of key nodes obtained in Step 1, first simplify the electromagnetic interference model of the two-stage grid-connected inverter, and then construct the common-mode circuit of the two-stage grid-connected inverter based on the simplified electromagnetic interference model of the two-stage grid-connected inverter, thereby determining the common-mode noise source of the circuit. The common-mode circuit is equivalent to an RLC series circuit, thereby obtaining the weighting coefficients of the common-mode noise sources of the front and rear stages, thereby determining the total common-mode noise source of the circuit and qualitatively predicting the noise distribution.

[0010] Step 3: Based on the estimated common-mode noise distribution, determine the spread spectrum modulation function to be used, and perform spread spectrum modulation on the Boost converter circuit and the full-bridge inverter circuit to make the noise spikes of the Boost converter circuit and the full-bridge inverter circuit staggered, thus completing the optimization.

[0011] Preferably, in step 1, the key nodes include: the anode of the diode and the drain of the four MOSFETs in the full-bridge inverter circuit.

[0012] Preferably, in step 2, the specific method for simplifying the electromagnetic interference model of the two-stage grid-connected inverter is as follows:

[0013] In a boost converter circuit, the MOSFET is equivalent to an equivalent voltage source whose voltage is equal to its drain-source voltage, and the freewheeling diode is equivalent to an equivalent current source whose current is equal to the current flowing through the diode.

[0014] In a full-bridge inverter circuit, a MOSFET is equivalent to a voltage source with the same waveform as its own operating voltage; a diode is equivalent to a current source with the same current as its own operating current.

[0015] By cascading the equivalent power supply models of the Boost circuit and the full-bridge inverter circuit, the electromagnetic interference model of the two-stage grid-connected inverter is finally obtained.

[0016] Preferably, in step 2, the specific method for constructing the common-mode circuit of the inverter is as follows:

[0017] Based on the superposition theorem, the equivalent current source of the diode in the Boost circuit is treated as a short circuit, and the equivalent current source of the freewheeling diode in the Boost circuit is treated as an open circuit, thus constructing the common-mode circuit of the inverter.

[0018] Based on the common-mode circuit of the inverter, the conducted emission noise path of the two-stage circuit is obtained. The source of common-mode excitation is then determined from this path, including: the common-mode noise source u of the Boost converter circuit. CM1 For switch S dc voltage u sDc The common-mode noise source u of the full-bridge inverter circuit CM2 For (u A +u B ) / 2, where u A and u B This is the voltage at the midpoint between the two bridge arms.

[0019] Preferably, in step 2, the common-mode circuit is equivalent to an RLC series circuit in the following way:

[0020] By applying the Thevenin equivalent to the branch containing the sampling resistor in the common-mode circuit, the RLC series circuit can be represented as a voltage source u. eq Series equivalent capacitance C eq In the form of, where the equivalent voltage u eq and capacitor C eq Calculate as follows:

[0021]

[0022] In the formula, C cm1 C is the capacitance to ground at the anode of the diode in a full-bridge inverter circuit. cm2 It is the capacitance to ground of the MOSFET drain on one branch of the full-bridge inverter circuit, C. cm3 It is the capacitance to ground of the MOSFET drain on another branch of the full-bridge inverter circuit.

[0023] Preferably, in step 3, the Boost converter circuit and the full-bridge inverter circuit are spread spectrum modulated using trigonometric functions. The trigonometric function formula is as follows:

[0024] Spread spectrum modulation function of Boost converter circuit:

[0025] ω dc (t)=ω sdc +Δω s cos(ω fdc t)

[0026] In the formula, ω dc (t) represents the angular frequency of the Boost converter circuit, ω sdc Δω is the center angular frequency of the frequency fluctuation in the Boost converter circuit. s To represent the frequency fluctuation range, take 10% of the switching frequency, ω fdc ω is the angular frequency of the frequency fluctuation of the Boost converter circuit, and t is time.

