A vibration state simulation monitoring system for dry-type reactors

Abstract

The present invention relates to the technical field of reactor vibration analysis, and specifically to a vibration state simulation monitoring system for dry-type reactors, the system comprising: assembling parts into a reactor model through the assembly function of drawing software; a model import module, used to import the drawn reactor model into finite element software, and set parameters in each part; a modal analysis module, used to determine and analyze the vibration characteristics of the reactor structure; a diagnosis module, used to obtain the vibration response characteristics of the reactor under power-on conditions according to the vibration characteristics of each part, and find out the vibration signal with obvious fault characteristics. The model of the relationship between the reactor fault and each characteristic quantity of its vibration signal established by the present invention can obtain the vibration characteristics of the reactor during the fault latent period from the model, so as to study the direct or indirect relationship between its latent characteristics and the fault characteristics, and adjust the relevant parameters of the reactor in advance, so as to prevent the occurrence of faults to a certain extent.

Description

Technical Field

[0001] The invention relates to the technical field of reactor vibration analysis, and in particular to a vibration state simulation monitoring system for a dry-type reactor. Background Art

[0002] With the increase of power grid capacity and the improvement of power quality requirements, the use of reactors has continued to increase, and their failures have increased accordingly. This situation has directly threatened the safe operation of urban power grids and has attracted people's attention. As a key equipment for power transmission and transformation of power systems, the reliable operation of reactors is of great significance to the safe, economical and stable operation of power systems.

[0003] Therefore, the reactor is the most important equipment in the power system, and its safe and reliable operation is related to the stability of the system operation. Various anti-fault studies must be carried out on it.

[0004] There are many types of failures in reactors, such as failures caused by lax quality control during the bolt assembly process, structural design defects, loose components caused by long-term vibration, breakage problems, poor welding, insulation defects, component quality problems, poor core grounding, and unqualified on-site installation processing technology.

[0005] There are many reasons for mechanical vibration defects. Among the known reasons, lax quality control of bolt assembly and structural design defects can lead to internal mechanical looseness defects during the long-term operation of the reactor due to the existence of vibration, thus causing mechanical failure. Therefore, studying the vibration issues of the reactor and ensuring the safe and reliable operation of the reactor are the key to maintaining the stability of the power system.

[0006] Vibration-based diagnostic methods are relatively mature in the application of rotating machinery (generators, large pumps, etc.), and can be used for reference and learning, but they have been applied to reactors relatively late. Research on them has been very hot in recent years, and there have been a number of literature reports on related research.

[0007] Based on the above literature reports, although there are many diagnostic methods for reactors, and they have made important contributions to reactor fault diagnosis, the overall effect is not very ideal and cannot meet the requirements of modern power systems, especially the requirements of smart grid development. This is mainly due to the complex internal structure of the reactor and the strong interference at the reactor operation site, which makes monitoring and diagnosis difficult. Therefore, few people in the prior art have studied the relationship between reactor winding deformation, winding looseness and other faults caused by reactor vibration, and thus the prior art cannot analyze the vibration changes caused by changes in the mechanical properties of the reactor winding.

[0008] For example, the invention patent with publication number CN117912198A discloses an early warning prediction system for reactors, which monitors the phase difference between zero-sequence current and zero-sequence voltage through an early warning discrimination module to perform adaptive compensation and analysis under different fault conditions. The system can monitor the status and data of the reactor in real time and can provide reliable early warnings and future trend predictions to determine future faults, and can operate stably and for a long time in the environment of high electromagnetic fields and high currents inside the reactor, but it still has not established the relationship between the reactor structure and the characteristic quantity of the vibration signal, and the accuracy needs to be improved.

[0009] For example, the invention patent with publication number CN117875082A discloses a method and system for simulating inter-turn short-circuit faults of dry-type air-core reactors, which establishes a geometric model of the dry-type air-core reactor, but the factors considered are only the electromagnetic field, temperature field and fluid field, and the influence of the reactor structure itself on the vibration is not considered. Therefore, the prediction result cannot be guaranteed to be accurate.

[0010] For example, a paper published in High Voltage Technology, titled Research on Vibration and Noise Characteristics of Ultra-High Voltage Shunt Reactor Core, discloses a method that combines magnetic-structural-acoustic multi-field coupling analysis to simulate and calculate the Maxwell force and magnetostrictive force acting on the core. The structural field calculation results are fast Fourier transformed and used as load excitation to calculate the noise generated by the vibration of the reactor, and the effects of magnetostrictive force and Maxwell force on the vibration of the reactor core are analyzed separately. However, it does not give a quantitative relationship between magnetostriction and vibration, and it is impossible to obtain parameters such as the frequency set when the reactor is working without any doubt through the force on the core.

[0011] For example, the invention patent with publication number CN116306212A discloses a method, device, equipment and storage medium for predicting reactor faults, which forms sample pairs with the acquired reactor operating status values, inputs them into a pre-trained LSTM-based twin network to predict reactor faults and obtain prediction results. Due to the complex structure of the reactor, it is obviously impossible to obtain accurate prediction results by only using a neural network model.

[0012] For example, the invention patent with publication number CN114894415A discloses a method for determining the optimal measuring point of a sensor based on the vibration signal on the surface of a transformer box. Starting from the analysis of the vibration and sound source, a vibration and sound propagation model is established, the propagation attenuation of the vibration and sound in the medium is analyzed, the attenuation coefficients of different measuring points are theoretically calculated, and the response of different positions of the box to the actual vibration and sound is quantitatively calculated from a theoretical analysis perspective. Therefore, the model and analysis method corresponding to this application are quite complex, involving many parameters, and are difficult to apply in actual production. Summary of the invention

[0013] Purpose of the invention: In order to overcome the deficiencies of the prior art, the present invention provides a vibration state simulation monitoring system for a dry-type inductor, which solves the blank problem in the prior art of realizing vibration signal monitoring and fault diagnosis through the structure of the inductor.

[0014] Technical solution: The present invention provides a vibration state simulation monitoring system for a dry-type reactor, the system comprising:

[0015] The model drawing module is used to draw the reactor model using drawing software. The drawing of the reactor model first draws the structural rods, insulating spacers, iron cores and coils, and then assembles the parts into the reactor model through the assembly function of the drawing software.

[0016] In this embodiment, there is no restriction on the drawing software, and conventional CAD (Computer Aided Design) three-dimensional drawing software or three-dimensional drawing software such as Pro / Engineer can be used to draw the reactor model.

[0017] The model import module is used to import the drawn reactor model into the finite element software and set the parameters of each part.

[0018] The finite element software used in this embodiment is the finite element software ANSYS. This embodiment sets the parameters in each part, specifically including: the setting of the material parameters of each part in the model is a very critical part for finite element analysis, and the nonlinear characteristics of the insulating pad will cause its elastic modulus to change with the change of the clamping force. Due to the material of the pad and its loose structure, when the pressure load applied to the pad changes, this makes the elastic coefficient of the insulating pad related to the degree of compression after assembly. Therefore, it is necessary to calculate the elastic modulus of the insulating pad under different preload forces to simulate the influence of winding looseness on the reactor.

[0019] The modal analysis module is used to determine and analyze the vibration characteristics of the reactor structure, wherein the vibration characteristics of the reactor structure include the vibration characteristics of the reactor winding, the vibration characteristics of the reactor core, and the vibration characteristics of the reactor box surface.

[0020] Specifically, in this embodiment, modal analysis is used to determine the vibration characteristics of the structure, and its essence is to solve the motion equation under the undamped and external load-free state with a finite number of degrees of freedom, and obtain the eigenvalues ​​and eigenvectors of the equation, that is, the natural frequencies and modal vibration shapes of each mode.

[0021] This is to determine the different effects of various frequencies and vibration modes on the structure, and through reasonable design, the natural frequency of the reactor is kept away from the excitation force frequency to avoid resonance, and thereby predict the actual vibration response of the reactor winding structure under external excitation within its natural frequency range.

[0022] Since vibration characteristics determine the response of the structure to various dynamic loads, any form of dynamic analysis must be based on modal analysis.

[0023] Harmonic response analysis is to determine the steady-state response of a structure under a sinusoidal load of known frequency and amplitude, calculate the vibration of the structure at different frequencies, and obtain the vibration amplitude curve.

[0024] When the reactor is working, under the excitation of electromagnetic force, the vibration of the coil can be regarded as simple harmonic motion. Harmonic response analysis can determine the continuous dynamic performance of the winding to avoid resonance caused by forced vibration. The modal superposition method is used in the analysis, which multiplies the vibration mode obtained by modal analysis by a factor and calculates the response of the structure by summing. The modal superposition method is faster and less expensive than the complete method, includes the prestress effect, and can aggregate the solution according to the natural frequency of the structure, which can produce a smoother and more accurate response curve.

[0025] The diagnostic module is used to obtain the vibration response characteristics of the reactor under power-on conditions based on the vibration characteristics of each part, so as to find out the vibration signal with obvious fault characteristics, and compare it with the test results when the winding is working, so as to further diagnose the fault of the reactor.

[0026] Further, including:

[0027] The analysis steps of the reactor winding vibration characteristics include:

[0028] Establish the vibration model of the reactor winding: When the reactor is working, the current in the winding interacts with the leakage magnetic field, generating electromagnetic force acting on the coil. Under the excitation of the electromagnetic force, a single coil is regarded as performing simple harmonic vibration in mechanics, thus representing the overall dynamic model of the winding;

[0029] Based on the winding overall dynamic model, the vibration acceleration and angular frequency of the reactor winding dynamic model are expressed;

[0030] The relationship between the magnitude of the preload force and the vibration acceleration and angular frequency is analyzed to find relevant measuring points to avoid resonance.

