A multi-channel electromagnetic induction detection method and system based on uniform magnetic field confinement

Abstract

This invention discloses a multi-channel electromagnetic induction detection method and system based on uniform magnetic field constraint. A uniform magnetic field is constructed within the detection area of ​​the structure under test. Electromagnetic induction response signals are acquired through multiple channels under the constraint of the uniform magnetic field, and the spatial coordinates of each acquisition channel are marked. The raw electromagnetic induction signals acquired by each channel are preprocessed, including noise reduction filtering, envelope extraction, and amplitude normalization. A deep learning model is then used to extract structural features from the preprocessed multi-channel electromagnetic response signals. Based on the spatial position information of each acquisition channel, magnetic field uniformity parameters, and response characteristics, a channel correlation weight matrix is ​​constructed to describe the spatial consistency and physical coupling relationships between channels. A multi-channel joint optimization model is constructed, and the equivalent electromagnetic parameter distribution inside the structure under test is reconstructed through iterative solution. This invention improves the reliability, repeatability, and engineering applicability of electromagnetic induction detection results.

Description

Technical Field

[0001] This invention relates to the field of nondestructive testing technology, and in particular to a multi-channel electromagnetic induction detection method and system based on uniform magnetic field constraint. Background Technology

[0002] Electromagnetic induction testing technology, due to its non-contact nature, strong penetration capability, and sensitivity to conductive materials, is widely used in the non-destructive testing of reinforced concrete, steel fiber reinforced concrete, and other metal-reinforced materials. Existing electromagnetic induction testing systems typically generate an alternating magnetic field in the structure under test using an excitation coil, and then use the induction coil to collect the induced response caused by the conductive components within the structure, thereby retrieving information about the material's internal distribution. However, existing technologies still have many limitations in engineering applications.

[0003] First, most existing electromagnetic induction detection devices use a single coil or a simple coil combination to generate the excitation magnetic field. The magnetic field distribution typically exhibits significant spatial non-uniformity, with the magnetic induction intensity varying considerably with position. This magnetic field non-uniformity directly affects the induction response at different spatial locations due to differences in excitation intensity. Consequently, the acquired signal simultaneously contains both structural and magnetic field distribution information, increasing uncertainty in the inversion process and reducing the stability and repeatability of the detection results.

[0004] Secondly, to improve spatial resolution, existing technologies have attempted to introduce multi-channel or array-type electromagnetic induction acquisition structures. However, most solutions simply superimpose or analyze the multi-channel signals independently, lacking a systematic modeling of the spatial and physical coupling relationships between channels. Due to differences in the location, magnetic field conditions, and response sensitivity of different induction channels, significant inconsistencies often exist between multi-channel signals. If inversion is performed directly without constraints, artifacts or local misjudgments are easily introduced, making it difficult to obtain stable and reliable imaging results.

[0005] Furthermore, existing electromagnetic induction imaging systems generally lack the ability to assess and control the magnetic field state in real time. During the detection process, the distribution of the excitation magnetic field may shift or become distorted due to changes in the size and shape of the measured structure or the surrounding environmental factors. Traditional systems typically treat the magnetic field as a fixed background condition and cannot dynamically correct for the uniformity of the magnetic field, resulting in detection results that are highly sensitive to environmental and setup conditions.

[0006] On the other hand, when inspecting large components or complex structures, a single acquisition is often insufficient to cover the entire detection area, and technical bottlenecks remain in terms of spatial consistency and fusion imaging of data acquired from different locations or channels. Existing inversion algorithms mostly rely on empirical models or local fitting, lacking constraints on the spatial continuity and consistency between multi-channel sensing responses, which limits the stability, resolution, and anti-interference capabilities of the final imaging results.

