High-low grade silicon steel hybrid magnetic circuit transformer iron core design method and application

Abstract

The invention relates to a design method and application of a hybrid magnetic circuit transformer core based on high-grade and low-grade silicon steel. The design method comprises the following steps that a core limb and a yoke of the transformer core are made of two different materials, namely a high-grade silicon steel material and a low-grade silicon steel material respectively; constructing a high-grade silicon steel material database and a low-grade silicon steel material database; materials used by the iron core columns are obtained from a high-grade silicon steel material database, materials used by the iron yokes are obtained from a low-grade silicon steel material database, and constraints of a flux density constraint module on the premise that the optimal cross section of the iron core is obtained through an intelligent optimization algorithm are met. The total cost and the total no-load loss of the whole transformer iron core composed of the iron core columns and the iron yokes are the lowest. According to the hybrid magnetic circuit transformer iron core manufactured based on the design method, on the premise that it is ensured that the electromagnetic performance reaches the standard, the material cost and / or loss are / is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of transformer core design, in particular to a design method and application of a transformer core based on a high-low grade silicon steel hybrid magnetic circuit. BACKGROUND

[0002] Silicon steel material has characteristics of high saturation magnetic flux density, high magnetic permeability, low coercive force, high thermal stability and low cost, and its manufacturing process is simple and low in cost, so it still occupies a large market, and is generally used in laminated core structure, and is usually used in power frequency transformers.

[0003] For the laminated core, it is of great significance to improve the effective cross-sectional area as much as possible under the condition of given core diameter and sheet width series, for reducing no-load loss and saving wire material. In recent years, manufacturers have adopted an oval structure type on small and medium-sized distribution transformers in order to save raw materials and reduce manufacturing costs. This structure is easy to shape when winding, does not need to be pressed, and reduces the manufacturing process. However, there is less optimization research on the oval cross section. The Chinese patent with publication number CN211125312U discloses a kind of elliptical transformer core, which aims at the defects of high cost of existing standard elliptical core transformer and poor dynamic stability of long circular core, and improves the ability to withstand short circuit by changing the thickness of the first two layers of laminated sheets, but the change of cross section may cause the change of operating magnetic density, and it is not necessarily able to reduce the loss.

[0004] There are still some technical problems to be solved in the current transformer core design: in terms of structural design, the optimization design method of the oval cross section core is not perfect; in terms of economy, the cost of core manufacturing is high; in terms of energy efficiency, the problem of no-load loss is still prominent; in terms of material selection, there is a lack of optimal material selection scheme considering electromagnetic performance and economy. These key problems restrict the further improvement of transformer energy efficiency and economic benefit. SUMMARY

[0005] In view of the deficiencies of the prior art, the technical problem to be solved by the present application is to provide a design method and application of a transformer core based on a high-low grade silicon steel hybrid magnetic circuit, which realizes material cost reduction and / or loss reduction under the premise of ensuring electromagnetic performance.

[0006] The technical solution adopted by the present application to solve the technical problem is:

[0007] In a first aspect, the present application provides a design method of a transformer core based on a high-low grade silicon steel hybrid magnetic circuit, which includes the following contents:

[0008] The core column and the yoke of the transformer core are made of two different materials of high-grade silicon steel material and low-grade silicon steel material respectively.

[0009] The type, BH curve, BP fitting curve, unit price and no-load loss additional coefficient of various high-grade silicon steel materials are packaged to form a high-grade silicon steel material database;

[0010] The type, BH curve, BP fitting curve, unit price and no-load loss additional coefficient of various low-grade silicon steel materials are packaged to form a low-grade silicon steel material database;

[0011] The determination process of the BP fitting curve of each material in the two databases is as follows: the point with the maximum first derivative and zero second derivative on the BH curve is found as the knee point K, a 0.1T interval is moved on the BH curve, the slope in each interval is calculated, and when the slope of the first three intervals is less than 0.3 times the maximum first derivative, the starting point of the first interval that meets the condition is determined as the saturation point Sa; the magnetic flux density values corresponding to the knee point K and the saturation point Sa are determined, and then the BP curve is fitted in sections according to the magnetic flux density values B K , B Sa , and the final BP fitting curve is represented by formula (1):

[0012]

[0013] Wherein, p is the unit loss, B is the magnetic flux density, a1, b1, k, c1, d1, p s , p k are fitting parameters;

[0014] Start optimization: input the long axis and short axis of the circumscribed ellipse, the core series, the high-grade silicon steel material database, the low-grade silicon steel material database, the initial no-load loss Loss0, and the initial cost cost0;

