Induction motor electromagnetic vibration and noise suppression optimization method

CN118940581APending Publication Date: 2024-11-12JIANGSU UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411022795.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-29
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The electromagnetic vibration and noise problems in electric vehicle drive induction motors have led to a decrease in service life, impact on ride comfort and driving safety. The existing technology has not been effectively carried out from the two aspects of electromagnetic force waves and inherent mode analysis of motor structures. Comprehensive noise reduction design.

Method used

The time-step finite element method is used to calculate electromagnetic force waves, and the mechanical structure modal analysis and rotor chute design optimization are combined with the random switching frequency pulse width modulation strategy (RFPWM) to weaken electromagnetic vibration and noise.

Benefits of technology

While ensuring the torque and efficiency of the motor, it effectively weakens electromagnetic vibration and noise, improves the performance and service life of the motor, and enhances ride comfort and driving safety.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118940581A_ABST
    Figure CN118940581A_ABST
Patent Text Reader

Abstract

The invention discloses an optimization design method for electromagnetic vibration and noise of an induction motor, and the method comprises the specific steps: 1, calculating an electromagnetic force wave of the induction motor for driving an electric vehicle through employing a time-step finite element method, and considering the influences of rotor motion, motor slotting and saturation through the proper encryption of a finite element subdivision grid; 2, the vibration mode and the inherent frequency of the anisotropic material of the motor are solved through mechanical structure modal analysis; step 3, adjusting air gap harmonic waves through different inclined distances of the rotor guide bars, and determining an optimal inclined slot distance; 4, determining the optimal inclination of the motor rotor, and optimizing the optimal inclination of the rotor by adopting a genetic algorithm; and step 5, adopting a random switching frequency pulse width modulation (RFPWM) strategy to uniformly disperse harmonic peak values concentrated at the switching frequency and the frequency multiplication position within the whole random switching frequency range, reducing the peak value of harmonic energy, and weakening electromagnetic vibration and noise of the motor.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of improving the efficiency of induction motors, and more specifically, relates to an optimization design method for electromagnetic vibration and noise of induction motors. Background Art

[0002] The motor vibration noise level is an important indicator to measure the performance of induction motors. According to the noise generation mechanism, motor noise can be divided into mechanical noise, electromagnetic noise and aerodynamic noise. Mechanical noise is determined by the machining accuracy and assembly quality of the motor. Mechanical noise can be reduced by improving the stability of the motor bearing-rotor system and suppressing the resonance of motor components. Electromagnetic noise is mainly related to the vibration caused by the electromagnetic force of the motor. The electromagnetic vibration of the electric vehicle drive motor will affect the performance of the motor and reduce the service life of the motor. The noise of the motor will affect the comfort of the ride and the safety of driving. Excessive noise will seriously affect the living environment and quality of life of people around. With the pursuit of quality of life and the improvement of environmental awareness, a low electromagnetic vibration and low noise electric vehicle drive system is bound to increase the competitiveness of electric vehicles in the market and meet people's pursuit of high-quality life. Therefore, the electromagnetic vibration and noise problems of electric vehicle drive motors have important theoretical research significance and engineering application value.

[0003] Electromagnetic noise is mainly caused by the electromagnetic force waves caused by the harmonics of the air gap magnetic density when the motor is powered on, which act on the motor structure. Therefore, reducing the source of the electromagnetic force wave can reduce the electromagnetic noise. The accuracy of the electromagnetic force wave division directly affects the prediction accuracy and noise reduction effect of the electromagnetic noise. The electromagnetic vibration and noise of the induction motor are not only related to the radial electromagnetic force wave in the motor air gap, but also closely related to the mechanical structure of the motor. When the vibration mode of the radial electromagnetic force wave in the induction motor is the same as the modal vibration mode of the induction motor, or the frequency of the radial electromagnetic force wave is close to the natural frequency of the mechanical structure of the induction motor, the motor will resonate, causing serious electromagnetic vibration and noise. Modal analysis of the induction motor can obtain the modal vibration modes of each order of the induction motor and the corresponding natural frequencies. Modal analysis can be performed in three ways: numerical calculation, finite element simulation, and modal experiment. Compared with numerical calculation, finite element simulation can take into account the irregularity of the structure and the accuracy of the calculation results is high; compared with modal experiments, there is no need to build an experimental platform and the cost is low, so it is widely used.

[0004] The rapid development of modern computer technology and signal analysis technology has promoted the in-depth research on the problem of electromagnetic noise in motors. Electromagnetic force waves and the inherent modes of motor structure are two key factors that determine the size of motor electromagnetic noise. In recent years, the research on the problem of electromagnetic noise in motors has also mainly focused on these two aspects.

