A method and system for evaluating the corrosion of steel reinforcement in gantry structures of a substation

Abstract

The present application relates to the field of nondestructive testing, and discloses a substation gantry steel corrosion detection method and evaluation system, the specific steps are: using a double field multiplexing RF shielding and magnetic flux focusing integrated probe to synchronously acquire X-ray backscattering image data and original complex impedance data of the steel; extracting the real three-dimensional physical topology boundary of the steel; substituting it into the Maxwell control equation as a forced geometric constraint condition, carrying out eddy current field space inversion on the complex impedance data, obtaining the real conductivity three-dimensional distribution matrix of the stripping surface deformation interference; extracting the local equivalent diameter deviation based on the topology boundary, extracting the real conductivity deviation based on the three-dimensional distribution matrix, to construct a two-dimensional feature matrix; according to the feature matrix, carrying out multi-dimensional fusion judgment of the corrosion state. The present application binds the macroscopic structure deformation and the microscopic material variation in situ, accurately distinguishes the surface false thickening, the substantial thinning and the physical deformation, and completely avoids the misjudgment and the missed judgment of the conventional detection.

Description

Technical Field

[0001] This invention relates to the field of non-destructive testing technology, specifically to a method and evaluation system for detecting corrosion of reinforcing steel bars in substation gantry frames. Background Technology

[0002] Substation gantry frames are fundamental support structures for power systems, and the structural integrity of their internal steel reinforcement affects the operational safety of the power grid. Being exposed to the outdoor environment for extended periods, the steel reinforcement is prone to corrosion and deterioration. Currently, non-destructive testing for steel reinforcement corrosion primarily employs conventional electromagnetic testing methods.

[0003] In practical engineering applications, existing detection technologies face various interferences. Changes in humidity and salinity of outdoor concrete layers can cause fluctuations in the spatial dielectric constant, resulting in signal drift and phase crosstalk. More significantly, the surface of corroded steel bars can develop irregular local expansion or pitting. Conventional electromagnetic testing typically treats the target object as a smooth columnar body, ignoring the edge distortion and scattering interference of electromagnetic signals caused by the aforementioned surface deformation. This leads to the impedance data obtained from measurements failing to accurately reflect the true material parameters inside the steel bars.

[0004] Furthermore, existing single-mode inspection equipment mostly uses a fixed excitation frequency, and the detection depth cannot be dynamically adjusted according to the actual corrosion thickness, easily leading to detection blind spots. Due to the lack of multi-dimensional physical data support, existing technologies struggle to effectively distinguish between the false volume increase caused by surface rust and the substantial thinning caused by internal material loss when facing complex working conditions. The final qualitative test results often rely on human experience, which carries the risk of misjudgment and missed detection.

[0005] More importantly, existing technologies are limited to single physical field detection or simple data splicing, lacking a mechanism to deeply correlate and decouple geometric morphology information from material degradation information. As a result, they cannot effectively distinguish between the three easily confused states of surface pseudo-coarsening, real physical deformation, and internal substantial corrosion at the underlying physical logic level. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method and evaluation system for detecting and evaluating corrosion of steel bars in substation gantry frames. This solves the problems of existing electromagnetic detection methods being affected by irregular deformation of the steel bar surface and interference from environmental dielectric fluctuations, which leads to distortion in the measurement of internal material parameters and makes it difficult to accurately distinguish between surface pseudo-coarsening and internal substantial corrosion.

[0007] To achieve the above objectives, the present invention provides the following technical solution: The first aspect of the present invention provides a method for detecting corrosion of reinforcing steel bars in a substation gantry frame, the method comprising the following steps: A dual-field multiplexed RF shielding and magnetic flux focusing integrated probe was used to perform a unilateral translational scan of the target steel bar, and X-ray backscattered image data and original complex impedance data of the target steel bar were acquired simultaneously. Extract the true three-dimensional physical topological boundary of the target steel reinforcement from the X-ray backscatter image data; Substituting the real three-dimensional physical topological boundary as a forced geometric constraint into Maxwell's control equations, the original complex impedance data is subjected to eddy current field spatial inversion to obtain the real three-dimensional conductivity distribution matrix stripped of surface deformation interference. Based on the real three-dimensional physical topological boundary and nominal outer diameter, local equivalent diameter deviation is extracted; based on the real three-dimensional conductivity distribution matrix and nominal conductivity, real conductivity deviation is extracted to construct a two-dimensional feature matrix composed of the local equivalent diameter deviation and the real conductivity deviation. A loop is used to perform multi-dimensional fusion judgment on the corrosion state of the target steel bar based on the two-dimensional feature matrix.

[0008] As an optional implementation, the dual-field multiplexing RF shielding and magnetic flux focusing integrated probe specifically includes: a probe substrate made of high-density X-ray shielding material; a collimation channel for X-ray backscattered photon emission and reception is opened inside the substrate; the outer contour of the substrate is processed into a magnetic yoke for physically focusing the eddy current alternating magnetic field, so that the X-ray detection main axis and the central axis of the eddy current alternating magnetic field are strictly coaxially aligned in situ in the three-dimensional coordinate system.

