Steel pipe-ecc concrete pier crack detection device and detection method thereof

Abstract

The application relates to the technical field of bridge health monitoring, and discloses a steel pipe-ECC concrete pier crack detection device and a detection method thereof, which comprises a multi-modal sensing module integrated with a high-frequency ultrasonic probe array and an electromagnetic induction sensor group; the high-frequency ultrasonic probe array is composed of a plurality of focused piezoelectric ceramic probes which are distributed in a spiral shape on a flexible base; and the electromagnetic induction sensor group is composed of a plurality of micro planar coils. The steel pipe-ECC concrete pier crack detection device and the detection method thereof integrate the high-frequency ultrasonic array and the electromagnetic induction sensor group, construct an acoustic-electric multi-modal collaborative sensing mechanism, can synchronously acquire the acoustic response of internal cracks of concrete and the electromagnetic characteristic change of interface debonding, realize all-around and penetrating detection of hidden micro-cracks of the steel pipe-ECC pier, and effectively overcome the blind area and the misjudgment problem of single-mode detection.

Description

Technical Field

[0001] This invention relates to the field of bridge health monitoring technology, specifically to a steel pipe-ECC concrete bridge pier crack detection device and its detection method. Background Technology

[0002] In bridge engineering, steel-tube-ECC concrete composite piers are widely used in complex and harsh environments due to their excellent crack resistance, high ductility, and durability. The overall integrity of this composite structure is highly dependent on the interfacial bonding performance between the steel tube and the ECC material. Micron-level early debonding and microcracks that initiate at the interface are significant safety hazards that are difficult to detect directly.

[0003] Current methods for detecting cracks in underwater and outdoor bridge piers primarily employ single-physical-principle detection devices such as optical vision and point-based ultrasound. These methods have significant limitations when dealing with steel pipe-ECC composite structures characterized by small acoustic impedance differences, crack sizes often below 100 micrometers, and complex interface behavior: The detection mode is singular, failing to simultaneously capture the coordinated acoustic and electromagnetic multi-physics responses triggered by microcrack propagation, making it difficult to effectively identify hidden interface defects; sensor positioning mechanisms are mostly rigid structures, making it difficult to dynamically conform to curved bridge piers and maintain a constant quasi-non-contact detection distance, resulting in a significant decrease in signal stability and accuracy when the pier surface is corroded or uneven; data processing lacks spatiotemporal correlation and intelligent fusion criteria for multi-source heterogeneous signals, leading to highly subjective and poor repeatability of detection results. Therefore, developing dedicated detection devices and methods capable of adapting to complex curved surfaces and achieving multi-modal collaborative sensing and intelligent analysis has become a critical technical challenge urgently needing to be addressed in this field. Summary of the Invention

[0004] To address the problems mentioned in the background section, the present invention provides the following technical solution: a steel pipe-ECC concrete bridge pier crack detection device, comprising:

[0005] The multimodal sensing module integrates a high-frequency ultrasonic probe array and an electromagnetic induction sensor group. The high-frequency ultrasonic probe array consists of multiple focused piezoelectric ceramic probes arranged in a spiral pattern on a flexible substrate. The electromagnetic induction sensor group consists of multiple miniature planar coils, which are interleaved and embedded between the spirally distributed probes. The multimodal sensing module is used to perform synchronous quasi-non-contact scanning of the interior of the ECC concrete of the bridge pier and the steel pipe-ECC interface area to acquire ultrasonic echo signals and electromagnetic impedance change data, respectively.

[0006] An adaptive positioning module includes an annular guide rail that can be arranged around the outer wall of the pier, a displacement servo motor that drives the multimodal sensing module to move along the annular guide rail, and a distance closed-loop control unit for maintaining a constant detection distance between the multimodal sensing module and the surface of the pier.