[0005] Chinese invention CN 112803648A discloses a low electromagnetic noise brushless motor and its use method, including a bottom plate, a housing is fixedly arranged on the upper end of the bottom plate, a plurality of stators are evenly fixedly arranged on the inner side wall of the housing, a rotor is movably arranged on the inner middle part of the housing, a coil is wound around the surface of the stator and the rotor, connecting rods are fixedly arranged at both ends of the rotor, a bearing is arranged between the connecting rod and the housing, a fixing ring is fixedly sleeved on the outer surface of the connecting rod on the outer surface of the housing, a sealing mechanism is arranged between the fixing ring and the connecting rod, and the sealing mechanism includes a first sealing groove, a first sealing ring, a second sealing groove, a second sealing ring, a first connecting pipe, and a liquid injection pipe. The present invention has good use effect, and the sealing property of the bearing can be improved by the sealing mechanism, so that dust can be effectively prevented from entering the bearing, so that the bearing can be effectively prevented from being contaminated, resulting in unnecessary noise when the device rotates at high speed. Invention CN 218161956 U discloses an outer rotor brushless motor with improved electromagnetic noise, which relates to the field of motor technology. It includes a shell, an iron core is installed in the shell, and multiple magnets are installed between the shell and the iron core. The magnets are arc-shaped structural blocks, and the multiple magnets are arranged in a circular array in sequence and fit between the inner wall of the shell and the circumference of the iron core; a machine base is installed in the middle of one side of the iron core, and two annular disks are installed in the machine base, one of which is equipped with an inner ring on one side of the annular disk and a fixed ring on one side of the other annular disk, and a bearing is installed in the middle of the two annular disks, the inner ring and the fixed ring. The above inventions all reduce electromagnetic vibration and noise from the design of the motor body, and do not analyze the design method of reducing electromagnetic noise from the two aspects of electromagnetic force wave and motor inherent mode analysis. Summary of the invention

[0006] In response to the above problems, this paper will start with the analysis of electromagnetic force waves and the inherent modes of the motor structure, study the main influences on the electromagnetic noise of the induction motor used to drive electric vehicles, and suppress the electromagnetic vibration and noise of the motor from two aspects: motor body design and control strategy.

[0007] The technical solution of the present invention is: an optimization design method for electromagnetic vibration and noise of an induction motor, and the specific process is: step 1, using the time-step finite element method to calculate the electromagnetic force wave of the induction motor used for electric vehicle driving, and taking into account the influence of rotor motion, motor slotting and saturation by appropriately encrypting the finite element mesh; step 2, solving the vibration mode and natural frequency of the anisotropic material of the motor through mechanical structure modal analysis; step 3, adjusting the air gap harmonics by tilting the rotor bars at different distances to determine the optimal tilt slot distance; step 4, determining the optimal inclination of the motor rotor, and optimizing the optimal inclination of the rotor by using a genetic algorithm; step 5, using a random switching frequency pulse width modulation strategy (RFPWM) to evenly disperse the harmonic peaks concentrated at the switching frequency and its multiples within the entire random switching frequency range, reduce the peak value of the harmonic energy, and weaken the electromagnetic vibration and noise of the motor.

[0008] The beneficial effects of the present invention are

[0009] (1) Rotor skew slots are an effective way to weaken the electromagnetic vibration and noise of the motor. When the rotor is skewed, the harmonics in the air gap flux are weakened, thereby weakening the electromagnetic vibration and noise of the motor. The present invention optimizes the slot matching and rotor skew slot distance of the induction motor for electric vehicles. When the number of stator slots remains unchanged, the number of rotor slots and the skew slot distance of the rotor bars are optimized, thereby weakening the electromagnetic vibration and noise of the motor while ensuring the motor torque, efficiency and other performance.

[0010] (2) The current output by the inverter contains harmonics of the switching frequency and its sidebands. These harmonics have the characteristics of high frequency and narrow bandwidth, which will cause electromagnetic vibration and narrowband noise in the motor.

[0011] (RFPWM) improves the energy distribution characteristics concentrated in the switching frequency harmonic spectrum, disperses the harmonic energy, and makes it have a wider bandwidth, thereby achieving the purpose of reducing high-frequency electromagnetic vibration and narrowband noise. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 Flow chart of electromagnetic force wave calculation using time-step finite element method

[0013] Figure 2 This is the process of optimizing the optimal rotor pitch using genetic algorithms.

[0014] Figure 3 It is a fixed switching frequency SVPWM sampling diagram

[0015] Figure 4 is the switching waveform of the induction motor vector control

[0016] Figure 5 This is the optimization design flow chart for suppressing electromagnetic vibration and noise of induction motor

[0017] Figure 6 This is the overall flow chart of the optimization design for suppressing electromagnetic vibration and noise of the induction motor proposed by the present invention. DETAILED DESCRIPTION

[0018] As people pay more attention to noise pollution, low noise has gradually become an important performance indicator of electric vehicles. The drive motor is a key component of electric vehicles. It has the design characteristics of high current and high magnetic density, wide speed regulation range, complex working conditions and variable frequency speed regulation. These characteristics increase the possibility of large electromagnetic noise in the motor. The performance requirements of the drive motor and the particularity of the working environment make the electromagnetic noise problem of the motor more prominent and the difficulty of noise reduction increases. In order to improve the overload capacity, the electromagnetic density of the electric vehicle drive motor is relatively high, which makes the motor saturated at a high load working point and increases the content of electromagnetic force waves. The motor structural parameters such as the yoke thickness of the drive motor are also smaller than those of general industrial drive motors, which reduces the stiffness of the motor structure and increases the amplitude of electromagnetic vibration. The narrow installation space and harsh working environment put forward higher requirements on the structural design and heat dissipation design of the drive motor, resulting in a more compact and complex structure of the drive motor, which increases the difficulty of analyzing the inherent modes of the motor structure. Variable frequency speed regulation increases the content of power supply harmonics, making the content of electromagnetic force waves richer. At the same time, the wide speed regulation range increases the difficulty of avoiding electromagnetic resonance. Induction motor is one of the main types of electric vehicle drive motors. The electromagnetic force waves acting on the surface of the motor stator and the natural modes of the motor structure are two key factors that determine the electromagnetic noise of the induction motor. The present invention studies and improves the electromagnetic noise problem of the induction motor used to drive electric vehicles from these two aspects.

