Rotating magnet magnetic field vector interference compensation method based on adaptive spectrum tracking

Abstract

A rotating magnet magnetic field vector interference compensation method based on adaptive spectrum tracking, comprising: establishing a reference digital model of a rotating magnetic component, obtaining a raw magnetic measurement vector signal sequence of a three-component magnetic sensor at a preset observation point, and constructing a time-domain observation vector; establishing a finite harmonic model of each axial component; establishing a spatial modulation function; combining the finite harmonic model with the spatial modulation function to obtain a modulated time-domain signal; performing fast Fourier transform on the modulated time-domain signal, adaptively updating second-order IIR notch filter parameters under candidate frequency point set constraints, and then performing inverse fast Fourier transform to obtain a noise-reduced time-domain magnetic field signal; obtaining a non-stationary magnetic field signal across rotating speeds based on the noise-reduced time-domain magnetic field signal; performing vector fusion and coordinate unification on the non-stationary magnetic field signal to output a final dynamic magnetic vector signal mapping after interference compensation. The present application can improve the filtering effect on noise frequency points and improve the interference compensation effect.

Description

Technical Field

[0001] This invention relates to the field of digital signal processing technology for magnetic measurement signals, and specifically to a method for compensating for magnetic field vector interference of a rotating magnet based on adaptive spectrum tracking. Background Technology

[0002] In applications such as weak magnetic signal measurement, magnetic anomaly detection, and magnetic field reconstruction on mobile platforms, three-component magnetic sensors are often used to acquire vector sequences of ambient magnetic field signals. However, rotating magnetic components (such as rotating bodies composed of permanent magnets or magnetized structures) present near the platform can generate periodic magnetic interference with significant amplitude at the sensor. This interference may partially overlap with the target weak magnetic signal in the frequency band, leading to a decrease in signal-to-noise ratio and feature distortion.

[0003] Existing suppression methods mostly employ notch filtering with fixed parameters or empirical filtering, which typically implicitly assumes that the interference frequency is constant. When the rotational angular velocity changes, the dominant interference frequency and its harmonics drift, making it difficult for a fixed notch filter to consistently align with the noise frequency, resulting in suppression failure or mis-suppressing of target components. Furthermore, when there is a fixed angle between the sensor's three-axis coordinates and the interference source's reference axis, or when changes in the platform's attitude alter the projection relationship, coupling occurs between the three-axis components, further increasing the difficulty of interference extraction and compensation. Summary of the Invention

[0004] In view of this, the present invention provides a method for compensating for magnetic field vector interference of a rotating magnet based on adaptive spectrum tracking, so as to improve the filtering effect of noise frequency points and improve the interference compensation effect.

[0005] A method for compensating for magnetic field vector interference in a rotating magnet based on adaptive spectrum tracking, comprising:

[0006] Step S1: Establish a reference digital model of the rotating magnetic component, obtain the original magnetic measurement vector signal sequence of the three-component magnetic sensor at the preset observation point, and perform segmented framing and normalization according to the rotation period or sliding window to construct the time-domain observation vector and its steady-state mapping for subsequent processing.

[0007] Step S2: Perform short-time spectrum tracking on the time-domain observation vector to estimate the rotating fundamental frequency and phase, and establish a finite harmonic model for each axial component under the constraint of the fundamental frequency to obtain the amplitude and phase parameters that vary with angular velocity.

[0008] Step S3: Establish a spatial modulation function characterizing circumferential non-uniformity and amplitude-phase fluctuations, and generate a set of candidate frequency points for the main peak and side bands accordingly.

[0009] Step S4: Combine the finite harmonic model with the spatial modulation function to obtain the modulated time-domain signal containing the main peak-sideband structure and its frequency domain expression.

[0010] Step S5: Perform fast Fourier transform on the modulated time-domain signal, and adaptively update the parameters of the second-order IIR notch filter under the constraint of the candidate frequency point set to realize the online positioning and suppression of the main peak and side band frequency points. Then, obtain the noise-suppressed time-domain magnetic field signal through inverse fast Fourier transform.

[0011] Step S6: The noise-suppressed time-domain magnetic field signals of different angular velocity ranges are resampled in the same angular domain and stitched together with a unified time axis to obtain non-stationary magnetic field signals across rotational speeds.

[0012] Step S7: Define a fixed orthogonal rotation matrix from the rotating reference axis to the magnetic sensor axis, perform vector fusion and coordinate unification on the non-stationary magnetic field signal across rotational speeds, and output the final dynamic magnetic vector signal mapping and related parameters after interference compensation.

