Planar coreless transformer parameter optimization method based on genetic algorithm

Abstract

The invention relates to the technical field of power electronics and semiconductors, in particular to a planar coreless transformer parameter optimization method based on a genetic algorithm, which performs parameter optimization simulation by combining simulation data with the genetic algorithm, and performs design and parameter adjustment of an actual circuit according to an optimal result. The specific process is as follows: firstly, building a coil model in Maxwell, performing field-circuit joint simulation of the coil model and an external circuit in Simplorer, obtaining energy transmission efficiency under the condition of different parameter combinations through several times of comparison simulation and analysis, further respectively calculating the fitness of combination individuals in MATLAB according to a genetic algorithm principle, and calculating the fitness of the combination individuals according to the genetic algorithm principle. And an optimal combination is finally obtained through operations such as selection, intersection and variation, so that the optimal transmission efficiency is obtained, and data analysis and optimization are completed. Finally, circuit parameters are designed or adjusted based on the optimization result, and the problems that an existing planar coreless transformer is low in energy transmission efficiency, and the process of designing the parameters of the transformer and the circuit is complex and long in consumed time are solved.

Description

Technical Field

[0001] This invention relates to the field of power electronics and semiconductor technology, and specifically to a method for optimizing the parameters of a planar coreless transformer based on a genetic algorithm. Background Technology

[0002] A planar coreless transformer is a transformer design that flattens out the traditional coil windings and eliminates the iron core. Extensive research has been conducted on planar coils in the field of magnetically coupled or near-field coupled wireless power transmission. However, existing planar coreless transformers still suffer from the following problems:

[0003] 1. Electromagnetic interference problem: The magnetic field generated by a coreless planar coil usually has a large distribution range, especially at the edges, which can easily interfere with other nearby electronic systems, resulting in low efficiency and high crosstalk.

[0004] 2. Low transmission efficiency and low output power: Some planar coreless transformers, such as PCB high-frequency coreless transformers, have extremely high operating frequencies, usually reaching several MHz to tens of MHz. However, due to electromagnetic interference and the influence of matching circuit design, the transformer has low transmission efficiency and low output power in the secondary circuit, which limits its application in high-power scenarios.

[0005] 3. High design and manufacturing costs: The design process of planar coreless transformers is very complex, requiring consideration of various factors such as magnetic flux distribution, coupling efficiency, and parasitic parameter control, resulting in high design costs. Simultaneously, the manufacturing process requires high-precision technology and equipment, further increasing manufacturing costs. Summary of the Invention

[0006] The purpose of this invention is to provide a parameter optimization method for planar coreless transformers based on genetic algorithms, which aims to solve the problems of low transmission efficiency, cumbersome circuit design process, and long time consumption of existing planar coreless transformers.

[0007] To achieve the above objectives, this invention provides a method for optimizing the parameters of a planar coreless transformer based on a genetic algorithm, comprising the following steps:

[0008] Step 1: Use Maxwell to build a transformer coil model, complete the eddy current simulation of the transformer coil model, build the circuit in Simplier, and perform field-circuit co-simulation with the transformer coil model.

[0009] Step 2: Using the distance H between the two coils, the signal frequency ƒ0, and the secondary load R... L Operating frequency ƒ forms parameter combinations (ƒ0, H, R) L , ƒ) and collect the corresponding transmission efficiency Data, change H and R respectively L Values ​​are determined, parameter co-simulation is performed, and simulation data is collected to form a dataset.

[0010] Step 3: training and testing of the fitness evaluation model using the data set, and designing a fitness function using the generated fitness evaluation model;

[0011] Step 4: optimization based on the combination of the fitness function and the GA optimization algorithm, and output of the final optimization result;

[0012] Step 5: design or adjustment of actual circuit parameters according to the final optimization result.

[0013] Optionally, in step 3, the fitness evaluation model is trained and tested by random forest regression, and the optimal model is saved.

[0014] Optionally, in the fitness function, the transmission efficiency is taken as the fitness target, and the trained model is used to complete the prediction of the transmission efficiency of the current individual, and when the working frequency f is close to the signal frequency f0, i.e. , the obtained transmission efficiency is the maximum.

[0015] Optionally, in step 4, the GA optimization algorithm is set to independently run multiple starting points, the initial population of each run is changed, and the highest transmission efficiency obtained after each run is compared with the current highest transmission efficiency to determine whether to update the optimal individual information.

