A method for calculating and optimizing overall power loss of a three-level converter

Abstract

The application provides a three-level converter overall power loss calculation and optimization method, and belongs to the technical field of power electronic converters.The application is aimed at the problem that the traditional three-level converter ignores the filter inductance current in loss evaluation, and is based on accurate modeling of the amplitude of the filter inductance current ripple component, and models the overall power loss of the three-level power electronic converter as a mathematical function about the switching frequency, which is composed of the conduction loss power and switching loss power of the semiconductor power device, the filter inductance loss power and the bus capacitor loss power.The application can be used not only for converter loss evaluation, but also for guiding the switching frequency optimization, and further improves the operation efficiency of the converter.

Description

Technical Field

[0001] This invention belongs to the field of power electronic converter technology, specifically relating to a method for calculating and optimizing the overall power loss of a three-level converter. Background Technology

[0002] As a core component in power conversion between renewable energy generation and the power transmission and distribution network, the power loss of power electronic converters is crucial, as reducing power loss directly equates to reducing carbon emissions. With the annual increase in renewable energy installed capacity, even a small reduction in the power loss of power electronic converters can bring significant economic and environmental benefits throughout their entire lifecycle. Furthermore, under the same heat dissipation conditions, reduced power loss in power electronic converters also means less heat accumulation inside the unit, which helps extend the lifespan of power components and promotes long-term safe and stable operation of the equipment.

[0003] In power conversion applications, three-level converter topologies are generally superior to traditional two-level topologies due to their lower electromagnetic interference, lower operating power losses, and lower output voltage and current harmonics. These characteristics enable three-level converters to achieve higher power density and lower cost in high-power applications, leading to their increasingly widespread adoption.

[0004] Currently, power loss analysis of three-level converters largely focuses on the semiconductor devices themselves, with existing research using simulation analysis or mathematical modeling to obtain the conduction and switching losses of these devices. However, the losses of the filter inductor in three-level converters are often neglected, resulting in a lack of a comprehensive power loss calculation and optimization method from the perspective of the entire converter. Furthermore, existing methods generally suffer from the following shortcomings: a lack of effective methods for predicting filter inductor current ripple, and failure to incorporate its impact on overall loss into a unified modeling framework; insufficient consideration of the nonlinear variation of the filter inductor value with current; and a lack of comprehensive optimization of key parameters such as switching frequency from the perspective of overall system power loss. Therefore, there is a lack of a practical overall power loss model for three-level converters that can accurately model the filter inductor current ripple, precisely characterize semiconductor device losses and filter inductor losses, and guide the selection of switching frequency. Summary of the Invention

[0005] To address existing challenges, this invention proposes a practical overall loss model for power electronic converters to guide switching frequency selection. Based on accurate modeling of the filter inductor current ripple, it precisely characterizes semiconductor device losses and filter inductor losses, obtaining a mathematical function of the overall loss of a three-level power electronic converter with respect to the switching frequency. Thus, with the goal of minimizing the overall loss of the power electronic converter, the optimal switching frequency is selected, effectively improving the operating efficiency of the power electronic converter.

[0006] To achieve the above technical objectives, this invention proposes a method for calculating and optimizing the overall power loss of a three-level converter, including:

[0007] S1. Calculate the amplitude of the ripple component of the filter inductor current based on the switching frequency of the three-level converter and the inductance value, voltage across the two ends, and instantaneous current of the filter inductor of the three-level converter.

[0008] S2, calculate the conduction loss power of each semiconductor power device based on the instantaneous current and the conduction voltage drop of the semiconductor power device; add up the conduction loss power of all semiconductor power devices in the three-level converter to obtain the total conduction loss power;

[0009] The switching loss power of each semiconductor power device is calculated by combining the amplitude of the ripple component and the switching frequency; the switching loss power of all semiconductor power devices in the three-level converter is added together to obtain the total switching loss power.

[0010] S3, calculate the filter inductor loss power based on the switching frequency and the amplitude of the ripple component; calculate the bus capacitor loss power based on the voltage fluctuation across the three-level converter bus capacitor, the voltage fluctuation angular frequency, the capacitance value, and the loss angle.

