A multi-physics coupling design method for shaft-clamped synchronous generators

Abstract

The present application discloses a design method for a multi-physics field coupled shaft-clamped synchronous generator, which relates to the field of shaft-clamped synchronous generators. The method performs discrete sampling within the parameter range of each design parameter of the shaft-clamped synchronous generator to construct multiple sampling combinations. Then, through coupling analysis of the electromagnetic field, structural mechanics field and thermal field, considering the coupling correlation of electromagnetic force, vibration and heat conduction in each physical field, a multi-physics field integrated design comprehensive optimization framework is formed through parametric design to calculate the electromagnetic weight, temperature rise reference value and radial vibration response under each sampling combination. Then, through the Kriging prediction model of these three motor performance parameters, multi-objective optimization is performed based on the optimal comprehensive performance obtained by weighted calculation on the basis of satisfying the constraints, and the performance balance and comprehensive improvement are achieved in terms of vibration control, electromagnetic weight and temperature rise management, thereby ensuring the efficient and stable operation of the shaft-clamped synchronous generator.

Description

Technical Field

[0001] The present application relates to the field of shaft-clamped synchronous generators, and in particular to a design method for shaft-clamped synchronous generators with multi-physics field coupling. Background Art

[0002] In the ship's power system, the shaft-mounted synchronous generator makes the system structure compact by tightly integrating the generator on the ship's propulsion shaft, greatly saving the ship's internal space while reducing complex mechanical connections. Therefore, its application is of extremely important practical significance.

[0003] With the rise of smart ships and green shipping concepts, research on shaft-mounted synchronous generators is also moving towards integration with intelligent control systems. This effort aims to achieve dynamic adjustment and energy-saving management of power generation systems through advanced algorithms and optimized designs. Shipbuilding companies and research institutions are also gradually beginning to focus on the application of shaft-mounted generator technology. Currently, research on shaft-mounted generator power generation technology focuses on optimizing electromagnetic performance, structural vibration, and heat dissipation. Improved electromagnetic performance ensures efficient operation of shaft-mounted synchronous generators under low-speed and variable-speed conditions, significantly improving the stability of the ship's power system. Structural vibration optimization reduces energy loss and equipment wear caused by mechanical vibration on ships, extending the generator's service life. Temperature rise control and optimized heat dissipation ensure reliable operation of shaft-mounted synchronous generators at high power densities, thereby reducing maintenance requirements and energy loss.

[0004] However, the electromagnetic performance, structural vibration performance and heat dissipation performance of the shaft-clamped synchronous generator will affect each other. Under this complex coupling relationship, the current practice is generally to optimize one of the performances, but this approach often leads to a decline in the performance of the other two, affecting the overall operating effect of the shaft-clamped synchronous generator. Summary of the Invention

[0005] In response to the above-mentioned problems and technical requirements, this application proposes a design method for a shaft-clamped synchronous generator with multi-physics field coupling. The technical solution of this application is as follows:

[0006] A multi-physics field coupled shaft-clamped synchronous generator design method, the shaft-clamped synchronous generator design method comprising:

[0007] Discrete sampling is performed within the parameter range of each design parameter of the shaft-clamped synchronous generator to construct multiple sampling combinations, each sampling combination includes K sampling values ​​of different design parameters, and the integer parameter K is greater than or equal to 2;

[0008] For each sampling combination, the electromagnetic weight W of the shaft-clamped synchronous generator under the sampling combination is determined. The electromagnetic power loss P under the sampling combination is determined based on the electromagnetic field model of the shaft-clamped synchronous generator. The electromagnetic power loss P is input into the temperature field as a heat source, and the temperature rise reference value ΔT considering the electromagnetic characteristics under the sampling combination is calculated using the thermal resistance method. The electromagnetic weight W and the temperature rise reference value ΔT are substituted into the system vibration equation and solved to obtain the radial vibration response x considering the temperature rise characteristics and the electromagnetic weight W under the sampling combination. radial (t);

[0009] The electromagnetic weight W of each sampling combination is used as output and the sampling combination is used as input to establish a convergent electromagnetic weight Kriging prediction model; the temperature rise reference value ΔT of each sampling combination is used as output and the sampling combination is used as input to establish a convergent temperature rise reference value Kriging prediction model; the radial vibration response x of each sampling combination is used as output and the sampling combination is used as input to establish a convergent temperature rise reference value Kriging prediction model. radial (t) establishing a convergent radial vibration response Kriging prediction model with the sample combination as output and input;