[0009] As an optional implementation, during the acquisition of the original complex impedance data, closed-loop dynamic control is performed on the excitation frequency of the eddy current alternating magnetic field: The spatial gray-level gradient in the X-ray backscattered image data is calculated, and the spatial gray-level gradient is converted into the effective detection depth of the eddy current electromagnetic field through a preset nonlinear mapping function. Based on the skin effect physical equation, the effective detection depth is used as an input variable to solve in reverse and dynamically adjust the excitation frequency of the eddy current alternating magnetic field in real time. :

[0010] In the formula, The permeability of free space, The relative permeability of the material being tested. The nominal electrical conductivity of the target steel reinforcement.

[0011] As an optional implementation, before inverting the original complex impedance data, hardware-level bias calibration is performed based on an environmental dielectric perturbation compensation model: during the scanning acquisition process, the intensity of the disordered scattering background signal directed towards the non-metallic region is simultaneously extracted, and energy dispersive spectral analysis is performed on it to generate an environmental dielectric constant perturbation term characterizing the water absorption rate and porosity of the target environment. The environmental dielectric constant perturbation term is multiplied by the system-calibrated cross-field coupling coefficient matrix and then input to the impedance plane zeroing circuit. The output is the calibrated original complex impedance data after eliminating interference from the non-metallic environmental layer. :

[0012] In the formula, This refers to the uncompensated raw complex impedance data acquired by the probe. This is the cross-field coupling coefficient matrix calibrated for the system.

[0013] As an optional implementation, the real three-dimensional physical topological boundary is substituted as a forced geometric constraint into Maxwell's governing equations to perform eddy current field spatial inversion on the original complex impedance data. The specific process includes: introducing a partial differential equation for the magnetic vector potential containing unknown internal spatial conductivity into the quasi-static electromagnetic field Maxwell's governing equations.

[0014] In the formula, The angular frequency of the eddy current excitation. The source current density of the probe. The imaginary unit, The spatial magnetic vector potential, The true conductivity distribution inside the target steel reinforcement is to be inverted and solved. At the extracted true three-dimensional physical topological boundary, in-situ continuity boundary conditions are applied, including the magnetic vector potential continuity condition at the interface between the air domain and the steel reinforcement domain, and the source current density continuity condition. The partial differential equation of the magnetic vector potential is solved using an iterative inversion algorithm with boundary correction, and the true three-dimensional conductivity distribution matrix is ​​output.

[0015] As an optional implementation, the two-dimensional feature matrix is ​​constructed under a unified spatial grid system. Specifically, the difference between the actual diameter and the nominal outer diameter of the target grid is calculated as the local equivalent diameter deviation; the difference between the actual conductivity derived from the target grid and the nominal conductivity is calculated as the actual conductivity deviation. In the three-dimensional distribution matrix of the actual conductivity, the actual conductivity deviation is further divided into the surface conductivity deviation of the shallow boundary grid and the core conductivity deviation of the core region grid.

[0016] As an optional implementation, a loop is used to perform multi-dimensional fusion determination of the corrosion state of the target steel reinforcement based on the two-dimensional feature matrix. Specific rules include: If the local equivalent diameter deviation shows a positive increase and is greater than the preset diameter expansion threshold, and the surface conductivity deviation is lower than the negative preset threshold and the core conductivity deviation remains within the tolerance range, then the target steel bar is determined to be in a false thickening state caused by the accumulation of surface oxide layer. If the local equivalent diameter deviation shows a negative decrease and is lower than the preset negative diameter reduction threshold, and the actual electrical conductivity deviation of the entire cross-section area is lower than the negative preset threshold, then the target steel bar is determined to be in a substantial thinning state due to deep material deterioration or pitting loss. If the absolute value of the local equivalent diameter deviation exceeds the preset tolerance range, but the corresponding actual conductivity deviation remains within the tolerance range, it is determined to be a non-corrosive physical deformation state caused by mechanical stress or environmental compression.

[0017] A second aspect of the present invention provides a system for assessing the corrosion of reinforcing steel bars in substation gantry frames, comprising: The collaborative acquisition module is used to simultaneously acquire X-ray backscattered image data and raw complex impedance data of the target steel bar through a dual-field multiplexed radio frequency shielded and magnetic flux focusing integrated probe. The topological boundary extraction module is used to extract the true three-dimensional physical topological boundary of the target steel bar from the X-ray backscatter image data. The eddy current field spatial decoupling module is used to substitute the real three-dimensional physical topological boundary as a forced geometric constraint into Maxwell's control equations, perform spatial inversion calculation on the original complex impedance data, and output the real three-dimensional conductivity distribution matrix. The feature matrix construction module is used to extract local equivalent diameter deviation based on the real three-dimensional physical topological boundary and nominal outer diameter, extract real conductivity deviation based on the real three-dimensional conductivity distribution matrix and nominal conductivity, and construct a two-dimensional feature matrix composed of the local equivalent diameter deviation and the real conductivity deviation. The comprehensive judgment and evaluation module is used to perform a multi-dimensional fusion judgment on the corrosion status of the target steel bar based on the two-dimensional feature matrix in a loop.

