Multi-disciplinary and multi-working-condition intelligent collaborative optimization design method for generator

Abstract

The invention discloses a multi-disciplinary and multi-working-condition intelligent collaborative optimization design method for a generator. The method comprises the following steps: determining initial geometric and physical model parameters of the generator; geometric parametric modeling is carried out on related parts of the generator; carrying out parametric modeling on material attributes, physical processes and operation conditions of the generator; establishing a generator model, preliminarily screening design parameters, and checking whether the generator meets feasibility constraints or not under the multi-constraint condition; determining an optimization target and an optimization space; selecting a sampling algorithm and a parameter optimization range, and constructing a test design sample library; selecting an intelligent optimization design method to find a global final solution meeting the design requirements, meeting the design requirements of multiple targets and multiple working conditions of the generator, and completing the optimization process; and determining various parameters and design schemes of the generator, and performing performance evaluation and verification. According to the invention, multi-parameter, multi-target, multi-working-condition and multi-disciplinary optimal design searching of the generator is realized.

Description

Technical Field

[0001] The present application relates to the technical field of multi-parameter coupling systems for generators, and in particular to a multi-disciplinary and multi-operating-condition intelligent collaborative optimization design method for generators. Background Art

[0002] With the continuous advancement of renewable energy power generation and consumption, motors, from the power generation side (such as wind turbines) to the power consumption side (such as electric vehicles), face increasingly stringent design requirements and significant challenges. Traditional design methods such as "large margins" and "analog design," or design concepts that only aim to create "feasible" motors, are no longer able to meet the competitive demands of the industry. Entering an era of comprehensive extreme design, motor design must continuously surpass previous design limits and achieve design requirements such as improved output performance, low cost, lightweight, and high reliability. Improving the overall performance of a generator requires optimizing its multi-parameter coupled system, which presents a complex design challenge encompassing multiple objectives, multiple operating conditions, and multiple disciplines. Traditional design methods face numerous design objectives, high design difficulty, high design costs, and long design cycles, posing significant challenges to current motor design. Therefore, efficiently, quickly, and accurately finding the optimal values ​​of this multi-parameter coupled system to achieve comprehensive improvements in motor design performance is a pressing issue.

[0003] In response to the above problems, many researchers have studied the multi-objective optimization design of motors, using various optimization algorithms and models to improve the design performance of motors. The invention patent with publication number CN115374571B discloses a multi-objective optimization method and system for an embedded double-layer tangential pole rare earth permanent magnet motor. The patent protects a method for optimizing a double-layer tangential pole permanent magnet motor. It mainly establishes a phase optimization objective equation by introducing a weight coefficient and proposes a path to seek optimal parameters. However, the motor involved in this method is quite different from the motor topology of this application, and the method does not mention multi-condition optimization. Its effect is not obvious when facing higher-dimensional optimization problems. The invention patent with publication number CN114757112A proposes a motor parameter design method and system based on the gray wolf algorithm. The method constructs a finite element model of the motor, determines the parameter set to be optimized of the motor finite element model and its associated optimization objective function set, and automatically optimizes the parameter set to be optimized based on the gray wolf algorithm. However, this method fails to simultaneously support the optimization space of multiple disciplines, fails to adequately account for the strong coupling of multiple physical fields in motors, and its engineering practicality needs to be improved. In summary, selecting the right strategy for intelligent optimization of multi-parameter coupled systems presents a significant challenge for motor design. Summary of the Invention

[0004] In view of this, the present application provides a multi-disciplinary and multi-working condition intelligent collaborative optimization design method for generators. Based on the electromagnetic, heat transfer and mechanical design theories of generators, it comprehensively considers the output performance requirements of the generators such as cost, weight and reliability, and combines a multi-objective genetic optimization design method to ultimately achieve a multi-objective, multi-working condition and multi-disciplinary intelligent optimization design of the generator.

[0005] This application discloses a multi-disciplinary and multi-operating-condition intelligent collaborative optimization design method for a generator, which includes: Step 1: According to the application background and design requirements of the generator, determine the generator initialization geometry and physical model parameters; Step 2: Perform geometric parametric modeling on the relevant components of the generator; the relevant components include the stator and rotor; Step 3: Parameterize the material properties, physical processes, and operating conditions of the generator for multi-objective optimization of its electromagnetic, thermal, and mechanical designs. Step 4: Establish a generator model, preliminarily screen the design parameters, and verify whether the generator meets the feasibility constraints under multiple constraints; Step 5: Comprehensively select the design optimization variables of the generator and determine the optimization target and optimization space; Step 6: Select appropriate sampling algorithm and parameter optimization range to build the experimental design sample library; Step 7: Select an intelligent optimization design method to find the final global solution that meets the design requirements, meet the multi-objective and multi-operating design requirements of the generator, and complete the optimization process; Step 8: Determine various parameters and design schemes of the generator, and conduct performance evaluation and verification.