[0019] At present, there are three main types of electric vehicle drive motors: induction motors, permanent magnet motors and switched reluctance motors. Induction motors have the characteristics of simple structure and high reliability, and have been widely used in electric vehicles. Induction motors for electric vehicle drive are small induction motors, generally using a squirrel cage structure. The electromagnetic force wave characteristics of induction motors for electric vehicle drive can be calculated by the finite element method for the harmonic flux density generated by the stator and rotor under different speeds and loads. The change law of electromagnetic force waves with motor speed and load and the main design parameters affecting the amplitude, frequency and number of electromagnetic force waves are analyzed. For ordinary industrial drive induction motors, the motor mainly works at the rated speed point. Generally, the resonant frequency of the motor structure is far from the electromagnetic force wave frequency, and the load current and no-load current are not much different in value. Therefore, the noise caused by load and speed changes does not change much, and the electromagnetic noise is small. Unlike ordinary industrial drive motors, electric vehicle drive motors have a large overload multiple and a wide speed regulation range. The motor noise under different speeds and load conditions varies greatly, and there may be large electromagnetic noise in some working ranges. Therefore, it is necessary to study the electromagnetic noise characteristics of induction motors used to drive electric vehicles and analyze the reasons for large changes in electromagnetic noise and excessive local electromagnetic noise. The main cause of electromagnetic vibration and noise in motors is radial electromagnetic force waves, which can be analyzed by analytical calculation or finite element calculation. Optimizing the matching of stator and rotor slots and rotor skew slots are effective means to weaken radial electromagnetic force waves.

[0020] Mode is the inherent and overall vibration characteristic of mechanical structure. Each order mode has its specific vibration form and natural frequency. By performing modal analysis on the mechanical structure, the modal characteristics of the mechanical structure within a certain frequency range can be obtained, and then the vibration response of the mechanical structure under internal and external excitation can be predicted. In the design process of induction motor, the volume of the motor stator and rotor is determined, so the modal of each order of the motor can be determined. Therefore, electromagnetic design can be used to weaken the amplitude of the radial electromagnetic force wave with the same modal vibration mode and the amplitude of the radial electromagnetic force wave with a frequency close to the natural frequency to avoid the generation of resonance. The electromagnetic vibration of the motor mainly comes from the radial vibration of the motor, so the axial vibration of the motor can be ignored. Modal analysis can be carried out through three methods: numerical calculation, finite element simulation and modal experiment. Among them, finite element simulation has high accuracy and low cost, so it is widely used. The accuracy of the three-dimensional model of the motor directly affects the accuracy of the modal analysis results, so it is necessary to establish an accurate three-dimensional model of the motor. The stator and rotor cores are made of laminated silicon steel sheets, and the stator winding is made of copper wire. Therefore, these two materials are not isotropic materials, and the determination of their elastic mechanical parameters is a difficult point in modal analysis.

[0021] The stator and rotor harmonic magnetic fields of induction motors are mainly generated by factors such as the current distribution of the stator and rotor conductors, the slotting of the stator and rotor, and the saturation of the core. When the speed and load of the motor change, the current distribution in the stator and rotor conductors of the motor and the saturation of the core change accordingly, causing the stator and rotor harmonic magnetic fields to change, causing the electromagnetic force waves to change, thereby causing the electromagnetic noise to change. Therefore, the present invention analyzes the electromagnetic force wave characteristics of the induction motor used to drive electric vehicles from the time-step finite element method and the natural mode of the induction motor.

[0022] 1. The time-stepping finite element method is used to calculate the electromagnetic force waves of the induction motor used to drive electric vehicles. The effects of rotor motion, motor slotting and saturation are considered by appropriately encrypting the finite element mesh.

[0023] Figure 1 This is a flow chart of the time-step finite element method for calculating electromagnetic force waves. Induction motors used in electric vehicle driving are generally designed with high current and high magnetic flux density. The motor has a high degree of saturation at the maximum load operating point. The use of the magnetic potential multiplied by the magnetic permeance method to calculate electromagnetic force waves cannot well consider the influence of rotor motion, motor slotting and motor saturation. The calculation accuracy is low and some key electromagnetic force wave components may be missed, which is not conducive to accurately analyzing the cause of electromagnetic noise. The finite element method is currently the most widely used numerical calculation method with high analysis accuracy. The time-step finite element method is used to calculate the electromagnetic force waves of induction motors used in electric vehicle driving. By appropriately encrypting the finite element mesh, the influence of rotor motion, motor slotting and saturation is well considered. The calculation steps are as follows: Figure 1 As shown. First, the electromagnetic field of the induction motor is calculated using the time-step finite element method, and then the air gap magnetic flux distribution and harmonic waves of the stator and rotor are analyzed by finite element simulation. Finally, the amplitude, order and frequency of the electromagnetic force wave are analyzed by the Maxwell stress equation. The Maxwell stress equation of the electromagnetic force wave density on the surface of the stator core is:

[0024]

[0025] In formula (1), B μi is the air gap magnetic flux harmonic amplitude generated by the stator; B λj is the harmonic amplitude of the air gap flux density generated by the rotor; μ0 is the harmonic coefficient of the air gap flux density; r is the force wave order; ψ r is the initial phase angle of the force wave; ω r is the force wave angular frequency.