[0018] As an optional implementation, the system further includes a frequency domain adaptive slicing feedback module that is communicatively connected to the collaborative acquisition module; The module is configured to: calculate the spatial gray-level gradient of the X-ray backscattered image data, and convert the spatial gray-level gradient into the effective detection depth of the eddy current electromagnetic field through a preset nonlinear mapping function; solve the effective detection depth as an input variable in reverse according to the skin effect physical equation, and implement closed-loop dynamic adjustment of the excitation frequency of the eddy current alternating magnetic field in the collaborative acquisition module in real time.

[0019] As an optional implementation, the system also includes an environmental dielectric disturbance compensation module located at the front end of the collaborative acquisition module hardware; this module is configured to analyze the intensity of the disordered scattering background signal returned to the non-metallic region in real time, convert it into an environmental dielectric constant disturbance term, and input it as a bias parameter to the impedance plane zeroing circuit to filter out the signal drift caused by the external dielectric environment in real time.

[0020] This invention provides a method and evaluation system for detecting corrosion of reinforcing steel bars in substation gantry frames. It has the following beneficial effects: 1. At the physical implementation layer, this invention uses a dual-field multiplexing integrated probe to force coaxial bonding and utilizes the real-time X-ray sensing of the steel surface density to drive the adaptive adjustment of the eddy current excitation frequency, eliminating detection blind spots while ensuring the synchronization and fusion of multi-modal data from the hardware source. At the data decoupling layer, it creatively uses the real irregular physical boundary extracted by X-rays as a forced geometric constraint of Maxwell's control equations, so that the conductivity matrix calculated by the spatial inversion of the eddy current field is free from signal distortion caused by surface pit deformation. At the state discrimination layer, a two-dimensional feature matrix composed of "local equivalent diameter deviation" and "real conductivity deviation" is constructed, and multi-dimensional cross-criteria are executed based on this to technically and accurately distinguish three easily confused states: "mechanical extrusion physical deformation", "false thickening of surface rust layer" and "substantial thinning of internal pitting corrosion".

[0021] 2. This invention introduces an environmental dielectric disturbance compensation mechanism. By analyzing the disordered scattering background signal to extract the environmental dielectric disturbance term, zero-bias calibration is performed directly at the hardware acquisition layer, effectively filtering out signal crosstalk caused by changes in humidity and porosity of outdoor concrete in substations. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the modular architecture of the system of the present invention; Figure 2 This is a schematic diagram of the method flow of the present invention; Figure 3 This is a sub-flowchart of the cross-modal dynamic acquisition and hardware-level environmental compensation mechanism of the present invention; Figure 4 This is a diagram illustrating the spatial decoupling and inversion process of the eddy current field based on three-dimensional topological constraints according to the present invention. Figure 5This is a flowchart of the comprehensive corrosion state determination logic based on a two-dimensional feature matrix according to the present invention. Figure 6 This is a hardware structure block diagram of the electronic device of the present invention; Figure 7 This is a schematic diagram of the physical structure of the integrated probe of the present invention. Detailed Implementation

[0023] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see the appendix Figure 1 - Appendix Figure 7 This invention provides a method and evaluation system for detecting and assessing corrosion of steel reinforcement in substation gantry frames. The substation gantry frame steel reinforcement corrosion assessment system may include: a collaborative acquisition module, a topology boundary extraction module, an eddy current field spatial decoupling module, a feature matrix construction module, and a comprehensive judgment and evaluation module. These modules cooperate to construct a complete execution logic chain from acquiring front-end physical signals to qualitatively judging the structural state at the back end.

[0025] The collaborative acquisition module is equipped with a dual-field multiplexing RF shielding and magnetic flux focusing integrated probe at the hardware front end. The probe substrate is integrally molded from high-density X-ray shielding material. Inside the central axis region of the substrate, a collimating channel with a specific angle is machined to strictly constrain the emission path and receiving field of view of X-ray backscattered photons.

[0026] Simultaneously, the outer contour of the probe substrate is milled into a magnetically conductive yoke with a specific geometric configuration. During hardware operation, the high-density substrate itself acts as a shield against radiation and suppresses stray photons in space, while its peripheral magnetically conductive yoke structure directly serves as the physical energy-concentrating and conducting loop for the eddy current high-frequency alternating magnetic field. This shared substrate structure forcibly regulates the positional relationship between the X-ray detection main axis and the central axis of the eddy current alternating magnetic field in the three-dimensional coordinate system, achieving a rigid physical binding and spatial coaxial alignment of the detection axis.