[0006] Furthermore, in step 1, the different working conditions are any combination of the following items: Generator light load condition, generator rated condition, generator overload condition, generator fault condition.

[0007] Furthermore, the step 2 includes: The relevant components of the generator are geometrically parameterized and modeled using simulation software; the simulation software includes CAE software platforms; CAE software platforms include JMAG, FLUX, and ANSYS.

[0008] Furthermore, the step 3 includes: Based on the application background and design requirements of the generator, the multi-objective optimization design of the generator's electromagnetic, heat transfer and mechanical aspects is determined. The multi-objectives include electromagnetic performance, heat transfer performance, mechanical performance, weight, cost, reliability and environmental protection.

[0009] Furthermore, the step 3 includes: Taking JMAG as an example, the generator parameter modeling methods include: Both JMAG-Designer and the geometry editor have the function of creating variables; parameter modeling is at the "Model level" or "Study level"; parameter variables are set as "value variables" or "expression variables"; Geometric parametric modeling is a two-dimensional or three-dimensional model, which is imported through third-party software or directly modeled in the geometry editor and parametrically associated; material property parametric modeling includes: soft magnetic materials, hard magnetic materials and conductor materials; physical process parametric modeling includes: motion effects and eddy current effects; operating condition parametric modeling includes: generator light load condition, generator rated condition, generator overload condition, generator fault condition.

[0010] Furthermore, the step 4 includes: The generator model consists of an optimization model and a parameterized model. The optimization model includes relevant requirements and uses post-processing scripts to set reasonable constraints. It can be combined with finite element methods to automatically model, calculate, process, evaluate, and control the generation of design solutions. Optimization algorithms can also be used to improve iteration speed and optimization efficiency. Relevant requirements include design parameters, key performance, and cost. A two-dimensional parametric model of the generator was constructed using JMAG software. Electromagnetic and heat transfer bidirectional coupling was set up based on the magnetic permeability and electrical conductivity of the generator material. The generator parameters to be designed were preliminarily screened. The established software model was then used to calculate whether the generator's output performance met feasibility constraints under multiple design parameters and constraints. These constraints included electromagnetic, heat transfer, mechanical, output performance, cost, weight, and reliability constraints. Electromagnetic constraints refer to the generator meeting electromagnetic requirements including torque quality, electromagnetic power, electrical loss, harmonic content, current, and magnetic flux density; heat transfer constraints refer to the generator meeting heat transfer requirements including insulation level, temperature rise limit, operating environment and cooling medium; mechanical constraints refer to the generator meeting mechanical requirements including mechanical strength, dynamic characteristics, fatigue life and NVH characteristics; output performance constraints refer to the generator needing to meet all output performance requirements in the design requirements; cost constraints refer to the generator meeting cost requirements including material cost, manufacturing cost, maintenance cost and risk cost; weight constraints refer to the generator meeting the maximum and minimum weight constraints in the design requirements; reliability constraints refer to the generator meeting reliability constraints including stable output performance during operation and operating life.

[0011] Furthermore, the step 5 includes: Based on the output performance under multiple constraints, the optimization variables of the generator design are determined. The optimization variables specifically include the geometric size parameters, material parameters and operating condition parameters of the generator. The optimization target and optimization space are determined based on the application background and design requirements of the generator. The optimization target refers to the goal that the generator can achieve through optimization, including electromagnetic performance, heat transfer performance, mechanical performance, weight, cost, reliability and environmental protection; the optimization space refers to the electromagnetic space, heat transfer space and mechanical space.

[0012] Furthermore, the step 6 includes: Determine the feasibility range and value range of various design optimization variable parameters of the generator; use sampling algorithms to construct an experimental design sample library, the sampling algorithms include uniform Latin hypercube algorithm (ULH), Bayesian sampling, basic sampling method, rejection sampling, importance sampling, Metropolis sampling algorithm, Metropolis-Hasting sampling algorithm, sampling method and semi-random sampling method; the experimental design sample library refers to the different values ​​of various design parameters obtained by the above sampling algorithms, and the different values ​​of various design parameters constitute the experimental design sample library.