[0015] First, the first high-grade silicon steel material is assigned to the core column, the total cross section of the laminations is maximized as the optimization target, the genetic algorithm is used for core column cross section optimization, and the optimal cross section S0 of the core column of the current high-grade silicon steel material is obtained; the operating magnetic flux density of the current core column is calculated according to S0, the effective value of the excitation voltage, and the number of turns on the secondary side of the transformer; the corresponding high-grade silicon steel material with a running magnetic flux density less than the magnetic flux density corresponding to the knee point K is determined in the high-grade silicon steel material database;

[0016] After determining the high-grade silicon steel material selected for the core column, the low-grade silicon steel material database is traversed, and the corresponding low-grade silicon steel material with a running magnetic flux density less than the magnetic flux density corresponding to the knee point K is determined in the low-grade silicon steel material database; the transformer core is composed of the high-grade silicon steel material that meets the requirements as the core column and the low-grade silicon steel material that meets the requirements as the yoke, and the corresponding unit loss is calculated according to the material operating magnetic flux density using the BP fitting curve, and then the no-load loss and cost of the entire transformer core are calculated;

[0017] The scheme 1 is the scheme of minimum no-load loss of output transformer, the scheme 2 is the scheme of minimum cost of transformer, and the scheme 3 is the scheme of minimum no-load loss and cost of transformer.

[0018] Further, the core is a three-phase three-column distribution transformer core.

[0019] Further, the data in the high-grade silicon steel material library is coded as: [Amaterial1, Abh1, Abp1, AK1, Amaterial2, Abh2, Abp2, AK2, …Abh T , Abp T , AK T ], wherein Amaterial1 represents the model of the first high-grade silicon steel material, Abh1 represents the BH curve of the first high-grade silicon steel material, Abp1 represents the BP fitting curve of the first high-grade silicon steel material, AK1 represents the no-load loss process additional coefficient of the first high-grade silicon steel material, and the rest of the codes are similar, and T represents the number of types of high-grade silicon steel materials.

[0020] The data in the low-grade silicon steel material library is coded as: [Bmaterial1, Bbh1, Bbp1, BK1, Bmaterial2, Bbh2, Bbp2, BK2, …Bbh W , Bbp W , BK W ]; wherein Bmaterial1 represents the model of the first low-grade silicon steel material, Bbh1 represents the BH curve of the first low-grade silicon steel material, Bbp1 represents the BP fitting curve of the first low-grade silicon steel material, BK1 represents the no-load loss additional coefficient of the first low-grade silicon steel material, and the rest of the codes are similar, and W represents the number of types of low-grade silicon steel materials.

[0021] The parameters of each level of the laminations are coded as: [width1, thickness1, width2, thickness2, …, width i , thickness i , …, width n , thickness n ], wherein width1 represents the width of the 1st level of laminations, thickness1 represents the thickness of the 1st level of laminations, width n represents the width of the n-th level of laminations, thickness n represents the thickness of the n-th level of laminations, n is the number of levels of the core, width i represents the width of the i-th level of laminations, and thickness iThis represents the stack thickness of the i-th layer, in mm; i = 1, 2, 3…n.

[0022] Furthermore, the constraints of the genetic algorithm are as follows:

[0023] Set the width of the last level. n If the minimum value and the minimum thickness of a single stack of the first-level stack are both the minimum values, then the width of the last-level stack and the thickness of a single stack of the first-level stack are not less than their minimum values.

[0024] The width of the image decreases in a decreasing pattern, with the width of level i being greater than the width of level i+1.

[0025] The stacking thickness shows an increasing trend, with the stacking thickness of layer i being less than that of layer i+1, and the stacking thickness of layer n being less than the major axis of the ellipse.

[0026] If the remainder when the width of each piece is divided by 5 is less than or equal to 2.5, then round down; if the remainder when the width of each piece is divided by 5 is greater than 2.5 but less than 5, then round up. The width of each piece satisfies formula (8):

[0027]

[0028] Where, d i width of the i-th level piece i The remainder when divided by 5;

[0029] If the rectangular stacked pieces are located inside the circumscribed ellipse, then they satisfy formula (9):

[0030]

[0031] Secondly, the present invention provides a hybrid magnetic circuit transformer core, wherein the hybrid magnetic circuit transformer core is obtained by any of the design methods described in claims 1-3, the core column of the hybrid magnetic circuit transformer core is made of high-grade silicon steel, the yoke of the hybrid magnetic circuit transformer core is made of low-grade silicon steel, and the core column and the yoke have the same number of laminations and the same cross-section.