[0007] A signal fusion analysis module, which is communicatively connected to the multimodal sensing module, is configured to: perform wavelet packet decomposition on the ultrasonic echo signal to extract energy entropy features; calculate the phase angle change rate of the electromagnetic impedance change data; and construct a spatial coordinate decision map with the energy entropy features and the phase angle change rate as dual parameters. The spatial coordinate decision map uses a system-preset dynamic threshold for anomaly detection, which includes an energy entropy dynamic threshold and a phase angle change rate dynamic threshold. When the energy entropy features and phase angle change rate corresponding to the same spatial coordinate point exceed the corresponding dynamic thresholds, and three or more consecutive adjacent sampling points meet the above conditions, a potential crack is determined to exist.

[0008] Preferably, the high-frequency ultrasonic probe array is a focused piezoelectric ceramic probe with a center frequency of 5 to 10 MHz, which is spirally distributed on a flexible substrate, and the spacing between adjacent probes is 15 to 25 mm; the electromagnetic induction sensor group is composed of multiple miniature planar coils, each with a diameter of 8 to 12 mm, and its excitation frequency is set in the range of 50 to 200 kHz.

[0009] Preferably, the inner side of the annular guide rail of the adaptive positioning module is provided with a magnetic attachment layer; the distance closed-loop control unit includes a laser ranging sensor and a PID controller. The laser ranging sensor is used to measure the distance between the multimodal sensing module and the surface of the bridge pier in real time. The PID controller adjusts the length of a telescopic arm according to the distance signal to drive the multimodal sensing module to perform displacement compensation.

[0010] Preferably, the signal fusion analysis module is configured to: perform bandpass filtering and time-series alignment preprocessing on the original ultrasonic echo signal and electromagnetic impedance change data; perform 4-level wavelet packet decomposition on the ultrasonic echo signal using the db6 wavelet basis to extract the energy entropy of a specific frequency band as a damage-sensitive feature; calculate the impedance phase angle change rate on the electromagnetic impedance change data; normalize the energy entropy and the phase angle change rate, and generate a "energy entropy-phase angle change rate" dual-parameter decision map in the same spatial coordinate system; when both feature values ​​corresponding to a certain spatial coordinate point exceed the system's preset dynamic threshold, and three or more consecutive adjacent sampling points are marked as abnormal, it is determined that there is a potential crack in the area.

[0011] Preferably, the signal fusion analysis module also has a built-in machine learning classifier, which is trained using a dataset containing simulated interface crack response patterns with different widths, depths and inclination angles, and is used to classify the severity of the identified potential cracks.

[0012] Preferably, it also includes a detachable protective shell made of carbon fiber composite material, which integrates a temperature and humidity sensor and an anti-condensation heating element. When the ambient humidity or temperature reaches a preset condition, the heating element is automatically activated.

[0013] A method for detecting cracks in steel pipe-ECC concrete bridge piers includes the following steps:

[0014] S1: Install the device on the outer surface of the bridge pier, and drive the multimodal sensing module to move along a preset path through the adaptive positioning module to scan the surface of the bridge pier;

[0015] S2: During the scanning process, the position of the multimodal sensing module is dynamically adjusted by the distance closed-loop control unit to keep the detection distance between it and the surface of the bridge pier constant;

[0016] S3: The ultrasonic echo signal and electromagnetic impedance change data of the scanning area are synchronously acquired through the multimodal sensing module and transmitted to the signal fusion analysis module;

[0017] S4: The signal fusion analysis module performs fusion analysis on the ultrasonic echo signal and electromagnetic impedance change data to identify and locate cracks inside the ECC concrete of the bridge pier and in the steel pipe-ECC interface area.

[0018] Preferably, the preset path in step S1 adopts a hierarchical scanning strategy: first, a coarse scan of the bridge pier is performed with a first preset step distance; when a suspected crack signal is found in the coarse scan, it automatically switches to a high-density encrypted scan around the suspected area with a second preset step distance smaller than the first preset step distance.

[0019] Preferably, during the high-density encrypted scanning, the high-frequency ultrasonic probe array is also controlled to emit ultrasonic waves in a multi-angle tilt incident manner, and the tilt angle information of the crack is determined based on the acquired multi-angle echo signals.