[0013] Furthermore, the step 7 includes: Based on the generator design background and requirements, intelligent optimization design methods are selected to find the global optimal solution that meets the generator electromagnetic and heat transfer design requirements. Among them, intelligent optimization design methods include: particle swarm optimization (PSO), ant colony optimization (ACO), simulated annealing algorithm, artificial bee colony algorithm (ABC), artificial fish swarm algorithm (AFSA), shuffle frog leaping algorithm (SFLA), firework algorithm (FWA), bacterial foraging optimization (BFO), firefly algorithm (FA), evolutionary algorithm (EC), genetic algorithm (GA), fuzzy logic, swarm intelligence (SI) algorithm, artificial immune system (AIS), artificial neural network (ANN), response surface method (RSM), support vector machine (SVM) and gradient-based algorithm. Based on the intelligent optimization design method, while meeting relevant design requirements, the optimization target value of the n-th generation scheme is calculated through the established generator software model; it is judged whether the optimization result meets the expected value. If it does not meet the expected value, the n+1-th generation scheme is generated to continue the optimization. If it meets the expected value, the optimization process is completed; relevant design requirements include multiple objectives and multiple operating conditions.

[0014] Furthermore, the step 8 includes: Through multiple optimization calculations, the final solution is retained; the final solution is the optimized generator geometric size parameters, material parameters and operating condition parameters. Using the final solutions of various parameters, the generator processing plan is determined, and performance evaluation and verification are carried out. Among them, the performance evaluation includes: electromagnetic performance evaluation, heat transfer performance evaluation, mechanical performance evaluation, output performance evaluation, cost evaluation, weight evaluation and reliability evaluation.

[0015] Due to the adoption of the above technical solution, this application has the following advantages: Compared with traditional generator design methods, this method applies multi-objective optimization technology to the electromagnetic, heat transfer and mechanical design of generators. While satisfying constraints such as power, current, magnetic flux density, loss and harmonics, it solves the problem of excessively high optimization space dimension caused by the large number of optimization objectives and design parameters in the generator design process, and realizes the optimal design search for generators with multiple parameters, multiple objectives, multiple working conditions and multiple disciplines.

[0016] This method was used to optimize the design of a large wind turbine. A cost- and performance-optimization control strategy was developed, achieving an optimal balance between electromagnetic performance, structure, ventilation, and cost. This approach resulted in high overall efficiency, high torque density, and low losses for wind turbines operating under complex operating conditions. This optimization method, while maintaining the same load level and ensuring safety, achieved a 3.5% increase in torque, a 1.5% reduction in permanent magnet usage, an approximately 5% increase in overall torque density, and reduced torque fluctuation to less than 1%.

[0017] Based on the electromagnetic, heat transfer and mechanical design theories of generators, and taking into account the electromagnetic performance, heat transfer performance, mechanical performance, output performance, reliability, environmental protection, lightweight and cost requirements of generators, and combining multi-objective optimization design methods, the multi-objective, multi-working condition and multi-disciplinary intelligent collaborative optimization design of generators is finally achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments recorded in the embodiments of the present application. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0019] Figure 1 This is a flow chart of a multi-disciplinary and multi-operating-condition intelligent collaborative optimization design method for a generator according to an embodiment of the present application; Figure 2 A schematic diagram of a two-dimensional parameterized model according to an embodiment of the present application; Figure 3 This is a schematic diagram of an optimization model according to an embodiment of the present application; Figure 4 Schematic diagram of the analysis results of the correlation between the design parameters and performance indicators of the embodiment of the present application; Figure 5 This is a schematic diagram of the calculation results of an embodiment of the present application; Figure 6 This is a cost impact curve diagram of an embodiment of the present application. DETAILED DESCRIPTION

[0020] The present application is further described with reference to the accompanying drawings and embodiments. The embodiments described are only a part of the embodiments of the present application, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field should fall within the scope of protection of the embodiments of the present application.

