A noise reduction method for a three-dimensional wound core amorphous alloy transformer

Abstract

The application discloses a kind of three-dimensional volume iron core amorphous alloy transformer noise reduction method, belong to transformer technical field, this method includes: multiple measuring points are arranged on transformer, and vibration feature is extracted based on the equipment data and vibration signal of measuring point;Compensation unit is arranged in the core of transformer, and disturbance is applied to compensation unit, and response matrix is constructed;According to equipment data, magnetic flux margin, constraint coefficient and sound radiation weight are calculated to construct correction matrix, and the singular value decomposition of correction matrix is carried out, to obtain response base vector and parameter space matrix;The projection coefficient of response base vector in target blanking direction is calculated, and compensation magnetic potential parameters are determined based on projection coefficient, parameter space matrix and constraint coefficient;Compensation magnetic potential parameters are applied to control compensation unit, and weighted vibration vector after application is obtained, when weighted vibration vector after application two norms are less than current weighted vibration vector two norms, compensation magnetic potential parameters are retained;Through the application, the noise of transformer can be effectively reduced.

Description

Technical Field

[0001] This invention belongs to the field of transformer technology, specifically relating to a noise reduction method for a three-dimensional wound core amorphous alloy transformer. Background Technology

[0002] The three-dimensional wound amorphous alloy transformer uses amorphous alloy strip to form a three-phase three-dimensional wound core. The A-phase, B-phase, and C-phase cores form a vibration transmission path through clamps, an oil tank, and a tank cover. After the transformer is connected to AC power, an alternating current is generated in the primary winding. This alternating current excites an alternating magnetic field around the core, causing the magnetic flux density within the core to change periodically with the magnitude and direction of the alternating voltage. Because the three-dimensional wound core is made of amorphous alloy strip, the alternating magnetic field drives the magnetic domains inside the strip to repeatedly flip and move, undergoing a periodic magnetization process. During this magnetization process, the amorphous alloy strip, being a ferromagnetic material, undergoes microscopic dimensional changes, i.e., magnetostriction. This dimensional expansion and contraction accumulates continuously in the multi-layered strip core, eventually causing periodic deformation of the core. This periodic deformation is transmitted outward in the form of mechanical waves, thus triggering continuous vibration of the transformer.

[0003] When the grid frequency is 50 Hz, the magnetostrictive strain typically varies with the square of the magnetic flux density. The main frequency components of core vibration and noise usually include 100 Hz and its higher harmonics, such as 200 Hz and 300 Hz. In practical engineering, the magnetostrictive properties of amorphous alloys have a nonlinear relationship with magnetic flux density. Furthermore, changes in transformer tap position, operating voltage, and core temperature can cause changes in the core's working magnetic flux density and its distribution, which in turn alters the vibration amplitude of structural components such as clamps, tank walls, and tank covers at various target frequencies. These structural vibrations radiate outward through the tank walls and tank cover, potentially increasing transformer noise.

[0004] Existing technologies typically reduce transformer vibration by increasing the core clamping force, installing damping materials between the clamps or the core and the tank, and reinforcing the tank walls. These methods usually require increasing the number, size, or preload of clamps, clamping components, damping components, or tank reinforcement components, thus increasing the transformer's weight and manufacturing cost. Another technology applies reverse mechanical vibration to the vibration transmission ends, such as the tank walls, to suppress structural vibration. However, this method requires actuators to be configured for multiple vibration transmission areas. When tap position, operating voltage, load conditions, or core temperature change, the structural vibration response changes accordingly, and the preset mechanical vibration parameters are difficult to adapt to different operating conditions, resulting in unstable noise reduction effects. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a noise reduction method for a three-dimensional wound core amorphous alloy transformer, thereby resolving the issues present in the background art.

[0006] To achieve the aforementioned objectives, this invention proposes a noise reduction method for a three-dimensional wound-core amorphous alloy transformer, comprising: Multiple measuring points are set up on the transformer to obtain equipment data and vibration signals from multiple measuring points under normal operating conditions. Based on the equipment data and vibration signals, the vibration characteristics of each measuring point at each target frequency are extracted. A compensation unit is set in the core of the transformer. Positive and negative test disturbances are applied to the compensation unit. The first change, second change, symmetry coefficient and effective response of the vibration characteristics before and after the test are calculated, and the response matrix is ​​constructed. The magnetic flux margin and constraint coefficient are calculated based on the equipment data. The acoustic radiation weight is calculated based on the vibration characteristics and the position weight of the measuring point. A correction matrix is ​​constructed based on the response matrix, magnetic flux margin, constraint coefficient and acoustic radiation weight. Singular value decomposition is performed on the correction matrix to obtain the response basis vector and parameter space matrix. The current weighted vibration vector is obtained based on the vibration characteristics and acoustic radiation weight. The inverse of the current weighted vibration vector is defined as the target blanking direction. The projection coefficient of the response basis vector on the target blanking direction is calculated. The compensation magnetomotive force parameters are determined based on the projection coefficient, parameter space matrix and constraint coefficient. The control compensation unit applies compensation magnetomotive force parameters and obtains the weighted vibration vector after application. When the second norm of the applied weighted vibration vector is less than the second norm of the current weighted vibration vector, the compensation magnetomotive force parameters are retained to reduce transformer noise.

[0007] The beneficial effects of this invention are as follows: This invention first applies positive and negative test perturbations to the core of each phase to obtain the effective response of each compensated magnetomotive force amplitude to the vibration characteristics of each measuring point and each target frequency, and constructs a response matrix to characterize the vibration influence relationship of different compensation units on the main vibration structure of the transformer. Then, constraint coefficients are determined based on the flux margin of each phase core, and a correction matrix is ​​constructed by combining the vibration response of the main acoustic radiation structure and the response trends of the positive and negative test perturbations. This prioritizes suppressing high acoustic radiation vibrations while reducing the impact of flux density exceeding limits and nonlinear responses on the reliability of the compensation parameter solution. Subsequently, singular value decomposition is performed on the correction matrix, and each compensated magnetomotive force parameter is determined based on the projection of the response basis vector onto the target blanking direction of the current weighted vibration vector, enabling multiple compensation units to adjust collaboratively. Finally, the weighted vibration response and the total working flux density of each phase core after applying the compensated magnetomotive force are verified, and the compensated magnetomotive force parameters that satisfy the conditions of vibration reduction and flux density not exceeding limits are retained. This invention can obtain compensation parameters that balance vibration reduction effect and magnetic flux safety margin in response to changes in core magnetic flux and structural vibration under different operating conditions, reduce reliance on mechanical reinforcement, damping materials or end-reverse vibration devices, and improve the noise reduction stability of three-dimensional wound core amorphous alloy transformers. Attached Figure Description

[0008] Figure 1 This is a flowchart illustrating the steps of a noise reduction method for a three-dimensional wound-core amorphous alloy transformer according to the present invention. Detailed Implementation

[0009] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0010] This method includes a calibration phase and an operation phase. During the calibration phase, vibration sensors are temporarily placed at predetermined structural vibration measurement points on the transformer's clamps, tank walls, and tank cover to acquire vibration characteristics, construct a response matrix, and determine the compensation magnetomotive force parameters. During the operation phase, the compensation magnetomotive force parameters determined in the calibration phase are written into the control device. The control device outputs the corresponding compensation magnetomotive force according to the preset operating state to reduce transformer noise.

