Temperature prediction method and device for high-frequency transformer

Abstract

The invention provides a temperature prediction method and device for a high-frequency transformer, and belongs to the field of transformers. The method comprises the following steps: acquiring a temperature field parameter of a target high-frequency transformer; calculating initial magnetic core loss and initial winding loss according to the temperature field parameters; inputting the initial magnetic core loss and the initial winding loss into a pre-constructed thermal network model to obtain the temperature of each node in the pre-constructed thermal network model; based on the initial temperature of each node, correcting the initial magnetic core loss and the initial winding loss, and inputting the corrected initial magnetic core loss and initial winding loss into a pre-constructed thermal network model to recalculate the temperature of each node; and when the temperature of each node satisfies a preset condition, obtaining a temperature prediction value of each electromagnetic component of the target high-frequency transformer. According to the method, the problem that the high-frequency transformer is inaccurate due to inaccurate modeling and excessive conditions in a traditional method can be solved, and the accurate and effective temperature prediction method for the high-frequency transformer is provided.

Description

Technical Field

[0001] The present invention relates to the technical field of transformers, and in particular to a temperature prediction method and device for a high-frequency transformer. Background Art

[0002] High power density DC-DC converters are widely used in medium and low voltage DC power distribution and renewable energy power generation. As a key component for voltage conversion and energy transmission in DC converters, the volume and loss of high-frequency transformers directly determine the power density and transmission efficiency of electrical equipment. Under the same power level conditions, the mass and volume of high-frequency transformers made of nanocrystalline iron cores in LLC resonant circuits are reduced by about 85% compared with low-frequency transformers made of silicon steel sheets in dual active bridge (DAB) converters. Compared with industrial frequency power transformers, the core structure using high saturation magnetic materials has the advantages of small size and high power density.

[0003] However, the core material and windings of the transformer will produce high-frequency eddy current effects and stronger parasitic effects, which will lead to increased losses. The hysteresis and eddy current losses of the magnetic core, as well as the winding losses caused by resistance and eddy current effects when the coil is energized are the main reasons for the heating of high-frequency transformers. Under high frequency and large capacity conditions, the heat dissipation area of ​​high-frequency transformers is greatly reduced, resulting in heat dissipation difficulties and severe temperature rise. Therefore, reasonable thermal management is essential. The prediction of temperature rise can not only indirectly reflect the loss characteristics of the magnetic components of high-frequency transformers, but also serve as a standard for measuring whether the transformer can operate stably for a long time. At present, the temperature field modeling methods of high-frequency transformers mainly include empirical formula method, numerical calculation method, artificial intelligence algorithm and thermal network modeling method.

[0004] As a traditional transformer temperature rise prediction method, the empirical formula method does not consider the influence of transformer multi-physical field coupling and internal complex structure, and cannot meet the requirements of high-power and high-frequency transformers for temperature rise calculation accuracy. Numerical calculation methods use numerical models such as finite element method and finite volume method to effectively characterize the temperature field distribution, but most existing studies are based on two-dimensional and three-dimensional simplified models to analyze the temperature field and fluid field around the transformer, which is difficult to fully reflect the multi-physical field distribution of high-frequency transformers, and it is difficult to balance the calculation accuracy and calculation cost. Artificial intelligence algorithms need to be based on a large number of training samples to extract complex mapping relationships, but transformers in operation are rarely equipped with hot spot temperature measurement devices, so it is difficult to obtain enough internal state information of the transformer to meet the complex mapping relationships of machine learning algorithms.

[0005] Although the thermal network modeling method reasonably analyzes the heat dissipation conditions and hot spot distribution at different boundaries, it cannot accurately characterize the heat flow path within the magnetic core, especially the edge and corner areas of the laminated strips with anisotropic thermal conductivity, and there is a problem of repeated calculation of the heat dissipation area and the heat conduction path. Summary of the invention

[0006] The embodiments of the present invention provide a method and device for predicting the temperature of a high-frequency transformer, so as to solve the problem that the current method for predicting the temperature of a high-frequency transformer has limitations, resulting in inaccurate temperature prediction.

[0007] In a first aspect, an embodiment of the present invention provides a temperature prediction method for a high-frequency transformer, comprising: Obtain temperature field parameters of the target high-frequency transformer; Calculate the initial core loss and initial winding loss based on the temperature field parameters; Inputting the initial core loss and the initial winding loss into a pre-built thermal network model, and obtaining the temperature of each node in the pre-built thermal network model; Based on the initial temperature of each node, the initial core loss and the initial winding loss are corrected, and the corrected initial core loss and the initial winding loss are input into a pre-built thermal network model to recalculate the temperature of each node; When the temperature of each node meets the preset conditions, the temperature prediction value of each electromagnetic component of the target high-frequency transformer is obtained.

[0008] In a possible implementation, the initial core loss and the initial winding loss are corrected based on the initial temperature of each node, including: Based on the initial temperature of each node, calculate the convection thermal resistance and radiation thermal resistance; The initial core losses and initial winding losses are corrected based on the convection thermal resistance and radiation thermal resistance.

[0009] In a possible implementation, the initial core loss and the initial winding loss are corrected according to the convection thermal resistance and the radiation thermal resistance, including: Based on the convection thermal resistance and radiation thermal resistance, calculate the current temperature of each corresponding node; According to the current temperature of each node, determine the current loss parameter and the current conductivity; Current loss parameters, to correct the initial core loss; The initial winding losses are corrected for the current conductivity.

[0010] In a possible implementation, based on the initial temperature of each node, the convection thermal resistance and the radiation thermal resistance are calculated, including: Based on the initial temperature of each node, determine the corresponding convective heat transfer coefficient and temperature difference; Based on the convection heat transfer coefficient, calculate the convection thermal resistance; Based on the temperature difference and the initial temperature of each node, the radiation thermal resistance is calculated.

[0011] In a possible implementation, the initial core loss and the initial winding loss are input into a pre-built thermal network model to obtain the temperature of each node in the pre-built thermal network model, including: In the pre-built thermal network model, the equivalent thermal resistance is calculated according to the structure, initial core loss and initial winding loss of the pre-built thermal network model; Determine the temperature of each node in the pre-built thermal network model based on the equivalent thermal resistance.

[0012] In a possible implementation, the method further includes: When the temperature of each node does not meet the preset conditions, return to the step of inputting the initial core loss and the initial winding loss into the pre-built thermal network model to obtain the temperature of each node in the pre-built thermal network model.