[0031] Further, including:

[0032] The overall dynamic model of the winding is expressed as:

[0033]

[0034] Among them, z n is the displacement of the nth coil; C is the damping coefficient between adjacent coils of the reactor; K Bis the stiffness coefficient between the coil and the upper iron yoke; K is the stiffness coefficient between adjacent coils; K H F is the stiffness coefficient between the coil and the iron yoke at the lower end; n is the electromagnetic force on the nth coil unit; mg is the mass of a single coil and the mass of each coil is the same; is the inertia force of the nth coil unit; is the damping force on the nth coil unit; Kz n is the elastic force on the nth coil unit;

[0035] The method of expressing the vibration acceleration and angular frequency of the reactor winding dynamics model based on the winding overall dynamics model specifically includes:

[0036] The stiffness coefficient between adjacent coils corresponding to each coil, the displacement of the coil and the electromagnetic force received are made equal, so as to add each equation in formula (1) to obtain:

[0037]

[0038] Among them, M = nm is the mass of the coil, C' = nc is the damping coefficient of the reactor coil, K' = k B +k z =2k is the elastic coefficient of the insulating pad, z are the acceleration, velocity and displacement of the axial motion of the coil, respectively, and F=F1+F2+,…,+F n ;

[0039] According to the law of electromagnetic induction and the left-hand rule, assuming that the current I = I0cos(ω0t), the axial electromotive force on the winding is:

[0040]

[0041] Wherein, k is the calculation parameter; I0 is the current amplitude, ω0 is the angular frequency;

[0042] Combining formula (2) and formula (3), we get:

[0043]

[0044] Formula (4) is a non-homogeneous linear differential equation with constant coefficients. Its solution is the general solution of the homogeneous equation plus the special solution of the non-homogeneous equation. Therefore, the homogeneous equation of formula (4) is:

[0045]

[0046] The general solution is:

[0047]

[0048] Among them, p, is a parameter related to the initial state of the winding, f=ω / (2π) is the natural frequency of the winding and complies with 4MK'>>C' 2 ,therefore,

[0049] The non-homogeneous equation is decomposed into two:

[0050]

[0051] Find their special solutions respectively:

[0052]

[0053] Therefore, in formula (8):

[0054]

[0055] Therefore, the general solution of equation (8) is obtained:

[0056]

[0057] By taking the derivative of formula (9) twice, the corresponding vibration acceleration can be obtained:

[0058]

[0059] and

[0060] Further, including:

[0061] The analysis of the relationship between the magnitude of the preload force and the vibration acceleration and angular frequency, so as to find relevant measuring points to avoid resonance, includes:

[0062] According to formulas (10) and (11),

[0063] If K'>4Mω0 2 , that is, ω>2ω0, the amplitude of a decreases as the pressing force decreases;

[0064] If K'<4Mω0 2 , that is, ω<2ω0, the amplitude of a increases as the pressing force decreases;

[0065] If K'=4Mω0 2 , that is, ω=2ω0, resonance will occur at the fundamental frequency;

[0066] Therefore, if a measuring point where no resonance occurs can be found, that is, ω is always greater than 2ω0 or ω is always less than 2ω0, the influence of resonance can be avoided and winding looseness defect diagnosis can be achieved.

[0067] Further, including:

[0068] The analysis steps of the vibration characteristics of the reactor core include:

[0069] First, analyze the electromagnetic field generated when the primary coil of the iron core is energized:

[0070] The magnetic field inside the reactor winding is regarded as a uniform magnetic field, and the magnetic field strength inside the winding is expressed as:

[0071]

[0072] The magnetic induction intensity is expressed as:

[0073] B=μ0μ r H=μ0μ r (N / L0)I (13)

[0074] Where L0 is the average length of the core loop; μ0 = 4π × 10 -7 H / m, is the vacuum magnetic permeability; μ r is the relative magnetic permeability; N is the total number of turns of the primary winding, and I is the current inside the winding;

[0075] Secondly, calculate the magnetostriction coefficient ε of the core silicon steel sheet in the magnetic field:

[0076] The relationship between the linear deformation of silicon steel sheets in a magnetic field and the magnetic field strength is:

[0077]

[0078] In the formula, H c is the coercive force; s is the saturation magnetostriction of silicon steel sheet;

[0079] From the definition of magnetostriction:

[0080]

[0081] Where: L is the initial length of the silicon steel sheet; ΔL is the change in length of the silicon steel sheet; is the expansion coefficient of silicon steel sheet;

[0082] Therefore, the magnetostriction of silicon steel sheet is expressed as:

[0083]

[0084] Among them, V0 is the loading voltage of the reactor, B s is the magnetic flux density of the saturated reactor core, which can be expressed as: B s=B0 cos(ωt), ω is the angular frequency, B0 is the magnetic flux density amplitude under the harmonic, the phase value is simplified here, and according to Faraday's law of electromagnetic induction, the loading voltage of the reactor is proportional to the rate of change of the magnetic flux. According to formula (16), the magnetostriction of the silicon steel sheet is proportional to the square of the loading voltage of the reactor, and the frequency is twice the voltage frequency. Therefore, the large magnetostriction of the silicon steel sheet means that its size changes greatly under the action of the magnetic field, which will cause the transformer to generate more noise and vibration during operation.

[0085] Magnetostriction refers to the ratio of the length change of a material under the action of a magnetic field to its original length. For silicon steel sheets, the specific value of magnetostriction varies depending on factors such as material composition, production process, and working conditions, and is usually in the order of one millionth to one ten-thousandth.

[0086] Therefore, a vibration signal of a certain frequency is used to determine whether it is core vibration.

[0087] Further, including:

[0088] The analysis steps of the reactor case surface vibration characteristics include:

[0089] Constructing the reactor surface vibration signal model:

[0090] The surface vibration of the reactor is a response to the vibration of the winding. The winding vibration is transmitted to the reactor surface through the solid component path. The vibration signal will be attenuated to a certain extent during the transmission process. The reduction of the winding preload will cause the winding to loosen, which will weaken the transmission capacity of the winding vibration and reduce the stiffness of the reactor structure. The natural frequency of the overall structure of the reactor will decrease, and the resonant frequency will decrease accordingly. Therefore, the resonant frequency is the frequency corresponding to the peak value in the reactor frequency response curve. Based on the above analysis, the reactor surface vibration signal model is proposed:

[0091]

[0092] Where: v tank is the vibration signal of the reactor surface; α is the transmission capacity coefficient. When the preload is reduced, the winding becomes loose and the coefficient decreases; v e,base is the fundamental frequency component in the winding vibration; β is the resonance amplification factor, which is a variable related to the damping ratio, frequency and other parameters; v e,r It is the vibration part of the winding vibration at the resonant frequency of the reactor; It is the vibration part of the winding except the fundamental frequency and resonant frequency; v x It is random noise, which comes from environmental interference.

[0093] Further, including:

[0094] The step of analyzing the vibration characteristics of the reactor housing surface also includes:

[0095] The vibration characteristics are analyzed based on the constructed reactor surface vibration signal model:

[0096] From formula (17), we know that the surface vibration signal of the reactor contains four responses:

[0097] The first term is the fundamental frequency component v of the winding vibration e,base The response αv e,base , which is proportional to the square of the current;

[0098] The second term is the vibration of the reactor surface at the resonant frequency βαv e,r , even if the winding vibration contains only a small amount of vibration component v in the resonant frequency band e,r , which will also generate vibration on the surface of the reactor;

[0099] The third item is the vibration part of the winding except the fundamental frequency and resonant frequency band. Response

[0100] The fourth term is random noise;

[0101] When the resonant frequency and the fundamental frequency coincide, the first and second terms are combined into one response;

[0102] Therefore, the winding compression state affects the leakage magnetic field distribution, the winding vibration transmission capacity and the reactor resonance frequency, and there are differences in the surface vibration response of the reactor under different compression states.

[0103] Further, including:

[0104] The vibration signals with obvious fault characteristics include: in the vibration of the reactor winding, the vibration frequency that resonates at the fundamental frequency; in the vibration of the iron core, the loading voltage and vibration frequency that lead to a large magnetostriction coefficient; and in the vibration of the reactor surface, a vibration signal that makes the fundamental frequency component of the winding vibration respond larger, a vibration signal with a large vibration component in the resonance frequency band contained in the winding vibration, a vibration signal with a large random noise, and a vibration signal with a large response of the vibration part of the winding vibration other than the fundamental frequency and the resonance frequency band.

[0105] Further, including:

[0106] The system also includes: a noise simulation analysis module, which is used to adopt an electromagnetic-structural-noise multi-physics field coupling analysis method according to the vibration conditions of the reactor to achieve full-process simulation of the reactor vibration noise.

[0107] Further, including:

[0108] The simulation process comprises the following steps:

[0109] S1 Establish a 1:1 real data reactor 3D simulation model in SolidWorks;

[0110] S2 set up the transient magnetic field solver in Ansoft Maxwell, set the time solution step to 0.5ms, obtain the time domain waveform of the electromagnetic force through post-processing and record the data;

[0111] S3 performs FFT transformation on the electromagnetic force in MATLAB to obtain the frequency, amplitude and phase of the electromagnetic force;

[0112] S4 In ANSYS Workbench, the electromagnetic force calculated by the electromagnetic field is used as the load excitation, and the structure-noise coupling transfer file is defined at the same time;

[0113] S5 In ANSYS Acoustics, the coupled noise sources are defined, the structural field calculation results are received, the calculations are solved, and the calculated simulation results are compared with the experimental data.