[0032] Thirdly, the present invention provides a design system for a transformer core based on a hybrid magnetic circuit of high and low grade silicon steel, the system comprising:

[0033] A database of high-grade silicon steel materials, used to store relevant parameters of various high-grade silicon steel materials;

[0034] A database of low-grade silicon steel materials, used to store relevant parameters of various low-grade silicon steel materials;

[0035] Intelligent optimization algorithms are used to obtain the optimal cross-section of core columns made of different types of materials;

[0036] The magnetic density constraint module is used to determine whether the operating magnetic density is less than the magnetic density corresponding to the current material knee point K according to the average magnetic density passing through the core column section as the operating magnetic density, and if less, the current material is retained;

[0037] The material used for the core column is obtained from a high-grade silicon steel material database, the material used for the iron yoke is obtained from a low-grade silicon steel material database, and satisfies the constraint of the magnetic density constraint module under the premise of the optimal core section obtained by the intelligent optimization algorithm, so as to minimize the total cost and total no-load loss of the overall transformer core composed of the core column and the iron yoke.

[0038] Compared with the prior art, the beneficial effects of the present application are:

[0039] 1. Creatively, the present application differentiates the material configuration by using high-grade silicon steel for the core column and low-grade silicon steel for the iron yoke, and realizes the reduction of material cost and no-load loss under the premise of ensuring the electromagnetic performance up to the standard.

[0040] 2. In the present application, by integrating the complete magnetic characteristic database of high-grade and low-grade silicon steel materials, the optimal material combination scheme can be automatically recommended according to specific working condition parameters (such as working magnetic density, loss requirement, etc.), thereby providing data support for engineering design and production.

[0041] 3. In the material selection process, the magnetic density is checked, and the working magnetic density of each part of the core is monitored in real time to ensure that it is always controlled within the material knee point, thereby preventing local saturation of the core and the temperature rise problem caused thereby.

[0042] 4. For the nonlinear magnetization characteristics of silicon steel materials, the BP fitting curve is segmented based on the knee point and the saturation point, the function value and the first-order derivative at the characteristic points are strictly guaranteed to be continuous, a three-section analytical expression of quadratic polynomial-exponential function-linear function is adopted, the balance between calculation accuracy and efficiency is realized, and the effectiveness and accuracy of the design method are ensured. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 is a flowchart of an embodiment of the transformer core design method based on the mixed magnetic circuit transformer core of high and low-grade silicon steel according to the present application.

[0044] Figure 2 is a structural schematic diagram of the core column section in the present application.

[0045] Figure 3 is a schematic diagram of the core column section in the present application.

[0046] Figure 4 is a schematic diagram of the mixed magnetic circuit transformer core in the present application.

[0047] In the figure, 1 is an upper yoke, 2 is an A-phase core column, 3 is a B-phase core column, 4 is a C-phase core column, and 5 is a lower yoke. DETAILED DESCRIPTION

[0048] The application is further described in detail below with reference to the drawings and specific embodiments completed by the inventors according to the technical solutions provided by the application, but the embodiments are not used to limit the application, and any similar method and similar changes thereof shall be included in the protection scope of the application.

[0049] According to different operating magnetic densities, the transformer core columns and yokes are respectively selected from high-grade silicon steel materials and low-grade silicon steel materials. The high-grade silicon steel has improved loss performance and magnetization characteristics compared with the low-grade silicon steel, but the unit cost is higher. Based on a large amount of required magnetic density, loss, and unit price data, a high-grade silicon steel material database and a low-grade silicon steel material database are respectively established. By traversing the two databases, the knee point is constrained by the magnetic density, and the core column section is optimized to obtain the lowest no-load loss, the lowest cost, and the lowest no-load loss and cost.

[0050] The type, BH curve, BP fitting curve, unit price, and no-load loss additional coefficient of more than 100 high-grade silicon steel materials are packaged to form a high-grade silicon steel material database. The type, BH curve, BP fitting curve, unit price, and no-load loss additional coefficient of more than 100 low-grade silicon steel materials are packaged to form a low-grade silicon steel material database. The BH curve is a data curve with magnetic field intensity (unit: A / m) as the independent variable and magnetic flux density (unit: Tesla: T) as the dependent variable. The BP fitting curve is a data curve with magnetic flux density as the independent variable and unit loss (unit: W / kg) as the dependent variable.

[0051] The determination process of the BP fitting curve is as follows: find the point with the maximum first derivative and zero second derivative on the BH curve as the knee point K, move an interval of 0.1 T on the BH curve, and calculate the slope (first derivative) in the interval. When the slope of the first three intervals is less than 0.3 times the maximum first derivative for the first time, the starting point that meets the condition for the first time is determined as the saturation point Sa. The magnetic density values corresponding to the knee point K and the saturation point Sa are determined, and then the magnetic density values B K , B Sa Segment fitting BP curve, divide the BP curve into three intervals, and fit them with different mathematical models respectively.

