Permanent magnet synchronous motor cogging torque parameter optimization method
By combining genetic algorithms and particle group algorithms, and introducing disturbance mechanisms and scope update mechanisms, the problem of time -consuming and labor -intensive time -consuming and consuming of traditional methods has been solved, and more efficient alcarius torque optimization is achieved, reducing motor vibration.
Patent Information
- Application Number
- CN202411405371.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-09
- Publication Date
- 2025-01-24
AI Technical Summary
The traditional cogging torque optimization method of permanent magnet synchronous motors requires a lot of time and computing power when optimizing multi-objectives, and it is difficult to effectively improve the optimization effect of cogging torque.
Combining the genetic algorithm and the particle swarm algorithm, by introducing the disturbance mechanism and the range update mechanism, the global and local search capabilities of the algorithm in the iterative process are improved, the tendency of convergence toward the global optimal solution is improved, and the cogging torque is optimized.
This method greatly reduces time costs and computing power costs, improves the optimization effect of the alcile torque, reduces motor vibration, and increases work efficiency.
Smart Images
Figure CN119357535A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of permanent magnet motor performance optimization, and in particular to a method for optimizing cogging torque parameters of a permanent magnet synchronous motor. Background Art
[0002] With the rapid development of aerospace and new energy vehicles in recent years, permanent magnet synchronous motors have also been widely used. This is because permanent magnet synchronous motors have good power density and torque density, simple structure, high reliability, low failure rate, and can adjust the stator and rotor structure and optimize the motor design according to different application scenarios. However, all permanent magnet synchronous motors with cogging will have cogging torque. When the cogging torque is too large and causes vibration to the motor, it will have a serious impact on the motor operation.
[0003] In the early stages of research and development at home and abroad, a large number of scholars proposed to achieve effective reduction of cogging torque by adjusting the motor structure. C. Peng et al. proposed rotor segmentation and changing the symmetry of the pole width. On this basis, they analyzed the influence of the pole width on the cogging torque and finally determined the optimized motor parameters. Shi Songning et al. proposed an analytical expression for the cogging torque based on unequal tooth widths, which can significantly weaken the cogging torque under a certain tooth width. Yang Yubo analyzed different adjacent slot widths and clarified that this method can effectively reduce the component with the largest amplitude in the small harmonic under specific circumstances. In addition, there are methods such as skewed poles and skewed slots, which can greatly reduce the torque caused by the cogging effect and reduce the torque pulsation caused by it.
[0004] In recent years, unlike traditional single-objective decision-making, multi-objective decision-making is constructed by introducing decision theory in the process of motor parameter optimization, assigning different weights to each output variable, establishing the connection between different parameters, and establishing constraints between parameters with the structure as the solution range, and performing sufficient iterations. The same method is also applicable to introducing optimization models and conditional constraints into heuristic algorithms, and has become a hot topic for many scholars. KPBadgujar et al. changed the thickness of the permanent magnet of the surface-mounted permanent magnet synchronous motor so that the thickness of the entire permanent magnet increased from both sides to the middle, and found that the cogging torque could be reduced within a certain range. Zhu Ziqiang et al. found through research that changing the pole arc-pole pitch ratio can significantly optimize the cogging torque; in addition to effectively improving the cogging torque in motor design. W. Ren et al. optimized the structure of the traditional built-in permanent magnet synchronous motor, adjusted the boundary of the rotor for placing permanent magnets, and then used the parameter scanning method to determine the motor parameters and verified the effectiveness of the method through simulation.
[0005] Using algorithms to optimize parameters can greatly improve the calculation efficiency while keeping the calculation error within a certain range compared to using finite element software. In recent years, many scholars have studied new methods of combining algorithms with simulation. Among them, a large number of algorithms such as genetic algorithms, particle swarm algorithms and Taguchi methods have improved the iteration speed and accuracy in the process of combining with parameter optimization, and verified the effectiveness of the methods through finite element analysis. V. Simón-Sempere et al. proposed a reverse approach different from the traditional method. First, they assumed that the cogging torque was effectively weakened, obtained the waveform data under this state, derived the corresponding structural parameters and verified them through finite element analysis to obtain the optimized structural parameters. Gao Jian et al. found that harmonics are generated by injecting current, and the cogging torque comes from the combined effect of the magnetic flux harmonics generated by the stator slots and the air gap permeance harmonics. Therefore, the harmonics generated by the injected current are just harmonics of the same order, the same magnitude and opposite phase, which offset the harmonics of the cogging torque. Yan Hongliang and others studied the auxiliary slots on the stators and rotors, and by parameterizing the slot dimensions, they found that both the slot depth and slot width will have a suppressive effect on the cogging torque.