[0013] In a possible implementation, the preset condition is: The difference between the current temperature of each node and the temperature of each node calculated last time is less than a preset threshold.

[0014] In one possible implementation, the pre-built thermal network model is constructed based on the three-dimensional structure and heat transfer path of the target high-frequency transformer; The pre-built thermal network model is a quarter model of the target high-frequency transformer.

[0015] In a possible implementation, the temperature field parameters include thermal resistance, heat flow, temperature difference, thermal conductivity and heat capacity.

[0016] In a second aspect, an embodiment of the present invention provides a temperature prediction device for a high-frequency transformer, comprising: An acquisition module is used to obtain the temperature field parameters of the target high-frequency transformer; A loss calculation module, used to calculate the initial core loss and the initial winding loss according to the temperature field parameters; A temperature calculation module, used for inputting the initial core loss and the initial winding loss into a pre-built thermal network model to obtain the temperature of each node in the pre-built thermal network model; A correction module, used for correcting the initial core loss and the initial winding loss based on the initial temperature of each node, and inputting the corrected initial core loss and the initial winding loss into a pre-built thermal network model to recalculate the temperature of each node; The prediction module is used to obtain the temperature prediction value of each electromagnetic component of the target high-frequency transformer when the temperature of each node meets the preset conditions.

[0017] The embodiment of the present invention provides a temperature prediction method and device for a high-frequency transformer. Considering the limitations of various prediction algorithms at this stage, this embodiment uses a thermal network modeling method as a basis. Considering that the temperature rise of the high-frequency transformer during operation is caused by the core loss and the winding loss, the core loss and the winding loss are used as initial calculation parameters and input into a pre-built thermal network model to obtain the temperature of each node. Then, considering that the thermal network modeling method has the problem of repeated calculation of the heat dissipation area and the heat conduction path, this embodiment corrects the initial core loss and the initial winding loss based on the temperature of each node, and calculates the temperature of each node again based on the corrected core loss and winding loss to improve the accuracy of the calculation. Finally, the temperature of each node that meets the preset conditions is used as the temperature prediction value of the target high-frequency transformer. It can be seen that the embodiment of the present invention can overcome the problem of inaccurate high-frequency transformers caused by the shortcomings of inaccurate modeling and too many conditions in the traditional method, and provide an accurate and effective temperature prediction method for high-frequency transformers. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 is a flow chart of the implementation of the temperature prediction method of the high-frequency transformer provided by the embodiment of the present invention; Figure 2 It is a schematic diagram of the heat dissipation method of the high-frequency transformer; FIG3 (a)-FIG3 (b) are schematic diagrams of the three-dimensional structure and heat transfer path of a high-frequency transformer; Figure 4 It is a thermal network model architecture diagram provided by an embodiment of the present invention; Figure 5 is a schematic diagram of the stacked structure of nanocrystalline magnetic cores; Figure 6 Schematic diagram of a temperature-controllable AC magnetic property measurement system provided by an embodiment of the present invention; Figure 7 is the nanocrystalline material loss curve; Figure 8 is a temperature rise curve of each node determined based on a thermal network model provided by an embodiment of the present invention; Fig. 9 Schematic diagram of steady-state temperature distribution of the magnetic core under no-load condition; Fig.10 Schematic diagram of a high-frequency transformer temperature rise test system provided by an embodiment of the present invention; FIG. 11( a ) to FIG. 11( d ) are comparison diagrams of core temperature rise tests under no-load conditions provided by an embodiment of the present invention. Fig.12 is a flow chart of a method for predicting the temperature of a high-frequency transformer provided by another embodiment of the present invention; Fig.13 It is a schematic diagram of the structure of a temperature prediction device for a high-frequency transformer provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0021] In order to make the purpose, technical solutions and advantages of the present invention more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0022] Figure 1 is a flow chart of the temperature prediction method for a high-frequency transformer provided by an embodiment of the present invention, such as Figure 1 As shown, the method may include: Step 110: Obtain temperature field parameters of the target high-frequency transformer.

[0023] In this embodiment, the target high-frequency transformer may be a high-frequency transformer with a nanocrystalline iron core, and its temperature field parameters may include data such as thermal resistance, heat flow, temperature difference, thermal conductivity and heat capacity, and initial ambient temperature.

[0024] Step 120: Calculate the initial core loss and the initial winding loss according to the temperature field parameters.

[0025] In this embodiment, the core loss and the winding loss are the main heat sources of the temperature rise of the target high-frequency transformer. Therefore, it is particularly important to accurately predict the core loss and the winding loss.

[0026] Considering that the excitation conditions (i.e., operating conditions) of high-frequency transformers are completely different from those of power-frequency transformers, which is one of the characteristics of high-frequency transformers that is different from power transformers. Since the focus of the embodiment of the invention is the temperature rise change of the transformer core structure, square wave excitation is used as the research object to calculate the core loss.

[0027] For non-sinusoidal excitation waveform, the core loss is calculated using the improved generalized Steinmetz equation (IGSE), which takes into account both accuracy and convenience of calculation.

[0028] Since the initial temperature of each material in the transformer equipment can be approximately considered to be equal to the ambient temperature, and the temperature usually affects the resistivity and magnetic properties of the material. At the initial ambient temperature, it is generally possible to query the factory parameters of the material by using the table lookup method to obtain k , α、β The initial core loss and initial winding loss are calculated equally.

[0029] Accordingly, the initial core loss can be expressed as:

[0030]

[0031] in, T is the magnetization period, Δ B is the peak-to-peak value of the magnetic flux density during the magnetization cycle, k , α and β is the loss coefficient, which is related to the material type and temperature; dB / dt is the rate of change of magnetic flux density.

[0032] High-frequency transformers usually use Litz wire windings to minimize the impact of eddy current effects on losses, but due to the complex torsional structure of Litz wire, it is challenging to accurately and quickly calculate the windings. As a classic and relatively accurate method, this embodiment uses Dowell's formula to calculate Litz wire losses, which is the heat source of the windings. Accordingly, the initial winding loss can be expressed as:

[0033]

[0034]

[0035] in, I n Indicates the effective value of the current component of the nth harmonic; F r,n represents the AC resistance coefficient corresponding to the nth harmonic, R dc is the DC resistance of the winding, and the product of the two is the equivalent AC resistance of the winding in the nth harmonic; N s is the number of single strands of Litz wire contained in each bundle of Litz wires; m Indicates the maximum number of layers of continuous winding. D L is the outer diameter of the Litz wire harness, d s is the diameter of a single strand of Litz wire, d L is the distance between the Lidz wire bundles,h w is the height of the winding, h c is the height of the core window.