[0114] Further, including:

[0115] The electromagnetic force is obtained on the basis of the vector magnetic potential, and various physical quantities of the electromagnetic field are obtained through transformation and post-processing. It is assumed that the current density in the winding is constant, and the electromagnetic force is solved using the virtual displacement principle:

[0116]

[0117] Where: J is the current density, A is the vector magnetic potential, dg is the virtual displacement;

[0118] The vector magnetic potential A is obtained through the three-dimensional transient electromagnetic field equation:

[0119]

[0120] Where: v is the magnetic resistivity of the material, J is the winding current density, is the eddy current density caused by electromagnetic induction.

[0121] Further, including:

[0122] The method of finding a vibration signal with obvious fault characteristics and comparing it with the test result when the winding is working, and further diagnosing the fault of the reactor specifically includes:

[0123] Based on the reactor vibration experiment, a multi-channel dynamic signal acquisition analyzer is used to collect vibration signal data under different times and working conditions. The different working conditions include: winding vibration signal, box surface vibration signal and winding core vibration signal; different times refer to signal data divided by seasons, and the time span is at least two years;

[0124] The signal data obtained under different working conditions and at different times are compared with the vibration signal with obvious fault characteristics obtained by simulation calculation to obtain the accuracy of the vibration signal with obvious fault characteristics obtained by simulation calculation. If the accuracy is low, adjustments are made.

[0125] Further, including:

[0126] The signal data obtained under different working conditions and at different times are compared with the vibration signal with obvious fault characteristics obtained by simulation calculation, specifically including:

[0127] From the time dimension, under the same working conditions, the signal data within a year is divided into spring and autumn signal data, winter signal data and summer signal data, and the signal data in the same season are compared. If there is obvious abnormality, it is determined to be abnormal signal data;

[0128] From the working condition dimension, the signal data of different working conditions in the same time period are compared, and the corresponding abnormal signal data are obtained according to the relationship between different working conditions. In this embodiment, for example, the vibration of the reactor surface is caused by the vibration velocity component perpendicular to the reactor surface, that is, the axial electromagnetic force causes the top of the reactor to vibrate, and the radial electromagnetic force causes the surrounding of the reactor to vibrate. Therefore, collecting vibration signals on the reactor surface can also indirectly reflect the internal operating status of the winding.

[0129] The abnormal signal data of the corresponding working condition is numerically compared with the vibration signal with obvious fault characteristics obtained by simulation calculation. If the difference ratio is greater than the set threshold, the vibration signal data with obvious fault characteristics corresponding to the simulation result is adjusted; otherwise, the vibration signal with obvious fault characteristics is maintained and used as the fault reference value of the reactor.

[0130] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0131] This application analyzes and studies the generation mechanism and propagation path of the internal, core and surface vibrations of the reactor, thereby obtaining the corresponding models of the internal, core and surface vibrations of the reactor, laying a theoretical foundation for subsequent signal monitoring and fault diagnosis. On the basis of theoretical research, the vibration changes caused by changes in the mechanical properties of the reactor windings are analyzed to obtain vibration signals with obvious fault characteristics, and a model of the relationship between the reactor fault and the characteristic quantities of its vibration signal is established. Through a large number of theoretical and experimental studies, the reactor fault vibration model is optimized, and the model is simulated and calculated to verify the correctness of the theoretical research results.

[0132] The present invention establishes a model of the relationship between the reactor fault and the characteristic quantities of its vibration signal. From the model, the vibration characteristics of the reactor during the fault latent period can be obtained, so as to study the direct or indirect relationship between its latent characteristics and fault characteristics, and adjust the relevant parameters of the reactor in advance, thereby preventing the occurrence of faults to a certain extent.

[0133] This application combines laboratory and field data collection to perform multi-faceted parameter corrections on vibration signals with obvious fault characteristics obtained through vibration models and simulation analysis results, improves the diagnostic model, and thus improves the accuracy of diagnosis. Specifically: This application compares the obtained vibration signals with obvious fault characteristics with the actual measured data, and makes adaptive adjustments based on factors such as season to make them more accurate and more in line with actual production needs, as well as provide a basis for future monitoring of vibration signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0134] Figure 1 This is a schematic diagram of the structure of a vibration state simulation monitoring system for a dry-type reactor according to Embodiment 1 of the present invention;

[0135] Figure 2 This is a schematic diagram of a simplified kinetic model of a single wire cake according to Example 1 of the present invention;

[0136] Figure 3 This is a schematic diagram of the winding dynamics model described in Example 1 of the present invention;

[0137] Figure 4 This is a schematic diagram of the structure of a vibration state simulation monitoring system for a dry-type reactor according to Embodiment 2 of the present invention;

[0138] Figure 5 This is a flow chart of the simulation method described in Example 2 of the present invention;

[0139] Figure 6 A three-dimensional finite element equivalent model diagram of the reactor winding according to Embodiment 2 of the present invention;

[0140] Figure 7 This is a three-dimensional finite element equivalent model diagram of the reactor core described in Example 2 of the present invention. DETAILED DESCRIPTION

[0141] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0142] Example 1

[0143] The first aspect of the present invention, as Figure 1 As shown, the invention provides a vibration state simulation monitoring system for a dry-type reactor, the system comprising:

[0144] A model drawing module is used to draw a reactor model using drawing software. The reactor model is drawn by first drawing the structural rods, insulating spacers, iron cores and coils, and then assembling the parts into a reactor model through the assembly function of the drawing software;

[0145] Model import module, used to import the drawn reactor model into the finite element software and set the parameters of each part;

[0146] The modal analysis module is used to determine and analyze the vibration characteristics of the reactor structure, wherein the vibration characteristics of the reactor structure include the vibration characteristics of the reactor winding, the vibration characteristics of the reactor core, and the vibration characteristics of the reactor box surface.

[0147] In this embodiment, the vibration of the reactor is mainly generated by the cooling system and the body, and the vibration of the cooling system is caused by the rotation of the heat sink and the fan.

[0148] The vibrations generated by the reactor cooling system include the air flow around the heat sink and the vibrations caused by aerodynamic reasons when the fan rotates.

[0149] Among them, aerodynamic vibration is the main part, which is manifested as periodic exhaust vibration caused by pressure pulsation caused by the periodic rotation of the fan impeller to exhaust gas outward, and eddy current vibration caused by the splitting of gas vortex on the fan blade surface. Its vibration mode is relatively simple, and the frequency generated shows medium and low frequency characteristics, that is, the vibration frequency is below 100Hz, which is quite different from the vibration characteristics of the reactor body with 100Hz as the multiple frequency, and it is also easier to distinguish.

[0150] The vibration sources of the reactor body are the winding and the iron core, and the vibration signals of both are transmitted to the surface through the liquid medium and the solid medium. The winding contacts the pad and the iron core through the cardboard, and the vibrations generated by the two interact and influence each other with the solid medium in the middle as the transmission path, and their vibration signals are transmitted to the surface of the reactor.

[0151] The main reasons for the vibration of the reactor are the electromagnetic force on the winding, magnetostriction of the iron core, and the electromagnetic force on the iron core, which are transmitted to the surface through the reactor support components and other paths. The axial electromagnetic force of the winding and the electromagnetic attraction between the core cakes cause the top of the reactor to vibrate through solid transmission paths such as the yoke on the core and various clamps. In practice, the vibration of the reactor will continue to attenuate during the propagation process, and due to the complex structure of the reactor, it is easy to cause the refraction and reflection of the vibration wave, which brings difficulties to the analysis of the vibration propagation law of the reactor. Therefore, based on theoretical analysis, combined with experimental methods, the vibration characteristics of the reactor can be more effectively studied.

[0152] Similar to the structure of the reactor winding, the reactor winding is wound with copper wire and arranged in a self-adhesive transposition form, presenting a tangled pancake structure and a cylindrical shape. AC voltage is passed through both ends of the reactor winding, generating current and establishing a magnetic field in space. In the space where the leakage magnetic field is distributed, the magnetic lines of force can be decomposed into radial and axial lines along the winding. The radial magnetic field and current interact to generate axial electromotive force, and the axial magnetic field and current interact to generate radial electromotive force. Under the action of the electromotive force, the winding vibrates. In order to effectively analyze the vibration of the winding, it can be decomposed into axial and radial lines. Among them, the winding in this embodiment is composed of a coil-wound coil structure to form a mass unit, the insulating pads between the coils form a spring unit, and other external factors of the reactor form a damping unit. Therefore, the axial vibration of the winding can be equivalent to a multi-degree-of-freedom mass-spring-damping model. After analysis, it can be seen that if the clamping force is too small, the winding will not resonate, but it will also cause insufficient clamping force and looseness; if the preload force is too large, it is easy to approach the multiple of 100Hz while being away from the 100Hz frequency, causing resonance. Therefore, it is necessary to reasonably calculate and select an appropriate preload force to ensure the normal operation of the reactor winding without causing looseness.

[0153] Therefore, in summary, if the reactor winding is loose, the vibration of the winding will increase; when a momentary short circuit occurs and causes the winding insulation to break down, the vibration of the winding will also increase.

[0154] Similarly, after one of the high or low voltage windings is deformed, displaced or collapsed, the winding is not tightened enough, causing the height difference between the high and low voltage windings to gradually expand, resulting in increased ampere-turn imbalance of the winding, increased axial force caused by leakage magnetic flux, and thus increased vibration of the winding.