[0010] An optimization algorithm is used to optimize the comprehensive performance of the shaft-bracketed synchronous generator under each design parameter combination, and a design parameter combination that optimizes the comprehensive performance of the shaft-bracketed synchronous generator is obtained; wherein, for each design parameter combination, an electromagnetic weight prediction value is obtained based on the design parameter combination using an electromagnetic weight Kriging prediction model, a temperature rise prediction value is obtained based on the design parameter combination using a temperature rise reference value Kriging prediction model, and a radial vibration response prediction value is obtained based on the design parameter combination using a radial vibration response Kriging prediction model. The comprehensive performance of the shaft-bracketed synchronous generator under the design parameter combination is obtained by weighting the electromagnetic weight prediction value, the temperature rise prediction value, and the radial vibration response prediction value. The smaller the electromagnetic weight prediction value, the smaller the temperature rise prediction value, and the smaller the radial vibration response prediction value of the shaft-bracketed synchronous generator, the better the comprehensive performance of the shaft-bracketed synchronous generator.

[0011] A further technical solution is to use the thermal resistance method to calculate the temperature rise reference value ΔT considering the electromagnetic characteristics under the sampling combination:

[0012] ΔT=P×R total

[0013] Among them, the total thermal resistance R total =R1+R2+R3+R4, R1 is the thermal resistance of the winding, R2 is the contact thermal resistance between the winding and the iron core, R3 is the stator thermal resistance, and R4 is the convection heat dissipation resistance on the stator surface.

[0014] Its further technical solution is:

[0015] Where L is the effective length of the core, k efis the thermal conductivity of the core material, A1 is the cross-sectional area of ​​the stator winding; d is the average gap between the stator winding and the core, k2 is the thermal conductivity of the contact material between the stator winding and the core, A2 is the contact area between the stator winding and the core; D so is the stator outer diameter, D si is the stator inner diameter, k s is the thermal conductivity of the stator material; h is the convective heat transfer coefficient.

[0016] A further technical solution is to obtain the radial vibration response taking into account the temperature rise characteristics and electromagnetic weight under the sampling combination, including:

[0017] The system mass matrix M is constructed based on the electromagnetic weight, and the material stiffness matrix K(T)=K0(1-α T ×ΔT), where α T is the influence coefficient, K0 is the initial material stiffness matrix;

[0018] Substitute the material stiffness matrix K(T) and the system mass matrix M into the system vibration equation The radial vibration response x considering the temperature rise characteristics and electromagnetic weight is obtained by the modal superposition method. radial (t); where is x radial The first derivative of (t), is x radial (t), C is the system damping matrix; F radial (t) is the radial electromagnetic vibration force of the generator and is calculated based on the electromagnetic field parameters of the electromagnetic field model of the shaft-clamped synchronous generator.

[0019] A further technical solution is that the radial electromagnetic vibration force of the generator is calculated based on the electromagnetic field parameters of the electromagnetic field model of the shaft-holding synchronous generator:

[0020]

[0021] Where μ0 is the vacuum permeability, B is the magnetic induction intensity of the electromagnetic field and I is the motor output current, A radial is the effective bearing area of ​​the shaft-clamped synchronous generator and A radial =2πrL; N s is the number of stator winding turns, g is the air gap length, r is the average radius of the air gap, and L is the effective length of the core.

[0022] A further technical solution is to establish a Kriging prediction model for any motor performance parameter among electromagnetic weight, temperature rise reference value and radial vibration response, including:

[0023] Constructing an interpolation model with unknown model parameters Among them, Y(X) represents the motor performance parameters, X i represents the value of the i-th design parameter, β i is the interpolation coefficient corresponding to the i-th design parameter, β0 is the zero-order interpolation coefficient, Z(X) is the random process term, and the unknown model parameters include β0, β i and Z(X);

[0024] Substitute the sampled value of the i-th design parameter in each sampling combination into the X of the interpolation model i The motor performance parameters corresponding to the sampling combination are substituted into Y(X) and solved simultaneously to obtain the model parameters in the interpolation model, including β0, Z(X) and the interpolation coefficients β1~β corresponding to various design parameters. K ;

[0025] The model parameters are substituted into the interpolation model and the convergence of the interpolation model is judged. When the interpolation model does not meet the convergence conditions, the sampling combination is increased and the interpolation model is refitted until the interpolation model meets the convergence conditions. The Kriging prediction model of the motor performance parameters is established.

