Radar imaging optimization method for basin-type insulator defect detection

Abstract

The invention discloses a radar imaging optimization method for basin-type insulator defect detection, and belongs to the technical field of electrical equipment nondestructive testing. The method comprises the following steps: 1, collecting echo signals by adopting a UWB-MIMO radar, and calculating a multi-channel delay parameter based on a two-section propagation model; 2, generating a preliminary three-dimensional image by using a DAS algorithm; 3, compensating signal depth attenuation through TGC to obtain an intermediate image with balanced intensity; 4, PSCF weighting is applied to suppress sidelobe artifacts, and the image contrast and resolution are improved; and 5, performing normalization and visual display on the final image. According to the method, through the synergistic effect of the TGC and the PSCF, signal attenuation can be effectively compensated, imaging artifacts are inhibited, the contrast ratio and the resolution ratio of the image are remarkably improved while real defect details are reserved, and therefore the accuracy and the reliability of defect detection of the basin-type insulator are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of non-destructive testing of power equipment, in particular to a radar imaging optimization method for detecting defects of a basin-type insulator. BACKGROUND

[0002] Radar imaging technology has great application potential in the detection of defects of a basin-type insulator of a gas-insulated switchgear due to its non-contact and high penetration. In this technology, beamforming is a core link that determines the imaging quality. The widely used delay-and-sum (DAS) beamforming technology is simple and effective, but its inherent wide main lobe and high side lobe defects result in low spatial resolution and contrast of the image, which is difficult to meet the needs of high-precision detection. To overcome this defect, adaptive beamforming technology has emerged, and its core idea is to adaptively construct a weighting factor according to the relevant information extracted from the received signal to suppress artifacts and enhance real targets.

[0003] However, when the existing adaptive imaging method is directly applied to such specific and challenging scenarios as the detection of defects of a basin-type insulator, its effectiveness will be restricted by two aspects. First, as a dielectric target with a certain thickness, the electromagnetic wave propagates in the basin-type insulator and suffers significant attenuation. Especially for ultra-wideband (UWB) radar signals, not only the return amplitude of deep defects is much lower than that of shallow defects, but also the frequency-selective attenuation (dispersion attenuation) characteristics of the material will cause pulse distortion, further degrading the distance resolution, and most existing adaptive methods fail to effectively compensate for this. Second, the defects in the basin-type insulator are not simply point targets, and the complex scattering characteristics of a crack as an irregularly shaped extended target will cause inherent fluctuations in the return signals of different antenna channels. If the traditional coherence factor based on the strict signal consistency assumption is directly used for weighting, the algorithm will not be able to distinguish between the normal signal fluctuations caused by the shape of the target itself and the side lobes and background clutter formed by beam divergence, so it is easy to misjudge the former as the latter, thus excessively suppressing the real defect signal and causing loss of target energy and distortion of image shape. SUMMARY

[0004] The purpose of the present application is to provide an adaptive radar imaging optimization method for basin-type insulator defects that combines time gain compensation (TGC) and power spectrum coherence factor (PSCF). The present application aims to simultaneously solve the two core problems of signal depth attenuation and imaging artifacts through a collaborative processing procedure to obtain a high-quality basin-type insulator defect image with balanced intensity, clean background, and clear details, thereby improving the accuracy and reliability of detection.

[0005] To achieve the above purpose, the present application provides the following technical solutions:

[0006] A radar imaging optimization method based on polar coordinates includes the following steps:

[0007] Step 1: To improve the focusing accuracy and positioning precision of the imaging, this invention employs a precise two-stage propagation model when calculating the propagation delay parameters. This model decomposes the total path of the electromagnetic wave from the antenna to the imaging point into two independent paths: one in the air and the other in the basin-type insulator medium. Delay calculations are then performed for each of these two paths using their respective propagation velocities. Total propagation delay. The result is obtained by adding the delays of these two parts, as shown in formula (1):

[0008] (1)

[0009] Among them, the air propagation delay Calculated from equation (2)

[0010] (2)

[0011] Among them, the propagation delay in the basin insulator Calculated from equation (3)