[0011] like Figure 1 As shown, a noise reduction method for a three-dimensional wound-core amorphous alloy transformer includes: S1: Multiple measuring points are set up on the transformer to obtain equipment data and vibration signals from multiple measuring points under normal operating conditions. Based on the equipment data and vibration signals, the vibration characteristics of each measuring point at each target frequency are extracted.

[0012] The transformer's equipment data includes three-phase voltage signals, core temperature signals, and tap position status signals. The three-phase voltage signals provide the fundamental frequency and synchronization phase of the power grid, the core temperature signals determine the allowable magnetic flux density, and the tap position status signals determine the effective number of turns at the current tap position. Accelerometers are placed at multiple measuring points on the transformer to acquire vibration signals, specifically a sequence of vibration acceleration values. The sensor installation locations include the clamping surfaces corresponding to phase A, phase B, and phase C cores, the three-phase shared clamping area, and the long and short walls of the tank. The three-phase shared clamping area refers to the clamping part that is mechanically connected to at least two phase cores.

[0013] In this embodiment, vibration characteristics at each target frequency are extracted from each measuring point based on equipment data and vibration signals, including: The equipment data includes three-phase voltage signals. The A-phase voltage signal is extracted from the three-phase voltage signals. After performing a discrete Fourier transform on the A-phase voltage signal, the frequency corresponding to the frequency component with the largest amplitude is determined as the fundamental frequency. The second, fourth, and sixth harmonics of the fundamental frequency are determined as the target frequencies.

[0014] Before extracting vibration characteristics, the A-phase voltage signal is first extracted from the three-phase voltage signal. The A-phase voltage signal is a discrete voltage numerical sequence. The A-phase voltage signal is subjected to a discrete Fourier transform. The frequency component with the largest amplitude is determined from the transform result, and the frequency corresponding to this frequency component is determined as the fundamental frequency. The target frequency set is determined based on the fundamental frequency. The target frequencies include the second, fourth, and sixth harmonics of the fundamental frequency. For example, if the fundamental frequency is 50Hz, then the second, fourth, and sixth harmonics are 100Hz, 200Hz, and 300Hz, respectively.

[0015] The zero-crossing point of the phase A voltage signal in the same direction is extracted as the starting time. An analysis time window containing an integer number of fundamental frequency periods is set. The vibration signal of each measuring point is intercepted based on the analysis time window to obtain the discrete vibration sequence corresponding to each measuring point. The discrete Fourier transform result of each discrete vibration sequence at each target frequency is calculated to obtain the complex value of each measuring point at each target frequency. The magnitude of each complex value is calculated as the vibration amplitude, and the vibration amplitude is used as the vibration characteristic of the target frequency.

[0016] The length of the analysis time window is set. First, the number of fundamental frequency periods included in the analysis time window is determined so that the second, fourth, and sixth harmonics have an integer number of periods within the window. After setting, the length of the analysis time window is the product of the number of fundamental frequency periods included and the duration of each fundamental frequency period. The zero-crossing point of the A-phase voltage is used as the start time of the time window. The end time is determined based on the analysis time window. Vibration signals at each measuring point are captured at the start and end times to obtain discrete vibration sequences. Single-frequency complex amplitude extraction is performed on the discrete vibration sequences at each target frequency to obtain the complex values ​​at the target frequencies. Single-frequency complex amplitude extraction is accomplished by calculating the discrete Fourier transform frequency components of the discrete vibration sequences at the corresponding target frequencies.

[0017] The magnitude of the complex values ​​at each target frequency of the measuring point is taken as the vibration amplitude, and the vibration amplitude is taken as the vibration characteristic of the measuring point at the corresponding target frequency. Since the vibration characteristic is obtained under normal operating conditions, it is defined as the basic vibration characteristic. The vibration characteristic obtained through the above steps is used to reflect the vibration response magnitude of the clamping areas corresponding to different cores, the three-phase common clamping area, the tank wall area, and the tank cover area in the three-dimensional wound core amorphous alloy transformer at each target frequency.

[0018] S2: Set up a compensation unit in the core of the transformer, perform positive and negative test disturbances on the compensation unit, calculate the first change, second change, symmetry coefficient and effective response of the vibration characteristics before and after the test, and construct the response matrix.

[0019] In this embodiment, the compensation unit is subjected to positive and negative test perturbations, including: Based on the compensation magnetomotive force amplitude of the current test compensation unit, a preset amplitude is added to perform a positive test disturbance; based on the compensation magnetomotive force amplitude, a preset amplitude is reduced to perform a negative test disturbance; the difference between the test vibration characteristics after applying the positive test disturbance and the basic vibration characteristics is calculated to obtain the first change; the difference between the test vibration characteristics after applying the negative test disturbance and the basic vibration characteristics is calculated to obtain the second change.

[0020] To reduce noise, this embodiment sets up compensation units for phase A, phase B, and phase C cores respectively. Each compensation unit includes a compensation winding and a power drive circuit. The compensation winding is an independent coil wound on the corresponding phase core. Its winding position is located on the core column or core yoke section outside the primary and secondary windings of the corresponding phase core. The number of turns of the compensation winding is determined according to the required compensation magnetomotive force range and the maximum output current of the compensation drive circuit. The compensation winding is magnetically coupled to the primary and secondary windings only through the shared core magnetic circuit. The power drive circuit is used to output alternating current to the compensation winding.

