Assembly type hollow slab beam bridge hinge joint internal defect adaptive detection method

Abstract

The present application relates to a kind of assembled hollow slab beam bridge hinge joint internal defect adaptive detection method, the method obtains structural features and historical damage data, assesses and divides detection priority area;Different regions are dynamically configured ultrasonic frequency, pulse energy and signal gain, realize and structure characteristic adaptive matching;According to frequency, surface flatness and adjacent component misalignment state collaborative determination array layout strategy;In high priority area, the parameter of discrete point is optimized and position-parameter mapping library is established, optimal parameter is automatically called when continuously scanning along hinge joint, internal section image is reconstructed by synthetic aperture focusing technique;Based on the interference model correction image distortion of material attenuation, surface state and boundary effect etc., accurately determine the type, location and size of defect.The present application realizes the dynamic coupling optimization of detection parameter and structure characteristic, significantly improves defect detection rate and quantitative accuracy, reduces operation complexity, provides scientific and reliable data support for bridge maintenance.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of bridge engineering nondestructive testing, and in particular to a self-adaptive detection method for internal defects of a hinge joint of an assembled hollow slab beam bridge. BACKGROUND

[0002] The assembled hollow slab beam bridge is the most widely used small and medium span bridge type in the highway bridge system in China, and its transverse connection mainly relies on the hinge joint structure to achieve the transverse distribution of load. The compactness of the internal concrete of the hinge joint directly determines the overall stress performance and durability of the bridge. However, due to factors such as incomplete compaction during construction, concrete shrinkage, long-term vehicle load and environmental erosion, hidden defects such as cavities, looseness and cracks are easily produced in the internal hinge joint, resulting in transverse connection failure, single slab stress, and even serious bridge structure safety accidents.

[0003] At present, the detection of internal defects of the hinge joint mainly relies on manual experience judgment, conventional ultrasonic single-point scanning or impact echo method. The existing technology generally has the following problems: the detection parameters (frequency, gain, energy) are fixedly configured, and cannot be adaptively adjusted according to the differences in hinge joint structure thickness, concrete strength, reinforcement density and other characteristics, resulting in missed detection of deep defects or insufficient resolution of shallow defects; there is a lack of scientific detection planning strategy, and a uniform point distribution method is often used, which lacks pertinence for serious disease areas and has low detection efficiency; the influence of site conditions such as concrete surface flatness, adjacent slab beam misalignment and structure boundary effect on the detection signal is not fully considered, the probe coupling quality is unstable, and the defect misjudgment rate is high in the boundary area; the quantitative analysis accuracy is restricted by the coupling of multiple factors such as material attenuation, surface state and boundary interference, and the defect position and size determination error is large; the detection result depends on the subjective experience of the operator, the data comparability is poor, and it is difficult to form a traceable digital evaluation system. The above defects seriously restrict the accuracy, efficiency and reliability of the detection of internal defects of the hinge joint, and cannot meet the technical needs of high-precision and automated detection for modern bridge intelligent maintenance management. SUMMARY

[0004] Therefore, the present application provides a self-adaptive detection method for internal defects of a hinge joint of an assembled hollow slab beam bridge to solve the foregoing problems in the prior art.

[0005] To achieve the above-mentioned purpose, the present application provides a self-adaptive detection method for internal defects of a hinge joint of an assembled hollow slab beam bridge, comprising:

[0006] Step S1, obtaining the feature information and historical damage record of the structure to be detected, and dividing different detection priority areas by using a state evaluation method to output the area division result;

[0007] Step S2: Based on the region division results, the partition parameter configuration method is used to set corresponding detection frequency and energy parameter combinations for different priority regions, and the signal gain is adjusted according to the structural geometric characteristics to obtain a dynamic parameter configuration scheme.

[0008] Step S3: Based on the frequency setting, structural surface flatness, and relative positional relationship of adjacent components in the parameter configuration scheme, a parameter-condition collaborative determination method is used to determine the spatial layout strategy.

[0009] Step S4: According to the spatial layout strategy, perform discrete point parameter optimization in the high-priority area and establish a position-parameter mapping relationship. When scanning continuously along the structure, call the corresponding mapping parameters and obtain the internal cross-sectional image through array signal synthesis processing.

[0010] Step S5: Based on the abnormal features in the internal cross-sectional image and the geometric relationship between the probe and the structural boundary, a multi-factor interference compensation method is used to correct image distortion in order to obtain the target defect determination result.

[0011] Furthermore, the process of step S2 includes:

[0012] Based on the region division results, differentiated ultrasonic frequencies are configured for different priority regions to obtain region-adaptive frequency configurations.

[0013] Based on the estimated detection depth of different priority areas, the pulse energy of the corresponding areas is dynamically adjusted to obtain depth-adaptive energy parameters.

[0014] Based on the structural geometry and material attenuation characteristics, the signal gain is adjusted differently for regions of different thicknesses, and the gain benchmark is corrected by combining material characteristic parameters to obtain the structural response gain model.

[0015] The region-adaptive frequency configuration, depth-adaptive energy parameters, and structural response gain model are coupled with multiple parameters to generate the dynamic parameter configuration scheme.

[0016] Furthermore, the process of differentially adjusting the signal gain for regions of different thicknesses based on structural geometric characteristics and material attenuation characteristics, and then modifying the gain benchmark by combining material characteristic parameters to obtain the structural response gain model includes:

[0017] Obtain the geometric parameters of the hinge structure under test, establish the thickness distribution function along the detection path, and form a structural geometric model;

[0018] The concrete strength grade and reinforcement density are obtained as material property parameters, and a material attenuation coefficient mapping relationship is established to form a material attenuation model.

[0019] Based on the aforementioned structural geometric model, the signal gain benchmark is increased proportionally to the thickness increasing region, and the signal gain benchmark is decreased proportionally to the thickness decreasing region. A saturation constraint is applied to the gain value of the thin-walled region to limit the upper limit of the gain. At the same time, the pulse reflection interval parameter is configured to adjust the echo acquisition timing, thereby generating a thickness adjustment rule set.

[0020] Based on the material attenuation model, gain compensation correction factors are applied to the high-strength grade region and the high reinforcement density region respectively to generate a set of material correction factors.

[0021] The thickness adjustment rule set and the material correction factor set are weighted and coupled to generate gain configuration values ​​for each detection location, thus constructing a structural response gain model.

[0022] Furthermore, the process of step S3 includes:

[0023] Based on the frequency settings in the parameter configuration scheme, a correspondence between frequency and deployment method is established to obtain frequency adaptation mapping;

[0024] A mapping relationship between flatness and layout method is established based on the flatness evaluation results of the structural surface to obtain the flatness adaptation mapping;

[0025] A mapping relationship between misalignment state and layout method is established based on the relative positional relationship of adjacent components to obtain misalignment state adaptation mapping;

[0026] The frequency adaptation mapping, the flatness adaptation mapping, and the misalignment state adaptation mapping are collaboratively determined to generate a spatial deployment strategy.