[0012] (3)

[0013] in, An imaging point with an intensity to be calculated within a preset three-dimensional imaging space. The three-dimensional spatial coordinate vector, and They are the components of the first The transmitting and receiving antennas of each transceiver channel are three-dimensional spatial coordinate vectors in the same coordinate system. It is the speed at which electromagnetic waves propagate in the air medium. It is the propagation speed of electromagnetic waves inside the dielectric material of the basin-type insulator, which is determined by the propagation speed in the air. and the relative permittivity of the dielectric material Confirmed, the calculation formula is as follows: , and They are the first Channel connection to transmitting antenna With imaging point and connecting the receiving antenna With imaging point The coordinate vectors of the intersection points of the two actual propagation paths with the surface of the basin insulator.

[0014] Step 2, this step utilizes the delay parameters determined in step 1. The acquired raw echo signal The delay-and-sum beamforming is performed to focus and coherently superimpose the signals from different channels. The purpose of this step is to converge the signal energy from different channels to the corresponding spatial position by means of coherent superposition. As shown in equation (4):

[0015] (4)

[0016] The output of this step is the preliminary three-dimensional imaging result , which can preliminarily locate the defects, but has problems of signal intensity distortion and artifacts.

[0017] Step 3, the purpose of this step is to correct the intensity of the preliminary image to compensate for the depth attenuation of the signal. A gain factor is used to weight the preliminary image to obtain an intermediate image with balanced intensity . As shown in equation (5):

[0018] (5)

[0019] The present application proposes an optimized exponential gain model by introducing a reference distance for the near-field detection of the basin-type insulator, which is dominated by material absorption attenuation. By accurately calibrating the gain starting point (when ) to 1, the model effectively avoids unnecessary over-amplification of shallow signals, significantly enhancing the numerical stability of the algorithm. By setting as the starting distance of the imaging area, it can ensure that the gain factor starts from 1 and increases smoothly in the entire region of interest. is a gain factor based on material absorption attenuation compensation, and its expression is shown in equation (6):

[0020] (6)

[0021] wherein, is an equivalent round-trip propagation distance used to calculate the gain compensation, which is defined as the total path length of the electromagnetic wave propagating from the geometric center of the antenna array to the imaging point , and then returning from the imaging point to the geometric center of the antenna array, is the minimum round-trip propagation distance of the imaging area, is an attenuation coefficient related to the center frequency of the radar signal and the electromagnetic properties of the basin-type insulator material. The expression of is shown in equation (7):

[0022] (7)

[0023] wherein, is the center frequency of the radar signal, is the relative permittivity of the dielectric material, is the propagation speed of electromagnetic wave in air medium, is the loss tangent of the dielectric.

[0024] Step 4, image quality enhancement and artifact suppression. This step optimizes the intermediate image by applying a power spectrum coherence factor (PSCF) as adaptive weight to suppress artifacts and enhance contrast and resolution. Unlike traditional coherence factor, the PSCF method adopted by this invention has significant advantage for defects as extended targets: by analyzing the energy proportion within a wider low-frequency window of power spectrum, instead of only the zero-frequency component corresponding to perfect coherence, this method can better adapt to normal signal fluctuation caused by complex scattering characteristics of targets, while suppressing sidelobes and background clutter caused by beam divergence, thus avoiding excessive suppression and morphological distortion of real defects. The expression is shown in equation (8):

[0025] (8)

[0026] wherein, power spectrum is obtained by performing discrete Fourier transform on the coherent signal vector composed of complex-valued signals of imaging point after N-channel delay-and-sum focusing, and calculating the square of its amplitude. The value of this factor is close to 1 at real targets, and much less than 1 at artifacts.

[0027] wherein, the expression of the final optimized image is shown in equation (9):

[0028] (9)

[0029] wherein, is the intensity-equalized intermediate image obtained in step 3, is the power spectrum coherence factor calculated in step 4.

[0030] Step 5, this step visualizes the data volume of the final optimized image for user to analyze and diagnose the defects of post insulators.