[0021] See also Figure 1 The present application provides an embodiment of a multi-disciplinary and multi-operating-condition intelligent collaborative optimization design method for a generator, which includes: S1: According to the application background and design requirements of the generator, establish the generator initialization geometry and physical model parameters; S2: Perform geometric parameter modeling on the relevant components of the generator; the relevant components include the stator and rotor; S3: Targeting multi-objective optimization of the generator’s electromagnetic, heat transfer, and mechanical design, parameterized modeling of the generator’s material properties, physical processes, and operating conditions is performed. S4: Establish a generator model, preliminarily screen the design parameters, and verify whether the generator meets the feasibility constraints under multiple constraints; S5: Comprehensively select the design optimization variables of the generator and determine the optimization target and optimization space; S6: Select sampling algorithm and parameter optimization range to build experimental design sample library; S7: Select an intelligent optimization design method to find a global final solution that meets the design requirements, meet the design requirements of multiple objectives and multiple operating conditions of the generator, and complete the optimization process; S8: Determine various parameters and design schemes of the generator, and conduct performance evaluation and verification.

[0022] Optionally, in S1, the different working conditions are any combination of the following: Generator light load condition, generator rated condition, generator overload condition, generator fault condition.

[0023] Optionally, the S2 includes: The relevant components of the generator are geometrically parameterized and modeled using simulation software; the simulation software includes CAE software platforms; the CAE software platforms include JMAG, FLUX, and ANSYS.

[0024] Optionally, the S3 includes: Based on the application background and design requirements of the generator, the multi-objective optimization design of the generator's electromagnetic, heat transfer and mechanical aspects is determined. The multi-objectives include electromagnetic performance, heat transfer performance, mechanical performance, weight, cost, reliability and environmental protection.

[0025] Optionally, the S3 includes: Taking JMAG as an example, the generator parameter modeling methods include: Both JMAG-Designer and Geometry Editor have the function of creating variables, which can be used in both. Parametric modeling is done at the "Model level" or "Study level". Parametric variables are set as "value variables" or "expression variables". Geometric parametric modeling is a two-dimensional or three-dimensional model, which is imported through third-party software or directly modeled in the geometry editor and parametrically associated; material property parametric modeling includes: soft magnetic materials, hard magnetic materials and conductor materials, and its purpose is to serve as the basis for subsequent optimization design and simulation analysis; physical process parametric modeling includes: motion effects and eddy current effects, and its purpose is to serve as the basis for subsequent optimization design and simulation analysis; operating condition parametric modeling includes: generator light load condition, generator rated condition, generator overload condition, generator fault condition, and its purpose is to serve as the basis for subsequent optimization design and simulation analysis.

[0026] Optionally, the S4 includes: The generator model consists of an optimization model and a parameterized model. The optimization model includes design parameters, key performance, and cost requirements. Reasonable constraints are set using post-processing scripts. Finite element methods can be combined to automatically model, calculate, process, evaluate, and control the generation of design solutions. Optimization algorithms can also be used to improve iteration speed and optimization efficiency. A two-dimensional parametric model of the generator was constructed using JMAG software. Electromagnetic and heat transfer bidirectional coupling was set up based on the magnetic permeability and electrical conductivity of the generator material. The generator parameters to be designed were preliminarily screened. The established software model was then used to calculate whether the generator's output performance met feasibility constraints under multiple design parameters and constraints. These constraints included electromagnetic, heat transfer, mechanical, output performance, cost, weight, and reliability constraints. Electromagnetic constraints refer to the generator meeting electromagnetic requirements including torque quality, electromagnetic power, electrical loss, harmonic content, current, and magnetic flux density; heat transfer constraints refer to the generator meeting heat transfer requirements including insulation level, temperature rise limit, operating environment and cooling medium; mechanical constraints refer to the generator meeting mechanical requirements including mechanical strength, dynamic characteristics, fatigue life and NVH characteristics; output performance constraints refer to the generator needing to meet all output performance requirements in the design requirements; cost constraints refer to the generator meeting cost requirements including material cost, manufacturing cost, maintenance cost and risk cost; weight constraints refer to the generator meeting the maximum and minimum weight constraints in the design requirements; reliability constraints refer to the generator meeting reliability constraints including stable output performance during operation and operating life.

[0027] Optionally, the S5 includes: Based on the output performance under multiple constraints, the generator design optimization variables are determined. The optimization variables specifically include the generator's geometric size parameters, material parameters, and operating condition parameters. The purpose is to improve optimization efficiency and quality. According to the application background and design requirements of the generator, the optimization target and optimization space are determined; among them, the optimization target refers to the goal that the generator can achieve through optimization, including electromagnetic performance, heat transfer performance, mechanical performance, weight, cost, reliability and environmental protection, and its purpose is to achieve the optimal design effect; the optimization space refers to the electromagnetic space, heat transfer space and mechanical space, and its purpose is to achieve the optimal design effect.