[0021] The compensation unit applies an amplitude compensation magnetomotive force (MMF) to the corresponding iron core. This MMF is an alternating MMF synchronized with the fundamental frequency of the power grid, and its phase is determined based on the A-phase voltage signal. The compensation MMF generates a compensation flux in the magnetic circuit of the corresponding iron core, which is superimposed on the working flux formed by the winding voltage during normal transformer operation, thereby altering the waveform of the total magnetic flux density within the corresponding iron core. Since there is a non-linear relationship between the magnetostrictive strain of the amorphous alloy strip and the magnetic flux density within the iron core, changes in the amplitude of the total magnetic flux density result in corresponding changes in the amplitude of the magnetostrictive strain generated by the amorphous alloy strip at the target frequency and the amplitude of the resulting mechanical excitation. This mechanical excitation, transmitted through the iron core, clamps, and tank structure, alters the vibration characteristics of the clamps, tank walls, and tank cover at the target frequency. Therefore, the purpose of this invention is to reduce the vibration of noisy structural components such as the tank walls and tank cover by adjusting the compensation MMF output by each phase compensation unit.

[0022] In this embodiment, the parameters that the compensation unit needs to determine include the compensation magnetomotive force amplitude. When determining the compensation magnetomotive force amplitude, each compensation unit is subjected to test perturbation according to a preset test sequence. While testing one compensation unit, the remaining compensation units maintain their current compensation magnetomotive force amplitude unchanged. The test perturbation includes positive test perturbation and negative test perturbation, which set opposite directions of change for the same compensation magnetomotive force amplitude. For example, increasing the current compensation magnetomotive force amplitude by a preset amplitude completes one positive test perturbation, while decreasing the current compensation magnetomotive force amplitude by a preset amplitude completes one negative test perturbation.

[0023] The amplitude of the test disturbance is determined based on the working magnetic flux density of the iron core and the output capability of the compensation unit. The duration of the positive and negative test disturbances covers the fundamental period of the predetermined value. During each test disturbance, the vibration characteristics of multiple measurement points at each target frequency are obtained based on the analysis time window according to the method of the previous steps. Since the vibration characteristics are obtained under test conditions, they are defined as test vibration characteristics.

[0024] For the compensation magnetomotive force amplitude, the change in vibration characteristics caused by the positive test perturbation is obtained by subtracting the basic vibration characteristics of the j-th measuring point at the k-th target frequency before the positive test perturbation from the test vibration characteristics of the j-th measuring point at the k-th target frequency after the positive test perturbation is applied. This change is defined as the first change. Similarly, the change in vibration characteristics caused by the negative test perturbation is obtained by subtracting the basic vibration characteristics of the j-th measuring point at the k-th target frequency before the negative test perturbation from the test vibration characteristics of the j-th measuring point at the k-th target frequency after the negative test perturbation is applied. This change is defined as the second change. When the first and second changes are in opposite directions, it indicates that the vibration changes caused by increasing and decreasing the corresponding compensation magnetomotive force parameters are opposite.

[0025] The symmetry coefficient is determined based on the product of the first change and the second change. The first difference between the test vibration characteristics after applying a positive test disturbance and the test vibration characteristics after applying a negative test disturbance is calculated. The second difference between the positive test disturbance parameter value and the negative test disturbance parameter value is calculated. The effective response of the compensation magnetomotive force amplitude at the corresponding measurement point and target frequency is calculated based on the first difference and the second difference. The response matrix is ​​constructed based on the effective response, and the symmetry coefficient matrix is ​​constructed based on the symmetry coefficient.

[0026] Based on the first and second changes, calculate the symmetry coefficient of the i-th compensated magnetomotive force parameter with respect to the j-th measuring point at the k-th target frequency. When the product of the first and second vibration amplitude changes is less than zero, it indicates that the vibration amplitude changes caused by the positive and negative test disturbances are in opposite directions, and the symmetry coefficient of the positive and negative disturbance responses is determined as the first value. When the product is greater than or equal to zero, the symmetry coefficient of the positive and negative disturbance responses is determined as the second value. For example, the first value is 1, and the second value is a value greater than or equal to 0 and less than 1.

[0027] Based on the test vibration characteristics obtained after applying positive and negative test perturbations, the effective response of the i-th compensated magnetomotive force amplitude to the j-th measuring point at the k-th target frequency is calculated. The effective response is calculated based on the following formula: In the formula, For the effective response of the i-th compensated magnetomotive force amplitude to the j-th measuring point at the k-th target frequency, This describes the test vibration characteristics at the k-th target frequency of the j-th measurement point after applying a positive test disturbance. This describes the test vibration characteristics at the k-th target frequency of the j-th measuring point after applying a negative test disturbance. Let be the positive test perturbation parameter value corresponding to the i-th compensated magnetomotive force parameter. For the negative test disturbance parameter value corresponding to the i-th compensated magnetomotive force parameter, the compensated magnetomotive force parameters are arranged according to preset column numbers, the measurement point numbers and target frequency combinations are arranged according to preset row numbers, and each effective response is used as a matrix element to construct a response matrix. The element in the i-th row and j-th column of the response matrix represents the change in vibration amplitude corresponding to the i-th measurement point and target frequency combination when the j-th compensated magnetomotive force parameter undergoes a unit change.

[0028] Arrange the symmetry coefficients according to the same row and column numbers as the response matrix to obtain the symmetry coefficient matrix. The element in the i-th row and j-th column of the symmetry coefficient matrix represents the degree of opposite trend between the positive and negative test perturbation responses of the j-th compensation magnetomotive force parameter to the i-th measurement point and target frequency combination.

[0029] S3: Calculate the flux margin and constraint coefficient based on the equipment data, calculate the acoustic radiation weight based on the vibration characteristics and the position weight of the measuring point, construct a correction matrix based on the response matrix, flux margin, constraint coefficient and acoustic radiation weight, perform singular value decomposition on the correction matrix to obtain the response basis vector and parameter space matrix.

[0030] In this embodiment, a correction matrix is ​​constructed based on the response matrix, flux margin, constraint coefficient, and acoustic radiation weight, including: The working magnetic flux density of the core under positive and negative test disturbances is determined respectively. The change in peak magnetic flux density is determined based on the difference between the working magnetic flux density during the test and the working magnetic flux density before the test. The constraint coefficient is determined based on the ratio of magnetic flux margin to the change in peak magnetic flux density.