[0007] In summary, existing electromagnetic induction detection systems still have significant shortcomings in terms of magnetic field uniformity control, multi-channel signal consistency constraints, and fusion imaging stability. There is an urgent need for an electromagnetic induction imaging system that can achieve collaborative acquisition of multi-channel electromagnetic induction responses under controlled uniform magnetic field conditions and perform joint fusion and reconstruction through spatial constraint mechanisms, so as to improve the reliability, stability, and engineering applicability of detection results. Summary of the Invention

[0008] The purpose of this invention is to provide a multi-channel electromagnetic induction detection method and system based on uniform magnetic field constraint, so as to solve the problems of non-uniform excitation magnetic field space, poor consistency of multi-channel data acquisition, insufficient stability of inversion results, and limited imaging accuracy in the detection of large-size or complex structures in existing electromagnetic induction non-destructive testing technology, thereby improving the reliability, repeatability and engineering applicability of electromagnetic induction detection results.

[0009] To achieve the above objectives, the present invention provides the following technical solution:

[0010] A multi-channel electromagnetic induction detection method based on uniform magnetic field confinement mainly includes the following steps:

[0011] S1. Construct a uniform magnetic field with a uniform magnetic induction intensity distribution and consistent direction within the detection area of ​​the structure to be tested.

[0012] S2. Under the constraint of a uniform magnetic field, electromagnetic induction response signals generated from multiple locations of the structure under test are collected to form multi-channel electromagnetic induction response signals, and the spatial coordinates of each acquisition channel are marked.

[0013] S3. Preprocess the raw electromagnetic induction signals acquired from each channel, including noise reduction filtering, envelope extraction, amplitude normalization, and use a deep learning model to extract structural features from the preprocessed multi-channel electromagnetic response signals.

[0014] S4. Based on the spatial location information, magnetic field uniformity parameters and response characteristics of each acquisition channel, construct a channel association weight matrix to describe the spatial consistency relationship and physical coupling relationship between channels;

[0015] S5. Construct a multi-channel joint optimization model and reconstruct the distribution of equivalent electromagnetic parameters inside the structure under test through iterative solution.

[0016] Preferably, in step S1, the uniform magnetic field is constructed through a uniform magnetic field constraint module, which includes multiple sets of excitation coils, a magnetic field modulation unit, and a magnetic field feedback correction unit. The multiple sets of excitation coils are arranged in a symmetrical array around the structure under test and are all electrically connected to the magnetic field modulation unit. The magnetic field modulation unit jointly regulates the excitation parameters of each excitation coil to form the expected magnetic field distribution. The magnetic field feedback correction unit is deployed in the detection area of ​​the structure under test to monitor the magnetic field distribution in the detection area in real time and feeds back the monitoring results to the magnetic field modulation unit. The magnetic field modulation unit automatically adjusts the excitation parameters according to the feedback results to achieve dynamic maintenance of the uniform magnetic field.

[0017] Preferably, in step S5, the objective function of the multi-channel joint optimization model is:

[0018] ;

[0019] in, For the first The measured response of the channel, The total number of channels. For the first The response mapping operator of the channel, The first one to be solved The equivalent electromagnetic parameters of the channel. The first one to be solved The equivalent electromagnetic parameters of the channel. For the first Channel and the Channel association weights between channels This represents the probability of structural features or prior structural information obtained through deep learning models. For the magnetic field gradient term, , , These represent the weights of each constraint term during the optimization process.

[0020] Furthermore, this invention proposes a detection system for implementing the above-described multi-channel electromagnetic induction detection method based on uniform magnetic field confinement, which mainly includes the following modules:

[0021] The uniform magnetic field confinement module is used to stably form a uniform magnetic field with a uniform magnetic induction intensity distribution and consistent direction within the detection area of ​​the structure under test.

[0022] The multi-channel acquisition module is deployed in the detection area of ​​the structure under test. It consists of several spatially known and independent sensing units, which are used to acquire electromagnetic induction response signals generated by the structure under test under uniform magnetic field excitation from multiple locations. The sensing units support individual acquisition or parallel acquisition.