[0027] Spread spectrum modulation function of full-bridge inverter circuit:

[0028] ω ac (t)=ω sdc / 2+Δω s cos(ω fac t)

[0029] In the formula, ω ac (t) is the angular frequency of the full-bridge inverter circuit, ω fac The angular frequency of the frequency fluctuation in the full-bridge inverter circuit;

[0030] Spread spectrum modulation is performed to stagger the noise spikes of the Boost converter circuit and the full-bridge inverter circuit, thus achieving optimization.

[0031] Preferably, the method further includes step 4: by keeping the fluctuation period of the switching frequency consistent with the scanning period of the test receiver, noise spikes are reduced without changing the angular frequency range of the spread spectrum modulation function, and secondary optimization is performed.

[0032] Preferably, in step 4, the period of the spread spectrum modulation function is 0.1s.

[0033] A parameter optimization system for spread spectrum modulation of a microgrid interface inverter is also provided, comprising:

[0034] The simulation module is used to construct an electromagnetic interference model of a two-stage grid-connected inverter. The front stage is a Boost converter circuit, and the back stage is a full-bridge inverter circuit. Several key nodes are selected, and the voltage and current waveforms of the key nodes are obtained.

[0035] The equivalent transformation module is used to simplify the electromagnetic interference model of the two-stage grid-connected inverter based on the voltage and current waveforms of the key nodes. Then, based on the simplified electromagnetic interference model of the two-stage grid-connected inverter, the common-mode circuit of the two-stage grid-connected inverter is constructed to determine the common-mode noise source of the circuit. The common-mode circuit is equivalent to an RLC series circuit to obtain the weighting coefficients of the common-mode noise sources of the front and rear stages, thereby determining the total common-mode noise sources of the circuit and qualitatively predicting the noise distribution.

[0036] The calculation module is used to determine the spread spectrum modulation function to be used based on the estimated noise distribution, and to perform spread spectrum modulation on the Boost converter circuit and the full-bridge inverter circuit so that the noise spikes of the Boost converter circuit and the full-bridge inverter circuit are staggered, thus completing the optimization.

[0037] The beneficial effects of this invention are: Based on the voltage and current waveforms of key nodes, this invention obtains the voltage and current source equivalent operating state of a two-stage inverter, thereby obtaining the common-mode circuit of the conducted emission low-frequency inverter, estimating the noise distribution, determining the spread spectrum modulation function, and then obtaining the frequency domain distribution of the common-mode noise source through two-dimensional Fourier decomposition. Based on the frequency domain distribution characteristics, the noise spikes of the boost circuit and inverter circuit are staggered to reduce common-mode interference.

[0038] By matching the fluctuation period of the switching frequency with the scanning period of the test receiver, noise spikes can be further reduced without changing the range of the spread spectrum modulation function.

[0039] This invention can obtain spread spectrum modulation function parameters that can effectively reduce inverter conducted emissions without changing the fluctuation range of the spread spectrum modulation function. Attached Figure Description

[0040] Figure 1 This is a structural diagram of a typical two-stage inverter.

[0041] Figure 2 This is a simplified model diagram of a two-stage inverter.

[0042] Figure 3 This is the equivalent common-mode circuit diagram of a two-stage inverter.

[0043] Figure 4 This is the simplified equivalent circuit diagram of the common-mode circuit.

[0044] Figure 5 This is a comparison chart of the spread spectrum modulation function and the inverter output fundamental voltage.

[0045] Figure 6 This is a diagram showing the voltage distribution of the switch. Figure 6 middle Figure 6 (a) is a diagram showing the voltage distribution of the Boost switch. Figure 6 (b) is a diagram showing the voltage distribution at the midpoint of the bridge arm.

[0046] Figure 7 This is a comparison diagram of conducted emissions between the original circuit and the power supply equivalent circuit.

[0047] Figure 8 In the simulation experiment, ω f Graph of conducted emission test results under changing conditions

[0048] Figure 9 In the simulation experiment, ω f The diagram shows the conducted emission test results when π = 2π.

