Multi-source heterogeneous information fusion detection device and method for fatigue crack of steel box girder

Abstract

The invention discloses a multi-source heterogeneous information fusion detection device and method for fatigue cracks of a steel box girder, and relates to the technical field of bridge engineering. Comprising a signal generator, a power amplification module, a uniform field magnetization module, an array magnetic sensor module, a data acquisition card, an infrared thermal imaging module, a visible light imaging module and a detection terminal computer. According to the method, electromagnetic two-dimensional images and infrared and visible light imaging results are fused, and compared with a traditional eddy current and visual method for measuring the fatigue cracks of the steel box girder, the method has higher reliability; the structure of the uniform field magnetization module is optimized according to the geometrical characteristics of the steel box girder, and the situation that a formed interval uniform heating field shields the field of view of the infrared and visible light imaging module is avoided while the local magnetic flux is improved; a multi-source heterogeneous information fusion detection method is established, and the method can be suitable for defect identification of the surface of the steel box girder under complex conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of bridge engineering, and in particular to a multi-source heterogeneous information fusion detection device and method for fatigue cracks of a steel box girder. Background Art

[0002] Steel box girders are favored by engineers due to their light weight, high torsional stiffness, and convenient manufacturing and construction, and are widely used in cable-stayed bridges and suspension bridges. However, with the continuous increase in operating time, the increasingly prominent steel box girder diseases will bring more challenges to bridge maintenance. Under the combined effects of environmental factors and external loads, fatigue cracks are prone to sudden structural damage. Under the action of external loads or oxidation filling, early fatigue cracks appear in a closed form, and the cracks are relatively hidden and difficult to detect. Therefore, mastering the fatigue crack detection method of steel box girders, especially the multi-source heterogeneous information fusion non-destructive detection method that can make up for the information uncertainty of a single sensor in the process of crack detection and identification, is an important prerequisite for box girder maintenance and ensuring the safe service of large-span bridges. It has guiding significance and reference value for the future application of bridge inspection engineering.

[0003] At present, fatigue crack detection of steel box girders mainly adopts eddy current or visual methods. Eddy current detection is affected by the probe lift-off effect, while the visual method has low detection efficiency. The detection effect of other surface defect non-destructive detection methods such as optical imaging methods is affected by the light intensity and the surface conditions of the material. The eddy current thermal imaging method can record the surface / near-surface crack temperature distribution of metal materials and can detect early closed cracks in fatigue, but there are problems such as non-uniform heating and the field of view of the thermal imager being blocked by the excitation coil.

[0004] Therefore, it is an urgent problem for those skilled in the art to propose a multi-source heterogeneous information fusion detection device and method for fatigue cracks of steel box girders to solve the difficulties existing in the prior art. Summary of the invention

[0005] In view of this, the present invention provides a multi-source heterogeneous information fusion detection device and method for fatigue cracks in steel box girders, which can simultaneously carry out electromagnetic, eddy current thermal imaging, and optical image quantitative detection and evaluation of fatigue cracks on the surface of steel box girders in a specific area in one detection, fully reflect the fatigue crack distribution of the scanning path, and realize efficient and non-destructive detection and evaluation of fatigue cracks in bridge steel box girders.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A multi-source heterogeneous information fusion detection device for fatigue cracks of steel box beams, comprising: a signal generator, a power amplification module, a uniform field magnetization module, an array magnetic sensor module, a data acquisition card, an infrared thermal imaging module, a visible light imaging module and a detection terminal computer; wherein,

[0008] A signal generator, used to set a high-frequency signal;

[0009] A power amplifier module is used to amplify the high-frequency signal output by the signal generator to form a high-frequency alternating current signal;

[0010] The uniform field magnetization module is placed on the surface of the steel box girder plate to be tested, and is used to pass the high-frequency alternating current signal through the sensor coil to generate an induced magnetic field, thereby forming a uniform heating field on the surface of the steel box girder plate to be tested;

[0011] The array magnetic sensor module is located in the interval uniform heating field formed by the uniform field magnetization module, and is used to characterize the distribution difference of electromagnetic signal information of surface defects of the steel box girder plate to be detected;

[0012] Data acquisition card, used to transmit the recorded defect electromagnetic signal information to the detection terminal computer;

