Motor multi-objective optimization method and system based on improved dung beetle optimization algorithm

Abstract

The present invention is applicable to the technical field of permanent magnet synchronous motor performance improvement, and provides a motor multi-objective optimization method and system based on an improved dung beetle optimization algorithm. The method comprises: establishing a mathematical model of motor torque pulsation and dynamic response capability, and determining initial design parameters affecting the torque pulsation and dynamic response capability; determining an optimal angle satisfying a first constraint condition based on simulation data of the motor's stator skew slots, so as to establish a motor model; analyzing motor structural parameters affecting the sensitivity of the initial design parameters based on the motor model, and obtaining the motor structural parameters as the stator outer diameter, permanent magnet thickness, tooth width, and air gap length; in the present invention, research is conducted from the perspective of the stator skew slots to obtain an optimal angle at which torque pulsation is greatly reduced while dynamic response capability is minimized; performing sensitivity analysis on the initial design parameters to obtain structural parameters, and performing algorithm optimization after fitting, thereby achieving multi-objective optimization of the motor.

Description

Technical Field

[0001] The present invention belongs to the technical field of permanent magnet synchronous motor performance improvement, and specifically discloses a motor multi-objective optimization method and system based on an improved dung beetle optimization algorithm. Background Art

[0002] Due to its unique structure, where the rotor is external and the stator is internal, outer rotor permanent magnet synchronous motors (EPMSs) offer advantages such as high output torque and power density, making them widely used in robotic joints. However, due to the presence of cogging torque and the large rotor inertia, the motors exhibit large torque ripple and poor dynamic response. High torque ripple increases electromagnetic noise and system instability, while low dynamic response exacerbates torque ripple and increases mechanical vibration. Therefore, reducing motor torque ripple and improving dynamic response are pressing challenges.

[0003] To reduce torque pulsation, the motor structure is usually improved through methods such as stator slot optimization, stator slot skew, rotor pole skew, and pole slot matching. However, all of the above methods effectively reduce the motor cogging torque, but do not simultaneously consider the impact of structural improvements on dynamic response capabilities. Generally, research on improving the dynamic response capability of motors focuses on rotor structure design, the use of new materials, and other aspects. However, the dynamic response capability of a motor depends not only on the rotor inertia, but also on torque. In order to improve the performance of the motor, a common method is single-objective optimization. However, because the selected optimization objective conflicts with other objectives, it is often overlooked. Therefore, it is necessary to design a multi-objective optimization design method. Summary of the Invention

[0004] The present invention aims to provide a multi-objective motor optimization method and system based on an improved dung beetle optimization algorithm, aiming to address the issues mentioned in the background. Taking into account factors influencing dynamic response capability, the present invention selects an appropriate stator skew angle to reduce torque ripple, improves dynamic response capability by optimizing the mechanical time constant, and optimizes motor parameters using an improved dung beetle optimization algorithm based on non-dominated sorting. This improves the algorithm's convergence speed and optimization accuracy, ultimately obtaining a superior Pareto optimal solution set.

[0005] The present invention is implemented as follows: a motor multi-objective optimization method based on an improved dung beetle optimization algorithm, the method comprising the following steps:

[0006] Establish a mathematical model of motor torque ripple and dynamic response capability, and determine the initial design parameters that affect torque ripple and dynamic response capability;

[0007] Determining an optimal angle that satisfies a first constraint condition based on simulation data of a stator skew slot of a motor to establish a motor model; wherein the first constraint condition is to minimize a decrease in torque ripple and a decrease in dynamic response capability;

[0008] Based on the motor model, analyzing the motor structural parameters that affect the sensitivity of the initial design parameters, and obtaining the motor structural parameters as the stator outer diameter, permanent magnet thickness, tooth width, and air gap length;

[0009] The response surface methodology is used to construct a fitting model for the design goal, which is to reduce the torque ripple of the motor and improve its dynamic response capability.

[0010] Based on the fitting model, the improved dung beetle optimization algorithm is used to optimize the motor structural parameters.

