A method for calculating harmonic loss of a converter transformer based on multi-physical field coupling

Abstract

The application discloses a kind of based on the harmonic loss calculation method of converter transformer of multi-physical field coupling, comprising: the electromagnetic field, temperature field and mechanical field coupling simulation model of converter transformer are established;Harmonic source signal in the process of commutation is obtained and frequency spectrum analysis is carried out, to obtain each harmonic component;Each harmonic component is as excitation input coupling simulation model, obtains electromagnetic field distribution and temperature field distribution;According to temperature field distribution, electromagnetic field distribution and harmonic frequency, winding high-frequency resistance is corrected;According to core material parameters, electromagnetic field distribution and temperature field distribution, establish or update core nonlinear magnetic characteristic model;The copper loss and iron loss corresponding to each harmonic are calculated, and the total harmonic loss is obtained according to the coupling relationship between harmonics and is corrected superposition.This application can comprehensively consider high-frequency effect, core nonlinear characteristic and harmonic coupling influence, improve the accuracy of converter transformer harmonic loss calculation.

Description

Technical Field

[0001] This invention belongs to the field of power system transmission technology, specifically relating to a method for calculating harmonic losses of converter transformers based on multi-physics coupling. Background Technology

[0002] Converter transformers are one of the core devices in DC transmission systems. They are typically installed between the AC system and the converter valve to achieve voltage level transformation, electrical isolation, and energy transfer required for the operation of the converter system. Compared with ordinary AC transformers, converter transformers operate under more complex conditions. Their windings and cores not only bear the effects of power frequency voltage and current, but are also subject to harmonic voltages and harmonic currents generated by the nonlinear switching process of the converter over a long period of time.

[0003] In conventional DC transmission systems, the periodic switching on, off, and commutation processes of thyristor converters generate multiple harmonic components on both the valve side and the grid side. These harmonic components, upon entering the converter transformer, alter the current density distribution within the winding conductors and induce additional losses in the core, clamps, tank, and other metal structures. With the increasing capacity and voltage levels of DC transmission projects, the calculated losses of converter transformers not only affect equipment efficiency evaluation but also influence temperature rise verification, insulation life assessment, and operational safety margin judgment. Therefore, how to reasonably assess the losses of converter transformers under harmonic effects has always been a crucial issue in converter transformer design and operational analysis.

[0004] Currently, the common methods for calculating harmonic losses in engineering still require some simplification.

[0005] For example, some methods calculate the additional losses of converter transformers based on the total harmonic distortion rate or the content of major harmonics. While these methods are simple to calculate and convenient for engineering estimation, they typically treat the effects of harmonics of different frequencies, phases, and amplitudes as equivalents, making it difficult to reflect the actual differences in losses caused by different harmonics in different regions of the windings and core.

[0006] Other methods utilize electromagnetic transient simulation software to obtain harmonic voltages and currents during the operation of the converter transformer, then calculate the corresponding losses based on each harmonic component, and finally sum the results as the total harmonic loss. This type of method offers better adaptability to operating conditions compared to empirical estimation, but it often uses simplified resistance, magnetization characteristics, or loss coefficients in the loss calculation stage. In particular, it fails to adequately consider the uneven winding current density under high-frequency harmonics, the nonlinear changes in core material, and the mutual influence between different harmonics. For converter transformers, which have complex winding structures, strong leakage magnetic fields, and significant local variations in core magnetization, these simplifications can easily cause the calculation results to deviate from the actual operating conditions.

[0007] Empirical formulas based on material properties are also common. These methods typically calculate core losses based on parameters such as the magnetization curve of silicon steel sheets, hysteresis loss coefficient, and eddy current loss coefficient. However, material test data often comes from specific frequencies, specific magnetic flux density ranges, and relatively ideal sinusoidal excitation conditions. In actual operation, converter transformers experience a wider harmonic frequency range, uneven magnetic flux density distribution, and localized nonlinear magnetization in the core. Directly extrapolating material parameters under power frequency or low-frequency conditions often fails to accurately describe the decrease in core permeability, increase in eddy current loss, and localized loss concentration under high-frequency harmonic conditions.

[0008] Regarding winding losses, the shortcomings of existing methods are even more apparent. Converter transformer windings exhibit skin and proximity effects under the influence of harmonic currents. The higher the harmonic frequency, the more uneven the current distribution within the conductor, and the smaller the effective conductive area. Simultaneously, leakage magnetic field coupling between adjacent conductors further alters the local current density distribution. If the power frequency resistance or only temperature-corrected resistance is used as the basis for calculating copper losses for each harmonic, it cannot fully reflect the changes in high-frequency winding resistance with harmonic frequency, conductor arrangement, and local electromagnetic field variations.

[0009] Existing technology CN106202736A discloses a method for coupled calculation of electromagnetic field-fluid-temperature field of converter transformer. This method first performs spectral analysis on the winding current of the converter transformer to obtain the DC component and harmonic components; then, it measures the loss characteristics and magnetization characteristics of the laminated core under 50Hz sinusoidal excitation at different temperatures to establish electromagnetic field geometric models and fluid-temperature field geometric models; next, it uses a heterogeneous mesh node data mapping algorithm to map the losses in the electromagnetic field to the fluid-temperature field, and then maps the temperature results back to the electromagnetic field calculation unit, obtaining the temperature distribution through iterative iteration. This method can consider the data transfer between the electromagnetic field calculation results and the temperature field, and can also reflect the influence of temperature changes on the core magnetization and loss characteristics to a certain extent.

[0010] However, this method primarily focuses on obtaining the temperature distribution of the converter transformer. In its technical process, losses are treated more as heat sources in the fluid-temperature field calculation, rather than undergoing quantitative analysis of the composition of copper losses, iron losses, and total harmonic losses for each harmonic order. For the assessment of harmonic losses in converter transformers, simply obtaining the temperature distribution cannot directly reflect the contribution of different harmonic orders to winding copper losses, core iron losses, and ultimately, the total losses.

[0011] Furthermore, the loss and magnetization characteristics of the laminated core in this method are mainly measured based on 50Hz sinusoidal excitation conditions at different temperatures. While this data can reflect the influence of temperature on the core material properties under power frequency conditions, the harmonic components in the actual operation of converter transformers are not limited to the vicinity of power frequency. Higher frequency harmonics can cause changes in the core's eddy current effect, local magnetization state, and equivalent permeability. If the material property data under 50Hz sinusoidal excitation is relied upon primarily, the characterization of the nonlinear loss characteristics of the core under high-frequency harmonics remains insufficient.

[0012] Furthermore, while the existing technology performs spectral analysis on the winding current, its subsequent processing focuses on iterative calculations between the harmonic magnetic field and the fluid-temperature field. It does not fully reflect the impact of each harmonic frequency on the high-frequency resistance of the winding, nor does it establish corresponding correction relationships for the skin effect and proximity effect of the winding. For high-voltage, high-capacity converter transformers, the winding conductor size, arrangement, and leakage magnetic field distribution all affect the high-frequency current distribution. Without processing for the high-frequency resistance of the winding, the calculation of harmonic copper losses is prone to errors.

[0013] Meanwhile, existing methods typically treat losses caused by different harmonics as heat sources in temperature field calculations, while neglecting the coupling effects between different harmonic orders. Under complex harmonic operating conditions, different harmonic components may interact through winding leakage magnetic fields, local core magnetization states, and equivalent impedance relationships, causing the total loss to no longer be a simple linear superposition of harmonic losses. If such interactions between harmonics are ignored, errors may still occur in the calculation of total harmonic loss in operating scenarios with multiple harmonics acting together.

[0014] In summary, existing methods for calculating harmonic losses in converter transformers still suffer from several problems, including insufficient consideration of high-frequency winding resistance variations, inadequate description of the broadband nonlinear loss characteristics of the core, insufficiently detailed analysis of the composition of each harmonic loss, and insufficient consideration of the coupling effects between harmonics. As a result, the calculated harmonic losses of converter transformers under complex harmonic operating conditions still cannot meet the needs of refined design and operational evaluation. Summary of the Invention

[0015] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a method for calculating harmonic losses of converter transformers based on multi-physics coupling.

[0016] The objective of this invention can be achieved through the following technical solutions: This invention provides a method for calculating harmonic losses in converter transformers based on multi-physics coupling, comprising the following steps: Establish a multi-physics coupling simulation model for the converter transformer; Harmonic source signals during the commutation process are acquired, and the harmonic source signals are subjected to spectral analysis to obtain each harmonic component. By using the harmonic components as excitation inputs to the multiphysics coupling simulation model, the electromagnetic field distribution, temperature field distribution, and deformation information of the converter transformer under the action of each harmonic are obtained. Based on the temperature field distribution, electromagnetic field distribution and harmonic components, the high-frequency resistance of the converter transformer winding is corrected to obtain the corrected high-frequency resistance corresponding to each harmonic. Based on the electromagnetic field distribution, temperature field distribution, and core material parameters, establish or update the nonlinear magnetic characteristic model of the converter transformer core. Based on the harmonic components, the corrected high-frequency resistor, and the nonlinear magnetic characteristic model, the copper loss and iron loss corresponding to each harmonic are calculated in the multiphysics coupling simulation model. The coupling relationship between each harmonic is determined based on the electromagnetic field distribution, and the copper loss and iron loss corresponding to each harmonic are corrected and superimposed based on the coupling relationship to obtain the total harmonic loss of the converter transformer.