[0049] Figures 7 to 9 In the diagram, the vertical axis represents EMI (Total Interference), CM (Common Mode Interference), and DM (Differential Mode Interference). Detailed Implementation

[0050] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] A parameter optimization method for spread spectrum modulation of a microgrid interface inverter includes the following steps:

[0052] Step 1: See Figure 1 An electromagnetic interference model of a two-stage grid-connected inverter was constructed on simulation software. The front stage was a Boost converter circuit and the back stage was a full-bridge inverter circuit. The parasitic capacitance to ground of the system mainly came from the drain of the MOSFET near the heat sink. Several key nodes were selected and the voltage and current waveforms of the key nodes were obtained.

[0053] In this embodiment, the key nodes are the anode of diode D1 in the full-bridge inverter circuit and the drains of the four MOSFETs S1 to S4.

[0054] Step 2: Based on the voltage and current waveforms of key nodes obtained in Step 1, first simplify the electromagnetic interference model of the two-stage grid-connected inverter, and then construct the common-mode circuit of the two-stage grid-connected inverter based on the simplified electromagnetic interference model of the two-stage grid-connected inverter, thereby determining the common-mode noise source of the circuit. The common-mode circuit is equivalent to an RLC series circuit, thereby obtaining the weighting coefficients of the common-mode noise sources of the front and rear stages, thereby determining the total common-mode noise source of the circuit and qualitatively predicting the noise distribution.

[0055] See Figure 2 The specific way to simplify the electromagnetic interference model of a two-stage grid-connected inverter is as follows: In the Boost circuit, the MOSFET is equivalent to an equivalent voltage source with a voltage equal to its drain-source voltage, and the freewheeling diode is equivalent to an equivalent current source with a current equal to the diode current.

[0056] The freewheeling diode in the Boost circuit has no parallel MOSFET, but according to the superposition theorem, when the equivalent current source of the freewheeling diode acts alone, the voltage source is zero. Therefore, the current source is controlled by S. dc and u odc A short circuit has no effect on the current flowing into the 50Ω circuit, so it can be considered non-existent and therefore can be considered an open circuit.

[0057] In a full-bridge inverter circuit, a MOSFET is equivalent to a voltage source with the same waveform as its operating voltage; a diode is equivalent to a current source with the same current as its operating current. The equivalent current source current of the diode is the same as the current flowing through the diode in the original full-bridge circuit, ensuring that its output characteristics are equivalent. Since it is connected in parallel with the MOSFET, its output characteristics are in the form of a voltage source. In a full-bridge circuit, when the MOSFET and diode are connected in parallel, the equivalent current source model of the diode can be ignored. In a Boost circuit, the freewheeling diode exists alone, and the current of the equivalent current source is a trapezoidal wave.

[0058] By cascading the equivalent power supply models of the Boost circuit and the full-bridge inverter circuit, the electromagnetic interference (EMI) model of the two-stage grid-connected inverter is finally obtained. Under ideal switching conditions, the cascaded equivalent power supply models of the front-end Boost circuit and the rear-end full-bridge circuit can be approximated as the EMI model of the two-stage grid-connected inverter.

[0059] At this point, the noise flowing into the sampling resistor and the noise of the original circuit are approximately equivalent under the condition that the DC bus capacitance is large and the switching waveform is relatively ideal. That is, the equivalent circuit of the voltage and current source in the low-frequency band of conducted emission is the same as the original circuit.

[0060] The specific method for constructing the common-mode circuit of the inverter is as follows:

[0061] Using the superposition theorem, in the electromagnetic interference model of the two-stage grid-connected inverter, the equivalent current source of the Boost circuit diode is equivalent to a short circuit. When the equivalent current source of diode D1 in the full-bridge inverter circuit acts alone, it is directly short-circuited. Therefore, diode D1 has no effect on the common-mode voltage, and diode D1 is equivalent to an open circuit; thus, we obtain... Figure 3 The inverter equivalent common-mode circuit is shown.

[0062] Based on the common-mode circuit of the inverter, the conducted emission noise path of the two-stage circuit is obtained. From the conducted emission noise path of the two-stage circuit, the source of the common-mode excitation can be obtained: the common-mode noise source u of the Boost converter circuit. CM1 For switch S dc voltage u sDC The common-mode noise source u of the full-bridge inverter circuit CM2 For (u A +u B ) / 2, where u A and u B The voltage at the midpoint between the two bridge arms;

[0063] It should be noted that: because the DC bus capacitance is relatively large and has a strong ability to absorb high-frequency differential-mode noise, the differential-mode noise of the DC-DC link flows into the capacitor (i.e., Figure 2 C in dc In this invention, the differential mode measurement of the inverter stage is not affected, so the noise generated by the differential mode circuit is not considered, and only the common mode circuit needs to be considered.