[0020] Preferably, in step S4, the fusion analysis includes: extracting energy entropy features from the ultrasonic echo signal and extracting phase angle change rate features from the electromagnetic impedance change data; performing spatiotemporal alignment and normalization on the two features; identifying abnormal regions based on the dual-parameter joint criterion, and performing three-dimensional reconstruction of cracks and severity classification on the identified abnormal regions.

[0021] Compared with the prior art, the present invention provides a crack detection device and method for steel pipe-ECC concrete bridge piers, which has the following beneficial effects:

[0022] 1. The steel pipe-ECC concrete bridge pier crack detection device and its detection method integrate a high-frequency ultrasonic array and an electromagnetic induction sensor group to construct an acoustic-electric multimodal collaborative sensing mechanism. It can simultaneously acquire the acoustic response of internal concrete cracks and the electromagnetic characteristic changes of interface debonding, realizing all-round and penetrating detection of hidden microcracks in steel pipe-ECC bridge piers, effectively overcoming the blind spots and missed detection problems of single-modal detection.

[0023] 2. The steel pipe-ECC concrete bridge pier crack detection device and its detection method adopt an adaptive positioning module with magnetic adhesion and closed-loop control functions. Combined with laser ranging and PID real-time adjustment, it can dynamically compensate for the changes in detection distance caused by undulations, corrosion or uneven coating of the bridge pier surface, ensuring that the sensor is always at the optimal working distance, which greatly improves the signal-to-noise ratio and measurement repeatability of quasi-non-contact scanning.

[0024] 3. The steel pipe-ECC concrete pier crack detection device and its detection method establish an intelligent analysis model based on the spatiotemporal correlation of multi-source signals by introducing wavelet packet decomposition, feature fusion, and machine learning classifiers. This model can not only accurately identify the location of cracks, but also perform graded and quantitative assessments of crack width, depth, inclination angle, and severity, greatly reducing the interference of subjective human factors and improving the scientific nature and engineering guidance value of the health assessment results.

[0025] 4. The steel pipe-ECC concrete bridge pier crack detection device and its detection method feature a detachable carbon fiber protective shell that integrates temperature and humidity monitoring and automatic anti-condensation heating functions, ensuring reliable operation of the equipment in harsh outdoor environments such as high humidity and low temperature. The tiered scanning strategy (coarse scanning for positioning, fine scanning for analysis) and the design of the rapid installation structure optimize the balance between detection efficiency and accuracy, significantly improving the ease of rapid deployment and maintenance of the device on site.

[0026] 5. This steel-tube-ECC concrete bridge pier crack detection device and method systematically integrates a spirally arranged acoustic-electric multimodal sensor array, an adaptive ring guide rail with constant-distance closed-loop control, and a fusion analysis module based on a dual-parameter spatial coincidence criterion, forming a dedicated technical solution for detecting interface cracks in steel-tube-ECC bridge piers. The spiral arrangement ensures the continuity of curved surface scanning, while the staggered embedding enables synchronous sensing of acoustic-electric properties at the same spatial point. Constant-distance scanning ensures the consistency of the spatiotemporal reference between ultrasonic and electromagnetic signals during dynamic processes, providing a prerequisite for the subsequent accurate "spatial coordinate coincidence" criterion. This complementary structural embedding and spatiotemporal functional synergy jointly solve the technical challenges of blind spots in single-modal detection and dynamic scanning signal mismatch. Its technical effect far exceeds the simple superposition of the functions of each module, constituting a significant and substantial feature. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the process of the present invention;

[0028] Figure 2 This is a schematic diagram of the multi-circle scanning dual-parameter fusion monitoring of the present invention;

[0029] Figure 3 This is a schematic diagram of the spiral staggered arrangement of the high-frequency ultrasonic probe and the electromagnetic induction coil in the multimodal sensing module of the present invention.

