Method for designing performance optimization model of magnetic element under multiple factors

Abstract

The invention provides a magnetic element performance optimization model design method under multiple factors, and belongs to the technical field of electric energy conversion of power electronics. The technical problem that the accurate calculation of the magnetic core loss under different temperature conditions is low is solved. According to the technical scheme, the method comprises the following steps of S1, performing data measurement to form an original data set, and performing data processing to establish a new data set; s2, improving an iGSE magnetic core loss function; s3, fitting unknown parameters of the magnetic core loss function by a nonlinear least square method; s4, constructing a weighted single-target optimization model; and S5, solving the optimal solution of the target function by a fire bird optimization algorithm. The method has the beneficial effects that the improved iGSE magnetic core loss function can be suitable for various temperature changes, the constructed weighted single target model can more comprehensively optimize the performance of the magnetic element, and the fire bird optimization algorithm can more quickly converge the optimal solution of the target function.

Description

Technical Field

[0001] The invention relates to the technical field of electric energy conversion of power electronics, and in particular to a design method for a magnetic element performance optimization model under multiple factors. Background Art

[0002] With the development of the times and the advancement of science and technology, power conversion technology is also evolving. It can convert the electric energy we are familiar with from one form to another. Power conversion of power electronics technology plays an important role in many fields, especially in current hot fields such as new energy and electric vehicles, as well as future industrial automation and uninterruptible power supply (UPS) systems. In these applications, magnetic components in power converters play a vital role.

[0003] Core loss, that is, the power loss generated by magnetic materials under the action of high-frequency alternating magnetic fields, is a key indicator for evaluating the performance of magnetic components. Although there are empirical calculation formulas, there may still be significant errors in the calculation of core loss under different temperatures, excitation waveforms and magnetic materials. In addition, in order to optimize the overall performance of magnetic components in engineering practice, the indicator of transmitted magnetic energy must also be considered. Although the Steinmetz equation and its related correction equations have been proposed, there is currently no universal correction equation that can adapt to continuous temperature changes and other factors. For the two key evaluation indicators of core loss and transmitted magnetic energy, their reasonable integration method has not yet been fully developed.

[0004] In actual engineering applications, due to the variability of the situation, factors such as temperature, excitation waveform, material type, etc. may become the main factors affecting the performance of magnetic components. Therefore, in-depth analysis of these influencing factors is indispensable in order to more accurately evaluate and optimize the performance of magnetic components. Summary of the invention

[0005] The purpose of the present invention is to solve the technical problems of calculating the core loss at different temperatures and optimizing the performance evaluation index of magnetic components under multiple factors, and to propose a design method for a magnetic component performance optimization model under multiple factors.

[0006] In order to achieve the above-mentioned invention object, the technical scheme adopted by the present invention is as follows: a method for designing a magnetic component performance optimization model under multiple factors, comprising the following steps:

[0007] S1. Measure data under different working conditions (different temperatures, frequencies, excitation waveforms) to form the original data set W 0 ;

[0008] S2, construct an improved iGSE core loss function;

[0009] S3. Use the least square method to fit the unknown parameters K, α, β, γ, and θ in the core loss function under different working conditions;

[0010] S4. Combined with the transmission magnetic energy, a dual-objective optimization model is constructed to minimize the core loss and maximize the transmission magnetic energy;

[0011] S5. Solve the weighted single-objective optimization model using the Flamingo optimization algorithm to find its minimum value;

[0012] As a preferred technical solution of the present invention: the step S1 specifically comprises the following steps:

[0013] S1.1. Establish a core loss calculation model under different temperatures, different excitation waveforms, and different material types, measure data under different working conditions, and establish the original data set W 0 ={C,f,Waveform,P,Material,B}, where C is temperature data, f is frequency data, Waveform is excitation waveform data, P is core loss data, Material is material type data, and B is magnetic flux density data;

[0014] S1.2, for the magnetic flux density data set B under the i-th working condition i,j Peak value Where i = 1, 2, 3...12300; j represents the sampling point, j = 1, 2, 3...1024;

[0015] S1.3. Take the minimum value of the magnetic flux density data under the i-th working condition in the magnetic flux density data set B Then calculate the difference ΔB between the maximum and minimum values ​​of the magnetic flux density under the i-th working condition i , which is taken as the peak-to-peak value under the i-th working condition;