[0027] The substation gantry steel reinforcement corrosion assessment system also includes a frequency domain adaptive slicing feedback module that communicates with the collaborative acquisition module. This module is equipped with an independent programmable gate array (PGA) processor responsible for executing closed-loop dynamic control of the alternating excitation source. This module calculates the spatial grayscale gradient of the X-ray backscattered image data. And through a preset nonlinear mapping function The spatial grayscale gradient is converted into the effective detection depth of the eddy current electromagnetic field. The mathematical expression of this operational logic is:

[0028] In the formula, , , All of these are coordinate components in the system's three-dimensional spatial coordinate system.

[0029] Based on the skin effect physical equation, the microcontroller in the frequency domain adaptive slicing feedback module will determine the effective detection depth. The inverse solution is performed using the input variable, and the excitation frequency of the eddy current alternating magnetic field in the collaborative acquisition module is also considered. Real-time closed-loop dynamic adjustment is implemented. The real-time frequency driving equation is set as follows:

[0030] In the formula, The permeability of free space, The relative permeability of the target material is given. The nominal electrical conductivity of the target steel reinforcement is the design specification.

[0031] In addition, the system also includes an environmental dielectric disturbance compensation module located at the front end of the collaborative acquisition module hardware. This module is mainly designed to address the abrupt changes in spatial dielectric constant caused by fluctuations in the humidity and salinity of the concrete layer in the outdoor environment of the substation. This compensation module is equipped with a high-precision analog-to-digital conversion acquisition circuit, which separates and extracts the intensity of the disordered scattered background signal returning from the non-metallic region in real time. .

[0032] The environmental dielectric disturbance compensation module compensates for the disordered scattering background signal intensity. Perform energy dispersive spectroscopy analysis to generate environmental dielectric constant perturbation terms characterizing the water absorption rate and porosity of the target environment. :

[0033] Obtain the environmental dielectric constant perturbation term Then, the module multiplies the environmental dielectric constant perturbation term by the cross-field coupling coefficient matrix calibrated by the system. Subsequently, it is input as a bias parameter to the system's impedance plane zeroing circuit to filter out signal drift caused by the external dielectric environment in real time. This circuit returns the original complex impedance from the probe's receiving coil. Hardware-level reference shifting is performed to output calibrated raw complex impedance data that excludes interference from the non-metallic environmental layer. :

[0034] In the formula, The original complex impedance signal contains both amplitude and phase dimensions. Through this hardware biasing operation logic, the system filters out crosstalk to the eddy current signal phase caused by changes in the dielectric parameters of the external non-metallic environment at the low-level signal acquisition stage.

[0035] In this invention, in response to the problem in the prior art that irregular deformation of the steel bar surface causes edge distortion of electromagnetic signals, which in turn leads to distortion of internal material measurements, a method for detecting corrosion of steel bars in substation gantry frames is provided, which realizes the underlying fusion calculation of heterogeneous physical field data.

[0036] The method for detecting corrosion of steel reinforcement in the substation gantry frame may include the following steps: A dual-field multiplexed RF shielding and magnetic flux focusing integrated probe is used to perform a unilateral translational scan of the target rebar, simultaneously acquiring X-ray backscattered image data and raw complex impedance data of the target rebar. In this embodiment, the system uses the same probe substrate as the execution carrier, ensuring that the X-ray optical sensing circuit and the low-frequency electromagnetic detection circuit are rigidly triggered in a unified mechanical coordinate system, thereby guaranteeing that the acquired multimodal data have a strict in-situ correspondence in the time and spatial dimensions.

[0037] The true three-dimensional physical topological boundary of the target rebar is extracted from the X-ray backscattered image data. In this embodiment, the backscattered photon flow characteristics received by the collimated channel inside the radio frequency shielding substrate are analyzed to reconstruct the irregular undulations such as local expansion and pits on the surface of the rebar, generating a point cloud boundary set that reflects the objective physical shape of the inspected object.

[0038] Substituting the real three-dimensional physical topological boundary as a forced geometric constraint into Maxwell's governing equations, eddy current field spatial inversion is performed on the original complex impedance data to obtain the real three-dimensional conductivity distribution matrix stripped of surface deformation interference. In this embodiment, the bottom-level spatial decoupling process employs a quasi-static electromagnetic field partial differential equation containing unknown variables, and its governing equation is set as follows:

[0039] In the formula, The spatial magnetic vector potential, The permeability of free space, The relative permeability of the material being tested. Let be the excitation angular frequency of the eddy current alternating magnetic field. To determine the true conductivity distribution of the internal space to be inverted, The source current density applied to the probe coil, The imaginary unit is used. The system incorporates the previously extracted objective optical topological boundary into the computational domain of the equation for truncation and constraint, eliminating scattering perturbations caused by geometric distortion, and thus obtaining the true three-dimensional conductivity distribution matrix that reflects the pure electrochemical properties.

