Hybrid excitation permanent magnet motor optimization method for improving Taguchi method and electronic equipment

Abstract

The invention relates to the technical field of motor optimization design, and particularly discloses a hybrid excitation permanent magnet motor optimization method for improving a Taguchi method and electronic equipment, the hybrid excitation permanent magnet motor optimization method for improving the Taguchi method comprises the following steps: defining an optimization model, determining design parameters, an optimization range and an optimization target of a motor, and establishing a parameterized model; sensitivity analysis and parameter layering, wherein the design parameters are divided into sensitive parameters and non-sensitive parameters through sensitivity analysis; and performing sequential optimization on the non-sensitive parameters, performing staged optimization on the non-sensitive parameters according to the sensitivity priority, and reducing the design space. According to the hybrid excitation permanent magnet motor optimization method for improving the Taguchi method, layered optimization, noise factor quantification and fuzzy multi-target fusion are combined for the first time, the limitation that a traditional Taguchi method only depends on a static experiment is broken through, an efficient and robust solution is provided for complex motor design, and the method is suitable for large-scale popularization and application. And the method has remarkable technical progress and industrial popularization potential.

Description

Technical Field

[0001] The present invention relates to the technical field of motor optimal design, and particularly relates to an optimization method for a hybrid excitation permanent magnet motor and an electronic device that improves the Taguchi method. Background Art

[0002] Traditional motor optimization methods (such as genetic algorithms and particle swarm optimization) require a large number of finite element simulations, resulting in high computational costs. Moreover, axial-radial hybrid excitation motors have many parameters (such as permanent magnet size, air gap length, etc.), which leads to a sharp increase in the complexity of high-dimensional optimization problems. Additionally, traditional methods are difficult to balance multiple conflicting objectives (such as increasing torque often accompanied by an increase in torque ripple), which can cause the design results to favor a single performance. Summary of the Invention

[0003] The present invention aims to solve at least one of the technical problems in the related art to some extent. To this end, an object of the present invention is to propose an optimization method for a hybrid excitation permanent magnet motor and an electronic device that improves the Taguchi method, so as to improve the motor optimization efficiency and the robustness of the design results.

[0004] To achieve the above object, a first aspect embodiment of the present invention proposes an optimization method for a hybrid excitation permanent magnet motor that improves the Taguchi method, and the method includes the following steps:

[0005] Optimization model definition: Determine the design parameters, optimization range, and optimization objectives of the motor, and establish a parametric model;

[0006] Sensitivity analysis and parameter stratification: Divide the design parameters into sensitive parameters and non-sensitive parameters through sensitivity analysis;

[0007] Sequential optimization of non-sensitive parameters: Optimize the non-sensitive parameters in stages according to the sensitivity priority to narrow the design space;

[0008] Robust optimization of sensitive parameters: Combine orthogonal experimental design, noise factor analysis, and signal-to-noise ratio calculation to iteratively optimize the sensitive parameters until the convergence condition is met;

[0009] Output the final optimized parameter combination to complete the motor design.

[0010] In some embodiments of the present invention, the optimization model definition specifically includes the following steps:

[0011] Select the design parameters of the motor, including permanent magnet size, stator structure, rotor structure, and air gap parameters;

[0012] Determine the optimization range of each design parameter according to physical size limitations, performance constraint conditions, and manufacturing error tolerances;

[0013] Define multiple optimization objectives, including torque, torque ripple, core loss, and permanent magnet volume, and comprehensively quantify them through an objective function.

[0014] In some embodiments of the present invention, the sensitivity analysis calculates the design parameter x through the following formula i For the optimization objective f i (X) sensitivity:

[0015]

[0016] where E(f i (X) / x i ) is the average value of f i when the design parameter x i (X) is a constant, V(E(f i (X) / x i )) is the variance of E(f i (X) / x i ), V(f i (X)) is the total variance of f i (X);

[0017] Through the design parameter x i For the optimization objective f i (X) sensitivity The sensitivity of the parameter x i can be calculated, and the formula is as follows:

[0018]

[0019] where w i is the weighting factor.

[0020] In some embodiments of the present invention, the sequential optimization of the insensitive parameters includes the following steps:

[0021] Sort the insensitive parameters from low to high according to sensitivity;

[0022] Perform single-objective or multi-objective optimization on each insensitive parameter in turn, and fix its value after optimization;

[0023] Iteratively optimize until all insensitive parameters are optimized, and verify whether the overall performance meets the preset conditions.