[0013] The infrared thermal imaging module is arranged above the surface of the steel box girder plate to be inspected, and is used to record the surface temperature information of the steel box girder plate to be inspected and transmit it to the inspection terminal computer;

[0014] A visible light imaging module is arranged above the surface of the steel box girder plate to be inspected, and is used to record the optical information of the surface of the steel box girder plate to be inspected and transmit it to the inspection terminal computer;

[0015] The detection terminal computer is used to receive electromagnetic signal information, temperature information and optical information of the surface defects of the steel box girder plate to be detected, so as to realize accurate detection of fatigue cracks on the surface of the steel box girder plate to be detected.

[0016] In the above device, optionally, the array magnetic sensor module includes n coils containing ferrite cores, n≥4, and during the detection process, the uniform field magnetization module and the array magnetic sensor module are in non-contact with the surface of the steel box girder plate to be detected.

[0017] In the above device, optionally, the electromagnetic signal information and temperature information of the surface defects of the steel box girder plate to be detected have the same excitation generation source, and can be coupled at the physical end. The influence of the inspection speed on the heat diffusion is considered during the detection. At this time, the diffusion equation of the interval uniform heating field of the uniform field magnetization module is expressed as:

[0018]

[0019] Where ρ is the material density of the steel box girder, C p is the specific heat capacity of the steel box girder material, λ is the heat transfer coefficient, v is the inspection speed, T is the temperature field distribution function, σ is the electrical conductivity of the steel box girder, J e is the eddy current density and t is the heat diffusion time.

[0020] The above device can optionally analyze the flux variation law of the induced magnetic field in the uniform field magnetization module by constructing a magnetic circuit model, and the magnetic potential F is expressed as:

[0021] F=Φ·R=Φ1·R1+Φ2·R2

[0022] Φ1=Φ 11 +Φ 12

[0023] Φ2=Φ 21 +Φ 22

[0024] The magnetic resistance R of the magnetic circuit is related to the path length l through which the magnetic flux passes, the magnetic permeability μ of the material, and the effective lateral area A through which the magnetic flux flows in the path, that is:

[0025] R=l / μ·A

[0026] Therefore, the partial reluctance of the uniform field magnetization module is expressed as:

[0027] R1=R core1 +R Air1

[0028] R2=R core2 +R Air2 +R s

[0029] Since the magnetic permeability of the yoke is greater than the magnetic permeability of the air and the magnetic permeability of the specimen, the magnetic flux of the specimen is greater than the magnetic flux of other parts. When the lift-off of the uniform field magnetization module is not 0, the excitation effect on the surface of the steel box girder decays with the increase of the lift-off height.

[0030] Among them, Φ is the total magnetic flux generated by the coil with high-frequency current, Φ1 is the magnetic flux generated by the first part of the coil, which can characterize the magnetic flux distribution of the overall structure, Φ2 is the magnetic flux generated by the second part of the coil, and Φ 11 is the magnetic flux of the upper yoke structure, Φ 12 is the magnetic flux passing through the yoke structure to the object under test, R core1 is Φ 11 The reluctance of the yoke in the flux path, R Air1 is Φ 11 Air gap reluctance; R core2 is Φ 12 The reluctance of the yoke in the flux path, R Air2 is Φ 12 Air gap reluctance, R s is the magnetic resistance in the steel box beam specimen. Since the relative position of the yoke and the induction coil is fixed, R Air1 No change R Air2 It will change with the change of lift-off.

[0031] A method for detecting fatigue cracks of a steel box girder using multi-source heterogeneous information fusion, and a device for detecting fatigue cracks of a steel box girder using multi-source heterogeneous information fusion for performing any of the above, comprising the following steps:

[0032] Reconstruct electromagnetic 2D images, infrared 2D images and optical images based on line scanning;

[0033] Perform pixel matching operations on electromagnetic 2D images, infrared 2D images and optical images;

[0034] The segmentation window size is defined according to the proportion of the crack size in the electromagnetic two-dimensional image, the infrared two-dimensional image and the optical image. The electromagnetic two-dimensional image, the infrared two-dimensional image and the optical image are divided into a series of background information and crack information according to the window size using the dilution matrix decomposition method, and the redundant information is eliminated.