[0016] S1.4. From the original dataset W 0 Extract temperature data, excitation waveform data, frequency data, core loss data, material type data, add peak value and peak-to-peak value, and create a new data set

[0017] As a preferred technical solution of the present invention: the step S2 specifically includes the following steps:

[0018] S2.1. The original iGSE core loss function without considering the temperature factor is:

[0019]

[0020] Among them, K, α, β, θ are the 0The unknown parameters solved by fitting the magnetic flux density, excitation waveform and frequency data under different working conditions, ΔB i is the peak-to-peak value under the i-th working condition, T i is the cycle under the i-th working condition, is the rate of change of magnetic flux;

[0021] S2.2. The flux change rate in the original iGSE core loss function without considering the temperature factor Discretization, the discretization is: converting ΔB into the difference of magnetic flux density of 1024 sampling points under the i-th working condition, and converting Δt into 1024 equal parts of the cycle under the i-th working condition, which is specifically expressed as:

[0022]

[0023] S2.3. Improvements are made to the original iGSE core loss function that does not consider the temperature factor, by adding the temperature factor:

[0024]

[0025] Among them, γ is the number of 0 The unknown parameters solved by fitting the magnetic flux density, excitation waveform and frequency data under different working conditions, C i is the temperature under the i-th working condition.

[0026] As a preferred technical solution of the present invention: the step S3 specifically comprises the following steps:

[0027] S3.1. Extracting new dataset W 1 The excitation waveform and material type data in the filter can be used to filter out the temperature, frequency, peak value of magnetic flux density, peak-to-peak value of magnetic flux density, and core loss data under specific excitation waveform and material type;

[0028] S3.2, set the initial prediction value of the nonlinear least squares method, K = 1.0, α = 1.0, β = 1.0, γ = 0.1, θ = 0.1;

[0029] S3.3. Using the new dataset W 1 The temperature, frequency, peak value of magnetic flux density, peak-to-peak value of magnetic flux density, and core loss data extracted from the test results are used to perform nonlinear least squares fitting on the improved iGSE core loss function.

[0030] S3.4. Obtain the results through fitting and record the five unknown parameters K, α, β, γ, and θ for all combinations of excitation waveforms and material types.

[0031] As a preferred technical solution of the present invention: the step S4 specifically comprises the following steps:

[0032] S4.1. From the new dataset W 1 Extract the frequency f i , peak flux density data;

[0033] S4.2, combined with another magnetic component performance evaluation index, the frequency f i and peak flux density The product of is recorded as the transmitted magnetic energy;

[0034] S4.3. Construct a dual-objective optimization function to minimize core loss and maximize transmitted magnetic energy:

[0035]

[0036] S4.4. In order to make both of them in the optimal state, the dual-objective optimization function is transformed into a weighted single-objective optimization model:

[0037]

[0038] Minimize the difference between the core loss calculation function and the transmitted magnetic energy. 1 、s 2 are the weights of the core loss function and the transmitted magnetic energy function respectively;

[0039] As a preferred technical solution of the present invention: the step S5 specifically comprises the following steps:

[0040] S5.1, according to the input temperature C i , fitting parameters K, α, β, θ, γ, peak value of magnetic flux density Magnetic flux density peak-to-peak value ΔB i , frequency f i , weights 1 、s 2 , the specific objective function of its optimization is:

[0041]

[0042] S5.2. Based on the new dataset W 1 The initial flamingo population is randomly generated using the four temperatures, three excitation waveforms, and three material types. The specific formation form of the population is:

[0043] P pop =rand[0,2]*rand[0,3]*rand[0,3] (7)

[0044] The initial size of the population is set to N, and the maximum number of iterations of the population is M gen ;

[0045] S5.3. Specifying weights s1 、s 2 Each generation grows by g 1 , reduce g 2 The form changes dynamically, and the maximum value of the iterative algebra is M gen , dynamically update the population and find the global optimal individual Loss in the current generation gl_best ;

[0046] S5.4. Distinguish between foraging and migrating flamingos, mutate according to the difference between the global optimal individual and the current individual, generate a new individual and calculate its objective function value Loss new , and the objective function value Loss of the current individual cur For comparison, the ones with inferior objective function values ​​are migratory flamingos, and the others are foraging flamingos;