[0031] Both positive and negative test disturbances are applied within a preset test safety range to obtain the local vibration response of each compensated magnetomotive force amplitude under the current operating condition. Since the subsequently determined candidate compensated magnetomotive force parameters may be larger than the test disturbance amplitude, and multiple compensated magnetomotive force parameters may have a superimposed effect when applied simultaneously, this step determines the constraint coefficients for each compensated magnetomotive force parameter based on the magnetic flux margin under the current operating condition and the change in magnetic flux density caused by the test disturbance. The constraint coefficients are used to reduce the change in compensated magnetomotive force parameters with smaller magnetic flux margins during subsequent parameter determination, ensuring that the determined compensated magnetomotive force parameters balance both vibration suppression requirements and core magnetic flux density margin.

[0032] Specifically, within the analysis time window during normal transformer operation, the magnetic flux density of each core at each moment is calculated based on Faraday's law of electromagnetic induction, the effective number of turns, the effective cross-sectional area of ​​the core, and the phase voltage value. The specific calculation formula is existing technology and will not be described here. The magnetic flux densities of the core at each sampling moment within the analysis time window are arranged in chronological order to obtain the corresponding magnetic flux density sequence. The working magnetic flux density of each core is determined based on the maximum absolute value in the magnetic flux density sequence. Based on the material parameters of the amorphous alloy strip, the core temperature signal, and the preset saturation magnetic flux density data table, the allowable magnetic flux density of each core at the current core temperature is determined. The preset saturation magnetic flux density data table records the saturation magnetic flux density values ​​of the amorphous alloy strip at different temperatures. The allowable magnetic flux density is the saturation magnetic flux density multiplied by a preset safety factor. In this embodiment, the preset safety factor is set to 0.8. The allowable magnetic flux density is subtracted from the working magnetic flux density to obtain the magnetic flux margin of each core. The magnetic flux margin indicates how much usable margin is left between the magnetic flux density of the corresponding core under the current operating state and the allowable upper limit.

[0033] During both positive and negative test disturbances, the working magnetic flux density of the corresponding core during the positive and negative test disturbances is calculated using the same method. The working magnetic flux density during the positive test disturbance is subtracted from the working magnetic flux density before the test to obtain the magnetic flux density increment caused by the positive test disturbance. Similarly, the working magnetic flux density during the negative test disturbance is subtracted from the working magnetic flux density before the test to obtain the magnetic flux density increment caused by the negative test disturbance. The larger absolute value of the magnetic flux density increments caused by the positive and negative test disturbances is taken as the peak magnetic flux density change caused by the corresponding compensation magnetomotive force parameter. The peak magnetic flux density change represents the maximum change in core magnetic flux density that the compensation magnetomotive force parameter can cause under the test disturbance amplitude.

[0034] Based on the above data, the constraint coefficient of the j-th compensated magnetomotive force parameter is calculated. The constraint coefficient represents the proportion of the j-th compensated magnetomotive force parameter that can be used under the current magnetic flux density constraint. The constraint coefficient is calculated based on the following formula: In the formula, Cj is the constraint coefficient of the j-th compensated magnetomotive force parameter. Let be the flux margin of the iron core corresponding to the j-th compensated magnetomotive force parameter. Let be the change in peak magnetic flux density caused by the j-th compensated magnetomotive force parameter.

[0035] The weighted vibration amplitude is obtained by multiplying the basic vibration characteristics of the measuring point with the corresponding position weight. The sound radiation weight is determined based on the proportion of the weighted vibration amplitude of the measuring point to the sum of the weighted vibration amplitudes of all measuring points. The correction element is the product of the effective response in the response matrix with the corresponding sound radiation weight, constraint coefficient and symmetry coefficient. All correction elements are arranged according to the response matrix to form the correction matrix.

[0036] Then, based on the position weights and vibration characteristics of the measuring points, the sound radiation weights of each measuring point at each target frequency are determined. The position weights are used to reflect the relative ability of vibrations of different structural parts to radiate noise to the outside. In this embodiment, higher position weights are set for the measuring points on the long side wall of the tank, the short side wall of the tank, and the tank cover, while lower position weights are set for the clamp measuring points. This is because the tank wall and the tank cover face the external space, and the sound caused by their vibrations is easily transmitted to the outside. Their vibrations have a greater impact on external noise than the clamp vibrations. The clamps are mainly located inside the transformer, and their vibrations indirectly affect the outside through structural transmission, resulting in a weaker sound radiation capability. After obtaining the location weights, the acoustic radiation weight of the p-th measurement point at the h-th target frequency is calculated based on the following formula: In the formula, Let p be the acoustic radiation weight at the h-th target frequency for the p-th measurement point. The vibration characteristics of the p-th measuring point at the h-th target frequency are given. Let p be the position weight of the p-th measurement point. This represents the total number of measurement points. and These represent the vibration characteristics and position weights of the r-th measuring point at the h-th target frequency, respectively. Based on the above formula, the acoustic radiation weight represents the relative influence of the measuring point on the external noise of the transformer at the target frequency. A measuring point with larger vibration characteristics indicates a higher vibration level at that measuring point.

[0037] The acoustic radiation weights of the measurement point and target frequency corresponding to the i-th row in the response matrix are used as the row weights of the i-th row, and the constraint coefficients of the j-th compensation magnetomotive force parameter are used as the column constraint coefficients of the j-th column. Then, the correction matrix is ​​obtained based on the following formula. In the formula, To correct the element in the i-th row and j-th column of the matrix, For the element in the i-th row and j-th column of the response matrix, Let be the acoustic radiation weight of the measurement point and the target frequency corresponding to the i-th row. Let be the constraint coefficient for the j-th compensated magnetomotive force parameter. is the element in the i-th row and j-th column of the symmetric coefficient matrix. The correction matrix obtained based on the above formula is used to characterize the ability of changes in each compensated magnetomotive force parameter to regulate the vibration response of each measuring point at each target frequency under the joint constraints of the importance of acoustic radiation at the measuring point, the flux margin of the compensated magnetomotive force parameter, and the symmetry of the positive and negative test disturbance responses. Specifically, the larger the modulus of the elements in the correction matrix, the higher the priority of the corresponding compensated magnetomotive force parameter in regulating the weighted vibration response of the corresponding measuring point and target frequency combination under the combined effects of acoustic radiation weight, flux margin constraint coefficient, and positive and negative test disturbance response symmetry coefficient.

[0038] The above formula also introduces a symmetric coefficient matrix to correct the response matrix, which is used to evaluate whether the changes in structural vibration response caused by each compensated magnetic potential parameter under positive and negative test disturbances have the expected opposite trends.