[0026] r=μ i ±λ j (2)

[0027] In formula (2), μ i is the number of force waves generated by the stator; j is the number of force waves generated by the rotor.

[0028]

[0029] In formula (3), ω1 is the power frequency; k is an integer; s2 is the number of rotor slots; p is the number of pole pairs; s is the slip rate, and k1 is the saturation coefficient of the motor, which reflects the degree of magnetic saturation inside the motor.

[0030] It can be seen from formula (1) that as long as the amplitude, order and frequency of each harmonic of the air gap flux generated by the stator and rotor are calculated, the radial electromagnetic force waves acting on the surface of the stator core can be obtained.

[0031] 2. Solve the vibration mode and natural frequency of the anisotropic material of the motor through mechanical structure modal analysis.

[0032] Mechanical structure modal analysis utilizes the orthogonality of the modes, takes the matrix composed of each order modal vector as the transformation matrix, and transforms it linearly, thereby mapping the mutually coupled physical coordinates and parameters into independent modal equations. The motion equation of the mechanical structure is derived based on Hamilton's theorem:

[0033]

[0034] In the above formula (4), K is the stiffness matrix; R is the damping matrix; M is the mass matrix; μ is the node displacement vector; F represents the force vector of the external excitation node; represents the first and second differentials of the node displacement vector, which are the velocity and acceleration of the displacement, respectively. Modal analysis studies the free vibration of the mechanical structure, that is, the undamped vibration, so in equation (4), let F = {0}, and replace the time inverse d / dt with jω (where ω is the angular frequency), then the eigenvalue corresponding to the modal equation of the induction motor mechanical structure is:

[0035] (K-ω 2 M)μ={0} (5)

[0036] According to the theory of linear equations, the necessary and sufficient conditions for having non-zero solutions are:

[0037] |K-ω 2 M|=0 (6)

[0038] Combining equations (5) and (6) we can solve for the frequency ω 2 (r=1,2,…n) and the corresponding vibration order vector u r (r=1,2,…n), where ω, u r are the natural frequency and mode order of the mechanical structure respectively.

[0039] The stator and rotor of the induction motor are made of laminated silicon steel sheets, and the stator winding is made of copper wire. Therefore, the xy plane direction and the z direction perpendicular to it have different mechanical properties. This material is called orthotropic material. The stiffness matrix of orthotropic material is different from that of isotropic material. The relationship between its stress tensor σ and strain tensor ε is:

[0040]

[0041] Where D is the elastic matrix; σ is the stress tensor, which contains 6 components, namely the tensile or compressive stress σ along the x, y, and z directions. x , σ y , σ z , and the shear stress components τ along the yz, zx, and xy planes yz , τ zx , τ yx These components together determine the stress state of the material in different directions. The strain tensor ε contains six components, namely the linear strain ε along the x, y, and z directions. x ,ε y ,ε z , and the shear strain components γ along the yz, zx, and xy planes yz ,γ zx ,γ yx These components together determine the deformation state of the material in different directions.

[0042] The elastic parameters of orthotropic materials in the xy plane are the same, so the elastic matrix D can be simplified accordingly:

[0043]

[0044] where d 11 , d 12 , d 13 , d 33 , d 44 is the elastic coefficient of the orthotropic material, obtained by Formula 9.

[0045] Expressed as an engineering constant, the elastic modulus E is x =E y =E z , Poisson's ratio ν xy =v yx , ν xz =v zx , shear modulus G xy =E x / [2(1+V xy )], Therefore, the elastic coefficient of the orthotropic material can be obtained as:

[0046]

[0047] Substituting equation (9) into equation (8) can obtain the elastic matrix D of the anisotropic material, thereby solving the vibration mode and natural frequency of the anisotropic material.

[0048] 3. The air gap harmonics are adjusted by tilting the rotor bars at different distances to determine the optimal skew distance.

[0049] Rotor skew slots are an effective way to reduce electromagnetic vibration and noise in motors. When the rotor slots are skewed, the harmonics in the air gap magnetic field are weakened, and the electromagnetic force waves in the air gap change, which will affect the electromagnetic vibration and noise of the motor. A common motor design method is to offset the rotor slot position by the distance of a stator tooth, but this method is inaccurate. Rotor bar skews at different distances have different effects on air gap harmonics, so it is necessary to determine the optimal skew slot distance.