[0006] However, traditional parameter optimization can only be performed by parametric scanning of a single objective. If multi-objective optimization is required, the parameters can only be decoupled first, and then parametric scanning can be performed in sequence. The completion process requires a lot of time and computing power. Summary of the invention
[0007] The purpose of the present invention is to provide a method for optimizing the cogging torque parameters of a permanent magnet synchronous motor, which can complement the shortcomings of the algorithm by combining the genetic algorithm and the particle swarm algorithm, and introduce a disturbance mechanism and a range update mechanism, which is beneficial to improving the global and local search capabilities of the algorithm during the iteration process, effectively improving the trend of convergence to the global optimal solution, and improving the optimization effect of the cogging torque.
[0008] In order to achieve the above object, the technical solution adopted by the present invention is: a method for optimizing the cogging torque parameters of a permanent magnet synchronous motor, comprising the following steps; Step 1: Establish the cogging torque analytical model of the permanent magnet synchronous motor. The cogging torque is expressed as: Where, Q represents the number of stator slots, L Fe represents the axial length of the armature core, μ0 represents the air permeability, R2 represents the stator inner diameter, R1 represents the rotor outer diameter, p represents the number of pole pairs, and n represents an integer. is also an integer, α represents the angle between the center line of a permanent magnet and the center line of a tooth of the stator core, G n represents the Fourier coefficient after Fourier expansion of the corresponding air gap permeability; Step 2: Based on the optimization goal of reducing the cogging torque and taking the analyzed cogging torque as the model, the improved particle swarm algorithm is used to perform multiple optimization iterations of the permanent magnet synchronous motor parameters, and finally the parameters and cogging torque of the optimization model are output; The particle swarm algorithm and genetic algorithm are combined into an optimization algorithm, and the iterative process is re-assigned; combined with the global search capability of the genetic algorithm, in the early stage of the algorithm, only the genetic algorithm iteration is performed, and under the action of the perturbation mechanism, the fluctuation of the iteration curve is increased to fully accumulate sample data; when the genetic algorithm iteration is completed, the best individuals that appear in each generation of iteration are collected and processed to generate a new solution domain and initial model, and then the particle swarm algorithm iteration is started, and local optimization is performed on the best individuals collected in the early stage.
[0009] Preferably, in step 2, a disturbance mechanism is introduced to intervene when the iteration curve is in a long-term unchanged state to change the current iteration trend.
[0010] Preferably, in step 2, a sample closest to the current group optimal solution is selected from the data collected in the early stage for replacement and re-solution.
[0011] Preferably, in the perturbation mechanism in the early stage of the algorithm in step 2, in order to assign the group optimal solution to an inferior solution, the midpoint of the value range of each component of the sample is used as a reference, and the opposite component of the position component of the group optimal solution relative to this reference is calculated, and the generated new sample is the assigned inferior solution. X m =(AB) / 2 ΔX=|X gb -X m | Among them, X op represents the position component of the inverse solution; X m represents the midpoint in the solution domain; X gb Represents the position component of the group's historical optimal solution; A and B represent the upper and lower limits of the solution domain respectively; △X represents the absolute value of the change between the inverse solution and the current optimal solution position variable.
[0012] Preferably, after the genetic algorithm completes the iteration in step 2, the sample data is collected and sorted according to the fitness size, and the component information of each sample is counted to determine the upper and lower limits of each component in all the information, and finally the new upper and lower limits are synchronously updated to the particle swarm algorithm.
[0013] The beneficial effects of the present invention are:
[0014] This solution uses the method of model analysis and algorithm optimization to improve the efficiency of parameter optimization and simplify the optimization operation. By analyzing and establishing the mathematical model of the relevant motor, and combining the model and algorithm for iteration, the time cost and computing power cost are greatly reduced, and the optimization process with ideal effect can be completed in a relatively short time. On the basis of genetic algorithm and particle swarm algorithm, the shortcomings of the algorithm are supplemented by combining the two, and the disturbance mechanism and range update mechanism are introduced, which is conducive to improving the global and local search capabilities of the algorithm during the iteration process, effectively improving the trend of convergence to the global optimal solution, improving the optimization effect of the cogging torque, weakening the motor vibration caused by the cogging torque, and increasing work efficiency. Compared with the lengthy setup process and calculation process required for simulation, this method is easy to operate and can perform a large number of calculations by simply modifying the algorithm data, which is conducive to application to various production scenarios and improving productivity. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0016] Figure 1 Iteration curve of particle swarm algorithm.