[0013] The rotating magnet magnetic field vector interference compensation method based on adaptive spectrum tracking provided by the present invention has the following beneficial effects:

[0014] (1) This invention performs a fast Fourier transform on the modulated time-domain signal and constructs a second-order IIR notch filter that is linked to the rotational angular velocity. The second-order IIR notch filter detects the main peak and side band adaptively locates the frequency points that need to be suppressed, thereby obtaining the frequency-suppressed signal. This can effectively cope with the interference frequency drift caused by the rotational speed change, thereby improving the filtering effect of noise frequency points.

[0015] (2) Based on the finite harmonic model and spatial modulation function, this invention obtains the modulated time-domain signal and its frequency domain expression containing the main peak-side band structure. Furthermore, this invention defines a fixed orthogonal rotation matrix from the rotating reference axis to the magnetic sensor system. Based on the non-stationary magnetic field signal across rotational speed and the fixed orthogonal rotation matrix, the final dynamic magnetic vector signal mapping and related parameters after interference compensation are obtained. Therefore, this invention effectively incorporates the complex interference sources introduced by the rotating magnet, including spatial modulation caused by structural non-uniformity and interference factors such as sensor installation posture deviation, into the scope of consideration, thereby achieving effective suppression and compensation of non-stationary vector magnetic interference signals and improving the reliability of magnetic measurement data. Attached Figure Description

[0016] Figure 1 A flowchart illustrating a method for compensating for magnetic field vector interference of a rotating magnet based on adaptive spectrum tracking, provided in an embodiment of the present invention;

[0017] Figure 2 This is a schematic diagram of frequency domain filtering;

[0018] Figure 3 A comparison of the x-axis time-domain images before and after interference compensation;

[0019] Figure 4A comparison of the y-axis time-domain images before and after interference compensation;

[0020] Figure 5 This is a comparison of the z-axis time-domain images before and after interference compensation. Detailed Implementation

[0021] To facilitate understanding of the present invention, a more complete description will be given below with reference to various embodiments. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the specification of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0023] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain embodiments of the present invention, and should not be construed as limiting the present invention.

[0024] Please see Figure 1 The present invention provides a method for compensating for magnetic field vector interference of a rotating magnet based on adaptive spectrum tracking, comprising steps S1 to S7:

[0025] Step S1: Establish a reference digital model of the rotating magnetic component, obtain the original magnetic measurement vector signal sequence of the three-component magnetic sensor at the preset observation point, and perform segmentation and normalization according to the rotation period or sliding window to construct the time-domain observation vector and its steady-state mapping for subsequent processing.

[0026] In the specific implementation, the simulation model of the rotating magnetic component places three-component magnetic sensors directly below the rotation axis, with a horizontal radius of r and a vertical spacing of h. Triaxial magnetic measurement data are synchronously acquired at a sampling frequency Fs. The acquired data is segmented and framed according to the rotation period or a sliding window, and each frame undergoes detrending, amplitude normalization, and time alignment to ensure the robustness of subsequent spectrum tracking and parameter estimation. The observation data comes from calibration experiments, playback data, and numerical solution results.

[0027] Specifically, the time-domain observation vector and its steady-state mapping satisfy the following equation:

[0028]

[0029] in, express At all times, the rotational angular velocity The periodic steady-state magnetic field vector mapping of a rotating magnet. for At all times, the rotational angular velocity Measured by the lower magnetic sensor shaft electromagnetic signal components, for At all times, the rotational angular velocity Measured by the lower magnetic sensor shaft electromagnetic signal components, for At all times, the rotational angular velocity Measured by the lower magnetic sensor shaft electromagnetic signal components, This indicates transpose.

[0030] By using reference data and frame processing, the statistical characteristics and spectral structure (including main harmonics and higher harmonics) of the interference magnetic field signals introduced by the rotating magnetic component at different angular velocities can be obtained. This steady-state magnetic field vector mapping will be used to constrain subsequent short-time spectrum tracking, candidate frequency point set generation, and adaptive notch filter parameter updates, thereby improving the traceability and repeatability of non-stationary interference.

[0031] Step S2: Perform short-time spectrum tracking on the time-domain observation vector to estimate the rotating fundamental frequency and phase, and establish a finite harmonic model for each axial component under the constraint of the fundamental frequency to obtain the amplitude and phase parameters that vary with angular velocity.

[0032] For each type of rotational angular velocity obtained in step S1 The magnetic measurement signal data is used to perform short-time spectrum tracking on the three components of the magnetic field (sensor X, Y, and Z axes) to estimate the amplitude and phase parameters that change with angular velocity, and to establish a periodic steady-state finite-order harmonic model.