[0047] S5.5. Update the position of migrating flamingos according to the following formula:

[0048]

[0049] in, is the position of the i-th flamingo in the population after the m+1th iteration, is the position of the i-th flamingo in the population after the m-th iteration, Ind m is the location of the flamingo with the optimal objective function value in the mth iteration, and σ is a random number with n degrees of freedom, which is used to simulate the randomness of the behavior of individual flamingos during the entire migration process;

[0050] S5.6. Specify the range limit conditions for flamingo migration and determine the migration distance according to the following formula:

[0051]

[0052] Among them, D i m is the migration distance of the i-th flamingo in the m-th iteration, ξ 1 , 2 is a random number between -1 and 1, X 1 , X 2 is a random number that follows a normal distribution;

[0053] S5.7, continuously update the optimal objective function value Loss according to iteration lst_best Finally, the optimal performance index of the magnetic component at different temperatures and arbitrary excitation waveforms and its corresponding operating parameters are obtained.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] (1) In the original iGSE core loss model, the temperature factor was not fully considered, resulting in significant differences between the actual value of the core loss and the model prediction value under different temperature conditions, especially in the range from low temperature to high temperature. Traditional calculation methods often only consider factors such as excitation waveform and material type, ignoring the significant effect of temperature on core loss, thereby introducing large calculation errors. Although in previous studies, very few works have attempted to consider the effect of temperature changes on core loss (for example, the temperature factor introduced in CN201810616405.6 is ), but the calculation results of these improved methods still deviate greatly from the actual situation. Therefore, in order to improve the applicability and accuracy of the iGSE core loss model under different temperature conditions, the present invention introduces an innovative temperature factor. By performing box plot analysis on the measured data, we found that the core loss decreases with increasing temperature. Based on this finding, the present invention designs a temperature factor in the form of a decreasing function to accurately capture the effect of temperature changes on core loss. Compared with the prior art, the temperature factor of the present invention has a significant improvement in calculation accuracy, effectively compensating for the effect of temperature changes on the calculation of core loss, so that the improved iGSE core loss model can be more accurately applied to various temperature environments.

[0056] (2) In the prior art, the methods for evaluating the performance of magnetic components are relatively limited, especially when considering multiple influencing factors, the performance evaluation indicators that can be used are even more scarce. In order to improve the persuasiveness and applicability of the evaluation method, the present invention introduces another key performance evaluation indicator, the transmission magnetic energy, in addition to the traditional commonly used core loss indicator. In view of the fact that the transmission magnetic energy indicator is rarely used in the prior art, and there is a lack of methods to effectively combine core loss with transmission magnetic energy, the present invention proposes a dual-objective weighted method. Through this method, the two performance evaluation indicators can be reasonably integrated, thereby making the performance evaluation of the magnetic component more comprehensive. This integration not only covers a wider range of performance aspects, but also can show excellent evaluation effects when facing multiple influencing factors.

[0057] (3) In order to find an optimization algorithm that can perform global search and local refinement at the same time, the present invention adopts the Flamingo optimization algorithm in the optimization link of the objective function. Compared with the optimization algorithms widely used in the prior art, the Flamingo optimization algorithm not only has the ability of global search and local refinement, but also performs well in terms of convergence speed, optimization accuracy and stability. These characteristics jointly improve the efficiency of finding the optimal solution of the objective function. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 A flow chart of a method for designing a magnetic component performance optimization model under multiple factors is proposed for the present invention;

[0059] Figure 2 A flow chart of "data measurement to form an original data set and data processing to establish a new data set" for a method for designing a magnetic component performance optimization model under multiple factors is proposed for the present invention;

[0060] Figure 3 The present invention proposes a flow chart of "improved iGSE core loss function" for a magnetic component performance optimization model design method under multiple factors;

[0061] Figure 4 A flow chart of "nonlinear least square method for fitting unknown parameters of core loss function" for a magnetic component performance optimization model design method under multiple factors is proposed for the present invention;

[0062] Figure 5 A flow chart of "constructing a weighted single-objective optimization model" is provided for the present invention, which is a method for designing a magnetic component performance optimization model under multiple factors;

[0063] Figure 6 A flow chart of "Flamingo optimization algorithm to solve the optimal solution of objective function" for a magnetic component performance optimization model design method under multiple factors proposed by the present invention;