[0011] S4. The total conduction loss power, total switching loss power, filter inductor loss power, and bus capacitor loss power are added together to obtain the overall power loss of the three-level converter; the overall power loss is adjusted by changing the switching frequency until the desired overall power loss optimization target is achieved.

[0012] Further, S1 includes:

[0013] S101, fits the nonlinear relationship between the inductance value of the filter inductor and the instantaneous current into a cubic function;

[0014] S102, Construct the modulation wave function; Based on the switching frequency and the modulation wave function, calculate the duration of the current rise in the ripple component of the filter inductor current.

[0015] S103, the amplitude of the ripple component of the filter inductor current is calculated based on the cubic function, the duration, and the voltage across the filter inductor.

[0016] Furthermore, in S102, the duration of the current rise is obtained by dividing the modulation wave function at the current moment by the switching frequency.

[0017] Furthermore, in S103, the specific calculation formula for the amplitude of the ripple component of the filter inductor current is as follows:

[0018] ;

[0019] in The ripple component amplitude of the filter inductor current. For switching frequency, , The voltage across the filter inductor is... The inductance value of the filter inductor. This represents the duration of the rising portion of the current in the ripple component of the filter inductor current. This refers to the DC bus voltage of the converter. This is the modulation wave function.

[0020] Furthermore, in S2, the specific calculation method for the conduction loss power is as follows: within one fundamental frequency period, the instantaneous current at the current moment is multiplied by the conduction voltage drop to obtain the instantaneous conduction loss power, and then the instantaneous conduction loss power is multiplied by the modulation wave function at the current moment to obtain the calculation result;

[0021] After iterating through all times within a fundamental period and obtaining the calculation result for each time, the calculation results for all times are set to 0. Integrate the result to obtain the total value of the integral; divide the total value of the integral by... The conduction loss power is obtained.

[0022] Furthermore, the switching loss power includes the turn-on loss of the semiconductor power device and the turn-off loss of the semiconductor power device.

[0023] Furthermore, in S2, the switching loss power is specifically calculated as follows:

[0024] ;

[0025] in Let i be the switching power loss and i be the instantaneous current. This represents the amplitude of the ripple component of the filter inductor current. Where t is the switching frequency, and t is time. This is the reverse voltage drop that the power device withstands when it is turned off. For turn-off losses, T0 is one fundamental frequency period. This is for activation losses.

[0026] Furthermore, in S3, the power loss of the bus capacitor is specifically calculated by multiplying the square of the voltage fluctuation across the converter bus capacitor within one fundamental cycle by the voltage fluctuation angular frequency, then by the capacitance value of the converter bus capacitor, and finally by the tangent of the converter bus capacitor loss angle.

[0027] Furthermore, in S3, the power loss of the filter inductor is specifically calculated as follows:

[0028] S301, the change in magnetic flux density at the current moment is calculated based on electromagnetic theory and the amplitude of the ripple component;

[0029] S302, according to the Steinmetz equation, the current unit volume core loss power is calculated using the switching frequency and the change in magnetic flux density;

[0030] S303: The core loss power per unit volume is added up at all times within a fundamental period, multiplied by the core volume of the filter inductor, and then divided by the carrier ratio to obtain the filter inductor loss power; the carrier ratio is obtained by multiplying the switching frequency and the fundamental period.

[0031] Furthermore, in S4, the overall power loss optimization objective is to minimize the overall power loss of the three-level converter.

[0032] Based on the above technical solution, the present invention has the following beneficial technical effects:

[0033] (1) This invention improves the operating efficiency of power electronic converters by accurately modeling the mathematical function of the overall loss of a three-level power electronic converter with respect to the switching frequency and selecting the optimal switching frequency with the goal of minimizing the overall loss.

[0034] (2) This invention takes into account the nonlinear variation characteristics of the filter inductor value with the current and accurately models the current ripple component of the filter inductor, providing a basis for calculating the switching loss of semiconductor power devices and the loss of the filter inductor.