[0035] The likelihood image is obtained by using the decision-level fusion method to achieve accurate detection of fatigue cracks on the surface of the steel box girder to be detected.

[0036] The above method is optional, based on array magnetic sensor module detection, and obtains a magnetic two-dimensional image through Lagrangian interpolation method;

[0037] Acquire infrared two-dimensional images based on infrared thermal imaging module;

[0038] Acquire optical images based on visible light imaging module.

[0039] In the above method, optionally, the specific content of performing pixel matching operation on the electromagnetic two-dimensional image, the infrared two-dimensional image and the optical image is:

[0040] Use feature extraction algorithms to identify key points in the image;

[0041] Find the matching feature points in the two images through feature matching algorithm;

[0042] Transform one image into the size and spatial coordinate system of another image so that the two images are geometrically aligned. During the image transformation process, interpolation methods are used to calculate the pixel values ​​at the new position.

[0043] The above steps are repeated until the pixel matching operation of the electromagnetic two-dimensional image, the infrared two-dimensional image and the optical image is completed.

[0044] It can be seen from the above technical solutions that, compared with the prior art, the present invention provides a steel box girder fatigue crack multi-source heterogeneous information fusion detection device and method, which has the following beneficial effects:

[0045] (1) The present invention combines electromagnetic two-dimensional images with infrared and visible light imaging results, and has higher reliability than traditional eddy current and visual methods for measuring fatigue cracks in steel box girders;

[0046] (2) The uniform field magnetization module structure was optimized according to the geometric characteristics of the steel box girder, which not only improved the local magnetic flux but also avoided the formation of the interval uniform heating field to block the field of view of the infrared and visible light imaging modules;

[0047] (3) A multi-source heterogeneous information fusion detection method was established, which can be applied to the defect identification on the surface of steel box girders in complex situations. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0049] Figure 1 A structural block diagram of a multi-source heterogeneous information fusion detection device for fatigue cracks of a steel box girder provided by the present invention;

[0050] Figure 2 A schematic diagram of a magnetic circuit model of a uniform field magnetization module provided by the present invention;

[0051] Figure 3 A flow chart for reconstructing electromagnetic two-dimensional images, infrared two-dimensional images and optical images provided by the present invention;

[0052] Figure 4 A flow chart of a method for multi-source heterogeneous information fusion detection of fatigue cracks in a steel box girder provided by the present invention;

[0053] Figure 5 A schematic top view of the distribution of the induced eddy current field and temperature field during detection provided by the present invention;

[0054] Among them, 1-signal generator, 2-power amplification module, 3-uniform field magnetization module, 4-array magnetic sensor module, 5-data acquisition card, 6-infrared thermal imaging module, 7-visible light imaging module, 8-detection terminal computer, 9-steel box girder plate, 10-crack, 11-induced eddy current, 12-temperature distribution information. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] Reference Figure 1 As shown, the present invention discloses a steel box girder fatigue crack multi-source heterogeneous information fusion detection device, comprising: a signal generator 1, a power amplification module 2, a uniform field magnetization module 3, an array magnetic sensor module 4, a data acquisition card 5, an infrared thermal imaging module 6, a visible light imaging module 7 and a detection terminal computer 8; wherein,

[0057] A signal generator 1, used for setting a high-frequency signal;

[0058] The power amplifier module 2 is used to amplify the high-frequency signal output by the signal generator 1 to form a high-frequency alternating current signal;

[0059] The uniform field magnetization module 3 is placed on the surface of the steel box girder plate 9 to be tested, and is used to pass the high-frequency alternating current signal through the sensor coil to generate an induced magnetic field, thereby forming a uniform heating field on the surface of the steel box girder plate 9 to be tested;

[0060] The array magnetic sensor module 4 is located in the interval uniform heating field formed by the uniform field magnetization module 3, and is used to characterize the distribution difference of the electromagnetic signal information of the surface defects of the steel box girder plate 9 to be detected;

[0061] A data acquisition card 5 is used to transmit the recorded defect electromagnetic signal information to the detection terminal computer 8;

[0062] The infrared thermal imaging module 6 is arranged above the surface of the steel box girder plate 9 to be detected, and is used to record the surface temperature information of the steel box girder plate 9 to be detected and transmit it to the detection terminal computer 8;

[0063] A visible light imaging module 7 is disposed above the surface of the steel box girder plate 9 to be detected, and is used to record the optical information of the surface of the steel box girder plate 9 to be detected and transmit it to the detection terminal computer 8;

[0064] The detection terminal computer 8 is used to receive electromagnetic signal information, temperature information and optical information of the surface defects of the steel box girder plate 9 to be detected, so as to realize accurate detection of fatigue cracks on the surface of the steel box girder plate 9 to be detected.