[0028] Optionally, the S6 includes: Determine the feasibility range and value range of various design optimization variable parameters of the generator; use sampling algorithms to construct an experimental design sample library, the sampling algorithms include uniform Latin hypercube algorithm (ULH), Bayesian sampling, basic sampling method, rejection sampling, importance sampling, Metropolis sampling algorithm, Metropolis-Hasting sampling algorithm, sampling method and semi-random sampling method, the purpose of which is to ensure the quality of the sample library by improving sample indicators; the experimental design sample library refers to the different values ​​of various design parameters obtained by the above sampling algorithms, and the different values ​​of various design parameters constitute the experimental design sample library.

[0029] Optionally, the S7 includes: Based on the generator design background and design requirements, intelligent optimization design methods are selected to find the global optimal solution that meets the electromagnetic and heat transfer design requirements of the generator. Among them, intelligent optimization design methods include: particle swarm optimization (PSO), ant colony optimization (ACO), simulated annealing algorithm, artificial bee colony algorithm (ABC), artificial fish swarm algorithm (AFSA), shuffle frog leaping algorithm (SFLA), firework algorithm (FWA), bacterial foraging optimization (BFO), firefly algorithm (FA), evolutionary algorithm (EC), genetic algorithm (GA), fuzzy logic, swarm intelligence (SI) algorithm, artificial immune system (AIS), artificial neural network (ANN), response surface method (RSM), support vector machine (SVM) and gradient-based algorithm. The above purposes are to use various algorithms or algorithm combinations to efficiently find the global optimal solution, achieve the exploration and discovery of the global sample space, improve the screening speed and efficiency, obtain high-precision calculation results with the minimum computational cost, and complete the optimal design of the motor.

[0030] Based on the intelligent optimization design method, while meeting relevant design requirements, the optimization target value of the n-th generation scheme is calculated through the established generator software model; it is judged whether the optimization result meets the expected value. If it does not meet the expected value, the n+1-th generation scheme is generated to continue the optimization. If it meets the expected value, the optimization process is completed; relevant design requirements include multiple objectives and multiple operating conditions.

[0031] Optionally, the S8 includes: After multiple optimization calculations, the final solution is retained; the final solution is the optimized generator geometric size parameters, material parameters and operating condition parameters. Using the final solutions of various parameters, the generator processing plan is determined, and performance evaluation and verification are carried out. Among them, the performance evaluation includes: electromagnetic performance evaluation, heat transfer performance evaluation, mechanical performance evaluation, output performance evaluation, cost evaluation, weight evaluation and reliability evaluation, and its purpose is to verify the optimal design result.

[0032] The present application discloses a multi-disciplinary and multi-working condition intelligent collaborative optimization design method for a generator for a multi-disciplinary coupling system of a generator. The method is based on the electromagnetic, heat transfer and mechanical design theories of the generator, comprehensively considers the electromagnetic performance, heat transfer performance, mechanical performance, output performance, reliability, environmental protection, lightweight and cost requirements of the generator, and combines the multi-objective optimization design method to ultimately achieve the multi-objective, multi-working condition and multi-disciplinary intelligent collaborative optimization design of the generator.

[0033] Compared with traditional generator design methods, this method applies multi-objective optimization technology to the electromagnetic, heat transfer and mechanical design of generators. While meeting the constraints of power, current, magnetic flux density, loss and harmonics, it solves the problem of excessively high optimization space dimension caused by the large number of optimization objectives and design parameters in the generator design process, and realizes the optimal design search for generators with multiple parameters, multiple objectives, multiple working conditions and multiple disciplines.

[0034] For ease of understanding, this application provides a more specific embodiment: An 18 MW semi-direct drive wind turbine is used as an example for detailed description. The steps include: In step 1, the initial geometric and physical model parameters of the generator are established based on the application background and design requirements of the generator. This embodiment is based on electromagnetic, heat transfer, and mechanical design theories of the generator and empirical parameters of wind turbines to determine the initial geometric and physical model parameters of the generator.

[0035] In step 2, the generator geometry is parametrically modeled. In this embodiment, parametric geometric models of the generator's stator and rotor have been established. Model parameterization involves representing parameters such as the motor's geometry, size, topology, and operating conditions as a series of variables, which are then used in a multi-objective optimization model. Parametric representation of motor parameters as a mathematical model facilitates subsequent optimization design and simulation analysis. The parametric modeling process includes the following steps: Step 201: Select parameterized variables. Motor parameterization includes geometric shape parameterization, material property parameterization, and operating condition parameterization. Appropriate parameterized variables need to be selected based on the specific motor structure and optimization objectives.

