A bridge pier column crack automatic detection system based on a wall-climbing robot

Abstract

The application discloses a kind of bridge pier column crack automatic detection system based on wall-climbing robot, it is related to crack detection technical field, the closed loop operation of accurate detection from crack to intelligent repair is realized by crack detection module and repair material intelligent spraying detection module.The crack detection module uses multispectral imaging equipment, thermal imaging camera and laser scanning technology, can collect the multispectral difference value G (x,y) of crack area, surface temperature distribution T (x,y) and depth matrix D (x,y), extract crack boundary B (x,y) by high-precision algorithm and calculate crack volume Vc.The repair material intelligent spraying detection module calculates the volume Vm of repair material according to crack volume, optimizes spraying path and parameter using SLAM technology, to ensure uniform coverage and efficient use of repair material.The system greatly improves the accuracy of crack detection and repair, avoids material waste and missed repair.

Description

Technical Field

[0001] This invention relates to the field of crack detection technology, specifically to an automatic crack detection system for bridge piers based on a wall-climbing robot. Background Technology

[0002] Bridge engineering is a crucial branch of civil engineering, with its core objective being to ensure the long-term safety and stability of bridge structures. During bridge maintenance, bridge piers, as key components supporting the bridge load, directly impact the bridge's service life and safety due to their structural condition. However, coastal bridge piers, constantly exposed to complex natural environments (such as high humidity, salt spray corrosion, and strong winds), are highly susceptible to structural defects like cracks. With the development of intelligent robotics technology, the automatic detection and repair of cracks in bridge piers using wall-climbing robots has gradually become a research hotspot in bridge engineering. Among these advancements, the dynamic environmental adaptation and adsorption adjustment mechanism provides crucial support for the stable operation of wall-climbing robots in complex environments.

[0003] Currently, traditional methods for detecting and repairing cracks in bridge piers mainly rely on manual operation, including high-altitude work or underwater inspection. These methods are inefficient, costly, and pose significant safety risks. Furthermore, the application of automated inspection equipment in complex environments also has shortcomings: many traditional wall-climbing devices struggle to maintain stable adhesion under high wind speeds or slippery surfaces, easily leading to the equipment detaching from the pier surface; during inspection and repair, the inability to adjust adhesion force and operating posture according to real-time environmental conditions severely limits the operating range and accuracy, impacting the efficiency and reliability of automated inspection and repair. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an automatic detection system for cracks in bridge piers based on a wall-climbing robot, which solves the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solution: including a working environment detection module, a detection posture adjustment module, a crack detection module, a repair material intelligent spraying detection module, and a repair analysis module;

[0006] The working environment detection module is used to install a detection equipment group on the wall-climbing robot to collect working environment data on the surface of the coastal bridge piers and transmit the collected working environment data to the ground control system of the wall-climbing robot. The working environment data is then preprocessed and features are extracted to obtain an environmental feature set.

[0007] The detection posture adjustment module evaluates the deviation between the actual pressure FPac and the ideal required pressure FPid at the adsorption point of the wall-climbing robot. When the adsorption state is found to be abnormal, the module calculates and outputs the adsorption force coefficient FP based on the environmental feature set, and then calculates and outputs the number of adsorption points N based on the adsorption force coefficient FP.

[0008] The crack detection module collects crack data of the coastal bridge piers through imaging equipment, extracts the crack boundary B(x,y) through image processing algorithm to identify and locate the cracks in the coastal bridge piers, and collects the depth distribution data of the crack area through integrated scanning technology after location, and calculates and outputs the total crack volume Vc.

[0009] The intelligent spraying and detection module for repair materials calculates the volume of repair material Vm in real time based on the total crack volume Vc, monitors the repair effect through imaging, and outputs the repair integrity error Er.

[0010] The repair analysis module evaluates the repair error based on the output of the repair integrity error Er, analyzes the repair effect before and after the repair, and generates early warning information based on the repair error evaluation results.

[0011] Preferably, the working environment detection module includes a data acquisition unit, a preprocessing unit, and a feature extraction unit;

[0012] The data acquisition unit is used to install a detection equipment group on the wall-climbing robot to collect real-time working environment data of the wall-climbing robot when it is working on the surface of the bridge piers on the coast.