[0027] Furthermore, the process of establishing a mapping relationship between flatness and layout method based on the surface flatness evaluation results to obtain a flatness adaptation mapping includes:

[0028] Pre-treatment is performed on the bottom surface of the detection area to remove surface deposits and improve surface smoothness;

[0029] The flatness level is evaluated based on the pre-treated surface condition to form a quantitative grading result of flatness.

[0030] Establish a mapping rule between flatness level and layout method, assign high flatness level to horizontal layout to expand the detection coverage, and assign low flatness level to vertical layout to ensure probe coupling quality, thus forming the flatness adaptation mapping.

[0031] Furthermore, the process of step S4 includes:

[0032] In the high-priority area, representative measurement points are selected to perform discrete point test mode. Based on the characteristics of the echo signal of each measurement point, the gain, frequency, and pulse energy parameters are optimized and adjusted. The spatial coordinates of each measurement point and its corresponding optimal parameter combination are recorded, and a location-parameter mapping relationship library is established.

[0033] The scanning path direction and step size parameters are determined according to the spatial layout strategy. Continuous scanning tests are performed along the hinge structure. At each scanning position point, the corresponding optimal parameter configuration in the mapping relationship library is called.

[0034] The reflected echo signal was acquired by array transducer, and the multi-channel signal was delayed and superimposed by synthetic aperture focusing technology to reconstruct a two-dimensional cross-sectional image of the concrete component.

[0035] Multiple two-dimensional cross-sectional images acquired through continuous scanning are stitched together in spatial order to generate an overall internal structure image of the detection area.

[0036] Furthermore, the process of acquiring reflected echo signals through an array of transducers, and then using synthetic aperture focusing technology to perform time-delay superposition processing on the multi-channel signals to reconstruct a two-dimensional cross-sectional image of the concrete component includes:

[0037] Excitation and reception control are performed on the array transducer to acquire multi-channel reflected echo signals and form a multi-channel holographic dataset.

[0038] The propagation delay of each channel signal is calculated based on the wave velocity and sensor geometry, and a channel delay parameter set is established.

[0039] The signals of each channel in the multi-channel holographic dataset are precisely compensated according to the delay parameter to achieve time-domain alignment of signal components from the same target point, forming a time-domain aligned signal set.

[0040] The time-domain aligned signal set is coherently superimposed to enhance the target signal and suppress noise, thereby completing the reconstruction of the two-dimensional cross-sectional image inside the concrete.

[0041] Furthermore, the process of step S5 includes:

[0042] Anomaly features are extracted from the internal cross-sectional image to identify candidate defect regions, resulting in an anomaly feature set.

[0043] Obtain the probe position coordinates and structural boundary geometric parameters, establish a probe-boundary spatial relationship model, and calculate the distance parameters between the detection point and the boundary to obtain the geometric constraint parameter set;

[0044] Based on the material properties, surface condition and boundary effect of concrete, a multi-factor interference analysis model is constructed, and the abnormal feature set is modified by multiple factors to obtain the interference correction factor set.

[0045] The interference correction factor set is applied to the internal cross-sectional image to compensate and correct the positional deviation and size distortion of the candidate defect region to obtain a distortion-corrected image.

[0046] Based on the distortion-corrected image, pattern matching is performed using a preset defect feature library to determine the defect type, location, and geometric parameters to obtain the target defect determination result.

[0047] Furthermore, the construction of the multi-factor interference analysis model and the multi-factor correction of the abnormal feature set include:

[0048] Material properties, surface conditions and boundary conditions are obtained as multiphysics field influence parameters, and corresponding attenuation correction sub-models are established respectively.

[0049] The various attenuation correction sub-models are coupled to construct a multi-factor coupling analysis framework;

[0050] The comprehensive correction coefficients are calculated on the abnormal feature set using the multi-factor coupling analysis framework to obtain the interference correction factor set.

[0051] Furthermore, the process of compensating for and correcting the positional deviation and dimensional distortion of the candidate defect region to obtain a distortion-corrected image includes:

[0052] The position correction component, size correction amount, and intensity correction amount in the interference correction factor set are extracted, and geometric transformations are performed on the coordinate position, contour boundary, and reflection intensity of the candidate defect region in the internal cross-section image to obtain the position correction image, size correction image, and intensity correction image, respectively.

[0053] The position-corrected image, the scale-corrected image, and the intensity-corrected image are fused and edge-smoothed to generate the distortion-corrected image.

[0054] Compared with the prior art, the beneficial effect of the present invention is that by dynamically coupling and configuring the detection parameters with structural characteristics and environmental conditions, the present invention significantly improves the accuracy and reliability of detecting internal defects in hinge joints. To address the inherent differences in the response of defects of different depths to ultrasonic frequencies, adaptive matching optimizes near-surface resolution with high frequencies and enhances deep penetration with low frequencies, avoiding insufficient penetration or resolution loss due to fixed frequencies. By positively adjusting the gain in relation to structural thickness, energy loss due to natural attenuation of sound waves in concrete due to path length is compensated. Simultaneously, the gain benchmark is corrected based on material strength and reinforcement density to prevent artifacts caused by over-gain in thin-walled areas and missed defects in deep areas due to under-gain, ensuring the signal strength remains within the optimal recognition range. Surface flatness and misalignment directly determine the probe coupling effect; the mapping rule adaptively selects horizontal or vertical placement to maximize the contact area between the probe and concrete, fundamentally improving signal acquisition quality. Multi-channel synthetic aperture focusing, through precise calculation and weighted superposition of sound wave propagation delays, significantly improves the signal-to-noise ratio by utilizing the coherent enhancement characteristics of signals from different paths, enabling clear imaging of deep, weak defects. A multi-factor interference compensation model, based on the objective influence of material attenuation, surface condition, and boundary effects on sound waves, systematically corrects the distortion of defect location, size, and strength, eliminating misjudgments and missed detections caused by differences in detection conditions. This method achieves closed-loop feedback of dynamic optimization of detection parameters with structural characteristics, transforming the traditional experience-based manual debugging into automatic configuration based on physical propagation laws. While improving the defect detection rate and quantitative accuracy, it significantly shortens the on-site debugging time, reduces the dependence on the skill level of operators, and the generated digital detection results are comparable across working conditions, providing scientific, reliable, and traceable data support for bridge maintenance decisions. Attached Figure Description

[0055] Figure 1 A flowchart illustrating an adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges provided by this invention.