[0040] Based on the real three-dimensional physical topological boundary and the nominal outer diameter, local equivalent diameter deviation is extracted; based on the real three-dimensional conductivity distribution matrix and the nominal conductivity, real conductivity deviation is extracted, so as to construct a two-dimensional feature matrix composed of the local equivalent diameter deviation and the real conductivity deviation. In this embodiment, the system encapsulates macroscopic physical deformation data and microscopic electromagnetic parameters within a unified three-dimensional spatial grid. The mathematical construction of this two-dimensional feature matrix is ​​defined as follows:

[0041] In the formula, This is the difference between the actual equivalent outer diameter of the current target mesh and the nominal outer diameter of the reinforcement in the prior design, i.e., the local equivalent diameter deviation; The difference between the conductivity output by the current target grid inversion and the nominal conductivity of the prior design is the actual conductivity deviation.

[0042] Finally, a loop is used to perform multi-dimensional fusion judgment on the corrosion state of the target steel bar based on the two-dimensional feature matrix.

[0043] As can be seen from the above steps, this invention constructs a three-level linkage closed loop from physical coaxial detection, boundary constraint inversion to feature fusion judgment: First, the integrated probe achieves strict coaxial synchronous acquisition of multimodal data at the physical layer, and uses the X-ray spatial gray-scale gradient to adaptively adjust the eddy current excitation frequency, ensuring the effectiveness of data fusion from the source; Second, the real three-dimensional physical topological boundary extracted by X-rays serves as the sole geometric constraint for eddy current field inversion, fundamentally separating the influence of geometric deformation and material variation; Finally, based on the multi-dimensional cross-criteria of the two-dimensional feature matrix, the aforementioned decoupled geometric deviation and conductivity deviation are jointly identified, achieving accurate characterization of corrosion state that is unattainable by existing technologies, and completely avoiding misjudgment and omission of surface pseudo-coarsening and internal substantial thinning by conventional detection.

[0044] In this embodiment, in order to address the interference of multi-dimensional media overlap in substations on the phase of electromagnetic signals, and the limitation of conventional fixed excitation frequencies in taking into account different corrosion depths, the present invention provides a cross-modal dynamic acquisition and hardware-level environmental compensation mechanism in the process of acquiring multi-modal data.

[0045] During the acquisition of raw complex impedance data, the system implements millisecond-level closed-loop dynamic control of the excitation frequency of the eddy current alternating magnetic field. The system analyzes the backscattered photon stream received by the X-ray detector spindle, generates a spatial grayscale matrix containing the target density, and calculates its spatial grayscale gradient. Utilizing the system's internally preset nonlinear mapping function This spatial grayscale gradient is directly converted into the effective detection depth of the eddy current electromagnetic field in the current spatial coordinates. The mathematical expression of this computational logic is as follows:

[0046] In the formula, , , The three-dimensional spatial coordinate system components established for the system. In specific implementation, the nonlinear mapping function... It is based on a pre-scanned standard stepped corroded steel bar specimen with a known step depth, extracting the mapping relationship between the surface X-ray backscatter spatial gray-scale gradient characteristics and its actual physical depth, which is pre-determined through polynomial curve fitting or offline training of neural networks, and then stored in the system memory.

[0047] Based on the physical law of the skin effect of electromagnetic waves in good conductors, the system microprocessor will determine the effective detection depth mentioned above. As a feedforward input parameter, the transient excitation frequency that drives the eddy current alternating magnetic field in real time is solved in reverse. Its governing equations are set as follows:

[0048] In the formula, The permeability of free space, The relative permeability of the target steel reinforcement is given. The nominal conductivity of the target steel reinforcement in the prior design is used. This closed-loop adjustment process ensures that the penetration depth of the low-frequency alternating magnetic field always dynamically matches the actual physical thickness of the surface corrosion and deposits, avoiding detection blind spots caused by over-penetration or too shallow a skin layer.

[0049] In this invention, before performing algorithmic spatial inversion on the original complex impedance data, hardware-level bias calibration is performed based on an environmental dielectric disturbance compensation model. During the dual-field integrated cooperative translational scanning process, the main optical path of the detection unit extracts the target's physical features, while the lateral optical path synchronously separates and extracts the intensity of the disordered scattered background signal that is directed towards the non-metallic concrete region and generates irregular reflections. .

[0050] By analyzing the intensity of the disordered scattering background signal Perform energy dispersive spectroscopy analysis to obtain characteristic parameters reflecting dielectric abrupt changes in the external environment. The system calls a preset conversion function. This characteristic parameter is mapped to generate an environmental dielectric constant perturbation term characterizing the changes in water absorption and porosity. :

[0051] In the specific implementation process, the conversion function The system constructs a database by offline scanning and data scatter fitting of concrete test blocks with known standard moisture content and porosity, and then stores the data in the system database.

