A planar transformer design method based on finite element simulation software

Abstract

The application discloses a planar transformer design method based on finite element simulation software, relates to the technical field of transformers, and obtains the winding impedance matrix by using the finite element simulation software, and then adopts a theoretical calculation method to obtain a winding structure with minimum loss. The method combines the simulation tool with theoretical analysis, forms a relatively complete theoretical system, is applicable to any winding structure, has a wide application range, can quickly and accurately find the optimal arrangement mode of the transformer winding, can avoid a complex theoretical solving process compared with a solving method based on a theoretical loss model, and has high precision; and compared with a solving method based on the finite element simulation software alone, a large amount of simulation time can be saved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of transformer technology, in particular to a planar transformer design method based on finite element simulation software. BACKGROUND

[0002] With the development of DC / DC module power supply, the traditional wound transformer cannot meet the development trend of high frequency and flatness of switching power supply due to the problems of single coil structure, poor heat dissipation characteristics and poor parameter consistency, while the planar transformer overcomes the shortcomings of the traditional wound transformer and is used more and more widely.

[0003] The winding structure of the planar transformer has three types of series, parallel and series-parallel. For the series structure winding, the same current flows through each turn of the winding, but due to the different distribution positions of the winding in the magnetic element, the AC impedance of each turn of the winding is different, which further causes the loss of each turn of the winding to be different. For the parallel structure winding, when the closed path formed by the parallel windings passes through the alternating magnetic field, circulating current will be generated between the parallel windings, which will cause uneven current flow in the parallel windings, and even the current of some layers of windings may be reversed. The distribution position of the winding has a great influence on the current flow of the parallel winding, but when the number of windings is large, it is difficult to determine the distribution position of the parallel winding. For the series-parallel structure, the winding can be connected in parallel first and then in series, or connected in series first and then in parallel. The current distribution between the windings is very different under different connection modes, which has a great influence on the winding loss.

[0004] In order to optimize the winding arrangement in the planar transformer, a winding loss model can be established to obtain the optimal winding arrangement according to the lowest winding loss. In order to establish the winding loss model, the magnetic field distribution in the magnetic element needs to be accurately obtained. The magnetic field distribution in the magnetic element is related to many factors such as the shape of the magnetic core, the position of the air gap, the position and arrangement of the winding, and the skin effect and proximity effect of the current-carrying conductor. It is very difficult to accurately obtain the magnetic field distribution in the magnetic element, and the workload is huge. Therefore, it is difficult to optimize the winding arrangement based on the theoretical establishment of the winding loss model, and the accuracy is generally low.

[0005] Another way to obtain the winding loss is to use finite element simulation software to build a model for simulation. This method can obtain more accurate winding loss and save the complex theoretical research process. However, in order to obtain the optimal winding loss, a model needs to be built for each winding arrangement, and then simulation is performed. With the increase in the number of windings, the number of winding arrangements increases rapidly. It will take a lot of time to simulate so many winding arrangements.

[0006] Therefore, this invention proposes a transformer winding design method based on finite element simulation software. After obtaining the impedance matrix of each winding using finite element simulation software, theoretical calculation methods are used to obtain the winding structure with the minimum loss. This method combines simulation tools with theoretical analysis, forming a relatively complete theoretical system that can be applied to any winding structure and has a wide range of applications. Summary of the Invention

[0007] To address the above technical problems, this invention provides a planar transformer design method based on finite element simulation software, characterized by the following steps:

[0008] A1. Build a transformer model in finite element simulation software;

[0009] A2. When the primary and secondary currents of the transformer are sinusoidal, sinusoidal current excitation is given to the primary and secondary windings separately to obtain the winding impedance matrix.

[0010] When the primary and secondary currents of the transformer are non-sinusoidal, DC, fundamental, and harmonic excitations are injected into the transformer model in the finite element simulation software to obtain the winding impedance matrices Z(DC) and Z(f) at each frequency point. s ...Z(m·f s ), where Z(DC) is the impedance matrix at DC, Z(m·f s Let f be the impedance matrix at the m-th harmonic frequency. s Let m be the fundamental frequency and m be the harmonic order. Then, Fourier decomposition is performed on the primary and secondary currents.