[0017] Figure 2 Iterative principle diagram of particle swarm algorithm.
[0018] Figure 3 This is the iterative principle diagram of the genetic algorithm.
[0019] Figure 4 This is a sample diagram of genetic algorithm data.
[0020] Figure 5 This is an iterative principle diagram of the optimization algorithm of the present invention.
[0021] Figure 6 Flow chart of the perturbation mechanism of the present invention.
[0022] Figure 7 It is an iterative curve diagram of the genetic algorithm under the disturbance mechanism of the present invention.
[0023] Figure 8 It is an iterative curve diagram of the particle swarm algorithm under the disturbance mechanism of the present invention.
[0024] Fig. 9 It is the iteration curve and comparison diagram of the optimization algorithm of the present invention.
[0025] Fig.10 It is a simulation result diagram of the present invention. DETAILED DESCRIPTION
[0026] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.
[0027] The structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with the technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantive technical significance. Any structural modification, change in proportion or adjustment of size shall still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.
[0028] At the same time, it should be noted that: unless otherwise defined, the technical terms or scientific terms used in this article should be the ordinary meaning understood by people with ordinary skills in the technical field to which the present invention belongs. The meaning of "multiple" used in the patent application specification and claims of the present invention is two or more; the orientation or position relationship indicated by the terms "upper", "lower", "left", "right", "front end", "rear end", etc. is based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. Changes or adjustments in their relative relationships should also be regarded as the scope of the implementation of the present invention without substantially changing the technical content.
[0029] The present invention is further described in detail below in conjunction with embodiments and with reference to the accompanying drawings.
[0030] The present invention discloses a method for optimizing the cogging torque parameters of a permanent magnet synchronous motor. The embodiment includes the following steps: Step 1: Establish the cogging torque analytical model of the permanent magnet synchronous motor. The cogging torque is expressed as: Where, Q represents the number of stator slots, L Fe represents the axial length of the armature core, μ0 represents the air permeability, R2 represents the stator inner diameter, R1 represents the rotor outer diameter, p represents the number of pole pairs, and n represents an integer. is also an integer, α represents the angle between the center line of a permanent magnet and the center line of a tooth of the stator core, G n represents the Fourier coefficient after Fourier expansion of the corresponding air gap permeability; Step 2: Based on the optimization goal of reducing the cogging torque and taking the analyzed cogging torque as the model, the improved particle swarm algorithm is used to perform multiple optimization iterations of the permanent magnet synchronous motor parameters, and finally the parameters and cogging torque of the optimization model are output; The particle swarm algorithm and genetic algorithm are combined into an optimization algorithm, and the iterative process is re-assigned; combined with the global search capability of the genetic algorithm, in the early stage of the algorithm, only the genetic algorithm iteration is performed, and under the action of the perturbation mechanism, the fluctuation of the iteration curve is increased to fully accumulate sample data; when the genetic algorithm iteration is completed, the best individuals that appear in each generation of iteration are collected and processed to generate a new solution domain and initial model, and then the particle swarm algorithm iteration is started, and local optimization is performed on the best individuals collected in the early stage.