[0030] Figure 4 This is a cross-sectional view of the annular guide rail and telescopic arm of the adaptive positioning module of the present invention.

[0031] Figure 5 This is an example of a two-parameter spatial coincidence decision graph constructed by the signal fusion analysis module of the present invention. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] Please see Figure 1-5 The present invention provides a technical solution:

[0034] Example 1

[0035] See Figure 1 This embodiment provides a steel pipe-ECC concrete bridge pier crack detection device, which aims to solve the technical problems of insufficient detection capability, high false alarm and false alarm rate, and poor environmental adaptability of existing technologies at the interface between steel pipe and ECC concrete.

[0036] The device mainly includes a multimodal sensing module 100, an adaptive positioning module 200, a signal fusion analysis module 300, and a detachable protective housing 400.

[0037] It should be noted that the "quasi-non-contact scanning" described in this embodiment does not refer to the propagation of ultrasonic signals through long-distance air medium, but rather to the multimodal sensing module, driven by the adaptive positioning module, dynamically maintaining a constant and minute air gap (i.e., lift-off distance, which is constant at 5mm ± 0.5mm in this embodiment) between itself and the pier surface under the dynamic closed-loop control unit (described later). This minute gap avoids physical contact wear between the sensor and the pier surface, while not affecting the ability of high-frequency ultrasonic waves (5-10MHz) to penetrate ECC concrete over extremely short distances. Compared to traditional contact ultrasonic testing that requires the application of coupling agent, this solution achieves quasi-non-contact testing without coupling agent, ensuring both effective coupling of ultrasonic energy and meeting the engineering requirements for rapid scanning.

[0038] The multimodal sensing module 100 is the core detection unit of the device, integrating a high-frequency ultrasonic probe array and an electromagnetic induction sensor group. The high-frequency ultrasonic probe array uses 24 focused piezoelectric ceramic probes with a center frequency of 8MHz, arranged in a spiral pattern and fixed on a flexible polyimide substrate. This spiral arrangement better conforms to the curved surface of the bridge pier. The axial and circumferential spacing between adjacent probes is designed to be 20mm to ensure a continuous and blind-zone-free ultrasonic beam coverage area, specifically for penetrating and detecting the propagation of microcracks within the ECC material. The electromagnetic induction sensor group consists of 16 miniature planar coils with a diameter of 10mm. These coils are embedded in the gaps between the ultrasonic probes in an alternating arrangement, forming a spatially complementary layout. The excitation frequency of the coils is set to 100kHz, enabling sensitive sensing of minute changes in local electromagnetic impedance (especially phase) caused by debonding or microcracks at the interface between the steel pipe and the ECC. Through this synergistic arrangement of acoustic-electric heterogeneous sensors, simultaneous multi-physics sensing of defects within and at the interface of the composite material is achieved.

[0039] The adaptive positioning module 200 is responsible for driving the sensing module to perform precise and stable scanning motion on the pier surface. Its main body is a detachable, modular ring-shaped guide rail, with an inner layer covered with a high-coercivity neodymium iron boron magnetic material, allowing it to firmly adhere to the outer surface of the steel pier. It is suitable for piers with diameters from 800mm to 1500mm. A displacement servo motor is mounted on the guide rail, driving the sensing module to move circumferentially and axially along the pier. To overcome the influence of unevenness caused by rust and coating on the pier surface on non-contact detection, the module integrates a distance closed-loop control unit consisting of a laser rangefinder and a PID controller. The laser rangefinder 231 measures the distance between the sensor head and the pier surface in real time at a frequency of 1kHz and feeds the signal back to the PID controller 232 (parameters set as: proportional gain Kp = 1.2, integral time Ti = 0.05s, derivative time Td = 0.01s). The controller dynamically compensates for distance deviations by driving a carbon fiber telescopic arm 240 driven by a ball screw to adjust its length, thereby keeping the detection distance constant within the optimal range of (5±0.5) mm, ensuring the consistency of ultrasonic coupling and the signal-to-noise ratio of electromagnetic detection.