[0013] The testing equipment group includes an ultrasonic anemometer, a laser roughness measuring instrument, an infrared humidity sensor, a conductivity sensor, a barometer, and a strain gauge pressure sensor.

[0014] The operational environment data includes wind speed v, wind direction angle θ, surface roughness coefficient Cr, surface humidity H, salt concentration S, pressure change rate EPR, and actual pressure at the adsorption point FPac.

[0015] The preprocessing unit is used to integrate the collected working environment data into the central processing unit of the wall-climbing robot, connect the ground control system to the central processing unit of the wall-climbing robot via a wireless network, and transmit the real-time collected working environment data to the ground control system.

[0016] The collected operational environment data is preprocessed in the ground control system. The preprocessing methods include data cleaning and normalization to unify the dimensions of the operational environment data.

[0017] Preferably, the feature extraction unit preprocesses the work environment data and then extracts features from the wind speed v, wind direction angle θ, surface roughness coefficient Cr, surface humidity H, and salt concentration S in the work environment data to obtain an environmental feature set.

[0018] The environmental feature set includes wind shear force (WF), surface adhesion factor (SAF), pressure change rate (EPR), and actual adsorption point pressure (FPac).

[0019] The wind shear force WF is calculated and extracted from the wind speed v and wind direction angle θ in the working environment data. The specific algorithm formula is as follows: In the formula, ρ represents the air density, the initial setting value, and cos represents the cosine function;

[0020] The surface adhesion factor (SAF) is calculated and extracted by considering the surface roughness coefficient (Cr), surface humidity (H), and salt concentration (S). The specific algorithm formula is as follows: In the formula, k represents the humidity sensitivity coefficient, and e represents the exponential function.

[0021] Preferably, the detection posture adjustment module includes an adsorption force analysis unit, an adsorption force adjustment unit, and an adsorption point distribution optimization unit;

[0022] The adsorption force analysis unit extracts the actual pressure FPac of the adsorption point in the environmental feature set and compares it with the ideal demand pressure FPid initially set by the user to evaluate the deviation and analyze the adsorption force of the current wall-climbing robot during operation. The specific evaluation content is as follows.

[0023] When the actual adsorption pressure FPac - the ideal required pressure FPid ≥ F1, the adsorption state is considered normal and no adjustment is needed.

[0024] When the actual adsorption pressure FPac - the ideal required pressure FPid < F1, the adsorption state is determined to be abnormal and the adsorption state needs to be adjusted.

[0025] F1 represents the allowable deviation value, which is initially set by the user.

[0026] Preferably, the adsorption force adjustment unit is used to trigger when the adsorption state is determined to be abnormal. By extracting the wind shear force WF and surface adhesion factor SAF from the environmental feature set and combining them with the pressure change rate EPR, the adsorption force coefficient FP is calculated and output, and the rotor motor power is dynamically adjusted to optimize the adsorption pressure of the wall-climbing robot.

[0027] The adsorption force coefficient FP is calculated and output using the following algorithm formula;

[0028]

[0029] In the formula, γ represents the adjustment factor, η represents the adhesion efficiency coefficient, δ represents the dynamic environmental influence coefficient, and sin(φ) represents the inclination angle of the surface of the coastal bridge pier. This represents the inverse proportional term of the surface adhesion factor;

[0030] The adsorption point distribution optimization unit dynamically adjusts the adsorption force of each adsorption point of the wall-climbing robot by using the adsorption force coefficient FP. Then, it calculates and outputs the number of adsorption points N by combining the center of gravity and working position of the wall-climbing robot, and dynamically distributes the adsorption points.

[0031] The number of adsorption points N is calculated and output using the following algorithm formula;

[0032]

[0033] In the formula, Ftotal represents the load-bearing capacity of the wall-climbing robot, m represents the total number of adhesion points, and FP i d represents the adjusted adsorption force coefficient at the i-th adsorption point. i This represents the weight correction value resulting from the distance of the i-th adsorption point relative to the robot's center of gravity.

[0034] Preferably, the crack detection module includes a crack identification unit and a crack volume analysis unit;

[0035] The crack identification unit adjusts the suction force of the wall-climbing robot and then starts the wall-climbing robot to move from bottom to top on the pier of the coastal bridge. During the movement, the imaging equipment installed on the wall-climbing robot collects crack data on the surface of the pier of the coastal bridge.