[0031] Visualization processing can include: extracting a series of two-dimensional slice images along a certain coordinate axis (such as X, Y, Z axis) and displaying them in grayscale or pseudo-color map. Before display, the images can also be post-processed such as logarithmic compression or normalization to enhance the dynamic range and visual effect of the images.

[0032] wherein the normalized expression is shown as equation (10):

[0033] (10)

[0034] wherein, is the normalized final image, is the maximum intensity value of the optimized image, is the maximum intensity value of the preliminary image.

[0035] The present application has the following advantages and beneficial effects:

[0036] Firstly, the present application effectively compensates the signal propagation attenuation in the medium of the basin-type insulator by introducing time gain compensation (TGC). This step can enhance the intensity of the deep weak defect signal to the same amplitude as the shallow signal, fundamentally solving the diagnostic deviation caused by intensity distortion, making the defect characteristics of different depths have objective comparability, and thus ensuring the integrity of the detection. Secondly, through the adaptive analysis of the spectral coherence of each channel signal by power spectral coherence factor (PSCF), the present method can accurately suppress sidelobes and grating lobes and other artifacts, and better adapt to the imaging characteristics of the defect as an extended target, avoiding excessive suppression of the real signal. The final result of this synergistic effect is that the contrast and spatial resolution of the image are greatly improved, making the defect target clear and the background clean. In summary, the present application can generate high-quality and high-fidelity defect diagnosis images while ensuring computational efficiency, significantly improving the accuracy and reliability of basin-type insulator defect detection. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 is a flow chart of the radar imaging optimization method of the present application;

[0038] Figure 2 is a schematic diagram of the application scenario in the embodiment of the present application;

[0039] Figure 3 is a schematic diagram of the experimental model of the basin-type insulator crack defect detection in the present application;

[0040] Figure 4 is a slice image of the crack imaging result obtained by applying the traditional DAS algorithm;

[0041] Figure 5 is a slice image of the crack imaging result obtained by applying the method of the present application. DETAILED DESCRIPTION

[0042] The present application is further illustrated by the following examples. It should be noted that the examples are only used to illustrate and explain the present application, and those skilled in the art can make various modifications or supplements to the described specific examples or use similar ways to replace, as long as the modifications or supplements do not deviate from the concept of the present application or exceed the scope defined by the claims, and should be considered to fall within the protection scope of the present application.

[0043] Example: Adaptive radar imaging detection of surface cracks of 252kV basin insulator. This example aims to verify the effectiveness and superiority of the method of the present application in detecting typical crack defects of basin insulators.

[0044] (I) Experimental platform and object to be detected:

[0045] The experimental platform used in this example is shown in Figure 2 The core of the detection system is a portable, low-power pulse ultra-wideband multiple-input multiple-output (Pulse UWB-MIMO) radar array, which is composed of nine Novelda X4 radar transceiver modules that can work synchronously. The radar modules are fixed on a 3D printed array support to form a 3x3 physical antenna array, and are equivalent to a large-aperture array containing 81 virtual transceiver channels through virtual array technology, to improve the spatial resolution under the limited hardware size. The working center frequency of the radar system is 7.29 GHz.

[0046] The object to be detected is a ventilation-type basin insulator with a rated voltage level of 252kV, whose main material is epoxy resin, with a diameter of 450mm and a thickness of 120mm. To simulate the crack defects that may occur in actual operation, this example uses mechanical scribing to pre-prepare a crack with a size of 20mm in length, 2mm in width and 4mm in depth on the surface of the insulator, as shown in Figure 3 To suppress the electromagnetic scattering interference of non-target areas such as the central metal insert and the outer metal flange of the insulator, these areas are covered with wave-absorbing materials before the experiment. During detection, the radar array is placed opposite the crack area, 220mm away from the surface of the insulator.