[0064] Figure 7 Comparison chart of the predicted value and the true value of the original iGSE core loss function;

[0065] Figure 8 To improve the iGSE core loss function, the predicted value and the actual value comparison chart;

[0066] Fig. 9 This is a comparison chart between the predicted value and the actual value of the core loss function with the temperature factor added in the prior art. DETAILED DESCRIPTION

[0067] The present invention is further explained below in detail with reference to the accompanying drawings so that those skilled in the art can more deeply understand the present invention and be able to implement it. However, the following reference examples are only used to explain the present invention and are not intended to limit the present invention.

[0068] Example 1

[0069] See also Figure 1 As shown, this embodiment provides a method for designing a magnetic component performance optimization model under multiple factors, comprising the following steps:

[0070] S1. Measure data under different working conditions (different temperatures, frequencies, excitation waveforms) to form the original data set W 0 ;

[0071] S2, construct an improved iGSE core loss function;

[0072] S3. Use the least square method to fit the unknown parameters K, α, β, γ, and θ in the core loss function under different working conditions;

[0073] S4. Combined with the transmission magnetic energy, a dual-objective optimization model is constructed to minimize the core loss and maximize the transmission magnetic energy;

[0074] S5. Solve the weighted single-objective optimization model using the Flamingo optimization algorithm to find its minimum value;

[0075] like Figure 2 As shown, step S1 specifically includes the following steps:

[0076] S1.1. Establish a core loss calculation model under different temperatures, different excitation waveforms, and different material types, measure data under different working conditions, and establish the original data set W 0 ={C,f,Waveform,P,Material,B}, where C is temperature data, f is frequency data, Waveform is excitation waveform data, P is core loss data, Material is material type data, and B is magnetic flux density data;

[0077] S1.2, for the magnetic flux density data set B under the i-th working condition i,j Peak value Where i = 1, 2, 3...12300; j represents the sampling point, j = 1, 2, 3...1024;

[0078] S1.3. Take the minimum value of the magnetic flux density data under the i-th working condition in the magnetic flux density data set B Then calculate the difference ΔB between the maximum and minimum values ​​of the magnetic flux density under the i-th working condition i , which is taken as the peak-to-peak value under the i-th working condition;

[0079] S1.4. From the original dataset W 0 Extract temperature data, excitation waveform data, frequency data, core loss data, material type data, add peak value and peak-to-peak value, and create a new data set like Figure 3 As shown, step S2 specifically includes the following steps:

[0080] S2.1. The original iGSE core loss function without considering the temperature factor is:

[0081]

[0082] Among them, K, α, β, θ are the 0 The unknown parameters solved by fitting the magnetic flux density, excitation waveform and frequency data under different working conditions, ΔB iis the peak-to-peak value under the i-th working condition, T i is the cycle under the i-th working condition, is the rate of change of magnetic flux;

[0083] S2.2. The flux change rate in the original iGSE core loss function without considering the temperature factor Discretization, the discretization is: converting ΔB into the difference of magnetic flux density of 1024 sampling points under the i-th working condition, and converting Δt into 1024 equal parts of the cycle under the i-th working condition, which is specifically expressed as:

[0084]

[0085] S2.3. Improvements are made to the original iGSE core loss function that does not consider the temperature factor, by adding the temperature factor:

[0086]

[0087] Among them, γ is the number of 0 The unknown parameters solved by fitting the magnetic flux density, excitation waveform and frequency data under different working conditions, C i is the temperature under the i-th working condition.

[0088] like Figure 4 As shown, step S3 specifically includes the following steps:

[0089] S3.1. Extracting new dataset W 1 The excitation waveform and material type data in the filter can be used to filter out the temperature, frequency, peak value of magnetic flux density, peak-to-peak value of magnetic flux density, and core loss data under specific excitation waveform and material type;

[0090] S3.2, set the initial prediction value of the nonlinear least squares method, K = 1.0, α = 1.0, β = 1.0, γ = 0.1, θ = 0.1;

[0091] S3.3. Using the new dataset W 1 The temperature, frequency, peak value of magnetic flux density, peak-to-peak value of magnetic flux density, and core loss data extracted from the test results are used to perform nonlinear least squares fitting on the improved iGSE core loss function.