[0023] The signal processing module, connected to the multi-channel acquisition module, includes a preprocessing unit and a feature extraction unit. The preprocessing unit is used to preprocess the raw electromagnetic induction signals acquired from each channel, including noise reduction filtering, envelope extraction, and amplitude normalization. The feature extraction unit is used to extract structural features from the preprocessed multi-channel electromagnetic response signals using a deep learning model.

[0024] The spatial constraint and channel association module, connected to the signal processing module, is used to construct and generate a channel association weight matrix that describes the spatial consistency relationship and physical coupling relationship between channels based on the spatial location information, magnetic field uniformity parameters and response characteristics of each acquisition channel.

[0025] The signal fusion and inversion reconstruction module is connected to the spatial constraint and channel association module. It is used to construct a multi-channel joint optimization model and reconstruct the distribution of equivalent electromagnetic parameters inside the structure under test through iterative solution.

[0026] The display terminal is connected to the signal fusion and inversion reconstruction module and is used to generate imaging results and related parameter information of the internal structure of the test structure based on the reconstruction data of the equivalent electromagnetic parameter distribution inside the structure.

[0027] Preferably, the sensing unit in the multi-channel acquisition module is an induction coil sensor or a magnetoresistive sensor or a combination thereof.

[0028] Compared with the prior art, the present invention has the following advantages:

[0029] 1) This invention utilizes the coordinated control of multiple excitation coils and a magnetic field feedback correction mechanism to create a uniform magnetic field environment with a uniform magnetic induction intensity distribution and consistent direction within the detection area of ​​the structure under test. This effectively reduces the systematic errors introduced by the spatial non-uniformity of the magnetic field in existing electromagnetic induction detection methods. Compared with traditional single-coil or uncontrolled magnetic field excitation methods, this invention can significantly reduce the induction response deviation caused by differences in excitation intensity at different spatial locations, thereby improving the stability, repeatability, and engineering reliability of the detection results.

[0030] 2) This invention achieves simultaneous acquisition of multi-spatial position sensing responses under uniform magnetic field confinement. Compared to single-channel or simple array detection methods, this invention has higher spatial coverage and information redundancy, effectively improving the detection capability for large-sized components and areas with complex structural parameter distributions. Simultaneously, multi-channel collaborative acquisition provides a sufficient data foundation for subsequent fusion imaging, enabling the system to have stronger noise resistance and resistance to local anomalies in complex structure detection.

[0031] 3) This invention incorporates the spatial location information, magnetic field state parameters, and induction response characteristics of each sensing channel into a unified fusion reconstruction framework, enabling joint modeling and constraint solving of multi-channel electromagnetic induction signals. This approach effectively suppresses inconsistencies caused by differences in position, sensitivity, or local interference between different channels, avoiding artifacts and misjudgments easily generated by traditional independent inversion or simple superposition methods, thereby significantly improving the spatial consistency and physical rationality of the fusion imaging results. Furthermore, this invention introduces a uniform magnetic field constraint term during the fusion inversion process, making magnetic field homogeneity one of the important constraints for inversion reconstruction, thus enhancing the robustness of the imaging results to magnetic field disturbances and environmental changes. Compared to traditional inversion methods that rely on empirical models or local fitting, this invention maintains high imaging stability and result reliability even under complex structures, heterogeneous materials, and changing detection conditions.

[0032] 4) This invention can realize highly automated electromagnetic induction detection and also supports parameter adjustment and extended configuration according to actual engineering needs. It has good versatility and scalability, and provides a high-performance and highly reliable technical solution for non-destructive testing of materials and structures containing conductive components. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of the multi-channel electromagnetic induction detection method based on uniform magnetic field constraint according to the present invention.