[0065] Furthermore, the array magnetic sensor module 4 includes n coils containing ferrite cores, n≥4, and during the detection process, the uniform field magnetization module 3 and the array magnetic sensor module 4 are in non-contact with the surface of the steel box girder plate 9 to be detected.

[0066] Furthermore, the electromagnetic signal information and temperature information of the surface defects of the steel box girder plate 9 to be detected have the same excitation generation source and can be coupled at the physical end. The influence of the inspection speed on the heat diffusion is considered during the detection. At this time, the diffusion equation of the interval uniform heating field of the uniform field magnetization module 3 is expressed as:

[0067]

[0068] Where ρ is the material density of the steel box girder, C p is the specific heat capacity of the steel box girder material, λ is the heat transfer coefficient, v is the inspection speed, T is the temperature field distribution function, σ is the electrical conductivity of the steel box girder, J e is the eddy current density and t is the heat diffusion time.

[0069] Furthermore, by constructing a magnetic circuit model to analyze the flux variation law of the induced magnetic field in the uniform field magnetization module 3, the magnetic potential F is expressed as:

[0070] F=Φ·R=Φ1·R1+Φ2·R2

[0071] Φ1=Φ 11 +Φ 12

[0072] Φ2=Φ 21 +Φ 22

[0073] The magnetic resistance R of the magnetic circuit is related to the path length l through which the magnetic flux passes, the magnetic permeability μ of the material, and the effective lateral area A through which the magnetic flux flows in the path, that is:

[0074] R=l / μ·A

[0075] Therefore, the partial reluctance of the uniform field magnetization module 3 is expressed as:

[0076] R1=R core1 +R Air1

[0077] R2=R core2 +R Air2 +R s

[0078] Since the magnetic permeability of the magnetic yoke is greater than the magnetic permeability of the air and the magnetic permeability of the specimen, the magnetic flux of the specimen is greater than the magnetic flux of other parts. When the lift-off of the uniform field magnetization module 3 is not 0, the excitation effect on the surface of the steel box girder plate 9 decays as the lift-off height increases;

[0079] Among them, Φ is the total magnetic flux generated by the coil with high-frequency current, Φ1 is the magnetic flux generated by the first part of the coil, which can characterize the magnetic flux distribution of the overall structure, Φ2 is the magnetic flux generated by the second part of the coil, and Φ 11 is the magnetic flux of the upper yoke structure, Φ 12 is the magnetic flux passing through the yoke structure to the object under test, R core1 is Φ 11 The reluctance of the yoke in the flux path, R Air1 is Φ 11 Air gap reluctance; R core2 is Φ 12 The reluctance of the yoke in the flux path, R Air2 is Φ12 Air gap reluctance, R s is the magnetic resistance in the steel box beam specimen. Since the relative position of the yoke and the induction coil is fixed, R Air1 No change R Air2 It will change with the change of lift-off.

[0080] In a specific embodiment, the following are included

[0081] The signal generator 1 is connected to the power amplifier module 2 to provide a high-frequency alternating current driving signal for the uniform field magnetization module 3. A uniformly distributed eddy current field and temperature field are formed within the range of the uniform field magnetization module 3. During the detection process, the uniform field magnetization module 3 and the array magnetic sensor module 4 are lifted off the surface of the steel box girder plate 9 to be detected by 3 mm. When the uniform field magnetization module 3 scans the crack 10 at a fixed detection speed, the eddy current field and temperature field within the range of the uniform field magnetization module 3 are distorted. The array magnetic sensor module 4 is composed of four coil structures, which are distributed on one side of the yoke arm (a part of the uniform field magnetization module 3, not shown in the figure) to record the electromagnetic information of the defect, and then supplement the optical image and thermal image information. The shooting direction of the infrared thermal imaging module 6 and the visible light imaging module 7 is facing the surface of the steel box girder plate 9 surrounded by the uniform field magnetization module 3, and the output ends of the infrared thermal imaging module 6 and the visible light imaging module 7 are connected to the detection terminal computer 8.