[0047] (II) Specific implementation of imaging optimization method:

[0048] According to the technical solution of the present application, the collected radar echo data is processed, and the specific implementation steps are as follows:

[0049] Step 1, data acquisition and delay parameter determination:

[0050] First, control the radar array through the SPI communication protocol to synchronously collect the original echo signals of all 81 virtual channels Subsequently, a three-dimensional imaging space covering the crack defect region is defined. In this embodiment, the X-axis and Y-axis ranges of this imaging space are both set to [-0.1m, 0.1m], and the Z-axis (depth) range is set to [0.22m, 0.35m]. Finally, the multi-channel propagation delay parameters are calculated using the two-stage propagation model described in this invention. Specifically, the signal propagation delay in air is first calculated based on the distance between the radar and the insulator surface (220mm), and then the relative permittivity of the epoxy resin material is set to... Calculate the propagation delay of the signal after it enters the medium. Total propagation delay. The result is obtained by adding the delays of these two parts, as shown in formula (1):

[0051] (1)

[0052] Among them, the air propagation delay Calculated from equation (2)

[0053] (2)

[0054] Among them, the propagation delay in the basin insulator Calculated from equation (3)

[0055] (3)

[0056] in, An imaging point with an intensity to be calculated within a preset three-dimensional imaging space. The three-dimensional spatial coordinate vector, and They are the components of the first The transmitting and receiving antennas of each transceiver channel are three-dimensional spatial coordinate vectors in the same coordinate system. It is the speed at which electromagnetic waves propagate in the air medium. It is the propagation speed of electromagnetic waves inside the dielectric material of the basin-type insulator, which is determined by the propagation speed in the air. and the relative permittivity of the dielectric material Confirmed, the calculation formula is as follows: , and They are the first Channel connection to transmitting antenna With imaging point and connecting the receiving antenna With imaging point The coordinate vectors of the intersection points of the two actual propagation paths with the surface of the basin insulator.

[0057] Step 2: Beamforming of the preliminary image.

[0058] Using the delay parameter determined in step 1 , the 81-channel raw echo signals collected The delay-and-sum algorithm is applied to perform delay focusing and coherent summation. This process is performed for each two-dimensional imaging slice one by one, and all generated slice images are combined to form the preliminary three-dimensional imaging result . The slice imaging effect is shown in Figure 4 The expression of the preliminary three-dimensional imaging result is shown in equation (4):

[0059] (4)

[0060] Step 3: Propagation attenuation correction and intensity equalization.

[0061] The preliminary imaging result obtained in step 2 is subjected to time gain compensation to correct the signal propagation attenuation in the medium, and an intensity-equalized intermediate image is obtained. As shown in equation (5):

[0062] (5)

[0063] where the gain factor is shown in equation (6):

[0064] (6)

[0065] where is the equivalent round-trip propagation distance used to calculate the gain compensation, which is defined as the total path length of the electromagnetic wave propagating from the geometric center of the antenna array to the imaging point , and then returning from the imaging point to the geometric center of the antenna array, and is set to the starting round-trip distance of the imaging region, 0.44 m. The material attenuation coefficient is determined by equation (7), following a scientific method combining theoretical calculation and experimental optimization.

[0066] (7)

[0067] First, the typical electromagnetic parameters of epoxy resin (relative permittivity , loss tangent ) and the center frequency of the radar in this embodiment (7.29 GHz), the theoretical value of the attenuation coefficient is calculated to be about 3.05 Np / m. The theoretical calculation value is used as an initial reference for parameter selection, and is further optimized through calibration experiments on the same type of standard sample, and the effective attenuation coefficient applicable to this type of pot-type insulator is finally determined to be 4 Np / m. The preset value can be directly applied to the detection of subsequent unknown defects.

[0068] Step 4, image quality enhancement and artifact suppression. This step aims to further optimize the intensity-balanced intermediate image obtained in step 3 , by calculating and applying a power spectrum coherence factor (PSCF) as an adaptive weight, to suppress sidelobe artifacts and improve image contrast and resolution. Unlike traditional coherence factors based on strict signal consistency assumptions, the PSCF method used in this invention has a significant imaging advantage for irregularly shaped extended targets such as cracks. This is because PSCF can better tolerate normal amplitude and phase fluctuations caused by the complex scattering characteristics of extended targets by analyzing the energy proportion in the low-frequency window of the power spectrum. Therefore, by reasonably setting the width of the low-frequency window , this method can effectively identify and suppress artifacts while preserving the true shape and energy information of extended targets to the greatest extent, avoiding the over-suppression and image distortion problems that may be caused by traditional coherence factors. The expression of the power spectrum coherence factor (PSCF) is shown in equation (8):