[0034] Figure 2 This is a schematic diagram of the detection system of the present invention;

[0035] Figure descriptions: 1. Uniform magnetic field constraint module; 2. Multi-channel acquisition module; 3. Signal processing module; 4. Spatial constraint and channel association module; 5. Signal fusion and inversion reconstruction module; 6. Display terminal; 7. Structure under test; 8. Uniform magnetic field; 9. Excitation coil. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0037] This embodiment provides a multi-channel electromagnetic induction detection system based on uniform magnetic field confinement, such as... Figure 1 As shown, the system includes a uniform magnetic field constraint module 1, a multi-channel acquisition module 2, a signal processing module 3, a spatial constraint and channel association module 4, a signal fusion and inversion reconstruction module 5, and a display terminal 6, as detailed below.

[0038] 1. Uniform magnetic field construction module

[0039] The uniform magnetic field constraint module generates a uniform magnetic field with a consistent magnetic induction intensity distribution within the detection area of ​​the structure under test. It consists of multiple sets of excitation coils, a magnetic field feedback correction unit, and a magnetic field modulation unit. The multiple sets of excitation coils are arranged in a preset symmetrical array around the detection area. The magnetic field modulation unit jointly controls the amplitude, phase, and on / off state of the excitation current of each coil to form the desired magnetic field distribution. The magnetic field feedback correction unit is located within the detection area of ​​the structure under test and monitors the magnetic field distribution in real time. When the magnetic field uniformity deviates from a preset threshold, the magnetic field modulation unit automatically adjusts the excitation parameters based on the feedback result to dynamically maintain the uniform magnetic field. This real-time monitoring-feedback-correction method avoids the adverse effects of magnetic field spatial non-uniformity on the detection results in traditional electromagnetic induction detection, providing a stable physical basis for consistent acquisition of multi-channel induction responses.

[0040] Under ideal conditions, a uniform magnetic field satisfies the following relationship: ,in Indicates magnetic flux density. This represents the spatial gradient of the magnetic field. When the magnetic field gradient within the detection region satisfies... At this time, the detection area can be considered to be in an approximately uniform magnetic field environment, where This is a preset threshold for magnetic field uniformity.

[0041] 2. Multi-channel acquisition module

[0042] The multi-channel acquisition module is deployed within a uniform magnetic field region to acquire electromagnetic induction response signals induced by conductive components within the structure under test under the influence of a uniform magnetic field. This module consists of several sensing units with known spatial locations and independently set up, such as induction coil sensors, magnetoresistive sensors, or combinations thereof. Each sensing unit can independently or in parallel collect the induced voltage, phase change, or equivalent electromagnetic response parameters of a local area.

[0043] By using a multi-channel collaborative acquisition method, multi-source electromagnetic induction information covering a large detection area can be obtained in a single detection process, effectively overcoming the problems of limited coverage and insufficient spatial resolution of traditional single-channel detection.

[0044] 3. Signal processing module

[0045] The signal processing module is connected to the multi-channel acquisition module and includes a preprocessing unit and a feature extraction unit. Since the acquired raw electromagnetic response signal typically contains environmental noise, electromagnetic interference, and systematic errors, it needs to be preprocessed by the preprocessing unit before subsequent analysis. The preprocessing process includes noise reduction filtering, envelope extraction, and signal normalization to obtain a stable and representative signal representation. After obtaining a stable preprocessed signal, the feature extraction unit further utilizes a deep learning model to extract structural features from the multi-channel electromagnetic response signal to identify potential cracks, voids, or rebar distribution characteristics within the structure under test.

[0046] 4. Spatial Constraints and Channel Association Module

[0047] This module connects to the signal processing module and is used to construct and generate a channel association weight matrix based on the spatial location information, magnetic field uniformity parameters, and response characteristics of each acquisition channel. This matrix describes the spatial consistency and physical coupling relationships between the channels.

[0048] ;

[0049] in, Indicates channel With channel Spatial correlation weight, Indicates channel With channel Spatial distance between them This is the distance attenuation parameter.

[0050] This process provides a structured physical background for subsequent joint inversion, avoiding local misjudgments or artifacts that are easily generated by traditional independent inversion methods.