[0092] S3.4. Obtain the results through fitting and record the five unknown parameters K, α, β, γ, and θ for all combinations of excitation waveforms and material types.

[0093] like Figure 5 As shown, step S4 specifically includes the following steps:

[0094] S4.1. From the new dataset W 1 Extract the frequency fi , peak flux density data;

[0095] S4.2, combined with another magnetic component performance evaluation index, the frequency f i and peak flux density The product of is recorded as the transmitted magnetic energy;

[0096] S4.3. Construct a dual-objective optimization function to minimize core loss and maximize transmitted magnetic energy:

[0097]

[0098] S4.4. In order to make both of them in the optimal state, the dual-objective optimization function is transformed into a weighted single-objective optimization model:

[0099]

[0100] Minimize the difference between the core loss calculation function and the transmitted magnetic energy. 1 、s 2 are the weights of the core loss function and the transmitted magnetic energy function respectively;

[0101] like Figure 6 As shown, step S5 specifically includes the following steps:

[0102] S5.1, according to the input temperature C i , fitting parameters K, α, β, θ, γ, peak value of magnetic flux density Magnetic flux density peak-to-peak value ΔB i , frequency f i , weights 1 、s 2 , the specific objective function of its optimization is:

[0103]

[0104] S5.2. Based on the new dataset W 1 The initial flamingo population is randomly generated using the four temperatures, three excitation waveforms, and three material types. The specific formation form of the population is:

[0105] P pop =rand[0,2]*rand[0,3]*rand[0,3] (7)

[0106] The initial size of the population is set to N, and the maximum number of iterations of the population is M gen ;

[0107] S5.3. Specifying weights s 1 、s 2 Each generation grows by g1 , reduce g 2 The form changes dynamically, and the maximum value of the iterative algebra is M gen , dynamically update the population and find the global optimal individual Loss in the current generation gl_best ;

[0108] S5.4. Distinguish between foraging and migrating flamingos, mutate according to the difference between the global optimal individual and the current individual, generate a new individual and calculate its objective function value Loss new , and the objective function value Loss of the current individual cur For comparison, the ones with inferior objective function values ​​are migratory flamingos, and the others are foraging flamingos;

[0109] S5.5. Update the position of migrating flamingos according to the following formula:

[0110]

[0111] in, is the position of the i-th flamingo in the population after the m+1th iteration, is the position of the i-th flamingo in the population after the m-th iteration, Ind m is the location of the flamingo with the optimal objective function value in the mth iteration, and σ is a random number with n degrees of freedom, which is used to simulate the randomness of the behavior of individual flamingos during the entire migration process;

[0112] S5.6. Specify the range restrictions for flamingo migration and determine the migration distance according to the following formula:

[0113]

[0114] in, is the migration distance of the i-th flamingo in the m-th iteration, ξ 1 , 2 is a random number between -1 and 1, X 1 , X 2 is a random number that follows a normal distribution;

[0115] S5.7, continuously update the optimal objective function value Loss according to iteration lst_best Finally, the optimal performance index of the magnetic component at different temperatures and arbitrary excitation waveforms and its corresponding operating parameters are obtained.

[0116] In a specific implementation manner, the description of each symbol is shown in Table 1:

[0117] Table 1 Symbols

[0118]

[0119]

[0120] This embodiment provides a method for designing a magnetic component performance optimization model under multiple factors. This embodiment adds a temperature factor to the traditional iGSE core loss model. This improvement improves the accuracy of its core loss calculation and enables it to adapt to various temperature changes. At the same time, in order to optimize the performance of magnetic components, this example also combines another magnetic component performance evaluation index, namely the transmission magnetic energy, and more reasonably integrates the two in a weighted form to convert it into a weighted single-objective model. By comparing and analyzing the previous optimization methods for the optimal solution of the objective function, this example uses the Flamingo optimization algorithm to perform global optimal search and local utilization of the final objective function, which greatly improves the convergence speed of the optimal solution of the objective function.