[0056] Figure 2 This is a schematic diagram of the 25 to 40 kHz imaging in an adaptive detection method for internal defects in the hinge joints of a prefabricated hollow slab beam bridge provided by the present invention. Detailed Implementation

[0057] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0058] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0059] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0060] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0061] Please see Figure 1 and Figure 2 As shown, this invention provides an adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges, comprising:

[0062] Step S1: Obtain the feature information and historical damage records of the structure to be tested, and use the state assessment method to divide the regions with different detection priority to output the region division results;

[0063] Specifically, the structural parameters of the hinge joint (span, concrete strength, reinforcement details) were obtained by consulting bridge design drawings. On-site measurements of the actual materials and geometric conditions were conducted using a rebound hammer and a rebar cover tester. Damage records of the hinge joint (water seepage, cracking, exposed rebar location and extent) were extracted from past inspection reports. Visual inspection and photographic records were supplemented on-site to form a structural characteristic and damage dataset. A hinge joint condition assessment index system was constructed, comprising three levels: Level 1 indicators: structural importance coefficient (weight 0.3), severity coefficient (weight 0.4), and damage evolution rate coefficient (weight 0.3); Level 2 indicators: the structural importance coefficient includes span length, traffic load level, and hinge joint construction type; the severity coefficient includes water seepage range, crack width, and concrete spalling area; and the damage evolution rate coefficient includes damage discovery time interval and damage propagation rate. The weights of each indicator were determined using the analytic hierarchy process (AHP), and secondary indicators were quantitatively scored (0-100 points) using an expert scoring method. A weighted calculation was then performed to obtain the comprehensive status assessment value for each hinge section. Based on the assessment value, the inspection area was divided into high-priority areas (assessment value ≥ 70 points), medium-priority areas (assessment value 40-69 points), and low-priority areas (assessment value < 40 points). The output was a region division map containing the spatial range, priority level, and assessment basis for each region. Based on the region division results, surveying instruments were used to lay out each priority area, and inspection grid lines were drawn along the longitudinal direction of the hinge at equal or variable intervals. The grid size was determined according to priority: the longitudinal and transverse spacing of the grid was set to 0.5-1.0 meters for high-priority areas, 1.0-1.5 meters for medium-priority areas, and 1.5-2.0 meters for low-priority areas. Measurement point numbers were marked at the grid nodes, forming a traceable measurement point location coding system. An inspection work file containing measurement point coordinates, priority attributes, and number information was generated.

[0064] Step S2: Based on the region division results, the partition parameter configuration method is used to set corresponding detection frequency and energy parameter combinations for different priority regions, and the signal gain is adjusted according to the structural geometric characteristics to obtain a dynamic parameter configuration scheme.

[0065] Specifically, step S2 includes the following process:

[0066] Based on the region division results, differentiated ultrasonic frequencies are configured for different priority regions to obtain region-adaptive frequency configurations.

[0067] Specifically, the ultrasonic frequencies are configured according to the following rules based on the regional division results:

[0068] High-priority area: For near-surface fine defect detection, the frequency is set to 45-55KHz; if the detection depth is less than 10cm, the frequency is gradually increased to 60-80KHz to enhance the resolution of small defects.

[0069] Medium priority area: Taking into account both penetration depth and resolution, the frequency is set to 35-45KHz; if the estimated detection depth is in the range of 50-80cm, adjust to 30-40KHz.

[0070] Low priority area: For deep defect detection, the frequency is set to 25-35KHz; if the estimated detection depth exceeds 80cm, the frequency is reduced to 20-25KHz to maximize penetration capability.

[0071] The frequency adjustment step is 5kHz, with an initial default value of 50kHz, and is fine-tuned by ±5kHz based on the real-time echo clarity.

[0072] Based on the estimated detection depth of different priority areas, the pulse energy of the corresponding areas is dynamically adjusted to obtain depth-adaptive energy parameters.

[0073] Specifically, the pulse energy parameters (number of phases) are dynamically adjusted based on the estimated detection depth of each priority area (based on the design thickness and historical damage depth records):

[0074] Deep regions (detection depth > 50cm): The number of phases is gradually increased from the default value of 0.5 to 1.0-2.0, with a step of 0.5. The effective detection range is improved by increasing the pulse length to extend the energy.

[0075] Shallow region (detection depth <20cm): maintain the duration of the pulse for 0.5 or less, and reduce the pulse energy to reduce the surface blind zone (the depth of the blind zone is about 1 / 2 of the pulse wavelength).

[0076] Mid-layer region (detection depth 20-50cm): The number of periods is kept at 0.5, and signal attenuation is compensated primarily by gain adjustment.

[0077] Simultaneously configure the transmission pulse interval parameters: for thin structures (<20cm), set a pause of 10-20ms to eliminate multiple echo interference; for thick structures (>50cm), set a pause of 5-10ms to ensure detection efficiency.

[0078] Based on the structural geometry and material attenuation characteristics, the signal gain is adjusted differently for regions of different thicknesses, and the gain benchmark is corrected by combining material characteristic parameters to obtain the structural response gain model.

[0079] Specifically, the process of differentially adjusting the signal gain for regions of different thicknesses based on structural geometric characteristics and material attenuation characteristics, and then modifying the gain benchmark by combining material characteristic parameters to obtain the structural response gain model includes:

[0080] Obtain the geometric parameters of the hinge structure under test, establish the thickness distribution function along the detection path, and form a structural geometric model;

[0081] Specifically, a laser rangefinder was used to measure the structural thickness along the longitudinal direction of the hinge at 0.5-meter intervals, and the intervals were increased to 0.2 meters in the variable cross-section area to obtain a thickness sampling dataset; simultaneously, an ultrasonic thickness gauge was used to verify the measurement accuracy. The sampling data was imported into the geometric modeling module, and a continuous thickness distribution function h(x) along the detection path was established using the hinge origin as the coordinate origin and a cubic spline interpolation algorithm, forming a structural geometric model that includes the thickness values ​​along the path and their rate of change.

[0082] The concrete strength grade and reinforcement density are obtained as material property parameters, and a material attenuation coefficient mapping relationship is established to form a material attenuation model.

[0083] Specifically, on-site testing of concrete strength was conducted using a rebound hammer at 5-meter intervals along the hinge area, and the strength grade was recorded. Reinforcement density parameters were obtained from design drawings or verified using a rebar scanner. A material attenuation coefficient mapping table was established: the attenuation coefficient for C30 concrete foundations was 0.8 dB / cm, for C40 it was 0.6 dB / cm, and for C50 it was 0.5 dB / cm; when the reinforcement density exceeded 150 kg / m³, an additional attenuation coefficient of 0.1-0.2 dB / cm was applied. The strength and reinforcement parameters were mapped to a material attenuation coefficient α, forming a material attenuation model.

[0084] Based on the aforementioned structural geometric model, the signal gain benchmark is increased proportionally to the thickness increasing region, and the signal gain benchmark is decreased proportionally to the thickness decreasing region. A saturation constraint is applied to the gain value of the thin-walled region to limit the upper limit of the gain. At the same time, the pulse reflection interval parameter is configured to adjust the echo acquisition timing, thereby generating a thickness adjustment rule set.

[0085] Specifically, based on the structural geometry model, a baseline thickness of 20cm is set to correspond to a baseline gain of 30dB. A linear gain adjustment relationship is established: for every 10cm increase in thickness, the analog gain increases by 3-5dB, and the color gain increases by 3-4.5dB; for every 10cm decrease in thickness, the gain decreases proportionally. Saturation constraints are applied to thin-walled regions (thickness <15cm), limiting the analog gain to ≤35dB and the color gain to ≤30dB. Pulse reflection interval parameters are configured synchronously: for thicknesses <20cm, the interval is set to 15-20ms; for thicknesses >50cm, the interval is set to 5-10ms; intermediate thicknesses are calculated using linear interpolation. These rules generate a thickness adjustment rule set covering the entire detection path.

[0086] Based on the material attenuation model, gain compensation correction factors are applied to the high-strength grade region and the high reinforcement density region respectively to generate a set of material correction factors.