[0035] (3) This invention incorporates the precisely modeled filter inductor current ripple component into the switching loss model of semiconductor power devices, thereby achieving higher precision power device switching loss modeling and obtaining a mathematical function between switching loss and switching frequency.

[0036] (4) This invention incorporates the precisely modeled filter inductor current ripple component into the calculation of filter inductor loss in a three-level power electronic converter. Based on the accurate prediction of the filter inductor current ripple component, it accurately models the mathematical function between filter inductor loss and switching frequency.

[0037] (5) The accurate model of the overall loss of the three-level power electronic converter established in this invention can not only be used for converter loss assessment, thereby guiding the heat dissipation design of the converter, but also further improve the operating efficiency of the converter by optimizing the switching frequency. Furthermore, based on the optimized switching frequency, iterative design of components sensitive to the switching frequency, such as filter inductors, is carried out. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of a three-phase SiC-Si heterostructure ANPC three-level topology;

[0039] Figure 2 A schematic diagram of the modulation wave for injecting the third harmonic and the PWM2 modulation method;

[0040] Figure 3 This is a schematic diagram illustrating the nonlinear variation of the filter inductance value with current.

[0041] Figure 4 A schematic diagram showing the current ripple prediction results at different switching frequencies;

[0042] Figure 5 This is a schematic diagram showing the relationship between switching losses and current of a power device obtained through a double-pulse test.

[0043] Figure 6 This diagram illustrates the filter inductor loss, the power device switching loss, and their sum.

[0044] Figure 7 This is a schematic diagram showing the predicted and test results of the current ripple component at a switching frequency of 25kHz.

[0045] Figure 8 This is a schematic diagram showing the predicted and test results of the current ripple component at a switching frequency of 20kHz.

[0046] Figure 9 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0047] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below. Technical features in various embodiments of the present invention can be combined accordingly without mutual conflict.

[0048] like Figure 9 As shown, this invention models the overall loss of a three-level power electronic converter as a mathematical function of the switching frequency. The overall loss consists of four parts: the conduction power and switching power of the semiconductor power devices, the filter inductor loss, and the bus capacitor loss.

[0049] I. Modeling the current ripple component of the filter inductor

[0050] 1) As the current flowing through the filter inductor increases, its inductance will decrease, leading to a decline in filtering performance. Therefore, this invention fits the nonlinear variation characteristic of the filter inductance value with current into a cubic function characteristic curve, as shown in the following formula:

[0051]

[0052] Where L(i) is the inductance value of the filter inductor, and i is the instantaneous current flowing through the filter inductor. , , and The four coefficients of the cubic function characteristic curve can be obtained from the filter inductor's datasheet or through actual experiments.

[0053] Since the time of a switching cycle is extremely short, the fundamental component of the filter inductor current is approximately constant. It can be approximated as the instantaneous current i in the cubic function characteristic curve and used to estimate the inductance value L(i) of the filter inductor.

[0054] Let the fundamental component of the filter inductor current be... The following relation can be established:

[0055]

[0056] in This represents the peak current of the fundamental component. Let α be the fundamental angular frequency, α be the power factor angle, and t be the time.

[0057] 2) Assume the modulation scheme for grid-connected operation of the three-level converter is as follows: With grid-side peak voltage DC bus voltage The following relation can be established:

[0058]

[0059] Let the modulation waveform of the converter controller be... The voltage of the converter filter capacitor is The following relation can be established:

[0060]

[0061] Where D() is the normalized or standardized instantaneous modulation waveform, and the peak range is usually between [-1, 1].

[0062] 3) Model the ripple component of the filter inductor current. First, in the modulation function... In the context of D(t)≥0, the converter's machine-side phase voltage is between 0 and +V. dc Switching between / 2; when D(t) < 0, the converter machine-side phase voltage is between -V dcSwitching between / 2 and 0, since the modulation wave function is centrally symmetrical in both positive and negative half-cycles, this invention only considers the case where D(t)≥0, in which case the converter's machine-side phase voltage is between 0 and +V. dc Switch between / 2 and / 2.