[0017] Furthermore, the establishment of the multiphysics coupling simulation model of the converter transformer specifically includes: Based on the geometric and electromagnetic parameters of the windings, core, insulation structure, tank, and cooling medium of the converter transformer, an electromagnetic field sub-model is established based on Maxwell's equations. The electromagnetic field sub-model is used to receive each harmonic component as excitation, and outputs the magnetic flux density distribution, current density distribution, Joule heat source, iron loss heat source, and electromagnetic force distribution under the action of each harmonic after inputting each harmonic component. Based on the Joule heat source and iron loss heat source output by the electromagnetic field sub-model, a temperature field sub-model is established based on the heat conduction equation, and combined with the convective heat transfer boundary conditions between the cooling medium and the heat-generating components of the converter transformer, the temperature field distribution of the winding, core and surrounding medium is obtained. Based on the electromagnetic force distribution output by the electromagnetic field sub-model and the temperature field distribution output by the temperature field sub-model, a mechanical field sub-model is established. The electromagnetic force distribution is used as the force load, and the thermal expansion caused by the temperature field distribution is used as the thermal load to obtain the mechanical stress distribution and deformation information of the winding and the core. The temperature field distribution is fed back to the electromagnetic field sub-model to update the winding resistivity and core permeability, and the deformation information is fed back to the electromagnetic field sub-model to update the local geometric position and local electromagnetic field distribution of the winding, core and surrounding medium, thereby forming a coupled simulation model between electromagnetic field, temperature field and mechanical field.

[0018] Furthermore, the heat conduction equation satisfies: in, The density of the material; Specific heat capacity of the material; For temperature; For time; Thermal conductivity; It is a Joule heat source generated by the winding current; It is the heat source of iron loss generated by iron core loss; The thermal strain in the mechanical field sub-model satisfies: in, Thermal strain; The coefficient of linear expansion of the material; The temperature output by the temperature field submodel; This is a reference temperature.

[0019] Furthermore, the acquisition of harmonic source signals during the commutation process and the spectral analysis of the harmonic source signals to obtain each harmonic component specifically include: Based on the converter's pulsation number, firing angle, commutation angle, and AC system fundamental frequency, the range of harmonic orders to be analyzed during the switching process is determined, and the harmonics within the range of harmonic orders are used as the target harmonics to be input into the multiphysics coupling simulation model. The voltage time-domain signal and current time-domain signal of the AC side of the converter valve and / or the valve side of the converter transformer are acquired or simulated, and the DC component is removed, synchronous sampling and windowing are performed on the voltage time-domain signal and / or current time-domain signal to obtain the preprocessed harmonic source signal; A fast Fourier transform is performed on the preprocessed harmonic source signal to extract the harmonic frequency, effective value of harmonic voltage, effective value of harmonic current, and initial phase of each target harmonic. The harmonic frequency, the effective value of the harmonic voltage, the effective value of the harmonic current, and the initial phase of the harmonic are combined to form each harmonic component, and each harmonic component is used as the harmonic excitation input of the multiphysics coupling simulation model.

[0020] Furthermore, the step of using the harmonic components as excitation inputs to the multiphysics coupling simulation model to obtain the electromagnetic field distribution, temperature field distribution, and deformation information of the converter transformer under the action of each harmonic specifically includes: Based on the harmonic frequency, effective value of harmonic voltage, effective value of harmonic current, and initial phase of each harmonic component, a corresponding time-domain harmonic excitation is constructed, and the time-domain harmonic excitation is applied to the winding port or the valve side boundary of the converter transformer in the multi-physics coupling simulation model. The time-domain harmonic excitation is expressed as: in, for Time-domain voltage excitation or time-domain current excitation corresponding to the subharmonic; for The effective value of voltage or effective value of current corresponding to the subharmonic; for Second harmonic frequency; for The initial phase of the second harmonic; For each target harmonic, the frequency domain is solved in the electromagnetic field sub-model to obtain the magnetic vector potential, magnetic flux density distribution, current density distribution and electromagnetic force distribution under the action of the target harmonic; Wherein, the frequency domain solution satisfies: in, For curl operator; for Resistivity under the influence of subharmonics, and ; for Equivalent permeability under the action of subharmonics; for The magnetic vector potential corresponding to the subharmonic; The imaginary unit; for The angular frequency corresponding to the subharmonic; for Equivalent conductivity under the influence of subharmonics; for Current density corresponding to the subharmonic; Based on the current density distribution and magnetic flux density distribution output by the electromagnetic field sub-model, the winding Joule heat source and iron loss heat source corresponding to each target harmonic are determined respectively, and the winding Joule heat source and iron loss heat source corresponding to each target harmonic are superimposed to obtain the total heat source input to the temperature field sub-model. The total heat source is represented as: in, For spatial location The total heat source at the location; For the target harmonic order set; For spatial location Cause The winding Joule heat source generated by the subharmonic; For spatial location Cause Iron loss heat source generated by subharmonics; The total heat source is input into the temperature field sub-model, and the temperature field distribution of the winding, core and surrounding medium is obtained by combining the convective heat transfer boundary conditions of the cooling medium. The electromagnetic force distribution output by the electromagnetic field sub-model and the temperature field distribution output by the temperature field sub-model are input into the mechanical field sub-model to solve for the mechanical stress distribution and deformation information of the winding and the core. The winding resistivity and core permeability are updated based on the temperature field distribution. The local geometric positions of the winding, core and surrounding medium are updated based on the deformation information. The electromagnetic field sub-model is then resolved until the temperature field distribution and deformation information obtained from two adjacent iterations satisfy the first preset convergence condition. The electromagnetic field distribution and temperature field distribution of the converter transformer under each harmonic are then output.

[0021] Furthermore, the step of correcting the high-frequency resistance of the converter transformer winding based on the temperature field distribution, electromagnetic field distribution, and harmonic components to obtain the corrected high-frequency resistance corresponding to each harmonic specifically includes: Based on the temperature field distribution, the equivalent temperature of the conductor region of the converter transformer winding is extracted, and the power frequency resistance of the winding at the reference temperature is corrected according to the equivalent temperature to obtain the temperature-corrected resistance. The equivalent temperature and the temperature correction resistor are defined by the following formulas: in, This is the equivalent temperature of the winding conductor region; This refers to the winding conductor region; For spatial location Temperature at that location; Temperature correction resistor; Reference temperature The power frequency winding resistance is below; The temperature coefficient of resistance of the winding conductor; For reference temperature; Based on the harmonic frequencies in each harmonic component and the resistivity of the winding conductor at the equivalent temperature, the skin depth corresponding to each harmonic is determined, and the effective conductive area corresponding to each harmonic is determined based on the skin depth and the equivalent cross-sectional size of the winding conductor. The skin depth and the effective conductive area are defined by the following formulas: in, for Skin depth corresponding to the subharmonic; The resistivity of the winding conductor at the equivalent temperature; for The angular frequency corresponding to the subharmonic; for Equivalent permeability of the region where the winding conductor is located under the action of subharmonics; for The effective conductive area corresponding to the subharmonic; This is the equivalent diameter of the winding conductor; It is the error function; Based on the current density distribution in the winding region and the leakage magnetic field distribution in the adjacent winding in the electromagnetic field distribution, the proximity effect correction coefficient corresponding to each harmonic is determined, and the correction high-frequency resistance corresponding to each harmonic is obtained based on the temperature correction resistance, the effective conductive area and the proximity effect correction coefficient. Wherein, the corrected high-frequency resistor satisfies: in, for Corrected high-frequency resistors corresponding to subharmonics; This represents the geometric cross-sectional area of ​​the winding conductor; for The proximity effect correction coefficient corresponding to the subharmonic is determined by the degree of non-uniformity of winding current density in the electromagnetic field distribution and the leakage magnetic coupling strength between adjacent conductors.

[0022] Furthermore, the step of establishing or updating the nonlinear magnetic characteristic model of the converter transformer core based on the electromagnetic field distribution, temperature field distribution, and core material parameters specifically includes: Based on the electromagnetic field distribution, the magnetic flux density distribution and magnetic field strength distribution of the iron core region under the action of each harmonic are extracted, and based on the temperature field distribution, the local temperature distribution corresponding to the iron core region is extracted. Based on the magnetization curve, hysteresis loop, core conductivity, lamination thickness, and reference permeability in the core material parameters, the core region is divided into an unsaturated region, a transition region, and a saturated region, and corresponding piecewise permeability functions are configured for the unsaturated region, the transition region, and the saturated region, respectively. The piecewise permeability function is temperature-corrected based on the local temperature distribution, and the piecewise permeability function after temperature correction is frequency-corrected based on the harmonic frequencies in each harmonic component to obtain the equivalent nonlinear permeability corresponding to each harmonic. Based on the equivalent nonlinear permeability, the material properties of the iron core region in the electromagnetic field sub-model are updated, and the magnetic flux density distribution and magnetic field strength distribution of the iron core region are re-extracted based on the updated electromagnetic field distribution. When the change in equivalent nonlinear permeability of the core region before and after the update satisfies the second preset convergence condition, the current piecewise permeability function, frequency correction factor, hysteresis loss parameter and eddy current loss parameter are used as the nonlinear magnetic characteristic model of the converter transformer core.

[0023] Furthermore, the equivalent nonlinear permeability is expressed as: in, For spatial location Place The equivalent nonlinear permeability corresponding to the subharmonic; The piecewise permeability function is determined by the core magnetization curve and temperature field distribution; For spatial location Place The magnetic flux density corresponding to the subharmonic; For spatial location Local temperature of the iron core at that location; for Frequency correction factor corresponding to subharmonics; The frequency correction factor satisfies: in, for Frequency correction factor corresponding to subharmonics; for Second harmonic frequency; The preset high-frequency correction threshold is used; This is the frequency correction factor for the core material.