[0039] To facilitate understanding of the above calculation process, an example is given below. Assume the element in the i-th row and j-th column of the response matrix corresponds to the influence of the change in the second compensated magnetomotive force parameter on the vibration characteristics of the first measuring point on the long side wall of the fuel tank at the target frequency of 200Hz. Since this measuring point is located on the long side wall of the fuel tank, its corresponding acoustic radiation weight is relatively large. If the magnetic flux margin corresponding to the second compensated magnetomotive force parameter is sufficient, its constraint coefficient is large, indicating that the compensated magnetomotive force parameter has a large adjustable range. Furthermore, if the changes in vibration characteristics obtained after applying positive and negative test perturbations to this compensated magnetomotive force parameter are opposite, then the corresponding symmetry coefficient is large. Under the above conditions, the original response matrix elements are corrected by the acoustic radiation weights, constraint coefficients, and symmetry coefficients to obtain larger corrected matrix elements. These larger matrix elements indicate that, under the condition of satisfying the magnetic flux constraint, the second compensated magnetomotive force parameter has a more stable and stronger control over the vibration of the first measuring point on the long side wall of the oil tank at 200Hz, and therefore has a higher influence in the subsequent calculation of the compensated magnetomotive force parameter. Conversely, if a matrix element corresponds to a clamping measurement point, the clamping component has a weak direct radiation capability to external noise, resulting in a smaller acoustic radiation weight, a smaller constraint coefficient, or a smaller symmetry coefficient. In this case, the corresponding original response matrix element is reduced, thereby decreasing its influence in the subsequent calculation of the compensation magnetomotive force parameters. Therefore, by increasing the acoustic radiation weight, the influence of external noise radiation structures such as the tank wall and tank cover on the vibration response is improved. By reducing the constraint coefficient, the influence of the compensation magnetomotive force parameter with insufficient magnetic flux margin in the parameter calculation is reduced. Furthermore, by reducing the symmetry coefficient, the influence of the effective response, which does not have opposite trends to the positive test disturbance response and the negative test disturbance response, is reduced. This allows subsequent parameter calculations to prioritize vibration control directions with larger acoustic radiation contributions, relatively sufficient magnetic flux margin, and reliable test responses. As a result, the applicability of the compensation magnetomotive force parameter is improved, and the risk of reduced core magnetic flux density margin due to parameter adjustment is reduced.

[0040] After constructing the correction matrix, singular value decomposition (SVD) is performed on it to obtain the response space matrix, singular value matrix, and parameter space matrix. Singular value decomposition is a prior art technique and will not be described further here. The singular values ​​in the singular value matrix are arranged in descending order, and the vectors corresponding to the first preset number of singular values ​​in the response space matrix are used as response basis vectors. These response basis vectors correspond to the vibration characteristic changes at multiple target frequencies caused by a set of changes in the compensated magnetomotive force parameters. The first preset number is less than or equal to the total number of compensated magnetomotive force parameters and can be set based on the total number of measurement points and the number of target frequencies.

[0041] S4: Obtain the current weighted vibration vector based on the vibration characteristics and acoustic radiation weights, define the inverse of the current weighted vibration vector as the target blanking direction, calculate the projection coefficient of the response basis vector on the target blanking direction, and determine the compensation magnetomotive force parameters based on the projection coefficient, parameter space matrix and constraint coefficient.

[0042] In this embodiment, calculating the projection coefficients of the response basis vector in the target hiding direction includes: The product of the pre-test vibration features and the corresponding acoustic radiation weights is calculated to obtain the weighted vibration features. The weighted vibration features are arranged into the current weighted vibration vector. The opposite of each element in the current weighted vibration vector is taken to obtain the target blanking direction vector. The singular values ​​in the singular value matrix are arranged in descending order. The vectors corresponding to the first preset number of singular values ​​in the response space matrix are used as the response basis vectors. The product of each element in the response basis vector and the corresponding element in the target blanking direction vector is calculated, and all products are summed to obtain the projection coefficients corresponding to the response basis vectors.

[0043] Based on the fundamental vibration characteristics and corresponding acoustic radiation weights of multiple measuring points at each target frequency, the product of the fundamental vibration characteristics and corresponding acoustic radiation weights of each measuring point at each target frequency is calculated to obtain the weighted vibration characteristics. Following the row order of the correction matrix, the weighted vibration characteristics of all measuring points at all target frequencies are arranged sequentially to obtain the current weighted vibration vector. Each element in the current weighted vibration vector represents the weighted vibration amplitude of the corresponding measuring point at the corresponding target frequency.

[0044] By taking the opposite value of each element in the current weighted vibration vector, the target blanking direction vector is obtained. Each element in the target blanking direction vector represents the expected reduction in the weighted vibration amplitude at the corresponding measurement point and the corresponding target frequency. When the direction of the weighted vibration characteristic change caused by the change in the compensation magnetomotive force amplitude is consistent with the direction of the target blanking direction vector, the weighted vibration amplitude at the corresponding measurement point and the corresponding target frequency decreases, thereby reducing the L2 norm of the current weighted vibration vector, indicating that the vibration level of the main noise radiating structure at the target frequency is reduced.

[0045] Each response basis vector is projected onto the target blanking direction vector to obtain the projection coefficients corresponding to each response basis vector. For each response basis vector, the product of each element of the response basis vector with the corresponding element in the target blanking direction vector is calculated, and all products are summed to obtain the projection coefficients of the response basis vector. The projection coefficients are signed values; their sign indicates the direction of change of the corresponding response basis vector relative to the target blanking direction vector, and their absolute value indicates the magnitude of the component of the corresponding response basis vector on the target blanking direction vector. The larger the absolute value of the projection coefficient, the higher the adjustability of the corresponding response basis vector in reducing the L2 norm of the current weighted vibration vector.

[0046] In this embodiment, the compensation magnetomotive force parameters are determined based on the projection coefficients, parameter space matrix, and constraint coefficients, including: The response basis vectors are arranged in descending order of the absolute values ​​of the projection coefficients. The second-most preset number of response basis vectors are selected as candidate compensation basis vectors. All elements in the columns corresponding to the candidate compensation basis vectors are extracted from the parameter space matrix to obtain the parameter combination coefficients. The singular values ​​corresponding to the candidate compensation basis vectors are extracted from the singular value matrix. The ratio of the projection coefficients to the corresponding singular values ​​of the candidate compensation basis vectors is calculated to obtain the mode adjustment coefficients corresponding to the candidate compensation basis vectors.