[0011] Another object of the present invention is to provide a motor multi-objective optimization system based on an improved dung beetle optimization algorithm, the system comprising:

[0012] Mathematical model input module, used to establish a mathematical model of motor torque ripple and dynamic response capability, and determine the initial design parameters that affect torque ripple and dynamic response capability;

[0013] a simulation optimization module, configured to determine an optimal angle that satisfies a first constraint condition based on simulation data of the stator skew slots of the motor, so as to establish a motor model; wherein the first constraint condition is to minimize the reduction in torque ripple and the reduction in dynamic response capability;

[0014] a sensitivity analysis module, configured to analyze, based on the motor model, motor structural parameters that affect the sensitivity of the initial design parameters, and obtain the motor structural parameters as stator outer diameter, permanent magnet thickness, tooth width, and air gap length;

[0015] The target solution module is used to construct a fitting model of the design target through the response surface method. The design target is to reduce the torque ripple of the motor and improve the dynamic response capability;

[0016] A parameter optimization module is used to optimize the motor structure parameters based on the fitting model using an improved dung beetle optimization algorithm.

[0017] The present invention provides a multi-objective motor optimization method based on an improved dung beetle optimization algorithm. This method, through simulation research from the perspective of stator skew slots, significantly reduces torque pulsation while minimizing the increase in the mechanical time constant, helping to maintain the motor's good speed response capability. This allows the motor to operate smoothly while quickly responding to speed changes, improving its overall performance. The improved dung beetle optimization algorithm employed in the present invention is a non-dominated sorting dung beetle optimization algorithm, which optimizes the parameters of the fitting model, simultaneously addressing the motor's design goals of low torque pulsation and high dynamic response capability while obtaining a non-dominated solution. The algorithm's excellent global search capability enables it to avoid local optima in high-dimensional space and approach global optimal parameters. Its unique mechanism effectively prevents premature convergence, continuously exploring new solution spaces, and ultimately accurately optimizes the fitting model parameters, achieving optimization of the motor's structural parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flow chart illustrating a multi-objective motor optimization method based on an improved dung beetle optimization algorithm provided by an embodiment of the present invention;

[0019] Figure 2 This is a diagram showing the optimization target sensitivity analysis results in an embodiment of the present invention;

[0020] Figure 3 This is a response surface diagram of the optimization target in an embodiment of the present invention;

[0021] Figure 4 This is a diagram of the non-dominated sorting principle in an embodiment of the present invention;

[0022] Figure 5 1 is an optimization step diagram of the dung beetle optimization algorithm based on non-dominated sorting in an embodiment of the present invention;

[0023] Figure 6 is the optimized Pareto frontier graph in the embodiment of the present invention;

[0024] Figure 7 This is a comparison diagram of torque ripple before and after optimization in an embodiment of the present invention;

[0025] Figure 8 This is a comparison chart of dynamic response capabilities before and after optimization in an embodiment of the present invention;

[0026] Figure 9 A schematic structural diagram of an outer rotor permanent magnet synchronous motor used in an embodiment of the present invention;

[0027] Figure 10 Schematic diagram of a simulation model of a stator skew slot according to an embodiment of the present invention;

[0028] Figure 11A flowchart of a multi-objective optimization method for a motor based on an improved dung beetle optimization algorithm provided by an embodiment of the present invention;

[0029] Figure 12 A structural block diagram of a motor multi-objective optimization system based on an improved dung beetle optimization algorithm provided in an embodiment of the present invention.

[0030] In the accompanying drawings: 101 - stator; 102 - stator skew slot; 103 - rotor; 104 - permanent magnet. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0032] In reality, however, a motor's dynamic response capability depends not only on rotor inertia; torque also plays a key role. The dynamic response capability of an outer-rotor permanent-magnet synchronous motor includes both torque and speed response capabilities, which are related to the electrical and mechanical time constants, respectively. Because the mechanical time constant of an outer-rotor permanent-magnet synchronous motor is much greater than the electrical time constant, this embodiment of the present invention optimizes only the mechanical time constant.

[0033] To improve the performance of an outer rotor permanent magnet synchronous motor, this embodiment employs a multi-objective optimization design method to optimize design objectives. This motor multi-objective optimization method typically employs a linear weighted sum method to transform the multi-objective function into a single objective function. Using a multi-objective optimization algorithm can simplify the optimization process when searching for a Pareto optimal trade-off solution.