[0064] The common-mode circuit is equivalent to an RLC series circuit. The RLC series circuit directly affects the final noise reception result. However, since the circuit has been simplified to a voltage source series impedance structure, the magnitude of the common-mode noise source is directly positively correlated with the measured noise (i.e., the current of the series circuit). The influence of line impedance on noise is mainly reflected in the weighting coefficient of parasitic capacitance on the two common-mode noise sources in the circuit, as shown in equation (1) below.

[0065] u sDC and u CM The common-mode current is determined by the common-mode circuit, which, excluding the branch containing the sampling resistor, represents the rest of the common-mode circuit as a voltage source u using Thevenin's theorem. eq Series equivalent capacitance C eq In the form of equivalent voltage u eq and equivalent capacitance C eq Calculate as follows:

[0066]

[0067] Where C cm1 It is the capacitance at the anode of the diode in the full-bridge inverter circuit, C. cm2It is the capacitance of the MOSFET drain on one branch of the full-bridge inverter circuit, C. cm3 It is the capacitance of the MOSFET drain on the other branch of the full-bridge inverter circuit;

[0068] If we consider the parasitic capacitance of the inductor, then C eq Based on the above formula, the parallel inductor is equivalent to the series capacitor.

[0069] get Figure 4 The simplified common-mode RLC series circuit shown illustrates the predicted noise distribution. The resonant point is typically much larger than the harmonics of the lower switching frequencies, and the excitation source is u. CM1 and u CM2 In a linear combination, the impedance depends on the capacitance in the low-frequency range and decreases as the frequency increases.

[0070] Step 3: Based on the estimated noise distribution, determine the spread spectrum modulation function to be used. Obtain the frequency domain distribution of the common-mode noise source through two-dimensional Fourier decomposition. Determine the center frequency of the spread spectrum modulation function based on the frequency domain distribution characteristics. Select the spread spectrum range to ensure that the common-mode noise of the preceding and following stages does not overlap. Then, substitute the spread spectrum modulation function to make the noise spikes of the Boost converter circuit and the full-bridge inverter circuit staggered, thus completing the optimization.

[0071] join Figure 5 In this embodiment, trigonometric functions are used for spread spectrum modulation, meaning the switching frequency varies in the time domain in the form of a sine function. The trigonometric function is:

[0072] ω(t)=ω s +Δω s cos(ω f t) (2)

[0073] In the formula, ω(t) represents the angular frequency, ω s Δω represents the center angular frequency of the frequency fluctuation. s ω represents the range of frequency fluctuations. f The angular frequency represents the frequency fluctuation, and t represents time;

[0074] The frequency domain distribution of the common-mode noise source is obtained through two-dimensional Fourier decomposition. Specific steps include:

[0075] First, calculate the switching voltage u of the Boost converter circuit. sDC Midpoint voltage u of bridge arm A and the voltage u at the negative terminal of the DC bus B The formula is as follows:

[0076]

[0077] In the formula, u odcThis is the DC bus output voltage, sign() represents the sign function, D is the high-level duty cycle of the Boost converter, and u c1 A frequency-converting triangular wave between 0 and 1, with a period of 2π, θ dc Indicates the carrier phase of the Boost converter; u m For the modulation amplitude, u c2 (x) represents a frequency-converting triangular wave between -1 and 1, with a period of 2π, and θ ac The carrier phase of the full-bridge inverter circuit is represented by the time-domain relationship between phase and frequency as follows: Figure 5 ω0 is the angular frequency of the modulating wave, and t is the time with the modulating wave phase as the reference zero phase; ω0 is the angular frequency of the modulating wave.