[0033] In this embodiment, the finite harmonic model of each axial component satisfies the following equation:

[0034]

[0035] in, Indicates axial index. , Represents rotational angular velocity corresponding Finite harmonic model of axial component, and The first First harmonic in the axial direction The amplitude and phase can be estimated using least squares. and , The highest order of finite harmonic expansion. Angular velocity of rotation The corresponding fundamental frequency; To model the residual term, which is used to characterize unmodeled factors such as finite harmonic cutoff error and measurement noise.

[0036] Step S3: Establish a spatial modulation function that characterizes circumferential non-uniformity and amplitude-phase fluctuations, and generate a set of candidate frequency points for the main peak and side bands accordingly.

[0037] To characterize the parameter differences in the circumferential magnetization units of the rotating magnetic component, a spatial modulation factor is introduced to explain the resulting interference side frequencies. The circumferential magnetization of the rotating magnet is not perfectly uniform; for example, if a magnet is slightly shorter, has slightly weaker remanence, or has different assembly gaps, the normally constant magnetic field amplitude / air gap magnetic flux density will vary with the rotational phase angle. It fluctuates slowly in the circumferential direction. , This is the mechanical angular velocity. The modulation can be considered as the phase angle. The low-order periodic fluctuations are used to generate candidate frequency sets for the main peak and side bands, thereby providing prior constraints for the frequency positioning of the subsequent adaptive notch filter.

[0038] Specifically, the spatial modulation function satisfies the following equation:

[0039]

[0040] in, Rotation phase angle The corresponding spatial modulation function, and The first Modulation depth and phase of spatial harmonics of order 1 It is the highest order of spatial modulation harmonics.

[0041] Step S4: Combine the finite harmonic model with the spatial modulation function to obtain the modulated time-domain signal containing the main peak-sideband structure and its frequency domain expression.

[0042] The modulated time-domain signal satisfies the following equation:

[0043]

[0044] in, This is the modulated time-domain signal.

[0045] Modulated time-domain signal The spectrum in Symmetrical lateral bands appear at this location. The modulation frequency is defined by the non-uniform circumferential magnetization of the rotating magnet, and the first-order approximation of the sideband amplitude is... Proportional.

[0046] Step S5: Perform a fast Fourier transform on the modulated time-domain signal, and adaptively update the parameters of the second-order IIR notch filter under the constraint of the candidate frequency point set to realize the online positioning and suppression of the main peak and side band frequency points. Then, obtain the noise-suppressed time-domain magnetic field signal through inverse fast Fourier transform.

[0047] Specifically, step S5 includes:

[0048] For the modulated time-domain signal Perform a Fast Fourier Transform to obtain the spectrum. ;

[0049] Constructing a second-order IIR notch filter that is linked to the rotational angular velocity The expression is:

[0050]

[0051]

[0052] in, For the first For the left and right side band frequencies, For magnetic field sampling period, To and The corresponding normalized angular frequency, for Transform the complex variable, As the rotational angular velocity The linked pole magnitude is used to control the notch filter bandwidth. The peak value is detected and tracked based on the candidate frequency point set, and the tracking result is adaptively updated to keep the notch center and bandwidth synchronized with the instantaneous fundamental frequency drift.

[0053] The main peak and sideband frequencies to be suppressed are detected using a second-order IIR notch filter. The expression is:

[0054]

[0055] in, The modulation frequency is defined by the non-uniform circumferential magnetization of the rotating magnet;

[0056] In the spectrum By removing frequency points, the frequency-suppressed signal is obtained. The expression is:

[0057]

[0058] in, The symbol for multiplication;

[0059] Signal after frequency suppression Perform an inverse fast Fourier transform to obtain the noise-suppressed time-domain magnetic field signal. .

[0060] Step S6: The noise-suppressed time-domain magnetic field signals of different angular velocity ranges are resampled in the same angular domain and spliced ​​with a unified time axis to obtain non-stationary magnetic field signals across rotational speeds.