[0011] A3. Confirm any winding arrangement method;

[0012] A4. List the port voltage and current constraints, calculate the current flowing through each turn of the winding, and obtain the current matrix;

[0013] A5. Calculate the winding loss based on the impedance matrix and current matrix of the winding. When the transformer primary and secondary currents are non-sinusoidal, calculate the loss under each harmonic and sum the losses to obtain the total loss.

[0014] A6. Exhaustively enumerate all primary and secondary winding stacked structures, repeating steps A3 to A5. When the transformer primary and secondary currents are sinusoidal, calculate the losses under each winding arrangement; when the transformer primary and secondary currents are non-sinusoidal, calculate the total losses under each winding arrangement.

[0015] A7. Based on the loss results calculated in step A6, when the transformer primary and secondary currents are sinusoidal, find the winding arrangement corresponding to the minimum loss; when the transformer primary and secondary currents are non-sinusoidal, find the winding arrangement corresponding to the minimum total loss, which is the optimized winding structure.

[0016] The technical solution further defined in this invention is:

[0017] Furthermore, in step A1, the transformer model is set to have n·i turns of winding on the primary side and k·j turns of winding on the secondary side. There are i branches in parallel on the primary side, each with n turns of winding connected in series. There are j branches in parallel on the secondary side, each with k turns of winding connected in series.

[0018] In the aforementioned planar transformer design method based on finite element simulation software, step A2 involves performing Fourier decomposition on the primary and secondary currents, as shown in the following equation.

[0019] i p =i p (DC)+i p (f s )+…i p (m·f s (18)

[0020] i s =i s (DC)+i s (f s )+…i s (m·f s (19)

[0021] Among them, i p i represents the primary current. p (DC) represents the primary side direct current, i p (m·f s () represents the m-th harmonic current on the primary side, where m represents the harmonic order; i s Indicates the secondary current, i s (DC) represents the secondary DC current, i s (m·f s ) represents the mth harmonic current on the secondary side.

[0022] In the aforementioned planar transformer design method based on finite element simulation software, in step A2, the harmonic order m is set to 3.

[0023] In the aforementioned planar transformer design method based on finite element simulation software, step A3 describes a transformer with n·i turns of winding on the primary side and k·j turns of winding on the secondary side. The primary side has i parallel branches, each with n turns of winding connected in series. The secondary side has j parallel branches, each with k turns of winding connected in series.

[0024] A winding arrangement is determined, specifically, the (x-1)·n+1 layer to the x·n layer is the xth branch winding of the primary side, where x = 1 to i; the i·n+(x-1)·k+1 layer to the i·n+x·k layer is the xth branch winding of the secondary side, where x = 1 to j.

[0025] In the previously described planar transformer design method based on finite element simulation software, step A4 describes a transformer primary winding with n·i turns. Layers (x-1)·n+1 to x·n are the xth branch windings of the primary winding, therefore the current in each layer is the same, denoted as I. px Where x = 1 to i; the secondary side has k·j turns of winding, and layers i·n+(x-1)·k+1 to i·n+x·k are the xth branch windings of the secondary side, so the current in each layer is the same, denoted as I. sx , where x = 1 ~ j;

[0026] The method for obtaining the current matrix includes the following steps:

[0027] Define the voltage per turn of the winding as U 11 ...U 1n U i1 ...U in U s11 ...U s1k U sj1 ... U sjk According to Ohm's law and Kirchhoff's laws, we can obtain

[0028]

[0029] I p1 +…+I pi =I p (27)

[0030] I s1 +…+I sj =I s (28)

[0031] U 11 +…+U 1n =U i1 +…+U in (29)

[0032] U s11 +…+U s1k =U sj1 +…+U sjk (30)

[0033] Among them, I p I is the total primary current of the transformer. s This represents the total secondary current of the transformer.