[0036] The imaging equipment includes a visible light camera, a thermal imaging camera, and a multispectral imaging device;

[0037] The crack data includes multispectral difference values ​​G(x,y), surface temperature distribution T(x,y), and crack region images;

[0038] The edge detection algorithm is used to calculate the pixel intensity gradient ▽I(x,y) of the crack area image. Then, based on the multispectral difference value G(x,y) and the surface temperature distribution T(x,y), the feature response of the crack area is calculated to form the crack boundary B(x,y). The crack location of the seaside bridge pier is identified and located. After the crack location is identified, the climbing robot automatically stops moving.

[0039] The crack boundary B(x, y) is calculated and output using the following algorithm formula;

[0040] B(x, y)=argmax[▽I(x, y)+k·G(x, y)·cos(λ·T(x, y))];

[0041] In the formula, argmax represents the upper limit function, (x, y) represents the coordinates in the two-dimensional image of the coastal bridge pier, k represents the multispectral feature enhancement factor, and λ represents the temperature difference sensitivity coefficient.

[0042] Preferably, the crack volume analysis unit is used to capture the depth matrix D(x, y) of the crack area by using the laser scanning technology of the wall-climbing robot after identifying and locating the crack in the pier of the coastal bridge, supplement the three-dimensional data of the crack area by using multi-time image reconstruction technology, and then calculate the total crack volume Vc by integration and discrete summation.

[0043] The total crack volume Vc is calculated and output using the following algorithm formula;

[0044]

[0045] In the formula, The crack boundary range is represented by the crack boundary B(x, y), dx represents a small change in the x-direction, dy represents a small change in the y-direction, and dxdy represents the area of ​​the small matrix region.

[0046] Preferably, the intelligent spraying and detection module for repair materials includes a repair material volume analysis unit and a repair integrity analysis unit;

[0047] The repair material volume analysis unit calculates and outputs the repair material volume Vm based on the obtained total crack volume Vc, adjusts the output parameters of the spraying system in real time, and then uses SLAM synchronous positioning and mapping technology to dynamically update the spraying trajectory and optimize the spraying path based on the three-dimensional data of the crack area.

[0048] The volume Vm of the repair material is calculated and output using the following algorithm formula;

[0049] Vm=π·Vc·R+α·sin(ξ)·ΔP;

[0050] In the formula, π represents the circumference ratio, with a value of 3.14, R represents the coverage ratio of the repair material, α represents the additional material compensation coefficient, sin(ξ) represents the spraying angle correction term, ξ represents the angle between the surface where the crack is located and the spraying equipment, and △P represents the spraying pressure difference.

[0051] Preferably, the repair integrity analysis unit scans the repaired area with a thermal imaging camera after the crack is repaired, records the actual temperature distribution Tr(x,y) after repair, and calculates the difference between the actual temperature and the ideal temperature to obtain the repair integrity error Er.

[0052] The repair integrity error Er is calculated and output using the following algorithm formula;

[0053]

[0054] In the formula, A represents the area of ​​the crack region, and Ts(x, y) represents the ideal temperature of the repair region.

[0055] Preferably, the specific repair error assessment content of the repair analysis module is as follows;

[0056] When the repair integrity error Er > 0, it indicates that the repair is incomplete, and a repair spraying adjustment warning is generated, which prompts the engineer to perform repair operations through the ground control system;

[0057] When the repair integrity error Er ≤ 0, it indicates that the repair quality is qualified and no warning needs to be generated.

[0058] This invention provides an automatic crack detection system for bridge piers based on a wall-climbing robot. It has the following advantages:

[0059] (1) This system comprehensively enhances the adaptability of the wall-climbing robot to complex coastal bridge pier working environments through the operation environment detection module and the detection posture adjustment module. The operation environment detection module uses devices such as ultrasonic wind speed sensors, laser roughness measuring instruments, and infrared humidity sensors to collect environmental data such as wind speed, humidity, salt concentration, and surface roughness in real time, and extracts feature sets after normalizing these data. Based on the feature sets, the detection posture adjustment module dynamically calculates the adsorption force coefficient FP and the number of adsorption points N, and adjusts the adsorption pressure and point distribution in real time. Compared with the traditional fixed adsorption mode, this system can significantly reduce the adsorption instability caused by slippery surfaces and strong wind interference, and effectively improve the reliability and safety of the robot's operation under high humidity, high salt spray, and complex surface conditions.