[0024] Furthermore, the calculation of copper and iron losses corresponding to each harmonic in the multiphysics coupling simulation model based on the harmonic components, the corrected high-frequency resistance, and the nonlinear magnetic characteristic model specifically includes: Based on the effective value of the harmonic current in each harmonic component and the corrected high-frequency resistance corresponding to each harmonic, the copper loss of the winding under the action of each harmonic is calculated. Wherein, the copper loss is represented as: in, for The winding copper loss corresponding to the subharmonic; for RMS value of subharmonic current; for Corrected high-frequency resistors corresponding to subharmonics; Based on the electromagnetic field distribution, the magnetic flux density distribution of the core region under each harmonic is extracted, and the hysteresis loss parameter, eddy current loss parameter, frequency correction factor and equivalent nonlinear permeability in the nonlinear magnetic characteristic model are called to calculate the iron loss of the core under each harmonic. Wherein, the iron loss is expressed as: in, for Core iron loss corresponding to subharmonics; For the iron core region; This is the hysteresis loss coefficient; for Second harmonic frequency; For spatial location Place The magnetic flux density corresponding to the subharmonic; The hysteresis loss index; The eddy current loss coefficient is used. The thickness of the iron core laminations; The copper and iron losses corresponding to each harmonic are output to the temperature field sub-model of the multiphysics coupling simulation model to update the temperature field distribution. The updated temperature field distribution is then fed back to correct the high-frequency resistance of the winding and the nonlinear magnetic characteristic model of the iron core until the copper and iron losses corresponding to each harmonic meet the third preset convergence condition.

[0025] Furthermore, the step of determining the coupling relationship between each harmonic based on the electromagnetic field distribution, and correcting and superimposing the copper loss and iron loss corresponding to each harmonic based on the coupling relationship to obtain the total harmonic loss of the converter transformer, specifically includes: Based on the electromagnetic field distribution, the magnetic field overlap distribution, current density coupling distribution, and equivalent impedance matrix of different harmonic orders in the winding region and iron core region are extracted, and the mutual impedance and self impedance between different harmonic orders are determined by the equivalent impedance matrix. Based on the mutual impedance and self-impedance, determine the harmonic coupling coefficient between any two different harmonic orders; The harmonic coupling coefficient is expressed as: in, for Second harmonics and Harmonic coupling coefficient between subharmonics; for Second harmonics and Mutual impedance between subharmonics; for Self-impedance corresponding to subharmonics; for Self-impedance corresponding to subharmonics; is the real part of the mutual impedance; Based on the copper loss, iron loss and harmonic coupling coefficient corresponding to each harmonic, the inter-harmonic coupling loss is calculated, and the copper loss, iron loss and inter-harmonic coupling loss corresponding to each harmonic are superimposed to obtain the total harmonic loss of the converter transformer. The total harmonic loss is expressed as: in, This refers to the total harmonic losses of the converter transformer; The set of target harmonic orders for loss calculation; for Copper loss corresponding to subharmonics; for Iron loss corresponding to subharmonics.

[0026] Compared with the prior art, the present invention has the following advantages: (1) Compared with existing technologies, existing methods for calculating harmonic losses in converter transformers typically focus on electromagnetic field or electromagnetic field-temperature field calculations, easily treating losses as intermediate heat sources in temperature field solutions, without fully considering the mutual influence between temperature changes, local electromagnetic field changes, and structural deformation. This invention establishes a multi-physics coupled simulation model including electromagnetic field, temperature field, and mechanical field, and feeds back temperature field distribution and deformation information to the electromagnetic field sub-model, enabling the winding resistivity, core permeability, and local geometric position to be updated with the operating state. This reduces field distribution deviations caused by unidirectional coupling or single-field calculations, making the harmonic loss calculation results closer to the thermal, electrical, and structural coupling state of the converter transformer in actual operation.

[0027] (2) Existing technologies often use power frequency resistance or only temperature-corrected winding resistance when calculating winding copper losses, which is difficult to reflect the uneven current density caused by skin effect and proximity effect under high-frequency harmonics. This invention corrects the high-frequency resistance of the winding based on temperature field distribution, electromagnetic field distribution, and each harmonic component to obtain the corrected high-frequency resistance corresponding to each harmonic. This avoids treating different frequency harmonics as the same resistance, thereby improving the accuracy of high-frequency harmonic copper loss calculation.

[0028] (3) Existing calculations of core losses mostly rely on the material magnetization curves and loss parameters under power frequency or low-frequency conditions. Under higher frequency harmonic effects, problems such as the decrease in core permeability, the increase in eddy current loss, and the enhancement of local nonlinearity are easily overlooked. This invention establishes or updates the nonlinear magnetic characteristic model of the core based on the electromagnetic field distribution, temperature field distribution, and core material parameters. It also corrects the equivalent nonlinear permeability at different harmonic frequencies through a frequency correction factor, enabling the core loss calculation to simultaneously reflect the combined effects of magnetic flux density, temperature, and frequency. This reduces the errors generated when only 50Hz material characteristics or a single empirical formula are extrapolated to high-frequency harmonic conditions.

[0029] (4) Existing technologies usually only provide the overall loss or temperature distribution results of the converter transformer, making it difficult to distinguish the specific contributions of different harmonic orders to copper loss and iron loss. Based on each harmonic component, the corrected high-frequency resistance, and the nonlinear magnetic characteristic model, this invention calculates the copper loss and iron loss corresponding to each harmonic separately, so that the total loss can be decomposed according to the harmonic order and loss type, making it easier to determine the main sources of loss.

[0030] (5) Existing methods often directly add up the losses of each harmonic when dealing with multiple harmonic losses, assuming that the harmonics do not affect each other. However, in converter transformers, different harmonics may interact through leakage magnetic fields, local magnetization states, and equivalent impedances, resulting in the total loss not being a simple linear superposition. This invention determines the coupling relationship between each harmonic based on the electromagnetic field distribution, calculates the harmonic coupling coefficient based on mutual impedance, and then corrects and superimposes the copper and iron losses corresponding to each harmonic. Through this process, the impact of mutual interference between harmonics on the total loss can be quantified, reducing the risk of underestimating or overestimating the total harmonic loss under complex harmonic operating conditions.

[0031] (6) Existing technologies often handle harmonic sources in a rather coarse manner, easily neglecting the influence of harmonic frequency, amplitude, and initial phase on subsequent field distribution and loss calculation. This invention obtains each harmonic component by performing spectral analysis on the harmonic source signal during the converter process, and uses it as the harmonic excitation input of the multi-physics coupling simulation model, so that the input operating conditions have a better correspondence with the actual converter process.

[0032] (7) In existing loss calculations, the temperature field, electromagnetic field, and loss results are often obtained step by step, which can easily lead to inconsistencies in the calculation parameters. This invention feeds back the copper and iron losses of each harmonic to the temperature field sub-model, and uses the updated temperature field distribution to further correct the winding high-frequency resistance and the nonlinear magnetic characteristic model of the iron core until the preset convergence condition is met, thereby improving the consistency and stability of the calculation process.

[0033] (8) In addition, the present invention also considers the influence of deformation information caused by electromagnetic force and thermal expansion on the local electromagnetic field distribution, which can reflect the effect of structural state changes on local loss concentration to a certain extent, and avoid the calculation deviation caused by always treating the converter transformer as a fixed geometric structure. Attached Figure Description

[0034] Figure 1 This is a flowchart of the converter transformer harmonic loss calculation method according to an embodiment of the present invention; Figure 2 This is a loss distribution diagram of the inner side of the converter transformer at 250Hz according to an embodiment of the present invention. Figure 3 This is a loss distribution diagram on the inner side of the lower surface of the casing at 250Hz, according to an embodiment of the present invention. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below 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 should fall within the scope of protection of the present invention.

[0036] Example 1: This embodiment provides a method for calculating harmonic losses in converter transformers based on multi-physics coupling. This method no longer estimates harmonic losses solely based on the total harmonic distortion rate or the superposition of single harmonic losses. Instead, it first extracts the actual harmonic components present during the conversion process, and then uses these components as excitation inputs to a multi-physics coupling simulation model. Through feedback between electromagnetic field distribution, temperature field distribution, and deformation information, it corrects the high-frequency resistance of the windings and the nonlinear magnetic characteristics of the core, thereby calculating the copper losses, iron losses, and inter-harmonic coupling losses corresponding to each harmonic. This avoids the significant deviations that occur when calculating harmonic losses using only power frequency resistance, power frequency magnetization characteristics, or simple linear superposition, making the calculation results closer to the actual operating state of the converter transformer under complex harmonic conditions.

[0037] like Figure 1 , Figure 2 , Figure 3 As shown, the method in this embodiment includes the following steps: S101. Establish a multi-physics coupling simulation model for the converter transformer; S102. Obtain the harmonic source signal during the commutation process, and perform spectrum analysis on the harmonic source signal to obtain the harmonic components of each order. S103. Using each harmonic component as the excitation input, a multi-physics coupling simulation model is used to obtain the electromagnetic field distribution and temperature field distribution of the converter transformer under the action of each harmonic. S104. Based on the temperature field distribution, electromagnetic field distribution and each harmonic component, the high-frequency resistance of the converter transformer winding is corrected to obtain the corrected high-frequency resistance corresponding to each harmonic. S105. Based on the electromagnetic field distribution, temperature field distribution, and core material parameters, establish or update the nonlinear magnetic characteristic model of the converter transformer core. S106. Based on the harmonic components, the corrected high-frequency resistance, and the nonlinear magnetic characteristic model, the copper loss and iron loss corresponding to each harmonic are calculated in the multi-physics field coupling simulation model. The coupling relationship between each harmonic is determined according to the electromagnetic field distribution. The copper loss and iron loss corresponding to each harmonic are corrected and superimposed to obtain the total harmonic loss of the converter transformer.

[0038] In step S101, a multi-physics coupling simulation model of the converter transformer is established.

[0039] In one specific implementation, a simulation model is established using a ±800kV converter transformer as the object. This converter transformer has a capacity of 500MVA, a rated voltage of ±800kV / 500kV, 1200 turns in the high-voltage winding, and 750 turns in the low-voltage winding. The winding conductors are copper, and the core is a laminated silicon steel structure. During modeling, the windings, core, insulation structure, tank, and cooling medium are treated as different computational regions in the finite element model, and corresponding material parameters such as conductivity, permeability, dielectric constant, thermal conductivity, specific heat capacity, density, and coefficient of linear expansion are assigned to each region.