[0063] Furthermore, within one switching cycle, the average value of the ripple component of the filter inductor current is 0. Therefore, the rising and falling portions of the ripple component are equal in magnitude. Only the rising portion of the current needs to be considered, i.e., when the converter's machine-side phase voltage is at +V. dc / 2 case.

[0064] Based on the above analysis, the voltage across the filter inductor at this time... The converter's machine-side phase voltage minus the converter's filter capacitor voltage. The voltage value after that can be used to establish the following relationship:

[0065]

[0066] The duration of the rising portion of the ripple component of the filter inductor current is denoted as... f s For the switching frequency, the following relationship can be established:

[0067]

[0068] set up The following relationship can be established for the amplitude of the ripple component of the filter inductor current:

[0069]

[0070] II. Establishing the conduction loss power function of semiconductor power devices

[0071] 1) The on-state voltage drop of semiconductor power devices Modeled as knee voltage The sum of resistive and resistive voltage drops can be used to establish the following relationship:

[0072]

[0073] in For equivalent impedance, This refers to the instantaneous current flowing through the semiconductor power device.

[0074] and This information can be obtained from the datasheet of the semiconductor power device or through actual testing. It's important to note that when the semiconductor power device is a MOSFET, the corresponding knee voltage V... CE0 It is zero.

[0075] 2) Since the conduction loss of a semiconductor power device varies continuously within one fundamental frequency cycle, the conduction loss of the power device is first calculated by multiplying the on-state voltage drop and the instantaneous current within the fundamental frequency cycle to obtain the instantaneous conduction loss; then, the instantaneous conduction loss within one fundamental frequency cycle is integrated and averaged to obtain the power device's conduction loss power. Due to the instantaneous current flowing through the semiconductor power device Equal to the current of the filter inductor, therefore adopt express Conduction loss power It can be represented as:

[0076]

[0077] Since a three-level power electronic converter typically contains multiple semiconductor power devices, the conduction loss power of each power device needs to be obtained. Then, summing these values ​​yields the total conduction loss power of the three-level power electronic converter. .

[0078] III. Constructing the switching loss power function of semiconductor power devices

[0079] 1) Reduce the switching losses of semiconductor power devices Modeling as about It is a quadratic function, and is related to the reverse voltage drop experienced by the power device when it is turned off. Proportional.

[0080]

[0081] Where a sw b sw and c sw The coefficients of the quadratic function can be obtained from the power device's datasheet or through actual experiments; This information can be obtained from the power device's datasheet or measured using actual switching loss testing. Replacement.

[0082] 2) Further reduce switching losses Divided into turn-on losses and shutdown losses .

[0083]

[0084]

[0085] in, These are the corresponding quadratic function coefficients, which can be obtained from the power device's datasheet or through actual experiments.

[0086] 3) Switching losses of semiconductor power devices The average value is obtained by accumulating the switching losses of all switches operating within one fundamental cycle, where the current flowing through the semiconductor power device is equal to the current of the filter inductor.

[0087] In actual testing, the difference between turn-on loss and turn-off loss is generally large. Due to the presence of the filter inductor current ripple component, the actual current flowing through the power device at the turn-on time is smaller than that at the turn-off time. Therefore, when modeling the switching loss power of the power device, this invention further considers the influence of the filter inductor current ripple component, rather than simply using the fundamental component of the current.

[0088] The current flowing through the semiconductor power device is equal to the current in the filter inductor, therefore the switching power loss is... It can be represented as:

[0089]

[0090] Where T0 is the duration of one fundamental frequency period. .

[0091] Since a three-level power electronic converter typically contains multiple semiconductor power devices, obtaining the switching loss power of each power device is crucial. Then, summing these values ​​yields the total switching loss power of the three-level power electronic converter. .

[0092] IV. Constructing the power loss function of the filter inductor

[0093] 1) According to the Steinmetz equation, under sinusoidal excitation, the core loss power per unit volume is... With switching frequency f s Change in magnetic flux density The following relationship can be established:

[0094]

[0095] Where k, α and The correlation coefficient can be obtained from the core material datasheet or through actual experiments. It's important to note that the Steinmetz equation is an empirical equation in the field of electrical engineering, used to calculate magnetic losses per unit volume.