[0034] By combining equations (26) to (30), the current flowing through each turn of the winding and the voltage it withstands can be obtained. The current flowing through each turn of the winding can be represented in matrix form, and the corresponding current matrix is ​​defined as M. in+jk Specifically

[0035]

[0036] Among them, M in+jk This represents the current matrix.

[0037] In the aforementioned planar transformer design method based on finite element simulation software, step A5, when the primary and secondary currents of the transformer are sinusoidal, the specific expression for the winding loss is as follows:

[0038]

[0039] Among them, M in+jk This represents the current matrix of the winding. For M in+jk The conjugate transpose of ;

[0040] When the primary and secondary currents of the transformer are non-sinusoidal, the losses are calculated using equation (25) to obtain the losses under each harmonic, and the losses are summed to obtain P. total The specific expression is

[0041] P total =P(DC)+P(f) s )+…P(m·f s (20)

[0042] Among them, P total P(DC) represents the total loss, and P(m·f) represents the DC loss. s ) represents the loss at the mth harmonic frequency, where m represents the harmonic number.

[0043] The beneficial effects of this invention are:

[0044] After obtaining the impedance matrix of each winding using finite element simulation software, theoretical calculation methods are used to obtain the winding structure with minimum loss. This method combines simulation tools with theoretical analysis, forming a relatively complete theoretical system that is applicable to any winding structure and has a wide range of applications. It can quickly and accurately find the optimal arrangement of transformer windings. Compared with the solution method based on theoretical loss models, it avoids the complex theoretical solution process and has higher accuracy. Compared with the solution method based solely on finite element simulation software, it can save a lot of simulation time. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of a general model of a planar transformer;

[0046] Figure 2 This is a simplified diagram of the transformer model when i = j = 2 in this invention;

[0047] Figure 3 This is one winding arrangement method in the present invention;

[0048] Figure 4 This is a schematic diagram of a transformer model in which the primary and secondary windings are connected in parallel, as described in this invention.

[0049] Figure 5 This is a schematic diagram of a transformer model in which the primary side is connected in series and the secondary side is connected in parallel in this invention;

[0050] Figure 6 This is a schematic diagram of a transformer model with multiple output windings in this invention;

[0051] Figure 7 This is a flowchart of the process for finding the optimal winding arrangement when the transformer primary and secondary currents are non-sinusoidal.

[0052] Figure 8 This is a schematic diagram of a transformer model in this invention, where the primary and secondary sides are each connected in parallel with three turns.

[0053] Figure 9 This is a bar chart showing the winding losses under different winding arrangements in this invention.

[0054] Figure 10 This is a comparison diagram of the current distribution in each layer under different winding arrangements in this invention. Detailed Implementation

[0055] like Figure 1 The figure shows a general model of a single-input single-output planar transformer. The embodiment uses a winding structure that is first connected in series and then in parallel as an example for illustration. The winding structure that is first connected in parallel and then in series is optimized using the same method.

[0056] The transformer has n·i turns of winding on the primary side and k·j turns of winding on the secondary side. There are i branches connected in parallel on the primary side, each with n turns of winding connected in series. There are j branches connected in parallel on the secondary side, each with k turns of winding connected in series. When n, i, j, and k take different values, this transformer can be applied to different situations. For ease of description, the winding design method of the planar transformer described in this invention is illustrated below using i=j=2 as an example. The simplified transformer model is as follows: Figure 2 As shown.

[0057] The design method for planar transformers includes the following steps:

[0058] Step 1: Build the transformer model in finite element simulation software, and apply sinusoidal current excitation to each of the 2n+2k turns of the primary and secondary windings separately to obtain a winding impedance matrix Z with 2n+2k rows and columns. 2n+2k ,

[0059]

[0060] The diagonal elements represent the self-impedance of the winding, such as Z. s11,s11 Characterizing the self-impedance of the S11 winding; off-diagonal elements characterize the mutual impedance between windings, such as Z. s21,s11 Characterizes the mutual impedance between the s21 winding and the s11 winding.