[0031] The cogging torque in step 1 is the torque generated by the interaction between the permanent magnet and the stator tooth slot when no power is applied. The magnetic permeance between the permanent magnet and the stator tooth slot will change with the relative motion of the rotor. According to the assumption that the magnetic lines of force emitted from the permanent magnet pass through the air gap into the tooth part, the magnetic permeance of the permanent magnet pole arc part remains unchanged when the rotor position changes. Due to the different positions of the rotor, the magnetic lines of force and magnetic permeance at both ends of the permanent magnet will also change, causing changes in the magnetic field energy. Therefore, the cogging torque can be expressed as:
[0032] Among them, W represents the magnetic field energy; α represents the angle between the center line of a permanent magnet and the center line of a tooth of the stator core. To simplify the model, the following assumptions are made: the magnetic permeability of the armature core is infinite, and the loss of magnetic lines of force when passing through the magnetic circuit of the armature core is very small, so the magnetic field energy at the armature core can be ignored, and the motor magnetic field energy is approximately the magnetic field energy of the air gap and the permanent magnet; to simplify the model, when calculating the air gap magnetic density, the angle between the center line of the permanent magnet and the center line of the core tooth is taken as 0; all permanent magnets are of the same shape and size, and the permanent magnet material and air magnetic permeability are the same. Therefore, the magnetic field energy is expressed as:
[0033] Among them, μ0 represents the air magnetic permeability, and the magnetic field energy W depends on the structural data of the motor. Therefore, along the circumferential direction of the rotor, the air gap magnetic flux distribution can be approximately expressed as:
[0034] Substituting (3) into (2) yields:
[0035] Among them, at θ = 0, where B r represents the residual magnetic flux density of the permanent magnet; h m represents the length of the permanent magnet in the magnetizing direction; g represents the effective air gap length. The sum is decomposed by Fourier. According to the assumption that all permanent magnets have the same shape and size, B r 2 The Fourier expansion of (θ) can be expressed as:
[0036] in:
[0037] Among them, α p Represents the pole arc coefficient of the permanent magnet pole. According to the above assumptions, at the position of θ = 0, The Fourier expansion of can be expressed as:
[0038] Among them, θ s0 represents the armature slot width, G0 and Gn represent the Fourier coefficients after the Fourier expansion of the corresponding air gap permeance. Substituting equations (5) and (8) into (4), the cogging torque is expressed as:
[0039] Where, Q represents the number of stator slots, L Fe represents the axial length of the armature core, R2 represents the inner diameter of the stator, R1 represents the outer diameter of the rotor, p represents the number of pole pairs, and n represents an integer. Also an integer.
[0040] The particle swarm algorithm in step 2 is an application of the bird predation process, showing the process of collaborative optimization between individuals and groups, with fast convergence speed, strong local search ability, and simple algorithm operation. For the same initial model, the particle swarm algorithm is iterated, such as Figure 1 As shown, each iterative curve decreases in a step-like manner from the starting point to the final convergence.
[0041] like Figure 2 As shown in the figure, the algorithm's iterative mechanism combines the individual's historical speed, the optimal position of the entity, and the historical optimal position of the group, and the corresponding inertia coefficient, self coefficient, and social coefficient. The three coefficients determine the weight of each element in each round of iteration, which in turn affects the sample iteration. Therefore, a reasonable allocation of the three coefficients can effectively improve the algorithm's convergence and speed up the iteration. However, the iterative mechanism also causes the algorithm to converge locally in a relatively large solution domain. Ultimately, all samples converge to the same local value, and cannot jump out of the local optimum and re-search for global optimization.
[0042] Genetic algorithms are widely used in various scientific researches due to their good global optimization ability. Figure 3 As shown in the figure, the crossover and mutation mechanisms play an important role. Through the exchange between the parent and the parent, after multiple iterations, more samples will become excellent individuals, enhancing their ability to find the best, and the combination of the mutation mechanism and the crossover mechanism also provides more possibilities for expanding the sample capacity.
[0043] Different from the good local optimization ability of particle swarm algorithm, genetic algorithm can solve larger sample data in the iterative process, expand the distribution of samples in the solution domain, and fully demonstrate the quality of the solution model in each interval, such as Figure 4 The figure shows the result of iterating a model. After multiple iterations, the probability of a sample becoming an excellent individual is greatly improved. Due to the existence of the mutation mechanism, the probability of the sample entering local convergence is reduced, and all samples are prevented from iterating to the same individual through the crossover mechanism. However, the existence of the crossover mechanism and the mutation mechanism will make the iterative curve need to be solved for a long time before it can finally converge, which obviously cannot improve the computational efficiency and reduce the computational resources.
[0044] For the same analytical model from Figure 1 , 4 It can be seen that although the genetic algorithm has good global search capabilities, it also has the disadvantages of slow convergence and weak local search capabilities. Although the particle swarm algorithm converges quickly, it is easy to enter the local optimum and lacks an effective mechanism to jump out of the local optimum and continue to search for global optimization. Therefore, this scheme will combine the two algorithms and introduce a perturbation mechanism and a range update mechanism to improve the algorithm's global optimization capability and optimization speed.
[0045] In order to improve the global search capability of the particle swarm algorithm and reduce the impact of local convergence during the algorithm iteration, the particle swarm algorithm is combined with the genetic algorithm and the iteration process is re-assigned, such as Figure 5 As shown. Combined with the global search capability of the genetic algorithm, in the early stage of the algorithm, only the genetic algorithm iteration is performed. At the same time, under the action of the perturbation mechanism, the fluctuation of the iteration curve is increased to fully accumulate sample data. When the genetic algorithm iteration is completed, the best individuals that appear in each generation of iteration are processed and collected to generate a new solution domain and initial model, and then the particle swarm algorithm iteration is started, and local optimization is performed for the best individuals collected in the early stage.