[0016] The present application provides a genetic algorithm-based planar coreless transformer parameter optimization method, which simulates parameter optimization by combining simulation data with genetic algorithms, and designs and adjusts actual circuit parameters according to the optimal result. The specific process is as follows: first, a coil model is built in Maxwell, and field-circuit combined simulation of the coil model and external circuit is performed in Simplorer. After several comparative simulations, the energy transmission efficiency under different parameter combinations is analyzed. Further, according to the principle of genetic algorithm, the fitness of the combined individual is calculated in MATLAB. After selection, crossover and mutation operations, the optimal combination is obtained to obtain the best transmission efficiency, and the data analysis and optimization are completed. Finally, the circuit parameters are designed or adjusted based on the optimization result, solving the problems of low energy transmission efficiency of existing planar coreless transformers, and complex and time-consuming process of designing transformer and circuit parameters. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0018] Figure 1 is a specific execution flow diagram of a plane coreless transformer parameter optimization method based on a genetic algorithm.

[0019] Figure 2 is a schematic diagram of the change of transformer coupling coefficient k caused by different H in the specific embodiment of the present application.

[0020] Figure 3 is a schematic diagram of the change of transformer transmission efficiency caused by different R L in the specific embodiment of the present application.

[0021] Figure 4 is a schematic diagram of the change of R L and ƒ with H = 0.1088 mm in the specific embodiment of the present application. DETAILED DESCRIPTION

[0022] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the drawings, wherein the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation of the present application.

[0023] The present application provides a plane coreless transformer parameter optimization method based on a genetic algorithm, comprising the following steps:

[0024] Step 1: use Maxwell to build a transformer coil model, complete eddy current simulation of the transformer coil model, build a circuit in Simplorer, and perform field-circuit joint simulation combined with the transformer coil model;

[0025] Step 2: form a parameter combination (f0, H, R L , ƒ) with the two-coil spacing H, signal frequency f0, secondary load R L , working frequency ƒ, and collect corresponding transmission efficiency data, change the values of H and R L respectively, perform parameter joint simulation, collect simulation data to form a data set;

[0026] ​​Step 3: Use the dataset to train and test the fitness evaluation model, and design a fitness function using the generated fitness evaluation model;

[0027] Step 4: Optimize based on a combination of fitness function and GA optimization algorithm, and output the final optimization result;

[0028] Step 5: Design or adjust the actual circuit parameters based on the final optimization results.

[0029] Please see Figures 1 to 4 The following description, in conjunction with specific embodiments and execution steps, provides further details:

[0030] The specific execution process is as follows: Figure 1 As shown.

[0031] In step 1, eddy current simulation of the transformer coil model is performed in Maxwell to analyze the coil parameters that affect the transformer's transmission efficiency. The analysis shows that changes in parameter H will affect the transformer's transmission efficiency. To have an impact, such as Figure 2 As shown. Then, in Simplier, perform a joint field-circuit simulation of the transformer coil model and the external circuit to obtain different H conditions and different R values. L Under different circuit parameter combinations (ƒ0, H, R) conditions L , ƒ) correspond to different transmission efficiencies of transformers ,like Figure 3 , Figure 4 As shown. Parametric simulations are performed on these parameters to obtain the values ​​containing (ƒ0, H, R). L , ƒ) and transmission efficiency The information is data, which constitutes a dataset.

[0032] Using the data, a 5-fold cross-validation method was employed to divide the training and testing sets. Random forest regression was then used to train and test the fitness evaluation model. The trained fitness evaluation model was then used, combined with practical observations: when ƒ approaches ƒ0, the transmission efficiency is... Only the largest, increase The hard constraints are defined to complete the definition of the fitness function. Appropriate GA function parameters are designed according to actual needs, completing the design of the fitness function and GA optimization algorithm. The fitness function can predict any combination (ƒ0, H, R). L Transmission efficiency of , ƒ) The transmission efficiency is obtained by the GA optimization algorithm. The largest composite individual. A holistic optimization algorithm is formed by combining the fitness function with the GA optimization algorithm, and multi-startup GA runs are added to avoid premature convergence. The final optimization algorithm outputs the most reliable result. The largest combination of individuals (ƒ0, H, R) L Based on the optimization results, actual circuit parameters can be designed and adjusted to complete the design of a high-efficiency planar coreless transformer circuit.

[0033] Furthermore, the present invention is further illustrated through specific experiments:

[0034] The actual sample was designed with a planar coil according to the parameters in Table 1, and a relevant model was built in electromagnetic simulation software to obtain different combinations and their corresponding transmission efficiency data. The data was optimized using the algorithm designed in this invention, and the coil's transmitted and received signals before and after circuit matching were compared based on the algorithm's results. Throughout the process, the circuit load remained constant, and the signal transmission efficiency was approximately represented by the ratio α between the peak-to-peak value of the secondary received signal and the peak-to-peak value of the primary transmitted signal.