[0155] The compression and deformation of the windings have a great relationship with the vibration of the reactor. Therefore, the compression of the reactor core and the windings can be reflected in the vibration of the reactor to a large extent. If the reactor core and winding compression are in good condition, the reactor vibration will be relatively stable, which also provides a basis for eliminating the fault of increased reactor vibration.

[0156] When the reactor is working stably, it can be considered that the vibration of the reactor body is caused by the vibration of the iron core, and the change of vibration is also caused by the iron core failure. When a short circuit occurs in the reactor, a large impact current flows through the winding coil, and the vibration signal of the winding is much larger than the core vibration. Therefore, under the short circuit condition, the vibration of the reactor body is mainly caused by the vibration of the winding. In this way, the vibration signal of the reactor body during / after the short circuit accident can be used to monitor whether the winding coil is deformed or loose.

[0157] Therefore, this embodiment studies and constructs a vibration model of the reactor, specifically including:

[0158] First, the analysis steps of the reactor winding vibration characteristics can include:

[0159] S1 establishes the reactor winding vibration model: Figure 2 As shown in the figure, the winding of a large power reactor is composed of a series of copper coils stacked together. Adjacent copper coils are separated by insulating pads. There are iron yokes on the upper and lower parts of the winding to compress the winding. When the reactor is working, the current in the winding interacts with the leakage magnetic field to generate electromagnetic force acting on the coil. Under the excitation of the electromagnetic force, a single coil can be regarded as performing simple harmonic vibration in mechanics.

[0160] Figure 2 In the equation, m is the mass of a single coil; K s The static stiffness of the insulating spacer between the coils; R is the damping coefficient between adjacent coils; Z(t) represents the displacement of a single coil; F(t) represents the electromagnetic force on the coil. K s is the static stiffness of the pads between each coil unit, and R is the damping coefficient related to other reactors. For the overall structure of the reactor winding, the following simplifications can be made: assume that all coil units of the winding have the same mass, represented by m; the damping coefficients between adjacent coils are the same, represented by C; the stiffness between adjacent coils is the same, represented by K; the stiffness of the connection between the coils at both ends of the winding and the iron yoke is different. Combining the above analysis with the single coil model, the overall dynamic model of the winding can be obtained, such as Figure 3 Shown

[0161] S2 represents the vibration acceleration and angular frequency of the reactor winding dynamic model based on the overall winding dynamic model.

[0162] In this embodiment, the overall dynamic model of the winding is expressed as:

[0163]

[0164] Among them, z n is the displacement of the nth coil; C is the damping coefficient between adjacent coils of the reactor; K B is the stiffness coefficient between the coil and the upper iron yoke; K is the stiffness coefficient between adjacent coils; K H F is the stiffness coefficient between the coil and the iron yoke at the lower end; n is the electromagnetic force on the nth coil unit; mg is the mass of a single coil and the mass of each coil is the same; is the inertia force of the nth coil unit; is the damping force on the nth coil unit; Kz n is the elastic force on the nth coil unit;

[0165] The method of expressing the vibration acceleration and angular frequency of the reactor winding dynamics model based on the winding overall dynamics model specifically includes:

[0166] The stiffness coefficient between adjacent coils corresponding to each coil, the displacement of the coil and the electromagnetic force received are made equal, so as to add each equation in formula (1) to obtain:

[0167]

[0168] Among them, M = nm is the mass of the coil, C' = nc is the damping coefficient of the reactor coil, K' = k B +k z =2k is the elastic coefficient of the insulating pad, z are the acceleration, velocity and displacement of the axial motion of the coil, respectively, and F=F1+F2+,…,+F n ;

[0169] According to the law of electromagnetic induction and the left-hand rule, assuming that the current I = I0cos(ω0t), the axial electromotive force on the winding is:

[0170]

[0171] Wherein, k is the calculation parameter; I0 is the current amplitude, ω0 is the angular frequency;

[0172] Combining formula (2) and formula (3), we get:

[0173]

[0174] Formula (4) is a non-homogeneous linear differential equation with constant coefficients. Its solution is the general solution of the homogeneous equation plus the special solution of the non-homogeneous equation. Therefore, the homogeneous equation of formula (4) is:

[0175]

[0176] The general solution is:

[0177]

[0178] Among them, p, is a parameter related to the initial state of the winding, f=ω / (2π) is the natural frequency of the winding and complies with 4MK'>>C' 2 ,therefore,

[0179] The non-homogeneous equation is decomposed into two:

[0180]

[0181] Find their special solutions respectively:

[0182]

[0183] Therefore, in formula (8):

[0184]

[0185] Therefore, the general solution of equation (8) is obtained:

[0186]

[0187] By taking the derivative of formula (9) twice, the corresponding vibration acceleration can be obtained:

[0188]

[0189] and

[0190] S3 analyzes the relationship between the magnitude of the preload force and the vibration acceleration and angular frequency, thereby finding relevant measuring points to avoid resonance.

[0191] Furthermore, in this embodiment, the relationship between the magnitude of the preload force and the vibration acceleration and angular frequency is analyzed to find relevant measuring points to avoid resonance, including:

[0192] According to formulas (10) and (11),

[0193] If K'>4Mω0 2 , that is, ω>2ω0, the amplitude of a decreases as the clamping force decreases (the winding gradually loosens);

[0194] If K'<4Mω0 2 , that is, ω<2ω0, the amplitude of a increases as the clamping force decreases (the winding gradually loosens);

[0195] If K'=4Mω0 2 , that is, ω=2ω0, resonance will occur at the fundamental frequency;

[0196] Therefore, if a measuring point where no resonance occurs can be found, that is, ω is always greater than 2ω0 or ω is always less than 2ω0, the influence of resonance can be avoided. That is, according to the above analysis, a vibration signal with obvious fault characteristics can be found, that is, the frequency and other related parameters when resonance is likely to occur. At the same time, the corresponding frequency when resonance is avoided can also be obtained, thereby realizing the diagnosis of winding looseness defects.

[0197] In this embodiment, the vibration of the reactor core is also considered, and the analysis steps of the vibration characteristics of the reactor core include:

[0198] First, analyze the electromagnetic field generated when the primary coil of the iron core is energized:

[0199] The magnetic field inside the reactor winding is regarded as a uniform magnetic field, and the magnetic field strength inside the winding is expressed as:

[0200]

[0201] The magnetic induction intensity is expressed as:

[0202] B=μ0μ r H=μ0μ r (N / L0)I (13)

[0203] Where L0 is the average length of the core loop; μ0 = 4π × 10 -7 H / m, is the vacuum magnetic permeability; μ r is the relative magnetic permeability; N is the total number of turns of the primary winding, and I is the current inside the winding;

[0204] Secondly, the magnetostriction coefficient ε of the core silicon steel sheet in the magnetic field is calculated. Due to the fact that the iron yoke and the core column of the reactor produced in recent years are bound with non-weft epoxy tape, and the core lamination method has been improved, the vibration caused by the electromagnetic force between the silicon steel sheets and the joints is relatively weak, and the vibration of the core part can be regarded as being caused only by the magnetostrictive effect.

[0205] Therefore, the relationship between the linear deformation of the silicon steel sheet in the magnetic field and the magnetic field strength is:

[0206]

[0207] In the formula, H c is the coercive force; s is the saturation magnetostriction of silicon steel sheet;

[0208] From the definition of magnetostriction:

[0209]

[0210] Where: L is the initial length of the silicon steel sheet; ΔL is the change in length of the silicon steel sheet; is the expansion coefficient of silicon steel sheet;

[0211] Therefore, the magnetostriction of silicon steel sheet is expressed as:

[0212]

[0213] Among them, V0 is the loading voltage of the reactor, B s is the magnetic flux density of the saturated reactor core, which can be expressed as: B s=B0 cos(ωt), ω is the angular frequency, B0 is the magnetic flux density amplitude under the harmonic, the phase value is simplified here, and according to Faraday's law of electromagnetic induction, the loading voltage of the reactor is proportional to the rate of change of the magnetic flux. According to formula (16), the magnetostriction of the silicon steel sheet is proportional to the square of the loading voltage of the reactor, and the frequency is twice the voltage frequency. In this embodiment, it can be expressed as:

[0214]

[0215] Among them, N1 and A are parameters, and ω is the angular frequency.

[0216] Therefore, the large magnetostriction of silicon steel sheets means that their dimensions change greatly under the action of the magnetic field, which will cause the transformer to generate more noise and vibration during operation.

[0217] In this embodiment, magnetostriction refers to the ratio of the length change of a material under the action of a magnetic field to its original length. For silicon steel sheets, the specific value of magnetostriction varies depending on factors such as material composition, production process, and working conditions, and is usually in the order of one millionth to one ten-thousandth.

[0218] Therefore, in this embodiment, since the 100Hz vibration signal comes from the winding vibration and the core vibration, it is not reliable to simply judge the mechanical properties of the winding by the fundamental frequency vibration signal, because the change of the 100Hz vibration signal may come from the core failure. Therefore, for the core vibration of the reactor, a certain frequency can be set to avoid the increase of magnetostriction and obtain the frequency corresponding to the vibration signal that makes the fault characteristics obvious.