[0040] Specifically, the electromagnetic field sub-model is established using the finite element method. The winding region is used to apply harmonic voltage or harmonic current excitation, the core region is used to characterize magnetization and iron loss characteristics, and the oil or air region is used to characterize the influence of the surrounding medium on the leakage magnetic field distribution. The electromagnetic field sub-model outputs the magnetic flux density distribution, current density distribution, Joule heat source, iron loss heat source, and electromagnetic force distribution under each harmonic action.

[0041] The temperature field sub-model receives the Joule heat source and iron loss heat source output from the electromagnetic field sub-model, and calculates the temperature field distribution of the winding, core, and surrounding medium by combining the heat conduction and convective heat transfer boundary conditions between the winding, core, and cooling medium. This temperature field distribution is not merely output as a result, but is further fed back to the electromagnetic field sub-model to update the winding resistivity and core permeability.

[0042] The mechanical field sub-model receives the electromagnetic force distribution output from the electromagnetic field sub-model and the temperature field distribution output from the temperature field sub-model. It treats the electromagnetic force as a force load and the thermal expansion caused by temperature changes as a thermal load to solve for the mechanical stress distribution and deformation information of the winding and core. The deformation information is further fed back to the electromagnetic field sub-model to update the local geometric positions of the winding, core, and surrounding medium, enabling the electromagnetic field solution to reflect the local structural state after thermal expansion and electromagnetic force action.

[0043] It should be noted that existing methods often treat the electromagnetic field, temperature field, and mechanical state separately, or only treat electromagnetic losses as a heat source for unidirectional temperature field transfer. For converter transformers, local heating caused by harmonic currents changes the winding resistivity, core temperature rise changes the permeability, and winding stress and thermal expansion may also change the local leakage magnetic field distribution. If these feedback relationships are not considered, subsequent loss calculations will use the initial geometry and material parameters, leading to deviations between the calculation results and the actual operating conditions. This embodiment improves the consistency of the calculation process by using cyclic feedback between the electromagnetic field, temperature field, and mechanical field, ensuring that subsequent high-frequency resistance correction, core magnetic characteristic update, and loss calculation are all based on the updated field distribution.

[0044] In step S102, the harmonic source signal during the commutation process is acquired, and the spectrum analysis of the harmonic source signal is performed to obtain the harmonic components.

[0045] In one specific implementation, the voltage time-domain and current time-domain signals of the AC side of the converter valve and / or the valve side of the converter transformer are acquired using electromagnetic transient simulation software or field measurement equipment. For a 12-pulse converter, its characteristic harmonic order can be determined according to the converter's pulse number; for example, the main characteristic harmonics may include the 11th, 13th, 23rd, and 25th harmonics. For a dual 12-pulse converter system, the range of harmonic orders to be analyzed can be further determined based on the converter's wiring and operating mode.

[0046] After acquiring the voltage and current time-domain signals, DC component removal, synchronous sampling, and windowing are performed to reduce the impact of DC bias, spectral leakage, and asynchronous sampling on the harmonic extraction results. Then, a Fast Fourier Transform is applied to the preprocessed signals to extract the harmonic frequency, RMS value of the harmonic voltage, RMS value of the harmonic current, and initial phase of each target harmonic, which are then combined to form the harmonic components.

[0047] In one implementation, The second harmonic component can be expressed as: in, for Second harmonic components; for Second harmonic frequency; for Effective value of subharmonic voltage; for RMS value of subharmonic current; for The initial phase of the second harmonic.

[0048] For example, in a specific calculation, by performing spectral analysis on the converter output current, the main harmonic components are identified as follows: the 11th harmonic, with a frequency of 550 Hz and an effective current value of 150 A; the 13th harmonic, with a frequency of 650 Hz and an effective current value of 75 A; and the 23rd harmonic, with a frequency of 1150 Hz and an effective current value of 30 A. These harmonic components serve as the harmonic excitation input for the subsequent multiphysics coupled simulation model.

[0049] It should be noted that harmonic components include not only amplitude but also frequency and initial phase. Frequency affects the skin depth of the winding, proximity effect, and core eddy current loss; amplitude affects copper loss, iron loss, and heat source intensity; initial phase affects the instantaneous electromagnetic field distribution when multiple harmonics are superimposed. Combining this information into a harmonic excitation input model can avoid the simplification of operating conditions caused by describing the harmonic source using only the total harmonic distortion rate or a single amplitude coefficient.

[0050] In step S103, each harmonic component is used as the excitation input to the multi-physics coupling simulation model to obtain the electromagnetic field distribution and temperature field distribution of the converter transformer under the action of each harmonic.

[0051] In one specific implementation, a corresponding time-domain harmonic excitation is constructed based on each harmonic component and applied to the winding port or the valve side boundary of the converter transformer. The time-domain excitation corresponding to the subharmonic can be expressed as: in, for Time-domain voltage excitation or time-domain current excitation corresponding to the subharmonic; for The effective value of voltage or effective value of current corresponding to the subharmonic; for Second harmonic frequency; For time.

[0052] For cases where the contribution of each harmonic needs to be analyzed separately, frequency domain solutions can be performed for each target harmonic to obtain the magnetic vector potential, magnetic flux density distribution, current density distribution and electromagnetic force distribution under the action of the target harmonic; for cases where multiple harmonics need to be analyzed together, multiple target harmonics can be simultaneously input into the model in a time-domain superposition manner, and the field distribution corresponding to each frequency component can be extracted after the solution.

[0053] In one implementation, the frequency domain equations in the electromagnetic field sub-model can be expressed as: in, For curl operator; for Resistivity under the influence of subharmonics, and ; for Equivalent permeability under the action of subharmonics; for The magnetic vector potential corresponding to the subharmonic; The imaginary unit; for The angular frequency corresponding to the subharmonic; for Equivalent conductivity under the influence of subharmonics; for Current density corresponding to the subharmonic; After obtaining the electromagnetic field distribution, the Joule heat source of the winding is determined based on the current density distribution in the winding region, and the iron loss heat source is determined based on the magnetic flux density distribution in the core region. The heat sources corresponding to each harmonic are then superimposed to form the input heat source for the temperature field sub-model. The total heat source can be expressed as: in, For spatial location The total heat source at the location; For the target harmonic order set; For spatial location Cause The winding Joule heat source generated by the subharmonic; For spatial location Cause Iron loss heat source generated by subharmonics; It should be noted that the purpose of the frequency domain equations described above is to directly incorporate harmonic frequencies into the electromagnetic field calculation, allowing the current density and magnetic flux density distributions at different frequencies to be obtained separately. As the frequency increases, the current density in the winding region may concentrate towards the conductor surface, and the magnetic flux distribution in the core region will also change, resulting in the heat source distribution no longer being the same as that at the power frequency. By inputting the heat sources of each harmonic into the temperature field sub-model, a temperature field distribution that more closely matches the harmonic operating conditions can be obtained, rather than treating the entire converter transformer as a uniform heat source.

[0054] In one specific implementation, a first preset convergence condition is used to determine whether the temperature field distribution and deformation information in the multiphysics coupled simulation model have stabilized. Specifically, in the first... After the first iteration, the temperature field distribution and deformation information obtained in this iteration are compared with those obtained in the second iteration. The results of the next iteration are compared; when the maximum change in temperature field distribution in two adjacent iterations is not greater than the first temperature convergence threshold, and the maximum change in deformation information is not greater than the first deformation convergence threshold, it is determined that the first preset convergence condition is met.

[0055] The first presupposed convergence condition can be expressed as: in, For the first Spatial position of the next iteration Temperature at that location; This is the calculation region for the temperature field; This is the first temperature convergence threshold; For the first Spatial position of the next iteration Deformation vector at; The calculation region for the mechanical field; This is the first deformation convergence threshold.

[0056] It should be noted that the first preset convergence condition is mainly used to avoid outputting calculation results prematurely before the electromagnetic field, temperature field, and mechanical field have reached a stable state. For example, when the temperature field is still changing significantly, the winding resistivity and core permeability will continue to change; when the deformation information is still changing significantly, the local geometric position between the winding and the core will also affect the leakage magnetic field distribution. Therefore, by simultaneously limiting the amount of temperature change and deformation, it can be ensured that subsequent high-frequency resistance correction, core magnetic property update, and loss calculation are all based on a relatively stable multi-physics field state. In engineering calculations, the first temperature convergence threshold can be taken as 0.5℃ to 1℃ according to the temperature rise assessment accuracy, and the first deformation convergence threshold can be set according to the finite element mesh size or the allowable error of structural deformation. The above values ​​do not constitute a limitation on the scope of protection of this invention.

[0057] In step S104, the high-frequency resistance of the converter transformer winding is corrected according to the temperature field distribution, electromagnetic field distribution and each harmonic component, so as to obtain the corrected high-frequency resistance corresponding to each harmonic.

[0058] In one specific implementation, the equivalent temperature of the winding conductor region is first extracted based on the temperature field distribution. Since the winding temperature is typically not uniformly distributed, taking only the temperature of a single point as the winding temperature can easily cause the resistance correction result to be affected by local hot spots or local low temperatures. Therefore, a volume average can be performed on the winding conductor region to obtain the winding equivalent temperature: in, This is the equivalent temperature of the winding conductor region; This refers to the winding conductor region; For spatial location Temperature at that location; Temperature correction resistor; Reference temperature The power frequency winding resistance is below; The temperature coefficient of resistance of the winding conductor; For reference temperature; Subsequently, the power frequency winding resistance at the reference temperature is corrected based on the winding equivalent temperature. in, for Corrected high-frequency resistors corresponding to subharmonics; This represents the geometric cross-sectional area of ​​the winding conductor; for The proximity effect correction coefficient corresponding to the subharmonic is determined by the degree of non-uniformity of winding current density in the electromagnetic field distribution and the leakage magnetic coupling strength between adjacent conductors.