[0046] In the early stage of the algorithm, the genetic algorithm will periodically enter the local optimum. Although it will eventually re-enter the global optimization through its own iterative mechanism, the time required for this process cannot be quantified, which is not conducive to improving computational efficiency or expanding global search. Therefore, when the iteration curve remains unchanged for a long time, a perturbation mechanism is introduced to intervene and change the current iteration trend, such as Figure 6-8 .
[0047] In the later stage of the algorithm, the solution of the particle swarm algorithm is based on the previous iterative data. After multiple iterations, the curve will gradually converge. In order to test the convergence value at this time and avoid entering local convergence, a sample closest to the current group's optimal solution is taken from the data collected in the early stage to replace it and solve it again.
[0048] In the early perturbation mechanism of the algorithm, in order to assign the group optimal solution to an inferior solution, the midpoint of the value range of each component of the sample is used as a reference, and the opposite component of the group optimal solution position component relative to this reference is calculated. The new sample generated is the assigned inferior solution. X m =(AB) / 2 (13) ΔX=|X gb -X m | (14)
[0049] Among them, Xop represents the position component of the inverse solution; Xm represents the midpoint in the solution domain; Xgb represents the position component of the group's historical optimal solution; A and B represent the upper and lower limits of the solution domain respectively; △X represents the absolute value of the change between the inverse solution and the current optimal solution position variable.
[0050] After the genetic algorithm completes the iteration, the sample data will be collected and sorted according to the fitness size, and the component information of each sample will be counted to determine the upper and lower limits of each component in all the information, and finally the new upper and lower limits will be synchronized to the particle swarm algorithm. Gradually narrowing the scope of the solution domain is conducive to improving the calculation efficiency during the later iteration of the algorithm, completing the global convergence at a faster speed, and reducing the iterative curve for local search.
[0051] This method aims to improve the efficiency of motor parameter optimization through algorithms, and reduce the excessive torque pulsation caused by excessive cogging torque during motor operation, which in turn affects the flight status and the comfort of the flight crew. Based on this, the algorithm has the following key points when it is written:
[0052] ① The disturbance mechanism in the early stage of the algorithm, how to define that the iteration curve has entered the local convergence. After many solution tests and combining the characteristics of the algorithm itself, it is proposed to use the optimal solution of the group in each generation as the standard. When this standard does not change in 30-50 generations, it is considered to have entered the local convergence state and the interference mechanism is implemented.
[0053] The interference mechanism is to re-assign the current group optimal solution. Due to the existence of the crossover mechanism, the group optimal solution is usually not unique among all samples, but there are multiple samples in the group optimal state at the same time, so it is necessary to assign values to all samples in this state to prevent the iterative curve from entering local convergence again. For the assignment, any inferior solution can be used. After testing, taking the inverse solution of the current group optimal solution as an example, the solution result differs by 2-3 orders of magnitude from the group optimal solution before the assignment, which provides sufficient optimization space for the next iterative curve.
[0054] ② In the large number of solution results accumulated in the early stage of the algorithm, although the interference mechanism is implemented, there are still many duplicate values. When sorting and collecting, only one of them needs to be retained. If there are the same fitness but different component values, they should also be retained, so that the final collected data set should be multiple samples with different component values.
[0055] ③ With the good local search capability of the particle swarm algorithm, it will converge faster than the genetic algorithm iteration. At this time, the algorithm needs to be located in a suitable solution domain, and the perturbation mechanism in the later stage of the algorithm provides a guarantee for the convergence of the algorithm. At this time, re-assignment should avoid using samples that have appeared in this stage, and at the same time, it should meet the requirement of being closest to the most possible solution of the group.
[0056] The iterative curve after perturbation will also gradually converge. How to define the global convergence depends on the convergence value before and after the perturbation and the error between the finite element simulation solution result. Usually, the solution results before and after the perturbation will not be equal, but should be kept at the same order of magnitude, and the relative error with the finite element simulation solution result should not exceed 5%. For the curve that meets the conditions, the iteration can be terminated as it has completed convergence.