[0026] A further technical solution is to judge the convergence of the interpolation model including:

[0027] The sampling values ​​of each design parameter in each sampling combination are respectively substituted into the interpolation model with known model parameters to obtain the predicted values ​​of the motor performance parameters of the sampling combination. When the relative error of the predicted values ​​of the motor performance parameters obtained twice in a row for the same sampling combination is less than the error threshold, it is determined that the interpolation model converges at the sampling combination; when the interpolation model converges at all K sampling combinations, it is determined that the interpolation model meets the convergence condition; otherwise, it is determined that the interpolation model does not meet the convergence condition.

[0028] A further technical solution is to increase the sampling combination and refit the interpolation model when the interpolation model does not meet the convergence condition, including:

[0029] When the interpolation model does not converge at at least one sampling combination X′, discrete sampling points are added within a predetermined range of sampling values ​​of at least one design parameter in the sampling combination X′ to add a new sampling combination at the sampling combination X′. The number of newly added sampling combinations is in, is the predicted value of the motor performance parameter obtained by substituting the sampling combination X′ into the interpolation model with known model parameters, Y(X′) is the motor performance parameter of the sampling combination X′, and η is the design threshold.

[0030] A further technical solution is to obtain the comprehensive performance Θ of the shaft-clamped synchronous generator under the design parameter combination based on the weighted electromagnetic weight prediction value, temperature rise prediction value, and radial vibration response prediction value:

[0031]

[0032] in, is the electromagnetic weight prediction value, is the predicted value of radial vibration response, is the predicted value of temperature rise, λ is a positive penalty coefficient, ΔT max is the temperature rise threshold, γ1, γ2, and γ3 are all weighted parameters and γ1+γ2+γ3=1.

[0033] Its further technical solution is to use an optimization algorithm to seek the best result based on the comprehensive performance of the shaft-clamped synchronous generator under each design parameter combination, and the optimization algorithm used is a multi-objective optimization genetic algorithm NSGA-Ⅱ.

[0034] The beneficial technical effects of this application are:

[0035] The present application discloses a design method for a shaft-clamped synchronous generator with multi-physics field coupling. The method analyzes the coupling of electromagnetic field, structural mechanics field and thermal field, considers the coupling correlation of electromagnetic force, vibration and heat conduction in each physical field, and forms a multi-physics field integrated design comprehensive optimization framework through parametric design. By establishing a Kriging prediction model of vibration, electromagnetic weight and temperature rise limit of the shaft-clamped synchronous generator, multi-objective optimization is performed based on the optimal comprehensive performance obtained by weighted calculation on the basis of satisfying the constraints, and performance balance and improvement are achieved in vibration control, electromagnetic weight and temperature rise management. The reduction of electromagnetic weight helps to improve the system power density, the control of radial vibration enhances mechanical stability, and the temperature rise limit ensures the heat dissipation performance of the generator under high load, thereby effectively ensuring the efficient and stable operation of the generator under different working conditions.

[0036] In marine applications, reducing electromagnetic weight allows generators to achieve higher power output within a limited space, providing higher power density for the overall propulsion system. Furthermore, reduced electromagnetic weight directly contributes to reduced fuel consumption and emissions, thus meeting the requirements for energy conservation and emission reduction for ships. However, radial vibration control primarily affects the mechanical stability and service life of the generator. High-frequency vibration can cause mechanical problems such as component fatigue and loose bolts, affecting the normal operation of the generator. This application significantly improves the mechanical stability of the generator by reducing radial vibration without affecting power output. This application achieves reasonable limits on stator winding temperature rise through coupled optimization of heat source analysis and temperature fields, thereby avoiding damage to insulation materials and generator components caused by excessive temperature rise. Through temperature-limiting design, the generator can maintain performance while extending its service life, improving safety and heat dissipation efficiency under high-load conditions. By optimizing electromagnetic weight, vibration, and temperature rise, the shaft-mounted synchronous generator achieves high efficiency, stability, and durability within a limited space, meeting the reliability, energy efficiency, and space utilization requirements of modern marine systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a flow chart of a method for designing a shaft-clamped synchronous generator according to an embodiment of the present application.