[0059] After obtaining the temperature-corrected resistor, the skin depth is calculated based on each harmonic frequency. The skin depth corresponding to the subharmonic can be expressed as: in, for Skin depth corresponding to the subharmonic; The resistivity of the winding conductor at the equivalent temperature; for The angular frequency corresponding to the subharmonic; for Equivalent permeability of the region where the winding conductor is located under the action of subharmonics.

[0060] When the skin depth is smaller than the equivalent cross-sectional area of ​​the conductor, the current density inside the conductor is no longer uniformly distributed. In this case, the high-frequency resistance needs to be corrected based on the effective conductive area. For an equivalent circular conductor cross-section, the effective conductive area can be expressed as: in, for The effective conductive area corresponding to the subharmonic; This is the equivalent diameter of the winding conductor; This is the error function.

[0061] Furthermore, based on the current density distribution in the winding region and the leakage magnetic field distribution in adjacent windings in the electromagnetic field distribution, the following is determined: The proximity effect correction coefficient for subharmonics When the leakage magnetic field between adjacent conductors is strong or the conductors are closely packed, The value of is increased accordingly to reflect the current distribution distortion caused by the proximity effect. Ultimately, The corrected high-frequency resistance corresponding to the subharmonic can be expressed as: in, for Corrected high-frequency resistors corresponding to subharmonics; This represents the geometric cross-sectional area of ​​the winding conductor; for The proximity effect correction coefficient corresponding to the subharmonic is determined by the degree of non-uniformity of winding current density in the electromagnetic field distribution and the leakage magnetic coupling strength between adjacent conductors.

[0062] For example, in a specific calculation, the power frequency winding resistance Given a reference temperature of 20℃ and an operating winding equivalent temperature of 75℃, and assuming a copper temperature coefficient of 0.00393 / ℃, the temperature correction resistor is approximately 0.365mΩ. For the 11th harmonic at a frequency of 550Hz, considering the skin effect and proximity effect, the high-frequency correction resistor differs significantly from the power frequency temperature correction resistor. This difference indicates that under high-frequency harmonic conditions, directly using the power frequency resistor to calculate copper losses will distort the calculation results for winding harmonic copper losses.

[0063] It should be noted that the skin depth formula links harmonic frequencies to the effective conductive area of ​​a conductor; the effective conductive area formula further quantifies the area in the conductor cross-section that actually participates in conduction; and the proximity effect correction coefficient is used to compensate for the influence of leakage magnetic coupling between adjacent conductors on the current density distribution. These three parts work together to enable the high-frequency resistor to reflect not only temperature changes, but also frequency changes and changes in the local electromagnetic field.

[0064] In step S105, a nonlinear magnetic characteristic model of the converter transformer core is established or updated based on the electromagnetic field distribution, temperature field distribution, and core material parameters.

[0065] In one specific implementation, the magnetic flux density distribution of the iron core region under the action of each harmonic is first extracted based on the electromagnetic field distribution. And magnetic field strength distribution, and extract the local temperature distribution of the iron core region based on the temperature field distribution. Then, based on the magnetization curve, hysteresis loop, core conductivity, lamination thickness, and reference permeability in the core material parameters, the core region is divided into an unsaturated region, a transition region, and a saturated region, and corresponding piecewise permeability functions are configured for each.

[0066] For different harmonic frequencies, the high-frequency nonlinear changes of the core material also need to be considered. When the harmonic frequency exceeds the preset high-frequency correction threshold, the equivalent permeability of the core material will decrease as the frequency increases. The equivalent nonlinear permeability corresponding to the second harmonic can be expressed as: in, For spatial location Place The equivalent nonlinear permeability corresponding to the subharmonic; The piecewise permeability function is determined by the core magnetization curve and temperature field distribution; For spatial location Place The magnetic flux density corresponding to the subharmonic; For spatial location Local temperature of the iron core at that location; for Frequency correction factor corresponding to subharmonics; The frequency correction factor can be determined as follows: when hour, ;when hour, in, for Second harmonic frequency; The preset high-frequency correction threshold is used; This is the frequency correction factor for the core material. For silicon steel sheet material, It can be determined based on material test data or broadband magnetization characteristic data provided by the manufacturer.

[0067] In a specific example, when the preset high-frequency correction threshold is 50Hz, When the value is 0.02, and the 11th harmonic frequency is 550Hz, the frequency correction factor is approximately 0.94. This result indicates that the equivalent permeability of the iron core decreases relative to the power frequency condition at this frequency. Introducing this frequency correction factor into the nonlinear magnetic characteristic model of the iron core can avoid the calculation deviation of iron loss caused by directly extrapolating high-frequency harmonic conditions from the power frequency magnetization curve.

[0068] It should be noted that the piecewise nonlinear magnetic property model does not simply assign a fixed permeability to the entire iron core. Instead, it determines the equivalent permeability of different regions based on the local magnetic flux density, local temperature, and harmonic frequencies of the iron core. In this way, in regions of the iron core that are close to saturation or have high temperature rise, the model automatically reflects the impact of the decrease in the material's magnetization on the magnetic field distribution and loss calculation.

[0069] In one specific implementation, the second preset convergence condition is used to determine whether the copper loss and iron loss corresponding to each harmonic have stabilized. Specifically, in the first... After calculating the copper and iron losses corresponding to each harmonic in the second iteration, the results of this iteration are compared with those of the third iteration. The results of the iterations are compared; when the relative changes in copper loss and iron loss corresponding to all target harmonic orders are not greater than the second loss convergence threshold, it is determined that the copper loss and iron loss corresponding to each harmonic order meet the second preset convergence condition.

[0070] The second presupposed convergence condition can be expressed as: in, The set of target harmonic orders for loss calculation; For the first During the next iteration Copper loss corresponding to subharmonics; For the first During the next iteration Iron loss corresponding to subharmonics; This is a preset loss baseline value used to avoid the denominator being zero; This is the second loss convergence threshold.

[0071] It should be noted that the second preset convergence condition focuses on whether the loss result itself is stable. Since the high-frequency resistance of the winding varies with temperature, skin effect, and proximity effect, and the iron loss of the core also varies with magnetic flux density, temperature, and equivalent permeability, if only the convergence of the temperature field is used as the criterion, there may be situations where the temperature change is small but the loss of a certain high-frequency harmonic still fluctuates significantly. Therefore, this embodiment further sets convergence criteria for the copper loss and iron loss of each harmonic, which can make the final output total harmonic loss more stable. In actual calculations, the second loss convergence threshold can be taken as 0.5% to 2%, and can also be adjusted according to engineering accuracy requirements.

[0072] In step S106, the copper loss, iron loss, and total harmonic loss corresponding to each harmonic are calculated.

[0073] In one specific implementation, the copper loss of the winding under the action of each harmonic is calculated based on the effective value of the harmonic current in each harmonic component and the corrected high-frequency resistance corresponding to each harmonic. The copper loss corresponding to the second harmonic can be expressed as: in, for The winding copper loss corresponding to the subharmonic; for RMS value of subharmonic current; for Corrected high-frequency resistors corresponding to subharmonics; For core iron loss, the magnetic flux density distribution of the core region under each harmonic action is extracted based on the electromagnetic field distribution, and the hysteresis loss parameters, eddy current loss parameters, frequency correction factor and equivalent nonlinear permeability in the nonlinear magnetic characteristic model are called to calculate the core iron loss under each harmonic action. The iron loss corresponding to the second harmonic can be expressed as: in, for Core iron loss corresponding to subharmonics; For the iron core region; This is the hysteresis loss coefficient; for Second harmonic frequency; For spatial location Place The magnetic flux density corresponding to the subharmonic; The hysteresis loss index; The eddy current loss coefficient is used. The thickness of the iron core laminations; It should be noted that the copper loss formula reflects the combined effects of the effective value of harmonic current and the corrected high-frequency resistance on winding losses. The corrected high-frequency resistance already incorporates the effects of temperature, skin effect, and proximity effect. In the iron loss formula, the first term describes hysteresis loss, and the second term describes eddy current loss; since the magnetic flux density is spatially distributed... Therefore, it can reflect the situation of concentrated loss in local areas of the iron core, rather than estimating the overall iron loss by only the average magnetic flux density.

[0074] After calculating the copper and iron losses corresponding to each harmonic, the coupling relationship between different harmonic orders is determined based on the electromagnetic field distribution. Specifically, the coupling relationship between different harmonic orders can be extracted based on the magnetic field overlap distribution, current density coupling distribution, and equivalent impedance matrix in the winding and core regions. Second harmonics and Mutual impedance between subharmonics and extract Subharmonics and Self-impedance corresponding to subharmonics The harmonic coupling coefficient can be expressed as: in, for Second harmonics and Harmonic coupling coefficient between subharmonics; for Second harmonics and Mutual impedance between subharmonics; for Self-impedance corresponding to subharmonics; for Self-impedance corresponding to subharmonics; is the real part of the mutual impedance; The total harmonic loss of the converter transformer can be expressed as: in, This refers to the total harmonic losses of the converter transformer; The set of target harmonic orders for loss calculation; for Copper loss corresponding to subharmonics; for Iron loss corresponding to subharmonics.

[0075] It should be noted that the harmonic coupling coefficient is used to reflect the not entirely independent relationship between different harmonic orders. For example, the 11th and 13th harmonics may have coupling effects in the winding leakage magnetic field and the local magnetization state of the core. If the losses of the 11th, 13th, and other harmonics are simply added together, the contribution of this interaction to the total loss cannot be reflected. After determining the coupling coefficient through mutual impedance and self-impedance, the inter-harmonic coupling loss term can be added to the total loss calculation, thereby reducing the possibility of underestimation or overestimation of the total harmonic loss under complex harmonic operating conditions.