[0024] In some embodiments of the present invention, the robust optimization of the sensitive parameters includes the following steps:

[0025] Determine the noise factor according to the manufacturing error, and calculate the parameter instability to screen the key noise factors;

[0026] Construct an orthogonal experiment table, conduct multi-level combination design on sensitive parameters, and introduce the influence of noise factors at each level;

[0027] Through signal-to-noise ratio analysis and fuzzy inference mechanism, calculate the characteristic index CI to evaluate the optimization effect of parameter level combinations;

[0028] Dynamically adjust the parameter range according to the convergence condition, and repeat the iteration until the convergence condition is met.

[0029] In some embodiments of the present invention, the signal-to-noise ratio is calculated by the following formula:

[0030]

[0031] where n is the number of noise experiments for each experiment in the orthogonal experiment table.

[0032] In some embodiments of the present invention, the characteristic index CI is determined by the following formula:

[0033]

[0034] where μ(x) is the membership degree of the input value x, ranging from [0, 1], x is the input quantity, c is the central value of the Gaussian function, taking the expectation of the experimental data of this optimization objective, and σ is the variance of the experimental data of this optimization objective;

[0035] Adopt a fuzzy inference mechanism to map the Gaussian membership function values to the characteristic index CI, and design the parameter x i The CI at different levels can be calculated by the following formula:

[0036]

[0037] where level j represents different levels in the orthogonal experiment table, k represents the kth level in the orthogonal design experiment as level j of the experiment, CI k =(x i =level j ) represents the CI value of the sample of the kth level in the orthogonal design experiment as level j , m represents the total number of experiments at the level of level j in the orthogonal design experiment.

[0038] In some embodiments of the present invention, the convergence condition is:

[0039] max{ΔS / N(f i )}<0.01

[0040] where ΔS / N(f i)Indicates the design parameter f i The signal-to-noise ratio range during the current iteration process.

[0041] In some embodiments of the present invention, the optimization objective is comprehensively optimized by fuzzy logic inference, specifically including the following steps:

[0042] Normalize each optimization objective;

[0043] Use weighted summation or a fuzzy rule base to generate a single comprehensive optimization objective function.

[0044] To achieve the above object, an embodiment of the second aspect of the present invention proposes an electronic device, including a memory, a processor, and a computer program stored on the memory. When the computer program is executed by the processor, the above-mentioned optimization method of the hybrid excitation permanent magnet motor using the improved Taguchi method is implemented.

[0045] The optimization method of the hybrid excitation permanent magnet motor using the improved Taguchi method and the electronic device according to the embodiments of the present invention combine hierarchical optimization, noise factor quantization, and fuzzy multi-objective fusion for the first time, breaking through the limitation of the traditional Taguchi method that only relies on static experiments, providing an efficient and robust solution for complex motor design, and having significant technological progress and industrial promotion potential. Description of the Drawings

[0046] Figure 1 is a schematic flowchart of the optimization method of the hybrid excitation permanent magnet motor using the improved Taguchi method according to an embodiment of the present invention;

[0047] Figure 2 is a schematic diagram of the parametric model of the axial-radial hybrid excitation permanent magnet synchronous motor according to an embodiment of the present invention;

[0048] Figure 3 is a schematic diagram of the sensitivity analysis result of the design parameters according to an embodiment of the present invention;

[0049] Figure 4 is a schematic diagram of the calculation result of the instability of the design parameters according to an embodiment of the present invention;

[0050] Figure 5 is a schematic diagram of the corresponding relationship between the Gaussian membership function and the characteristic index CI during the first iteration process of optimizing the sensitive parameters according to an embodiment of the present invention;

[0051] Figure 6 is the characteristic index CI of the design parameter x i at different levels during each iteration process according to an embodiment of the present invention;

[0052] Figure 7 is a schematic structural diagram of an electronic device according to another embodiment of the present invention. Detailed Embodiments

[0053] Embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where like or similar reference numerals denote like or similar elements or elements having like or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present invention and should not be construed as limiting the present invention.

[0054] The following describes an optimization method and an electronic device for a hybrid excitation permanent magnet motor using an improved Taguchi method according to an embodiment of the present invention with reference to the accompanying drawings.

[0055] Figure 1 It is a schematic flowchart of an optimization method for a hybrid excitation permanent magnet motor using an improved Taguchi method according to an embodiment of the present invention.

[0056] As Figure 1 shown, the optimization method for a hybrid excitation permanent magnet motor using an improved Taguchi method includes the following steps:

[0057] S1. Definition of the optimization model: Determine the design parameters, optimization range, and optimization objectives of the motor, and establish a parametric model.