[0038] Figure 2 It is a flowchart of calculating the electromagnetic weight, temperature rise reference value and radial vibration response of each sampling combination in one embodiment of the present application.

[0039] Figure 3 It is a flow chart of a Kriging prediction model for obtaining each motor performance parameter in one embodiment of the present application. DETAILED DESCRIPTION

[0040] The specific implementation of this application will be further described below with reference to the accompanying drawings.

[0041] This application discloses a multi-physics coupling shaft-clamping synchronous generator design method, please refer to Figure 1 The method includes the following steps:

[0042] Step 1: discrete sampling is performed within the parameter range of each design parameter of the shaft-clamped synchronous generator to construct a plurality of sampling combinations.

[0043] Each constructed sampling combination includes sampling values ​​of K different design parameters, where the integer parameter K ≥ 2. In one embodiment, random sampling is performed using Latin hypercube sampling, and in order to improve the accuracy of the Kriging prediction model in subsequent testing, the number of sampling times should be no less than 200.

[0044] The design parameters included in each sampling combination include but are not limited to: stator outer diameter D so 、Stator inner diameter D si , stator winding turns N s , stator winding cross-sectional area A1, iron core effective length L, air gap length g, average air gap radius r, stator material, iron core material, rotor pole width, pole depth, rotor pole shoe, rotor pole shoe depth, number of excitation winding turns, and shaft outer diameter, etc. Because a large number of parameter types will result in a relatively large subsequent optimization target, a sensitivity analysis can also be performed on the above parameter types. Parameter types with high sensitivity to electromagnetic performance can be selected as optimization targets, while the remaining parameter types use pre-set fixed values.

[0045] Step 2: For each sampling combination, calculate the electromagnetic weight W, temperature rise reference value ΔT and radial vibration response x of the shaft-clamped synchronous generator under the sampling combination. radial (t), please combine Figure 2 The process diagram shown is:

[0046] 1. Electromagnetic weight W: The electromagnetic weight under the sampling combination can be determined based on the electromagnetic field model of the shaft-clamped synchronous generator, and this application will not elaborate on this.

[0047] 2. Temperature rise reference value ΔT

[0048] This application uses electromagnetic power loss and thermal resistance to represent the temperature rise of the generator stator winding. On the one hand, it enhances the accuracy of the coupled calculation between electromagnetic loss and temperature rise, and on the other hand, it simplifies the complexity of temperature calculation. It includes: first, based on the electromagnetic field model of the shaft-clamped synchronous generator, the electromagnetic power loss P is determined under the sampling combination, and then the electromagnetic power loss P is input into the temperature field as a heat source, and the temperature rise reference value ΔT considering the electromagnetic characteristics under the sampling combination is calculated using the thermal resistance method. During the operation of the generator, significant resistive heat loss will be generated when the stator current passes through, and among all the heat losses, the heat loss of the generator stator winding accounts for the highest proportion, which is the main source of the total heat loss of the generator. Therefore, this application mainly considers the temperature rise reference value of the generator stator winding.

[0049] The method of establishing an electromagnetic field model to obtain electromagnetic power loss P is a relatively common method, and this application will not elaborate on it. After obtaining the electromagnetic power loss P, the temperature rise reference value ΔT can be calculated using the thermal resistance method according to the following formula:

[0050] ΔT=P×R total

[0051] Among them, the total thermal resistance R total=R1+R2+R3+R4, R1 is the winding thermal resistance, R2 is the contact thermal resistance between the winding and the core, R3 is the stator thermal resistance, and R4 is the convection heat dissipation resistance on the stator surface. The calculation methods are as follows:

[0052] (1) Winding thermal resistance R1 represents the thermal resistance of the stator winding, and the calculation formula is:

[0053]

[0054] Among them, k ef is the thermal conductivity of the core material, which can be determined based on the type of core material.

[0055] (2) The contact thermal resistance R2 between the winding and the core depends on the contact condition between the winding and the core. The calculation formula is:

[0056]

[0057] Where d is the average gap of the contact interface between the stator winding and the iron core, A2 is the contact area between the stator winding and the iron core, and k2 is the thermal conductivity of the contact material between the stator winding and the iron core.