[0076] In a specific example, the main harmonics are the 11th, 13th, and 23rd harmonics. The effective value of the 11th harmonic current is 150A, the 13th harmonic current is 75A, and the 23rd harmonic current is 30A. After high-frequency resistor correction, the corresponding copper losses are calculated; the corresponding iron losses are calculated using the nonlinear magnetic characteristic model of the iron core. If the coupling coefficient between the 11th and 13th harmonics is taken as 0.02, the coupling loss between them must also be included in the total harmonic loss. The total harmonic loss obtained in this way includes not only the copper and iron losses generated by each harmonic itself, but also the additional losses caused by the interaction between harmonics.

[0077] In this embodiment, the copper and iron losses corresponding to each harmonic can be re-inputted as heat sources into the temperature field sub-model to update the temperature field distribution. The updated temperature field distribution is then used to correct the high-frequency resistance of the winding and the nonlinear magnetic properties of the core, and the losses of each harmonic are recalculated. If the changes in the temperature field distribution, corrected high-frequency resistance, equivalent nonlinear permeability, or total harmonic loss obtained from the previous two iterations are less than a preset threshold, the calculation is considered to have met the convergence condition, and the final total harmonic loss, copper losses of each harmonic, iron losses of each harmonic, temperature field distribution, and electromagnetic field distribution are output.

[0078] For example, a convergence criterion can be set if the relative change in total harmonic loss between two adjacent iterations is less than 1%, or if the change in the highest winding temperature is less than 1°C. In practical applications, these thresholds can be adjusted according to the required accuracy of engineering calculations.

[0079] The method described in this embodiment moves beyond simply calculating the linear superposition of power frequency material parameters and harmonic losses. Instead, it incorporates harmonic source characteristics, multi-physics feedback, winding high-frequency resistance, core nonlinear magnetic characteristics, and inter-harmonic coupling relationships into the calculation process. This method provides more detailed calculation results for converter transformer design verification, operation optimization, temperature rise assessment, and loss analysis.

[0080] In one specific implementation, a third preset convergence condition is used to determine whether the equivalent nonlinear permeability of the core region has stabilized. Specifically, after updating the nonlinear magnetic property model of the core based on the electromagnetic field distribution, temperature field distribution, and core material parameters, the third... The equivalent nonlinear permeability obtained in the second iteration is the same as that obtained in the third iteration. The equivalent nonlinear permeability obtained from the next iteration is compared; when the relative change in equivalent nonlinear permeability in all target harmonic orders and the core region is not greater than the third permeability convergence threshold, the third preset convergence condition is determined to be satisfied.

[0081] The third pre-defined convergence condition can be expressed as: in, The set of target harmonic orders for loss calculation; For the iron core region; For the first Spatial position of the next iteration Place The equivalent nonlinear permeability corresponding to the subharmonic; This is a preset permeability reference value used to avoid the denominator being zero; This is the third permeability convergence threshold.

[0082] It should be noted that the third preset convergence condition is mainly used to constrain the iterative update of the nonlinear magnetic characteristic model of the iron core. For high-frequency harmonics and local saturation regions, the equivalent nonlinear permeability of the iron core is quite sensitive to changes in magnetic flux density and temperature. If the equivalent nonlinear permeability is not yet stable, directly calculating iron loss may lead to fluctuations in the results of hysteresis loss and eddy current loss. By setting the third preset convergence condition, it can be ensured that the piecewise permeability function, frequency correction factor, and related loss parameters used for iron loss calculation have reached a relatively stable state. In practical applications, the third permeability convergence threshold can be set to 0.5% to 2%, or it can be set according to the testing accuracy of the iron core material and the accuracy of the finite element calculation.

[0083] Example 2: The parts not mentioned in this embodiment are the same as in Embodiment 1.

[0084] This embodiment uses a ±800kV converter transformer as an example to further illustrate the method for calculating harmonic losses of converter transformers based on multi-physics coupling provided by this invention. The equipment parameters, model parameters, harmonic data, and calculation results in this embodiment are only used to illustrate the specific application process of the method of this invention and do not constitute a limitation on the scope of protection of this invention. Under other voltage levels, other capacities, other winding structures, or other converter topologies, the harmonic losses of converter transformers can also be calculated according to the process shown in this embodiment.

[0085] The converter transformer has a capacity of 500MVA and a rated voltage of ±800kV / 500kV. The high-voltage side winding has 1200 turns, and the low-voltage side winding has 750 turns. The winding conductors are copper, with a wire specification of 2×10mm. 2 The insulation thickness is 2mm. The core is made of 30ZH130 silicon steel sheets, and the magnetic permeability of the core material is 1.5×10⁻⁶. 4 H / m, conductivity 5×10 6 S / m, loss tangent of 0.003, core cross-sectional area of ​​0.8m² 2 The lamination gap is 0.1mm.

[0086] First, a multiphysics coupling simulation model of the converter transformer was established. Specifically, an electromagnetic field sub-model was built using the electromagnetic field finite element simulation module, with the winding, core, insulation structure, tank, and surrounding medium treated as different computational regions. Electromagnetic parameters such as conductivity, permeability, and dielectric constant were set according to the material properties of different regions. To better reflect the current density concentration phenomenon in the winding region under high-frequency harmonics, the mesh size of the winding region was set to 0.1 mm; to balance the calculation accuracy of the core flux distribution and the overall computational efficiency, the mesh size of the core region was set to 0.5 mm.

[0087] Simultaneously, a mechanical field sub-model is established using the mechanical field finite element simulation module, and the electromagnetic force output from the electromagnetic field sub-model is used as the mechanical field load. Specifically, it can be done according to... Estimate the electromagnetic force on the winding, where the magnetic flux density is... Take 1.2T, current Take 3000A as the equivalent stress length. A value of 10m was used. A temperature field sub-model was established using the fluid-temperature field simulation module, setting the convective heat transfer coefficient of the oil medium to 1000 W / (m²). 2The ambient temperature was 25℃. During the calculation, the Joule heat source and iron loss heat source output by the electromagnetic field sub-model were used as inputs to the temperature field sub-model. The temperature field distribution output by the temperature field sub-model was used to correct the winding resistivity and core permeability. The deformation information output by the mechanical field sub-model was used to correct the local geometric position, thus forming a coupled calculation relationship between the electromagnetic field, temperature field and mechanical field.

[0088] It should be noted that in the calculation of harmonic losses in converter transformers, the mesh size for both the winding and core regions should not be too coarse. If the mesh size for the winding region is too coarse, the local current density concentration caused by the skin effect and proximity effect will be difficult to reflect; if the mesh size for the core region is too coarse, local magnetic flux density variations and concentrated core loss areas will easily be averaged out. This embodiment uses different mesh sizes for the winding and core regions, which can improve the stability of local loss calculation while ensuring computational efficiency.

[0089] Subsequently, harmonic source signals during the converter process are acquired. In this embodiment, the converter output voltage and current are simulated using an electromagnetic transient simulation platform, or the voltage and current time-domain signals at the valve side of the converter transformer are acquired using a field measurement system. A Fast Fourier Transform (FFT) is then performed on the obtained time-domain signals to obtain the main harmonic components at the valve side of the converter transformer. Analysis results show that, under this operating condition, the main harmonics include the 11th, 13th, and 23rd harmonics. The 11th harmonic has a frequency of 550Hz and an effective current value of 150A; the 13th harmonic has a frequency of 650Hz and an effective current value of 75A; and the 23rd harmonic has a frequency of 1150Hz and an effective current value of 30A. The total harmonic distortion rate under this operating condition is 4.8%.

[0090] After obtaining each harmonic component, the harmonic components are used as harmonic excitation inputs to the multiphysics coupling simulation model.

[0091] It should be noted that the harmonic excitation described above does not simply use an equivalent current value, but rather uses the frequency, amplitude, and phase of different harmonics as inputs. The purpose of this is that different frequency harmonics correspond to different skin depths, different core eddy current loss levels, and different magnetic field distributions; if only the total harmonic distortion rate or equivalent current is used for calculation, it is impossible to distinguish the different contributions of the 11th, 13th, and 23rd harmonics to the losses.

[0092] Then, the high-frequency resistance of the winding is corrected. First, the power frequency winding resistance at 20℃ is measured. , measured Based on the output of the temperature field sub-model, the winding operating temperature is taken as 75℃, and the temperature coefficient of resistance of the copper conductor is... Taking 0.00393 / ℃, the temperature correction resistor is: in, Temperature correction resistor; Subsequently, skin effect correction was performed using the 11th harmonic as an example. The 11th harmonic frequency is 550Hz, and the resistivity of the copper conductor is taken as... Permeability according to Calculations show that for the 550Hz harmonic, the skin depth is approximately 0.21 mm.

[0093] Furthermore, the effective conductive area is determined based on the conductor's equivalent diameter and skin depth. In this embodiment, the effective conductive area at 550Hz is approximately 0.000314m². 2 The high-frequency resistance calculated based on this effective conductive area is approximately 0.185 Ω.

[0094] When considering the proximity effect, based on the winding arrangement and the leakage magnetic field coupling between adjacent conductors, the proximity effect correction factor for the 11th harmonic is set to 1.3. After the proximity effect correction, the corrected high-frequency resistance for the 11th harmonic is approximately 0.142Ω.

[0095] It should be noted that the temperature correction mainly reflects the effect of operating temperature rise on winding resistivity; the skin effect correction mainly reflects the change in effective conductive area caused by uneven distribution of high-frequency current within the conductor cross-section; and the proximity effect correction mainly reflects the effect of leakage magnetic field coupling between adjacent conductors on local current density. These three factors work together to make the corrected high-frequency resistance closer to the equivalent resistance when harmonic current actually flows through the winding.