[0121] Example 2

[0122] This example provides a comparative experiment of the original iGSE core loss function (1) and the improved iGSE core loss function (3) of the present invention, including the following steps:

[0123] S1. Based on Example 1, extract the data set W required for the comparative experiment 0 ;

[0124] S2. Use the least squares method to fit the unknown parameters K, α, β, and θ in the original iGSE core loss function under different working conditions;

[0125] S3, calculate the mean square error RMSE of the original iGSE core loss function and the improved iGSE core loss function, compare the two, and select the optimal core loss function;

[0126] like Figure 7 and Figure 8 As shown, Figure 7 This is a comparison chart of the predicted value and the actual value of the original iGSE core loss function. Figure 8 The comparison chart of the predicted value and the true value of the improved iGSE core loss function. It is not difficult to see from the figure that the mean square error RMSE of the original iGSE core loss function is 77692.61, which is obviously larger than the mean square error RMSE of the improved iGSE core loss function, 42520.87, and the true value and the predicted value are in Figure 7 More dispersed in Figure 8 The coincidence between the real value and the predicted value is obviously better. Therefore, it can be concluded that the improved iGSE core loss function of the present invention is better than the original iGSE core loss function.

[0127] Example 3

[0128] This example provides a comparative experiment of the core loss function with a temperature factor added in CN201810616405.6 and the improved iGSE core loss function (3) of the present invention, including the following steps:

[0129] S1. Based on Example 1, extract the data set W required for the comparative experiment 0 ;

[0130] S2. Using the least square method to fit the unknown parameters K, α, β, and θ in the core loss function with the temperature factor added in the prior art under different working conditions;

[0131] S3. Calculate the mean square error RMSE of the core loss function with the temperature factor added in the prior art, and compare it with the mean square error RMSE of the core loss function of the improved iGSE of the present invention on the basis of Example 2, and select the optimal core loss function;

[0132] like Fig. 9 As shown, Fig. 9 The figure is a comparison chart of the predicted value and the actual value of the core loss function with the temperature factor added in the prior art. Figure 8 and Fig. 9 By comparison, it is not difficult to see that the temperature factor added by CN201810616405.6 is not as good as the temperature factor in the form of a subtracting function added by the present invention. Obviously, the mean square error RMSE of the former is 194051.71, which means that the temperature factor in the form of a subtracting function introduced by the present invention has higher accuracy.

[0133] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for designing a magnetic component performance optimization model under multiple factors, characterized in that: The following steps are involved: S1, measure data under different working conditions, different temperatures, frequencies, and excitation waveforms to form an original data set W0; S2, construct an improved iGSE core loss function; S3. Use the least square method to fit the unknown parameters K, α, β, γ, and θ in the core loss function under different working conditions; S4. Combined with the transmission magnetic energy, a dual-objective optimization model is constructed, i.e., minimizing the core loss and maximizing the transmission magnetic energy; S5. Use the Flamingo optimization algorithm to solve the weighted single-objective optimization model and find its minimum value.

2. The method for designing a magnetic component performance optimization model under multiple factors according to claim 1, characterized in that: The step S1 comprises the following steps: S1.

1. Establish a core loss calculation model under different temperatures, different excitation waveforms, and different material types, measure data under different working conditions and establish the original data set W0 = {C, f, Waveform, P, Material, B}, where C is temperature data, f is frequency data, Waveform is excitation waveform data, P is core loss data, Material is material type data, and B is magnetic flux density data; S1.2, for the magnetic flux density data set B under the i-th working condition i,j Peak value Where i = 1, 2, 3...12300; j represents the sampling point, j = 1, 2, 3...1024; S1.

3. Take the minimum value of the magnetic flux density data under the i-th working condition in the magnetic flux density data set B Then calculate the difference ΔB between the maximum and minimum values ​​of the magnetic flux density under the i-th working condition i , which is taken as the peak-to-peak value under the i-th working condition; S1.

4. Extract temperature data, excitation waveform data, frequency data, core loss data, and material type data from the original data set W0, add peak values ​​and peak-to-peak values, and create a new data set 3. The method for designing a magnetic component performance optimization model under multiple factors according to claim 1, characterized in that: The step S2 comprises the following steps: S2.

1. The original iGSE core loss function without considering the temperature factor is: Among them, K, α, β, and θ are unknown parameters that need to be solved based on the magnetic flux density, excitation waveform, and frequency data under different working conditions in the original data set W0. i is the peak-to-peak value under the i-th working condition, T i is the cycle under the i-th working condition, is the rate of change of magnetic flux; S2.