[0058] (3) Stator thermal resistance R3 represents the radial thermal resistance from the core to the stator housing. The calculation formula is:

[0059]

[0060] Among them, k s is the thermal conductivity of the stator material, which can be determined based on the type of stator material.

[0061] (4) The convection heat dissipation resistance R4 on the stator surface is calculated based on the convection heat dissipation on the stator surface. The calculation formula is:

[0062]

[0063] Where h is the convective heat transfer coefficient.

[0064] 3. Radial vibration response x radial (t)

[0065] The electromagnetic weight will affect the vibration of the generator, and the temperature rise change due to the stiffness and vibration characteristics of the material will also affect the vibration of the generator. Therefore, the electromagnetic weight W and the temperature rise reference value ΔT are substituted into the system vibration equation to solve it, and the radial vibration response x considering the temperature rise characteristics and electromagnetic weight W under the sampling combination is obtained. radial (t).

[0066] The vibration equation of the system is as follows:

[0067]

[0068] Where M is the system mass matrix, C is the system damping matrix, and K(T) is the material stiffness matrix. radial (t) is the radial vibration response, is x radial The first derivative of (t), is x radial The second derivative of (t), F radial (t) is the radial electromagnetic vibration force of the generator.

[0069] The system mass matrix M is affected by the electromagnetic weight W. The system mass matrix M can be constructed based on the electromagnetic weight W. The specific construction method refers to the existing method and is not described in detail in this application. In addition, the system damping matrix C can also be constructed by referring to the existing method.

[0070] The material stiffness matrix K(T) represents the material stiffness parameters. Traditionally, it is generally believed that material stiffness is not necessary. However, in fact, material stiffness changes with temperature. The influence of temperature on the elastic modulus of the material is reflected by the material stiffness matrix K(T). In this embodiment, considering the influence of the change of material stiffness with temperature, the material stiffness matrix under the influence of the temperature rise reference value ΔT is obtained as follows:

[0071] K(T)=K0(1-α T ×ΔT)

[0072] Among them, α T is the influence coefficient, K0 is the initial material stiffness matrix.

[0073] Generator radial electromagnetic vibration force F radial (t) can be calculated based on the electromagnetic field parameters of the electromagnetic field model of the shaft-clamped synchronous generator. The calculation formula is:

[0074]

[0075] Where μ0 is the vacuum permeability, B is the magnetic induction intensity of the electromagnetic field and I is the motor output current, A radial is the effective bearing area of ​​the shaft-clamped synchronous generator and A radial =2πrL; N s is the number of stator winding turns, g is the air gap length, r is the average radius of the air gap, and L is the effective length of the core.

[0076] The system mass matrix M, system damping matrix C, material stiffness matrix K(T) and generator radial electromagnetic vibration force F are obtained above. radial (t) is substituted into the system vibration equation and solved by the modal superposition method to obtain the radial vibration response x considering the temperature rise characteristics and electromagnetic weight. radial (t).

[0077] Step 3: Establish a convergent electromagnetic weight Kriging prediction model with the electromagnetic weight W of each sampling combination as output and the sampling combination as input. Establish a convergent temperature rise reference value Kriging prediction model with the temperature rise reference value ΔT of each sampling combination as output and the sampling combination as input. radial (t) is the output and the sampling combination is the input to establish a convergent radial vibration response Kriging prediction model.

[0078] The methods for establishing the three Kriging prediction models are similar. For any motor performance parameter among electromagnetic weight, temperature rise reference value and radial vibration response, establishing the Kriging prediction model of the motor performance parameter includes the following steps. Please refer to Figure 3 The flow chart shown:

[0079] (1) Construct an interpolation model containing unknown model parameters as follows:

[0080]

[0081] Among them, Y(X) represents the motor performance parameters. When Y(X) represents the electromagnetic weight, this method is used to establish the electromagnetic weight Kriging prediction model. When Y(X) represents the temperature rise reference value, this method is used to establish the temperature rise reference value Kriging prediction model. When Y(X) represents the radial vibration response, this method is used to establish the radial vibration response Kriging prediction model.

[0082] X i represents the value of the i-th design parameter in the parameter combination, β i is the interpolation coefficient corresponding to the i-th design parameter, and β0 is the zero-order interpolation coefficient. Z(X) is a random process term used to characterize the nonlinear association and spatial correlation between input variables, mainly capturing the nonlinear association caused by model complexity and unconsidered factors. The unknown model parameters in the interpolation model include β0, β i and Z(X).