[0060] (2) This system achieves a closed-loop operation from accurate crack detection to intelligent repair through a crack detection module and a repair material intelligent spraying and detection module. The crack detection module uses multispectral imaging equipment, thermal imaging cameras, and laser scanning technology to collect multispectral difference values ​​G(x,y), surface temperature distribution T(x,y), and depth matrix D(x,y) of the crack area. It extracts the crack boundary B(x,y) and calculates the crack volume Vc using a high-precision algorithm. The repair material intelligent spraying and detection module dynamically calculates the repair material volume Vm based on the crack volume and optimizes the spraying path and parameters using SLAM technology to ensure uniform coverage and efficient utilization of the repair material. Compared with the traditional method of manual inspection and fixed spraying path, this system greatly improves the accuracy of crack detection and repair, avoids material waste and missed repairs, and ensures the long-term stability of the bridge structure.

[0061] (3) The system achieves real-time monitoring and closed-loop optimization of repair quality through a repair integrity analysis module. This module uses a thermal imaging camera to scan the repaired crack area, collects the actual temperature distribution Tr(x,y), and compares it with the ideal temperature Ts(x,y) of the repaired area to calculate the repair integrity error Er. When Er > 0, the system automatically generates a repair adjustment warning, prompting engineers to perform repair operations through the ground control system; when Er ≤ 0, the repair quality is deemed acceptable, and no further operation is required. Through this mechanism, the system effectively solves the drawbacks of traditional repair operations, such as the inability to verify quality or the need for repeated repairs, ensuring complete repair of the crack area, greatly improving the durability and safety of bridge piers, and providing intelligent support for bridge maintenance and management. Attached Figure Description

[0062] Figure 1 This is a schematic diagram of the automatic detection system for cracks in bridge piers based on a wall-climbing robot, according to the present invention. Detailed Implementation

[0063] 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0064] Example 1

[0065] Please see Figure 1 This invention provides an automatic crack detection system for bridge piers based on a wall-climbing robot. To achieve the above objectives, this invention is implemented through the following technical solutions: including a working environment detection module, a detection posture adjustment module, a crack detection module, a repair material intelligent spraying detection module, and a repair analysis module.

[0066] The working environment detection module is used to install a detection equipment group on the wall-climbing robot to collect working environment data on the surface of the coastal bridge piers and transmit the collected working environment data to the ground control system of the wall-climbing robot. The working environment data is then preprocessed and features are extracted to obtain an environmental feature set.

[0067] The attitude adjustment detection module assesses the deviation between the actual pressure FPac and the ideal required pressure FPid at the adsorption point of the wall-climbing robot. When the adsorption state is found to be abnormal, the module calculates and outputs the adsorption force coefficient FP based on the environmental feature set, and then calculates and outputs the number of adsorption points N based on the adsorption force coefficient FP.

[0068] The crack detection module collects crack data of the coastal bridge piers through imaging equipment, and then uses image processing algorithms to extract the crack boundary B(x,y) to identify and locate the cracks in the coastal bridge piers. After location, it collects the depth distribution data of the crack area through integrated scanning technology and calculates and outputs the total crack volume Vc.

[0069] The intelligent spraying and detection module for repair materials calculates the volume of repair material Vm in real time based on the total crack volume Vc, and detects the repair effect through imaging, outputting the repair integrity error Er.

[0070] The repair analysis module evaluates the repair error based on the output of the repair integrity error Er, analyzes the repair effect before and after repair, and generates early warning information based on the repair error evaluation results.