[0096] Next, the nonlinear magnetic characteristic model of the iron core is established and updated. In this embodiment, based on experimental data from 30ZH130 silicon steel sheets, the hysteresis loss coefficient is selected. Hysteresis loss index eddy current loss coefficient Thickness of iron core laminations Taking the 11th harmonic as an example, its frequency is 550Hz, and the initial value of the magnetic flux density is taken as... The hysteresis loss can be calculated using the following formula: in, For the first Hysteresis loss corresponding to the second harmonic; This is the hysteresis loss coefficient; For the first Second harmonic frequency; For the first The magnetic flux density corresponding to the subharmonic; This represents the hysteresis loss exponent. Substituting the above parameters, the hysteresis loss corresponding to the 11th harmonic is approximately... .

[0097] Eddy current losses can be calculated using the following formula: in, For the first Eddy current loss corresponding to subharmonics; The eddy current loss coefficient is used. For the first Second harmonic frequency; For the first The magnetic flux density corresponding to the subharmonic; The thickness of the silicon steel sheet lamination; Let be the effective volume of the iron core. Substituting the parameters, the eddy current loss corresponding to the 11th harmonic is approximately... .

[0098] Further consideration is given to high-frequency nonlinearity correction. As the frequency increases, the equivalent permeability of the silicon steel sheet decreases; therefore, this embodiment introduces a frequency correction factor. For the 550Hz harmonic, take the frequency correction factor. If the reference frequency is taken as 50Hz, then: The corrected magnetic flux density is: By recalculating the hysteresis loss and eddy current loss based on the corrected magnetic flux density, a core loss that better reflects the state of the core material at this harmonic frequency can be obtained. After correction, the increase in iron loss in this embodiment is approximately 5%.

[0099] It should be noted that directly calculating core losses at 550Hz, 650Hz, or 1150Hz using only power frequency magnetization characteristics or low-frequency loss parameters will neglect the changes in the equivalent permeability of the core at high frequencies and the enhancement of eddy current effects. By introducing a frequency correction factor, the nonlinear magnetic characteristic model of the core can be updated with changes in harmonic frequency, thereby improving the applicability of core loss calculations.

[0100] Subsequently, the winding copper loss is calculated based on the effective value of each harmonic current and the corresponding corrected high-frequency resistance.

[0101] In this embodiment, the 11th harmonic copper loss is approximately: The copper loss of the 13th harmonic is approximately: The copper loss of the 23rd harmonic is approximately: Therefore, the total copper loss is approximately 4183W.

[0102] The iron core loss is obtained by superimposing the hysteresis loss and eddy current loss corresponding to each harmonic. Taking the 11th harmonic as an example, the corrected iron loss is approximately 2.06W; combined with other major harmonic components, the total iron loss in this embodiment is approximately 15W. It should be noted that in this example, copper loss accounts for the main part of the total harmonic loss, but iron loss still affects the temperature field distribution and the judgment of local hot spots in the iron core, so it cannot be directly ignored in multiphysics coupling calculations.

[0103] Furthermore, the inter-harmonic coupling effect is considered. In this embodiment, the coupling coefficient between the 11th and 13th harmonics is taken as 0.02. According to the harmonic coupling relationship, the coupling loss between the 11th and 13th harmonics is approximately 32.6W. Adding the total copper loss, total iron loss, and inter-harmonic coupling loss, the total harmonic loss of the converter transformer in this embodiment is approximately: It should be noted that simply adding the copper and iron losses corresponding to the 11th, 13th, and 23rd harmonics directly would not reflect the mutual influence between the 11th and 13th harmonics caused by leakage magnetic fields, local magnetization of the iron core, or equivalent impedance. This embodiment introduces a harmonic coupling loss term, so that the final total harmonic loss includes not only the losses of each harmonic itself, but also the additional losses caused by the combined effect of different harmonics.

[0104] Finally, the calculation results were verified. Under the same operating conditions, harmonic measurement devices and power measurement devices were installed on the high-voltage side of the converter transformer to measure the actual harmonic loss of the converter transformer. The harmonic loss measured on-site was approximately 4180W, while the total harmonic loss calculated in this embodiment was approximately 4230.6W, with a relative error of approximately 1.2%.

[0105] The verification results above show that the calculation process provided in this embodiment can better reflect the loss level of the converter transformer under actual harmonic operating conditions. Compared with methods that only use total harmonic distortion rate estimation, power frequency resistance calculation, or simple superposition of harmonic losses, this embodiment, through multi-physics coupling, high-frequency resistance correction, core nonlinear magnetic characteristic correction, and harmonic coupling loss calculation, can more accurately obtain the total harmonic loss and its composition of the converter transformer.

[0106] In practical engineering applications, the system can also output, as needed, the copper loss, iron loss, winding temperature distribution, core temperature distribution, local loss density distribution, and harmonic coupling loss ratio corresponding to each harmonic. For example, when the copper loss corresponding to a certain harmonic is significantly large, the concentrated area of ​​winding current density under that harmonic can be further analyzed; when the local iron loss of the core is high, the magnetic flux density distribution can be combined to determine whether there is local saturation or high-frequency eddy current concentration. Therefore, this embodiment can not only provide the total loss value, but also provide more detailed analytical basis for the structural design, operating condition optimization, and loss assessment of converter transformers.

[0107] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0108] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for calculating harmonic losses in a converter transformer based on multi-physics coupling, characterized in that, Includes the following steps: Establish a multi-physics coupling simulation model for the converter transformer; Harmonic source signals during the commutation process are acquired, and the harmonic source signals are subjected to spectral analysis to obtain each harmonic component. By using the harmonic components as excitation inputs to the multiphysics coupling simulation model, the electromagnetic field distribution, temperature field distribution, and deformation information of the converter transformer under the action of each harmonic are obtained. Based on the temperature field distribution, electromagnetic field distribution and harmonic components, the high-frequency resistance of the converter transformer winding is corrected to obtain the corrected high-frequency resistance corresponding to each harmonic. Based on the electromagnetic field distribution, temperature field distribution, and core material parameters, establish or update the nonlinear magnetic characteristic model of the converter transformer core. Based on the harmonic components, the corrected high-frequency resistor, and the nonlinear magnetic characteristic model, the copper loss and iron loss corresponding to each harmonic are calculated in the multiphysics coupling simulation model. The coupling relationship between each harmonic is determined based on the electromagnetic field distribution, and the copper loss and iron loss corresponding to each harmonic are corrected and superimposed based on the coupling relationship to obtain the total harmonic loss of the converter transformer.

2. The method for calculating harmonic losses of a converter transformer based on multi-physics coupling according to claim 1, characterized in that, The establishment of the multi-physics coupling simulation model for the converter transformer specifically includes: Based on the geometric and electromagnetic parameters of the windings, core, insulation structure, tank, and cooling medium of the converter transformer, an electromagnetic field sub-model is established based on Maxwell's equations. The electromagnetic field sub-model is used to receive each harmonic component as excitation, and outputs the magnetic flux density distribution, current density distribution, Joule heat source, iron loss heat source, and electromagnetic force distribution under the action of each harmonic after inputting each harmonic component. Based on the Joule heat source and iron loss heat source output by the electromagnetic field sub-model, a temperature field sub-model is established based on the heat conduction equation, and combined with the convective heat transfer boundary conditions between the cooling medium and the heat-generating components of the converter transformer, the temperature field distribution of the winding, core and surrounding medium is obtained. Based on the electromagnetic force distribution output by the electromagnetic field sub-model and the temperature field distribution output by the temperature field sub-model, a mechanical field sub-model is established. The electromagnetic force distribution is used as the force load, and the thermal expansion caused by the temperature field distribution is used as the thermal load to obtain the mechanical stress distribution and deformation information of the winding and the core. The temperature field distribution is fed back to the electromagnetic field sub-model to update the winding resistivity and core permeability, and the deformation information is fed back to the electromagnetic field sub-model to update the local geometric position and local electromagnetic field distribution of the winding, core and surrounding medium, thereby forming a coupled simulation model between electromagnetic field, temperature field and mechanical field.

3. The method for calculating harmonic losses of a converter transformer based on multi-physics coupling according to claim 2, characterized in that, The heat conduction equation satisfies: in, The density of the material; Specific heat capacity of the material; For temperature; For time; Thermal conductivity; It is a Joule heat source generated by the winding current; It is the heat source of iron loss generated by iron core loss; The thermal strain in the mechanical field sub-model satisfies: in, Thermal strain; The coefficient of linear expansion of the material; The temperature output by the temperature field submodel; This is a reference temperature.

4. The method for calculating harmonic losses of a converter transformer based on multi-physics coupling according to claim 1, characterized in that, The acquisition of harmonic source signals during the commutation process, and the spectral analysis of the harmonic source signals to obtain each harmonic component, specifically includes: Based on the converter's pulsation number, firing angle, commutation angle, and AC system fundamental frequency, the range of harmonic orders to be analyzed during the switching process is determined, and the harmonics within the range of harmonic orders are used as the target harmonics to be input into the multiphysics coupling simulation model. The voltage time-domain signal and current time-domain signal of the AC side of the converter valve and / or the valve side of the converter transformer are acquired or simulated, and the DC component is removed, synchronous sampling and windowing are performed on the voltage time-domain signal and / or current time-domain signal to obtain the preprocessed harmonic source signal; A fast Fourier transform is performed on the preprocessed harmonic source signal to extract the harmonic frequency, effective value of harmonic voltage, effective value of harmonic current, and initial phase of each target harmonic. The harmonic frequency, the effective value of the harmonic voltage, the effective value of the harmonic current, and the initial phase of the harmonic are combined to form each harmonic component, and each harmonic component is used as the harmonic excitation input of the multiphysics coupling simulation model.