2. The flux change rate in the original iGSE core loss function without considering the temperature factor Discretization, the discretization is: converting ΔB into the difference of magnetic flux density of 1024 sampling points under the i-th working condition, and converting Δt into 1024 equal parts of the cycle under the i-th working condition, which is specifically expressed as: S2.

3. Improvements are made to the original iGSE core loss function that does not consider the temperature factor, by adding the temperature factor: Among them, γ is the unknown parameter that needs to be solved by fitting the magnetic flux density, excitation waveform, and frequency data under different working conditions in the original data set W0, C i is the temperature under the i-th working condition.

4. The method for designing a magnetic component performance optimization model under multiple factors according to claim 1, characterized in that: The step S3 comprises the following steps: S3.1, extract the excitation waveform and material type data in the new data set W1, and filter out the temperature, frequency, magnetic flux density peak value, magnetic flux density peak-to-peak value, and core loss data under specific excitation waveform and material type; S3.2, set the initial prediction value of the nonlinear least squares method, K = 1.0, α = 1.0, β = 1.0, γ = 0.1, θ = 0.1; S3.3, using the temperature, frequency, peak value of magnetic flux density, peak-to-peak value of magnetic flux density, and core loss data extracted from the new data set W1, the improved iGSE core loss function is fitted by nonlinear least squares method; S3.

4. Obtain the results through fitting and record the five unknown parameters K, α, β, γ, and θ for all combinations of excitation waveforms and material types.

5. The method for designing a magnetic component performance optimization model under multiple factors according to claim 1, characterized in that: The step S4 comprises the following steps: S4.

1. Extract frequency f from the new dataset W1 i , peak flux density data; S4.2, combined with another magnetic component performance evaluation index, the frequency f i and peak flux density The product of is recorded as the transmitted magnetic energy; S4.

3. Construct a dual-objective optimization function to minimize core loss and maximize transmitted magnetic energy: S4.

4. Convert the dual-objective optimization function into a weighted single-objective optimization model: Minimize the difference between the core loss calculation function and the transmitted magnetic energy, where s1 and s2 are the weights of the core loss function and the transmitted magnetic energy function, respectively.

6. The method for designing a magnetic component performance optimization model under multiple factors according to claim 1, characterized in that: The step S5 comprises the following steps: S5.1, according to the input temperature C i , fitting parameters K, α, β, θ, γ, peak value of magnetic flux density Magnetic flux density peak-to-peak value ΔB i , frequency f i , weights s1, s2, the specific objective function to be optimized is: S5.

2. Based on the four temperatures, three excitation waveforms, and three material types in the new data set W1, the initial flamingo population is randomly generated. The specific formation form of the population is: P pop =row[0,2]*row[0,3]*row[0,3] (7) The initial size of the population is set to N, and the maximum number of iterations of the population is M gen ; S5.

3. The weights s1 and s2 are specified to change dynamically in the form of increasing g1 and decreasing g2 in each generation. The maximum value of the iteration number is M. gen , dynamically update the population and find the global optimal individual Loss in the current generation gl_best ; S5.

4. Distinguish between foraging and migrating flamingos, mutate according to the difference between the global optimal individual and the current individual, generate a new individual and calculate its objective function value Loss new , and the objective function value Loss of the current individual cur For comparison, the ones with inferior objective function values ​​are migratory flamingos, and the others are foraging flamingos; S5.

5. Update the position of migrating flamingos according to the following formula: in, is the position of the i-th flamingo in the population after the m+1th iteration, is the position of the i-th flamingo in the population after the m-th iteration, Ind m is the location of the flamingo with the optimal objective function value in the mth iteration, and σ is a random number with n degrees of freedom, which is used to simulate the randomness of the behavior of individual flamingos during the entire migration process; S5.

6. Specify the range limit conditions for flamingo migration and determine the migration distance according to the following formula: in, is the migration distance of the i-th flamingo in the m-th iteration, ξ1 and ξ2 are random numbers between -1 and 1, and X1 and X2 are random numbers following the normal distribution; S5.7, continuously update the optimal objective function value Loss according to iteration lst_best Finally, the optimal performance index of the magnetic component at different temperatures and arbitrary excitation waveforms and its corresponding operating parameters are obtained.

Images