[0052] Interval [0, B K ): a quadratic polynomial p=a1*B^2+b1*B is used, and the parameters are determined by least squares fitting; interval [B K , B Sa ): an exponential function When ensuring that the function is continuous and the first derivative is continuous at the K point; interval [B Sa , B max ]: a linear function p = p s +k(B-B Sa ) is adopted.

[0053] The final BP fitting curve is obtained as follows:

[0054]

[0055] Wherein, p is the unit loss, B is the magnetic density, B K represents the magnetic density corresponding to the knee point, B sa represents the magnetic density corresponding to the saturation point; a1, b1, c1, d1, p s , p k are fitting parameters.`Among them, the determination of the fitting parameters ensures that the functions of the three intervals simultaneously satisfy the following continuity conditions at the connecting points (knee point K and saturation point Sa): function value continuity: at B = B K , the value of the quadratic polynomial is equal to that of the exponential function; at B = B Sa , the value of the exponential function is equal to that of the linear function; first derivative continuity: at B = B K , the first derivative of the quadratic polynomial is equal to that of the exponential function; at B = B Sa , the first derivative of the exponential function is equal to that of the linear function.

[0056] The BP fitting curve obtained by piecewise fitting is different for different materials.

[0057] Start the optimization process: input the external ellipse major axis, minor axis, core series, high-grade silicon steel material database, low-grade silicon steel material database, and initialize the no-load loss Loss0 and the cost cost0; The no-load loss Loss0 and the cost cost0 can be set to infinity, ensuring that the no-load loss and the cost calculated in the first iteration can be retained as new initial values.

[0058] First, assign the first high-grade silicon steel material to the core column, and use the genetic algorithm to optimize the core column section to obtain the optimal core column section S0; Then calculate the running magnetic density of the core column at this time;

[0059] Determine whether the calculated running magnetic density is less than the magnetic density corresponding to the knee point (K) of the BH fitting curve of the high-grade silicon steel material. If not, it means that the running magnetic density is too high, which can easily cause the core column to be oversaturated, and then return to replace the next high-grade silicon steel material. If yes, calculate the no-load loss and cost of the core column at this time;

[0060] The first low-grade silicon steel material is given to the iron yoke, S0 is given to the cross-sectional area of the iron yoke, the operating magnetic density of the iron yoke made of the low-grade silicon steel material is calculated, and it is judged whether the operating magnetic density of the iron yoke is less than the magnetic density corresponding to the knee point of the BH curve of the low-grade silicon steel material. If not, it means that the operating magnetic density is too high, which can easily cause the iron yoke to be oversaturated, and then the next low-grade silicon steel material is returned to replace. If yes, the no-load loss of the iron yoke at this time and the cost are calculated.

[0061] The sum of the core column no-load loss and the iron yoke no-load loss is the no-load loss of the entire transformer core, and the sum of the core column cost and the iron yoke cost is the cost of the entire transformer core. The initial values of the no-load loss and the cost are set to be very large, so the scheme in which the first core column and the iron yoke material both meet the operating magnetic density requirement can be retained, and the no-load loss and the cost of the scheme are taken as the new initial values of the no-load loss and the cost to continue to participate in iteration until all the materials in the material library are traversed.

[0062] The final output is three sets of schemes: the scheme 1 with the minimum transformer no-load loss, the scheme 2 with the minimum transformer cost, and the scheme 3 with both the minimum transformer no-load loss and the minimum transformer cost. Each set of schemes includes the optimized high-grade silicon steel material type, the optimized low-grade silicon steel material type, the optimized transformer core overall no-load loss, and the optimized transformer core overall cost.

[0063] Embodiment 1

[0064] Referring to Figure 1 the optimization process, the embodiment of the present application provides a three-phase three-column distribution transformer core with an externally connected ellipse having a long axis of 294 mm and a short axis of 178 mm, which includes an upper yoke 1, three core columns (an A-phase core column 2, a B-phase core column 3, and a C-phase core column 4), and a lower yoke 5. The upper and lower yokes are completely symmetrical along the horizontal line, and the three core columns are completely the same. The core stage is selected to be 10 stages, the transformer window height is 465 mm, and the center distance is 370 mm. The initial no-load loss and the cost are both set to be 10 8 . The following steps are adopted:

[0065] 1) Input initial quantities: the externally connected ellipse long axis a = 294, the short axis b = 178, the core stage n = 10, the high-grade silicon steel material database, the low-grade silicon steel material database, and the initial setting values of the no-load loss Loss0 and the cost cost0. The initial values of the no-load loss Loss0 and the cost cost0 can be infinite or any relatively large value.