[0036] Step 130: Input the initial core loss and the initial winding loss into a pre-built thermal network model to obtain the temperature of each node in the pre-built thermal network model.

[0037] This embodiment is divided into two steps, one is the construction process of the pre-constructed thermal network model, and the other is the use process of the pre-constructed thermal network model, that is, the process of outputting the temperature of each node. Here, this embodiment first describes the construction process of the thermal network model: The construction process of the thermal network model can include three processes: heat transfer principle analysis, parameter derivation, and model construction. Among them, the heat transfer principle analysis can be: In this embodiment, high-frequency transformers have broad application prospects in fields such as direct current transmission due to their advantages such as small size, high frequency, and high power density. Accurately predicting the temperature rise characteristics of magnetic components such as magnetic cores is crucial to the sustained and stable operation of high-frequency transformers. However, existing temperature rise prediction methods have problems such as imprecise modeling and many assumptions, which are not conducive to the evaluation of transformer operating conditions. Therefore, this embodiment proposes a thermal network model that reasonably considers the heat flow path of the transformer core. The model is derived based on the thermoelectric analogy method and analyzes the various components of the thermal resistance modeling method in detail.

[0038] Specifically, we first analyze the heat transfer principle of high-frequency transformers: According to the laws of thermodynamics, heat can be spontaneously transferred from an object with a high temperature to an object with a low temperature. When the high-frequency transformer starts to operate, the loss distribution of each component is different, resulting in a temperature difference, and the heat continues to diffuse, making the temperature of the high-frequency transformer in a dynamic equilibrium state. Heat transfer is an extremely complex process, which is composed of a combination of multiple methods. Under natural heat dissipation conditions, heat is mainly transferred through conduction, convection and radiation. Accordingly, the heat dissipation method of the high-frequency transformer can be as follows: Figure 2 shown.

[0039] exist Figure 2 middle, Q Indicates the heat flux transferred by heat dissipation, also known as heat transfer rate; T 1 and T 2 respectively represent the temperatures at both ends of the thermal conductive material; T s and T a They represent the temperature of the heat dissipation surface of the material and the temperature of the surrounding environment respectively. Among them, (a) represents transmission heat conduction; (b) represents horizontal upward plane convection heat dissipation; (c) represents plane convection heat dissipation; (d) represents horizontal downward plane convection heat dissipation; (e) represents radiation heat dissipation.

[0040] against Figure 2 The corresponding heat fluxes of various heat dissipation methods given in are expressed as: For the heat transfer in (a), the heat flux Q The cross-sectional area through which heat passes A is proportional to the heat flow direction (perpendicular to A ), which is the famous Fourier law:

[0041] in, Q is the heat flow; k is the thermal conductivity of the material; A is the cross-sectional area for heat conduction.

[0042] For (b)-(d), they are uniformly classified as convection heat dissipation in this embodiment. In the heat dissipation process of high-frequency transformers, the heat transfer effect of convection heat dissipation is weaker and is related to the heat transfer of the fluid boundary layer adjacent to the surface of the object. Q Determined by the type, velocity and viscosity of the adjacent fluid medium. Newton's law of cooling gives a simple expression for the convection heat dissipation process:

[0043] in, h conv is the convection heat dissipation coefficient of the material; A is the surface area of ​​the heat source.

[0044] Different from heat dissipation by conduction and convection, heat dissipation by radiation is transferred in the material medium by electromagnetic radiation. This process can be described by the Stefan-Boltzmann law:

[0045] in, ε is the emissivity of the radiating surface; σ is the Stefan-Boltzmann constant, usually 5.67×10-8W / (m2·K4); A is the radiation area; h rad is the radiation heat transfer coefficient.

[0046] Figure 3 (a)-Figure 3 (b) are schematic diagrams of the three-dimensional structure and heat transfer path of a high-frequency transformer, wherein Figure 3 (a) is the three-dimensional structure of the high-frequency transformer, and Figure 3 (b) is the heat transfer path of the high-frequency transformer. The heat transfer path is determined based on its three-dimensional structure.

[0047] Based on Figure 3 (a)-Figure 3 (b), it can be seen that for high-frequency transformers, the heat transfer paths mainly include the following situations: (1) Conduction heat transfer inside the core and winding; (2) Convective and radiative heat transfer from the core and winding surface to the surrounding environment; (3) Since the gaps between the core and the primary side of the winding, and between the primary side of the winding and the secondary side of the winding are relatively small, the heat transfer can be approximately considered to be conduction.

[0048] Accordingly, in an optional embodiment, the parameter derivation process may be: The thermal network model is usually built based on the thermoelectric analogy, because the distribution of heat flux lines in the thermal field is similar to the distribution of electric field lines. As shown in Table 1, the thermal network model is constructed by analogizing the thermal circuit parameters to the circuit parameters, thereby constructing a high-frequency transformer block parameter thermal network model.

[0049] Table 1 Analogy relationship between temperature field parameters and electric field parameters

[0050] As shown in Table 1, when constructing a thermal network model, thermal resistance can be compared to resistance, heat flow can be compared to current, temperature difference can be compared to voltage, thermal conductivity can be compared to electrical conductivity, and heat capacity can be compared to capacitance. The corresponding thermal resistance can refer to the calculation method of resistance, that is, R th =Δ T / P It can be compared to R = U / I .

[0051] Accordingly, in an optional embodiment, the model building process may be: Based on the two processes of heat transfer principle analysis and parameter derivation, the construction of the thermal network model needs to be based on the following settings: (1) The core loss is evenly distributed in the core, and the winding loss is evenly distributed in the primary and secondary windings; (2) Considering that the structure of the high-frequency transformer is completely symmetrical, the thermal network model only needs to be constructed based on one quarter of the structure of the high-frequency transformer; (3) Heat conduction only exists in the iron core, winding, epoxy resin and air environment.

[0052] Based on the above analysis, we can get Figure 4 Thermal network model architecture diagram. Figure 4 In the thermal network model, there are 12 nodes, representing different electromagnetic components of the high-frequency transformer, to improve the calculation accuracy. Among them, there are 5 nodes that deserve more attention, namely Figure 4Several nodes are marked with numbers.

[0053] From ① to ⑤, these nodes represent the hot spots of the core center column, the hot spots of the core upper yoke, the hot spots of the core side yoke, the hot spots of the low-voltage winding and the high-voltage winding, respectively.