[0051] 4. Signal Fusion and Inversion Reconstruction Module

[0052] This module, connected to the spatial constraint and channel association module, is used to construct a multi-channel joint optimization model and reconstruct the distribution of equivalent electromagnetic parameters inside the structure under test through iterative solution. The reconstruction process follows the objective function:

[0053] ;

[0054] in, For the first The measured response of the channel, The total number of channels. For the first The response mapping operator of the channel, The first one to be solved The equivalent electromagnetic parameters of the channel. The first one to be solved The equivalent electromagnetic parameters of the channel. For the first Channel and the Channel association weights between channels This represents the probability of structural features or prior structural information obtained through deep learning models. For the magnetic field gradient term, , , These represent the weights of each constraint term in the optimization process. In the objective function, the first term is the data consistency term, used to ensure a good match between the electromagnetic parameter distribution obtained from the inversion and the multi-channel measured electromagnetic response signal; the second term is the spatial consistency constraint term, used to ensure the continuity of electromagnetic parameter changes between adjacent detection areas, thereby improving the spatial stability of the inversion results; the third term is the structural prior constraint term, which uses the probabilistic information of structural features identified by the deep learning model to provide prior guidance for the inversion results, thereby enhancing the system's ability to identify structural features such as cracks, voids, or rebar locations; the fourth term is the uniform magnetic field constraint term, used to limit the change of magnetic field gradient within the detection area, ensuring that the inversion calculation meets the uniform magnetic field condition, thereby improving the stability and accuracy of the imaging results.

[0055] 5. Display terminal

[0056] The display terminal is connected to the signal fusion and inversion reconstruction module, used to generate imaging results and related parameter information of the internal structure of the test structure based on the reconstruction data of the equivalent electromagnetic parameter distribution inside the structure. The display module synchronously outputs the reconstruction confidence parameter corresponding to each sensing channel, which is used to characterize the reliability and stability of the reconstruction results at each spatial location. The final imaging results support multi-view observation, slice analysis, and parameter comparison, enabling operators to intuitively determine the structural distribution state inside steel fiber reinforced concrete or metal composite materials. This system solves the technical problem of limited imaging accuracy in the detection of large-size complex structures through closed-loop logic of magnetic field confinement, multi-channel collaboration, and adaptive optimization.

[0057] like Figure 2As shown, the workflow of this invention is as follows: first, an excitation coil 9, a magnetic field feedback correction unit, and an induction unit are arranged in the detection area of ​​the structure to be tested, and then the modules are connected together. The uniform magnetic field constraint module 1 establishes a spatially stable and directional uniform magnetic field 8 environment within the detection area of ​​the structure under test 7. Under the constraint of the uniform magnetic field 8, a controlled eddy current field or equivalent electromagnetic response field is induced inside the structure under test 7. The multi-channel acquisition module 2 synchronously acquires electromagnetic response signals caused by changes in the conductivity distribution inside the structure, structural discontinuities, or defect features under unified timing control. At the same time, the signal processing module 3 preprocesses the acquired electromagnetic response signals and extracts structural features from the preprocessed multi-channel electromagnetic response signals using a deep learning model. The spatial constraint and channel association module 4 constructs a channel association weight matrix to describe the spatial consistency and physical coupling relationships between channels based on the spatial position information, magnetic field uniformity parameters, and response features of each acquisition channel. The signal fusion and inversion reconstruction module 5 constructs a multi-channel joint optimization model and reconstructs the equivalent electromagnetic parameter distribution inside the structure under test through iterative solution. Finally, the display terminal 6 generates the imaging results and related parameter information inside the structure under test based on the reconstructed data of the equivalent electromagnetic parameter distribution inside the structure.