[0096] 2) According to electromagnetic theory, the change in magnetic flux density and The size is directly proportional, as shown in the following formula:

[0097]

[0098] in denoted as the relative permeability of the core material, N as the number of turns in the filter inductor, and l as the magnetic flux path length.

[0099] 3) Filter inductor loss and power By averaging the power loss at each moment within a fundamental frequency period, the following relationship can be established:

[0100]

[0101] Where V is the volume of the filter inductor core.

[0102] V. Establish the bus capacitor loss power function

[0103] Bus capacitor loss power Voltage fluctuation across the bus capacitor Voltage fluctuation angular frequency Capacitance C and loss angle Related to the loss angle You can obtain this information from the capacitor's datasheet or through actual testing. The expression is as follows:

[0104]

[0105] VI. Establish the overall loss function of the three-level power electronic converter

[0106] Overall losses of three-level power electronic converter Power loss due to filter inductor Bus capacitor loss power The total power loss is obtained by adding the four parts: total conduction loss, total switching loss, and total power loss. The formula is as follows:

[0107]

[0108] Adjust the switching frequency according to the overall loss until the overall loss is minimized, where only... and Subject to switching frequency The influence is related, therefore, based on and The sum of the two is smaller. However, by adjusting the switching frequency as a specific optimization target, the overall losses can be effectively reduced while improving the efficiency of the converter.

[0109] To demonstrate the accuracy of the filter inductor current ripple modeling proposed in this invention, as well as the implementation effect of the calculation method for the overall loss of the three-level power electronic converter and the optimization method based on the switching frequency, verification was conducted on a grid-connected converter experimental platform based on a SiC-Si heterogeneous ANPC three-level topology. The power devices used were the Si IGBT IKY75N120CH7 and the SiC MOSFET C3M0016120K. The platform power was 215kW, the DC bus voltage was 1330V, and the grid-side line voltage RMS was 690V. The invention will be further described in detail below using the 100% inverter operating mode as an example, in conjunction with the accompanying drawings.

[0110] Figure 1 This is a schematic diagram of a three-phase SiC-Si heterogeneous ANPC three-level topology, where the power device T is highlighted in red. x2 and T x3 (x = a, b, c) represents a SiC MOSFET, and the remaining power devices marked in blue are Si IGBTs. Accordingly, the following are used... Figure 2 The PWM2 modulation method shown is used, and a third harmonic is injected into the modulation wave to improve the utilization rate of the DC bus. In this embodiment, T x2 and T x3 (x = a, b, c) performs high-frequency switching, while the other power devices only perform power-frequency switching. Therefore, the switching losses of the power devices only need to be calculated using T. x2 and T x3 .

[0111] Figure 3 To obtain the nonlinear characteristic curve of the filter inductance value L as a function of instantaneous current i obtained through actual experiments, a cubic function is used for fitting, resulting in:

[0112]

[0113] First, calculate the amplitude of the ripple component of the filter inductor current. .

[0114] Based on the symmetry between the phase arms of the three-phase converter, the following analysis will take phase A as an example. The modulation wave is injected with a third harmonic, therefore:

[0115]

[0116]

[0117] ripple component amplitude of filter inductor current The cycle period is half a power frequency cycle, so we only need to consider 0 ≤ ≤ In this situation, we can obtain the following results: Figure 4The current ripple envelopes at different switching frequencies are shown.

[0118] Secondly, calculate the total conduction loss power of the semiconductor power device. .

[0119] Based on the symmetry of the upper and lower arms of the three-level ANPC topology, taking the power device of the upper arm as an example for analysis, the power factor angle α ranges from [- , It can be known that T a1 The IGBT conduction angle is [α, The conduction angle of the anti-parallel diode is [0, α]; T a2 The MOSFET conduction angle is [α, The conduction angle of the body diode is []. ]; T a5 The IGBT conduction angle is [ , 2 The conduction angle of the anti-parallel diode is [ ]. , Obtain V from the power device's datasheet. CE0 With r0. Where the knee voltage V of the SiC MOSFET is... CE0 It is zero.