[0083] (2) Substitute the sampled value of the i-th design parameter in each sampling combination into the X of the interpolation model i , the motor performance parameters corresponding to the sampling combination are substituted into Y(X), and then the equations under all sampling combinations are solved to obtain the model parameters in the interpolation model, including β0, Z(X) and the interpolation coefficients β1~β corresponding to each design parameter. K .

[0084] (3) Substitute the model parameters into the interpolation model and judge the convergence of the interpolation model, including:

[0085] For each sampling combination X, the sampled values ​​of each design parameter in the sampling combination X are respectively substituted into the interpolation model with known model parameters to obtain the predicted values ​​of the motor performance parameters of the sampling combination X. When the same sampling combination X is obtained twice in a row, the predicted value of the motor performance parameter When the relative error is less than the error threshold, that is, when the following conditions are met, it is determined that the interpolation model converges at the sampling combination X:

[0086]

[0087] in, It is The predicted values ​​of the motor performance parameters obtained by the iteration It is The predicted values ​​of the motor performance parameters obtained by the iteration is the set allowable error, generally set to 10 -4 .

[0088] When the interpolation model converges at all K sampling combinations, it is determined that the interpolation model meets the convergence condition; otherwise, it is determined that the interpolation model does not meet the convergence condition, and the sampling combination X′ that does not meet the convergence condition is determined.

[0089] (4) When the interpolation model does not meet the convergence conditions, increase the sampling combination and refit the interpolation model until the interpolation model meets the convergence conditions, and then establish a Kriging prediction model for the motor performance parameters, including:

[0090] When the interpolation model does not converge at at least one sampling combination X′, discrete sampling points are added within a predetermined range of sampling values ​​of at least one design parameter in the sampling combination X′ to add a new sampling combination at the sampling combination X′. The number of newly added sampling combinations is:

[0091]

[0092] in, is the predicted value of the motor performance parameter obtained by substituting the sampling combination X′ into the interpolation model with known model parameters, Y(X′) is the motor performance parameter of the sampling combination X′, and η is the design threshold.

[0093] For the newly added sampling combination, the electromagnetic weight W, temperature rise reference value ΔT and radial vibration response x of the shaft-clamped synchronous generator under this sampling combination are calculated using step 2. radial (t). Then, the original sampling combination and the newly added sampling combination are combined to determine the interpolation model again according to step 3.

[0094] The above methods can be used to obtain converged electromagnetic weight Kriging prediction models, temperature rise reference value Kriging prediction models and radial vibration response Kriging prediction models respectively.

[0095] Step 4: Optimize the comprehensive performance of the shaft-clamped synchronous generator under each design parameter combination using an optimization algorithm to obtain a design parameter combination that optimizes the comprehensive performance of the shaft-clamped synchronous generator. In one embodiment, the optimization algorithm used is a multi-objective optimization genetic algorithm NSGA-II.

[0096] Among them, for each design parameter combination, the values ​​of each design parameter in the design parameter combination are first substituted into the electromagnetic weight Kriging prediction model to obtain the electromagnetic weight prediction value, and the values ​​of each design parameter in the design parameter combination are substituted into the temperature rise reference value Kriging prediction model to obtain the temperature rise prediction value, and the values ​​of each design parameter in the design parameter combination are substituted into the radial vibration response Kriging prediction model to obtain the radial vibration response prediction value.

[0097] The comprehensive performance of the shaft-clamped synchronous generator under a design parameter combination is then weighted based on the electromagnetic weight prediction value, temperature rise prediction value, and radial vibration response prediction value. The smaller the electromagnetic weight prediction value, the smaller the temperature rise prediction value, and the smaller the radial vibration response prediction value of the shaft-clamped synchronous generator, the better the comprehensive performance of the shaft-clamped synchronous generator. In one embodiment, the comprehensive performance Θ of the shaft-clamped synchronous generator under the design parameter combination is:

[0098]

[0099] in, is the electromagnetic weight prediction value, is the predicted value of radial vibration response. is the predicted value of temperature rise, λ is a positive penalty coefficient, ΔT max is a pre-set temperature rise threshold. When the predicted temperature rise reaches the threshold, a penalty is imposed to reduce the adverse impact on the generator. γ1, γ2, and γ3 are all weighted parameters, and γ1 + γ2 + γ3 = 1.