[0040] The signal fusion analysis module is integrated into the device's embedded industrial computer. It synchronously receives signals from the ultrasonic probe and electromagnetic coil via a high-speed data acquisition card (sampling rate 10MS / s, 14-bit resolution). The processing flow includes: first, bandpass filtering and timing alignment preprocessing of the raw signal; then, four-level wavelet packet decomposition based on the db6 wavelet basis is performed on the ultrasonic echo signal to extract the energy entropy of a specific frequency band as a damage-sensitive feature; simultaneously, the rate of change of the impedance phase angle of the electromagnetic signal is calculated. These two feature parameters are normalized, and a dual-parameter decision graph of "energy entropy - rate of change of phase angle" is generated in the same spatial coordinate system. The system presets a dynamic threshold; when both feature values ​​corresponding to a spatial coordinate point exceed the threshold, it is marked as an anomaly. If three or more consecutive adjacent sampling points are marked as anomalies, a potential crack is determined to exist in the area, triggering an alarm.

[0041] The detachable protective housing 400 is made of high-strength carbon fiber composite material, providing physical protection and electromagnetic shielding for the internal precision components. The housing integrates a temperature and humidity sensor and an anti-condensation heating element. When the system detects an ambient relative humidity greater than 80% or a temperature below 5°C, the heating element automatically activates, maintaining the internal temperature at least 3°C ​​above the dew point, effectively preventing condensation on the sensor surface and ensuring stable operation even in humid, low-temperature outdoor bridge environments. The bottom of the housing features a quick-locking buckle for easy on-site installation and disassembly, improving the equipment's engineering applicability and maintenance convenience.

[0042] The workflow is briefly described as follows: First, the device is magnetically fixed to the bottom of the bridge pier, and the initial scanning path is set (axial step distance 300mm, circumferential sampling points 36 / round). After startup, the adaptive positioning module 200 drives the multimodal sensing module 100 to move along the surface of the bridge pier, and the distance closed-loop control unit works in real time to maintain a constant lift-off distance. The sensing module 100 continuously collects data and wirelessly transmits it to the ground monitoring terminal or the analysis module 300 built into the device. The signal fusion analysis module 300 processes the data in real time. Once a suspected crack is detected in the coarse scan, the system can automatically instruct the positioning module to switch to a high-density scanning mode in the suspected area (axial step distance increased to 50mm, circumferential sampling points increased to 72 / round) and perform multi-angle ultrasonic detection to obtain three-dimensional information of the crack. Finally, the detection results are output in the form of a three-dimensional reconstructed image and a crack parameter report.

[0043] This embodiment significantly improves the detection rate, positioning accuracy, and evaluation objectivity of microcracks (with a width as low as 0.1 mm) at the interface of steel pipe-ECC bridge piers through a three-in-one design of multimodal collaborative sensing, adaptive constant distance scanning, and intelligent signal fusion, effectively overcoming the limitations of traditional single detection methods.

[0044] Example 2

[0045] Based on Example 1, this example further enhances the intelligent discrimination function of the signal fusion analysis module 300. In addition to the basic two-parameter threshold judgment, the system incorporates an offline-trained support vector machine (SVM) classifier model.

[0046] The training dataset for this classifier model comes from a large amount of laboratory simulation and finite element simulation data, containing more than 500 samples. The samples cover interface cracks with different characteristics: crack widths from 0.05 mm to 0.3 mm, depths from 1 mm to 10 mm, and tilt angles from 0° to 60°. Each sample corresponds to the "ultrasonic energy entropy feature sequence" and "electromagnetic phase angle change rate feature sequence" acquired by its multimodal sensing module, as well as its spatial coordinates.