[0054] In addition, these five nodes are also the areas where the high-frequency transformer heat source injects heat flow, that is, the central column heat source P cc , upper yoke heat source P cy 、External yoke heat source P ce , primary winding heat source P lv and the secondary winding heat source P hv The entire thermal network can be divided into three parts: the internal conduction thermal resistance of the magnetic core represented by the block diagram where node ② is located, the internal conduction thermal resistance represented by the block diagram where nodes ④ and ⑤ are located, and the collection of all convection and radiation thermal resistances inside the high-frequency transformer composed of other parts.

[0055] Considering the shell structure of the high-frequency transformer in this embodiment, the central area of ​​the magnetic core is usually surrounded by windings or insulation structures, and the area of ​​its surface in direct contact with a large amount of air is reduced, and the convection heat transfer process is hindered. If the magnetic core is surrounded by an epoxy resin insulation layer and cannot even come into contact with the air, the heat in the central area must first be conducted through the epoxy resin to the surface area in contact with the air, and then dissipated through convection and radiation. However, the thermal conductivity of epoxy resin itself is very low, and this structure is not conducive to the heat dissipation of the central pillar of the transformer, which will cause the hot spot temperature to rise.

[0056] The different heating and cooling conditions of the various components of the high-frequency transformer will produce temperature differences, so the heat will radiate between the core center, primary winding, secondary winding and core side yoke. Since thermal radiation is omnidirectional, this energy can be directly transmitted along the line of sight between the surfaces inside the transformer.

[0057] The steady-state temperature of the nodes in the thermal network model can be calculated by analogy with the node voltage equation of the circuit, that is, the total heat flowing into the node is equal to the total heat flowing out of the node:

[0058] in A represents the thermal conductivity matrix, T is the temperature matrix of the nodes, U is the heat source matrix, B is the coefficient matrix corresponding to the heat source matrix. Therefore, to calculate the temperature matrix, we first need to obtain the thermal resistance of each branch of the heat network and the heat source injected into the node. Accordingly, the following is an explanation of the use process of the thermal network model, that is, the process of outputting the temperature of each node: Accordingly, in a possible implementation, in step 130, inputting the initial core loss and the initial winding loss into a pre-built thermal network model to obtain the temperature of each node in the pre-built thermal network model may include: In the pre-built thermal network model, the equivalent thermal resistance is calculated based on the structure, initial core loss, and initial winding loss of the pre-built thermal network model.

[0059] Determine the temperature of each node in the pre-built thermal network model based on the equivalent thermal resistance.

[0060] In this embodiment, since the high-frequency transformer has high-voltage windings and low-voltage windings, the initial winding loss includes the primary low-voltage winding loss and the primary high-voltage winding loss. Therefore, the initial core loss can be injected into node ① in the thermal network model, the primary low-voltage winding loss can be injected into node ④ in the thermal network model, and the primary high-voltage winding loss can be injected into node ⑤ in the thermal network model, and the thermal balance equation analogous to the node current equation is followed. That is, the sum of the heat received by the node is equal to the sum of the heat dissipated by the node and the heat stored. Since this model focuses on the value after the temperature stabilizes, the thermal capacity calculation that has almost no effect on the final result is not considered. The corresponding mathematical expression is as follows:

[0061] where ∑ Q in and∑ Q out are the heat flows into and out of the node, Δ Q storage is the thermal storage change, which is usually ignored in steady-state analysis.

[0062] For node temperature calculation, a method similar to circuit analysis is used to find the temperature of each node by solving a set of linear / nonlinear equations. Generally, it can be solved by an iterative method such as Newton-Raphson. The node temperature calculation equation is: , you can also use the Simulink module in MATLAB software to assist in solving.

[0063] Accordingly, when determining the temperature of each node, it is necessary to consider the equivalent thermal resistance after injecting the initial core loss and the initial winding loss. Then, based on the equivalent thermal resistance and the structure of the high-frequency transformer, the anisotropic thermal resistance of the core and winding is derived to obtain the temperature of each node.

[0064] Correspondingly, the calculation formula of equivalent thermal resistance can be:

[0065] in, The thermal conductivity coefficient indicates the amount of heat that can be conducted by the material under unit temperature difference per unit time. is the cross-sectional area of ​​heat conduction, It is the thermal conductivity of the heat transfer material, which indicates the heat conduction capacity of the material per unit length and per unit temperature difference. l cond is the length of the material in the direction of conduction.

[0066] Based on the equivalent thermal resistance and the structure of the high-frequency transformer, the following derivation is performed: The magnetic core of the high-frequency transformer targeted by this embodiment is a nanocrystalline iron core. The layered stacking of nanocrystalline strips is the main reason for the anisotropy of the thermal performance of the magnetic core. This anisotropy causes the thermal conductivity of the magnetic core to be direction-dependent, and the heat transfer efficiency within the strip surface is significantly higher than the heat transfer efficiency between surfaces. Therefore, the thermal network model needs to consider the material properties, stacking geometry, and environmental cooling conditions of the nanocrystalline strips to predict the temperature distribution of the magnetic core under workload.

[0067] Figure 5 is a schematic diagram of the stacked structure of nanocrystalline magnetic cores. Figure 5 In the figure, the heat in the rectangular cube magnetic core column area mainly flows in a longitudinal straight line. Considering that the heat in the rectangular magnetic core mainly flows in a ring shape along the direction of the strip material, the magnetic core is divided into a rectangular magnetic core column, a yoke structure and a fan-shaped corner structure in the thermal network model proposed in this embodiment. Accordingly, the equivalent thermal resistance of the rectangular magnetic core column can be expressed as follows after transformation: (20) in, h is the height of the magnetic core column, b is the core column width, a is the thickness of the magnetic core column.

[0068] Since the core is wound with strip material, the thickness of the four sides of the core is equal. a , the equivalent thermal resistance of the rectangular yoke can be expressed as follows after transformation:

[0069] in, c is the length of the yoke heat flow path, that is, the yoke length.