[0058] The definition of the optimization model is the basic step of the optimization design method for the axial-radial hybrid excitation permanent magnet synchronous motor proposed by the present invention. As an example, the definition of the optimization model specifically includes the following steps:

[0059] Select the design parameters of the motor, including the size of the permanent magnet, the stator structure, the rotor structure, and the air gap parameters;

[0060] Determine the optimization range of each design parameter according to the physical size limit, performance constraint conditions, and manufacturing error tolerance;

[0061] Define multiple optimization objectives, including torque, torque ripple, core loss, and permanent magnet volume, and comprehensively quantify them through the objective function.

[0062] S2. Sensitivity analysis and parameter stratification: Divide the design parameters into sensitive parameters and non-sensitive parameters through sensitivity analysis.

[0063] The main purposes of performing sensitivity analysis and parameter stratification are to improve the optimization efficiency, reduce the computational complexity, and enhance the robustness and accuracy of the optimization results. In multi-objective optimization design, the number of design parameters of the motor is usually large, such as the size of the permanent magnet, the air gap length, the stator slot depth, etc. The influence degrees of these parameters on the motor performance are different, which directly have an important impact on the accuracy of the optimization results and the computational cost.

[0064] Sensitivity analysis is used to evaluate the influence degree of each design parameter on the target performance (such as torque, loss, and torque ripple), identify the sensitive parameters that have a significant impact on the motor performance, and the non-sensitive parameters that have a relatively small impact on the motor performance. The change of sensitive parameters has a greater response to the objective function, so key attention is needed, and fine optimization methods are adopted to ensure the accuracy of the optimization results.

[0065] S3. Sequential optimization of non-sensitive parameters: The non-sensitive parameters are optimized in stages according to the sensitivity priority to narrow the design space.

[0066] As an example, the sequential optimization of non-sensitive parameters includes the following steps:

[0067] Sort the non-sensitive parameters from low to high sensitivity;

[0068] Perform single-objective or multi-objective optimization on each non-sensitive parameter in turn, and fix its value after the optimization is completed;

[0069] Iteratively optimize until all non-sensitive parameters are optimized, and verify whether the overall performance meets the preset conditions.

[0070] S4. Robust optimization of sensitive parameters: Combining orthogonal experimental design, noise factor analysis, and signal-to-noise ratio calculation, iteratively optimize the sensitive parameters until the convergence condition is met.

[0071] As an example, the robust optimization of sensitive parameters includes the following steps:

[0072] Determine the noise factors according to the manufacturing errors, and calculate the parameter instability to screen the key noise factors;

[0073] Construct an orthogonal experimental table, conduct a multi-level combination design for the sensitive parameters, and introduce the influence of noise factors at each level;

[0074] Through signal-to-noise ratio analysis and fuzzy inference mechanism, calculate the characteristic index CI to evaluate the optimization effect of the parameter level combination;

[0075] Dynamically adjust the parameter range according to the convergence condition, and repeat the iteration until the convergence condition is met.

[0076] S5. Output the final optimized parameter combination to complete the motor design.

[0077] In the present invention, the optimization method of the hybrid excitation permanent magnet motor based on the improved Taguchi method significantly improves the efficiency of the motor optimization design and the robustness of the results through sensitivity analysis and hierarchical optimization strategies, and solves the deficiencies of traditional optimization methods in terms of high computational cost, sensitivity to manufacturing errors, and performance optimization conflicts; moreover, this method realizes multi-objective motor performance optimization and manufacturing error compensation through parameter sensitivity analysis, staged optimization, and multi-objective collaborative strategies, and is applicable to the design scenarios of high-performance motors with complex structures and obvious optimization objective conflicts.

[0078] In some embodiments of the present invention, the definition of the optimization model includes the following four key links:

[0079] S11. Determination of motor design parameters. In the motor optimization design, the selection of design parameters determines the efficiency of the motor optimization and the effectiveness of the optimization process. The design parameters of the motor cover multiple parts such as permanent magnets, stators, rotors, and air gaps. The selection of design parameters is generally determined according to actual needs and experience, and is represented by the matrix X.

[0080] S12. Determination of the optimization range of design parameters. Considering the design data and the manufacturing process of the actual production of the motor, the optimization range of the design parameters should cover all key operating conditions of the motor and manufacturing errors. The setting of the parameter range follows the following rules:

[0081] Physical size limitation: The sizes of permanent magnets and air gap lengths need to conform to the actual manufacturing process capabilities;

[0082] Performance constraint conditions: Motor performance indicators such as torque, loss, and torque ripple need to be within the technical specification range;

[0083] Manufacturing error tolerance: Considering the actual error range in the manufacturing process, ensure the applicability of the optimization results in actual production;

[0084] The upper and lower boundaries of the design parameters are usually represented by the matrices X up and X low respectively.