[0082] The electromagnetic signal information and temperature information of the surface defects of the steel box girder plate 9 to be detected have the same excitation generation source and can be coupled at the physical end. The influence of the inspection speed on the heat diffusion is considered during the detection. At this time, the diffusion equation of the interval uniform heating field of the uniform field magnetization module 3 is expressed as:

[0083]

[0084] Where ρ is the material density of the steel box girder, C p is the specific heat capacity of the steel box girder material, λ is the heat transfer coefficient, v is the inspection speed, T is the temperature field distribution function, σ is the electrical conductivity of the steel box girder, J e is the eddy current density and t is the heat diffusion time.

[0085] like Figure 2 As shown in the figure, by constructing a magnetic circuit model to analyze the flux change law of the induced magnetic field in the uniform field magnetization module, the magnetic potential F is expressed as:

[0086] F=Φ·R=Φ1·R1+Φ2·R2

[0087] Φ1=Φ 11 +Φ 12

[0088] Φ2=Φ 21 +Φ 22

[0089] The magnetic resistance R of the magnetic circuit is related to the path length l through which the magnetic flux passes, the magnetic permeability μ of the material, and the effective lateral area A through which the magnetic flux flows in the path, that is:

[0090] R=l / μ·A

[0091] Therefore, the partial reluctance of the uniform field magnetization module is expressed as:

[0092] R1=R core1 +R Air1

[0093] R2=R core2 +R Air2 +R s

[0094] Since the magnetic permeability of the yoke is greater than the magnetic permeability of the air and the magnetic permeability of the specimen, the magnetic flux of the specimen is greater than the magnetic flux of other parts. When the lift-off of the uniform field magnetization module is not 0, the excitation effect on the surface of the steel box girder decays with the increase of the lift-off height.

[0095] Among them, Φ is the total magnetic flux generated by the coil with high-frequency current, Φ1 is the magnetic flux generated by the first part of the coil, which can characterize the magnetic flux distribution of the overall structure, Φ2 is the magnetic flux generated by the second part of the coil, and Φ 11 is the magnetic flux of the upper yoke structure, Φ 12 is the magnetic flux passing through the yoke structure to the object under test, R core1 is Φ 11 The reluctance of the yoke in the flux path, R Air1 is Φ 11 Air gap reluctance; R core2 is Φ 12 The reluctance of the yoke in the flux path, R Air2 is Φ 12 Air gap reluctance, R s is the magnetic resistance in the steel box beam specimen. Since the relative position of the yoke and the induction coil is fixed, R Air1 No change R Air2 It will change with the change of lift-off.

[0096] In another specific embodiment, the following are included

[0097] In order to improve the scanning efficiency, a rectangular scanning method is used to uniformly detect the steel box girder plate 9. The solid line is the detection route, and the dotted line is the return route. During the scanning process, the array magnetic sensor module 4 inputs the multi-channel electrical signal of the steel box girder plate 9 to the data acquisition card 5 in real time. The mutual inductance between the array probes will cause the test signal to be distorted. Therefore, a high magnetic permeability shielding cover is required around the single detector. Through computer software processing, a two-dimensional magnetic image is obtained. The processing method is shown in Figure 3 As shown in the figure. After interpolation processing, the resolution of the electromagnetic image is improved. The frequency domain characteristics of the electromagnetic signal are extracted by using the adaptive wavelet separation method, and the electromagnetic noise is separated by setting the threshold, which improves the imaging quality of the electromagnetic image.

[0098] The present invention is simulated by using a multi-physics finite element simulation analysis method. The simulation analysis results of the surface crack detection of the steel box girder plate 9 are as follows: Figure 5 As shown. There are uniformly distributed induced eddy currents 11 and temperature distribution information 12 within the interval of the uniform field magnetization module 3. The eddy current field distribution around the crack 10 on the steel box girder plate 9 is affected by the crack 10 and gathers at the two ends of the crack 10 to form a local high temperature distribution. The verification results show the feasibility of the present invention.