[0070] In the core corner area, the nanocrystalline ribbon is bent into a fan-shaped surface and forms a fan-shaped column, which is a fan-shaped corner structure. In this structure, the heat flow can approximately flow along the surface in the fan-shaped structure. This assumption simplifies the three-dimensional heat flow problem into a two-dimensional problem along the surface, thereby simplifying the heat transfer analysis. At this time, the heat flow calculation in the fan-shaped column structure can be performed using the heat conduction formula along the surface. It is worth noting that this approximation method is most accurate when the temperature gradient is mainly along the surface direction. Accordingly, the equivalent thermal resistance of the fan-shaped corner structure can be expressed as follows after transformation:

[0071] in, is the thermal conductivity of the nanocrystalline material. This calculation method assumes that the inner and outer surfaces of the four corners of the core coincide at one point. r i and r o Represent the circumference radius of the inner and outer angles respectively, π ( r o + r i ) represents the average heat transfer path length of the fan-shaped column stack structure. Here, it should be noted in advance that when the convection heat dissipation surface of the winding structure is equivalent to a cylinder with a radius of D, the convection heat transfer coefficient can be expressed as follows: .

[0072] Since the side surface area of ​​the nanocrystalline core is significantly larger than the top and bottom surface areas, the heat dissipation effect of conduction in the axial direction of the winding area and convection on the upper and lower surfaces can be ignored. Based on this assumption, the winding can be treated as a composite structure of a square copper conductor wire and an epoxy resin insulation layer, so that the winding can be divided into two different regions: one is the region where the copper conductor and epoxy resin are connected in series, and the other is a region composed entirely of epoxy resin. By calculating the thermal resistance of these two parallel regions, the overall radial equivalent thermal resistance of the winding structure can be obtained. The equivalent thermal resistance calculation formula for the composite region is as follows:

[0073] in, is the thermal resistance of the copper conductor in the area where the copper conductor and epoxy resin are in series, is the thermal resistance of the epoxy resin in this area, R o and R i Respectively represent the outer diameter and inner diameter of the horizontal cross section of the copper winding; k cu and k epoxy denote the thermal conductivity of copper conductor and epoxy resin respectively;d Indicates the thickness of epoxy resin on the copper winding surface; L 1 represents the total height of the mixing zone in the vertical direction.

[0074] Using the same calculation method, the total height is L The radial equivalent thermal resistance of the epoxy resin area of ​​2 can be equivalently expressed as:

[0075] By calculating the thermal resistance of the composite area and the epoxy resin area, we can get the total thermal resistance of the coil R th,coil As shown in the following formula, it is R th,com and R th,epoxy The reciprocal of the sum of the reciprocals of , that is:

[0076] Based on the equivalent thermal resistances obtained above, the temperature of each node in the pre-constructed thermal network model, that is, the temperature corresponding to each electromagnetic component in the target high-frequency transformer, is determined through optimization iteration.

[0077] Step 140: Based on the initial temperature of each node, the initial core loss and the initial winding loss are corrected, and the corrected initial core loss and the initial winding loss are input into the pre-built thermal network model to recalculate the temperature of each node.

[0078] Figure 6 Schematic diagram of a temperature-controllable AC magnetic property measurement system provided by an embodiment of the present invention; Figure 7 is the loss curve of nanocrystalline material, combined with Figure 6 as well as Figure 7 This embodiment is described: In this embodiment, the AC magnetic property measurement system includes a first programmable DC power supply, an inverter circuit, a second programmable DC power supply, a constant temperature box, a ring sample, a power analyzer and a display. The ring sample is set in the constant temperature box, the first programmable DC power supply and the second programmable DC power supply adjust the temperature of the constant temperature box through the inverter circuit, and the power analyzer performs sampling and analysis on the ring sample at different temperatures to obtain magnetic density loss data, and the analysis results are displayed based on the display. Among them, the temperature setting range of the constant temperature box can be 30-150°C, and the frequency setting can be 10kHz.

[0079] Figure 7 The curve in is based on Figure 6 The data collected is analyzed, with the horizontal axis representing magnetic flux and the vertical axis representing loss. Figure 7It can be seen that the core loss per unit mass will gradually increase with the increase of temperature. When the core is close to the saturation flux density, this phenomenon will become more and more obvious. When the temperature reaches 100°C and the magnetic flux density reaches around 0.9T, the loss value can rise to about 1.3 times that at 30°C. Therefore, it is necessary to repeatedly correct the initial core loss and initial winding loss during the calculation of the thermal network model to improve the accuracy and scientificity of temperature prediction.

[0080] Accordingly, in an optional embodiment, in step 140, correcting the initial core loss and the initial winding loss based on the initial temperature of each node may include: Based on the initial temperature of each node, the convection thermal resistance and radiation thermal resistance are calculated.

[0081] The initial core losses and initial winding losses are corrected based on the convection thermal resistance and radiation thermal resistance.

[0082] In this embodiment, theoretically, if the length and cross-sectional area of ​​the conduction path are known in the thermal network model, the conduction thermal resistance can be obtained relatively simply, and the initial core loss and initial winding loss can be corrected. However, in the process of calculating the conduction thermal resistance, the calculation of the convection thermal resistance and the radiation thermal resistance is involved, and the calculation of the convection thermal resistance and the radiation thermal resistance is relatively complicated, and the heat transfer coefficient needs to be calculated separately.

[0083] Accordingly, in an optional embodiment, based on the initial temperature of each node, calculating the convection thermal resistance and the radiation thermal resistance may include: Based on the initial temperature of each node, the corresponding convective heat transfer coefficient and temperature difference are determined.

[0084] Based on the convection heat transfer coefficient, calculate the convection thermal resistance.

[0085] Based on the temperature difference and the initial temperature of each node, the radiation thermal resistance is calculated.

[0086] In this embodiment, in order to calculate the convection thermal resistance and the radiation thermal resistance, it is necessary to first derive the calculation formulas of the convection thermal resistance and the radiation thermal resistance respectively, so as to determine the important parameters designed for calculating the convection thermal resistance and the radiation thermal resistance: The calculation formulas for convection thermal resistance and radiation thermal resistance are as follows:

[0087]

[0088] in, h conv and h rad are the convective heat transfer coefficient and the radiative heat transfer coefficient, respectively. and are the heat transfer areas for convection heat transfer and radiation heat transfer, respectively.

[0089] The convective heat transfer coefficient is proportional to the Nusselt number, then:

[0090] in, Nu is the Nusselt number, which is the ratio of the convective heat transfer to the conductive heat transfer on the boundary; L is the characteristic length. The characteristic length of the horizontal plane is equal to the surface area divided by the perimeter, and the characteristic length of the vertical plane is equal to the height of the heat dissipation surface. k is the thermal conductivity of the fluid. The heat dissipation fluid of the high-frequency transformer is usually air.