[0034] Figure 9 A schematic structural diagram of an outer rotor permanent magnet synchronous motor used in an embodiment of the present invention; Figure 10 This figure is a schematic diagram of a simulation model of the stator slots in an embodiment of the present invention. The outer rotor permanent magnet synchronous motor comprises a stator 101, stator slots 102 formed on the outer circumference of the stator 101 for mounting coils, a rotor 103 sleeved outside the stator 101, and permanent magnets 104 fixed to the inner surface of the outer rotor permanent magnet synchronous motor. The motor parameters are selected for an outer rotor permanent magnet synchronous motor with 20 poles and 24 slots, and a rated speed of 3000 rpm. The motor adopts a surface-mount structure, which helps reduce rotor inertia and improve the motor's dynamic response capability.

[0035] Figure 1 The principle flow chart of a motor multi-objective optimization method based on an improved dung beetle optimization algorithm provided in an embodiment of the present invention specifically includes:

[0036] Determine design variables and optimization objectives;

[0037] Conduct sensitivity analysis;

[0038] Determine whether it is significant, if not, perform single parameter optimization;

[0039] If so, then the experimental points were selected based on Box-Behnken (response surface methodology);

[0040] Calculate motor response based on finite element analysis;

[0041] Response surface methodology was used to construct a fitting model for the design objectives;

[0042] Multi-objective optimization is performed based on the improved dung beetle optimization algorithm and response surface model;

[0043] Obtain the optimal design from the Pareto solution set;

[0044] Build finite element analysis models and test performance.

[0045] Figure 11 In this embodiment, a flowchart of a motor multi-objective optimization method based on an improved dung beetle optimization algorithm is provided; the method includes the following steps S101 to S105;

[0046] Step S101, establishing a mathematical model of motor torque ripple and dynamic response capability, and determining initial design parameters that affect torque ripple and dynamic response capability;

[0047] In the outer rotor permanent magnet synchronous motor, the dq axis inductance is not much different, so the cogging torque T cog It has a great influence on torque ripple. The cogging torque expression is as follows:

[0048]

[0049] Among them, L a is the axial length; Z is the number of slots; μ0 is the vacuum permeability; R3 is the rotor inner diameter; R2 is the stator outer diameter; n is the is an integer; α is the angle between the pole centerline and the tooth centerline; G n is the Fourier decomposition coefficient of the square of the relative air gap permeability; is the Fourier decomposition coefficient of the air gap magnetic flux;

[0050]

[0051] Among them, h m is the thickness of the permanent magnet; g is the air gap length; θ s0 Indicates the arc value corresponding to the stator slot width; α p is the polar arc coefficient; B r is the distribution of the permanent magnet's remanence along the circumferential direction, and 2p is the number of poles.

[0052] In this embodiment, the mathematical model of the motor torque ripple is:

[0053]

[0054] Among them, T max is the output torque T N Maximum value; T min is the output torque T N Minimum value; T av is the average torque value.

[0055] The mathematical model of the motor's dynamic response capability is:

[0056]

[0057] Among them, ρ w is the conductor resistivity; L E is the average half-turn length; N is the number of turns in series; A 01 is the cross-sectional area of ​​the enameled wire; a1 is the number of parallel branches; B av is the average magnetic flux density; D a is the diameter of the magnetic steel surface; l is the length of the conductor in the magnetic field; ρ m is the permanent magnet density; R1 is the rotor outer diameter; ρ Fe is the rotor core density; h m is the thickness of the permanent magnet; g is the air gap length; R2 is the stator outer diameter.

[0058] From the above analysis of formulas (2) to (5), we can know that the stator outer diameter R2 and the permanent magnet thickness h can be selected m , slot width B s0 、Tooth width B s1 , air gap length g, pole arc coefficient α p , groove depth H s0 , slot height H s1 Table 1 shows the reasonable value range of the initial design parameters, which is used for the subsequent sensitivity analysis.

[0059] Table 1 Initial design parameter values

[0060]

[0061] Step S102: determining an optimal angle that satisfies a first constraint condition based on simulation data of the stator skew slots of the motor to establish a motor model; wherein the first constraint condition is to minimize the reduction in torque ripple and the reduction in dynamic response capability;

[0062] The simulation data of the stator skew slots was simulated by establishing a finite element model. The finite element model was solved at different skew degrees. Based on the stator skew slot model of the motor, the torque ripple and mechanical time constant obtained are shown in Table 2.

[0063] Table 2 Torque ripple and mechanical time constant when the skew angle is 0-10°

[0064]

[0065] Table 2 shows that torque ripple decreases with increasing skew, while the mechanical time constant increases with increasing skew. Since the mechanical time constant increases significantly with increasing skew after 10°, it is not considered.