[0066] 2) Encoding: encode the data of the high-grade silicon steel material library: [Amaterial1, Abh1, Abp1, AK1, Amaterial2, Abh2, Abp2, AK2, … Abh T, Abp T , AK T ]. Wherein, Amaterial1 represents the model of the first high-grade silicon steel material, Abh1 represents the BH curve of the first high-grade silicon steel material, Abp1 represents the BP fitting curve of the first high-grade silicon steel material, AK1 represents the no-load loss process additional coefficient of the first high-grade silicon steel material, and the rest of the codes are similar, T represents the number of types of high-grade silicon steel materials.

[0067] The data of the low-grade silicon steel material library is coded as: [Bmaterial1, Bbh1, Bbp1, BK1, Bmaterial2, Bbh2, Bbp2, BK2, …Bbh W , Bbp W , BK W ]. Wherein, Bmaterial1 represents the model of the first low-grade silicon steel material, Bbh1 represents the BH curve of the first low-grade silicon steel material, Bbp1 represents the BP fitting curve of the first low-grade silicon steel material, BK1 represents the no-load loss additional coefficient of the first low-grade silicon steel material, and the rest of the codes are similar, W represents the number of types of low-grade silicon steel materials.

[0068] The parameters of each level of laminations are coded as: [width1, thickness1, width2, thickness2, …, width i , thickness i , …, width n , thickness n ]. As shown in Figure 2 , wherein width1 represents the width of the first level of laminations, thickness1 represents the thickness of the first level of laminations, and the unit is mm, wherein each level of laminations refers to two laminations that are symmetrically arranged above and below, the width and thickness of the two laminations of the same level are completely equal in value and symmetrically arranged along the center line; width n represents the width of the n-th level of laminations, thickness n represents the thickness of the n-th level of laminations, Figure 2 , the sum of the thicknesses of the uppermost and lowermost laminations is an ellipse, and the unit is mm; n is the number of core levels; width i represents the width of the i-th level of laminations, thickness i represents the thickness of the i-th level of laminations, and the unit is mm; i = 1, 2, 3…n.

[0069] 3) Initially, the first high-grade silicon steel material Amaterial1 is assigned to the core column, and the transformer core section is optimized based on the genetic algorithm to maximize the filling factor. It includes the following steps:

[0070] ① Population initialization:

[0071] Within the allowed range of parameters (major and minor axes, series, etc.), M individuals are randomly generated (M = 5000 in this embodiment), or useful data accumulated in previous studies are input into the original population to replace the random initialization process, thereby accelerating the optimization process.

[0072] ② Objective function and constraints

[0073] To maximize the core cross-sectional fill factor, the objective function is set to maximize the total cross-sectional area of ​​the 10-stage lamination.

[0074]

[0075] S represents the total cross-sectional area of ​​the n-stage stacked sheets, width i Let the thickness be the width of the i-th level stacked film. i Let be the stack thickness of the i-th level stack.

[0076] Establish constraints:

[0077] First, the boundaries of the sheet width and stack thickness are defined. In this embodiment, the minimum thickness of a single sheet in the first-level stack is 26mm, and the width of the last-level sheet is... n If the minimum value is 20mm, then:

[0078] thickness1 ≥ 26 × 2 (3)

[0079] width n ≥20 (4)

[0080] Since the width of the image decreases in a decreasing pattern, the width of image i is always greater than the width of image i+1.

[0081] width i >width i+1 (5)

[0082] The stacking thickness exhibits an increasing trend; the stacking thickness of layer i is always less than the stacking thickness of layer i+1, and the stacking thickness of layer n is always less than the major axis of the ellipse.

[0083]

[0084] The width of each element must be a multiple of 5. There are two cases: if the remainder when each element width is divided by 5 is less than or equal to 2.5, round down; if the remainder when each element width is divided by 5 is greater than 2.5 but less than 5, round up.

[0085]

[0086] Where, di the remainder of the width of the i-th level divided by 5.