[0085] S13. Selection of optimization objectives. In the motor optimization process, the selection of optimization objectives needs to comprehensively consider various performance requirements of the motor, usually including but not limited to the torque, torque ripple, loss, production cost, etc. of the motor. The motor optimization objective can be expressed as;

[0086] min:{f i (X)}, i = 1, 2,..., n (1)

[0087] where f i (X) is the i-th optimization objective of the motor.

[0088] S14. Establishment of the parameter model. Construct a parametric model of the motor based on finite element software. This parametric model will be used to obtain the performance indicators of the motor through simulation under given design parameters and working conditions during the optimization process, providing a reference for the optimal design of the motor.

[0089] In some embodiments of the present invention, the sensitivity analysis calculates the design parameter x through the following formula i for the optimization objective f i (X) sensitivity:

[0090]

[0091] where, E(f i (X) / x i ) is the average value of f i (X) when the design parameter x i is a constant, V(E(f i (X) / x i )) is the variance of E(f i (X) / x i ), V(f i (X)) is the total variance of f i (X);

[0092] The sensitivity of the optimization objective f i (X) with respect to the design parameter x i can be calculated, and the formula is as follows: The sensitivity of the parameter x i can be obtained, and the formula is as follows:

[0093]

[0094] where, w i is the weighting factor, and the size of the weighting factor is independently selected by the designer according to the degree of attention of different optimization objectives in this optimization process.

[0095] Generally speaking, all design parameters are divided into sensitive parameters and non-sensitive parameters according to the sensitivity analysis results, and a phased optimization strategy is adopted according to the hierarchical results. First, optimize the non-sensitive parameters to reduce the search range of the parameter space, narrow the design variables, and reduce the complexity of the optimization calculation. Subsequently, perform high-precision optimization on the sensitive parameters based on the improved Taguchi method, thereby accelerating the optimization convergence speed and improving the design efficiency.

[0096] In some embodiments of the present invention, the sequential optimization of non-sensitive parameters is a phased optimization strategy. It aims to gradually narrow the design space, reduce the amount of optimization calculations, and lay a foundation for the precise optimization of key sensitive parameters by preferentially optimizing the parameters that have less impact on the motor performance. The core objective of the sequential optimization of non-sensitive parameters is to reduce the computational complexity and improve the optimization convergence speed. This method follows the following principles:

[0097] ① Sensitivity priority sorting principle: According to the sensitivity analysis results, sort the non-sensitive parameters in ascending order of their impact on the motor performance, and optimize them one by one to ensure that the optimization process starts from the parameter with the least impact and gradually narrows the design parameter space;

[0098] ② Step-by-step fixation strategy: Optimize one non-sensitive parameter each time. After the optimization is completed, fix its value to reduce the number of design variables in subsequent optimization steps and lower the computational complexity.

[0099] The sequential optimization method of non-sensitive parameters can be further refined into the following steps:

[0100] S31. Parameter sorting and initialization. According to the sensitivity analysis results, sort the non-sensitive parameters by priority and start optimizing from the parameter with the least impact on the target performance. In the initialization stage, set the initial values of all design parameters, usually the middle value of their design ranges or the recommended values of the manufacturing process;

[0101] S32. Single-parameter optimization. Start from the non-sensitive parameter with the lowest sensitivity, keep other parameters fixed, and perform single-objective or multi-objective optimization on the current non-sensitive parameter. When optimizing, use finite element analysis (FEA) or a design space simulation model to evaluate the impact of the optimized parameter on the performance target. Construct an objective function for the optimization target:

[0102]

[0103] Determine the optimal value of the current parameter according to the optimization objective function and fix this parameter;

[0104] S33. Optimize the remaining non-sensitive parameters in sequence. Repeat the process of step S32 for the next non-sensitive parameter in the sorting. After each round of optimization, fix the optimized parameter value to narrow the design space and reduce the search range in subsequent optimization processes. Loop until all non-sensitive parameters are optimized;

[0105] S34. Result evaluation and convergence judgment. After all non-sensitive parameters are optimized, comprehensively evaluate the performance of the entire motor, including key performance indicators such as average torque, torque ripple, and core loss. If the performance indicators do not meet the preset requirements, adjust the optimization range of non-sensitive parameters or re-conduct sensitivity analysis. If the performance indicators reach the expected goals, enter the fine optimization stage of sensitive parameters.