[0078] With both carrier and modulation phases set as independent variables, the Boost switching voltage and the midpoint voltage of the bridge arm are expressed as functions of the carrier phase x and the modulation phase y, respectively. The Boost switching voltage is as follows: Figure 6 As shown in (a), the voltage at the midpoint of the bridge arm is as follows: Figure 6 As shown in (b), the output inside the boundary is high, and the output outside the boundary is zero.

[0079] Next, perform a two-dimensional Fourier decomposition of the switching voltage function with respect to x and y. The specific calculation formula is as follows:

[0080]

[0081] In the formula u dc For DC power supply voltage, u m M represents the modulation amplitude, m represents the modulation ratio, m represents the harmonic order, indicating that the harmonics are located near the center of each switching frequency, and n represents the sideband order, indicating that harmonics also exist at the switching frequency and its vicinity. n This represents the first nth-order Bessel function, where n represents the sideband order. For example, n is 1 for harmonics near the first switching frequency.

[0082] The carrier phase in the above formula is obtained by integrating the spread spectrum modulation function, and the specific calculation formula is as follows:

[0083]

[0084] In equation (5), f sdc and f sac Δf represents the average value of the switching frequency fluctuations of the Boost circuit and the inverter, respectively. dc and Δf ac The fluctuation range of the switching frequency of the Boost circuit and the inverter, respectively, ω fdc The angular frequency ω represents the ripple of the Boost circuit. fac This represents the angular frequency of the inverter circuit fluctuations.

[0085] Based on the composite trigonometric function of equation (4), trigonometric function decomposition is performed using Bessel series, and the DC component is ignored to obtain the common-mode noise source u of the Boost converter circuit. CM1 The specific calculation formula is as follows:

[0086]

[0087] In equation (6), ω sdc ω is the center angular frequency of the frequency fluctuation in the DC spread spectrum modulation function. fdc To obtain twice the power frequency, Δω dc represents the spread spectrum range of the Boost circuit, which does not overlap with the common-mode noise, and k represents the order of the Bessel decomposition in equation (5).

[0088] Next, calculate the common-mode noise source u of the full-bridge inverter circuit. CM2 The specific calculation formula is as follows:

[0089]

[0090] Where ω sac The center angular frequency, ω, of the frequency fluctuation in the AC spread spectrum modulation function. fac Take twice the power frequency, Δω ac This represents the spread spectrum range of the full-bridge circuit, which does not overlap with the common-mode noise, and m is an odd number while n is an even number.

[0091] Based on equations (6) and (7), u CM1 The spectrum is mainly distributed in mω sdc +kω fdc in, u CM2 The spectrum is mainly distributed in mω sac +nω0+kω fac Furthermore, m can only be odd numbers;

[0092] The frequency fluctuation range is taken as 10% of the switching frequency. It is known that the common-mode voltage of the inverter circuit is only distributed in the odd-order sidebands, while the common-mode voltage of the Boost circuit is distributed in all sidebands, and u in the common-mode loop... CM1 The generated common-mode current should be less than u CM2 u CM1 The effect on noise spikes is achieved through interaction with u CM2 Superposition, therefore take ω sdc =2ω sac ,Right now

[0093] Spread spectrum modulation function for the Boost converter circuit:

[0094] ω dc (t)=ω sdc +Δω s cos(ωfdc t) (8)

[0095] In the formula, ω dc (t) represents the angular frequency of the Boost converter circuit, ω sdc Δω is the center angular frequency of the frequency fluctuation in the Boost converter circuit. s To represent the frequency fluctuation range, take 10% of the switching frequency, ω fdc ω is the angular frequency of the frequency fluctuation of the Boost converter circuit, and t is time.

[0096] Spread spectrum modulation function of full-bridge inverter circuit:

[0097] ω ac (t)=ω sdc / 2+Δω s cos(ω fac t) (9)

[0098] In the formula, ω ac (t) is the angular frequency of the full-bridge inverter circuit, ω fac The angular frequency of the frequency fluctuation in the full-bridge inverter circuit;

[0099] Perform spread spectrum modulation to complete the optimization.