[0061] Step 2, determine a winding arrangement, such as Figure 3 As shown, layers 1 to n are the first branch windings of the primary side connected in series, layers n+1 to 2n are the second branch windings of the primary side connected in series, layers 2n+1 to 2n+k are the first branch windings of the secondary side connected in series, and layers 2n+k+1 to 2n+2k are the second branch windings of the secondary side connected in series. The two primary windings are connected in series and then in parallel, and the two secondary windings are connected in series and then in parallel.

[0062] Step 3: List the port voltage and current constraints, and calculate the current flowing through each turn of the winding;

[0063] Since layers 1 to n are primary windings, the winding currents of layers 1 to n are the same, denoted as I. p1 Similarly, the current in the (n+1)th to 2nth winding layers is the same, denoted as I. p2 The winding currents from layer 2n+1 to 2n+k are the same, denoted as I. s1 The winding currents from layer 2n+k+1 to 2n+2k are the same, denoted as I. s2 Define the voltage per turn of the winding as U. 11 ...U 1n U 21 ...U 2n U s11 ...U s1k U s21 ...U s2k According to Ohm's law and Kirchhoff's laws, we can obtain:

[0064]

[0065] I p1 +I p2 =I p (2)

[0066] I s1 +I s2 =I s (3)

[0067] U 11 +…+U 1n =U 21 +…+U 2n (4)

[0068] U s11 +…+U s1k =U s21 +…+U s2k (5)

[0069] Among them, I p I is the total primary current of the transformer. s This represents the total secondary current of the transformer.

[0070] By combining equations (1) to (5), the current flowing through each turn of the winding and the voltage it withstands can be obtained. For ease of description, the current flowing through each turn of the winding is represented in matrix form, and the corresponding current matrix is ​​defined as M. 2n+2k The specific expression is:

[0071]

[0072] Step 4: Given the impedance matrix and current matrix, the winding loss can be calculated, specifically as follows:

[0073]

[0074] in, For M 2n+2k The conjugate transpose of .

[0075] Step 5: Exhaustively enumerate all primary and secondary stacked structures, and then repeat steps 2 to 4 to calculate the loss P for each structure;

[0076] Step 6: Based on the loss results calculated in Step 5, find the winding structure corresponding to the minimum loss, which is the optimized winding structure.

[0077] The method proposed in this invention only requires one simulation using finite element simulation software in step 1. Subsequent steps can be completed through programming, which can save a lot of simulation time. Furthermore, steps 2 to 6 can be automatically iterated and filtered by a computer through a small program.

[0078] For different transformer structures, only the corresponding port voltage and current constraints need to be changed. The following are examples of several common winding structures. Their design steps are the same as those above, except that the port voltage and current constraints in step 3 are different.

[0079] Transformer Structure 1: As shown Figure 4As shown, when n = k = 1, the primary windings and secondary windings of the transformer are connected in parallel. This winding structure is suitable for applications with low voltage levels and high currents on both the primary and secondary sides. The currents flowing through the primary and secondary windings in this structure are defined as I1, I2...I... i and I s1 I s2 ...I sj The voltages of the primary winding and the secondary winding are respectively expressed as U p and U S At this point, the constraints on the port voltage and current are:

[0080]

[0081] I1+I2+…+I i =I p (9)

[0082] I s1 +I s2 +…+I sj =I s (10)

[0083] Combining equations (8) to (10), the current matrix M can be obtained. i+j .