[0106] The robust optimization method for sensitive parameters aims at design parameters that have a significant impact on motor performance. Considering factors such as manufacturing errors, changes in operating environment, and conflicts in performance goals, it optimizes the optimal combination of design parameters to ensure the overall performance of the motor remains stable under different operating conditions. This method is based on the theory of robust optimization. Through orthogonal experimental design, signal-to-noise ratio analysis, and multi-objective decision-making algorithms, it improves the optimal solution of performance goals while minimizing performance fluctuations.

[0107] In some embodiments of the present invention, during the optimization process of sensitive parameters, the errors in the motor manufacturing process are taken into account. List the possible magnitudes of sensitive parameter noises according to the actual processing conditions. To find the noise factor that has the greatest impact on the motor, the following instability calculation criterion is used to determine the influence of the noise factor on motor performance:

[0108]

[0109] where Δx i is the manufacturing tolerance of x i and x i_initial is the initial value of x i . By comparing the magnitudes of different instabilities, determine the design factor that has the greatest impact on motor performance. Then, design an orthogonal experimental table that includes all sensitive design parameters. Since the design factor that has the greatest impact on motor performance needs to consider the error during manufacturing, the parameter x k needs to consider its results after being affected by noise at each level of the orthogonal experimental table, that is, it has two values at each level:

[0110]

[0111] This means that the current orthogonal experimental design table needs to be experimented twice under different noise effects.

[0112] It should be noted that the numerical values of the design parameters corresponding to the same level in different iteration processes are not the same. After each iteration process ends, the numerical values corresponding to the design parameter levels will be updated according to the optimization results, and the updated numerical values will be input into the parameter optimization of the next iteration process.

[0113] In some embodiments of the present invention, the signal-to-noise ratio is used instead of the average value of the performance during the optimization process, making the optimization result have better stability. The signal-to-noise ratios of different optimization objectives are calculated by the following formula:

[0114]

[0115] where n is the number of noise experiments for each experiment in the orthogonal experiment table, and the value of n is 2 in this embodiment. For the optimization objective that needs to be maximized, its reciprocal can be taken first and then optimized. The Gaussian membership function value is calculated through the value of the signal-to-noise ratio.

[0116] In some embodiments of the present invention, the characteristic index CI is determined by the following formula. First, the Gaussian membership function is obtained by the following formula:

[0117]

[0118] where μ(x) is the membership degree of the input value x, with a range of [0, 1], x is the input quantity, c is the central value of the Gaussian function, taking the expectation of the experimental data of this optimization objective, and σ is the variance of the experimental data of this optimization objective, which determines the width of the Gaussian membership function. The larger the value, the "flatter" the function, and the smaller the value, the "steeper" the function.

[0119] For example, taking the input value x corresponding to the Gaussian membership function μ(x) = 0.5 as the boundary, the Gaussian membership function is divided into three levels. The fuzzy inference mechanism is used to map the Gaussian membership function values to the characteristic index CI, and the design parameter x i The CI at different levels can be calculated by the following formula:

[0120]

[0121] where level j represents different levels in the orthogonal experiment table, k represents the kth level as level j in the orthogonal design experiment, CI k =(x i =level j ) represents the CI value of the sample with the kth level as level j in the orthogonal design experiment, and m represents the total number of experiments with the level of level j in the orthogonal design experiment.

[0122] It should be noted that different CI values represent the quality of the design parameter xi at the current level. The higher the CI value, the higher the motor performance at the current level. By combining the design parameters at different levels, the optimal design of the axial-radial hybrid-excitation permanent magnet synchronous motor can be obtained.

[0123] In some embodiments of the present invention, after the optimization process of the current iteration is completed, it is necessary to determine whether the convergence condition is satisfied. The convergence condition is:

[0124] max{ΔS / N(f i )}<0.01 (10)

[0125] where, ΔS / N(f i ) represents the signal-to-noise ratio range of the design parameter f i in the current iteration process. If the convergence condition is satisfied, the optimization process ends. The combination of the design parameters generated in the current iteration process at different levels is output as the optimal design of the final axial-radial hybrid-excitation permanent magnet synchronous motor. If the convergence condition is not satisfied, it is necessary to further narrow the optimization range and generate new levels, and the new optimization range and the levels of the design parameters will be input into the next iteration process, and the iteration process continues until the convergence condition is satisfied.

[0126] In some embodiments of the present invention, the optimization objectives are comprehensively integrated through fuzzy logic reasoning, specifically including the following steps:

[0127] Normalize each optimization objective;

[0128] Use weighted summation or a fuzzy rule base to generate a single comprehensive optimization objective function.