[0050] Assume that the stator subharmonic winding coefficient of the induction motor with straight slots is K dpv' When the skew degree is C, the harmonic winding coefficient after the skew is K. dpcv' As shown in formula (10):

[0051] K dpcv' =K dpv' ×K cv' (10)

[0052] In formula (10), K cv' Indicates the skew coefficient of the v' harmonic when the skew degree is equal to C, K cv' The expression can be written as shown in formula (4.9):

[0053]

[0054] In formula (11), C represents the skewness, q is a constant, and q = 1, 2, 3, ... The expression of the slope is shown in formula (12):

[0055]

[0056] In formula (12), b sk represents the skew distance, and t1 represents the stator tooth pitch. Therefore, according to formula (10-12), the degree of attenuation of different harmonics by the skew degree can be calculated. Considering that a large skew degree will cause a large transverse current, thereby generating axial force, increasing additional losses, and reducing torque, the skew degree is selected between 0 and 2.

[0057] 4. Determine the optimal inclination of the motor rotor and use genetic algorithm to optimize the optimal inclination of the rotor.

[0058] In order to determine the optimal slope of the motor rotor, the present invention uses a genetic algorithm to optimize the optimal slope of the rotor. Figure 2 This is the process of optimizing the optimal inclination of the rotor using the genetic algorithm. The main process of optimizing the C value using the genetic algorithm includes: initializing the population, encoding, gene crossover and mutation, and population fitness analysis. The main process is analyzed and introduced below.

[0059] (1) Initialize the population

[0060] The initialization of population individuals is achieved by imitating the principle of random generation of individuals in the biological world. Because the generation of population individuals is random, the randomly generated population contains individuals that meet and do not meet the individual conditions. In each generation of population screening, the individuals that meet the conditions will be retained, and the individuals that do not meet the conditions will be eliminated. In the present invention, the population number of the C value is selected as 100, and the number of iterative evolutions is 100 generations. The numerical values ​​of the population number and iterative evolution meet the requirements of subsequent selection, crossover and mutation.

[0061] (2) Coding

[0062] The function of encoding is to encode the values ​​of random individuals into binary, simulating the DNA genetic information of organisms. Encoding includes the conversion of decimal to binary and binary to decimal. The crossover and mutation in the process can be operated directly on the binary-encoded gene fragments.

[0063] (3) Genetic crossover and mutation

[0064] The purpose of binary coding crossover and mutation is to improve the diversity of the population. Each generation of the population is a crossover and mutation of all individuals, which can prevent the target value from falling into local premature maturity.

[0065] (4) Population fitness analysis

[0066] The purpose of population fitness is to evaluate the individuals in the population. Similar to nature, individuals with higher fitness have a greater chance of survival. In the present invention, the objective function is K dpcv' In order to make K dpcv' The smaller the value of is, the fitness evaluation function is selected as 1 / K dpcv' The larger the value of the evaluation function, the higher the probability of individual survival. dpcv' The smaller it is.

[0067] When adjusting the system parameters, set the C parameter to the range of 0-2. dpcv' As the objective function optimized by genetic algorithm. Genetic algorithm can quickly obtain the objective function K through iterative evolution. dpcv' The minimum value of C is obtained and the optimal C value of the system is adjusted.

[0068] 5. Adopt the random switching frequency pulse width modulation strategy (RFPWM) to evenly disperse the harmonic peaks concentrated at the switching frequency and its multiples within the entire random switching frequency range, reduce the peak value of harmonic energy, and weaken the electromagnetic vibration and noise of the motor.

[0069] In the vector control system of an induction motor for electric vehicles, adopting a fixed switching frequency modulation strategy will generate a narrow-band harmonic cluster concentrated at the switching frequency and its multiples in the current waveform of the frequency converter. This harmonic energy is concentrated and easily causes the motor to generate relatively large-amplitude electromagnetic vibration and harsh narrow-band noise. The present invention adopts the random switching frequency pulse width modulation strategy (RFPWM) to evenly disperse the harmonic peaks concentrated at the switching frequency and its multiples within the entire random switching frequency range, reduce the peak value of harmonic energy, and weaken the electromagnetic vibration and noise of the motor. The regular sampling method samples the sine signal with the triangular carrier wave as the period, and uses a straight line with the same average amplitude as the sine wave to replace the modulation wave. Then, the intersection point of this straight line and the triangular carrier wave can replace the intersection point of the modulation wave and the carrier wave. The relationship between the triangular carrier wave and the modulation wave of the regular sampling method is plotted as Figure 3 shown. Figure 3 In it, the amplitude of the triangular carrier wave is U cm , and the amplitude of the sine modulation wave is U sm , and the expression is r s (t) = U sm sin(ω1t), where ω1 = 2π / T s represents the angular frequency of the modulation wave. The ratio of the amplitudes of the modulation wave and the carrier wave is M = U sm / U cm (0 < M < 1), which is called the modulation ratio. At the sampling time t e , there is:

[0070]

[0071] Among them, T s is the sampling time, U cm is the amplitude of the triangular carrier wave, U sm is the amplitude of the sine modulation wave, and the expression is r s (t) = U sm sin(ω1t), ω1 = 2π / T s represents the angular frequency of the modulation wave, and t on is the conduction time.