[0047] In real-time detection, after the two-parameter decision map initially identifies potential crack areas, the system inputs the corresponding multi-dimensional feature vector (including feature values, spatial distribution patterns, signal evolution trends over time, etc.) into the loaded SVM classifier for fine-grained identification. The classifier outputs a three-level discrimination result for crack severity: "Minor" (width < 0.1 mm, no debonding), "Moderate" (width 0.1-0.2 mm, local debonding), and "Severe" (width > 0.2 mm or expansion of debonding area).

[0048] To verify the effectiveness of the SVM classifier, we conducted a comparative test. A mixed dataset of 520 samples from both real-world and simulated tests was constructed, with 70% used for training and 30% for blind testing. The three-level labels—"minor," "moderate," and "severe"—were calibrated based on high-precision industrial CT scan results. In the blind test set, the pure threshold method (relying solely on whether the two parameters exceed a fixed threshold) achieved a 78.2% accuracy rate in classifying crack severity into three levels. However, the SVM classifier trained using this embodiment (with an RBF kernel function, a grid search method to determine the penalty factor C=10, and kernel parameter γ=0.1) achieved a 91.4% accuracy rate on the same blind test set. Error analysis showed that the main misclassifications occurred in the boundary region between "minor" and "moderate" (crack width 0.1-0.12 mm), with an overall Kappa coefficient of 0.87, indicating a high degree of consistency in the classification results. This demonstrates that introducing a machine learning classifier significantly improves the intelligence and objectivity of the assessment.

[0049] Example 3

[0050] This embodiment focuses on an optimized hierarchical scanning detection method, which can be executed by the device described in Embodiment 1 or 2, and aims to achieve the optimal balance between detection efficiency and detail capture capability.

[0051] The method includes the following steps:

[0052] Step 1 (Coarse Scan Mode): Starting from the bottom of the pier, the device controls the adaptive positioning module to perform a layer-by-layer surround scan with a large axial step size (e.g., 300mm) to complete a rapid survey of the entire target section. This mode is designed to efficiently locate suspected defect areas.

[0053] Step 2 (Intelligent Judgment and Mode Switching): The signal fusion analysis module processes the coarse scan data in real time. Once a potential crack signal that meets the dual-modal anomaly coincidence criterion is identified in a certain scan layer (e.g., the Nth layer), the system immediately records the spatial coordinates of that layer.

[0054] Step 3 (Encrypted Scanning Mode): The system automatically switches the scanning mode to high-density encrypted scanning. The adaptive positioning module, within the Nth layer where anomalies have been detected and its adjacent vertical range of 150mm above and below (i.e., layers N-1, N, and N+1), densifies the axial scanning step size to 50mm. Simultaneously, within this encrypted area, the number of circumferential sampling points is increased from 36 points per revolution to 72 points (i.e., one sampling point every 5°).

[0055] Step Four (Multi-Angle Detection): During the encrypted scanning process, for abnormal areas, an additional high-frequency ultrasonic probe array is controlled to alternately emit and receive ultrasonic waves at different preset tilt angles (e.g., 0°, ±15°, ±30°). By comparing and analyzing the differences in ultrasonic echo signal amplitude and flight time under different incident angles, and combining this with the time-of-flight diffraction (TOFD) principle, the strike, dip angle, and depth information of the crack can be more accurately determined.

[0056] Step 5 (Result Synthesis and Output): The data acquired from all scanning modes (coarse scan, fine scan, multi-angle scan) are fused together. The algorithm of the signal fusion analysis module is used to reconstruct the three-dimensional morphology of the crack, and a complete inspection report including the crack location, length, depth, dip angle and severity level is output.

[0057] This hierarchical scanning strategy, through "global rapid positioning and local fine analysis," enables the comprehensive inspection time of a typical high bridge pier to be controlled within 90 minutes, while ensuring millimeter-level spatial resolution for key defect areas, thus achieving a good balance between efficiency and accuracy in the inspection task.