[0047] The response basis vectors are arranged in descending order of the absolute values ​​of their corresponding projection coefficients. The first two preset number of response basis vectors are selected as candidate compensation basis vectors. The projection coefficients corresponding to the candidate compensation basis vectors retain their positive or negative signs and are used to determine the direction of increase or decrease of the corresponding compensation magnetomotive force amplitude. The candidate compensation basis vectors represent a vibration regulation mode formed by the coordinated change of the compensation magnetomotive force amplitude of multiple compensation units. For example, when the compensation magnetomotive force amplitude of phase A is increased, the compensation magnetomotive force amplitude of phase B is decreased, and the compensation magnetomotive force amplitude of phase C is increased, the weighted vibration characteristics of multiple measuring points at the target frequency change together in the decreasing direction.

[0048] For each candidate compensation basis vector, the corresponding parameter space vector is extracted from the parameter space matrix obtained by singular value decomposition. For example, for the Lth candidate compensation basis vector, all elements in the Lth column of the parameter space matrix are extracted to obtain the parameter combination coefficients corresponding to the Lth candidate compensation basis vector. The parameter combination coefficients represent the relative change direction and relative change ratio of the A-phase compensation magnetomotive force amplitude, the B-phase compensation magnetomotive force amplitude, and the C-phase compensation magnetomotive force amplitude when the corresponding vibration regulation mode is formed.

[0049] For each candidate compensation basis vector, the singular value corresponding to that candidate compensation basis vector is obtained. The ratio of the projection coefficient of the candidate compensation basis vector to the corresponding singular value is calculated to obtain the mode adjustment coefficient of the candidate compensation basis vector. The mode adjustment coefficient represents the normalized adjustment amount required for the vibration adjustment mode to change the current weighted vibration vector along the target blanking direction vector. The mode adjustment coefficient of each candidate compensation basis vector is multiplied by the corresponding parameter combination coefficient in the parameter space matrix to obtain the compensation magnetomotive force amplitude component corresponding to each candidate compensation basis vector. The compensation magnetomotive force amplitude components corresponding to all candidate compensation basis vectors are added together to obtain the initial change amount of the corresponding compensation magnetomotive force amplitude.

[0050] The compensation magnetomotive force amplitudes of the compensation unit corresponding to all candidate compensation basis vectors are synthesized to obtain the initial change in the compensation magnetomotive force amplitude of the compensation unit. Based on the initial change in the compensation magnetomotive force amplitude and the constraint coefficient corresponding to the compensation unit, the actual change in the compensation magnetomotive force amplitude of the compensation unit is determined. Based on the current compensation magnetomotive force amplitude and the actual change in the compensation magnetomotive force amplitude, the compensation magnetomotive force parameters are determined.

[0051] Multiply the initial change in the amplitude of the j-th compensation magnetomotive force by the constraint coefficient of the j-th compensation magnetomotive force amplitude to obtain the actual change in the amplitude of the j-th compensation magnetomotive force amplitude. The initial change is the normalized change in the amplitude of the compensation magnetomotive force amplitude, and the actual change is the change in the amplitude of the compensation magnetomotive force amplitude determined under the constraints of the corresponding constraint coefficient. Add the current compensation magnetomotive force amplitude of each compensation unit to the corresponding actual change to obtain the compensation magnetomotive force amplitude to be applied. To facilitate understanding of the above calculation process, an example is given below. After performing singular value decomposition on the correction matrix, the resulting response basis vector is used to represent the structural vibration coordinated adjustment mode formed when different compensation magnetomotive force amplitudes change in tandem. For example, the first candidate compensation basis vector can represent an adjustment mode in which, when the amplitude of the A-phase compensation magnetomotive force is appropriately increased, the amplitude of the B-phase compensation magnetomotive force is decreased, and the amplitude of the C-phase compensation magnetomotive force is increased, the weighted vibration characteristics of the first measuring point on the long side wall of the tank, the measuring point on the short side wall of the tank, and the measuring point on the tank cover at the target frequency of 200Hz all change in the direction of decrease.

[0052] The first candidate compensation basis vector is projected onto the target hidden direction vector to obtain the projection coefficients corresponding to the first candidate compensation basis vector. A larger absolute value of the projection coefficients indicates that the cooperative adjustment mode has a larger component in the current weighted vibration characteristic direction that needs to be reduced, that is, the adjustment mode has a higher ability to reduce the L2 norm of the current weighted vibration vector.

[0053] The parameter combination coefficients corresponding to the first candidate compensation basis vector may include, for example, the coefficients for the A-phase compensation magnetomotive force amplitude, the B-phase compensation magnetomotive force amplitude, and the C-phase compensation magnetomotive force amplitude, which are 0.6, -0.3, and 0.4, respectively. These parameter combination coefficients indicate that, to form the corresponding coordinated adjustment mode, the A-phase compensation magnetomotive force amplitude is adjusted in the increasing direction, the B-phase compensation magnetomotive force amplitude is adjusted in the decreasing direction, and the C-phase compensation magnetomotive force amplitude is adjusted in the increasing direction, with relative adjustment ratios of 0.6, 0.3, and 0.4.

[0054] If the projection coefficient of the first candidate compensation basis vector is P1, and the corresponding singular value is S1, then the ratio of P1 to S1 is calculated to obtain the mode adjustment coefficient K1 of the first candidate compensation basis vector. Multiplying the mode adjustment coefficient K1 by 0.6, -0.3, and 0.4 respectively yields the A-phase compensation magnetomotive force amplitude components, B-phase compensation magnetomotive force amplitude components, and C-phase compensation magnetomotive force amplitude components corresponding to the candidate compensation basis vector.

[0055] If multiple candidate compensation basis vectors exist, for each compensation magnetomotive force amplitude, the compensation magnetomotive force amplitude component corresponding to each candidate compensation basis vector is obtained, and the compensation magnetomotive force amplitude components belonging to the same compensation magnetomotive force amplitude are added together to obtain the initial change in the compensation magnetomotive force amplitude. For example, if the A-phase compensation magnetomotive force amplitude component corresponding to the first candidate compensation basis vector is 0.6K1, and the A-phase compensation magnetomotive force amplitude component corresponding to the second candidate compensation basis vector is -0.2K2, then the initial change in the A-phase compensation magnetomotive force amplitude is 0.6K1 - 0.2K2.