[0072] Simplifying the above formula can obtain the conduction time t on as

[0073]

[0074] The expression of the duty cycle d:

[0075]

[0076] The above is the expression of the duty cycle of sinusoidal pulse width modulation. For space vector pulse width modulation technology, the modulation waveform of space vector pulse width modulation technology is equivalent to injecting the third harmonic of 1 / 6 of the fundamental amplitude into the fundamental wave. Similarly, the expression of its duty cycle can be obtained as follows:

[0077]

[0078] The output voltage of the inverter depends on the duty cycle within a switching cycle, and has nothing to do with the switching frequency and the conduction position, but the switching frequency and the conduction position can affect the spectrum of the output voltage. After adopting the random switching frequency or random conduction position modulation strategy, the harmonics with large amplitudes of the carrier frequency multiplication and its sidebands will be evenly dispersed in the random switching frequency or conduction position, thereby weakening the amplitude of the harmonics, and then weakening the electromagnetic vibration of the motor, avoiding the generation of harsh narrowband noise.

[0079] In vector modulation, the duty cycle has nothing to do with the action time of the zero vector, but is only related to the action time of the effective vector. According to the modulation principle of space vector, there are 8 basic voltage vectors in each switching cycle, including two zero vectors T 00 and T 07 , and its corresponding three-phase switch signal waveform is as follows Figure 4 shown.

[0080] The present invention proposes Figure 5 The random switching frequency pulse width modulation method shown in the figure assumes that the center frequency f c The corresponding switching cycle time is T, and the time corresponding to the switching frequency change Δf is ξ. Then the RFPWM under vector control can be expressed as follows: Figure 5 As shown. Random switching frequency pulse width modulation technology (RFPWM) achieves random switching frequency by randomly changing the frequency of the triangular carrier. In the inverter using RFPWM modulation strategy, the frequency of the triangular carrier is:

[0081] f c =f c0 +R i Δf (17)

[0082] In formula (17), f c0 represents the switching center frequency, i.e. the average value of the switching frequency. Δf represents the change in the switching frequency, which determines the upper and lower limits of the random switching frequency. R i Represents a random number distributed in the range [-1, +1]. The harmonic distribution of the inverter output voltage and current spectrum depends on the random switching frequency f c , f cThe wider the range of change, the wider and more uniform the distribution of the voltage and current spectrum. However, it should also be noted that the upper frequency limit is limited by the performance of the power switching device and the switching loss. c If the lower limit is too low, the output voltage and current ripple will be too large, which will affect the quality of the voltage and current waveforms.

[0083] Figure 6 This is the overall flow chart of the optimization design of electromagnetic vibration and noise suppression of induction motor proposed by the present invention. The electromagnetic force wave of the induction motor is analyzed by the time-step finite element method, taking into account the influence of rotor rotation, motor slotting and saturation. The electromagnetic force wave characteristics of the induction motor are studied by analyzing the harmonic magnetic field generated by the motor body.

[0084] The present invention adopts a modulation strategy of fixed switching frequency instead of RFPWM modulation strategy. The inverter output current waveform is concentrated at the fixed switching frequency and its multiples. The narrow bandwidth harmonic clusters are evenly dispersed within the random switching frequency range. The amplitude of the harmonics is effectively weakened, thereby effectively weakening the electromagnetic vibration of the motor and avoiding the generation of narrowband noise. The skew degree of the rotor is analyzed, and the electromagnetic vibration and noise are reduced by the design of the rotor skew. The specific implementation of the invention is not limited to the above-mentioned method. As long as various non-substantial improvements are made using the method concept and technical solution of the present invention, they are within the protection scope of the present invention.

Claims

1. An optimization design method for electromagnetic vibration and noise of an induction motor, characterized in that: The specific process is: Step 1: The electromagnetic force waves of the induction motor used to drive the electric vehicle are calculated using the time-step finite element method. The effects of rotor motion, motor slotting and saturation are considered by appropriately encrypting the finite element mesh. Step 2, solving the vibration mode and natural frequency of the anisotropic material of the motor through mechanical structure modal analysis; Step 3, adjusting the air gap harmonics by tilting the rotor bars at different distances to determine the optimal skew distance; Step 4, determining the optimal slope of the motor rotor, and optimizing the optimal slope of the rotor using a genetic algorithm; Step 5, using random switching frequency pulse width modulation strategy (RFPWM), the harmonic peaks concentrated at the switching frequency and its multiples are evenly dispersed within the entire random switching frequency range, reducing the peak value of the harmonic energy and weakening the electromagnetic vibration and noise of the motor.