[0091] In general, convection and conduction heat flows are parallel to each other and to the surface normals of the boundary surfaces, both of which are perpendicular to the mean fluid flow. Therefore, the calculation of the Nusselt number varies depending on the spatial location. For a horizontal upward heat dissipation plane, the Nusselt number can be expressed as:

[0092] in Ra , Gr and Pr are the Rayleigh number, the Grashof number and the Prandtl number, respectively, which can be calculated by the following formula:

[0093]

[0094] in, g represents the acceleration due to gravity, c p represents the specific heat capacity, ν and µ Corresponding to kinematic viscosity and dynamic viscosity respectively.

[0095] For the horizontal downward heat dissipation surface, the Nusselt number can be expressed as:

[0096] There are many expressions for the Nusselt number of the vertical surface. Under natural convection conditions, the surface of the high-frequency transformer is in a laminar state (i.e. 10 4 < Ra <10 9 ), therefore, the Nusselt number of the vertical plane can be calculated based on the following formula to improve the calculation accuracy:

[0097] The calculation method of radiation heat transfer coefficient is as follows:

[0098] Thermal radiation is a key factor in the heat dissipation of an object. Ignoring the radiation heat dissipation in the transformer model will lead to a serious overestimation of the surface temperature. Therefore, the radiation thermal resistance must be considered and verified in the thermal analysis to ensure accurate prediction.

[0099] Based on the above analysis, the heat transfer coefficient of the convective thermal resistance is related to the Nusselt number and depends on the properties of the fluid (such as viscosity and density), which are affected by temperature. For example, as the temperature increases, the viscosity of the air decreases and the density also decreases, which may increase the convective heat transfer coefficient and thus reduce the convective thermal resistance. The parameters such as the kinematic viscosity, dynamic viscosity and specific heat capacity of the air at the corresponding temperature are obtained by the table lookup method. The radiation thermal resistance is related to the radiation heat exchange capacity of the surface of the object. According to the Stefan-Boltzmann law, the radiation thermal resistance is proportional to the fourth power of the temperature difference. Substituting the fixed ambient temperature, emissivity, Stefan-Boltzmann constant, and the node temperature and heat dissipation surface area that change with the calculation into the above formula can calculate the radiation thermal resistance at the current node temperature.

[0100] Therefore, when determining the convection thermal resistance and radiation thermal resistance, the kinematic viscosity, dynamic viscosity and specific heat capacity of the air corresponding to the current temperature can be determined according to the initial temperature of each node to determine the convection heat transfer coefficient. Then, the corresponding temperature difference can be determined according to the initial temperature of each node to determine the radiation thermal resistance of each node.

[0101] In an optional embodiment, correcting the initial core loss and the initial winding loss according to the convection thermal resistance and the radiation thermal resistance may include: Based on the convection thermal resistance and radiation thermal resistance, the current temperature of each corresponding node is calculated.

[0102] According to the current temperature of each node, the current loss parameter and the current conductivity are determined.

[0103] Current loss parameters, correcting the initial core loss.

[0104] The initial winding losses are corrected for the current conductivity.

[0105] In this embodiment, since the corresponding loss parameters and conductivity are different at different temperatures, when the loss parameters and conductivity change, the core loss and winding loss will also change accordingly.

[0106] For example, after n iterations, the temperature of each node corresponding to the nth iteration, that is, the initial temperature of each node, is obtained, and the corresponding equivalent thermal resistance is determined based on the initial temperature of each node to determine the convection thermal resistance and radiation thermal resistance. In order to improve the calculation accuracy, the temperature of each node after the n+1th iteration, that is, the current temperature of each node, is calculated based on the convection thermal resistance and radiation thermal resistance corresponding to the nth iteration. According to the current temperature of each node, the corresponding current loss parameter and current conductivity are determined to correct the initial core loss and initial winding loss.

[0107] After obtaining the corrected initial core loss and the corrected initial winding loss, these two sets of data are re-input into the nodes corresponding to the pre-built thermal network model, and iterative optimization is re-performed to obtain the temperature of each node.

[0108] Step 150: When the temperature of each node meets the preset conditions, the temperature prediction value of each electromagnetic component of the target high-frequency transformer is obtained.

[0109] In this embodiment, this embodiment adopts an optimization iteration method to determine the temperature of each node. When the temperature of each node does not meet the preset conditions, it returns to the step of inputting the initial core loss and the initial winding loss into the pre-constructed thermal network model to obtain the temperature of each node in the pre-constructed thermal network model, until the temperature of each node meets the preset conditions, the temperature of each node is output, and the temperature of each node is used as the temperature prediction value of each electromagnetic component of the target high-frequency transformer.

[0110] In this embodiment, the preset condition may be that the difference between the current temperature of each node and the temperature of each node calculated last time is less than a preset threshold, wherein the preset threshold may be 0.01, 0.011, 0.012, etc., and may be set as needed.

[0111] Figure 8 is a temperature rise curve of each node determined based on the thermal network model provided by an embodiment of the present invention, wherein the horizontal axis is time and the vertical axis is temperature, wherein T cc Represents the core center temperature, T cy Represents the core top temperature, T ce Represents the core edge temperature, T p and T s Represent the primary and secondary winding temperatures respectively. Figure 8 It can be seen that after running for about 40,000 seconds, the core temperature reaches a stable state, the core center hot spot temperature is the highest, 127.15 degrees Celsius; the upper yoke hot spot temperature is second only to the center hot spot temperature, 105.04 degrees Celsius; the outer hot spot temperature is the lowest, 83.92 degrees Celsius. In general, the heat flow and temperature distribution inside the core conform to the heat transfer law.

[0112] In order to verify the scientificity, accuracy and effectiveness of the method provided in this embodiment, this embodiment is verified through finite element simulation and transformer no-load temperature rise test.

[0113] Specifically, the embodiment of the present invention uses COMSOL Multiphysics software to perform two-dimensional temperature field finite element calculation based on magnetothermal coupling. Since the focus of this study is the temperature rise in the core area, the transformer under no-load operation is modeled based on a high-frequency transformer prototype.

[0114] Considering the unique properties of nanocrystalline layered structures that lead to anisotropic magnetic and thermal conductivities, the core is modeled in different regions. Cartesian coordinates are used in the columnar region, while cylindrical coordinates are used in the angular region for accurate calculations. Table 2 lists the thermal parameters that are critical to the finite element simulation.