[0071] In this embodiment, the system utilizes a working environment detection module to collect real-time environmental data from the surface of the coastal bridge piers using multiple sensor devices. This data is then extracted to form an environmental feature set, providing a precise basis for subsequent attitude adjustment and operational decisions. The attitude adjustment module dynamically adjusts the adsorption force coefficient FP and the number of adsorption points N based on the real-time monitored deviation between the actual pressure FPac and the ideal pressure FPid at the adsorption point, ensuring stable adsorption and safe operation of the robot in complex environments. The crack detection module uses imaging equipment combined with image processing algorithms to extract the crack boundary B(x,y) with high precision and calculates the total crack volume Vc through depth scanning, achieving accurate crack identification and 3D modeling. The intelligent spraying detection module for repair materials adjusts the repair material volume Vm and spraying path in real-time based on the crack volume Vc, ensuring uniform material coverage of the crack area. Simultaneously, it monitors the repair effect using imaging equipment and outputs the repair integrity error Er. The repair analysis module generates a repair warning based on the evaluation result of Er, prompting for respraying adjustments or confirming the repair is qualified. This system, through the organic integration of the five modules mentioned above, achieves fully automated operation of bridge pier crack detection and repair, offering the following significant advantages: First, it significantly enhances the robot's adaptability in complex working environments, ensuring operational stability and safety; second, it improves the accuracy of crack detection and repair, avoiding omissions or material waste; and third, through closed-loop control of repair quality, it effectively guarantees the reliability and durability of the repairs. The application of this system has greatly improved the efficiency and quality of bridge maintenance work, reduced the safety risks of manual operation, and provided intelligent protection for the long-term safe operation of bridges.

[0072] Example 2

[0073] This embodiment is an explanation based on Embodiment 1. Please refer to it. Figure 1 Specifically, the working environment detection module includes a data acquisition unit, a preprocessing unit, and a feature extraction unit;

[0074] The data acquisition unit is used to install a detection equipment group on the wall-climbing robot to collect real-time operating environment data of the wall-climbing robot when it is working on the surface of the bridge piers on the coast.

[0075] The testing equipment group includes an ultrasonic anemometer, a laser roughness meter, an infrared humidity sensor, a conductivity sensor, a barometer, and a strain gauge pressure sensor.

[0076] The operational environment data includes wind speed v, wind direction angle θ, surface roughness coefficient Cr, surface humidity H, salt concentration S, pressure change rate EPR, and actual pressure at the adsorption point FPac;

[0077] Wind speed v and wind direction angle θ are measured by using an ultrasonic wind speed sensor installed on the top of the wall-climbing robot to measure wind speed and wind direction by utilizing the change in the speed at which ultrasonic waves propagate in the air.

[0078] The surface roughness coefficient Cr was obtained by scanning the surface of the bridge pier with a laser roughness measuring instrument to obtain the micro-geometric characteristics of the surface.

[0079] Surface humidity H, salt concentration S, pressure change rate EPR, and actual adsorption point pressure FPac were directly acquired using an infrared humidity sensor, a conductivity sensor, a barometer, and a strain gauge pressure sensor, respectively.

[0080] The preprocessing unit is used to integrate the collected working environment data into the central processing unit of the wall-climbing robot, and to connect the ground control system with the central processing unit of the wall-climbing robot via a wireless network to transmit the real-time collected working environment data to the ground control system.

[0081] The collected operational environment data is preprocessed in the ground control system. The preprocessing methods include data cleaning and normalization to unify the dimensions of the operational environment data.

[0082] The feature extraction unit preprocesses the operational environment data and then extracts features from the wind speed v, wind direction angle θ, surface roughness coefficient Cr, surface humidity H, and salt concentration S to obtain an environmental feature set.

[0083] The environmental feature set includes wind shear force (WF), surface adhesion factor (SAF), pressure change rate (EPR), and actual adsorption point pressure (FPac).

[0084] Wind shear force (WF) is calculated and extracted from the wind speed (v) and wind direction angle (θ) data in the working environment. The specific algorithm formula is as follows: In the formula, ρ represents the air density, the initial setting value, and cos represents the cosine function;

[0085] The surface adhesion factor (SAF) is calculated and extracted by considering the surface roughness coefficient (Cr), surface humidity (H), and salt concentration (S). The specific algorithm formula is as follows: In the formula, k represents the humidity sensitivity coefficient, and e represents the exponential function.