[0066] In order to make the torque ripple drop greatly and the mechanical time constant increase slightly, Here, β1 = ΔT1 - ΔT2 represents the torque ripple decrease value, and β2 = τ2 - τ1 represents the mechanical time constant increase value. The larger it is, the better the effect is. Please refer to Table 3 for details.

[0067] Table 3 Variation of different skew degrees

[0068]

[0069] It can be seen from Table 3 that when the chute angle is 2 0 becomes 3 0 hour, This indicates that the use of 3 0 The skew degree can minimize the mechanical time constant while minimizing the torque ripple, so a skew degree of 3 is selected. 0 Proceed with subsequent model building.

[0070] Step S103, analyzing the motor structural parameters that affect the sensitivity of the initial design parameters based on the motor model, and obtaining the motor structural parameters as the stator outer diameter, permanent magnet thickness, tooth width, and air gap length;

[0071] In the step of analyzing the motor structural parameters that affect the sensitivity of the initial design parameters based on the motor model, Sobol sensitivity analysis is adopted; this can reduce the number of subsequent tests and alleviate the complexity of calculations.

[0072] Among them, in Table 1, sensitivity analysis is performed on several optimization parameters, and the final sensitivity analysis results are as follows: Figure 2 As shown;

[0073] The higher the absolute value of the sensitivity factor, the greater the influence of the optimization variable on the optimization objective, and vice versa. Figure 2It can be seen that the significant parameters that have a greater impact on the two optimization objectives are R2 and h m 、B s1 and g, and the above four parameters are used as motor structure parameters.

[0074] The significant variable levels after sensitivity analysis are shown in Table 4. Finite element simulation software was used to conduct the experiment, and a Box-Behnken (BBD, response surface methodology) design with four variables required 29 experiments. Table 4 shows the code and level value of each design variable;

[0075] Table 4 Significant variable levels

[0076]

[0077] The final regression equation of the optimization objective can be expressed as:

[0078] ΔT=29.82901-0.803543×R2-3.2429×h m -2.28548×B s1 -9.2697×g+0.095288×R2×h m -0.175492×R2×B s1 -0.050172×R2×g-1.07744×h m ×B s1 +0.676725×h m ×g-0.61125×B s1 ×g+0.010817×R2 2 -0.060553×h m 2 +3.50163×B s1 2 +5.22456×g 2 (6);

[0079] τ m =6.43702-0.13528×R2-3.48663×h m -0.6747×B s1 +2.88003×g+0.0555×R2×h m -0.1395×R2×B s1 -0.063281×R2×g-0.524×h m ×B s1 -1.195×h m ×g+0.84×B s1 ×g+0.004898×R2 2 +0.862×h m2 +1.562×B s1 2 +1.14766×g 2 (7);

[0080] Step S104, constructing a fitting model of the design goal by using the response surface method, wherein the design goal is to reduce the torque ripple of the motor and improve the dynamic response capability;

[0081] In step S104, the steps of constructing a fitting model of the design target by using the response surface method specifically include:

[0082] Obtain a table of significant variable levels after sensitivity analysis, conduct finite element simulation tests, and record each group of optimized parameters and the corresponding observed values ​​of the response variables;

[0083] Construct a response surface regression equation and use the recorded experimental data to fit and solve the response surface regression equation:

[0084] Among them, when performing fitting solution, the second-order polynomial regression method is used to establish the fitting model, and the established fitting model is:

[0085]

[0086] Where: y is the optimization target; b0 is the constant term; x i 、x ij corresponds to different optimization parameters; b i 、b ij and b ii are the coefficients of the first and second terms of the optimization parameters; ε is the random error;

[0087] In this step, the approximation of the fitted model is determined by the multiple correlation coefficient R 2 express;

[0088] The regression equation of the optimization target is analyzed by significance test; the approximation degree of the fitting model is determined by the multiple correlation coefficient R 2 express;

[0089] R 2 =1-S SE / S ST (9);

[0090]

[0091] where Y i is the true value of the test set, y i is the fitting value, n is the number of experiments, S SE is the residual sum of squares, S ST is the total squared dispersion. 2The closer the value of is to 1, the closer the test data of the fitted model is to the actual value. The correlation coefficients of the two responses of torque ripple and mechanical time constant are named

[0092] and The values ​​of are 0.973 and 0.996 respectively. It can be seen that the R 2 The values ​​of are all greater than 0.9, which have good applicability and good correlation.