[0057] In the specific application, taking the 18s / 16p surface-mount permanent magnet synchronous motor as an example, after algorithm optimization, finite element simulation is performed based on the solution results. Fig. 9 As shown, Fig. 9 (1) is the improved particle swarm algorithm iteration curve; (2) is the local enlargement diagram of the early stage of the algorithm; (3) is the local enlargement diagram of the late stage of the algorithm; (4) is the particle swarm algorithm iteration curve and local enlargement diagram; (5) is the genetic algorithm iteration curve and local enlargement diagram. From the algorithm optimization results, it can be seen that the convergence value under the improved algorithm is much smaller than the results of the genetic algorithm or particle swarm algorithm respectively.
[0058] like Fig.10 As shown, Fig.10 (1) is the waveform of the air gap flux density; (2) is the spatial harmonic distribution diagram; (3) is the waveform of the optimized model cogging torque; (4) is the waveform of the original model cogging torque; (5) is the waveform of the optimized model back electromotive force; (6) is the waveform of the original model back electromotive force. It can be seen from the simulation data of the optimized model that the peak-to-peak value of the cogging torque is reduced by 25%. After the air gap flux density is transformed by Fourier, the third and fifth harmonics in the frequency amplitude diagram are significantly weakened, which is consistent with the conclusion that the cogging torque is reduced by reducing the tooth harmonics. Afterwards, there is no obvious change in the simulation results of the electromotive force, that is, it has no effect on the power output of the motor.
[0059] It should be noted that the parts not described in detail in the above embodiments are all prior art.
[0060] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technician familiar with this profession can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention should be covered within the protection scope of the present invention.
Claims
1. A method for optimizing the cogging torque parameters of a permanent magnet synchronous motor, characterized in that: The steps include: Step 1: Establish the cogging torque analytical model of the permanent magnet synchronous motor. The cogging torque is expressed as: Where, Q represents the number of stator slots, L Fe represents the axial length of the armature core, μ0 represents the air permeability, R2 represents the stator inner diameter, R1 represents the rotor outer diameter, p represents the number of pole pairs, and n represents an integer. is also an integer, α represents the angle between the center line of a permanent magnet and the center line of a tooth of the stator core, G n represents the Fourier coefficient after Fourier expansion of the corresponding air gap permeability; Step 2: Based on the optimization goal of reducing the cogging torque and taking the analyzed cogging torque as the model, the improved particle swarm algorithm is used to perform multiple optimization iterations of the permanent magnet synchronous motor parameters, and finally the parameters and cogging torque of the optimization model are output; The particle swarm algorithm and genetic algorithm are combined into an optimization algorithm, and the iterative process is re-assigned; combined with the global search capability of the genetic algorithm, in the early stage of the algorithm, only the genetic algorithm iteration is performed, and under the action of the perturbation mechanism, the fluctuation of the iteration curve is increased to fully accumulate sample data; when the genetic algorithm iteration is completed, the best individuals that appear in each generation of iteration are collected and processed to generate a new solution domain and initial model, and then the particle swarm algorithm iteration is started, and local optimization is performed on the best individuals collected in the early stage.
2. A method for optimizing the cogging torque parameters of a permanent magnet synchronous motor according to claim 1, characterized in that: In step 2, a disturbance mechanism is introduced to intervene when the iteration curve is in a long-term unchanged state, so as to change the current iteration trend.
3. A method for optimizing the cogging torque parameters of a permanent magnet synchronous motor according to claim 1, characterized in that: In step 2, a sample closest to the optimal solution of the current group is selected from the data collected in the early stage to replace the sample and solve it again.
4. A method for optimizing the cogging torque parameters of a permanent magnet synchronous motor according to claim 1, characterized in that: In the perturbation mechanism in the early stage of the algorithm in step 2, in order to assign the group optimal solution to an inferior solution, the midpoint of the value range of each component of the sample is used as a reference, and the opposite component of the group optimal solution position component relative to this reference is calculated. The generated new sample is the assigned inferior solution. X m =(A-B) / 2 ΔX=|X gb -X m Among them, X op represents the position component of the inverse solution; X m represents the midpoint in the solution domain; X gb Represents the position component of the group's historical optimal solution; A and B represent the upper and lower limits of the solution domain respectively; △X represents the absolute value of the change between the inverse solution and the current optimal solution position variable.
5. The method for optimizing the cogging torque parameters of a permanent magnet synchronous motor according to claim 1, characterized in that: After the genetic algorithm completes the iteration in step 2, the sample data is collected and sorted according to the fitness size, and the component information of each sample is counted to determine the upper and lower limits of each component in all the information, and finally the new upper and lower limits are synchronously updated to the particle swarm algorithm.