[0100] From the properties of the Bessel function, we know that J k (x)(Jk(x) is the first-order k-th Bessel function of the first kind, x is the independent variable. This function is used to further decompose the harmonic expression when the spreading function is a nonlinear function to obtain the Fourier expansion of the spreading modulation. k represents the order of the series decomposition of equation (4) when θdc and θac are sine functions.) When x is large and k is small, the amplitude is close to Therefore, the larger Δω is, the more ω f The smaller the value, the smaller the amplitude of each sideband noise. However, two-dimensional Fourier decomposition represents the frequency domain characteristics of a signal over an infinite time period. In actual measurements, EMI test receivers perform spectrum analysis by windowing a short-time signal, so it is not ω. f The smaller the value, the lower the noise intensity measured. For example, when the real-time time-domain scan time of the test receiver is 100ms, using a 1Hz spread spectrum modulation function frequency will obviously reduce the equivalent spread spectrum range.

[0101] By aligning the fluctuation period of the switching frequency with the scanning period of the test receiver, noise spikes are further reduced without altering the range of the spread spectrum modulation function, thus achieving secondary optimization. In practical applications, the function range refers to the fluctuation range of the switching frequency, which is the difference between the maximum and minimum values ​​of the function (i.e., the switching frequency). This range is limited by the performance of the switching devices, system thermal management, and grid current harmonics requirements, and typically needs to be confined to a certain range. This invention further reduces the conducted emission noise of the system by adjusting the fluctuation period of the switching frequency without changing its fluctuation range.

[0102] In this embodiment, 2π / ω is taken. f Consistent with the real-time time-domain scan period of the test receiver, i.e., ω fdc and ω fac It should be consistent with the real-time scanning period. The scanning time of the test receiver is longer than its corresponding period. Taking into account the scanning time of most test receivers in the conducted transmission range, the period of the spread spectrum modulation function is taken as 0.1s.

[0103] A parameter optimization system for spread spectrum modulation of a microgrid interface inverter is also provided, comprising:

[0104] Simulation module: Used to construct an electromagnetic interference model of a two-stage grid-connected inverter. The front stage is a Boost converter circuit, and the back stage is a full-bridge inverter circuit. Several key nodes are selected to obtain the voltage and current waveforms of the key nodes.

[0105] Equivalent transformation module: Based on the voltage and current waveforms of key nodes, it first simplifies the electromagnetic interference model of the two-stage grid-connected inverter, and then constructs the common-mode circuit of the two-stage grid-connected inverter based on the simplified electromagnetic interference model, thereby determining the common-mode noise source of the circuit. The common-mode circuit is equivalent to an RLC series circuit, thereby obtaining the weighting coefficients of the common-mode noise sources of the front and rear stages, thus determining the total common-mode noise sources of the circuit and qualitatively predicting the noise distribution.

[0106] Calculation module: Based on the estimated noise distribution, it determines the spread spectrum modulation function to be used, and performs spread spectrum modulation on the Boost converter circuit and the full-bridge inverter circuit to stagger the noise spikes of the Boost converter circuit and the full-bridge inverter circuit, thus completing the optimization.

[0107] Example

[0108] Building in PSIM Figure 1 The two-stage grid-connected inverter shown and Figure 3 The common-mode circuit shown has a 100pF drain-to-ground capacitance for the Boost circuit switch and a 50pF drain-to-ground capacitance for the inverter circuit switch. It operates at a constant frequency f. sdc =150kHz, f sacAt 100kHz, the simulation results are as follows: Figure 7 As shown.

[0109] Simulation results show that the voltage values ​​of the original circuit and the power supply equivalent circuit are close at most noise spikes. In the actual circuit, according to power conservation, the DC voltage source has fluctuations of twice the power frequency. If the bus capacitor voltage is constant, the switching transistor voltage will have a DC component and higher harmonics. Therefore, the bus capacitor will also necessarily have a component of twice the power frequency. Thus, the conducted emissions of the power supply equivalent circuit are less than those of the actual circuit in some regions. However, since this value is much smaller than the voltage change caused by switching action, its impact on noise spikes can be ignored. Therefore, the equivalent power supply model derived in this invention can accurately characterize the conducted emissions measurement of a two-stage inverter in the low-frequency range of conducted emissions, and can provide a model reference for the design of EMI filters.