[0099] and Figure 1 Corresponding to the above device, the embodiment of the present invention also provides a method for detecting fatigue cracks of steel box beams by multi-source heterogeneous information fusion, the flow chart of which is as follows: Figure 4 As shown, the following steps are included:

[0100] Reconstruct electromagnetic 2D images, infrared 2D images and optical images based on line scanning;

[0101] Perform pixel matching operations on electromagnetic 2D images, infrared 2D images and optical images;

[0102] The segmentation window size is defined according to the proportion of the crack size in the electromagnetic two-dimensional image, the infrared two-dimensional image and the optical image. The electromagnetic two-dimensional image, the infrared two-dimensional image and the optical image are divided into a series of background information and crack information according to the window size using the dilution matrix decomposition method, and the redundant information is eliminated.

[0103] The likelihood image is obtained by using the decision-level fusion method to achieve accurate detection of fatigue cracks on the surface of the steel box girder to be detected.

[0104] Furthermore, based on the detection of the array magnetic sensor module, a magnetic two-dimensional image is obtained through the Lagrangian interpolation method;

[0105] Acquire infrared two-dimensional images based on infrared thermal imaging module;

[0106] Acquire optical images based on visible light imaging module.

[0107] Furthermore, the specific contents of the pixel matching operation for the electromagnetic two-dimensional image, the infrared two-dimensional image and the optical image are as follows:

[0108] Use feature extraction algorithms to identify key points in the image;

[0109] Find the matching feature points in the two images through feature matching algorithm;

[0110] Transform one image into the size and spatial coordinate system of another image so that the two images are geometrically aligned. During the image transformation process, interpolation methods are used to calculate the pixel values ​​at the new position.

[0111] The above steps are repeated until the pixel matching operation of the electromagnetic two-dimensional image, the infrared two-dimensional image and the optical image is completed.

[0112] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0113] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A steel box girder fatigue crack multi-source heterogeneous information fusion detection device, characterized in that: include: Signal generator, power amplifier module, uniform field magnetization module, array magnetic sensor module, data acquisition card, infrared thermal imaging module, visible light imaging module and detection terminal computer; among which, A signal generator, used to set a high-frequency signal; A power amplifier module is used to amplify the high-frequency signal output by the signal generator to form a high-frequency alternating current signal; The uniform field magnetization module is placed on the surface of the steel box girder plate to be tested, and is used to pass the high-frequency alternating current signal through the sensor coil to generate an induced magnetic field, thereby forming a uniform heating field on the surface of the steel box girder plate to be tested; The array magnetic sensor module is located in the interval uniform heating field formed by the uniform field magnetization module, and is used to characterize the distribution difference of electromagnetic signal information of surface defects of the steel box girder plate to be detected; Data acquisition card, used to transmit the recorded defect electromagnetic signal information to the detection terminal computer; The infrared thermal imaging module is arranged above the surface of the steel box girder plate to be inspected, and is used to record the surface temperature information of the steel box girder plate to be inspected and transmit it to the inspection terminal computer; A visible light imaging module is arranged above the surface of the steel box girder plate to be inspected, and is used to record the optical information of the surface of the steel box girder plate to be inspected and transmit it to the inspection terminal computer; The detection terminal computer is used to receive electromagnetic signal information, temperature information and optical information of the surface defects of the steel box girder plate to be detected, so as to realize accurate detection of fatigue cracks on the surface of the steel box girder plate to be detected.

2. The multi-source heterogeneous information fusion detection device for fatigue cracks of steel box beams according to claim 1 is characterized in that: The array magnetic sensor module includes n coils containing ferrite cores, n≥4. During the detection process, the uniform field magnetization module and the array magnetic sensor module are in non-contact with the surface of the steel box girder plate to be detected.