[0087] The rectangular laminations must be located within the circumscribed ellipse:

[0088]

[0089] The optimal scheme of the laminations of each level of the elliptical cross-section of the core column is obtained as follows:

[0090] The maximum geometric cross-sectional area of the core column is 39215 mm 2 , which is the optimal cross-section S0 of the current grade core column, and the geometric cross-sectional area of the circumscribed ellipse is 41101.46 mm 2 , and the filling coefficient is 95.4%, and the width of each level and the stacking thickness are:

[0091] The first level, the width is 175 mm, and the stacking thickness is 52.8 mm;

[0092] The second level, the width is 170 mm, and the stacking thickness is 37.2 mm;

[0093] The third level, the width is 165 mm, and the stacking thickness is 29.8 mm;

[0094] The fourth level, the width is 155 mm, and the stacking thickness is 26.6 mm;

[0095] The fifth level, the width is 145 mm, and the stacking thickness is 27.6 mm;

[0096] The sixth level, the width is 130 mm, and the stacking thickness is 25.8 mm;

[0097] The seventh level, the width is 115 mm, and the stacking thickness is 24.9 mm;

[0098] The eighth level, the width is 95 mm, and the stacking thickness is 24.1 mm;

[0099] The ninth level, the width is 70 mm, and the stacking thickness is 20.6 mm;

[0100] The tenth level, the width is 40 mm, and the stacking thickness is 16.5 mm;

[0101] 4) Determine whether the operating magnetic density of the core column under the first high-grade silicon steel material A material1 is less than the knee point:

[0102] The operating magnetic density B

[0103] U = 4.44 x f x N x B x S (10) U refers to the effective value of the excitation voltage, f is the operating frequency, N is the number of turns on the secondary side of the transformer, and S is the optimal cross-section S0 of the silicon steel sheet obtained in step 3), and the operating magnetic density here refers to the average magnetic density passing through the cross-section of the core column.

[0104] The specific data of Abh1 is that the independent variable x of Abh1 is the magnetic field intensity H, and the dependent variable y is the magnetic flux density B. It is judged whether the calculated operating magnetic density is less than or equal to the magnetic density value corresponding to the knee point. If the calculated operating magnetic density is less than or equal to the magnetic density value corresponding to the knee point, the next step is entered, otherwise, it is returned to 3) assigning the second high-grade silicon steel material Amaterial2 to the core column, and calculating the optimal cross section of the core column of the second high-grade silicon steel material again, and then judging the magnetic density constraint until the high-grade silicon steel material meeting the magnetic density constraint is found.

[0105] 5) In the case where the operating magnetic density is less than the magnetic density value corresponding to the knee point, the no-load loss and the cost of the core column of the current high-grade silicon steel material are calculated according to formulas (11) and (12):

[0106] P0=K0×p0×G0 (11)

[0107] cost0=G0×price0×10 -3 (12)

[0108] Wherein, P0 represents the no-load loss of the part (here, the core column part), K0 represents the process additional coefficient of the no-load loss of the part, p0 represents the unit loss of the silicon steel sheet of the part (which will change with the operating magnetic density), and the corresponding unit loss can be obtained through the BP curve formula (1) at the determined magnetic density; cost0 represents the cost of the part (in ten thousand yuan), and price0 represents the unit price of the material of the part (in ten thousand yuan per ton). G0 represents the total weight of the part (in kilograms).

[0109] The first high-grade silicon steel material A material1 meets the magnetic density constraint, at this time the no-load loss of the core column is p core1 , and the cost is cost core1 . The total weight G core1 of the three core columns is calculated by the following formula:

[0110]

[0111] density is the density of the silicon steel material, in kg / m 3 ; height is the window height, in mm.

[0112] 6) Initially, the first low-grade silicon steel material Bmaterial1 is assigned to the yoke, the yoke column cross section ratio is set to 1, the yoke cross section area is equal to the core column cross section area, and S0 is assigned to the yoke.

[0113] 7) Determine whether the magnetic flux density of the yoke is less than the knee point under the first type of low-grade silicon steel material Bmaterial1:

[0114] According to formula (10), when the effective value of excitation voltage U, operating frequency f, number of secondary turns N, and cross-sectional area S (S=S0) are known, the operating magnetic flux density of the yoke can be calculated (at this time, the operating magnetic flux density of the yoke and the operating magnetic flux density of the core column are the same).

[0115] Call the specific data of Bbh1, where the independent variable x of Bbh1 is the magnetic field strength H and the dependent variable y is the magnetic flux density B. Determine whether the calculated operating magnetic flux density is less than or equal to the magnetic flux density value corresponding to the knee point. If the calculated operating magnetic flux density is less than or equal to the magnetic flux density value corresponding to the knee point, proceed to the next step; otherwise, return to step 6). Assign the second type of low-grade silicon steel material Bmaterial2 to the iron yoke, and then determine the magnetic flux density constraint until a low-grade silicon steel material that meets the magnetic flux density constraint is found.