[0100] The above description is only a preferred embodiment of the present application, and the present application is not limited to the above embodiments. It is understood that other improvements and variations directly derived or imagined by those skilled in the art without departing from the spirit and concept of the present application should be considered to be included in the scope of protection of the present application.

Claims

1. A multi-physics field coupled shaft-clamped synchronous generator design method, characterized in that: The shaft-clamping synchronous generator design method includes: Discrete sampling is performed within the parameter range of each design parameter of the shaft-clamped synchronous generator to construct multiple sampling combinations, each sampling combination includes K sampling values ​​of different design parameters, and the integer parameter K is greater than or equal to 2; For each sampling combination, the electromagnetic weight W of the shaft-clamped synchronous generator under the sampling combination is determined, the electromagnetic power loss P under the sampling combination is determined based on the electromagnetic field model of the shaft-clamped synchronous generator, and the electromagnetic power loss P is input into the temperature field as a heat source. The temperature rise reference value ΔT considering the electromagnetic characteristics under the sampling combination is calculated using the thermal resistance method; the electromagnetic weight W and the temperature rise reference value ΔT are substituted into the system vibration equation for solution to obtain the radial vibration response x considering the temperature rise characteristics and the electromagnetic weight W under the sampling combination. radial (t); The electromagnetic weight W of each sampling combination is used as output and the sampling combination is used as input to establish a convergent electromagnetic weight Kriging prediction model; the temperature rise reference value ΔT of each sampling combination is used as output and the sampling combination is used as input to establish a convergent temperature rise reference value Kriging prediction model; the radial vibration response x of each sampling combination is used as output and the sampling combination is used as input to establish a convergent temperature rise reference value Kriging prediction model. radial (t) establishing a convergent radial vibration response Kriging prediction model with the sample combination as output and input; An optimization algorithm is used to optimize the comprehensive performance of the shaft-bracketed synchronous generator under each design parameter combination, and a design parameter combination that optimizes the comprehensive performance of the shaft-bracketed synchronous generator is obtained; wherein, for each design parameter combination, an electromagnetic weight prediction value is obtained based on the design parameter combination using an electromagnetic weight Kriging prediction model, a temperature rise prediction value is obtained based on the design parameter combination using a temperature rise reference value Kriging prediction model, and a radial vibration response prediction value is obtained based on the design parameter combination using a radial vibration response Kriging prediction model. The comprehensive performance of the shaft-bracketed synchronous generator under the design parameter combination is obtained by weighting the electromagnetic weight prediction value, the temperature rise prediction value, and the radial vibration response prediction value. The smaller the electromagnetic weight prediction value, the smaller the temperature rise prediction value, and the smaller the radial vibration response prediction value of the shaft-bracketed synchronous generator, the better the comprehensive performance of the shaft-bracketed synchronous generator.

2. The design method of a shaft-clamped synchronous generator according to claim 1, characterized in that: The temperature rise reference value ΔT considering the electromagnetic characteristics under the sampling combination is calculated using the thermal resistance method as follows: ΔT=P×R total Among them, the total thermal resistance R total =R1+R2+R3+R4, R1 is the thermal resistance of the winding, R2 is the contact thermal resistance between the winding and the iron core, R3 is the stator thermal resistance, and R4 is the convection heat dissipation resistance on the stator surface.

3. The design method of a shaft-clamped synchronous generator according to claim 2, characterized in that: Where L is the effective length of the core, k ef is the thermal conductivity of the core material, A1 is the cross-sectional area of ​​the stator winding; d is the average gap between the stator winding and the core, k2 is the thermal conductivity of the contact material between the stator winding and the core, A2 is the contact area between the stator winding and the core; D so is the stator outer diameter, D si is the stator inner diameter, k s is the thermal conductivity of the stator material; h is the convective heat transfer coefficient.