[0219] In this embodiment, the step of analyzing the vibration characteristics of the reactor housing surface includes:

[0220] Constructing the reactor surface vibration signal model:

[0221] The surface vibration of the reactor is a response to the vibration of the winding. The winding vibration is transmitted to the reactor surface through the solid component path. The vibration signal will be attenuated to a certain extent during the transmission process. The reduction of the winding preload will cause the winding to loosen, which will weaken the transmission capacity of the winding vibration and reduce the stiffness of the reactor structure. The natural frequency of the overall structure of the reactor will decrease, and the resonant frequency will decrease accordingly. Therefore, the resonant frequency is the frequency corresponding to the peak value in the reactor frequency response curve. Based on the above analysis, the reactor surface vibration signal model is proposed:

[0222]

[0223] Where: v tank is the vibration signal of the reactor surface; α is the transmission capacity coefficient. When the preload is reduced, the winding becomes loose and the coefficient decreases; ve,base is the fundamental frequency component in the winding vibration; β is the resonance amplification factor, which is a variable related to the damping ratio, frequency and other parameters; v e,r It is the vibration part of the winding vibration at the resonant frequency of the reactor; It is the vibration part of the winding except the fundamental frequency and resonant frequency; v x It is random noise, which comes from environmental interference.

[0224] The vibration characteristics are analyzed based on the constructed reactor surface vibration signal model:

[0225] From formula (17), we know that the surface vibration signal of the reactor contains four responses:

[0226] The first term is the fundamental frequency component v of the winding vibration e,base The response αv e,base , which is proportional to the square of the current;

[0227] The second term is the vibration of the reactor surface at the resonant frequency βαv e,r , even if the winding vibration contains only a small amount of vibration component v in the resonant frequency band e,r , which will also generate vibration on the surface of the reactor;

[0228] The third item is the vibration part of the winding except the fundamental frequency and resonant frequency band. Response

[0229] The fourth term is random noise;

[0230] When the resonant frequency and the fundamental frequency coincide, the first and second terms are combined into one response;

[0231] Therefore, the winding compression state affects the leakage magnetic field distribution, the winding vibration transmission capacity and the reactor resonance frequency, and there are differences in the surface vibration response of the reactor under different compression states.

[0232] In this embodiment, the vibration signals with obvious fault characteristics mentioned above specifically include: in the vibration of the reactor winding, a vibration frequency that resonates at the fundamental frequency; in the vibration of the core, a loading voltage and vibration frequency that lead to a large magnetostriction coefficient; and in the vibration of the reactor surface, a vibration signal that makes the fundamental frequency component of the winding vibration respond larger, a vibration signal with a larger vibration component in the resonant frequency band contained in the winding vibration, a vibration signal with a larger random noise, and a vibration signal with a larger response of the vibration part of the winding vibration except the fundamental frequency and the resonant frequency band.

[0233] The diagnostic module is used to obtain the vibration response characteristics of the reactor under power-on conditions based on the vibration characteristics of each part, so as to find out the vibration signal with obvious fault characteristics, and compare it with the test results when the winding is working, so as to further diagnose the fault of the reactor.

[0234] In this embodiment, the method of finding a vibration signal with obvious fault characteristics and comparing it with the test result when the winding is working to further diagnose the fault of the reactor specifically includes:

[0235] S41 uses a multi-channel dynamic signal acquisition analyzer to collect vibration signal data under different times and working conditions based on the reactor vibration experiment. The different working conditions include: winding vibration signal, box surface vibration signal and winding core vibration signal; different times refer to signal data divided by seasons, and the time span is at least two years;

[0236] That is, in this embodiment, the data that needs to be collected include at least: the winding vibration signal, box surface vibration signal and winding core vibration signal data in the spring and autumn of the first year, the winding vibration signal, box surface vibration signal and winding core vibration signal data in the summer of the first year, the winding vibration signal, box surface vibration signal and winding core vibration signal data in the winter of the first year, the winding vibration signal, box surface vibration signal and winding core vibration signal data in the spring and autumn of the second year, the winding vibration signal, box surface vibration signal and winding core vibration signal data in the summer of the second year, and the winding vibration signal, box surface vibration signal and winding core vibration signal data in the winter of the second year, and the corresponding signal data are all marked with time.

[0237] In this embodiment, based on the reactor vibration experiment, it includes: the reactor vibration experiment platform is composed of hardware and software, the hardware includes a piezoelectric acceleration vibration sensor, a signal acquisition instrument, and a laptop computer, and the software includes a DHDAS software platform, etc.

[0238] This embodiment does not limit the specific model of the sensor, and the sensor is selected according to actual needs. In one case, two types of current sensors are selected. One type is a clamp-type current sensor, which is easy to use but has relatively low accuracy and is not suitable for measuring fast transient signals.

[0239] The other type is the Hall-type current sensor from LEM, which has particularly good dynamic performance and high accuracy. This subject uses this sensor in experiments that measure transient current signals and require high accuracy.

[0240] The selection of voltage sensor is similar to that of current sensor.

[0241] Vibration sensor: This application simultaneously measures the vibration of the reactor in different states. Based on the previous research on vibration signals, the characteristics of vibration signals were fully considered, and finally the piezoelectric acceleration vibration sensor model IA212E was selected.

[0242] The IA212E piezoelectric acceleration vibration sensor has a sensitive chip inside, which converts and outputs a charge signal proportional to the acceleration amplitude, which is displayed as a voltage amplitude on the interface; the sensor itself has a high resonant frequency, which can avoid the generation of resonant components and ensure the high quality of the original signal without signal distortion; it is light in weight, ensuring that the vibration signal at the installation location will not be changed; it has high sensitivity, and the output signal has a high signal-to-noise ratio with obvious details, which is conducive to signal analysis and subsequent fault diagnosis; the output signal line uses a shielded coaxial cable to avoid interference from the strong magnetic environment outside.

[0243] The IA212E piezoelectric acceleration vibration sensor has a magnetic disk at the bottom, which can be firmly adsorbed on the surface of the reactor from the outside (considering insulation issues), and the sensor is fixed to the core and windings internally by gluing and bundling. In the vibration measurement environment, no movement or vibration occurs, ensuring that the collected data is true and reliable.

[0244] The vibration signal acquisition and analysis instrument in this embodiment is the core hardware part of the reactor vibration experiment, taking into account multiple functions such as signal acquisition, recording, filtering, analysis and conditioning. At the same time, in order to reduce the number of points, the channels are as many as possible. After considering the above issues, the DH5992D multi-channel dynamic signal acquisition and analysis instrument was finally selected.

[0245] DH5992D multi-channel dynamic signal acquisition analyzer has 16 channels to collect vibration signals simultaneously, which meets the needs of multiple measurement points to the greatest extent, reduces the trouble of point layout and the error caused by manual movement of sensors; the sampling frequency is as high as 256kHz, collects a lot of data, and has high resolution, which is conducive to discovering the difference between subtle faults and normal conditions; it collects a variety of signal types and can realize the test and analysis of acceleration, velocity and displacement at the same time; the data preservation and portability are good, and it can output a variety of file formats; it has good stability and strong anti-interference ability.

[0246] The laptop computer is used as a terminal device. The driver and related software of the DH5992D multi-channel dynamic signal acquisition and analysis instrument are installed inside the laptop computer. It is connected to the acquisition and analysis instrument to display the vibration signal waveform and other analysis results in real time, and store the collected data in the computer.

[0247] In this embodiment, no special limitation is made to the software program, and the system requirements can be met, and the relevant functions can be increased or decreased according to the analysis and processing. In one case, the software part is an important part of the reactor vibration experiment, specifically including communication protocols, underlying drivers, etc., integrating data acquisition, order analysis, basic analysis, experimental modal analysis, impact waveform detection, on-site dynamic balance, acoustic analysis and other analysis and processing and engineering application functions. The use of partitioned management modules makes it easier and more convenient to use. It has programmable instrument range, identification system parameters and setting sampling parameters, and can complete real-time signal acquisition, analysis and processing, functional modular management, virtual instrument and other operations. It can realize data processing functions such as spectrum analysis, 3D model diagram, color cloud diagram, color waterfall diagram, cepstrum analysis, digital filtering, strain stress, mathematical operations, drawing, signal generation, modal parameter identification, etc., to meet the needs of reactor vibration signal analysis to the greatest extent.

[0248] Therefore, in this embodiment, based on the above-mentioned software and hardware, relevant signal data is collected. In one case, the vibration of the reactor is a simple harmonic motion from the inside to the outside, and the vibration signals of its normal state and fault state are different in the time domain and frequency domain. The vibration characteristics can be studied through signal processing methods, and the internal faults of the reactor can be diagnosed based on the vibration method.

[0249] In this embodiment, in order to comprehensively and specifically analyze the vibration characteristics of the reactor and the internal fault research based on the vibration method, the collected data includes at least: under different working conditions, including: winding vibration signals, box surface vibration signals and winding core vibration signals; different times refer to signal data differentiated by season, and the time span is at least two years.

[0250] In one case of this embodiment, in order to collect vibration signals inside and outside the reactor. The piezoelectric acceleration vibration sensor is fixed inside the reactor and on the surface of the reactor by bundling adhesion and magnetic adsorption. The signal acquisition instrument converts the charge signal output by the acceleration vibration sensor into a voltage signal, and then transmits and stores the data in a computer. In one case, the sampling frequency of this piezoelectric acceleration vibration sensor is set to 16000Hz. The acquisition process adopts a step-by-step voltage increase method, which is set to 10 voltage levels of 20%, 40%, 60%, 80%, 100%, 110%, 120%, 130%, 140%, and 150% of the rated voltage.