[0036] Step 202: Determine the parameterized variable range. Determine the feasibility range and value range of other parameters such as the motor's geometric dimensions, material parameters, and operating condition parameters to ensure the feasibility and effectiveness of the optimization.

[0037] Step 203: Establish a parametric model. Use the set motor parametric variables for parametric geometry modeling, material settings, and boundary condition settings. In post-processing, performance results can also be set as parameters and exported through scripts.

[0038] In step 3, the generator's material properties, physical processes, and operating conditions are parametrically modeled. The motor's structural parameters were optimized using a parametric approach in step 2. This model is applicable to the optimized design of various motors. Based on this model, the various material properties within the generator, such as soft magnetic materials, hard magnetic materials, and conductor materials, are modeled; various physical processes within the generator, such as motion effects and eddy current effects, are modeled; and various operating conditions within the generator, such as light load conditions, rated conditions, overload conditions, and fault conditions, are modeled.

[0039] In step 4, a generator model is built, and design parameters are preliminarily screened to verify whether the generator satisfies feasibility constraints under multiple constraints. Design parameters a, b, c, …, n are selected, and the designed generator is verified to meet electromagnetic constraints, heat transfer constraints, mechanical constraints, output performance constraints, cost constraints, weight constraints, and reliability constraints under various constraints.

[0040] The generator model consists of two parts: parameterized model and optimization model. Figure 2 This is a schematic diagram of the two-dimensional parametric model of this application. Figure 3 This is a schematic diagram of the optimization model for an embodiment of this application. The optimization model includes design parameters, key performance, and cost requirements. A post-processing script is used to set reasonable constraints. This allows for automated modeling, calculation, processing, evaluation, and control of design solution generation using finite element methods. Optimization algorithms can also be used to improve iteration speed and optimization efficiency.

[0041] In step 5, the design optimization variables for the generator are selected, and the optimization target and optimization space are determined. The selection of optimization variables affects the optimization efficiency and optimization results of the motor. The selection range of design optimization variables includes geometric parameters and operating condition parameters. For example, the inner and outer diameters of the motor stator and rotor, air gap length, core length, number of slots in the rotor and stator, slot shape, width and thickness of the magnetic steel, etc. are geometric parameters. The selection of these parameters directly affects the electromagnetic performance, mechanical properties, and effective material usage of the motor. The operating condition parameters of the motor include the amplitude and phase angle of the current, speed, and operating voltage. The selection of these parameters will affect the power, efficiency, and temperature of the motor.

[0042] When selecting motor design optimization variables, it is necessary not only to comprehensively consider the above aspects, but also to consider the mutual influence between the optimization variables to avoid conflicts and contradictions between the optimization variables, so as to achieve the best effect of motor optimization design.

[0043] When selecting motor design optimization variables, it is necessary not only to comprehensively consider the above aspects, but also to consider the mutual influence between the optimization variables to avoid conflicts and contradictions between the optimization variables, so as to achieve the best effect of motor optimization design.

[0044] In step 6, a sampling algorithm and parameter optimization range are selected to construct an experimental design sample library. In engineering design optimization, a series of tests are required in which the input variable values ​​vary within a given range to analyze the impact on the output response. When optimizing a motor, it is necessary to perform initial sampling of areas of interest in the design space, such as electromagnetic space, heat transfer space, and mechanical space. Design of Experiments (DOE) is a technique used to effectively guide experimental selection. Using data points created with a suitable DOE can make the input parameter configuration well distributed, thereby reducing the correlation between inputs. There are many sampling methods. The embodiment of this application uses the ULH (Uniform Latin Hypercube) algorithm as an example for illustration, and it is not intended to limit this application. ULH is an advanced constrained Monte Carlo sampling that can generate random numbers that conform to a uniform distribution.

[0045] In step 7, an appropriate intelligent optimization design method is selected to find the global optimal solution that meets the design requirements. For example, a deterministic algorithm is used to refine the Pareto optimal set to meet the multi-objective and multi-operating condition design requirements of the generator, completing the optimization process. Multi-objective design requirements for the motor include, but are not limited to, electromagnetic performance, heat transfer performance, mechanical performance, output performance, reliability, environmental friendliness, lightweight, and cost. Multi-operating conditions for the generator include, but are not limited to, light-load conditions, rated conditions, overload conditions, fault conditions, and other conditions.