[0058] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-channel electromagnetic induction detection method based on uniform magnetic field confinement, characterized in that, Includes the following steps: S1. Construct a uniform magnetic field with a uniform magnetic induction intensity distribution and consistent direction within the detection area of ​​the structure to be tested. S2. Under the constraint of a uniform magnetic field, electromagnetic induction response signals generated from multiple locations of the structure under test are collected to form multi-channel electromagnetic induction response signals, and the spatial coordinates of each acquisition channel are marked. S3. Preprocess the raw electromagnetic induction signals acquired from each channel, including noise reduction filtering, envelope extraction, amplitude normalization, and use a deep learning model to extract structural features from the preprocessed multi-channel electromagnetic response signals. S4. Based on the spatial location information, magnetic field uniformity parameters and response characteristics of each acquisition channel, construct a channel association weight matrix to describe the spatial consistency relationship and physical coupling relationship between channels; S5. Construct a multi-channel joint optimization model and reconstruct the distribution of equivalent electromagnetic parameters inside the structure under test through iterative solution.

2. The multi-channel electromagnetic induction detection method according to claim 1, characterized in that, In step S1, the uniform magnetic field is constructed through a uniform magnetic field constraint module, which includes multiple sets of excitation coils, a magnetic field modulation unit, and a magnetic field feedback correction unit. The multiple sets of excitation coils are arranged in a symmetrical array around the structure under test and are all electrically connected to the magnetic field modulation unit. The magnetic field modulation unit jointly regulates the excitation parameters of each excitation coil to form the expected magnetic field distribution. The magnetic field feedback correction unit is deployed in the detection area of ​​the structure under test to monitor the magnetic field distribution in the detection area in real time and feeds back the monitoring results to the magnetic field modulation unit. The magnetic field modulation unit automatically adjusts the excitation parameters according to the feedback results to achieve dynamic maintenance of the uniform magnetic field.

3. The multi-channel electromagnetic induction detection method according to claim 2, characterized in that, In step S5, the objective function of the multi-channel joint optimization model is: ； in, For the first The measured response of the channel, The total number of channels. For the first The response mapping operator of the channel, The first one to be solved The equivalent electromagnetic parameters of the channel. The first one to be solved The equivalent electromagnetic parameters of the channel. For the first Channel and the Channel association weights between channels This represents the probability of structural features or prior structural information obtained through deep learning models. For the magnetic field gradient term, , , These represent the weights of each constraint term during the optimization process.

4. A detection system implementing the method as described in claims 1-3, characterized in that, include: The uniform magnetic field confinement module is used to stably form a uniform magnetic field with a uniform magnetic induction intensity distribution and consistent direction within the detection area of ​​the structure under test. The multi-channel acquisition module is deployed in the detection area of ​​the structure under test. It consists of several spatially known and independent sensing units, which are used to acquire electromagnetic induction response signals generated by the structure under test under uniform magnetic field excitation from multiple locations. The sensing units support individual acquisition or parallel acquisition. The signal processing module, connected to the multi-channel acquisition module, includes a preprocessing unit and a feature extraction unit. The preprocessing unit is used to preprocess the original electromagnetic induction signals acquired by each channel, including noise reduction filtering, envelope extraction, and amplitude normalization. The feature extraction unit is used to extract structural features from the preprocessed multi-channel electromagnetic response signal using a deep learning model. The spatial constraint and channel association module, connected to the signal processing module, is used to construct and generate a channel association weight matrix that describes the spatial consistency relationship and physical coupling relationship between channels based on the spatial location information, magnetic field uniformity parameters and response characteristics of each acquisition channel. The signal fusion and inversion reconstruction module is connected to the spatial constraint and channel association module. It is used to construct a multi-channel joint optimization model and reconstruct the distribution of equivalent electromagnetic parameters inside the structure under test through iterative solution. The display terminal is connected to the signal fusion and inversion reconstruction module and is used to generate imaging results and related parameter information of the internal structure of the test structure based on the reconstruction data of the equivalent electromagnetic parameter distribution inside the structure.

5. The detection system according to claim 4, characterized in that, The sensing unit in the multi-channel acquisition module adopts an induction coil sensor, a magnetoresistive sensor, or a combination thereof.

Images