[0058] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A crack detection device for steel pipe-ECC concrete bridge piers, characterized in that, include: The multimodal sensing module integrates a high-frequency ultrasonic probe array and an electromagnetic induction sensor group. The high-frequency ultrasonic probe array consists of multiple focused piezoelectric ceramic probes arranged in a spiral pattern on a flexible substrate. The center frequency of the focused piezoelectric ceramic probes is 5–10 MHz, and the spacing between adjacent probes is 15–25 mm. The electromagnetic induction sensor group consists of multiple miniature planar coils arranged in an alternating manner among the spirally distributed probes. The diameter of a single miniature planar coil is 8–12 mm, and its excitation frequency is 50–200 kHz. The multimodal sensing module is used to perform synchronous non-contact scanning of the interior of the ECC concrete of the bridge pier and the steel pipe-ECC interface area to acquire ultrasonic echo signals and electromagnetic impedance change data, respectively. An adaptive positioning module includes an annular guide rail that can be arranged around the outer wall of a bridge pier, a displacement servo motor that drives the multimodal sensing module to move along the annular guide rail, and a distance closed-loop control unit for maintaining a constant detection distance between the multimodal sensing module and the surface of the bridge pier. A magnetic attachment layer is provided on the inner side of the annular guide rail. The distance closed-loop control unit includes a laser rangefinder and a PID controller. The laser rangefinder is used to measure the distance between the multimodal sensing module and the surface of the bridge pier in real time. The PID controller adjusts the length of a telescopic arm according to the distance signal to drive the multimodal sensing module to perform displacement compensation. The signal fusion analysis module is communicatively connected to the multimodal sensing module. The signal fusion analysis module is configured to: perform wavelet packet decomposition on the ultrasonic echo signal to extract energy entropy features; calculate the phase angle change rate of the electromagnetic impedance change data; and construct a spatial coordinate decision map with the energy entropy features and the phase angle change rate as two parameters; when both exceed a preset threshold in the two-parameter decision map and coincide in the spatial coordinates, a potential crack is determined to exist.

2. The steel pipe-ECC concrete bridge pier crack detection device according to claim 1, characterized in that, The signal fusion analysis module also has a built-in machine learning classifier, which is trained using a dataset containing simulated interface crack response patterns with different widths, depths and inclination angles, and is used to classify the severity of identified potential cracks.

3. The steel pipe-ECC concrete bridge pier crack detection device according to claim 1, characterized in that, It also includes a detachable protective shell made of carbon fiber composite material, which integrates a temperature and humidity sensor and an anti-condensation heating element. When the ambient humidity or temperature reaches a preset condition, the anti-condensation heating element is automatically activated.

4. A method for detecting cracks in steel pipe-ECC concrete bridge piers using the detection device described in any one of claims 1-3, characterized in that, Includes the following steps: S1: The device is installed on the outer surface of the bridge pier. The adaptive positioning module drives the multimodal sensing module to move along a preset path to scan the surface of the bridge pier. The preset path adopts a hierarchical scanning strategy. First, the bridge pier is coarsely scanned as a whole with a first preset step distance. When a suspected crack signal is found in the coarse scan, it automatically switches to a high-density encrypted scan around the suspected area with a second preset step distance smaller than the first preset step distance. S2: During the scanning process, the position of the multimodal sensing module is dynamically adjusted by the distance closed-loop control unit to keep the detection distance between the multimodal sensing module and the surface of the pier constant; S3: The ultrasonic echo signal and electromagnetic impedance change data of the scanning area are synchronously collected by the multimodal sensing module and transmitted to the signal fusion analysis module; during the high-density encrypted scanning, the high-frequency ultrasonic probe array is controlled to emit ultrasonic waves in a multi-angle tilt incident manner, and the inclination information of the crack is determined based on the acquired multi-angle echo signal. S4: The signal fusion analysis module extracts energy entropy features from the ultrasonic echo signal and phase angle change rate features from the electromagnetic impedance change data. It performs spatiotemporal alignment and normalization on the two features, identifies abnormal regions based on the dual-parameter joint criterion, and performs three-dimensional reconstruction and severity classification of cracks in the identified abnormal regions.

Images