[0056] Furthermore, if the flux margin corresponding to the A-phase compensating magnetomotive force amplitude is large, then the constraint coefficient corresponding to the A-phase compensating magnetomotive force amplitude is also large, allowing a larger proportion of the initial change in the A-phase compensating magnetomotive force amplitude to be retained as the actual change. Conversely, if the difference between the working magnetic flux density and the allowable magnetic flux density corresponding to the B-phase compensating magnetomotive force amplitude is small, then the constraint coefficient corresponding to the B-phase compensating magnetomotive force amplitude is small, allowing a smaller proportion of the initial change in the B-phase compensating magnetomotive force amplitude to be determined as the actual change. Therefore, when reducing the weighted vibration characteristics of structural vibration measurement points such as the first measurement point on the long side wall of the oil tank, the possibility of large adjustments to the compensating magnetomotive force amplitude with a small flux margin can be reduced.

[0057] Finally, the current compensation magnetomotive force amplitude of each compensation unit is added to the corresponding actual change to obtain the compensation magnetomotive force amplitude to be applied. For example, if the current parameter value of the compensation magnetomotive force amplitude of phase A is M and the actual change is m, then the compensation magnetomotive force amplitude to be applied to phase A is M+m.

[0058] In this embodiment, after obtaining the compensated magnetomotive force parameters, the method further includes: When the compensation magnetic potential parameter is greater than the preset upper limit, the compensation magnetic potential parameter is set to the preset upper limit; when the compensation magnetic potential parameter is less than the preset lower limit, the compensation magnetic potential parameter is set to the preset lower limit.

[0059] The calculated amplitude of the compensating magnetomotive force to be applied is compared with the preset upper limit. When the amplitude of the compensating magnetomotive force to be applied is greater than the preset upper limit, the amplitude of the compensating magnetomotive force to be applied is determined as the preset upper limit. When the amplitude of the compensating magnetomotive force to be applied is less than the preset lower limit, the amplitude of the compensating magnetomotive force to be applied is determined as the preset lower limit.

[0060] S5: Control the compensation unit to apply the compensation magnetomotive force parameter, obtain the weighted vibration vector after application, and retain the compensation magnetomotive force parameter when the second norm of the applied weighted vibration vector is less than the second norm of the current weighted vibration vector, so as to reduce the noise of the transformer.

[0061] Based on the determined compensation magnetomotive force parameters, the compensation units corresponding to the A-phase core, B-phase core and C-phase core are controlled to perform compensation respectively. During the compensation process, the compensation magnetomotive force amplitude is kept constant, and the measurement is performed after a preset stabilization time. The preset stabilization time is not less than the length of one analysis time window.

[0062] After the preset settling time, the vibration characteristics of each structural vibration measurement point at each target frequency are reacquired. Based on the reacquired vibration characteristics of each structural vibration measurement point, the weighted vibration vector after applying the compensation magnetomotive force is obtained. When the L2 norm of the weighted vibration vector after applying the compensation magnetomotive force is less than the L2 norm of the current weighted vibration vector before applying the compensation magnetomotive force, it indicates that the weighted vibration response of the main acoustic radiation structure is reduced. Based on this, it is determined that the current compensation magnetomotive force parameter has a noise reduction effect. When the L2 norm of the weighted vibration vector after applying the compensation magnetomotive force is greater than or equal to the L2 norm of the current weighted vibration vector before applying the compensation magnetomotive force, the current compensation magnetomotive force parameter is canceled, so that the corresponding compensation unit is restored to the output state before applying the current compensation magnetomotive force parameter.

[0063] When the applied compensation magnetomotive force amplitude meets the vibration reduction condition, and the working magnetic flux density of each phase core recalculated based on the applied compensation magnetomotive force amplitude is not greater than the corresponding allowable magnetic flux density, the applied compensation magnetomotive force amplitude is retained; when the working magnetic flux density of any phase core is greater than the corresponding allowable magnetic flux density, the applied compensation magnetomotive force amplitude is canceled, and the corresponding compensation unit is restored to the output state before the applied compensation magnetomotive force amplitude was applied.

[0064] After retaining the amplitude of the compensating magnetomotive force to be applied, the current tap position is determined based on the current tap position status signal, the current temperature category is determined based on the preset temperature range of the current core temperature signal, and the current voltage category is determined based on the preset voltage range of the current three-phase voltage signal. The current tap position, current temperature category, and current voltage category are collectively determined as the current operating state category. The current operating state category, the amplitude of the compensating magnetomotive force to be applied, the vibration amplitude of each structural vibration measuring point at each target frequency after applying the compensating magnetomotive force, the foundation vibration characteristics, and the response matrix are associated and stored to form the compensation parameter combination for the transformer under the current operating state category.

[0065] During operation, the current tap position status signal, current core temperature signal, and current three-phase voltage signal are acquired. Based on these signals, the current tap position, current temperature category, and current voltage category are determined. These are then compared with the tap position, temperature category, and voltage category in each of the stored operating status categories. If each of these matches a stored operating status category, the compensation magnetomotive force amplitude from the corresponding compensation parameter group is retrieved, and the corresponding compensation unit outputs the compensation magnetomotive force. If no match is found, no compensation is performed.

[0066] It should be noted that the parameters such as the test disturbance amplitude, analysis time window length, and location weight in this paper can be determined based on the transformer's rated capacity, core structure, tank structure, winding parameters, compensation unit parameters, expected noise reduction frequency band, and test results. Those skilled in the art can select, calibrate, and adjust the above parameters based on conventional tests, simulation analysis, or engineering experience without affecting the implementation of the technical solution of this invention.

[0067] The acoustic radiation weights, constraint coefficients, or other weight parameters involved in the normalization process described in this article can be normalized according to the maximum, minimum, mean, sum, or preset benchmark values ​​of the corresponding parameters to ensure that the corresponding parameters are within the preset numerical range. The specific normalization benchmark and numerical range can be determined according to the actual application requirements.

[0068] For calculations involving denominators in the formulas presented in this paper, special cases where the denominator is zero or its absolute value is less than a preset threshold are not explained in detail for the sake of clarity. In practical implementation, alternative methods can be used to avoid division by zero or numerical anomalies, such as setting the corresponding result to zero, retaining the calculation result from the previous cycle, using a preset default value, skipping the corresponding measurement point or target frequency, adding a preset minimum positive number to the denominator, or using other equivalent numerical stabilization methods.