[0120] Integrating each component yields T. a1 The sum of the conduction losses of the IGBT and the anti-parallel diode T a2 The sum of the conduction losses of the MOSFET and the body diode T a5 The sum of the conduction losses of the IGBT and the anti-parallel diode .

[0121]

[0122]

[0123]

[0124]

[0125] Next, calculate the total switching loss power of the semiconductor power device. .

[0126] Since Si IGBT power devices only perform power frequency switching, their switching losses are negligible; only the switching losses of SiCMOSFET power devices are considered. Based on the symmetry of the upper and lower arms of the three-level ANPC topology, the power device T of the upper arm... a2 We will analyze this using an example. Based on the actual double-pulse test, the power device T... a2 The curves of turn-on loss and turn-off loss with respect to current are as follows: Figure 5 As shown, by fitting with a quadratic function, we can obtain:

[0127]

[0128]

[0129] Considering the amplitude of the current ripple component The effect of turn-on and turn-off losses can be obtained by averaging all switching losses over one fundamental cycle:

[0130]

[0131] Then, calculate the filter inductor loss P. core .

[0132] Obtain the core loss per unit volume from the datasheet of the filter inductor core. With switching frequency Change in magnetic flux density The function.

[0133]

[0134] Based on the amplitude of the filter inductor current ripple component The change in magnetic flux density was calculated. .

[0135]

[0136] The power loss of the filter inductor is obtained by summing the power loss at each moment within a fundamental frequency period and then averaging the sums. .

[0137]

[0138] Furthermore, the bus capacitance loss was calculated. The loss angle δ is obtained from the capacitor's datasheet, and the following calculation is performed:

[0139]

[0140] Finally, the overall losses of the converter were calculated. .

[0141]

[0142] Overall losses of the converter Switching losses of power devices only and filter inductor loss Both parts are related to the switching frequency; therefore, the optimal switching frequency is selected with the goal of minimizing their sum and ensuring that switching losses are not excessive. For example... Figure 6 As shown, 25kHz is the optimal switching frequency, which ensures that the overall loss of the converter is almost the lowest while ensuring that the switching loss of the power devices is low, that is, the switching stress of the power devices is within a reasonable range.

[0143] To verify the accuracy of the filter inductor current ripple component modeling proposed in this invention, as well as the implementation effect of the calculation method for the overall loss of a three-level power electronic converter and the optimization method based on the switching frequency, the overall converter loss calculated by the method of this invention and the difference between the actual loss and the calculated loss were further compared when the switching frequency was 20kHz and 25kHz.

[0144] Experimental results are as follows Figure 7 and Figure 8 As shown, under the conditions of switching frequencies of 25kHz and 20kHz, the predicted results obtained by the filter inductor current ripple modeling proposed in this invention are in good agreement with the measured results, which verifies the accuracy of the current ripple modeling method.

[0145] As shown in Table 1, under the conditions of switching frequencies of 25kHz and 20kHz, the calculated results of the overall converter loss proposed in this invention are in good agreement with the measured results, with an error of only about 1%. Furthermore, when the switching frequency is increased from 20kHz to 25kHz, the overall loss obtained from the experiment is reduced by 167W, which verifies the effectiveness of the method for optimizing the overall operating efficiency of the converter based on the switching frequency proposed in this invention.

[0146] Table 1 Comparison of Calculated and Experimental Results of Overall Converter Losses

[0147]

[0148] The above description of the embodiments is provided to enable those skilled in the art to understand and apply the present invention. It will be apparent to those skilled in the art that various modifications can be easily made to the above embodiments, and the general principles described herein can be applied to other embodiments without inventive effort. Therefore, the present invention is not limited to the above embodiments. Improvements and modifications made to the present invention by those skilled in the art based on the disclosure of the present invention should be within the scope of protection of the present invention, such as using different converter hardware topologies, modulation methods, software control code, etc. Therefore, the claims are intended to cover all variations within the true concept and scope of the present invention.