[0116] 8) Calculate the no-load loss and cost of the yoke:

[0117] The total weight of the two yokes, G yoke1 for:

[0118]

[0119] Where distance is the center-to-center distance between two adjacent core columns.

[0120] The unit loss of the corresponding low-grade steel material can be obtained by using the BP fitting curve formula (1) when the magnetic flux density is determined. Then, the no-load loss p of the yoke that meets the knee point requirement can be calculated by formulas (11) and (12). yoke1 and cost yoke1 .

[0121] 9) Comparative Iterative Storage: Judgment Whether it is true or false: If 1) is true but 2) is false, then p core1 +p yoke1 Assign it to Loss0, cost core1 +cost yoke1 Assigning cost0, storing the corresponding solution in solution library 1 (aimed at minimizing loss), and continuing to iterate for the next material solution; if 1) is false but 2) is true, then p core1 +p yoke1 Assign it to Loss0, cost core1 +cost yoke1 Assigning cost to 0, storing the corresponding solution in the solution library 2 (aimed at minimizing cost), and continuing to iterate for the next material solution; if both 1) and 2) are true, then p core1 +p yoke1Loss0, cost core1 +cost yoke1 Assign cost0, store the corresponding scheme into the scheme database for the purpose of minimum loss and minimum cost, continue to iterate the next material scheme; until all materials in the high-grade silicon steel material database are iterated.

[0122] Three scheme databases are established: scheme database 1 retains the scheme with minimum no-load loss of the transformer, scheme database 2 retains the scheme with minimum cost of the transformer, and scheme database 3 retains the scheme with both minimum no-load loss and minimum cost of the transformer. In scheme database 3, the no-load loss can be greater than that in scheme database 1, and the cost can be less than that in scheme database 1, or the no-load loss can be less than that in scheme database 2, and the cost can be greater than that in scheme database 2. The three scheme databases are retained for the convenience of users to select as needed. In the three scheme databases, it is determined that the core column selects which high-grade silicon steel material and the iron yoke selects which low-grade silicon steel material, and the target can be achieved.

[0123] 10) Output three sets of optimized schemes: scheme 1 with minimum no-load loss of the transformer, scheme 2 with minimum cost of the transformer, and scheme 3 with both minimum no-load loss and minimum cost of the transformer.

[0124] The unmentioned part of the present application is applicable to the prior art.

Claims

1. A design method for transformer cores based on a hybrid magnetic circuit of high and low grade silicon steel, characterized in that, The design method includes the following: The core column and yoke of the transformer core are made of two different materials: high-grade silicon steel and low-grade silicon steel, respectively. A database of high-grade silicon steel materials is formed by packaging various high-grade silicon steel materials with their model, BH curve, BP fitting curve, unit price, and no-load loss additional coefficient. The model, BH curve, BP fitting curve, unit price, and no-load loss additional coefficient of various low-grade silicon steel materials are packaged to form a database of low-grade silicon steel materials. The process for determining the BP fitting curves for each material in the two databases is as follows: On the BH curve, find the point where the first derivative is maximum and the second derivative is zero as the knee point K. Move the BH curve by an interval of 0.1T, calculating the slope within each interval. When the slope of three consecutive intervals is less than 0.3 times the maximum value of the first derivative for the first time, determine the first starting point that meets the condition as the saturation point Sa. Determine the magnetic flux density values ​​corresponding to the knee point K and the saturation point Sa, and then use the magnetic flux density values ​​B corresponding to the knee point K and the saturation point Sa as the basis for the determination. K B Sa Piecewise fitting of the BP curve, the final BP fitting curve is expressed by formula (1): Where p is the unit loss, B is the magnetic flux density, and a1, b1, k, c1, d1, p s p k These are the fitting parameters; Start optimization: Input the major axis and minor axis of the circumscribed ellipse, the number of core stages, the database of high-grade silicon steel materials, the database of low-grade silicon steel materials, the initial no-load loss Loss0, and the initial cost cost0; First, the first type of high-grade silicon steel material is assigned to the core column. With the maximum total cross-section of the laminated sheets as the optimization objective, a genetic algorithm is used to optimize the core column cross-section to obtain the optimal cross-section S0 of the core column with the current high-grade silicon steel material. Based on S0, the effective value of the excitation voltage, and the number of turns on the secondary side of the transformer, the operating magnetic flux density of the current core column is calculated. Then, in the high-grade silicon steel material database, the corresponding high-grade silicon steel material with an operating magnetic flux density less than the magnetic flux density corresponding to the knee point K is determined. After determining the high-grade silicon steel material to be selected for the core column, the database of low-grade silicon steel materials is traversed to identify the corresponding low-grade silicon steel material with an operating magnetic flux density less than the knee point K. The high-grade silicon steel material that meets the requirements is used as the core column, and the low-grade silicon steel material that meets the requirements is used as the yoke to form the transformer core. The corresponding unit loss is calculated using the BP fitting curve based on the operating magnetic flux density of the materials, and then the no-load loss and cost of the entire transformer core are calculated. Scheme 1, which minimizes the no-load loss of the output transformer; Scheme 2, which minimizes the transformer cost; and Scheme 3, which minimizes both the no-load loss and cost of the transformer.