2. The method for optimizing the design of electromagnetic vibration and noise of an induction motor according to claim 1, characterized in that: The specific process of step 1 is: Firstly, the electromagnetic field of the induction motor is calculated by the time-step finite element method, and then the air gap magnetic flux distribution and harmonic wave of the stator and rotor are simulated and analyzed by finite element method. Finally, the amplitude, frequency and frequency of the electromagnetic force wave are analyzed by the Maxwell stress equation. The Maxwell stress equation of the electromagnetic force wave density on the surface of the stator core is: In formula (1), B μi is the harmonic amplitude of the air gap flux density generated by the stator; B λj is the air gap flux harmonic amplitude generated by the rotor; μ0 is the air gap flux harmonic coefficient; r is the force wave order; ψ r is the initial phase angle of the force wave; ω r is the force wave angular frequency; r=μ i ±λ j (2) In formula (2), μ i is the number of force waves generated by the stator; j is the number of force waves generated by the rotor; In formula (3), ω1 is the power frequency; k is an integer; s2 is the number of rotor slots; p is the number of pole pairs; s is the slip rate, and k1 is the saturation coefficient of the motor, which reflects the degree of magnetic saturation inside the motor; It can be seen from formula (1) that as long as the amplitude, order and frequency of each harmonic of the air gap flux generated by the stator and rotor are calculated, the radial electromagnetic force waves acting on the surface of the stator core can be obtained.

3. The method for optimizing the design of electromagnetic vibration and noise of an induction motor according to claim 1, characterized in that: The specific process of step 2 is: Mechanical structure modal analysis utilizes the orthogonality of the modes and uses the matrix composed of each order modal vector as a transformation matrix to perform linear transformations, thereby mapping the mutually coupled physical coordinates and parameters into independent modal equations. Based on Hamilton's theorem, the motion equation of the mechanical structure is derived as follows: In the above formula (4), K is the stiffness matrix; R is the damping matrix; M is the mass matrix; μ is the node displacement vector; F represents the force vector of the external excitation node; represents the first and second differentials of the node displacement vector, which are the velocity and acceleration of the displacement, respectively. Modal analysis studies the free vibration of the mechanical structure, that is, the undamped vibration. Therefore, in equation (4), let F = {0}, and the time inverse d / dt is replaced by jω, where ω is the angular frequency. The eigenvalue corresponding to the modal equation of the induction motor mechanical structure is: (K-oh 2 M)μ={0} (5) According to the theory of linear equations, the necessary and sufficient conditions for having non-zero solutions are: |K-oh 2 M|=0 (6) Combining equations (5) and (6) we can solve for the frequency ω 2 (r=1,2,…n) and the corresponding vibration order vector u r (r=1,2,…n), where ω, u r are the natural frequency and mode order of the mechanical structure respectively; The stator and rotor of the induction motor are made of laminated silicon steel sheets, and the stator winding is made of copper wire. Therefore, the xy plane direction and the z direction perpendicular to it have different mechanical properties. This material is called orthotropic material. The stiffness matrix of orthotropic material is different from that of isotropic material. The relationship between its stress tensor σ and strain tensor ε is: Where D is the elastic matrix; σ is the stress tensor, which contains 6 components, namely the tensile or compressive stress σ along the x, y, and z directions. x , σ y , σ z , and the shear stress components τ along the yz, zx, and xy planes yz , τ zx , τ yx , these components jointly determine the stress state of the material in different directions. The strain tensor ε contains six components, namely the linear strain ε along the x, y, and z directions. x ,ε y ,ε z , and the shear strain components γ along the yz, zx, and xy planes yz ,γ zx ,γ yx , these components jointly determine the deformation state of the material in different directions; The elastic parameters of orthotropic materials in the xy plane are the same, so the elastic matrix D can be simplified accordingly to obtain the elastic coefficients of the orthotropic materials, thereby solving the vibration mode and natural frequency of the anisotropic materials.

4. The method for optimizing the design of electromagnetic vibration and noise of an induction motor according to claim 1, characterized in that: The specific process of step 3 is: Rotor skew is an effective way to weaken the electromagnetic vibration and noise of the motor. When the rotor is skewed, the harmonics in the air gap magnetic field are weakened, and the electromagnetic force wave in the air gap changes, which will affect the electromagnetic vibration and noise of the motor. The rotor bars have different effects on the air gap harmonics when they are skewed at different distances, so it is necessary to determine the optimal skew distance. Assume that the stator subharmonic winding coefficient of the induction motor with straight slots is K dpv' When the skew degree is C, the harmonic winding coefficient after the skew is K. dpcv' As shown in formula (10): K dpcv' =K dpv' ×K cv' (10) In formula (10), K cv' Indicates the skew coefficient of the v' harmonic when the skew degree is equal to C, K cv' The expression can be written as shown: In formula (11), C represents the skewness, q is a constant, and q = 1, 2, 3, ... The expression of the slope is shown in formula (12): In formula (12), b sk Represents the skew distance, t1 represents the stator tooth pitch, so according to formula (10-12), the degree of attenuation of different harmonics by the skew degree can be calculated; considering that a large skew degree will cause a large transverse current, thereby generating axial force, increasing additional losses and reducing torque, the slope is selected between 0 and 2.