[0087] Specifically, based on the material attenuation model, a gain compensation correction factor is determined: for high-strength regions with a strength grade ≥ C50, a correction factor of 0.85-0.90 is applied to reduce the gain; for highly reinforced regions with a reinforcement density > 150 kg / m³, a correction factor of 1.10-1.15 is applied to compensate for energy scattering loss. A set of material correction factors γ(x) corresponding to the detection location coordinates is formed.

[0088] The thickness adjustment rule set and the material correction factor set are weighted and coupled to generate gain configuration values ​​for each detection location, thus constructing a structural response gain model.

[0089] Specifically, at each detection location, the gain value G_thickness(x) calculated from the thickness adjustment rule set is multiplied by the material correction factor set γ(x) to obtain the initial gain configuration value. A three-point moving average method is used to smooth the gain values ​​of adjacent measurement points, avoiding parameter jumps. A field coupling fine-tuning amount from -5dB to +5dB is superimposed to generate the final simulated gain, color gain, and pulse interval parameters for each location. A complete structural response gain model file is output, including the detection location, thickness value, gain configuration value, and corresponding pulse parameters.

[0090] The region-adaptive frequency configuration, depth-adaptive energy parameters, and structural response gain model are coupled with multiple parameters to generate the dynamic parameter configuration scheme.

[0091] Specifically, a three-dimensional parameter matrix (frequency × energy × gain) is established, with each detection grid node corresponding to a set of parameter vectors. Discrete point tests are performed in high-priority areas to verify the effectiveness of the parameter combination. If the echo signal-to-noise ratio is <20dB or the defect identification rate is <70%, the gain is adjusted by ±5dB or the frequency by ±5kHz. A dynamic parameter configuration scheme file containing the coordinates, frequency values, period values, gain values, and pulse interval values ​​of each grid node is generated.

[0092] Step S3: Based on the frequency setting, structural surface flatness, and relative positional relationship of adjacent components in the parameter configuration scheme, a parameter-condition collaborative determination method is used to determine the spatial layout strategy.

[0093] Specifically, step S3 includes the following process:

[0094] Based on the frequency settings in the parameter configuration scheme, a correspondence between frequency and deployment method is established to obtain frequency adaptation mapping;

[0095] Specifically, the frequency parameter configuration scheme establishes a mapping relationship according to the following rules:

[0096] High-frequency configuration (≥50KHz): Corresponding to horizontal layout, arrange the array transducer perpendicular to the hinge joint direction to obtain detection data within a range of 15 - 20 cm on both sides of the center line of the hinge joint, and optimize the resolution of micro cracks and peeling defects with a depth of 5 - 20 cm near the surface.

[0097] Medium-frequency configuration (35 - 45KHz): Select horizontal or vertical layout according to the on-site flatness, and preferentially use horizontal layout; if the flatness is poor, switch to vertical layout.

[0098] Low-frequency configuration (≤35KHz): Corresponding to vertical layout, arrange the array transducer parallel to the hinge joint direction and along the edge of the hollow slab beam. The penetration depth can reach 60 - 80 cm, and optimize the detection ability of deep cavities and loose defects.

[0099] Encode this mapping relationship into a parameter table to generate a frequency adaptation mapping file of frequency - layout method.

[0100] Establish a mapping relationship between flatness and layout method based on the evaluation result of the structural surface flatness to obtain a flatness adaptation mapping;

[0101] Specifically, the process of establishing a mapping relationship between flatness and layout method based on the evaluation result of the structural surface flatness to obtain a flatness adaptation mapping includes:

[0102] Perform preprocessing on the bottom surface of the detection area to remove surface attachments and improve surface finish;

[0103] Specifically, use an angle grinder or sandblasting equipment to grind in longitudinal segments to remove surface laitance, dust, oil stains and loose particles, and clean the dust.

[0104] Evaluate the flatness level based on the surface state after preprocessing to form a quantified flatness grading result;

[0105] Specifically, after preprocessing, use a 2-meter straightedge and a feeler gauge to measure the flatness: Place the straightedge closely against the bottom surface in both longitudinal and transverse directions of the hinge joint, and use the feeler gauge to measure the maximum gap between the straightedge and the concrete surface. Take every 1 meter longitudinally and every 0.5 meter transversely as a measurement area, measure 3 - 5 points in each measurement area, and record the maximum gap value d_max. Perform quantified grading according to d_max:

[0106] High flatness level (Level I): d_max ≤ 2mm, the surface is flat and no special treatment is required;

[0107] Medium flatness level (Level II): 2mm < d_max ≤ 5mm, with local micro - concavities and convexities, which can be compensated by a coupling agent;

[0108] Low flatness level (Level III): d_max>5mm, the surface has large undulations, and the probe is difficult to couple stably.

[0109] Generate a flatness measurement and grading result table containing the survey area number, location mileage, flatness grade, and d_max value.

[0110] Establish a mapping rule between flatness level and layout method, assign high flatness level to horizontal layout to expand the detection coverage, and assign low flatness level to vertical layout to ensure probe coupling quality, thus forming the flatness adaptation mapping.

[0111] Specifically, a mapping rule is established between flatness level and array layout method:

[0112] Level I (High Flatness) → Lateral Layout: Arrange the 48-channel array transducers perpendicular to the hinge direction, covering a range of 15-20cm to the left and right of the hinge centerline, with a lateral step of 200mm to maximize the detection coverage area and improve detection efficiency.

[0113] Level II (Medium flatness) → Horizontal layout + step size adjustment: Horizontal layout is preferred, but the horizontal step size is reduced to 100mm. Data integrity in local uneven areas is ensured by dense layout. If local d_max>5mm, vertical layout is switched in that section.

[0114] Level III (Low Flatness) → Longitudinal Layout: The array transducers are arranged parallel to the hinge joint direction, along the straight area of ​​the edge of the hollow slab beam on one side, covering only a 10-15cm range on one side of the hinge joint, ensuring probe coupling consistency, sacrificing coverage for signal quality.

[0115] The mapping rule is encoded into executable logic: input flatness level, output layout method, step size, and coverage width parameters to form a flatness adaptation mapping database.

[0116] A mapping relationship between misalignment state and layout method is established based on the relative positional relationship of adjacent components to obtain misalignment state adaptation mapping;

[0117] Specifically, a steel ruler is used to measure the height difference between the bottom surfaces of adjacent hollow slab beams. A height difference ≤ 5mm is considered no misalignment, 5-15mm is considered slight misalignment, and > 15mm is considered significant misalignment. In areas with no misalignment, a transverse layout is used, allowing for stable scanning directly below the hinge joint. In areas with slight misalignment, an inclined transverse layout is used, adjusting the array angle to make it parallel to the bottom surface of the lower side slab beam, or switching to a longitudinal layout. In areas with significant misalignment, a forced longitudinal layout is used, arranged along the straight area of ​​the edge of a single side slab beam to avoid misalignment interference. The correspondence between misalignment status and layout method is encoded into a misalignment status adaptation mapping file.

[0118] The frequency adaptation mapping, the flatness adaptation mapping, and the misalignment state adaptation mapping are collaboratively determined to generate a spatial deployment strategy.