[0251] In one case of this embodiment, in order to collect the vibration signal of the winding core, the internal and external vibration signals when the winding core is not loose, the internal and external vibration signals when the winding core is 60% loose, and the internal and external vibration signals when the reactor winding core is completely loose can be collected. The relevant data in these three cases. In this embodiment, the collection of the internal and external vibration signals when the reactor winding core is 60% loose is based on the internal structural characteristics of the special reactor equivalent model, and the pre-tightening force of the loose through-core screw is used to simulate the fault state of the winding core being loosened by 60%. The collection of the internal and external vibration signals when the reactor winding core is completely loose is based on the internal structural characteristics of the special reactor equivalent model, and the pre-tightening force of the through-core screw is set to 0 to simulate the fault state of the winding core being completely loose.

[0252] S42 compares the signal data obtained under different working conditions and at different times with the vibration signal with obvious fault characteristics obtained by simulation calculation to obtain the accuracy of the vibration signal with obvious fault characteristics obtained by simulation calculation. If the accuracy is low, adjustments are made.

[0253] In this embodiment, the collected signal data is compared with the vibration signals with obvious fault characteristics obtained from different reactor structures in a multi-dimensional and all-round manner. This is determined by the particularity of the collected signal data. First, the base temperature of the reactor is different in different seasons. The base temperature of the reactor used in summer is high, and the base temperature of the reactor used in winter is low. Changes in ambient temperature have a significant impact on the vibration characteristics. Temperature changes indirectly change the vibration characteristics by affecting the physical properties of the material, especially the elastic modulus and density of the material, which are key factors in determining the vibration frequency and mode. Therefore, in addition to frequency, temperature changes also affect the shape of the vibration mode. At high temperatures, the reduction in the elastic modulus of the material may lead to a softening of the vibration mode, that is, the frequencies of the low-order modes are closer, which may lead to enhanced coupling between modes and affect the dynamic response of the structure.

[0254] Few people in the prior art have considered this situation, but the reactor is a device that is relatively sensitive to temperature, and the temperature will rise during use. After comprehensive consideration of various situations, this application believes that the abnormal signal data cannot be uniformly set as a value, but it is necessary to analyze the signal data of the reactor in different seasons or under different environmental influences. The abnormal signal data in this embodiment can be the size of the abnormal noise caused by the corresponding vibration when used at the current frequency under the current spring preload.

[0255] From the time dimension, under the same working condition, the signal data within a year is divided into spring and autumn season signal data, winter season signal data and summer season signal data, and the signal data in the same season are compared. If it is obviously abnormal, it is determined to be abnormal signal data; in this embodiment, this consideration is mainly due to the influence of temperature on the reactor. The basic temperature of the reactor used in summer is high, and the basic temperature of the reactor used in winter is low. Therefore, the abnormal signal data cannot be uniformly set as a value, but the signal data of the reactor under different seasons or environmental influences must be analyzed. The abnormal signal data in this embodiment can be the size of the abnormal noise corresponding to the use at the current frequency under the spring preload.

[0256] Based on the above description, in this embodiment, when designing and using a three-phase reactor, it is necessary to consider the influence of the reactor's temperature on its inductance value, and to eliminate the abnormal noise that occurs in this case in real time, mainly by taking corresponding measures to ensure the stability and accuracy of the reactor. For example, high-temperature insulating materials can be used to improve the high-temperature resistance of the reactor, or a temperature sensor can be added inside the reactor to monitor the temperature change of the reactor in real time, so as to adjust the working state of the reactor in time. And in this embodiment, if abnormal signal data appears under the current preload force and frequency, and the staff processes the abnormal signal data, it will return to normal at the next moment. Otherwise, more abnormal signal data will continue to be displayed, so as to issue an early warning and remind the staff to make corresponding processing. Regardless of the situation, the corresponding abnormal signal data can be collected.

[0257] From the working condition dimension, the signal data of different working conditions in the same time period are compared, and the corresponding abnormal signal data are obtained according to the relationship between different working conditions; as in the coupling situation involved in Example 2, the vibration of the reactor surface is caused by the vibration velocity component perpendicular to the reactor surface, that is, the axial electromagnetic force causes the vibration of the top of the reactor, and the radial electromagnetic force causes the vibration around the reactor. Therefore, collecting vibration signals on the reactor surface can also indirectly reflect the internal operating status of the winding.

[0258] In this embodiment, on the basis of the above, the abnormal signal data or the average value of the abnormal signal of the corresponding working condition is numerically compared with the vibration signal with obvious fault characteristics obtained by simulation calculation under the same parameter conditions. The average value of the abnormal signal in this embodiment refers to the existence of multiple corresponding abnormal signal data under the same working condition in the same time period, such as the abnormal signal data set from 8 am to 12 pm every day or every few days in spring and autumn. In this case, the average value is taken. Otherwise, if there is only one abnormal signal data, the current value is used for calculation. In this case, if the difference ratio is greater than 5%, the corresponding vibration signal data with obvious fault characteristics is adjusted. The abnormal signal data value shows that the acceleration value under the current preload and frequency is 0.1g, and the acceleration value of the vibration signal with obvious fault characteristics obtained by the simulation experiment under the above parameter conditions is 0.102g, then the difference ratio is |0.1-0.102| / 0.1=2%, and the result is less than 5%. The simulation data has high accuracy and no adjustment is required.

[0259] However, in the embodiment, it is necessary to limit that even if the parameter conditions are different, the difference ratio shall not be higher than 20%, otherwise, it is determined that the vibration signal value with obvious fault characteristics obtained by the simulation experiment is wrong. For example, if under different parameter conditions, the acceleration value of the vibration signal with obvious fault characteristics obtained by the simulation experiment is 0.15g under the above parameter conditions, then the difference ratio is 50%, and it is determined that the vibration signal value with obvious fault characteristics obtained by the simulation experiment is wrong. According to the season and other parameter conditions, the corresponding vibration signal data with obvious fault characteristics shall be adjusted; otherwise, the vibration signal with obvious fault characteristics shall be maintained and used as the fault reference value of the reactor.

[0260] In this embodiment, the seasonal conditions are such as high temperature in summer, the base temperature of the reactor will increase accordingly, and the corresponding vibration signal will also be different, but as long as the difference with the historical signal data under the same conditions is small, it can still be considered as normal data.

[0261] Example 2

[0262] like Figure 4 As shown, the system also includes: a noise simulation analysis module, which is used to adopt an electromagnetic-structural-noise multi-physics field coupling analysis method according to the vibration condition of the reactor to achieve a full-process simulation of the reactor vibration noise. This module is considered in the diagnosis module in Example 1.

[0263] like Figure 5 The simulation process of this embodiment includes the following steps:

[0264] S1 In SolidWorks, a 1:1 real data reactor 3D simulation model is established.

[0265] The accuracy of the finite element vibration characteristic analysis of the reactor lies in the establishment of the model. After comparing various 3D modeling software, SolidWorks software was finally selected.

[0266] SolidWorks software adopts a graphical user interface and intelligent parametric design concepts, has excellent geometric modeling capabilities and analysis functions, fast calculation speed, simple operation process, and convenient design process. It has become the mainstream software for 3D modeling and design in structural design and mechanical drawing, and has been widely used in various fields such as industrial production, machinery manufacturing, aerospace, etc.

[0267] This application uses SolidWorks software to establish the three-dimensional model of the reactor for the following reasons and makes full use of the following features of SolidWorks software:

[0268] (1) The reactor has a complex structure, with many dimensions and various parameters, making it difficult to use coordinate drawing software. SolidWorks is a software based on parametric CAD design. During the modeling process, the part dimensions can be expressed in the form of parameters, and the parameter values ​​can be modified at any time to directly change the part's external dimensions, which greatly facilitates users to improve the model design in a timely manner.

[0269] (2) The reactor has a complex structure and it is difficult to draw the entire model directly on one operating interface at the same time. It needs to be modeled separately and then assembled. The assembly modeling of SolidWorks meets the needs very well. In the process of creating an assembly, you can complete the assembly by selecting the vertices, edges, and surfaces of each part to match the parts; automatically assemble accessories and fasteners of each part; automatically assemble common hardware, parts, and required equipment features; customize the mechanical properties and relationships between parts, etc., to achieve the perfect assembly of a complex model.

[0270] (3) After the reactor 3D model is completed, it needs to be imported into finite element analysis software such as Maxwell and ANSYS. SolidWorks software provides a variety of formats for saving. First, you can directly link to ANSYS in the SolidWorks operation interface to create independent Geometry; second, save the 3D model in Parasolid (*.x_t) format and read the 3D model from the finite element software.

[0271] (4) SolidWorks can generate engineering drawings directly from 3D models, including dimensions, annotations, and views.

[0272] (5) Friendly user interface and excellent user experience. SolidWorks provides a complete set of mouse drag control and dynamic interface functions. The "full-motion" user interface eliminates unnecessary animation boxes, reduces redundant design steps, and avoids disorder and clutter in the interface.

[0273] In this embodiment, the special reactor winding structure is mainly composed of coils, insulating pads between adjacent coils, and upper and lower pads. The coils are wound by multiple turns of copper wire, and there is a transposition phenomenon. When establishing the three-dimensional finite element model of the winding, the winding should be equivalent to a circle of solid rings according to the winding structure, ignoring the influence of inter-turn insulation and coil transposition. However, due to the high axial height and large radial radius of the winding, it is difficult for the computer to run in the subsequent finite element simulation. On the basis of fully considering that it does not affect the calculation results, the reactor winding is finally equivalent to a cylindrical shape, such as Figure 6 shown.