[0061] To construct a dynamic magnetic interference sample covering takeoff, landing, acceleration, and deceleration processes for cross-speed transient analysis, the noise-suppressed time-domain magnetic field sequence needs to be stitched together sequentially along a unified time axis to obtain the cross-speed time-domain magnetic field signal. Each data segment is first resampled to an equiangular domain according to its estimated phase, and then stitched back along the unified time axis. This reduces the impact of instantaneous frequency stretching caused by speed changes on time-domain stitching and subsequent compensation table establishment. The non-stationary magnetic field signal across speeds satisfies the following equation:

[0062]

[0063]

[0064]

[0065] in, Indicates axial direction The non-stationary magnetic field signal across rotational speeds, For the first Each time interval is a dividing point. For the first Each time interval is a dividing point. Represents rotational angular velocity Duration, mechanical rotation speed The time within the duration, The total number of points in the time interval. express The result is obtained by splicing the noise-suppressed time-domain magnetic field sequence on a unified time axis according to the time sequence corresponding to the time period.

[0066] Step S7: Define a fixed orthogonal rotation matrix from the rotating reference axis to the magnetic sensor axis, perform vector fusion and coordinate unification on the non-stationary magnetic field signal across rotational speeds, and output the final dynamic magnetic vector signal mapping and related parameters after interference compensation.

[0067] In this context, considering the fixed angle between the magnetic sensor axis system and the rotating reference axis system, and treating it as a constant geometric relationship under rigid mounting, the relationship from the rotating reference axis system is defined. To magnetic sensor system Fixed orthogonal rotation matrix , , It is a special orthogonal group in three dimensions. It only redistributes amplitude and phase across the three axes without changing the frequency set.

[0068] To ensure that the model output is completely consistent with the triaxial readings of the magnetic sensor in the coordinate system, the reconstructed signal in the rotating reference frame needs to be... pass With unified body coordinates, the final dynamic magnetic vector signal mapping satisfies the following equation:

[0069]

[0070] in, For the final dynamic magnetic vector signal mapping, it completes the interference compensation.

[0071] The effect of this is that after frequency suppression and cross-speed reconstruction are completed, the vector results under the reference axis can still be made consistent to the sensor reading coordinates by fixing the rotation, thereby absorbing the triaxial cross-coupling introduced by the installation error, and ensuring that the relative position of the main peak and the side band, the triaxial energy relationship and geometric consistency are maintained throughout the process.

[0072] The present invention will now be verified and tested, such as... Figure 2 As shown, compared to the original spectrum obtained after the Fast Fourier Transform, after removing the frequency points that need to be suppressed using the method of this invention, the main frequency that is retained corresponds to the basic period and main harmonics of the rotating magnetic field of the rotating magnet. It belongs to the stable and physically meaningful part of the signal, while other sideband frequencies caused by modulation or noise have been filtered out.

[0073] like Figures 3 to 5 As shown, at speeds of 1200 rpm, 1600 rpm, and 800 rpm, before signal extraction and compensation, the time-domain image exhibits obvious envelope fluctuations and occasional spikes. After compensation, sidebands and unsteady-state energy are largely removed, retaining only the baseline harmonic groups related to the speed and the regular envelope generated by the non-orthogonal matrix. This waveform characteristic indicates that modulation effects unrelated to the speed or induced by circumferential inhomogeneity have been effectively stripped away, retaining only components strictly in sync with the rotational frequency.

[0074] In summary, the adaptive spectrum tracking-based rotating magnet magnetic field vector interference compensation method according to the above embodiments has the following beneficial effects:

[0075] (1) This invention performs a fast Fourier transform on the modulated time-domain signal and constructs a second-order IIR notch filter that is linked to the rotational angular velocity. The second-order IIR notch filter detects the main peak and side band adaptively locates the frequency points that need to be suppressed, thereby obtaining the frequency-suppressed signal. This can effectively cope with the interference frequency drift caused by the rotational speed change, thereby improving the filtering effect of noise frequency points.

[0076] (2) Based on the finite harmonic model and spatial modulation function, this invention obtains the modulated time-domain signal and its frequency domain expression containing the main peak-side band structure. Furthermore, this invention defines a fixed orthogonal rotation matrix from the rotating reference axis to the magnetic sensor system. Based on the non-stationary magnetic field signal across rotational speed and the fixed orthogonal rotation matrix, the final dynamic magnetic vector signal mapping and related parameters after interference compensation are obtained. Therefore, this invention effectively incorporates the complex interference sources introduced by the rotating magnet, including spatial modulation caused by structural non-uniformity and interference factors such as sensor installation posture deviation, into the scope of consideration, thereby achieving effective suppression and compensation of non-stationary vector magnetic interference signals and improving the reliability of magnetic measurement data.