5. The method for optimizing the design of electromagnetic vibration and noise of an induction motor according to claim 1, characterized in that: The specific process of step 4 is: In order to determine the optimal inclination of the motor rotor, a genetic algorithm is used to optimize the optimal inclination of the rotor. The main process of genetic algorithm for optimizing the C value includes: initializing the population, encoding, gene crossover and mutation, and population fitness analysis: (1) Initialize the population The initialization of population individuals is achieved by imitating the principle of random generation of individuals in the biological world. Because the generation of population individuals is random, the randomly generated population contains individuals that meet and do not meet the individual conditions. In each generation of population screening, those that meet the conditions will be retained, and those that do not meet the conditions will be eliminated. The population size of the C value is selected as 100, and the number of iterative evolutions is 100 generations. The population size and iterative evolution values ​​meet the requirements of subsequent selection, crossover and mutation. (2) Coding The function of encoding is to encode the values ​​of random individuals into binary, simulating the DNA genetic information of organisms; encoding includes the conversion of decimal to binary and binary to decimal; crossover and mutation in the process can be directly operated on the binary-coded gene fragments; (3) Genetic crossover and mutation The purpose of binary coding crossover and mutation is to improve the diversity of the population; each generation of the population is a crossover and mutation of all individuals, which can prevent the target value from falling into local premature maturity; (4) Population fitness analysis Population fitness is to evaluate the individuals in the population. Similar to nature, individuals with higher fitness are more likely to survive. The objective function is K dpcv' , in order to make K dpcv' The smaller the value of is, the fitness evaluation function is selected as 1 / K dpcv' , the larger the value of the evaluation function, the higher the probability of individual survival, K dpcv' The smaller it is; When adjusting the system parameters, set the C parameter to the range of 0-2. dpcv' As the objective function optimized by genetic algorithm; genetic algorithm can quickly obtain the objective function K through iterative evolution dpcv' The minimum value of C is obtained and the optimal C value of the system is adjusted.

6. The method for optimizing the design of electromagnetic vibration and noise of an induction motor according to claim 1, characterized in that: The specific process of step 5 is: The random switching frequency pulse width modulation strategy (RFPWM) is adopted to evenly disperse the harmonic peaks concentrated at the switching frequency and its multiples within the entire random switching frequency range, reduce the peak value of the harmonic energy, and weaken the electromagnetic vibration and noise of the motor; The regular sampling method samples the sine signal with the triangular carrier wave as the period, and uses a straight line with the same average amplitude as the sine wave to replace the modulation wave. Then the intersection points of this straight line and the triangular carrier wave can replace the intersection points of the modulation wave and the carrier wave, and the relationship between the triangular carrier wave and the modulation wave of the regular sampling method is plotted; the amplitude of the triangular carrier wave is U cm , and the amplitude of the sine modulation wave is U sm, The expression is r s (t) = U sm sin(ω1t), where ω1 = 2π / T s represents the angular frequency of the modulation wave, and the ratio of the amplitude of the modulation wave to that of the carrier wave is M = U sm / U cm (0 < M < 1) is called the modulation ratio. At the sampling time t e there is: Among them, T s is the sampling time, U cm is the amplitude of the triangular carrier, U sm is the amplitude of the sinusoidal modulation wave, expressed as r s (t) = U sm sin(ω1t),ω1=2π / T s represents the angular frequency of the modulated wave, t on is the on-time; Simplifying the above formula, we can get the conduction time t on for The expression of duty cycle d is: The above is the expression of the duty cycle of sinusoidal pulse width modulation. For space vector pulse width modulation technology, the modulation waveform of space vector pulse width modulation technology is equivalent to injecting the third harmonic of 1 / 6 of the fundamental amplitude into the fundamental wave. Similarly, the expression of its duty cycle can be obtained as follows: The output voltage of the inverter depends on the duty cycle within a switching cycle, and has nothing to do with the switching frequency and the conduction position, but the switching frequency and the conduction position can affect the spectrum of the output voltage; after adopting the random switching frequency or random conduction position modulation strategy, the harmonics with large amplitudes of the carrier frequency multiplication and its sidebands will be evenly dispersed in the random switching frequency or conduction position, thereby weakening the amplitude of the harmonics, and then weakening the electromagnetic vibration of the motor, avoiding the generation of harsh narrowband noise; In vector modulation, the duty cycle has nothing to do with the action time of the zero vector, but is only related to the effective vector action time. According to the modulation principle of space vector, there are 8 basic voltage vectors in each switching cycle, of which there are two zero vectors T 00 and T 07 ; Using random switching frequency pulse width modulation method, assuming that the center frequency f c The corresponding switching cycle time is T, and the time corresponding to the switching frequency change Δf is ξ. The random switching frequency pulse width modulation technology (RFPWM) realizes the randomization of the switching frequency by randomly changing the frequency of the triangular carrier. In the inverter using the RFPWM modulation strategy, the frequency of the triangular carrier is: f c =f c0 +R i Δf (17) In formula (17), f c0 represents the switching center frequency, i.e. the average value of the switching frequency. Δf represents the change in the switching frequency, which determines the upper and lower limits of the random switching frequency. R i Represents a random number distributed in the range of [-1, +1]. The harmonic distribution of the inverter output voltage and current spectrum depends on the random switching frequency f c , f c The wider the range of change, the wider and more uniform the distribution of the voltage and current spectrum. However, it should also be noted that the upper frequency limit is limited by the performance of the power switching device and the switching loss. c If the lower limit is too low, the output voltage and current ripple will be too large, which will affect the quality of the voltage and current waveforms.

Citation Information

Patent Citations

  • Low electromagnetic noise brushless motor and use method thereof

    CN112803648A

  • Outer rotor brushless motor capable of improving electromagnetic noise

    CN218161956U