[0129] Here, in combination with Figures 2 - 6 , the improved Taguchi method-based hybrid-excitation permanent magnet motor optimization method of the above embodiments of the present invention is further described in detail.

[0130] Embodiment 1:

[0131] Step 1: Optimization model definition

[0132] As Figure 2 shown, a parametric model of the motor is established based on the finite element software Jmag. In the optimization design of the axial-radial hybrid-excitation permanent magnet synchronous motor, the design parameters include the structural dimensions of multiple key components such as permanent magnets, stators, rotors, and air gaps. Considering the physical dimensions, manufacturing processes, and performance constraints of the motor, a reasonable optimization range is set. The selected design parameters and optimization range in this embodiment are shown in Table 1.

[0133]

[0134]

[0135] Table 1 Design parameters and optimization range

[0136] Here, the optimization objectives of the motor include the torque (T), torque ripple (T r ), loss (Pfe ) and the volume of the permanent magnet (V PM ):

[0137]

[0138] Step 2: Sensitivity analysis and parameter stratification

[0139] Design an orthogonal experiment table, and calculate the sensitivity of all design parameters based on the parametric model and the sensitivity calculation formulas (2)-(3), where the weight factors w1, w2, w3, and w4 corresponding to the torque (T), torque ripple (T r ), loss (P fe ), and the volume of the permanent magnet (V PM ) are defined as 0.3, 0.2, 0.3, and 0.2 respectively. Then, the sensitivity analysis results of the design parameters as shown Figure 3 are obtained. According to the results of the sensitivity analysis, the design parameters are divided into: sensitive parameters P3 (PM1 h ), P5 (PM1 ro ), P6 (PM2 α ), P7 (PM2 g ), P8 (PM2 h ), and non-sensitive parameters P1 (PM1 α ), P2 (PM1 g ), P4 (PM1 ri ), P9 (PM2 l ).

[0140] Step 3: Sequential optimization of non-sensitive parameters

[0141] Based on the optimization objectives selected in this case, the optimization function of the non-sensitive parameters is as follows:

[0142]

[0143] Fix other parameters unchanged, and optimize from the design parameter P1 (PM1 α ) with higher sensitivity in turn. The optimal value of P1 (PM1 α ) is obtained as 26 mm. Update the value of P1 (PM1 α ) to 26 mm, fix other design parameters except P2 (PM1 g ) unchanged, and optimize the value of P2 (PM1 g ). The optimal value is obtained as 1 mm. According to this method, the optimal values of P9 (PM2 l ) are obtained as 51 mm, and the optimal value of P4 (PM1 ri ) is 56 mm.

[0144] Step 4: Optimization of sensitive parameters

[0145] Considering the processing standards, the noises of different sensitive design parameters are shown in the table. The calculation results of the instability of the design parameters calculated according to formula (5) are as Figure 4 shown. Select P8 (PM2 h ) as the design parameter most sensitive to noise according to the calculation results.

[0146] Serial number Parameter Unit Noise P6 <![CDATA[PM2 α > ° ±0.02 P5 <![CDATA[PM1 ro > ° ±0.02 P8 <![CDATA[PM2 h > mm ±0.02 P3 <![CDATA[PM1 h > mm ±0.02 P7 <![CDATA[PM2 g > mm ±0.02

[0147] Table 2 Noise of sensitive parameters

[0148] Design an orthogonal experiment table as shown in Table 3. The specific values corresponding to different levels of the design parameters in the first iteration process are shown in Table 4. Since P8 (PM2 h ) needs to be experimented twice under the influence of the noise factor at each level, a total of 25×2 = 50 simulation experiments are required in this iteration process.

[0149]

[0150]

[0151] Table 3 Orthogonal experiment table of L25(5 5 )

[0152] Parameter Level 1 Level 2 Level 3 <![CDATA[PM2 α > 34 30 26 <![CDATA[PM1 ro > 98 102 106 <![CDATA[PM2 h > 3 3.5 4 <![CDATA[PM1 h > 3 3.5 4 <![CDATA[PM2 g > 0.6 1 1.4

[0153] Table 4 Actual values corresponding to different levels of design parameters in the first iteration process

[0154] Calculate the signal-to-noise ratio through formula (7), calculate the Gaussian membership function value through formula (8), and use the fuzzy inference mechanism to map the Gaussian membership function value to the characteristic index CI. The corresponding relationship between the Gaussian membership function and the characteristic index CI in this iteration process is as Figure 5 shown. Calculate the characteristic index CI of the design parameter x i at different levels according to formula (9), and obtain Figure 6 the result, and judge the convergence condition max{ΔS / N(f i )} > 0.01. Therefore, reduce the parameter optimization range according to the optimization result shown in Figure (6) and continue the Taguchi optimization. When the iteration reaches the 4th generation, the convergence condition is satisfied and the iteration process stops, and the result is output. The characteristic index CI of the design parameter xi at different levels in each iteration process is as Figure 6 shown, and the levels of different design parameters in each iteration process are shown in Table 5.