[0115] Table 2 Thermal characteristic parameters of high frequency transformer materials

[0116] In Table 2, these parameters include the thermal conductivity, specific heat capacity and density parameters of nanocrystals in the parallel and vertical directions, Litz wire in the axial and radial directions, and epoxy resin. These parameters play an important role in determining the thermal behavior of the core. When the temperature rises to a steady state, the temperature distribution of the entire core cross section is calculated and displayed. Based on the parameters in Table 2, the steady-state temperature distribution of the core of the high-frequency transformer prototype under no-load condition is obtained as shown in the figure. Fig. 9 shown.

[0117] exist Fig. 9 It can be seen that the temperature distribution in different areas is relatively uniform. This observation confirms the assumption of uniform temperature distribution in various parts of the thermal network model, thus verifying its rationality. Under no-load conditions, the stable temperature range of the core is between 82.9 and 136 degrees Celsius. The highest and lowest temperatures are located at the center and outer edge of the core, respectively. The errors of the proposed thermal network model in the highest and lowest temperatures are 6.5% and 1.2%, respectively.

[0118] In order to measure the temperature rise change of the high-frequency transformer under no-load conditions, this embodiment constructs the following Fig.10 The high-frequency transformer temperature rise test system shown in Fig.10 The invention comprises a first programmable DC power supply, a second programmable DC power supply, an inverter circuit, a high-frequency transformer prototype, a temperature recorder and a power analyzer.

[0119] The first programmable DC power supply and the second programmable DC power supply are used to power the high-frequency transformer prototype and the power analyzer through the inverter circuit. When the high-frequency transformer prototype is running at no load, the temperature and power changes of the high-frequency transformer prototype are recorded by the temperature recorder and the power analyzer. Among them, the thermocouples are set at the determined key hot spots. These thermocouples can be used as accurate temperature sensors to measure the temperature changes of these key positions in real time and accurately. In addition, a thermal imager is used to comprehensively monitor the high-frequency transformer.

[0120] Figures 11 (a) to 11 (d) are comparison diagrams of the core temperature rise test under no-load conditions provided by an embodiment of the present invention, which show the temperature distribution of the high-frequency transformer at different times during the no-load test measured by a thermal imager, where (a) is the temperature after 3 hours of no-load operation, (b) is the temperature after 5 hours of no-load operation, (c) is the temperature after 6 hours of no-load operation, and (d) is the temperature after 8 hours of no-load operation. It can be clearly seen from the figure that the temperature at the center of the core is significantly higher than the temperature on both sides of the yoke. This observed pattern is very consistent with the temperature distribution predicted by the thermal network model, and the prediction error of the highest temperature point is 6.7%, which further proves the reliability of the model provided by the embodiment of the present invention in predicting thermal characteristics.

[0121] In addition, it is worth noting that after three hours of no-load operation, the temperature of the transformer rose to 110.9 degrees Celsius, reaching 81.3% of the steady-state temperature. This observation shows that the temperature of the high-frequency transformer rises faster at the beginning of operation, a phenomenon that can be attributed to the thermal capacitance effect of the various components of the transformer.

[0122] Fig.12 FIG. 1 is a flow chart of a method for predicting the temperature of a high-frequency transformer according to another embodiment of the present invention; Fig.12 As shown, the method may include: The temperature field parameters are obtained, the current thermal resistance and thermal capacitance in the temperature field parameters are used as input, and the loss parameters and electrical conductivity are determined based on parameters such as heat flow, temperature difference, and thermal conductivity, and the initial core loss and initial winding loss are calculated.

[0123] The initial core loss and initial winding loss are input into the corresponding nodes of the thermal network model, and the temperature of each representative node is calculated, that is, the central column heat source P cc , upper yoke heat source P cy 、External yoke heat source P ce , primary winding heat source P lv and the secondary winding heat source P hv .

[0124] Recalculate the corresponding convection thermal resistance and radiation thermal resistance according to the temperature of each node.

[0125] The initial core loss and initial winding loss are corrected based on the calculated convection thermal resistance and radiation thermal resistance.

[0126] The corrected initial core loss and initial winding loss are re-entered into the nodes corresponding to the thermal network model, and the optimization iteration is repeated. It is determined whether the temperature of each node obtained each time is in a stable state, that is, whether the preset conditions are met. If not, the initial core loss and initial winding loss are returned to the nodes corresponding to the thermal network model, and the temperature of each representative node is calculated. This step is repeated until the temperature of each node is in a stable state.

[0127] The temperature of each node calculated at this time is output to determine the temperature prediction value of each electromagnetic component in the high-frequency transformer.

[0128] In summary, the embodiment of the present invention provides a thermal network model of a high-frequency transformer, which reduces the cost required for modeling through local analysis modeling, improves the calculation accuracy of thermal resistance in the edge area, and also takes into account the influence of the magnetic properties and thermal anisotropy of nanocrystals on temperature rise prediction. The effectiveness of the model is confirmed by finite element simulation and temperature rise test experiments.

[0129] In addition, the embodiment of the present invention comprehensively explains and analyzes the calculation method of the conduction, convection and radiation thermal resistance of each structure of the transformer, as well as the design principle of the thermal network model. Compared with complex finite element modeling, the thermal network model greatly shortens the calculation time, and its analytical model properties make it suitable for subsequent optimization design research of high-frequency transformers.

[0130] Finally, based on the comparison with the finite element simulation and the transformer hot spot temperature rise experimental data, the errors of the model provided by the embodiment of the present invention and the two methods are 6.5% and 6.7% respectively, which shows that the method provided by the embodiment of the present invention shows a high calculation accuracy. The provided model can provide a theoretical reference of certain value for other thermal design tasks.

[0131] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0132] The following is an embodiment of the device of the present invention. For details not described in detail therein, reference may be made to the corresponding method embodiment described above.

[0133] Fig.13 The schematic diagram of the structure of the temperature prediction device of the high-frequency transformer provided by the embodiment of the present invention is shown. For the convenience of explanation, only the part related to the embodiment of the present invention is shown, which is described in detail as follows: like Fig.13As shown, the temperature prediction device 13 of the high-frequency transformer includes: An acquisition module 31 is used to obtain temperature field parameters of a target high-frequency transformer; A loss calculation module 32, used to calculate the initial core loss and the initial winding loss according to the temperature field parameters; The temperature calculation module 33 is used to input the initial core loss and the initial winding loss into the pre-built thermal network model to obtain the temperature of each node in the pre-built thermal network model; A correction module 34, for correcting the initial core loss and the initial winding loss based on the initial temperature of each node, and inputting the corrected initial core loss and the initial winding loss into a pre-built thermal network model to recalculate the temperature of each node; The prediction module 35 is used to obtain the temperature prediction value of each electromagnetic component of the target high-frequency transformer when the temperature of each node meets the preset conditions.