[0052] Obtain the environmental dielectric constant perturbation term. Then, the system combines it with a pre-calibrated cross-field coupling coefficient matrix. Multiplying these parameters generates a bias compensation parameter. This parameter is converted into a voltage control signal in real time and injected into the impedance plane zeroing circuit at the hardware front end, adjusting the uncompensated original complex impedance obtained by the probe receiving coil. Perform a baseline translation:

[0053] In the formula, This is the calibrated raw complex impedance data after eliminating interference from the non-metallic environmental layer. Through this hardware-level cancellation operation, the system filters out electromagnetic phase crosstalk caused by environmental humidity fluctuations at the source, outputting pure raw complex impedance data, providing a high-confidence input source for subsequent underlying electromagnetic topology inversion.

[0054] In this embodiment, addressing the technical challenge of irregular corrosion on the surface of the target object causing edge distortion of the alternating magnetic field and resulting in inaccurate measurement of internal material parameters, this invention elaborates in detail the specific physical process of substituting three-dimensional topological constraints into Maxwell's control equations to perform spatial decoupling and inversion of the eddy current field.

[0055] In this invention, the system establishes a partial differential equation for the magnetic vector potential, which includes the unknown three-dimensional internal spatial conductivity, within the quasi-static electromagnetic field computational domain. The underlying governing equation is defined as follows:

[0056] In the formula, The spatial magnetic vector potential, The permeability of free space, The relative permeability of the target material is given. The angular frequency of the eddy current excitation. The imaginary unit, The source current density applied to the source coil of the integrated probe. The three-dimensional distribution of the true conductivity of the peeled surface after deformation disturbance is the solution to be found.

[0057] In conventional electromagnetic inverse problem solving, the target object is usually idealized as a smooth cylindrical geometry. This assumption can lead to significant truncation errors under severe corrosion conditions. This invention breaks with this convention by directly extracting the real three-dimensional physical topological boundary obtained from ray-mapping analysis in the previous stage and representing it as an irregular spatial closed surface. .

[0058] The system will close the irregular spatial surface The system is forced to use rigid geometric constraints for mesh generation in electromagnetic finite element calculations. At the aforementioned real three-dimensional physical topology boundaries, the system strictly applies in-situ continuity boundary conditions, which include the tangential continuity condition of the magnetic vector potential at the interface between the air domain and the reinforced concrete domain, as well as the normal continuity condition of the source current density.

[0059] The mathematical expression for the above in-situ continuity boundary conditions is as follows:

[0060]

[0061] In the formula, This refers to the extracted exterior normal vector at any point on the surface of the irregular topological boundary. and These are the magnetic vector potentials on the air domain side and the steel reinforcement domain side, respectively. and These are the current densities on the air domain side and the steel reinforcement domain side, respectively.

[0062] After establishing the geometric truncation boundary and governing equations, the system executes an iterative inversion algorithm with boundary correction. This algorithm uses the calibrated complex impedance, which has undergone hardware-level environmental compensation, as the target observation. Through forward calculation and residual differentiation, it dynamically updates the conductivity values ​​within the grid in each iteration.

[0063] As the residual of the objective function gradually approaches and converges, the system eventually decouples the true three-dimensional conductivity distribution matrix, stripped of external physical deformation factors. Through this deep integration of optical and electromagnetic mechanisms, the system fundamentally eliminates the scattering interference of surface structural abrupt changes such as pits and expansion on the low-frequency eddy current detection path.

[0064] In this embodiment, to address the engineering assessment challenge of easily confusing "false coarsening of surface rust" with "internal hidden material deterioration" under complex working conditions, this invention provides a qualitative state determination logic based on two-dimensional feature tensors, achieving a deep physical binding between macroscopic geometric morphology and microscopic electrochemical properties.

[0065] In this invention, the system establishes a unified spatial mesh generation system within the previously decoupled three-dimensional coordinate framework. Based on the extracted real three-dimensional physical topological boundary, the system calculates the difference between the actual equivalent diameter of the target mesh and the prior-defined nominal outer diameter for each mesh, defining this difference as the local equivalent diameter deviation. .

[0066] Simultaneously, based on the true three-dimensional conductivity distribution matrix obtained from the spatial inversion of the eddy current field, the system extracts the difference between the inverted conductivity corresponding to the target grid and the prior nominal conductivity, defining it as the true conductivity deviation. The system encapsulates the physical quantities of the above two dimensions into two-dimensional feature tensors. :

[0067] In the specific spatial distribution matrix analysis, the system determines the deviation of the true conductivity based on the topological hierarchy of the grid depth. Targeted decomposition is performed. The conductivity difference in the shallow mesh region located at the 3D physical boundary is extracted as the surface conductivity deviation; while the conductivity difference in the core mesh region inside the target object is extracted as the core conductivity deviation.

[0068] Before executing the final multi-dimensional fusion judgment loop, the system configures a series of benchmark discrimination parameters. The system presets a positively increasing diameter expansion threshold. and the negatively decreasing diameter reduction threshold This is used to measure the increase or decrease in structural volume; a negative surface conductivity threshold is also preset. and negative core conductivity threshold It is used to measure the degree of material degradation at different levels.