Claims

1. A method for overall power loss calculation and optimization of a three-level converter, characterized in that, Includes the following steps: S1. Calculate the amplitude of the ripple component of the filter inductor current based on the switching frequency of the three-level converter and the inductance value, voltage across the two ends, and instantaneous current of the filter inductor of the three-level converter. S1 includes: S101, fits the nonlinear relationship between the inductance value of the filter inductor and the instantaneous current into a cubic function; S102, Construct the modulation wave function; Based on the switching frequency and the modulation wave function, calculate the duration of the current rise in the ripple component of the filter inductor current. S103, the amplitude of the ripple component of the filter inductor current is calculated based on the cubic function, the duration, and the voltage across the filter inductor. S2, calculate the conduction loss power of each semiconductor power device based on the instantaneous current and the conduction voltage drop of the semiconductor power device; add up the conduction loss power of all semiconductor power devices in the three-level converter to obtain the total conduction loss power; The switching loss power of each semiconductor power device is calculated by combining the amplitude of the ripple component and the switching frequency; the switching loss power of all semiconductor power devices in the three-level converter is added together to obtain the total switching loss power. S3, calculate the filter inductor loss power based on the switching frequency and the amplitude of the ripple component; calculate the bus capacitor loss power based on the voltage fluctuation across the three-level converter bus capacitor, the voltage fluctuation angular frequency, the capacitance value, and the loss angle. S4. The total conduction loss power, total switching loss power, filter inductor loss power, and bus capacitor loss power are added together to obtain the overall power loss of the three-level converter; the overall power loss is adjusted by changing the switching frequency until the desired overall power loss optimization target is achieved.

2. The method of claim 1, wherein, In S102, the duration of the current rise is obtained by dividing the modulation wave function at the current moment by the switching frequency.

3. The method of claim 1, wherein, In S103, the ripple component amplitude of the filter inductor current is specifically calculated using the following formula: ； wherein is the amplitude of the ripple component of the filtered inductor current, is the switching frequency, , is the voltage across the filtered inductor, is the inductance value of the filtered inductor, is the duration of the current rise portion of the ripple component of the filtered inductor current, is the DC bus voltage of the converter, is the modulation wave function.

4. The method of claim 1, wherein, In S2, the specific calculation method for the conduction loss power is as follows: within one fundamental frequency period, the instantaneous current at the current moment is multiplied by the conduction voltage drop to obtain the instantaneous conduction loss power, and then the instantaneous conduction loss power is multiplied by the modulation wave function at the current moment to obtain the calculation result. Traverse all time in one fundamental wave period, get the calculation result corresponding to each time, then integrate all time calculation results in 0 to Get the integral total value; divide the integral total value by Get the conduction loss power.

5. The method of claim 1, wherein, The switching loss power includes the turn-on loss of the semiconductor power device and the turn-off loss of the semiconductor power device.

6. The method of claim 5, wherein, The switching loss power is calculated as follows: ； wherein is the switching loss power, i is the instantaneous current, is the ripple component amplitude of the filter inductor current, is the switching frequency, t is the time, is the reverse voltage drop experienced by the power device when it is turned off, is the turn-off loss, To is one fundamental period, is the turn-on loss.

7. The method of claim 1, wherein, In S3, the power loss of the bus capacitor is specifically calculated by multiplying the square of the voltage fluctuation across the converter bus capacitor within one fundamental cycle by the voltage fluctuation angular frequency, then by the capacitance value of the converter bus capacitor, and finally by the tangent of the converter bus capacitor loss angle.

8. The method of claim 1, wherein, In S3, the power loss of the filter inductor is calculated as follows: S301, the change in magnetic flux density at the current moment is calculated based on electromagnetic theory and the amplitude of the ripple component; S302, according to the Steinmetz equation, the current unit volume core loss power is calculated using the switching frequency and the change in magnetic flux density; S303, adding the unit volume magnetic core loss power at all time points in one fundamental period, multiplying the filter inductance magnetic core volume, and further dividing by the carrier ratio to obtain the filter inductance loss power; the carrier ratio is obtained by multiplying the switching frequency and the fundamental period.

9. The method of claim 1, wherein, In S4, the overall power loss optimization target is to minimize the overall power loss of the three-level converter.

Images