[0084] Transformer Structure 2: When i=1 and k=1, the primary side has only one branch with n turns of winding connected in series, and the secondary side has j windings connected in parallel, with each parallel branch having only one turn of winding. In this case, the transformer model can be simplified to... Figure 5 The series-parallel structure shown is characterized by n turns of primary winding connected in series and j turns of secondary winding connected in parallel. This winding structure is often used in applications with high primary voltage and high secondary current. The currents flowing through the primary and secondary windings of the transformer are defined as I0 and I0, respectively. p and I s1 I s2 ...I sj The voltages of the primary winding and the secondary winding are respectively expressed as U p1 U p2 ...U pn and U S At this point, the constraints on the port voltage and current are:

[0085]

[0086] I s1 +I s2 +…+I sj =I s (12)

[0087] Combining equations (11) to (12), the current matrix M can be obtained. n+j .

[0088] Transformer Structure 3: In addition to the single-input single-output transformers mentioned above, the design method proposed in this invention is also applicable to transformers with multiple outputs. To simplify the analysis, as follows... Figure 6 As shown, the analysis takes two output windings as an example. The primary winding has two parallel branches, each consisting of n turns of winding connected in series. Each secondary winding consists of k turns of winding connected in parallel. The currents flowing through the primary and secondary windings of the transformer are defined as I0 and I0, respectively. p1 I p2 and I s11 I s12 ...I s1k and I s21 I s22 ...I s2k The voltages of the primary winding and the secondary winding are U and U, respectively. p11 U p12 ...U p2n and U S1 U S2 At this point, the constraints on the port voltage and current are:

[0089]

[0090] I p1 +I p2 =I p (14)

[0091] I s11 +…+I s1k =I s1 (15)

[0092] I s21 +…+I s2k =I s2 (16)

[0093] U p11 +…+U p1n =U p21 +…+U p2n (17)

[0094] By combining equations (13) to (17), the current matrix M can be obtained. 2n+2k .

[0095] The method mentioned above is based on the impedance matrix, which is obtained after injecting a sinusoidal excitation of a certain frequency. When the frequency of the excitation changes, the impedance matrix will also change. Therefore, the above analysis method is only applicable when the primary and secondary currents of the transformer are sinusoidal or approximately sinusoidal.

[0096] When the primary and secondary currents of a transformer are non-sinusoidal, such as the primary and secondary current waveforms in a phase-shifted full-bridge converter, it is necessary to inject DC, fundamental, and harmonic frequency excitations to obtain the impedance matrix at the DC, fundamental, and harmonic frequencies. Then, Fourier decomposition is performed on this non-sinusoidal wave, and steps 2 to 5 are executed to calculate the losses. The losses generated by each harmonic are then summed to obtain the final total loss. Finally, the optimal winding structure is found based on the losses to minimize the winding losses.

[0097] like Figure 7 As shown, when the excitation waveform is non-sinusoidal, the specific steps to find the optimal winding arrangement are as follows:

[0098] The first step is to inject DC, fundamental, and harmonic excitations into the transformer model in the finite element simulation software to obtain the winding impedance matrix Z(DC) and Z(f) at each frequency point. s ...Z(m·f s ), where f s Given the fundamental frequency, the value of the harmonic order m is determined based on the magnitude of the harmonic content in the excitation current. The larger m is, the more accurate the result, but the more complex the calculation; conversely, the smaller m is, the greater the error in the result, but the simpler the calculation. In practical applications, specific considerations should be taken into account. Generally, a harmonic order of m = 3 is sufficient to meet the accuracy requirements.

[0099] The second step is to perform Fourier decomposition on the primary and secondary currents, as shown below:

[0100] i p =i p (DC)+i p (f s )+…i p (m·f s (18)

[0101] i s =i s (DC)+i s (f s )+…i s (m·f s (19)

[0102] Among them, i p i represents the primary current. s This represents the secondary current, and m represents the harmonic order.

[0103] The third step is to determine a winding arrangement.

[0104] The fourth step is to list the port voltage and current constraints, and calculate the winding current matrix under each harmonic according to the method in step 3.