[0046] The multi-objective genetic algorithm (MOGA-II) uses intelligent and efficient multi-search elitism, which can retain excellent solutions and prevent premature convergence to local optima. Elitism improves the convergence of the algorithm and ensures that the fitness of the new generation is greater than that of the parent generation. In the problem of generator optimization design, the embodiment of this application will use ULH to perform DOE experimental design, and then select MOGAII for multi-objective genetic algorithm optimization to obtain the optimization value target of the solution of the nth generation. If the result does not meet the expected value, the solution of the n+1th generation will be generated for further optimization. The embodiment of this application uses a deterministic algorithm (such as a gradient-based algorithm) to refine the Pareto optimal set as an example for illustration, and is not intended to limit this application.

[0047] Figure 4 The results of the correlation analysis between the design parameters and performance indicators of the embodiment of this application are as follows: Figure 5 This is a distribution diagram of the calculation results of an embodiment of the present application, where X is cost, Y is power, color is power factor, and size is core length.

[0048] Figure 6 This is a cost impact curve for an embodiment of this application, defining a lower power limit and an upper current limit. Each line of the broken line graph represents a design solution. Using the cost as the color scale on the far right, we can observe the impact of design parameters on cost: larger diameters and shorter cores result in lower costs; smaller air gaps also result in lower costs; yoke height, magnet angle, and pole arc radius have little impact on cost; thinner pole shoes result in lower costs; and thinner magnets also result in lower costs.

[0049] In step 8, the optimized design solution is determined, and performance evaluation and design verification are conducted. Based on the above optimization results, the motor can be optimized as follows: appropriately increase the diameter and shorten the core to reduce cost; appropriately reduce the air gap to reduce cost; the yoke height, magnet angle, and pole arc radius have little impact on cost, and the optimal selection can be made based on harmonics and performance; appropriately thin the pole shoes to reduce cost, but pay attention to the increase in rotor parasitic losses; appropriately thin the magnets to reduce cost, but balance the demagnetization capacity.

[0050] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and not to limit them. Although the present application has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present application can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present application should be included in the scope of protection of the claims of the present application.

Claims

1. A multi-disciplinary and multi-operating-condition intelligent collaborative optimization design method for generators, characterized by: include: Step 1: According to the application background and design requirements of the generator, determine the generator initialization geometry and physical model parameters; Step 2: Perform geometric parametric modeling on the relevant components of the generator; the relevant components include the stator and rotor; Step 3: Parameterize the material properties, physical processes, and operating conditions of the generator for multi-objective optimization of its electromagnetic, thermal, and mechanical designs. Step 4: Establish a generator model, preliminarily screen the design parameters, and verify whether the generator meets the feasibility constraints under multiple constraints; Step 5: Comprehensively select the design optimization variables of the generator and determine the optimization target and optimization space; Step 6: Select the sampling algorithm and parameter optimization range to build the experimental design sample library; Step 7: Select an intelligent optimization design method to find the final global solution that meets the design requirements, meet the multi-objective and multi-operating design requirements of the generator, and complete the optimization process; Step 8: Determine various parameters and design schemes of the generator, and conduct performance evaluation and verification.

2. The method according to claim 1, characterized in that In step 1, different working conditions are any combination of the following items: Generator light load condition, generator rated condition, generator overload condition, generator fault condition.

3. The method according to claim 1, characterized in that The step 2 includes: The relevant components of the generator are geometrically parameterized and modeled using simulation software; the simulation software includes CAE software platforms; the CAE software platforms include JMAG, FLUX, and ANSYS.

4. The method according to claim 1, wherein The step 3 includes: Based on the application background and design requirements of the generator, the multi-objective optimization design of the generator's electromagnetic, heat transfer and mechanical aspects is determined. The multi-objectives include electromagnetic performance, heat transfer performance, mechanical performance, weight, cost, reliability and environmental protection.

5. The method according to claim 1, wherein The step 3 includes: Taking JMAG as an example, the generator parameter modeling methods include: Both JMAG-Designer and the geometry editor have the function of creating variables; parameter modeling is at the "Model level" or "Study level"; parameter variables are set as "value variables" or "expression variables"; Geometric parametric modeling is a two-dimensional or three-dimensional model, which is imported through third-party software or directly modeled in the geometry editor and parametrically associated; material property parametric modeling includes: soft magnetic materials, hard magnetic materials and conductor materials; physical process parametric modeling includes: motion effects and eddy current effects; operating condition parametric modeling includes: generator light load condition, generator rated condition, generator overload condition, generator fault condition.