[0119] Specifically, a priority weighting method is adopted, assigning a weight of 0.4 to flatness mapping, 0.4 to misalignment mapping, and 0.2 to frequency mapping. For the same detection area, if the flatness and misalignment mapping results conflict (e.g., high flatness but significant misalignment), the misalignment mapping result takes precedence (vertical layout); if the three mapping results are consistent, the result is directly adopted; if they are partially consistent, the final layout method is determined by weighted voting. A segmented layout plan file is generated, recording the starting mileage, layout method (lateral / vertical), array angle, step size parameters, and decision basis for each segment. The single-sided coverage direction is marked for the longitudinal layout segment, and the scanning width is marked for the lateral layout segment, forming a spatial layout strategy that can directly guide on-site operations.

[0120] Step S4: According to the spatial layout strategy, perform discrete point parameter optimization in the high-priority area and establish a position-parameter mapping relationship. When scanning continuously along the structure, call the corresponding mapping parameters and obtain the internal cross-sectional image through array signal synthesis processing.

[0121] Specifically, step S4 includes the following process:

[0122] In the high-priority area, representative measurement points are selected to perform discrete point test mode. Based on the characteristics of the echo signal of each measurement point, the gain, frequency, and pulse energy parameters are optimized and adjusted. The spatial coordinates of each measurement point and its corresponding optimal parameter combination are recorded, and a location-parameter mapping relationship library is established.

[0123] Specifically, discrete measuring points are placed every 0.3m within a 0.5m radius of the location of apparent defects (water seepage, cracks); in areas without apparent defects, they are evenly spaced at 2-3m intervals. A single-point test in REVIEW mode is performed on each measuring point: the array transducer is fixed at the measuring point location, A-Scan echo signals are acquired, and the echo waveform quality is observed. If the defect echo amplitude is less than 30% of full scale or the signal-to-noise ratio is less than 20dB, the analog gain is gradually increased by 2-5dB or the color gain by 3-5dB; if clutter appears in the near-surface dead zone, the gain is reduced or the pulse interval is adjusted. Record the optimal parameter combination (frequency, gain, number of periods, pulse interval) at the measurement point when the defect echo amplitude reaches 40%-60% of the full range and the signal-to-noise ratio is >25dB. Then, use an RTK measuring instrument or laser rangefinder to accurately determine the three-dimensional coordinates (longitudinal mileage, lateral offset, elevation) of the measurement point. Associate and store the coordinates (x, y, z) with the optimal parameter vector P (frequency, gain, pulse energy, pulse interval) to form a position-parameter mapping relationship library (format: measurement point ID, coordinates, optimal parameters).

[0124] The scanning path direction and step size parameters are determined according to the spatial layout strategy. Continuous scanning tests are performed along the hinge structure. At each scanning position point, the corresponding optimal parameter configuration in the mapping relationship library is called.

[0125] Specifically, when laying out the markings laterally, the marking lines are perpendicular to the hinge joint direction, marking the center line position and left and right boundaries of the array; when laying out the markings longitudinally, the marking lines are parallel to the hinge joint direction, arranged along the edge of the single-sided beam. The scanning step size is determined according to priority: the horizontal step size is set to 100mm and the vertical step size is 100mm in high-priority areas; the horizontal step size is set to 200mm and the vertical step size is 100mm in medium-priority areas. Continuous scanning in MAP mode is started, and the device moves automatically along the baseline. Each time it reaches a scanning position point (grid node), the coordinates of that point are read, and the nearest measuring point (search radius 0.5m) is searched in the mapping relationship library. The optimal parameter configuration for that measuring point is automatically called; if there is no mapped measuring point within 1m of a certain position point, the parameters of that point are generated by linear interpolation of the parameters of adjacent measuring points.

[0126] The reflected echo signal was acquired by array transducer, and the multi-channel signal was delayed and superimposed by synthetic aperture focusing technology to reconstruct a two-dimensional cross-sectional image of the concrete component.

[0127] Specifically, the process of acquiring reflected echo signals through an array of transducers, and then using synthetic aperture focusing technology to perform time-delay superposition processing on the multi-channel signals to reconstruct a two-dimensional cross-sectional image of the concrete component includes:

[0128] Excitation and reception control are performed on the array transducer to acquire multi-channel reflected echo signals and form a multi-channel holographic dataset.

[0129] Specifically, an array transducer consisting of 48 shear wave sensors in 4 sets of ×12 arrays is used. Sequential excitation is executed by the control unit: the sensors in rows 1 to 12 are used as transmitting units, each transmitting a narrow pulse ultrasonic wave with a duration of 50 μs and a center frequency configured according to step S2 (e.g., 50 kHz); the remaining 11 rows of sensors synchronously receive reflected echo signals, with a sampling frequency of 25 MHz and 2048 sampling points, recording complete time-domain waveform data. After each transmit-receive cycle, 132 channels (12 transmits × 11 receive) of holographic data are acquired, with a single scan point acquisition time of approximately 1.5 seconds, forming a multi-channel holographic dataset containing the amplitude, phase, and propagation time of each channel's time-domain signal.

[0130] The propagation delay of each channel signal is calculated based on the wave velocity and sensor geometry, and a channel delay parameter set is established.

[0131] Specifically, based on the concrete shear wave velocity V calibrated in step S2 (typically 2500-3000 m / s) and the sensor geometric arrangement parameters (row spacing 10 mm, column spacing 8 mm), a three-dimensional spatial coordinate system is established: with the array center as the origin, the depth direction as the z-axis, the array transverse direction as the x-axis, and the longitudinal direction as the y-axis. For each imaging point ( , )( Coordinates along the array direction, (Using depth coordinates), calculate the total propagation path distance for transmission in the m-th row and reception in the n-th row. Thus, the propagation time delay is obtained. The delay was calculated for all 132 channels and imaging regions (depth range 0-80cm, lateral coverage 40cm, resolution 5mm×5mm) to form a channel delay parameter set containing the time delay matrix of each channel-imaging point.

[0132] The signals of each channel in the multi-channel holographic dataset are precisely compensated according to the delay parameter to achieve time-domain alignment of signal components from the same target point, forming a time-domain aligned signal set.

[0133] Specifically, for each channel signal in the multi-channel holographic dataset According to the channel delay parameter set Digital signal delay compensation is performed: an FIR interpolation filter (order 128) is used to achieve precise delay with sub-sampling accuracy (0.01μs), compensating for the delay of each channel signal. Shift to the reference time axis with the target imaging point as the zero point, i.e. After completing delay compensation for all channels, the defect reflection signal components from the same imaging point are precisely aligned to the same time in the time domain, forming a time-domain aligned signal set with consistent signal phase for each channel.

[0134] The time-domain aligned signal set is coherently superimposed to enhance the target signal and suppress noise, thereby completing the reconstruction of the two-dimensional cross-sectional image inside the concrete.