[0069] Based on the set thresholds, the system executes strict multi-dimensional conditional discrimination logic. First, if the local equivalent diameter deviation shows a positive increase and exceeds the preset diameter expansion threshold... Furthermore, the surface conductivity deviation is lower than the negative preset threshold. If the core conductivity deviation remains within the tolerance range (in this embodiment, the tolerance range is preferably set to fluctuate within ±2% to ±5% of the nominal core conductivity), the system directly determines that surface oxide layer accumulation has occurred in this area and outputs the result of the false thickening state determination.

[0070] Second, if the local equivalent diameter deviation shows a negative decrease and is below the preset negative diameter reduction threshold. Furthermore, the actual conductivity deviation of the entire cross-sectional area (including the surface and core) is lower than the negative preset threshold. If the target is found to have undergone deep material degradation or pitting corrosion, the system will output a substantial thinning and corrosion status.

[0071] Third, if the absolute value of the local equivalent diameter deviation exceeds the preset tolerance range, but the corresponding actual conductivity deviation remains within the tolerance range, the system determines that the area is a non-corrosive physical deformation state caused by mechanical tensile or compressive stress or environmental extrusion.

Claims

1. A method for detecting corrosion of reinforcing steel bars in a substation gantry frame, characterized in that, Includes the following steps: A dual-field multiplexing RF shielding and magnetic flux focusing integrated probe was used to perform a single-sided scan of the target steel bar, and X-ray backscattered image data and original complex impedance data of the target steel bar were acquired simultaneously. Extract the true three-dimensional physical topological boundary of the target steel reinforcement from the X-ray backscatter image data; The true three-dimensional physical topological boundary is substituted into Maxwell's governing equations as a forced geometric constraint, and the original complex impedance data is used for eddy current field spatial inversion to obtain the true three-dimensional conductivity distribution matrix stripped of surface deformation interference. The specific process includes: In Maxwell's governing equations for quasi-static electromagnetic fields, a partial differential equation for the magnetic vector potential containing unknown internal spatial conductivity is introduced. At the extracted real three-dimensional physical topology boundary, in-situ continuity boundary conditions are applied, including the magnetic vector potential continuity condition and the source current density continuity condition at the interface between the air domain and the steel reinforcement domain. The magnetic vector potential partial differential equation is solved using an iterative inversion algorithm with boundary correction. This algorithm uses the original complex impedance data as the target observation value, and dynamically updates the conductivity value in the grid in each iteration cycle through forward calculation and residual differentiation. Finally, it converges and outputs the true three-dimensional conductivity distribution matrix. Based on the real three-dimensional physical topological boundary and nominal outer diameter, local equivalent diameter deviation is extracted; based on the real three-dimensional conductivity distribution matrix and nominal conductivity, real conductivity deviation is extracted to construct a two-dimensional feature matrix composed of the local equivalent diameter deviation and the real conductivity deviation. A loop is used to perform multi-dimensional fusion judgment on the corrosion state of the target steel bar based on the two-dimensional feature matrix.

2. The method for detecting corrosion of reinforcing steel bars in substation gantry frames according to claim 1, characterized in that, The dual-field multiplexing RF shielding and magnetic flux focusing integrated probe specifically includes: The probe substrate is made of high-density X-ray shielding material; the substrate has a collimation channel inside for X-ray backscattered photon emission and reception, and the outer contour of the substrate is processed into a magnetic yoke for physical energy focusing of the eddy current alternating magnetic field, so that the X-ray detection main axis and the central axis of the eddy current alternating magnetic field are strictly coaxially aligned in the three-dimensional coordinate system.

3. The method for detecting corrosion of reinforcing steel bars in substation gantry frames according to claim 1, characterized in that, The process of acquiring the original complex impedance data also includes closed-loop dynamic control of the excitation frequency of the eddy current alternating magnetic field, specifically including: The spatial gray-level gradient in the X-ray backscattered image data is calculated, and the spatial gray-level gradient is converted into the effective detection depth of the eddy current electromagnetic field through a preset nonlinear mapping function. Based on the physical equation of the skin effect, the effective detection depth is used as the input variable to solve in reverse and dynamically adjust the excitation frequency of the eddy current alternating magnetic field in real time.

4. The method for detecting corrosion of reinforcing steel bars in substation gantry frames according to claim 1, characterized in that, Before inverting the original complex impedance data, a hardware-level bias calibration process based on the environmental dielectric disturbance compensation model is also included, specifically including: During the acquisition process, the intensity of the disordered scattering background signal directed toward the non-metallic region is extracted simultaneously. By performing energy spectrum dispersion analysis on the intensity of the disordered scattering background signal, an environmental dielectric constant perturbation term characterizing the water absorption rate and porosity of the target environment is generated. After multiplying the environmental dielectric constant disturbance term by the cross-field coupling coefficient matrix calibrated by the system, the result is input into the impedance plane zeroing circuit, which outputs the calibrated original complex impedance data after eliminating interference from the non-metallic environmental layer.