[0155]

[0156]

[0157] Table 5 The levels of different design parameters in each iteration process. The final optimization results of the motor obtained through non-sensitive parameter optimization and sensitive parameter optimization are shown in Table 6.

[0158]

[0159]

[0160] Table 6 Motor optimization results

[0161] Corresponding to the above embodiments, the present invention also provides an electronic device.

[0162] As Figure 7 shown in the structural schematic diagram of an electronic device in the present invention, the electronic device 200 includes: a processor 201 and a memory 203. Among them, the processor 201 and the memory 203 are connected, such as connected through a bus 202. Optionally, the electronic device 200 may further include a transceiver 204. It should be noted that in practical applications, the transceiver 204 is not limited to one, and the structure of the electronic device 200 does not constitute a limitation to the embodiments of the present invention.

[0163] The processor 201 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logic blocks, modules, and circuits described in connection with the disclosure of the present invention. The processor 201 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0164] The bus 202 may include a path for transmitting information between the above components. The bus 202 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus 202 may be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 7 only a thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus.

[0165] The memory 203 is used to store a computer program corresponding to the optimization method of the hybrid excitation permanent magnet motor by the improved Taguchi method of the above embodiments of the present invention, and the computer program is controlled and executed by the processor 201. The processor 201 is used to execute the computer program stored in the memory 203 to implement the content shown in the foregoing method embodiments.

[0166] Among them, the electronic device 200 includes but is not limited to: mobile terminals such as laptop computers, PADs (tablet computers), etc., and fixed terminals such as desktop computers, etc. Figure 7 The illustrated electronic device 200 is merely an example and should not impose any limitations on the functions and usage scope of the embodiments of the present invention.

[0167] The electronic device 200 of the embodiments of the present invention has the following advantages:

[0168] ① High optimization efficiency: The present invention adopts a sensitivity analysis method, classifies the motor design parameters into sensitive parameters and non-sensitive parameters, and adopts a hierarchical optimization strategy, significantly reducing the number of simulations of finite element analysis (FEA), reducing the computational complexity of high-dimensional optimization problems, and improving the optimization efficiency;

[0169] ② Strong multi-objective optimization ability: The present invention combines multi-objective optimization theory, and through fuzzy logic reasoning and signal-to-noise ratio analysis, synthesizes multiple performance indicators such as torque, torque ripple, core loss, and permanent magnet volume into a single optimization target, realizes the collaborative optimization of multiple objectives, and balances the comprehensive performance of the motor;

[0170] ③ Robustness-enhanced design: The present invention introduces the orthogonal experimental design and noise factor analysis in the Taguchi method. Considering manufacturing errors, through signal-to-noise ratio analysis during the optimization process, it ensures the stability and robustness of the design results in actual production, and effectively reduces the impact of manufacturing errors on the motor performance;

[0171] ④ Strong parameter adaptability: The present invention adopts an adaptive parameter adjustment mechanism, automatically adjusts the search range and experimental design scheme according to the sensitivity of the design parameters and the convergence of the optimization results during the optimization process, improves the adaptability and expansion ability of the optimization design method, and is applicable to various complex motor design tasks;

[0172] ⑤ High industrial application value: The optimization design method of the present invention significantly reduces the R & D and production costs of the motor while considering design performance and manufacturing errors, is widely applicable to the optimization design of high-performance motors in electric vehicle drive motors, industrial automation equipment, and new energy power generation equipment, and has good industrial application prospects.

[0173] Note that the logic and / or steps represented in the flowchart or described otherwise herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, electrically optimize the design, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of the computer-readable medium include the following: an electrical connection part with one or more wirings (electronic device), a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which a program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or other suitable processing as necessary, and then stored in a computer memory.