[0135] Specifically, amplitude-weighted coherent superposition is performed on all 132 channels of the time-domain aligned signal set: for each channel signal... The amplitude was obtained by sampling at the target imaging time t=0. , by amplitude weighting factor (The greater the distance, the greater the energy attenuation.) Weighted values ​​are then summed to obtain the total amplitude of the imaging point. Where M represents the number of rows of transmitting sensors and N represents the number of rows of receiving sensors (for a 48-channel array, typically M=12, N=11). Through this superposition operation, the target point signal amplitude is enhanced by approximately 20log10(132)≈42dB, while the random noise is only enhanced by approximately 10dB due to incoherent superposition, resulting in a net increase in signal-to-noise ratio of approximately 32dB. The superimposed amplitude I( , The image is mapped to grayscale values ​​(0-255), arranged by depth-horizontal coordinates, and a two-dimensional B-Scan cross-sectional image is generated. The dynamic range is compressed to 60dB to enhance the visualization effect, thus completing the reconstruction of the two-dimensional cross-sectional image inside the concrete.

[0136] Multiple two-dimensional cross-sectional images acquired through continuous scanning are stitched together in spatial order to generate an overall internal structure image of the detection area.

[0137] Specifically, grayscale feature points of the hinge centerline in each image are extracted as registration references, and affine transformation is used to correct image distortion caused by equipment movement deviations. The images are seamlessly stitched longitudinally (along the hinge direction) according to the scanning step size, and the average grayscale value of overlapping areas is used to eliminate seams. After stitching, an overall internal structure image of the detection area (C-Scan view) is generated, with the horizontal axis representing the longitudinal mileage of the hinge and the vertical axis representing depth, and color coding indicating reflection intensity. Simultaneously, a 3D view is generated, allowing observation from any cross-section, and the coordinates and geometric dimensions of candidate defect locations are labeled.

[0138] Step S5: Based on the abnormal features in the internal cross-sectional image and the geometric relationship between the probe and the structural boundary, a multi-factor interference compensation method is used to correct image distortion in order to obtain the target defect determination result.

[0139] Specifically, step S5 includes the following process:

[0140] Anomaly features are extracted from the internal cross-sectional image to identify candidate defect regions, resulting in an anomaly feature set.

[0141] Specifically, an adaptive Otsu algorithm is used to calculate the globally optimal threshold T. Regions with image grayscale values ​​> 1.5T and connected pixel areas > 10 mm² are marked as initial outliers. Morphological opening operations (kernel size 3×3 pixels) are performed on these initial outliers to remove isolated noise, and closing operations (kernel size 5×5 pixels) are performed to connect broken regions, forming connected components. Geometric parameters are calculated for each connected component: centroid coordinates (x, z), area S, major axis length L, minor axis length W, aspect ratio γ = L / W, average grayscale value G, and maximum grayscale gradient ∇G. The following screening conditions are set: G > 200, ∇G > 50, and γ < 5. Regions meeting these conditions are identified as candidate defect regions, and their feature vectors F = (x, z, S, L, W, γ, G) are recorded to form an anomaly feature set.

[0142] Obtain the probe position coordinates and the geometric parameters of the structural boundary, establish a probe-boundary spatial relationship model, and calculate the distance parameters between the detection points and the boundary to obtain a geometric constraint parameter set;

[0143] Specifically, during the scanning process, the real-time position coordinates (x_p, y_p) of the probe are recorded by the built-in encoder of the device, and the precise coordinates (x_b, y_b) of the hinge joint structure boundary (side, end) are measured by a total station. Establish a spatial relationship model: Calculate the distance d_b = |x_p - x_b| (lateral boundary) or the longitudinal distance d_l = |y_p - y_b| from the centroid of the candidate defect area to the nearest structural boundary. For each scan line, record the starting position, ending position of the probe and the boundary coordinates, and form a geometric constraint parameter set including the measurement point position, boundary distance, and member thickness.

[0144] Based on the concrete material properties, surface conditions, and boundary effects, construct a multi-factor interference analysis model, and perform multi-factor correction on the abnormal feature set to obtain an interference correction factor set;

[0145] Specifically, the construction of the multi-factor interference analysis model and the multi-factor correction of the abnormal feature set include:

[0146] Obtain the material properties, surface conditions, and boundary conditions as multi-physical field influence parameters, and establish corresponding attenuation correction sub-models respectively;

[0147] Specifically, the material attenuation sub-model: Use a rebound hammer to detect a measurement point every 5 meters along the hinge joint to obtain the concrete strength grade C (C30 - C50); Use a steel bar scanner to sample and measure the reinforcement density ρ (unit kg / m³). Establish a material attenuation coefficient calculation model: . This model reflects the additional attenuation of ultrasonic energy by high-strength and high-reinforcement concrete, and needs to be calibrated and verified in advance by standard test blocks, with the error controlled within ±5%.

[0148] The surface energy loss sub-model: Use a feeler gauge and a 2-meter straightedge to measure the maximum gap d_max on the preprocessed surface, and evaluate the flatness grade: Grade I (d_max ≤ 2mm), Grade II (2mm < d_max ≤ 5mm), Grade III (d_max > 5mm). Record the thickness t_c of the coupling agent applied (thin layer < 1mm, thick layer ≥ 1mm). Establish a surface correction model: = [1, 1.1, 1.2][flatness grade] × [1, 0.9][coupling agent thickness], quantifying the energy loss caused by surface roughness and coupling quality.

[0149] Boundary distortion correction sub-model: The coordinates of the hinge side boundary and end boundary are accurately measured using a total station, and the distance from the centroid of the candidate defect region to the nearest boundary is calculated. (Unit: cm). Establish boundary influence model: ,when When the depth is less than 10cm, the linear correction is increased to a maximum of 1.5 to compensate for the positioning distortion caused by the reduction of boundary wave velocity and the narrowing of the acoustic path.

[0150] The various attenuation correction sub-models are coupled to construct a multi-factor coupling analysis framework;

[0151] Specifically, a multiplication coupling mechanism is used to construct the analysis framework: This framework assumes that material attenuation, surface loss, and boundary effects are independent physical processes that act sequentially on the ultrasonic signal, conforming to the basic laws of sound wave propagation. In the computational flow, the coefficients of the three sub-models are first calculated in parallel, then multiplication is performed to form a batch-processable coupled analysis module. Threshold protection is embedded in the framework: when… An error alarm is triggered when the value is >2.0 or <0.8, prompting a re-verification of the input parameters to avoid image distortion caused by extreme corrections.

[0152] The comprehensive correction coefficients are calculated on the abnormal feature set using the multi-factor coupling analysis framework to obtain the interference correction factor set.

[0153] Specifically, for each candidate defect region in the anomaly feature set, its spatial coordinates (x,z) are extracted as input, and the coupling analysis framework is invoked to calculate the coordinates at that location. Simultaneously decompose and output the coefficients of the three components. Dedicated correction components are assigned to different characteristic parameters: for position deviation correction. Size distortion correction Intensity attenuation correction The comprehensive correction coefficient and its component coefficients for each candidate defect region are packaged into a correction factor vector. After traversing all candidate regions, an interference correction factor set covering the entire detection area is formed and stored as a data table associated with the defect coordinates.

[0154] The interference correction factor set is applied to the internal cross-sectional image to compensate and correct the positional deviation and size distortion of the candidate defect region to obtain a distortion-corrected image.