[0110] The time-domain scan time is set to 0.1s. When f sdc =200kHz, f sac Simulation results of conducted emission tests at 200kHz, with fixed-frequency modulation, and spread spectrum modulation functions having periods of 0.1s and 0.01s, are as follows: Figure 8 As shown, spread spectrum modulation achieved a 20dB attenuation compared to fixed-frequency modulation, and a spread spectrum modulation function period of 0.1s achieved a 5dB attenuation compared to 0.01s. Based on the analysis of noise excitation and loops, increasing the spread spectrum modulation function period further reduced the conducted emission measurements while maintaining the same spread spectrum range. Since the spread spectrum range fluctuation is only ±5%, its impact on system loss and distortion is minimal.

[0111] Further reducing the spread spectrum modulation function frequency to 1Hz, the frequency fluctuation period is much larger than the scanning period, as shown in the simulation results. Figure 9 As shown, at this time although ω f While the effective spread spectrum range is reduced, since time-domain scanning can only analyze signals within 0.1s, the conducted emission measurement actually increases, making the effect less than the former. This demonstrates that the parameter optimization method provided by this invention is effective.

[0112] This invention is not limited to the above embodiments. Based on the technical solutions disclosed in this invention, those skilled in the art can make some substitutions and modifications to some of the technical features without creative effort, and all such substitutions and modifications are within the protection scope of this invention.

Claims

1. A parameter optimization method for spread spectrum modulation of a microgrid interface inverter, characterized in that, Includes the following steps: Step 1: Construct an electromagnetic interference model of a two-stage grid-connected inverter in simulation software. The front stage is a Boost converter circuit, and the back stage is a full-bridge inverter circuit. Select several key nodes and obtain the voltage and current waveforms of the key nodes. Step 2: Based on the voltage and current waveforms of key nodes obtained in Step 1, first simplify the electromagnetic interference model of the two-stage grid-connected inverter, and then construct the common-mode circuit of the two-stage grid-connected inverter based on the simplified electromagnetic interference model of the two-stage grid-connected inverter, thereby determining the common-mode noise source of the circuit. The common-mode circuit is equivalent to an RLC series circuit, thereby obtaining the weighting coefficients of the common-mode noise sources of the front and rear stages, thereby determining the total common-mode noise source of the circuit and qualitatively predicting the noise distribution. Step 3: Based on the estimated common-mode noise distribution, determine the spread spectrum modulation function to be used, and perform spread spectrum modulation on the Boost converter circuit and the full-bridge inverter circuit to make the noise spikes of the Boost converter circuit and the full-bridge inverter circuit staggered, thus completing the optimization.

2. The parameter optimization method for spread spectrum modulation of a microgrid interface inverter according to claim 1, characterized in that, In step 1, the key nodes include: the anode of the diode and the drain of the four MOSFETs in the full-bridge inverter circuit.

3. The parameter optimization method for spread spectrum modulation of a microgrid interface inverter according to claim 1, characterized in that, In step 2, the specific method for simplifying the electromagnetic interference model of the two-stage grid-connected inverter is as follows: In a boost converter circuit, the MOSFET is equivalent to an equivalent voltage source whose voltage is equal to its drain-source voltage, and the freewheeling diode is equivalent to an equivalent current source whose current is equal to the current flowing through the diode. In a full-bridge inverter circuit, a MOSFET is equivalent to a voltage source with the same waveform as its own operating voltage; a diode is equivalent to a current source with the same current as its own operating current. By cascading the equivalent power supply models of the Boost circuit and the full-bridge inverter circuit, the electromagnetic interference model of the two-stage grid-connected inverter is finally obtained.