[0105] Fifth, based on the current matrix obtained in the fourth step, calculate the loss according to equation (7) in step 4 to obtain the loss under each harmonic, and sum the losses to obtain P. total The specific expression is:

[0106] P total =P(DC)+P(f) s )+…P(m·f s (20)

[0107] Among them, P total Indicates total loss;

[0108] Step 6: Exhaustively list all possible winding arrangements, and repeat steps 3 through 5 to calculate the total loss.

[0109] The seventh step is to find the winding arrangement that corresponds to the minimum total loss, which is the optimized winding structure.

[0110] To verify the effectiveness of the method proposed in this invention, a planar transformer with three turns in parallel on the primary side and three turns in parallel on the secondary side is used as an example for design. Figure 8 This is a diagram of the transformer model used in this example, where the 6-turn winding is distributed on a 6-layer PCB.

[0111] Step 1: After building the transformer model in the finite element simulation software, inject a sinusoidal excitation with a frequency of 1MHz into each of the six windings to obtain the winding matrix Z6, as follows:

[0112]

[0113] Step 2: Determine a winding arrangement, specifically that the primary windings of layers 1, 2, and 3 are connected in parallel, and the secondary windings of layers 4, 5, and 6 are connected in parallel.

[0114] Step 3: Solve for the current matrix M6 using the port voltage and current constraints.

[0115]

[0116] I1 + I2 + I3 = I p (twenty two)

[0117] I4 + I5 + I6 = I s (twenty three)

[0118] Where I p =30sin(ωt), I s = -30sin(ωt), solving for:

[0119]

[0120] Step 4, calculate the winding loss under this winding arrangement:

[0121]

[0122] Step 5: Exhaustively list all possible winding arrangements. There are 10 different winding arrangements, as shown in Table 1. Calculate the current matrix for each winding arrangement, and then calculate the losses for each winding arrangement. Figure 9 As shown.

[0123] Numbering Winding structure Numbering Winding structure 1 PPPSSS 6 PSPSPS 2 PPSPSS 7 PSPSSP 3 PPSSPS 8 PSSPPS 4 PPSSSP 9 PSSPSP 5 PSPPSS 10 PSSSPP

[0124] Table 1 Winding Arrangement

[0125] Step 6: Select an optimized winding arrangement based on losses, such as... Figure 9 As shown, the two structures with the lowest winding losses are PSPSPS and PSSPPS, which is consistent with practical engineering experience. Figure 10 The distribution of current in each turn is given under different winding arrangements. Figure 10 It can also be seen that the current distribution in each layer is the most uniform under the two winding arrangement methods of PSPSPS and PSSPPS, which further verifies the effectiveness of the proposed method.

[0126] This invention utilizes finite element simulation software to obtain the impedance matrix of each winding, and then employs theoretical calculation methods to obtain the winding structure with minimum loss. This method combines simulation tools with theoretical analysis, forming a relatively complete theoretical system that is applicable to any winding structure and has a wide range of applications. It can quickly and accurately find the optimal arrangement of transformer windings. Compared with the solution method based on theoretical loss models, it avoids the complex theoretical solution process and has higher accuracy. Compared with the solution method based solely on finite element simulation software, it can save a significant amount of simulation time.

[0127] In addition to the embodiments described above, the present invention may have other implementations. All technical solutions formed by equivalent substitution or equivalent transformation fall within the protection scope claimed by the present invention.