[0155] Specifically, the process of compensating for and correcting the positional deviation and dimensional distortion of candidate defect regions to obtain a distortion-corrected image includes:

[0156] The position correction component, size correction amount, and intensity correction amount in the interference correction factor set are extracted, and geometric transformations are performed on the coordinate position, contour boundary, and reflection intensity of the candidate defect region in the internal cross-section image to obtain the position correction image, size correction image, and intensity correction image, respectively.

[0157] Specifically, according to the interference correction factor set The three independent correction components are decoupled and extracted according to the following rules:

[0158] Position correction component Δx: directly derived from boundary correction coefficients calculate, ,in The boundary effect empirical coefficient (taken as 5 mm). This is the distance from the centroid of the candidate defect region to the boundary. When < 10cm, > 1, Δx is a positive value, indicating that it needs to be translated away from the boundary for correction.

[0159] Size correction amount : Derived from the material attenuation coefficient calculate, When the concrete has high strength and dense reinforcement ( When > 1.2), < 1 indicates that the defect size is amplified due to energy decay and needs to be reduced for correction.

[0160] Strength correction amount : From the surface state coefficient With coupling agent correction factor Calculation of the product, When the surface is rough or the coupling agent is thick ( > 1.1, When = 0.9), ≈ 1.0, to compensate for energy loss.

[0161] For each candidate defect region, extract its specific correction component vector β = (Δx, , Store it in a temporary cache.

[0162] Specifically, for the centroid coordinates (x, z) of each candidate defect region in the internal cross-sectional image, a translation transformation x' = x + Δx and z' = z are applied (the depth direction is not corrected for the time being). A bicubic interpolation algorithm is used to remap the coordinates of all pixels in the connected domain of the defect, generating a position-corrected image that only corrects the positional deviation, ensuring that the defect positioning error near the boundary is reduced from ±15mm to ±5mm.

[0163] The set of pixels representing the contour boundaries of the candidate defect region {( , Perform a scale transformation: using the centroid ( , Based on ) , A conformal transformation algorithm is used to avoid contour distortion, and the area of ​​the region is recalculated. This generates a scale-corrected image that corrects dimensional distortion, reducing dimensional measurement error from ±25% to ±8%.

[0164] The original grayscale values ​​of each pixel within the candidate defect region Execute amplitude adjustment Simultaneously, linear stretching (mapping the corrected grayscale range to 0-255) is performed to maintain image contrast. Pixels in non-candidate regions retain their original values, generating an intensity-corrected image to compensate for energy attenuation, thereby improving the consistency of defect reflection intensity under different surface conditions by 40%.

[0165] The position-corrected image, the scale-corrected image, and the intensity-corrected image are fused and edge-smoothed to generate the distortion-corrected image.

[0166] Specifically, the three component images are superimposed according to their importance weights: I_fused = 0.5 × I_position + 0.3 × I_scale + 0.2 × I_intensity. Positional error has the greatest impact on structural safety assessment (weight 0.5), size error affects quantitative accuracy (weight 0.3), and intensity error affects defect nature judgment (weight 0.2). Gaussian smoothing filtering (kernel size 5×5, σ=1) is applied to the boundaries of each candidate defect region in the fused image to eliminate jagged edges caused by geometric transformations; for boundaries where the distance between adjacent defect regions is <5mm, morphological dilation-erosion operation is used to merge connected components to avoid oversegmentation. Histogram equalization and contrast-limited adaptive histogram equalization (CLAHE) are performed on the fused image to improve the overall image visual quality and generate the final distortion-corrected image, which has smooth defect contours, accurate positioning, and uniform grayscale.

[0167] Based on the distortion-corrected image, pattern matching is performed using a preset defect feature library to determine the defect type, location, and geometric parameters to obtain the target defect determination result.

[0168] Specifically, a pre-defined defect feature library is established, containing feature vector templates for four typical defect types: voids (γ≈1, high G, circular), looseness (γ≈2-3, medium-high G, irregular), cracks (γ>4, high G, linear), and peeling (γ≈2-4, medium-high G, layered). A nearest neighbor classification algorithm is used to calculate the similarity distance D between the feature vector F' of the candidate defect region in the distortion-corrected image and the feature library templates. A threshold of D<0.3 is set for a successful match, thus determining the defect type. The defect location (depth z' from the hinge surface, longitudinal position x') is determined based on the centroid coordinates (x', z'). Based on Δx, ... , Calculate the equivalent diameter of the defect Output the target defect determination result (type, location, size).

[0169] Specifically, this invention significantly improves the accuracy and reliability of detecting internal defects in hinge joints by dynamically coupling detection parameters with structural characteristics and environmental conditions. To address the inherent differences in the response of defects of different depths to ultrasonic frequencies, adaptive matching optimizes near-surface resolution with high frequencies and enhances deep penetration with low frequencies, avoiding insufficient penetration or resolution loss due to fixed frequencies. By positively adjusting the gain in relation to structural thickness, energy loss due to natural attenuation of sound waves in concrete due to path length is compensated. Simultaneously, the gain benchmark is corrected based on material strength and reinforcement density to prevent artifacts caused by over-gain in thin-walled areas and missed defects in deep areas due to under-gain, ensuring the signal strength remains within the optimal recognition range. Surface flatness and misalignment directly determine the probe coupling effect; the mapping rule adaptively selects horizontal or vertical placement to maximize the contact area between the probe and concrete, fundamentally improving signal acquisition quality. Multi-channel synthetic aperture focusing, through precise calculation and weighted superposition of sound wave propagation delays, significantly improves the signal-to-noise ratio by utilizing the coherent enhancement characteristics of signals from different paths, enabling clear imaging of deep, weak defects. A multi-factor interference compensation model, based on the objective influence of material attenuation, surface condition, and boundary effects on sound waves, systematically corrects the distortion of defect location, size, and strength, eliminating misjudgments and missed detections caused by differences in detection conditions. This method achieves closed-loop feedback of dynamic optimization of detection parameters with structural characteristics, transforming the traditional experience-based manual debugging into automatic configuration based on physical propagation laws. While improving the defect detection rate and quantitative accuracy, it significantly shortens the on-site debugging time, reduces the dependence on the skill level of operators, and the generated digital detection results are comparable across working conditions, providing scientific, reliable, and traceable data support for bridge maintenance decisions.

[0170] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0171] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges, characterized in that, include: Step S1: Obtain the feature information and historical damage records of the structure to be tested, and use the state assessment method to divide the regions with different detection priority to output the region division results; Step S2: Based on the region division results, the partition parameter configuration method is used to set corresponding detection frequency and energy parameter combinations for different priority regions, and the signal gain is adjusted according to the structural geometric characteristics to obtain a dynamic parameter configuration scheme. Step S3: Based on the frequency setting, structural surface flatness, and relative positional relationship of adjacent components in the parameter configuration scheme, a parameter-condition collaborative determination method is used to determine the spatial layout strategy. Step S4: According to the spatial layout strategy, perform discrete point parameter optimization in the high-priority area and establish a position-parameter mapping relationship. When scanning continuously along the structure, call the corresponding mapping parameters and obtain the internal cross-sectional image through array signal synthesis processing. Step S5: Based on the abnormal features in the internal cross-sectional image and the geometric relationship between the probe and the structural boundary, a multi-factor interference compensation method is used to correct image distortion in order to obtain the target defect determination result.