3. The steel box girder fatigue crack multi-source heterogeneous information fusion detection device according to claim 1 is characterized in that: The electromagnetic signal information and temperature information of the surface defects of the steel box girder plate to be detected have the same excitation source and can be coupled at the physical end. The influence of the inspection speed on the heat diffusion is considered during the detection. At this time, the diffusion equation of the interval uniform heating field of the uniform field magnetization module is expressed as: Where ρ is the material density of the steel box girder, C p is the specific heat capacity of the steel box girder material, λ is the heat transfer coefficient, v is the inspection speed, T is the temperature field distribution function, σ is the electrical conductivity of the steel box girder, J e is the eddy current density and t is the heat diffusion time.

4. The multi-source heterogeneous information fusion detection device for fatigue cracks of steel box beams according to claim 1 is characterized in that: By constructing a magnetic circuit model to analyze the flux change law of the induced magnetic field in the uniform field magnetization module, the magnetic potential F is expressed as: F=Φ·R=Φ1·R1+Φ2·R2 Φ1=Φ 11 +F 12 Φ2=Φ 21 +F 22 The magnetic resistance R of the magnetic circuit is related to the path length l through which the magnetic flux passes, the magnetic permeability μ of the material, and the effective lateral area A through which the magnetic flux flows in the path, that is: R=l / μ·A Therefore, the partial reluctance of the uniform field magnetization module is expressed as: R1=R core1 +R Air1 R2=R core2 +R Air2 +R s Since the magnetic permeability of the yoke is greater than the magnetic permeability of the air and the magnetic permeability of the specimen, the magnetic flux of the specimen is greater than the magnetic flux of other parts. When the lift-off of the uniform field magnetization module is not 0, the excitation effect on the surface of the steel box girder decays with the increase of the lift-off height. Among them, Φ is the total magnetic flux generated by the coil with high-frequency current, Φ1 is the magnetic flux generated by the first part of the coil, which can characterize the magnetic flux distribution of the overall structure, Φ2 is the magnetic flux generated by the second part of the coil, and Φ 11 is the magnetic flux of the upper yoke structure, Φ 12 is the magnetic flux passing through the yoke structure to the object under test, R core1 is Φ 11 The reluctance of the yoke in the flux path, R Air1 is Φ 11 Air gap reluctance; R core2 is Φ 12 The reluctance of the yoke in the flux path, R Air2 is Φ 12 Air gap reluctance, R s is the magnetic resistance in the steel box beam specimen. Since the relative position of the yoke and the induction coil is fixed, R Air1 No change R Air2 It will change with the change of lift-off.

5. A steel box girder fatigue crack multi-source heterogeneous information fusion detection method, characterized in that: The multi-source heterogeneous information fusion detection device for fatigue cracks of a steel box girder according to any one of claims 1 to 4 comprises the following steps: Reconstruct electromagnetic 2D images, infrared 2D images and optical images based on line scanning; Perform pixel matching operations on electromagnetic 2D images, infrared 2D images and optical images; The segmentation window size is defined according to the proportion of the crack size in the electromagnetic two-dimensional image, the infrared two-dimensional image and the optical image. The electromagnetic two-dimensional image, the infrared two-dimensional image and the optical image are divided into a series of background information and crack information according to the window size using the dilution matrix decomposition method, and the redundant information is eliminated. The likelihood image is obtained by using the decision-level fusion method to achieve accurate detection of fatigue cracks on the surface of the steel box girder to be detected.

6. A steel box girder fatigue crack multi-source heterogeneous information fusion detection method according to claim 5, characterized in that: Based on the detection of array magnetic sensor modules, a magnetic two-dimensional image is obtained through the Lagrange interpolation method; Acquire infrared two-dimensional images based on infrared thermal imaging module; Acquire optical images based on visible light imaging module.

7. The method for detecting fatigue cracks of steel box beams by multi-source heterogeneous information fusion according to claim 5, characterized in that: The specific contents of pixel matching operation for electromagnetic two-dimensional images, infrared two-dimensional images and optical images are as follows: Use feature extraction algorithms to identify key points in the image; Find the matching feature points in the two images through feature matching algorithm; Transform one image into the size and spatial coordinate system of another image so that the two images are geometrically aligned. During the image transformation process, interpolation methods are used to calculate the pixel values ​​at the new position. The above steps are repeated until the pixel matching operation of the electromagnetic two-dimensional image, the infrared two-dimensional image and the optical image is completed.

Images