5. The method for detecting corrosion of reinforcing steel bars in substation gantry frames according to claim 1, characterized in that, The two-dimensional feature matrix is ​​defined under a unified spatial grid system, and its specific construction and decomposition methods are as follows: The difference between the actual equivalent diameter and the nominal outer diameter of the section containing the target mesh is calculated as the local equivalent diameter deviation; The difference between the actual conductivity and the nominal conductivity derived from the target mesh is calculated as the actual conductivity deviation. In the true conductivity three-dimensional distribution matrix, the true conductivity deviation is further divided into the surface conductivity deviation of the shallow boundary grid and the core conductivity deviation of the core region grid.

6. The method for detecting corrosion of reinforcing steel bars in substation gantry frames according to claim 5, characterized in that, The corrosion state of the target steel reinforcement is comprehensively evaluated and determined based on the two-dimensional feature matrix, and the specific rules include: If the local equivalent diameter deviation shows a positive increase and is greater than the preset diameter expansion threshold, and the surface conductivity deviation is lower than the negative preset threshold and the core conductivity deviation remains within the tolerance range, then the target steel bar is determined to be in a false thickening state caused by the accumulation of surface oxide layer. If the local equivalent diameter deviation shows a negative decrease and is lower than the preset negative diameter reduction threshold, and the actual electrical conductivity deviation of the entire cross-section area is lower than the negative preset threshold, then the target steel bar is determined to be in a substantial thinning state due to deep material deterioration or pitting loss. If the absolute value of the local equivalent diameter deviation exceeds the preset tolerance range, but the corresponding actual conductivity deviation remains within the tolerance range, then the target steel bar is determined to be in a non-corrosive physical deformation state caused by mechanical stress or environmental compression.

7. A system for assessing the corrosion of reinforcing steel bars in substation gantry frames, characterized in that, include: The collaborative acquisition module is used to perform the simultaneous acquisition of X-ray backscatter image data and original complex impedance data of the target steel bar using the dual-field multiplexing radio frequency shielding and magnetic flux focusing integrated probe as described in claim 1. The topology boundary extraction module is used to perform the extraction of the true three-dimensional physical topology boundary of the target steel bar from the X-ray backscattered image data as described in claim 1. The eddy current field spatial decoupling module is used to perform the spatial inversion calculation of the original complex impedance data as described in claim 1, by substituting the real three-dimensional physical topological boundary as a geometric constraint into Maxwell's control equations, and outputting the real three-dimensional conductivity distribution matrix. Specifically, the eddy current field spatial decoupling module is configured to perform the following process: In Maxwell's governing equations for quasi-static electromagnetic fields, a partial differential equation for the magnetic vector potential containing unknown internal spatial conductivity is introduced. At the extracted real three-dimensional physical topology boundary, in-situ continuity boundary conditions are applied, including the magnetic vector potential continuity condition and the source current density continuity condition at the interface between the air domain and the steel reinforcement domain. The magnetic vector potential partial differential equation is solved using an iterative inversion algorithm with boundary correction. This algorithm uses the original complex impedance data as the target observation value, and dynamically updates the conductivity value in the grid in each iteration cycle through forward calculation and residual differentiation. Finally, it converges and outputs the true three-dimensional conductivity distribution matrix. The feature matrix construction module is used to perform the extraction of local equivalent diameter deviation based on the real three-dimensional physical topological boundary and nominal outer diameter as described in claim 1, and the extraction of real conductivity deviation based on the real three-dimensional conductivity distribution matrix and nominal conductivity, and to construct a two-dimensional feature matrix composed of the local equivalent diameter deviation and the real conductivity deviation. The comprehensive judgment and evaluation module is used to perform the comprehensive evaluation and judgment of the corrosion state of the target steel bar based on the two-dimensional feature matrix as described in claim 1.

8. The substation gantry steel reinforcement corrosion assessment system according to claim 7, characterized in that, It also includes a frequency domain adaptive slicing feedback module that is communicatively connected to the collaborative acquisition module; The frequency domain adaptive slicing feedback module is configured to: calculate the spatial gray-level gradient of the X-ray backscattered image data, and convert the spatial gray-level gradient into the effective detection depth of the eddy current electromagnetic field through a preset nonlinear mapping function; solve the effective detection depth as an input variable in reverse according to the skin effect physical equation, and implement closed-loop dynamic adjustment of the excitation frequency of the eddy current alternating magnetic field in the collaborative acquisition module in real time.

9. The substation gantry steel reinforcement corrosion assessment system according to claim 7, characterized in that, It also includes an environmental dielectric disturbance compensation module located at the front end of the hardware of the collaborative acquisition module; The environmental dielectric disturbance compensation module is configured to: analyze the intensity of the disordered scattering background signal returned from the non-metallic region in real time, convert it into an environmental dielectric constant disturbance term, and input it as a bias parameter into the impedance plane zeroing circuit to filter out the signal drift caused by the external dielectric environment in real time.

Images