[0174] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0175] In the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0176] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0177] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. An optimization method for a hybrid-excited permanent magnet motor that improves the Taguchi method, characterized in that, The method includes the following steps: Optimize the model definition, determine the design parameters, optimization range, and optimization objectives of the motor, and establish a parametric model; Sensitivity analysis and parameter stratification, dividing the design parameters into sensitive parameters and non-sensitive parameters through sensitivity analysis; Sequential optimization of non-sensitive parameters, performing staged optimization on non-sensitive parameters according to the sensitivity priority to narrow the design space; Robust optimization of sensitive parameters, combining orthogonal experimental design, noise factor analysis, and signal-to-noise ratio calculation to perform iterative optimization on sensitive parameters until the convergence condition is met; Output the final optimized parameter combination to complete the motor design.

2. The optimization method of the hybrid excitation permanent magnet motor using the improved Taguchi method according to claim 1, characterized in that, The specific steps of the optimization model definition include the following: Select the design parameters of the motor, including the permanent magnet size, stator structure, rotor structure, and air gap parameters; Determine the optimization range of each design parameter according to physical size limitations, performance constraint conditions, and manufacturing error tolerances; Define multiple optimization objectives, including torque, torque ripple, core loss, and permanent magnet volume, and comprehensively quantify them through the objective function.

3. The optimization method of the hybrid excitation permanent magnet motor by the improved Taguchi method according to claim 1, characterized in that The sensitivity analysis calculates the design parameter x through the following formula i for the optimization objective f i (X) sensitivity: where E(f i (X) / x i ) is the mean value of f i (X) when the design parameter xi is a constant, V(E(f i (X) / x i )) is the variance of E(f i (X) / x i ), and V(f i (X)) is the total variance of f i (X); Through the design parameter x i For the optimization objective f i (X) sensitivity The sensitivity of the parameter x can be calculated i as follows: where w i is a weighting factor.

4. The optimization method of the hybrid excitation permanent magnet motor by the improved Taguchi method according to claim 1, characterized in that, The sequential optimization of the non-sensitive parameters includes the following steps: Sort the non-sensitive parameters from low to high according to sensitivity; Perform single-objective or multi-objective optimization on each non-sensitive parameter in turn, and fix its value after optimization; Iteratively optimize until all non-sensitive parameters are optimized, and verify whether the overall performance meets the preset conditions.

5. The optimization method of the hybrid excitation permanent magnet motor by the improved Taguchi method according to claim 1, characterized in that The robust optimization of the sensitive parameters includes the following steps: Determine the noise factors according to manufacturing errors, and calculate the parameter instability to screen the key noise factors; Construct an orthogonal experimental table, perform multi-level combination design on sensitive parameters, and introduce the influence of noise factors at each level; Calculate the characteristic index CI through signal-to-noise ratio analysis and fuzzy inference mechanism to evaluate the optimization effect of the parameter level combination; Dynamically adjust the parameter range according to the convergence condition, and repeat the iteration until the convergence condition is met.

6. The optimization method of the hybrid excitation permanent magnet motor using the improved Taguchi method according to claim 5, characterized in that The signal-to-noise ratio is calculated by the following formula: where n is the number of noise experiments for each experiment in the orthogonal experimental table.

7. The optimization method of the hybrid excitation permanent magnet motor using the improved Taguchi method according to claim 5, characterized in that The characteristic index CI is determined by the following formula: where μ(x) is the membership degree of the input value x, ranging from [0,1], x is the input quantity, c is the central value of the Gaussian function, taking the expectation of the experimental data of this optimization objective, and σ is the variance of the experimental data of this optimization objective; The Gaussian membership function values are mapped to the characteristic index CI by using a fuzzy inference mechanism, and the parameter x is designed. i The CI at different levels can be calculated by the following formula: Among them, level j represents different levels in the orthogonal experiment table, and k represents the k-th level in the orthogonal design experiment being level j of the experiment, and CI k =(x i =level j ) represents the CI value of the sample where the k-th level in the orthogonal design experiment is level j , and m represents the total number of experiments at level j in the orthogonal design experiment.

8. The optimization method of the hybrid excitation permanent magnet motor by the improved Taguchi method according to claim 5, characterized in that The convergence condition is: max{ΔS / N(f i )} < 0.01 where, ΔS / N(f i ) represents the signal-to-noise ratio range of the design parameter f i in the current iteration process.

9. The optimization method of the hybrid excitation permanent magnet motor using the improved Taguchi method according to claim 1, characterized in that, The multi-objective synthesis of the optimization objectives is performed through fuzzy logic inference, and the specific steps include the following: Normalize each optimization objective; Adopt weighted summation or fuzzy rule base to generate a single comprehensive optimization objective function.

10. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored on the memory. When the computer program is executed by the processor, it implements the hybrid-excitation permanent magnet motor optimization method of the improved Taguchi method as described in any one of claims 1-9.

Images