4. The design method of a shaft-clamped synchronous generator according to claim 1, characterized in that: The radial vibration response considering the temperature rise characteristics and electromagnetic weight under the sampling combination includes: The system mass matrix M is constructed based on the electromagnetic weight, and the material stiffness matrix K(T)=K0(1-α T ×ΔT), where α T is the influence coefficient, K0 is the initial material stiffness matrix; Substitute the material stiffness matrix K(T) and the system mass matrix M into the system vibration equation The radial vibration response x considering the temperature rise characteristics and electromagnetic weight is obtained by the modal superposition method. radial (t); where is x radial The first derivative of (t), is x radial (t), C is the system damping matrix; F radial (t) is the radial electromagnetic vibration force of the generator and is calculated based on the electromagnetic field parameters of the electromagnetic field model of the shaft-clamped synchronous generator.

5. The design method of a shaft-clamped synchronous generator according to claim 4, characterized in that: The radial electromagnetic vibration force of the generator calculated based on the electromagnetic field parameters of the electromagnetic field model of the shaft-clamping synchronous generator is: Where μ0 is the vacuum permeability, B is the magnetic induction intensity of the electromagnetic field and I is the motor output current, A radial is the effective bearing area of ​​the shaft-clamping synchronous generator and A radial =2πrL; N s is the number of stator winding turns, g is the air gap length, r is the average radius of the air gap, and L is the effective length of the core.

6. The design method of a shaft-clamped synchronous generator according to claim 1, characterized in that: For any one motor performance parameter of electromagnetic weight, temperature rise reference value, and radial vibration response, establishing a Kriging prediction model of the motor performance parameter includes: Constructing an interpolation model with unknown model parameters Among them, Y(X) represents the motor performance parameters, X i represents the value of the i-th design parameter, β i is the interpolation coefficient corresponding to the i-th design parameter, β0 is the zero-order interpolation coefficient, z(X) is the random process term, and the unknown model parameters include β0, β i and Z(X); Substitute the sampled value of the i-th design parameter in each sampling combination into the X of the interpolation model i The motor performance parameters corresponding to the sampling combination are substituted into Y(X) and solved simultaneously to obtain the model parameters in the interpolation model, including β0, Z(X) and the interpolation coefficients β1~β corresponding to various design parameters. K ; The model parameters are substituted into the interpolation model and the convergence of the interpolation model is judged. When the interpolation model does not meet the convergence conditions, the sampling combination is increased and the interpolation model is refitted until the interpolation model meets the convergence conditions, thereby establishing a Kriging prediction model for the motor performance parameters.

7. The design method of a shaft-clamped synchronous generator according to claim 6, characterized in that: The convergence judgment of the interpolation model includes: The sampled values ​​of each design parameter in each sampling combination are respectively substituted into the interpolation model with known model parameters to obtain the predicted values ​​of the motor performance parameters of the sampling combination. When the relative error of the predicted values ​​of the motor performance parameters obtained twice in a row for the same sampling combination is less than the error threshold, it is determined that the interpolation model converges at the sampling combination; when the interpolation model converges at all K sampling combinations, it is determined that the interpolation model meets the convergence condition; otherwise, it is determined that the interpolation model does not meet the convergence condition.

8. The design method of a shaft-clamped synchronous generator according to claim 7, characterized in that: When the interpolation model does not meet the convergence condition, adding sampling combinations and refitting the interpolation model includes: When the interpolation model does not converge at at least one sampling combination X′, discrete sampling points are added within a predetermined range of sampling values ​​of at least one design parameter in the sampling combination X′ to add a new sampling combination at the sampling combination X′. The number of the newly added sampling combinations is in, is the predicted value of the motor performance parameter obtained by substituting the sampling combination X′ into the interpolation model with known model parameters, Y(X′) is the motor performance parameter of the sampling combination X′, and η is the design threshold.

9. The design method of a shaft-clamped synchronous generator according to claim 1, characterized in that: The comprehensive performance Θ of the shaft-clamped synchronous generator under the design parameter combination is obtained based on the weighted electromagnetic weight prediction value, temperature rise prediction value, and radial vibration response prediction value: in, is the electromagnetic weight prediction value, is the predicted value of radial vibration response, is the predicted value of temperature rise, λ is a positive penalty coefficient, ΔT max is the temperature rise threshold, γ1, γ2, and γ3 are all weighted parameters and γ1+γ2+γ3=1.

10. The design method of a shaft-clamped synchronous generator according to claim 1, characterized in that: When optimizing the comprehensive performance of the shaft-clamped synchronous generator under each design parameter combination using an optimization algorithm, the optimization algorithm used is the multi-objective optimization genetic algorithm NSGA-Ⅱ.

Images