5. The method for calculating harmonic losses of a converter transformer based on multi-physics coupling according to claim 1, characterized in that, The process of using the harmonic components as excitation inputs to the multiphysics coupling simulation model to obtain the electromagnetic field distribution, temperature field distribution, and deformation information of the converter transformer under the action of each harmonic specifically includes: Based on the harmonic frequency, effective value of harmonic voltage, effective value of harmonic current, and initial phase of each harmonic component, a corresponding time-domain harmonic excitation is constructed, and the time-domain harmonic excitation is applied to the winding port or the valve side boundary of the converter transformer in the multi-physics coupling simulation model. The time-domain harmonic excitation is expressed as: in, for Time-domain voltage excitation or time-domain current excitation corresponding to the subharmonic; for The effective value of voltage or effective value of current corresponding to the subharmonic; for Second harmonic frequency; for The initial phase of the second harmonic; For each target harmonic, the frequency domain is solved in the electromagnetic field sub-model to obtain the magnetic vector potential, magnetic flux density distribution, current density distribution and electromagnetic force distribution under the action of the target harmonic; Wherein, the frequency domain solution satisfies: in, For curl operator; for Resistivity under the influence of subharmonics, and ; for Equivalent permeability under the action of subharmonics; for The magnetic vector potential corresponding to the subharmonic; The imaginary unit; for The angular frequency corresponding to the subharmonic; for Equivalent conductivity under the influence of subharmonics; for Current density corresponding to the subharmonic; Based on the current density distribution and magnetic flux density distribution output by the electromagnetic field sub-model, the winding Joule heat source and iron loss heat source corresponding to each target harmonic are determined respectively, and the winding Joule heat source and iron loss heat source corresponding to each target harmonic are superimposed to obtain the total heat source input to the temperature field sub-model. The total heat source is represented as: in, For spatial location The total heat source at the location; For the target harmonic order set; For spatial location Cause The winding Joule heat source generated by the subharmonic; For spatial location Cause Iron loss heat source generated by subharmonics; The total heat source is input into the temperature field sub-model, and the temperature field distribution of the winding, core and surrounding medium is obtained by combining the convective heat transfer boundary conditions of the cooling medium. The electromagnetic force distribution output by the electromagnetic field sub-model and the temperature field distribution output by the temperature field sub-model are input into the mechanical field sub-model to solve for the mechanical stress distribution and deformation information of the winding and the core. The winding resistivity and core permeability are updated based on the temperature field distribution. The local geometric positions of the winding, core and surrounding medium are updated based on the deformation information. The electromagnetic field sub-model is then resolved until the temperature field distribution and deformation information obtained from two adjacent iterations satisfy the first preset convergence condition. The electromagnetic field distribution and temperature field distribution of the converter transformer under each harmonic are then output.

6. The method for calculating harmonic losses of a converter transformer based on multi-physics coupling according to claim 1, characterized in that, The step of correcting the high-frequency resistance of the converter transformer winding based on the temperature field distribution, electromagnetic field distribution, and harmonic components to obtain the corrected high-frequency resistance corresponding to each harmonic specifically includes: Based on the temperature field distribution, the equivalent temperature of the conductor region of the converter transformer winding is extracted, and the power frequency resistance of the winding at the reference temperature is corrected according to the equivalent temperature to obtain the temperature-corrected resistance. The equivalent temperature and the temperature correction resistor are defined by the following formulas: in, This is the equivalent temperature of the winding conductor region; This refers to the winding conductor region; For spatial location Temperature at that location; Temperature correction resistor; Reference temperature The power frequency winding resistance is below; The temperature coefficient of resistance of the winding conductor; For reference temperature; Based on the harmonic frequencies in each harmonic component and the resistivity of the winding conductor at the equivalent temperature, the skin depth corresponding to each harmonic is determined, and the effective conductive area corresponding to each harmonic is determined based on the skin depth and the equivalent cross-sectional size of the winding conductor. The skin depth and the effective conductive area are defined by the following formulas: in, for Skin depth corresponding to the subharmonic; The resistivity of the winding conductor at the equivalent temperature; for The angular frequency corresponding to the subharmonic; for Equivalent permeability of the region where the winding conductor is located under the action of subharmonics; for The effective conductive area corresponding to the subharmonic; This is the equivalent diameter of the winding conductor; It is the error function; Based on the current density distribution in the winding region and the leakage magnetic field distribution in the adjacent winding in the electromagnetic field distribution, the proximity effect correction coefficient corresponding to each harmonic is determined, and the correction high-frequency resistance corresponding to each harmonic is obtained based on the temperature correction resistance, the effective conductive area and the proximity effect correction coefficient. Wherein, the corrected high-frequency resistor satisfies: in, for Corrected high-frequency resistors corresponding to subharmonics; This represents the geometric cross-sectional area of ​​the winding conductor; for The proximity effect correction coefficient corresponding to the subharmonic is determined by the degree of non-uniformity of winding current density in the electromagnetic field distribution and the leakage magnetic coupling strength between adjacent conductors.

7. The method for calculating harmonic losses of a converter transformer based on multi-physics coupling according to claim 1, characterized in that, The step of establishing or updating the nonlinear magnetic characteristic model of the converter transformer core based on the electromagnetic field distribution, temperature field distribution, and core material parameters specifically includes: Based on the electromagnetic field distribution, the magnetic flux density distribution and magnetic field strength distribution of the iron core region under the action of each harmonic are extracted, and based on the temperature field distribution, the local temperature distribution corresponding to the iron core region is extracted. Based on the magnetization curve, hysteresis loop, core conductivity, lamination thickness, and reference permeability in the core material parameters, the core region is divided into an unsaturated region, a transition region, and a saturated region, and corresponding piecewise permeability functions are configured for the unsaturated region, the transition region, and the saturated region, respectively. The piecewise permeability function is temperature-corrected based on the local temperature distribution, and the piecewise permeability function after temperature correction is frequency-corrected based on the harmonic frequencies in each harmonic component to obtain the equivalent nonlinear permeability corresponding to each harmonic. Based on the equivalent nonlinear permeability, the material properties of the iron core region in the electromagnetic field sub-model are updated, and the magnetic flux density distribution and magnetic field strength distribution of the iron core region are re-extracted based on the updated electromagnetic field distribution. When the change in equivalent nonlinear permeability of the core region before and after the update satisfies the second preset convergence condition, the current piecewise permeability function, frequency correction factor, hysteresis loss parameter and eddy current loss parameter are used as the nonlinear magnetic characteristic model of the converter transformer core.

8. The method for calculating harmonic losses of a converter transformer based on multi-physics coupling according to claim 7, characterized in that, The equivalent nonlinear permeability is expressed as: in, For spatial location Place The equivalent nonlinear permeability corresponding to the subharmonic; The piecewise permeability function is determined by the core magnetization curve and temperature field distribution; For spatial location Place The magnetic flux density corresponding to the subharmonic; For spatial location Local temperature of the iron core at that location; for Frequency correction factor corresponding to subharmonics; The frequency correction factor satisfies: in, for Frequency correction factor corresponding to subharmonics; for Second harmonic frequency; The preset high-frequency correction threshold is used; This is the frequency correction factor for the core material.

9. The method for calculating harmonic losses of a converter transformer based on multi-physics coupling according to claim 1, characterized in that, The calculation of copper and iron losses corresponding to each harmonic in the multiphysics coupling simulation model based on the harmonic components, the corrected high-frequency resistance, and the nonlinear magnetic characteristic model specifically includes: Based on the effective value of the harmonic current in each harmonic component and the corrected high-frequency resistance corresponding to each harmonic, the copper loss of the winding under the action of each harmonic is calculated. Wherein, the copper loss is represented as: in, for The winding copper loss corresponding to the subharmonic; for RMS value of subharmonic current; for Corrected high-frequency resistors corresponding to subharmonics; Based on the electromagnetic field distribution, the magnetic flux density distribution of the core region under each harmonic is extracted, and the hysteresis loss parameter, eddy current loss parameter, frequency correction factor and equivalent nonlinear permeability in the nonlinear magnetic characteristic model are called to calculate the iron loss of the core under each harmonic. Wherein, the iron loss is expressed as: in, for Core iron loss corresponding to subharmonics; For the iron core region; This is the hysteresis loss coefficient; for Second harmonic frequency; For spatial location Place The magnetic flux density corresponding to the subharmonic; The hysteresis loss index; The eddy current loss coefficient is used. The thickness of the iron core laminations; The copper and iron losses corresponding to each harmonic are output to the temperature field sub-model of the multiphysics coupling simulation model to update the temperature field distribution. The updated temperature field distribution is then fed back to correct the high-frequency resistance of the winding and the nonlinear magnetic characteristic model of the iron core until the copper and iron losses corresponding to each harmonic meet the third preset convergence condition.

10. The method for calculating harmonic losses of a converter transformer based on multi-physics coupling according to claim 1, characterized in that, The process of determining the coupling relationship between each harmonic based on the electromagnetic field distribution, and correcting and superimposing the copper loss and iron loss corresponding to each harmonic based on the coupling relationship to obtain the total harmonic loss of the converter transformer, specifically includes: Based on the electromagnetic field distribution, the magnetic field overlap distribution, current density coupling distribution, and equivalent impedance matrix of different harmonic orders in the winding region and iron core region are extracted, and the mutual impedance and self impedance between different harmonic orders are determined by the equivalent impedance matrix. Based on the mutual impedance and self-impedance, determine the harmonic coupling coefficient between any two different harmonic orders; The harmonic coupling coefficient is expressed as: in, for Second harmonics and Harmonic coupling coefficient between subharmonics; for Second harmonics and Mutual impedance between subharmonics; for Self-impedance corresponding to subharmonics; for Self-impedance corresponding to subharmonics; is the real part of the mutual impedance; Based on the copper loss, iron loss and harmonic coupling coefficient corresponding to each harmonic, the inter-harmonic coupling loss is calculated, and the copper loss, iron loss and inter-harmonic coupling loss corresponding to each harmonic are superimposed to obtain the total harmonic loss of the converter transformer. The total harmonic loss is expressed as: in, This refers to the total harmonic losses of the converter transformer; The set of target harmonic orders for loss calculation; for Copper loss corresponding to subharmonics; for Iron loss corresponding to subharmonics.

Images