[0069] (8)

[0070] In this embodiment, the calculation parameters of PSCF factor are set as follows: the Fourier transform length L is set to 128, which is the smallest integer power of 2 greater than the total number of channels 81, aiming to ensure the highest computational efficiency of fast Fourier transform (FFT). The low-frequency window half-width B is empirically preferred to be 2, which is wide enough to accommodate the spectral energy of extended targets such as cracks, and narrow enough to effectively distinguish artifacts. Finally, the suppression index is set to 2, which is a nonlinear weighting of the power of 2 that can achieve a robust balance between effectively suppressing sidelobes and preserving the energy of real targets, avoiding excessive loss of noisy signals that may be caused by high powers.

[0071] The expression of the final optimized image is shown in equation (9):

[0072] (9)

[0073] where, is the intensity-balanced intermediate image obtained in step 3, ​​Power spectrum coherence factor calculated in step 4.

[0074] In step 5, after normalization and logarithmic compression of the final optimized image, two-dimensional slice visualization is performed, and the effect is as shown in Figure 5

[0075] So far, a radar imaging optimization method for detecting defects of basin insulators is implemented.

[0076] (III) Result and effect analysis:

[0077] In order to intuitively verify the effectiveness and advancement of the method of the present application, Figure 4 and Figure 5 respectively show the imaging results after applying the traditional DAS algorithm (i.e. the preliminary result of step 2 of the present application) and the complete optimization method (i.e. DAS+TGC+PSCF) of the present application on the most representative z=220mm depth slice.

[0078] As shown in Figure 4 , the imaging result of the traditional DAS algorithm has obvious defects. Although the target area can be identified in the figure, the energy distribution is diffuse, the main lobe response is too wide, and the target core area is surrounded by a large number of sidelobes and grating lobes with similar intensity. The energy of these artifacts seriously pollutes the image background, resulting in low overall contrast of the image, unclear real profile of the crack defect, and difficulty in accurate morphological analysis and size evaluation.

[0079] Figure 5 The imaging result after applying the complete optimization method of the present application is shown. Compared with the traditional method, the image quality is significantly improved, and this improvement is due to the synergistic optimization effect of the time gain compensation (TGC) and the power spectrum coherence factor (PSCF) modules. This method first ensures that the original signals of defects at any depth can be fully compensated for energy through the TGC step, laying a high signal-to-noise ratio data foundation for subsequent accurate imaging. On this basis, the adaptive weighting of PSCF shows the most significant visual optimization effect, and the sidelobe and grating lobe artifacts surrounding the target are effectively suppressed, thereby realizing high-precision positioning of the crack defect. Finally, due to the deep purification of the background area, the intensity difference between the target and the background is greatly enhanced, resulting in significant improvement in the contrast and signal-to-noise ratio of the image.

[0080] The above is an exemplary description of the invention. Obviously, the specific implementation of the present application is not limited by the above method. As long as this non-essential improvement is made by using the method concept and technical solution of the present application, or the concept and technical solution of the present application is directly applied to other occasions without improvement, it is within the protection scope of the present application.​