[0093] The response surface diagram of the optimization objective is as follows Figure 3 As shown in Figure 2, when the size of a single parameter is changed, the torque ripple and the mechanical time constant will change at the same time. Therefore, in order to avoid conflicts between design goals, a multi-objective optimization algorithm is used.

[0094] Step S105 : optimizing the motor structural parameters using an improved dung beetle optimization algorithm based on the fitting model.

[0095] In this embodiment, to achieve the design goals of low torque ripple and high dynamic response capability, the Dung Beetle Optimizer (DBO) algorithm is used to iteratively optimize the fitting model. However, the original DBO algorithm (Dung Beetle Optimizer) has certain deficiencies in local search capabilities and handling multi-objective optimization problems. To overcome these deficiencies, this embodiment combines the DBO algorithm with a non-dominated sorting strategy to obtain an improved Dung Beetle Optimizer (NSDBO) algorithm; namely, the Non-dominated Sorting Dung Beetle Optimizer (NSDBO) algorithm. This algorithm can effectively improve the algorithm's convergence speed and optimization accuracy, ultimately obtaining a more superior Pareto optimal solution set.

[0096] In this embodiment, the method of this embodiment further includes: combining the dung beetle optimization algorithm with a non-dominated sorting strategy to obtain an improved dung beetle optimization algorithm; the non-dominated sorting strategy is configured as follows:

[0097] For each solution, calculate the number of solutions it is dominated by;

[0098] Solutions that are not dominated by any solution are classified as the first layer;

[0099] Among the solutions in the first layer, remove all solutions that dominate other solutions and assign the remaining solutions to the second layer;

[0100] Repeat the above process until all solutions are classified into different non-dominated layers;

[0101] Among the solutions in the same layer, they are sorted by the degree of congestion, which is the density of any solution from other solutions.

[0102] Output all solutions in the non-dominated layer as the Pareto optimal solution set.

[0103] In this embodiment, the non-dominated sorting is performed according to the degree of dominance of the Pareto optimal solution, such as Figure 4 As shown in Figure 1, solutions that are not dominated by any other solution are assigned rank 1, solutions dominated by only one solution are assigned rank 2, solutions dominated by only two solutions are assigned rank 3, and so on. Solutions are then selected based on their rank to improve the quality of the dung beetle population and maintain population diversity through crowding calculations.

[0104] like Figure 5 As shown, the algorithm optimization specifically includes steps S1) to S7);

[0105] S1) Initialize the population: Initialize the dung beetle population N = 200, and the initial maximum iteration P = 500;

[0106] S2) Calculate fitness: Calculate the fitness value of the individual according to the objective function;

[0107] S3) Screening the population by fast non-dominated sorting and crowding calculation to obtain the top N individuals as a new population;

[0108] S4) The position update formula for the ball-rolling dung beetle, the breeding dung beetle, the foraging dung beetle, and the thief dung beetle is:

[0109] X i t+1 =X i t +βkX i t-1 +b|X i t -X worst t | (12);

[0110] X i t+1 =X i t +tanθ|X i t -X i t-1 | (13);

[0111] B i t+1 =X gbest t +b1(B i t -Lb * )+b2(B i t -Ub* ) (14);

[0112] X i t+1 =X i t +C1(X i t -Lb l )+C2(X i t -Ub l ) 15);

[0113] X i t+1 =X lbest t +M(|X i t -X gbest t |+|X i t -X lbest t |) (16);

[0114] Formulas (12) and (13) are the formulas for the behavior stages of the rolling ball dung beetle, and formulas (14) to (16) are the formulas for the reproductive behavior stage, foraging behavior stage, and stealing behavior stage, respectively. Where t represents the current iteration number; X i t represents the position of the i-th dung beetle in the population at the t-th iteration; X worst t Indicates the worst position in the current population; |X i t -X worst t |Used to simulate the changes in light intensity; B i t+1 is the position of individual iteration during the foraging process; b1 and b2 are two independent random vectors of size 1×D, D represents the dimension of the problem, and M is a constant;

[0115] S5) recalculating the fitness value of each individual in the new population, determining the dominance relationship between individuals based on the non-dominated sorting and crowding distance calculation, and selecting the top N individuals as the parent population;

[0116] S6) Whether the algorithm termination condition is met, if so, stop iteration, otherwise repeat steps S2) to S5);

[0117] S7) The first N individuals obtained are the Pareto solution set.