[0274] Furthermore, in the present embodiment, the special reactor core is mainly composed of a core column and an iron yoke, and also includes components such as screws, clamps, pressure beams, upper and lower beams and feet. The core column is composed of a core cake and a marble pad. In practice, the core cake is made up of circular silicon steel sheets stacked up and down, and the silicon steel sheets are sealed by pouring epoxy resin. Therefore, it can be completely equivalent to a round cake shape during modeling, and the marble pad is built into a round cake shape of the same model according to the original size. The iron yoke is divided into an upper and lower yoke and a side yoke, which is also made up of silicon steel sheets stacked front and back. Epoxy resin is poured to fill the gap to complete the sealing. Therefore, it can be equivalent to a square frame structure during modeling. Taking full account of the accuracy of the results and the operating capacity of the computer, according to the size of other components of the reactor core, an original-size component model is established, and it is assembled into a complete core model, such as Figure 7 shown.

[0275] S2 set up the transient magnetic field solver in Ansoft Maxwell, set the time solution step to 0.5ms, obtain the time domain waveform of the electromagnetic force through post-processing and record the data.

[0276] In this embodiment, after constructing the overall three-dimensional finite element equivalent model of the special reactor, Ansoft Maxwell electromagnetic simulation software is used to analyze the magnetic field distribution and force characteristics of the reactor.

[0277] Ansoft Maxwell is an industrial-grade electromagnetic simulation software, which is widely used in various power equipment such as transformers, motors, sensors and other industrial control systems. It has a guided user interface, precision-driven adaptive segmentation technology and powerful post-processing functions, especially in 3D. It can automatically calculate parameters such as current density, loss, inductance, torque, etc., and provide cloud maps and vector diagrams of physical quantities such as magnetic field and electromagnetic force, as well as various curves. Therefore, it can well meet the requirements of reactor field-circuit coupling calculation.

[0278] S3 performs FFT transformation on the electromagnetic force in MATLAB to obtain the frequency, amplitude and phase of the electromagnetic force;

[0279] S4 In ANSYS Workbench, the electromagnetic force calculated by the electromagnetic field is used as the load excitation, and the structure-noise coupling transfer file is defined at the same time;

[0280] S5 In ANSYS Acoustics, the coupled noise sources are defined, the structural field calculation results are received, the calculations are solved, and the calculated simulation results are compared with the experimental data.

[0281] The field-circuit coupling analysis of the reactor mainly involves the coupling of two physical quantities, the electromagnetic field and the circuit. In the field-circuit coupling analysis, the electromagnetic field theory of the reactor satisfies Maxwell's equations, including Ampere's circuit law, Faraday's law of electromagnetic induction, Gauss's electric flux law, and Gauss's magnetic flux law.

[0282] (1) Ampere's circuit law

[0283] The line integral of the magnetic field intensity in the reactor magnetic field along any closed loop is equal to the sum of the currents enclosed by the surface determined by the loop. This current includes conduction current and displacement current, indicating that both current and changing electric field can generate magnetic field.

[0284]

[0285] Where: H is the magnetic field intensity, L is any closed loop, S is the surface determined by the closed loop, J is the conduction current density vector, is the displacement current density, and D is the electric flux density.

[0286] (2) Faraday’s law of electromagnetic induction

[0287] The magnitude of the induced potential in the closed loop of the reactor circuit is proportional to the rate of change of the magnetic flux passing through the loop, indicating that a changing magnetic field can generate an electric field.

[0288]

[0289] Where: E is the electric field intensity, B is the magnetic induction intensity.

[0290] (3) Gauss's law of flux

[0291] The electric flux through any closed surface in the reactor is equal to the amount of charge enclosed by the surface, indicating that charges can generate electric fields.

[0292]

[0293] Where: V is the volume area enclosed by the closed surface, and ρ is the charge density.

[0294] (4) Gauss's flux law

[0295] The magnetic flux passing through any closed surface in the reactor must be equal to zero, indicating that the magnetic field is a passive field and the magnetic lines of force are always closed.

[0296]

[0297] In actual operation, the two ends of the reactor winding are connected with industrial frequency AC voltage, and the magnetic field is time-varying. Ignoring the displacement current and skin effect, the three-dimensional transient electromagnetic field equation is obtained on the basis of satisfying Maxwell's equations:

[0298]

[0299] Where: A is the vector magnetic potential, v is the magnetic resistivity of the material, J is the winding current density, is the eddy current density caused by electromagnetic induction.

[0300] Based on the magnetic potential calculated by the finite element method, various physical quantities of the electromagnetic field are obtained through transformation and post-processing. Assuming that the current density in the winding is constant, the electromagnetic force is solved using the virtual displacement principle:

[0301]

[0302] Where: J is the current density, A is the vector magnetic potential, and dg is the virtual displacement.

[0303] Therefore, based on the above analysis, the following results are obtained: the core and winding models are imported into Maxwell, and the transient electromagnetic field solver is selected to calculate and analyze the reactor. The solution calculation is completed through material selection, excitation addition, boundary condition setting, etc. Specifically: the magnetic field of the reactor is distributed along the core, mainly concentrated on the inner side of the core cake and the iron yoke corner, with a maximum value of 1.4038T. During the operation of the reactor, the direction of the electromagnetic force does not change, because the electromagnetic force is generated by the joint action of the current and the magnetic field. In one cycle of the current, the magnetic field and the current keep changing at the same time. According to the left-hand rule, the direction of the electromagnetic force remains unchanged. Due to the high winding of the UHV reactor, the electromagnetic force is distributed outward, the axial electromagnetic force on both sides is greater than the middle, and the radial electromagnetic force on both sides is less than the middle. The vibration of the reactor surface is caused by the vibration velocity component perpendicular to the reactor surface, that is, the axial electromagnetic force causes the vibration of the top of the reactor, and the radial electromagnetic force causes the vibration of the reactor around. Therefore, collecting vibration signals on the surface of the reactor can reflect the operating state inside the winding.

[0304] Therefore, in order to detect minor faults early and accurately, this system proposes a signal data monitoring method, which can effectively suppress the interference of vibration signals. It also combines laboratory and field data collection to make multi-faceted parameter corrections to the established diagnostic model, thereby improving the diagnostic model and the accuracy of fault diagnosis.

[0305] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0306] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0307] Obviously, those skilled in the art can make various changes and modifications to the embodiments of the present invention without departing from the spirit and scope of the embodiments of the present invention. Thus, if these modifications and variations of the embodiments of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A vibration state simulation monitoring system for a dry-type reactor, characterized in that: The system includes: A model drawing module is used to draw a reactor model using drawing software. The reactor model is drawn by first drawing the structural rods, insulating spacers, iron cores and coils, and then assembling the parts into a reactor model through the assembly function of the drawing software; Model import module, used to import the drawn reactor model into the finite element software and set the parameters of each part; A modal analysis module, used to determine and analyze the vibration characteristics of the reactor structure, wherein the vibration characteristics of the reactor structure include the vibration characteristics of the reactor winding, the vibration characteristics of the reactor core, and the vibration characteristics of the reactor housing surface; The diagnostic module is used to obtain the vibration response characteristics of the reactor under power-on conditions based on the vibration characteristics of each part, so as to find out the vibration signal with obvious fault characteristics, and compare it with the test results when the winding is working, so as to further diagnose the fault of the reactor.

2. The vibration state simulation monitoring system of the dry-type reactor according to claim 1, characterized in that: The analysis steps of the reactor winding vibration characteristics include: Establish the vibration model of the reactor winding: When the reactor is working, the current in the winding interacts with the leakage magnetic field, generating electromagnetic force acting on the coil. Under the excitation of the electromagnetic force, a single coil is regarded as performing simple harmonic vibration in mechanics, thus representing the overall dynamic model of the winding; Based on the winding overall dynamic model, the vibration acceleration and angular frequency of the reactor winding dynamic model are expressed; The relationship between the magnitude of the preload force and the vibration acceleration and angular frequency is analyzed to find relevant measuring points to avoid resonance.

3. The vibration state simulation monitoring system of the dry-type reactor according to claim 2, characterized in that: The overall dynamic model of the winding is expressed as: Among them, z n is the displacement of the nth coil; C is the damping coefficient between adjacent coils of the reactor; K B is the stiffness coefficient between the coil and the upper iron yoke; K is the stiffness coefficient between adjacent coils; K H F is the stiffness coefficient between the coil and the iron yoke at the lower end; n is the electromagnetic force on the nth coil unit; mg is the mass of a single coil and the mass of each coil is the same; is the inertia force of the nth coil unit; is the damping force on the nth coil unit; Kz n is the elastic force on the nth coil unit; The method of expressing the vibration acceleration and angular frequency of the reactor winding dynamics model based on the winding overall dynamics model specifically includes: The stiffness coefficient between adjacent coils corresponding to each coil, the displacement of the coil and the electromagnetic force received are made equal, so as to add each equation in formula (1) to obtain: Among them, M = nm is the mass of the coil, C' = nc is the damping coefficient of the reactor coil, K' = k B +k z =2k is the elastic coefficient of the insulating pad, z are the acceleration, velocity and displacement of the axial motion of the coil, respectively, and F=F1+F2+,…,+F n ; According to the law of electromagnetic induction and the left-hand rule, assuming that the current I = I0cos(ω0t), the axial electromotive force on the winding is: Wherein, k is the calculation parameter; I0 is the current amplitude, ω0 is the angular frequency; Combining formula (2) and formula (3), we get: Formula (4) is a non-homogeneous linear differential equation with constant coefficients. Its solution is the general solution of the homogeneous equation plus the special solution of the non-homogeneous equation. Therefore, the homogeneous equation of formula (4) is: The general solution is: Among them, p, is a parameter related to the initial state of the winding, f=ω / (2π) is the natural frequency of the winding and complies with 4MK'>>C' 2 ,therefore, The non-homogeneous equation is decomposed into two: Find their special solutions respectively: Therefore, in formula (8): Therefore, the general solution of equation (8) is obtained: By taking the derivative of formula (9) twice, the corresponding vibration acceleration can be obtained:

4. The vibration state simulation monitoring system of the dry-type reactor according to claim 3, characterized in that: The analysis of the relationship between the magnitude of the preload force and the vibration acceleration and angular frequency, so as to find relevant measuring points to avoid resonance, includes: According to formulas (10) and (11), If K'>4Mω0 2 , that is, ω>2ω0, the amplitude of a decreases as the pressing force decreases; If K'<4Mω0 2 , that is, ω<2ω0, the amplitude of a increases as the pressing force decreases; If K'=4Mω0 2 , that is, ω=2ω0, resonance will occur at the fundamental frequency; Therefore, if a measuring point where no resonance occurs can be found, that is, ω is always greater than 2ω0 or ω is always less than 2ω0, the influence of resonance can be avoided and winding looseness defect diagnosis can be achieved.