2. The design method according to claim 1, characterized in that, The core is a three-phase, three-limb distribution transformer core. The data codes in the high-grade silicon steel material library are: [Amaterial1, Abh1, Abp1, AK1, Amaterial2, Abh2, Abp2, AK2, ... Abh T Abp T AK T ], where Amaterial1 represents the model of the first high-grade silicon steel material, Abh1 represents the BH curve of the first high-grade silicon steel material, Abp1 represents the BP fitting curve of the first high-grade silicon steel material, AK1 represents the no-load loss process additional coefficient of the first high-grade silicon steel material, and the rest of the codes follow the same pattern, and T represents the number of types of high-grade silicon steel materials. The data codes for low-grade silicon steel materials are: [Bmaterial1, Bbh1, Bbp1, BK1, Bmaterial2, Bbh2, Bbp2, BK2, ... Bbh W Bbp W BK W ]; where Bmaterial1 represents the model of the first low-grade silicon steel material, Bbh1 represents the BH curve of the first low-grade silicon steel material, Bbp1 represents the BP fitting curve of the first low-grade silicon steel material, BK1 represents the no-load loss additional coefficient of the first low-grade silicon steel material, and the remaining codes follow the same pattern, and W represents the number of types of low-grade silicon steel materials. Encode the stacking parameters at each level: [width1, thickness1, width2, thickness2, ..., width i thickness i , ..., width n thickness n ], where width1 represents the width of the first-level stack, and thickness1 represents the stack thickness of the first-level stack, both in mm. n The thickness represents the width of the nth-level stack. n This indicates the thickness of the nth lamination, in mm; n is the core stage number; width i The thickness represents the width of the i-th stacked piece. i This represents the stack thickness of the i-th layer, in mm; i = 1, 2, 3…n.

3. The design method according to claim 1, characterized in that, The constraints of the genetic algorithm are: Set the width of the last level. n If the minimum value and the minimum thickness of a single stack of the first-level stack are both the minimum values, then the width of the last-level stack and the thickness of a single stack of the first-level stack are not less than their minimum values. The width of the image decreases in a decreasing pattern, with the width of level i being greater than the width of level i+1. The stacking thickness shows an increasing trend, with the stacking thickness of layer i being less than that of layer i+1, and the stacking thickness of layer n being less than the major axis of the ellipse. If the remainder when the width of each piece is divided by 5 is less than or equal to 2.5, then round down; if the remainder when the width of each piece is divided by 5 is greater than 2.5 but less than 5, then round up. The width of each piece satisfies formula (8): Where, d i width of the i-th level piece i The remainder when divided by 5; If the rectangular stacked pieces are located inside the circumscribed ellipse, then they satisfy formula (9):

4. A hybrid magnetic circuit transformer core, characterized in that, The hybrid magnetic circuit transformer core is obtained by any of the design methods described in claims 1-3. The core column of the hybrid magnetic circuit transformer core is made of high-grade silicon steel, and the yoke of the hybrid magnetic circuit transformer core is made of low-grade silicon steel. The core column and the yoke have the same number of laminations and the same cross-section.

5. A design system for a transformer core based on a hybrid magnetic circuit of high and low grade silicon steel, characterized in that, The system includes: A database of high-grade silicon steel materials, used to store relevant parameters of various high-grade silicon steel materials; A database of low-grade silicon steel materials, used to store relevant parameters of various low-grade silicon steel materials; Intelligent optimization algorithms are used to obtain the optimal cross-section of core columns made of different types of materials; The magnetic flux density constraint module is used to determine whether the running magnetic flux density is less than the magnetic flux density corresponding to the knee point K of the current material, based on the average magnetic flux density through the cross section of the iron core column. If it is less, the current material is retained. The materials used for the core column are obtained from a database of high-grade silicon steel materials, while the materials used for the yoke are obtained from a database of low-grade silicon steel materials. The magnetic flux density constraint module is constrained under the premise of obtaining the optimal cross-section of the core by the intelligent optimization algorithm, with the goal of minimizing the total cost and total no-load loss of the overall transformer core composed of the core column and the yoke.

Images