Claims

1. A noise reduction method for a three-dimensional wound-core amorphous alloy transformer, characterized in that, include: Multiple measuring points are set up on the transformer to obtain equipment data and vibration signals from multiple measuring points under normal operating conditions. Based on the equipment data and vibration signals, the vibration characteristics of each measuring point at each target frequency are extracted. A compensation unit is set in the core of the transformer. Positive and negative test disturbances are applied to the compensation unit. The first change, second change, symmetry coefficient and effective response of the vibration characteristics before and after the test are calculated, and the response matrix is ​​constructed. The magnetic flux margin and constraint coefficient are calculated based on the equipment data. The acoustic radiation weight is calculated based on the vibration characteristics and the position weight of the measuring point. A correction matrix is ​​constructed based on the response matrix, magnetic flux margin, constraint coefficient and acoustic radiation weight. Singular value decomposition is performed on the correction matrix to obtain the response basis vector and parameter space matrix. The current weighted vibration vector is obtained based on the vibration characteristics and acoustic radiation weight. The inverse of the current weighted vibration vector is defined as the target blanking direction. The projection coefficient of the response basis vector on the target blanking direction is calculated. The compensation magnetomotive force parameters are determined based on the projection coefficient, parameter space matrix and constraint coefficient. The control compensation unit applies compensation magnetomotive force parameters and obtains the weighted vibration vector after application. When the second norm of the applied weighted vibration vector is less than the second norm of the current weighted vibration vector, the compensation magnetomotive force parameters are retained to reduce transformer noise.

2. The method according to claim 1, characterized in that, Based on equipment data and vibration signals, vibration characteristics at each target frequency are extracted from each measuring point, including: The equipment data includes three-phase voltage signals. The A-phase voltage signal is extracted from the three-phase voltage signals. After performing a discrete Fourier transform on the A-phase voltage signal, the frequency corresponding to the frequency component with the largest amplitude is determined as the fundamental frequency. The second, fourth, and sixth harmonics of the fundamental frequency are determined as the target frequencies. The zero-crossing point of the phase A voltage signal in the same direction is extracted as the starting time. An analysis time window containing an integer number of fundamental frequency periods is set. The vibration signal of each measuring point is intercepted based on the analysis time window to obtain the discrete vibration sequence corresponding to each measuring point. The discrete Fourier transform result of each discrete vibration sequence at each target frequency is calculated to obtain the complex value of each measuring point at each target frequency. The magnitude of each complex value is calculated as the vibration amplitude, and the vibration amplitude is used as the vibration characteristic of the target frequency.

3. The method according to claim 1, characterized in that, The compensation unit is subjected to positive and negative test perturbations, including: Based on the compensation magnetomotive force amplitude of the current test compensation unit, a preset amplitude is added to perform a positive test disturbance; based on the compensation magnetomotive force amplitude, a preset amplitude is reduced to perform a negative test disturbance; the difference between the test vibration characteristics after applying the positive test disturbance and the basic vibration characteristics is calculated to obtain the first change; the difference between the test vibration characteristics after applying the negative test disturbance and the basic vibration characteristics is calculated to obtain the second change. The symmetry coefficient is determined based on the product of the first change and the second change. The first difference between the test vibration characteristics after applying a positive test disturbance and the test vibration characteristics after applying a negative test disturbance is calculated. The second difference between the positive test disturbance parameter value and the negative test disturbance parameter value is calculated. The effective response of the compensation magnetomotive force amplitude at the corresponding measurement point and target frequency is calculated based on the first difference and the second difference. The response matrix is ​​constructed based on the effective response, and the symmetry coefficient matrix is ​​constructed based on the symmetry coefficient.

4. The method according to claim 1, characterized in that, A correction matrix is ​​constructed based on the response matrix, flux margin, constraint coefficients, and acoustic radiation weights, including: The working magnetic flux density of the core under positive and negative test disturbances is determined respectively. The change in peak magnetic flux density is determined based on the difference between the working magnetic flux density during the test and the working magnetic flux density before the test. The constraint coefficient is determined based on the ratio of magnetic flux margin to the change in peak magnetic flux density. The product of the basic vibration characteristics of the measuring point and the corresponding position weight is calculated to obtain the weighted vibration amplitude. The sound radiation weight is determined based on the proportion of the weighted vibration amplitude of the measuring point to the sum of the weighted vibration amplitudes of all measuring points. The product of the effective response in the response matrix and the corresponding sound radiation weight, constraint coefficient and symmetry coefficient is used as the correction element. All correction elements are arranged according to the response matrix to form the correction matrix.

5. The method according to claim 1, characterized in that, Calculate the projection coefficients of the response basis vectors onto the target hiding direction, including: The product of the pre-test vibration features and the corresponding acoustic radiation weights is calculated to obtain the weighted vibration features. The weighted vibration features are arranged into the current weighted vibration vector. The opposite of each element in the current weighted vibration vector is taken to obtain the target blanking direction vector. The singular values ​​in the singular value matrix are arranged in descending order. The vectors corresponding to the first preset number of singular values ​​in the response space matrix are used as the response basis vectors. The product of each element in the response basis vector and the corresponding element in the target blanking direction vector is calculated, and all products are summed to obtain the projection coefficients corresponding to the response basis vectors.

6. The method according to claim 5, characterized in that, The compensation magnetomotive force parameters are determined based on the projection coefficients, parameter space matrix, and constraint coefficients, including: The response basis vectors are arranged in descending order of the absolute values ​​of the projection coefficients. The second-most preset number of response basis vectors are selected as candidate compensation basis vectors. All elements in the columns corresponding to the candidate compensation basis vectors are extracted from the parameter space matrix to obtain the parameter combination coefficients. The singular values ​​corresponding to the candidate compensation basis vectors are extracted from the singular value matrix. The ratio of the projection coefficients to the corresponding singular values ​​of the candidate compensation basis vectors is calculated to obtain the mode adjustment coefficients corresponding to the candidate compensation basis vectors. The compensation magnetomotive force amplitudes of the compensation unit corresponding to all candidate compensation basis vectors are synthesized to obtain the initial change in the compensation magnetomotive force amplitude of the compensation unit. Based on the initial change in the compensation magnetomotive force amplitude and the constraint coefficient corresponding to the compensation unit, the actual change in the compensation magnetomotive force amplitude of the compensation unit is determined. Based on the current compensation magnetomotive force amplitude and the actual change in the compensation magnetomotive force amplitude, the compensation magnetomotive force parameters are determined.

7. The method according to claim 6, characterized in that, After obtaining the compensation magnetomotive force parameters, the following is also included: When the compensation magnetic potential parameter is greater than the preset upper limit, the compensation magnetic potential parameter is set to the preset upper limit; when the compensation magnetic potential parameter is less than the preset lower limit, the compensation magnetic potential parameter is set to the preset lower limit.

Images