2. The adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges according to claim 1, characterized in that, The process of step S2 includes: Based on the region division results, differentiated ultrasonic frequencies are configured for different priority regions to obtain region-adaptive frequency configurations. Based on the estimated detection depth of different priority areas, the pulse energy of the corresponding areas is dynamically adjusted to obtain depth-adaptive energy parameters. Based on the structural geometry and material attenuation characteristics, the signal gain is adjusted differently for regions of different thicknesses, and the gain benchmark is corrected by combining material characteristic parameters to obtain the structural response gain model. The region-adaptive frequency configuration, depth-adaptive energy parameters, and structural response gain model are coupled with multiple parameters to generate the dynamic parameter configuration scheme.

3. The adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges according to claim 2, characterized in that, The process of differentially adjusting the signal gain for regions of different thicknesses based on structural geometry and material attenuation characteristics, and then modifying the gain benchmark by combining material characteristic parameters to obtain the structural response gain model includes: Obtain the geometric parameters of the hinge structure under test, establish the thickness distribution function along the detection path, and form a structural geometric model; The concrete strength grade and reinforcement density are obtained as material property parameters, and a material attenuation coefficient mapping relationship is established to form a material attenuation model. Based on the aforementioned structural geometric model, the signal gain benchmark is increased proportionally to the thickness increasing region, and the signal gain benchmark is decreased proportionally to the thickness decreasing region. A saturation constraint is applied to the gain value of the thin-walled region to limit the upper limit of the gain. At the same time, the pulse reflection interval parameter is configured to adjust the echo acquisition timing, thereby generating a thickness adjustment rule set. Based on the material attenuation model, gain compensation correction factors are applied to the high-strength grade region and the high reinforcement density region respectively to generate a set of material correction factors. The thickness adjustment rule set and the material correction factor set are weighted and coupled to generate gain configuration values ​​for each detection location, thus constructing a structural response gain model.

4. The adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges according to claim 3, characterized in that, The process of step S3 includes: Based on the frequency settings in the parameter configuration scheme, a correspondence between frequency and deployment method is established to obtain frequency adaptation mapping; A mapping relationship between flatness and layout method is established based on the flatness evaluation results of the structural surface to obtain the flatness adaptation mapping; A mapping relationship between misalignment state and layout method is established based on the relative positional relationship of adjacent components to obtain misalignment state adaptation mapping; The frequency adaptation mapping, the flatness adaptation mapping, and the misalignment state adaptation mapping are collaboratively determined to generate a spatial deployment strategy.

5. The adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges according to claim 4, characterized in that, The process of establishing a mapping relationship between flatness and layout method based on the surface flatness evaluation results to obtain a flatness adaptation mapping includes: Pre-treatment is performed on the bottom surface of the detection area to remove surface deposits and improve surface smoothness; The flatness level is evaluated based on the pre-treated surface condition to form a quantitative grading result of flatness. Establish a mapping rule between flatness level and layout method, assign high flatness level to horizontal layout to expand the detection coverage, and assign low flatness level to vertical layout to ensure probe coupling quality, thus forming the flatness adaptation mapping.

6. The adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges according to claim 5, characterized in that, The process of step S4 includes: In the high-priority area, representative measurement points are selected to perform discrete point test mode. Based on the characteristics of the echo signal of each measurement point, the gain, frequency, and pulse energy parameters are optimized and adjusted. The spatial coordinates of each measurement point and its corresponding optimal parameter combination are recorded, and a location-parameter mapping relationship library is established. The scanning path direction and step size parameters are determined according to the spatial layout strategy. Continuous scanning tests are performed along the hinge structure. At each scanning position point, the corresponding optimal parameter configuration in the mapping relationship library is called. The reflected echo signal was acquired by array transducer, and the multi-channel signal was delayed and superimposed by synthetic aperture focusing technology to reconstruct a two-dimensional cross-sectional image of the concrete component. Multiple two-dimensional cross-sectional images acquired through continuous scanning are stitched together in spatial order to generate an overall internal structure image of the detection area.

7. The adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges according to claim 6, characterized in that, The process of acquiring reflected echo signals through an array of transducers, and then using synthetic aperture focusing technology to perform time-delay superposition processing on the multi-channel signals to reconstruct a two-dimensional cross-sectional image of the concrete component includes: Excitation and reception control are performed on the array transducer to acquire multi-channel reflected echo signals and form a multi-channel holographic dataset. The propagation delay of each channel signal is calculated based on the wave velocity and sensor geometry, and a channel delay parameter set is established. The signals of each channel in the multi-channel holographic dataset are precisely compensated according to the delay parameter to achieve time-domain alignment of signal components from the same target point, forming a time-domain aligned signal set. The time-domain aligned signal set is coherently superimposed to enhance the target signal and suppress noise, thereby completing the reconstruction of the two-dimensional cross-sectional image inside the concrete.

8. The adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges according to claim 7, characterized in that, The process of step S5 includes: Anomaly features are extracted from the internal cross-sectional image to identify candidate defect regions, resulting in an anomaly feature set. Obtain the probe position coordinates and structural boundary geometric parameters, establish a probe-boundary spatial relationship model, and calculate the distance parameters between the detection point and the boundary to obtain the geometric constraint parameter set; Based on the material properties, surface condition and boundary effect of concrete, a multi-factor interference analysis model is constructed, and the abnormal feature set is modified by multiple factors to obtain the interference correction factor set. The interference correction factor set is applied to the internal cross-sectional image to compensate and correct the positional deviation and size distortion of the candidate defect region to obtain a distortion-corrected image. Based on the distortion-corrected image, pattern matching is performed using a preset defect feature library to determine the defect type, location, and geometric parameters to obtain the target defect determination result.

9. The adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges according to claim 8, characterized in that, The construction of the multi-factor interference analysis model and the multi-factor correction of the abnormal feature set include: Material properties, surface conditions and boundary conditions are obtained as multiphysics field influence parameters, and corresponding attenuation correction sub-models are established respectively. The various attenuation correction sub-models are coupled to construct a multi-factor coupling analysis framework; The comprehensive correction coefficients are calculated on the abnormal feature set using the multi-factor coupling analysis framework to obtain the interference correction factor set.

10. The adaptive detection method for internal defects in hinge joints of prefabricated hollow slab beam bridges according to claim 9, characterized in that, The process of compensating for and correcting the positional deviation and dimensional distortion of candidate defect regions to obtain a distortion-corrected image includes: The position correction component, size correction amount, and intensity correction amount in the interference correction factor set are extracted, and geometric transformations are performed on the coordinate position, contour boundary, and reflection intensity of the candidate defect region in the internal cross-section image to obtain the position correction image, size correction image, and intensity correction image, respectively. The position-corrected image, the scale-corrected image, and the intensity-corrected image are fused and edge-smoothed to generate the distortion-corrected image.

Images