4. The parameter optimization method for spread spectrum modulation of a microgrid interface inverter according to claim 3, characterized in that, In step 2, the specific method for constructing the common-mode circuit of the inverter is as follows: Based on the superposition theorem, the equivalent current source of the diode in the Boost circuit is treated as a short circuit, and the equivalent current source of the freewheeling diode in the Boost circuit is treated as an open circuit, thus constructing the common-mode circuit of the inverter. Based on the common-mode circuit of the inverter, the conducted emission noise path of the two-stage circuit is obtained. The source of common-mode excitation is then determined from this path, including: the common-mode noise source u of the Boost converter circuit. CM1 For switch S dc voltage u sDc The common-mode noise source u of the full-bridge inverter circuit CM2 For (u A +u B ) / 2, where u A and u B This is the voltage at the midpoint between the two bridge arms.

5. The parameter optimization method for spread spectrum modulation of a microgrid interface inverter according to claim 4, characterized in that, In step 2, the common-mode circuit is equivalent to an RLC series circuit in the following way: By applying the Thevenin equivalent to the branch containing the sampling resistor in the common-mode circuit, the RLC series circuit can be represented as a voltage source u. eq Series equivalent capacitance C eq In the form of, where the equivalent voltage u eq and capacitor C eq Calculate as follows: In the formula, C cm1 C is the capacitance to ground at the anode of the diode in a full-bridge inverter circuit. cm2 It is the capacitance to ground of the MOSFET drain on one branch of the full-bridge inverter circuit, C. cm3 It is the capacitance to ground of the MOSFET drain on another branch of the full-bridge inverter circuit.

6. The parameter optimization method for spread spectrum modulation of a microgrid interface inverter according to claim 1, characterized in that, In step 3, trigonometric functions are used to perform spread spectrum modulation on the Boost converter circuit and the full-bridge inverter circuit. The trigonometric function formula is as follows: Spread spectrum modulation function of Boost converter circuit: oh dc (t)=ω sdc +See s cos(ω fdc t) In the formula, ω dc (t) represents the angular frequency of the Boost converter circuit, ω sdc Δω is the center angular frequency of the frequency fluctuation in the Boost converter circuit. s To represent the frequency fluctuation range, take 10% of the switching frequency, ω fdc ω is the angular frequency of the frequency fluctuation of the Boost converter circuit, and t is time. Spread spectrum modulation function of full-bridge inverter circuit: oh ac (t)=ω sdc / 2+See s cos(ω fac t) In the formula, ω ac (t) is the angular frequency of the full-bridge inverter circuit, ω fac The angular frequency of the frequency fluctuation in the full-bridge inverter circuit; Spread spectrum modulation is performed to stagger the noise spikes of the Boost converter circuit and the full-bridge inverter circuit, thus achieving optimization.

7. The parameter optimization method for spread spectrum modulation of a microgrid interface inverter according to claim 1, characterized in that, The process also includes step 4: by keeping the fluctuation period of the switching frequency consistent with the scanning period of the test receiver, noise spikes are reduced without changing the angular frequency range of the spread spectrum modulation function, and secondary optimization is performed.

8. The parameter optimization method for spread spectrum modulation of a microgrid interface inverter according to claim 7, characterized in that, In step 4, the period of the spread spectrum modulation function is 0.1s.

9. A parameter optimization system for spread spectrum modulation of a microgrid interface inverter, characterized in that, include: The simulation module is used to construct an electromagnetic interference model of a two-stage grid-connected inverter. The front stage is a Boost converter circuit, and the back stage is a full-bridge inverter circuit. Several key nodes are selected, and the voltage and current waveforms of the key nodes are obtained. The equivalent transformation module is used to simplify the electromagnetic interference model of the two-stage grid-connected inverter based on the voltage and current waveforms of the key nodes. Then, based on the simplified electromagnetic interference model of the two-stage grid-connected inverter, the common-mode circuit of the two-stage grid-connected inverter is constructed to determine the common-mode noise source of the circuit. The common-mode circuit is equivalent to an RLC series circuit to obtain the weighting coefficients of the common-mode noise sources of the front and rear stages, thereby determining the total common-mode noise sources of the circuit and qualitatively predicting the noise distribution. The calculation module is used to determine the spread spectrum modulation function to be used based on the estimated noise distribution, and to perform spread spectrum modulation on the Boost converter circuit and the full-bridge inverter circuit so that the noise spikes of the Boost converter circuit and the full-bridge inverter circuit are staggered, thus completing the optimization.

Images