6. The method according to claim 1, characterized in that The step 4 comprises: The generator model consists of an optimization model and a parameterized model. The optimization model includes relevant requirements and uses post-processing scripts to set reasonable constraints. It can be combined with finite element methods to automatically model, calculate, process, evaluate, and control the generation of design solutions. Optimization algorithms can also be used to improve iteration speed and optimization efficiency. Relevant requirements include design parameters, key performance, and cost. A two-dimensional parametric model of the generator was constructed using JMAG software. Electromagnetic and heat transfer bidirectional coupling was set up based on the magnetic permeability and electrical conductivity of the generator material. The generator parameters to be designed were preliminarily screened. The established software model was then used to calculate whether the generator's output performance met feasibility constraints under multiple design parameters and constraints. These constraints included electromagnetic, heat transfer, mechanical, output performance, cost, weight, and reliability constraints. Electromagnetic constraints refer to the generator meeting electromagnetic requirements including torque quality, electromagnetic power, electrical loss, harmonic content, current, and magnetic flux density; heat transfer constraints refer to the generator meeting heat transfer requirements including insulation level, temperature rise limit, operating environment and cooling medium; mechanical constraints refer to the generator meeting mechanical requirements including mechanical strength, dynamic characteristics, fatigue life and NVH characteristics; output performance constraints refer to the generator needing to meet all output performance requirements in the design requirements; cost constraints refer to the generator meeting cost requirements including material cost, manufacturing cost, maintenance cost and risk cost; weight constraints refer to the generator meeting the maximum and minimum weight constraints in the design requirements; reliability constraints refer to the generator meeting reliability constraints including stable output performance during operation and operating life.

7. The method according to claim 1, characterized in that The step 5 comprises: Based on the output performance under multiple constraints, the optimization variables of the generator design are determined. The optimization variables specifically include the geometric size parameters, material parameters and operating condition parameters of the generator. The optimization target and optimization space are determined based on the application background and design requirements of the generator. The optimization target refers to the goal that the generator can achieve through optimization, including electromagnetic performance, heat transfer performance, mechanical performance, weight, cost, reliability and environmental protection; the optimization space refers to the electromagnetic space, heat transfer space and mechanical space.

8. The method according to claim 1, characterized in that The step 6 comprises: Determine the feasibility range and value range of various design optimization variable parameters of the generator; use sampling algorithms to construct an experimental design sample library, the sampling algorithms include uniform Latin hypercube algorithm, Bayesian sampling, basic sampling method, rejection sampling, importance sampling, Metropolis sampling algorithm, Metropolis-Hasting sampling algorithm, sampling method and semi-random sampling method; the experimental design sample library refers to the different values ​​of various design parameters obtained by the above sampling algorithms, and the different values ​​of various design parameters constitute the experimental design sample library.

9. The method according to claim 1, characterized in that The step 7 comprises: Based on the generator design background and requirements, intelligent optimization design methods are selected to find the global optimal solution that meets the generator's electromagnetic and heat transfer design requirements. These intelligent optimization design methods include: particle swarm optimization, ant colony optimization, simulated annealing, artificial bee colony optimization, artificial fish swarm optimization, shuffle frog leaping algorithm, fireworks algorithm, bacterial foraging optimization, firefly algorithm, evolutionary algorithm, genetic algorithm, fuzzy logic, swarm intelligence algorithm, artificial immune system, artificial neural network, support vector machine, response surface method, and gradient-based algorithm. Based on the intelligent optimization design method, while meeting relevant design requirements, the optimization target value of the n-th generation scheme is calculated through the established generator software model; it is judged whether the optimization result meets the expected value. If it does not meet the expected value, the n+1-th generation scheme is generated to continue the optimization. If it meets the expected value, the optimization process is completed; relevant design requirements include multiple objectives and multiple operating conditions.

10. The method according to claim 1, characterized in that The step 8 comprises: Through multiple optimization calculations, the final solution is retained; the final solution is the optimized generator geometric size parameters, material parameters and operating condition parameters. Using the final solutions of various parameters, the generator processing plan is determined, and performance evaluation and verification are carried out. Among them, the performance evaluation includes: electromagnetic performance evaluation, heat transfer performance evaluation, mechanical performance evaluation, output performance evaluation, cost evaluation, weight evaluation and reliability evaluation.

Images