[0077] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for compensating for magnetic field vector interference in a rotating magnet based on adaptive spectrum tracking, characterized in that, include: Step S1: Establish a reference digital model of the rotating magnetic component, obtain the original magnetic measurement vector signal sequence of the three-component magnetic sensor at the preset observation point, and perform segmented framing and normalization according to the rotation period or sliding window to construct the time-domain observation vector and its steady-state mapping for subsequent processing. Step S2: Perform short-time spectrum tracking on the time-domain observation vector to estimate the rotating fundamental frequency and phase, and establish a finite harmonic model for each axial component under the constraint of the fundamental frequency to obtain the amplitude and phase parameters that vary with angular velocity. Step S3: Establish a spatial modulation function characterizing circumferential non-uniformity and amplitude-phase fluctuations, and generate a set of candidate frequency points for the main peak and side bands accordingly. Step S4: Combine the finite harmonic model with the spatial modulation function to obtain the modulated time-domain signal containing the main peak-sideband structure and its frequency domain expression. Step S5: Perform fast Fourier transform on the modulated time-domain signal, and adaptively update the parameters of the second-order IIR notch filter under the constraint of the candidate frequency point set to realize the online positioning and suppression of the main peak and side band frequency points. Then, obtain the noise-suppressed time-domain magnetic field signal through inverse fast Fourier transform. Step S6: The noise-suppressed time-domain magnetic field signals of different angular velocity ranges are resampled in the same angular domain and stitched together with a unified time axis to obtain non-stationary magnetic field signals across rotational speeds. Step S7: Define a fixed orthogonal rotation matrix from the rotating reference axis to the magnetic sensor axis, perform vector fusion and coordinate unification on the non-stationary magnetic field signal across the rotational speed, and output the final dynamic magnetic vector signal mapping and related parameters after interference compensation. In step S1, the time-domain observation vector and its steady-state mapping satisfy the following equation: in, express At all times, the rotational angular velocity The periodic steady-state magnetic field vector mapping of a rotating magnet. for At all times, the rotational angular velocity Measured by the lower magnetic sensor Axis electromagnetic signal sequence, for At all times, the rotational angular velocity Measured by the lower magnetic sensor Axis electromagnetic signal sequence, for At all times, the rotational angular velocity Measured by the lower magnetic sensor Axis electromagnetic signal sequence, Indicates transpose; In step S2, the finite harmonic models of each axial component satisfy the following equation: in, Indicates axial index. , Represents rotational angular velocity corresponding Finite harmonic model of axial component, and The first First harmonic in the axial direction The amplitude and phase on, The highest order of finite harmonic expansion. Angular velocity of rotation The corresponding fundamental frequency, To model the residual terms; In step S3, the spatial modulation function satisfies the following equation: in, Rotation phase angle The corresponding spatial modulation function, and The first Modulation depth and phase of spatial harmonics, This is the highest order of spatially modulated harmonics; Step S5 specifically includes: For the modulated time-domain signal Perform a Fast Fourier Transform to obtain the spectrum. ; Constructing a second-order IIR notch filter that is linked to the rotational angular velocity The expression is: in, For the first For the left and right side band frequencies, For magnetic field sampling period, To and The corresponding normalized angular frequency, for Transform the complex variable, As the rotational angular velocity The magnitude of the linked poles; The main peak and sideband frequencies to be suppressed are detected using a second-order IIR notch filter. The expression is: in, The modulation frequency is defined by the non-uniform circumferential magnetization of the rotating magnet; In the spectrum By removing frequency points, the frequency-suppressed signal is obtained. The expression is: in, The symbol for multiplication; Signal after frequency suppression Perform an inverse fast Fourier transform to obtain the noise-suppressed time-domain magnetic field signal. ; In step S7, the final dynamic magnetic vector signal mapping satisfies the following equation: in, For the final dynamic magnetic vector signal mapping, To rotate from the reference axis To magnetic sensor system A fixed orthogonal rotation matrix, Indicates axial direction The non-stationary magnetic field signal across rotational speeds.

2. The method for compensating for magnetic field vector interference of a rotating magnet based on adaptive spectrum tracking according to claim 1, characterized in that, In step S4, the modulated time-domain signal satisfies the following equation: in, This is the modulated time-domain signal.

3. The method for compensating for magnetic field vector interference of a rotating magnet based on adaptive spectrum tracking according to claim 2, characterized in that, In step S6, the non-stationary magnetic field signal across rotational speeds satisfies the following equation: in, For the first Each time interval is a dividing point. For the first Each time interval is a dividing point. Angular velocity of rotation The time within the duration, The total number of points in the time interval. express The result is obtained by splicing the noise-suppressed time-domain magnetic field sequence on a unified time axis according to the time sequence corresponding to the time period.

Images