[0035] Table 1

[0036]

[0037]

[0038] Calculate the first α of the match max The value is 0.75, and the α value after matching is... smax The efficiency is 1.62, a significant improvement of approximately 2.16 times. Based on the optimized algorithm analysis results of this invention, a matching circuit and a filtering circuit were designed. The voltage values ​​of the secondary winding output after rectification and filtering before and after matching were measured, and the measurement results are shown in Table 2. When a DC voltage is input to the primary circuit of the transformer, after circuit resonance, the voltage is transmitted to the secondary winding via magnetic coupling through the coil. When the parameters of the secondary rectification and filtering circuit are adjusted to achieve maximum transmission efficiency, a ±15V control voltage can be output through the push-pull circuit via the boost circuit.

[0039] Table 2

[0040]

[0041] V in the table inmax V is the effective value of the peak-to-peak input voltage of the primary coil. outmax This represents the effective value of the peak-to-peak output voltage at the secondary load. This relates to the transformer's energy transfer efficiency. From the sample circuit testing process and measurement results, it's clear that for transformers with the same coil design parameters, the circuit can achieve its maximum transfer efficiency through frequency matching. However, differences in operating frequency and circuit load will cause variations in the maximum transfer efficiency. The measurement results show that when the load current is 13.8mA, the transformer's transfer efficiency is 54.8% higher than that of the unmatched circuit. The highest transmission efficiency is 71.7%, which is much higher than the maximum transmission efficiency of the frequency matching circuit. In other words, the maximum transmission efficiency of the frequency matching circuit under the current load condition can be achieved However, the circuit load size also affects the transmission efficiency, and when the load is not properly selected, The transmission efficiency is not equal to the maximum transmission efficiency of the transformer.

[0042] In summary, the present application has the following beneficial effects:

[0043] From the simulation verification, the parameter combination given by the optimization algorithm predicts the transmission efficiency of 86.3%, and the transformer transmission efficiency obtained by applying the parameters for simulation is 88.6%, which is not much different, and the method of the present application is feasible; from the actual sample circuit test, when the sample circuit, the current flowing through the load is 13.8mA, the transformer transmission efficiency is improved by 54.8% compared with the transmission efficiency of the unmatched circuit, and the transmission efficiency The highest transmission efficiency is 71.7%, which is much higher than the maximum transmission efficiency of the frequency matching circuit, and also verifies the feasibility of the method of the present application.

[0044] The above only discloses a preferred embodiment of the present application, and of course cannot limit the scope of the present application, and those skilled in the art can understand that the above-mentioned embodiment can be implemented in whole or in part, and the equivalent changes made according to the claims of the present application still belong to the scope covered by the present application.

Claims

1. A method for optimizing the parameters of a planar coreless transformer based on a genetic algorithm, characterized in that, Includes the following steps: Step 1: Use Maxwell to build a transformer coil model, complete the eddy current simulation of the transformer coil model, build the circuit in Simplier, and perform field-circuit co-simulation with the transformer coil model. Step 2: Using the distance H between the two coils, the signal frequency ƒ0, and the secondary load R... L Operating frequency ƒ forms parameter combinations (ƒ0, H, R) L , ƒ) and collect the corresponding transmission efficiency Data, change H and R respectively L Values ​​are determined, parameter co-simulation is performed, and simulation data is collected to form a dataset. Step 3: Use the dataset to train and test the fitness evaluation model, and design a fitness function using the generated fitness evaluation model; Step 4: Optimize based on a combination of fitness function and GA optimization algorithm, and output the final optimization result; Step 5: Design or adjust the actual circuit parameters based on the final optimization results.

2. The method for optimizing the parameters of a planar coreless transformer based on a genetic algorithm as described in claim 1, characterized in that, In step 3, the model is trained and tested using random forest regression, and the fitness evaluation model is saved.

3. The method for optimizing the parameters of a planar coreless transformer based on a genetic algorithm as described in claim 2, characterized in that, The fitness function includes transmission efficiency. As a fitness objective, the already trained model is used to improve the transmission efficiency of the current individual. The prediction is that when the operating frequency ƒ approaches the signal frequency ƒ0, i.e. At that time, the obtained transmission efficiency Only then is it the biggest.

4. The method for optimizing the parameters of a planar coreless transformer based on a genetic algorithm as described in claim 3, characterized in that, In step 4, the GA optimization algorithm is set up with multiple starting points and GAs running independently. The initial population is changed for each run, and the highest transmission efficiency obtained after each run is recorded. With the highest current transmission efficiency The comparison determines whether to update the optimal individual information.

Images