[0118] The optimized Pareto front is as follows Figure 6As shown in the figure, according to the set constraints and combined with the goals of low torque ripple and high dynamic response capability of the motor, the obtained Pareto frontier is screened and finally selected. Figure 6 The solution marked with a triangle is the optimal solution. The final values ​​of the motor optimization parameters are shown in Table 5.

[0119] Table 5 Final values ​​of optimized parameters

[0120]

[0121] The comparison of torque ripple and dynamic response capability before and after optimization is shown in the figure below. Figure 7 、 Figure 8 shown.

[0122] In another embodiment, Figure 12 As shown, a motor multi-objective optimization system based on an improved dung beetle optimization algorithm is used in the method described above, and the system includes:

[0123] Mathematical model input module 100, used to establish a mathematical model of motor torque ripple and dynamic response capability, and determine initial design parameters that affect torque ripple and dynamic response capability;

[0124] A simulation optimization module 200 is configured to determine an optimal angle that satisfies a first constraint condition based on simulation data of the stator slots of the motor, so as to establish a motor model; wherein the first constraint condition is to minimize the reduction in torque ripple and the reduction in dynamic response capability;

[0125] a sensitivity analysis module 300 for analyzing, based on the motor model, motor structural parameters that affect the sensitivity of the initial design parameters, and obtaining the motor structural parameters as the stator outer diameter, permanent magnet thickness, tooth width, and air gap length;

[0126] The target solving module 400 is used to construct a fitting model of the design target by using the response surface method. The design target is to reduce the torque ripple of the motor and improve the dynamic response capability;

[0127] The parameter optimization module 500 is used to optimize the motor structure parameters based on the fitting model using an improved dung beetle optimization algorithm.

[0128] The above embodiment provides a motor multi-objective optimization method based on an improved dung beetle optimization algorithm, and based on this motor multi-objective optimization method based on the improved dung beetle optimization algorithm, provides a motor multi-objective optimization system based on the improved dung beetle optimization algorithm. This motor multi-objective optimization method based on the improved dung beetle optimization algorithm simulates and analyzes different stator slot angles and optimizes the motor structural parameters using a dung beetle optimization algorithm based on non-dominated sorting (i.e., an improved dung beetle optimization algorithm). This method can more specifically address motor speed response issues and improve the overall performance of the motor. The optimal angle of the stator slots is analyzed to minimize the increase in the mechanical time constant while significantly reducing torque ripple, helping to maintain the motor's good speed response capability. This allows the motor to operate smoothly and respond quickly to speed changes, improving its overall performance. The improved dung beetle optimization algorithm is used to optimize the parameters of the fitting model, simultaneously addressing the motor's design goals of low torque ripple and high dynamic response capability while obtaining non-dominated solutions. The algorithm's excellent global search capability enables it to avoid local optima in high-dimensional space and approach global optimal parameters. It can effectively avoid premature convergence, continuously explore new solution spaces, and ultimately accurately optimize the fitting model parameters to achieve optimization of motor structure parameters.

[0129] It should be understood that, although the various steps in the flow chart of each embodiment of the present invention are shown in sequence according to the indication of the arrows, these steps are not necessarily performed in sequence according to the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in order, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0130] 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 and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A motor multi-objective optimization method based on an improved dung beetle optimization algorithm, characterized in that: The method comprises: Establish a mathematical model of motor torque ripple and dynamic response capability, and determine the initial design parameters that affect torque ripple and dynamic response capability; Determining an optimal angle that satisfies a first constraint condition based on simulation data of a stator skew slot of a motor to establish a motor model; wherein the first constraint condition is to minimize a decrease in torque ripple and a decrease in dynamic response capability; Based on the motor model, analyzing the motor structural parameters that affect the sensitivity of the initial design parameters, and obtaining the motor structural parameters as the stator outer diameter, permanent magnet thickness, tooth width, and air gap length; The response surface methodology is used to construct a fitting model for the design goal, which is to reduce the torque ripple of the motor and improve its dynamic response capability. Based on the fitting model, the motor structural parameters are optimized using an improved dung beetle optimization algorithm; The method further includes: combining the dung beetle optimization algorithm with a non-dominated sorting strategy to obtain an improved dung beetle optimization algorithm; the non-dominated sorting strategy is configured as follows: For each solution, calculate the number of solutions it is dominated by; Solutions that are not dominated by any solution are classified as the first layer; Among the solutions in the first layer, remove all solutions that dominate other solutions and assign the remaining solutions to the second layer; Repeat the above process until all solutions are classified into different non-dominated layers; Among the solutions in the same layer, they are sorted by the degree of congestion, which is the density of any solution from other solutions. Output all solutions in the non-dominated layer as the Pareto optimal solution set.