[0134] In a possible implementation, the correction module 34 is specifically configured to: Based on the initial temperature of each node, calculate the convection thermal resistance and radiation thermal resistance; The initial core losses and initial winding losses are corrected based on the convection thermal resistance and radiation thermal resistance.

[0135] In a possible implementation, the correction module 34 is specifically configured to: Based on the convection thermal resistance and radiation thermal resistance, calculate the current temperature of each corresponding node; According to the current temperature of each node, determine the current loss parameter and the current conductivity; Current loss parameters, to correct the initial core loss; The initial winding losses are corrected for the current conductivity.

[0136] In a possible implementation, the correction module 34 is specifically configured to: Based on the initial temperature of each node, determine the corresponding convective heat transfer coefficient and temperature difference; Based on the convection heat transfer coefficient, calculate the convection thermal resistance; Based on the temperature difference and the initial temperature of each node, the radiation thermal resistance is calculated.

[0137] In a possible implementation, the temperature calculation module 33 is specifically configured to: In the pre-built thermal network model, the equivalent thermal resistance is calculated according to the structure, initial core loss and initial winding loss of the pre-built thermal network model; Determine the temperature of each node in the pre-built thermal network model based on the equivalent thermal resistance.

[0138] In a possible implementation, the prediction module 35 is further configured to: When the temperature of each node does not meet the preset conditions, return to the step of inputting the initial core loss and the initial winding loss into the pre-built thermal network model to obtain the temperature of each node in the pre-built thermal network model.

[0139] In a possible implementation, the preset condition is: The difference between the current temperature of each node and the temperature of each node calculated last time is less than a preset threshold.

[0140] In one possible implementation, the pre-built thermal network model is constructed based on the three-dimensional structure and heat transfer path of the target high-frequency transformer; The pre-built thermal network model is a quarter model of the target high-frequency transformer.

[0141] In a possible implementation, the temperature field parameters include thermal resistance, heat flow, temperature difference, thermal conductivity and heat capacity.

[0142] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0143] Those of ordinary skill in the art will appreciate that the templates, units, and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0144] If the module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned temperature prediction method embodiments of each high-frequency transformer can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory, random access memory, electrical carrier signal, telecommunication signal and software distribution medium.

[0145] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A temperature prediction method for a high-frequency transformer, characterized in that: include: Obtain temperature field parameters of the target high-frequency transformer; Calculating initial core loss and initial winding loss according to the temperature field parameters; Inputting the initial core loss and the initial winding loss into a pre-built thermal network model to obtain the temperature of each node in the pre-built thermal network model; Based on the initial temperature of each node, the initial core loss and the initial winding loss are corrected, and the corrected initial core loss and initial winding loss are input into the pre-built thermal network model to recalculate the temperature of each node; When the temperature of each node meets the preset conditions, the temperature prediction value of each electromagnetic component of the target high-frequency transformer is obtained.

2. The temperature prediction method of a high-frequency transformer according to claim 1, characterized in that: The correcting the initial core loss and the initial winding loss based on the initial temperature of each node includes: Calculating convection thermal resistance and radiation thermal resistance based on the initial temperature of each node; The initial core loss and the initial winding loss are corrected according to the convection thermal resistance and the radiation thermal resistance.

3. The temperature prediction method of a high-frequency transformer according to claim 2, characterized in that: The correcting the initial core loss and the initial winding loss according to the convection thermal resistance and the radiation thermal resistance includes: Based on the convection thermal resistance and the radiation thermal resistance, calculate the current temperature of each corresponding node; Determining a current loss parameter and a current conductivity according to the current temperature of each node; The current loss parameter is used to correct the initial core loss; The initial winding losses are corrected according to the current conductivity.

4. The temperature prediction method of a high-frequency transformer according to claim 2, characterized in that: The calculating of the convection thermal resistance and the radiation thermal resistance based on the initial temperature of each node includes: Based on the initial temperature of each node, determine the corresponding convective heat transfer coefficient and temperature difference; Calculating the convective thermal resistance according to the convective heat transfer coefficient; The radiation thermal resistance is calculated according to the temperature difference and the initial temperature of each node.

5. The temperature prediction method of a high-frequency transformer according to claim 1, characterized in that: The step of inputting the initial core loss and the initial winding loss into a pre-built thermal network model to obtain the temperature of each node in the pre-built thermal network model comprises: In a pre-constructed thermal network model, calculating an equivalent thermal resistance according to the structure of the pre-constructed thermal network model, the initial magnetic core loss and the initial winding loss; The temperature of each node in the pre-built thermal network model is determined according to the equivalent thermal resistance.

6. The temperature prediction method of a high-frequency transformer according to claim 1, characterized in that: The method further comprises: When the temperature of each node does not meet the preset condition, return to the step of inputting the initial core loss and the initial winding loss into the pre-constructed thermal network model to obtain the temperature of each node in the pre-constructed thermal network model.

7. The temperature prediction method of a high-frequency transformer according to claim 1, characterized in that: The preset conditions are: The difference between the current temperature of each node and the temperature of each node calculated last time is less than a preset threshold.

8. The temperature prediction method of a high-frequency transformer according to claim 1, characterized in that: The pre-built thermal network model is built based on the three-dimensional structure and heat transfer path of the target high-frequency transformer; The pre-built thermal network model is a quarter model of the target high-frequency transformer.

9. The temperature prediction method of a high-frequency transformer according to claim 1, characterized in that: The temperature field parameters include thermal resistance, heat flow, temperature difference, thermal conductivity and heat capacity.

10. A temperature prediction device for a high-frequency transformer, characterized in that: include: An acquisition module is used to obtain the temperature field parameters of the target high-frequency transformer; A loss calculation module, used for calculating the initial core loss and the initial winding loss according to the temperature field parameters; A temperature calculation module, used for inputting the initial core loss and the initial winding loss into a pre-built thermal network model to obtain the temperature of each node in the pre-built thermal network model; A correction module, configured to correct the initial core loss and the initial winding loss based on the initial temperature of each node, and input the corrected initial core loss and initial winding loss into the pre-built thermal network model to recalculate the temperature of each node; The prediction module is used to obtain the temperature prediction value of each electromagnetic component of the target high-frequency transformer when the temperature of each node meets the preset conditions.

Images