Claims

1. A planar transformer design method based on finite element simulation software, characterized in that: Includes the following steps A1. Build a transformer model in finite element simulation software; A2. When the primary and secondary currents of the transformer are sinusoidal, sinusoidal current excitation is given to the primary and secondary windings separately to obtain the winding impedance matrix. When the primary and secondary currents of the transformer are non-sinusoidal, DC, fundamental, and harmonic excitations are injected into the transformer model in the finite element simulation software to obtain the winding impedance matrix Z(DC) at each frequency point. ,..., Where Z(DC) is the impedance matrix at DC. Let be the impedance matrix at the m-th harmonic frequency. Where is the fundamental frequency, and m is the harmonic order; Next, Fourier decomposition is performed on the primary and secondary currents, as shown in the following equation. （1） （2） in, Represents the primary current. Represents the primary side DC current. This represents the mth harmonic current on the primary side, where m represents the harmonic order. Indicates the secondary current. Indicates the secondary DC current. This represents the m-th harmonic current on the secondary side; A3. Confirm any winding arrangement method; A4. List the port voltage and current constraints, calculate the current flowing through each turn of the winding, and obtain the current matrix; A5. Calculate the winding loss based on the impedance matrix and current matrix of the winding. When the transformer primary and secondary currents are non-sinusoidal, calculate the loss under each harmonic and sum the losses to obtain the total loss. A6. Exhaustively enumerate all primary and secondary winding stacked structures, repeating steps A3 to A5. When the transformer primary and secondary currents are sinusoidal, calculate the losses under each winding arrangement; when the transformer primary and secondary currents are non-sinusoidal, calculate the total losses under each winding arrangement. A7. Based on the loss results calculated in step A6, when the transformer primary and secondary currents are sinusoidal, find the winding arrangement corresponding to the minimum loss; when the transformer primary and secondary currents are non-sinusoidal, find the winding arrangement corresponding to the minimum total loss, which is the optimized winding structure.

2. The planar transformer design method based on finite element simulation software according to claim 1, characterized in that: In step A1, the transformer model is set to have n·i turns of winding on the primary side and k·j turns of winding on the secondary side. There are i branches in parallel on the primary side, each with n turns of winding connected in series. There are j branches in parallel on the secondary side, each with k turns of winding connected in series.

3. The planar transformer design method based on finite element simulation software according to claim 1, characterized in that: In step A2, the harmonic order m is set to 3.

4. The planar transformer design method based on finite element simulation software according to claim 1, characterized in that: In step A3, the transformer has n·i turns of winding on the primary side and k·j turns of winding on the secondary side. There are i branches in parallel on the primary side, each with n turns of winding connected in series. There are j branches in parallel on the secondary side, each with k turns of winding connected in series. A winding arrangement is determined, specifically, the (x-1)·n+1 layer to the x·n layer is the xth branch winding of the primary side, where x=1~i; the i·n+(x-1)·k+1 layer to the i·n+x·k layer is the xth branch winding of the secondary side, where x=1~j.

5. The planar transformer design method based on finite element simulation software according to claim 1, characterized in that: In step A4, the primary winding of the transformer has n·i turns. Layers (x-1)·n+1 to x·n are the xth branch windings of the primary winding; therefore, the current in each layer is the same, expressed as... Where x = 1 to i; the secondary side has k·j turns of winding, and layers i·n+(x-1)·k+1 to i·n+x·k are the xth branch windings of the secondary side, so the current in each layer is the same, expressed as , where x = 1 ~ j; The method for obtaining the current matrix includes the following steps: Define the voltage per turn of the winding as ... , ... , ... , ... According to Ohm's law and Kirchhoff's laws, we can obtain （3） （4） （5） （6） （7） in, This represents the total primary current of the transformer. This represents the total secondary current of the transformer. By combining equations (3) to (7), the current flowing through each turn of the winding and the voltage it withstands can be obtained. The current flowing through each turn of the winding can be represented in matrix form, and the corresponding current matrix is ​​defined as follows: Specifically （8） in, This represents the current matrix.

6. The planar transformer design method based on finite element simulation software according to claim 1, characterized in that: In step A5, when the transformer primary and secondary currents are sinusoidal, the specific expression for winding losses is as follows: （9） in, This represents the current matrix of the winding. for The conjugate transpose of ; When the primary and secondary currents of the transformer are non-sinusoidal, the losses are calculated using equation (9) to obtain the losses under each harmonic, and the losses are summed to obtain the final value. The specific expression is （10） in, P(DC) represents the total loss, and P(DC) represents the DC loss. This represents the loss at the mth harmonic frequency, where m represents the harmonic number.

Images