5. The vibration state simulation monitoring system of the dry-type reactor according to claim 1, characterized in that: The analysis steps of the vibration characteristics of the reactor core include: First, analyze the electromagnetic field generated when the primary coil of the iron core is energized: The magnetic field inside the reactor winding is regarded as a uniform magnetic field, and the magnetic field strength inside the winding is expressed as: The magnetic induction intensity is expressed as: B=μ0μ r H=μ0μ r (N / L0)I (13) Where L0 is the average length of the core loop; μ0 = 4π × 10 -7 H / m, is the vacuum magnetic permeability; μ r is the relative magnetic permeability; N is the total number of turns of the primary winding, and I is the current inside the winding; Secondly, calculate the magnetostriction coefficient ε of the core silicon steel sheet in the magnetic field: The relationship between the linear deformation of silicon steel sheets in a magnetic field and the magnetic field strength is: In the formula, H c is the coercive force; s is the saturation magnetostriction of silicon steel sheet; From the definition of magnetostriction: Where: L is the initial length of the silicon steel sheet; ΔL is the change in length of the silicon steel sheet; is the expansion coefficient of silicon steel sheet; Therefore, the magnetostriction of silicon steel sheet is expressed as: Among them, V0 is the loading voltage of the reactor, B s is the magnetic flux density of the saturated reactor core, which can be expressed as: B s =B0cos(ωt), ω is the angular frequency, B0 is the magnetic flux density amplitude under the harmonic, and according to Faraday's law of electromagnetic induction, the loading voltage of the reactor is proportional to the rate of change of the magnetic flux. According to formula (16), the magnetostriction of the silicon steel sheet is proportional to the square of the loading voltage of the reactor, and the frequency is twice the voltage frequency. Therefore, the large magnetostriction of the silicon steel sheet means that its size changes greatly under the action of the magnetic field, which will cause the transformer to generate more noise and vibration during operation.

6. The vibration state simulation monitoring system of the dry-type reactor according to claim 5, characterized in that: The analysis steps of the reactor case surface vibration characteristics include: Constructing the reactor surface vibration signal model: The surface vibration of the reactor is a response to the vibration of the winding. The winding vibration is transmitted to the reactor surface through the solid component path. The vibration signal will be attenuated to a certain extent during the transmission process. The reduction of the winding preload will cause the winding to loosen, which will weaken the transmission capacity of the winding vibration and reduce the stiffness of the reactor structure. The natural frequency of the overall structure of the reactor will decrease, and the resonant frequency will decrease accordingly. Therefore, the resonant frequency is the frequency corresponding to the peak value in the reactor frequency response curve. Based on the above analysis, the reactor surface vibration signal model is proposed: Where: v tank is the vibration signal of the reactor surface; α is the transmission capacity coefficient. When the preload is reduced, the winding becomes loose and the coefficient decreases; v e,base is the fundamental frequency component in the winding vibration; β is the resonance amplification factor, which is a variable related to the damping ratio, frequency and other parameters; v e,r It is the vibration part of the winding vibration at the resonant frequency of the reactor; It is the vibration part of the winding except the fundamental frequency and resonant frequency; v x It is random noise, which comes from environmental interference.

7. The vibration state simulation monitoring system of the dry-type reactor according to claim 6, characterized in that: The step of analyzing the vibration characteristics of the reactor housing surface further includes: The vibration characteristics are analyzed based on the constructed reactor surface vibration signal model: From formula (17), we know that the surface vibration signal of the reactor contains four responses: The first term is the fundamental frequency component v of the winding vibration e,base The response αv e,base , which is proportional to the square of the current; The second term is the vibration of the reactor surface at the resonant frequency βαv e,r , even if the winding vibration contains only a small amount of vibration component v in the resonant frequency band e,r , which will also generate vibration on the surface of the reactor; The third item is the vibration part of the winding except the fundamental frequency and resonant frequency band. Response The fourth term is random noise; When the resonant frequency and the fundamental frequency coincide, the first and second terms are combined into one response; Therefore, the winding compression state affects the leakage magnetic field distribution, the winding vibration transmission capacity and the reactor resonance frequency, and there are differences in the surface vibration response of the reactor under different compression states.

8. The vibration state simulation monitoring system of a dry-type reactor according to claim 1, characterized in that: The vibration signals with obvious fault characteristics include: in the vibration of the reactor winding, the vibration frequency that resonates at the fundamental frequency; in the vibration of the iron core, the loading voltage and vibration frequency that lead to a large magnetostriction coefficient; and in the vibration of the reactor surface, a vibration signal that makes the fundamental frequency component of the winding vibration respond larger, a vibration signal with a large vibration component in the resonance frequency band contained in the winding vibration, a vibration signal with a large random noise, and a vibration signal with a large response of the vibration part of the winding vibration other than the fundamental frequency and the resonance frequency band.

9. The vibration state simulation monitoring system of the dry-type reactor according to claim 8, characterized in that: The method of finding a vibration signal with obvious fault characteristics and comparing it with the test result when the winding is working, and further diagnosing the fault of the reactor specifically includes: Based on the reactor vibration experiment, a multi-channel dynamic signal acquisition analyzer is used to collect vibration signal data under different times and working conditions. The different working conditions include: winding vibration signal, box surface vibration signal and winding core vibration signal; different times refer to signal data divided by seasons, and the time span is at least two years; The signal data obtained under different working conditions and at different times are compared with the vibration signal with obvious fault characteristics obtained by simulation calculation to obtain the accuracy of the vibration signal with obvious fault characteristics obtained by simulation calculation.

10. The vibration state simulation monitoring system of the dry-type reactor according to claim 9, characterized in that: The signal data obtained under different working conditions and at different times are compared with the vibration signal with obvious fault characteristics obtained by simulation calculation, specifically including: From the time dimension, under the same working condition, the signal data within a year is divided into spring and autumn signal data, winter signal data and summer signal data, and the signal data in the same season are compared. If it is obviously abnormal, it is determined to be abnormal signal data; the abnormal signal data is the abnormal noise size corresponding to the current frequency under the spring preload; From the working condition dimension, the signal data of different working conditions in the same time period are compared, and the corresponding abnormal signal data are obtained according to the relationship between different working conditions; The abnormal signal data of the corresponding working condition is numerically compared with the vibration signal with obvious fault characteristics obtained by simulation calculation. If the difference ratio is greater than the set threshold, the corresponding vibration signal data with obvious fault characteristics is adjusted; otherwise, the vibration signal with obvious fault characteristics is maintained and used as the fault reference value of the reactor.

11. The vibration state simulation monitoring system for a dry-type reactor according to any one of claims 1 to 10, characterized in that: The system also includes: a noise simulation analysis module, which is used to adopt an electromagnetic-structural-noise multi-physics field coupling analysis method according to the vibration conditions of the reactor to achieve full-process simulation of the reactor vibration noise.

12. The vibration state simulation monitoring system of the dry-type reactor according to claim 11, characterized in that: The simulation process comprises the following steps: S1 Establish a 1:1 real data reactor 3D simulation model in SolidWorks; S2 set up the transient magnetic field solver in Ansoft Maxwell, set the time solution step to 0.5ms, obtain the time domain waveform of the electromagnetic force through post-processing and record the data; S3 performs FFT transformation on the electromagnetic force in MATLAB to obtain the frequency, amplitude and phase of the electromagnetic force; S4 In ANSYS Workbench, the electromagnetic force calculated by the electromagnetic field is used as the load excitation, and the structure-noise coupling transfer file is defined at the same time; S5 In ANSYS Acoustics, the coupled noise sources are defined, the structural field calculation results are received, the calculations are solved, and the calculated simulation results are compared with the experimental data.

13. The vibration state simulation monitoring system of the dry-type reactor according to claim 12, characterized in that: The electromagnetic force is obtained on the basis of the vector magnetic potential, and various physical quantities of the electromagnetic field are obtained through transformation and post-processing. It is assumed that the current density in the winding is constant, and the electromagnetic force is solved using the virtual displacement principle: Where: J is the current density, A is the vector magnetic potential, dg is the virtual displacement; The vector magnetic potential A is obtained through the three-dimensional transient electromagnetic field equation: Where: v is the magnetic resistivity of the material, J is the winding current density, is the eddy current density caused by electromagnetic induction.

Images