Claims

1. A radar imaging optimization method for detecting defects in a pot insulator, characterized in that, The method comprises the following steps: Step 1, data acquisition and delay parameter determination: first, the UWB-MIMO radar antenna array is used to collect N-channel original echo signals. Then, the preset three-dimensional imaging space is divided into a series of two-dimensional imaging slices along a specific coordinate axis, and a delay calculation model considering the two-stage propagation path of electromagnetic waves in air and the insulator medium is used to calculate the multi-channel propagation delay parameters of all pixel points based on the spatial geometric relationship between each antenna channel and each pixel point on the imaging slice ; Step 2, the multi-channel propagation delay calculated in step 1 , the N-channel raw echo signals collected are processed using a delay-and-sum algorithm to generate a preliminary image for each two-dimensional imaging slice, and all slice images are combined to form a preliminary three-dimensional imaging result ; Step 3, propagation attenuation correction: on the preliminary imaging result Time gain compensation is performed to solve the problem of signal attenuation with the thickness of the pot-type insulator; the gain factor of the compensation Adaptively determined according to the radar center frequency, the relative dielectric constant of the medium, and the signal propagation distance; by multiplying the gain factor with the preliminary image, an intermediate image with balanced intensity is generated ; Step 4, Adaptive enhancement of image sharpness: [The text abruptly ends here, likely due to an incomplete sentence or a formatting error.] A power spectrum coherence factor weighting is applied, with the weighting intensity adaptively adjusted to optimize the spectral characteristics of common defects in basin insulators as an extended target. This aims to improve image contrast and resolution while suppressing sidelobe artifacts, resulting in the final optimized image. ; Step 5, imaging result generation and display: normalizing and visualizing the final optimized image.

2. The method of claim 1, wherein, In the step 1, a delay calculation model of two-stage propagation path is adopted, wherein the total propagation delay of the first channel is added by the propagation delay in the air and the propagation delay in the post insulator, as shown in the formula (1). ​​​​ wherein the air propagation delay calculated from equation (2) wherein the propagation delay in the basin insulator calculated from equation (3) in, An imaging point with an intensity to be calculated within a preset three-dimensional imaging space. The three-dimensional spatial coordinate vector, and They are the components of the first The transmitting and receiving antennas of each transceiver channel are three-dimensional spatial coordinate vectors in the same coordinate system. It is the speed at which electromagnetic waves propagate in the air medium. It is the propagation speed of electromagnetic waves inside the dielectric material of the basin-type insulator, which is determined by the propagation speed in the air. and the relative permittivity of the dielectric material Confirmed, the calculation formula is as follows: , and They are the first Channel connection to transmitting antenna With imaging point and connecting the receiving antenna With imaging point The coordinate vectors of the intersection points of the two actual propagation paths with the surface of the basin insulator.

3. The method of claim 2, wherein, In step 2, the preliminary three-dimensional imaging result As shown in equation (4): is the first raw echo data of the channel, is the first propagation delay of the channel, and N is the total number of channels. This process is a delay-and-sum beamforming.

4. The method of claim 3, wherein, In step 3, the intermediate image time gain compensation, as shown in equation (5): wherein is a preliminary three-dimensional imaging result obtained in step 2, is a gain factor based on material absorption attenuation compensation, which is expressed as formula (6): in, The equivalent round-trip propagation distance, used for calculating gain compensation, is defined as the distance the electromagnetic wave travels from the geometric center of the antenna array to the imaging point. Then from the imaging point The total path length back to the geometric center of the antenna array. The minimum round-trip propagation distance for the imaging area. The attenuation coefficient is related to the center frequency of the radar signal and the electromagnetic properties of the basin-type insulator material. The expression is shown in formula (7): wherein is the center frequency of the radar signal, is the relative permittivity of the dielectric material, is the propagation velocity of the electromagnetic wave in air, is the loss tangent of the dielectric.

5. The method of claim 4, wherein, In step 4, the power spectrum coherence factor The expression of the power spectrum coherence factor is shown in equation (8): where the power spectrum is obtained by performing a discrete Fourier transform on the complex-valued signal vector of the N coherent signals after time-delay focusing in all N channels and calculating the square of the amplitudes; is the frequency index of the spectrum; is the window half-width defining the low-frequency energy range; is the calculation length of the discrete Fourier transform; is the suppression exponent.

6. The method of claim 5, wherein, In step 4, the final optimized image is obtained The expression is shown as equation (9): wherein is the intensity balanced intermediate image obtained in step 3, is the power spectral coherence factor calculated in step 4.

7. The method of claim 6, wherein, In step 5, the image is normalized to adjust the overall dynamic range of the final optimized image, and the expression is shown in formula (10): wherein, is the final image normalized, is the maximum intensity value of the optimized image, is the maximum intensity value of the preliminary image.

Images