2. The method according to claim 1, characterized in that The mathematical model of the motor torque ripple is: Among them, T max is the output torque T N Maximum value; T min is the output torque T N Minimum value; T av is the average torque value.

3. The method according to claim 2, characterized in that The mathematical model of the motor's dynamic response capability is: Among them, ρ w is the conductor resistivity; L E is the average half-turn length; N is the number of turns in series; A 01 is the cross-sectional area of ​​the enameled wire; a1 is the number of parallel branches; B av is the average magnetic flux density; D a is the diameter of the magnetic steel surface; l is the length of the conductor in the magnetic field; ρ m is the permanent magnet density; R1 is the rotor outer diameter; ρ Fe is the rotor core density; h m is the thickness of the permanent magnet; g is the air gap length; R2 is the stator outer diameter.

4. The method according to claim 1, wherein The steps of constructing a fitting model for the design objective through response surface methodology include: Obtain a table of significant variable levels after sensitivity analysis, conduct finite element simulation tests, and record each group of optimized parameters and the corresponding observed values ​​of the response variables; Construct a response surface regression equation and use the recorded experimental data to fit and solve the response surface regression equation; Among them, when performing fitting solution, the second-order polynomial regression method is used to establish the fitting model, and the established fitting model is: Where: y is the optimization target; b0 is the constant term; x i 、x ij corresponds to different optimization parameters; b i 、b ij and b ii are the coefficients of the first and second terms of the optimization parameters; ε is the random error; The regression equation of the optimization target is analyzed by significance test; the approximation degree of the fitting model is determined by the multiple correlation coefficient R 2 express; R 2 =1-S SE / S ST ; where Y i is the true value of the test set, y i is the fitting value, n is the number of experiments, S SE is the residual sum of squares, S ST is the total dispersion squared.

5. The method according to claim 1, wherein In the step of analyzing the motor structural parameters that affect the sensitivity of the initial design parameters based on the motor model, Sobol sensitivity analysis is adopted.

6. A motor multi-objective optimization system based on an improved dung beetle optimization algorithm, used in the method according to any one of claims 1 to 5, characterized in that: The system includes: Mathematical model input module, used to establish a mathematical model of motor torque ripple and dynamic response capability, and determine the initial design parameters that affect torque ripple and dynamic response capability; a simulation optimization module, configured to determine an optimal angle that satisfies a first constraint condition based on simulation data of the stator skew slots of the motor, so as to establish a motor model; wherein the first constraint condition is to minimize a decrease in torque ripple and a decrease in dynamic response capability; a sensitivity analysis module, configured to analyze, based on the motor model, motor structural parameters that affect the sensitivity of the initial design parameters, and obtain the motor structural parameters as the stator outer diameter, permanent magnet thickness, tooth width, and air gap length; The target solution module is used to construct a fitting model of the design target through the response surface method. The design target is to reduce the torque ripple of the motor and improve the dynamic response capability; A parameter optimization module, configured to optimize the motor structural parameters using an improved dung beetle optimization algorithm based on the fitting model; The improved dung beetle optimization algorithm is obtained by combining the dung beetle optimization algorithm with the non-dominated sorting strategy; the non-dominated sorting strategy is configured as follows: For each solution, calculate the number of solutions it is dominated by; Solutions that are not dominated by any solution are classified as the first layer; Among the solutions in the first layer, remove all solutions that dominate other solutions and assign the remaining solutions to the second layer; Repeat the above process until all solutions are classified into different non-dominated layers; Among the solutions in the same layer, they are sorted by the degree of congestion, which is the density of any solution from other solutions. Output all solutions in the non-dominated layer as the Pareto optimal solution set.

Images