[0086] In this embodiment, the system comprises a data acquisition unit, a preprocessing unit, and a feature extraction unit through an operational environment detection module, comprehensively realizing the acquisition, processing, and feature extraction of complex environmental parameters. The data acquisition unit collects operational environment data from the surface of coastal bridge piers using a detection equipment group. This data is transmitted to the ground control system via a wireless network. The preprocessing unit cleans and normalizes the data, unifying dimensions to ensure standardization and accuracy. The feature extraction unit then forms a complete set of environmental features based on the environmental data, providing precise basis for adjusting the wall-climbing robot's adhesion and executing tasks. Through the design and implementation of this module, the system achieves the following goals and effects: First, it realizes multi-dimensional, real-time, and high-precision monitoring of complex environments, ensuring the comprehensiveness and reliability of the data; second, the feature extraction unit transforms complex raw environmental data into core feature parameters, improving the efficiency of subsequent analysis and decision-making. Supported by this module, the wall-climbing robot can operate stably in coastal environments with high wind speeds, complex surface roughness, and high humidity and salt spray concentrations, significantly improving the system's environmental adaptability and operational stability. This not only reduces the risk of detachment due to insufficient adhesion, but also lays a solid data foundation for bridge crack detection and repair, achieving a comprehensive improvement in work quality and efficiency.

[0087] Example 3

[0088] This embodiment is an explanation based on Embodiment 2. Please refer to it. Figure 1 Specifically: the detection attitude adjustment module includes an adsorption force analysis unit, an adsorption force adjustment unit, and an adsorption point distribution optimization unit;

[0089] The adsorption force analysis unit extracts the actual pressure FPac of the adsorption point in the environmental feature set and compares it with the ideal demand pressure FPid initially set by the user to evaluate the deviation and analyze the adsorption force of the current wall-climbing robot during operation. The specific evaluation content is as follows.

[0090] When the actual adsorption pressure FPac - the ideal required pressure FPid ≥ F1, the adsorption state is considered normal and no adjustment is needed.

[0091] When the actual adsorption pressure FPac - the ideal required pressure FPid < F1, the adsorption state is determined to be abnormal and the adsorption state needs to be adjusted.

[0092] F1 represents the allowable deviation value, which is initially set by the user.

[0093] The adsorption force adjustment unit is triggered when an abnormal adsorption state is detected. It calculates and outputs the adsorption force coefficient FP by extracting the wind shear force WF and surface adhesion factor SAF from the environmental feature set and combining them with the pressure change rate EPR. It then dynamically adjusts the rotor motor power to optimize the adsorption pressure of the wall-climbing robot.

[0094] The adsorption force coefficient FP is calculated and output using the following algorithm formula;

[0095]

[0096] In the formula, γ represents the adjustment factor, used to correct the adsorption force to adapt to task requirements, such as different load weights or environmental changes; η represents the adhesion efficiency coefficient, which depends on the adhesion surface material; δ represents the dynamic environmental influence coefficient; and sin(φ) represents the inclination angle of the surface of the coastal bridge pier. This represents the inverse term of the surface adhesion factor. When the surface roughness of the coastal bridge pier is high, the humidity is low, and the salt concentration is low, the surface adhesion is stronger and the adsorption force requirement is reduced.

[0097] The adsorption point distribution optimization unit dynamically adjusts the adsorption force of each adsorption point of the wall-climbing robot by using the adsorption force coefficient FP. Then, it calculates and outputs the number of adsorption points N by combining the center of gravity and working position of the wall-climbing robot, and dynamically distributes the adsorption points.

[0098] The number of adsorption sites N is calculated and output using the following algorithm formula;

[0099]

[0100] In the formula, Ftotal represents the load-bearing capacity of the wall-climbing robot, represents the total weight of the wall-climbing robot and the gravitational component of the task applied to the adhesion points, m represents the total number of adhesion points, and FP... i d represents the adjusted adsorption force coefficient at the i-th adsorption point. i The weight correction value represents the distance of the i-th adsorption point relative to the robot's center of gravity. The closer the adsorption point is to the center of gravity, the smaller its weight correction factor is, and the lower the torque it provides. The farther the adsorption point is from the center of gravity, the larger the weight correction factor is, and the more significant its contribution to stability. The correction factor is used to adjust the distribution of adsorption points, so that the torque effect is more uniform, and to prevent the robot from tipping over or slipping due to the shift of the center of gravity.

[0101] In this embodiment, the system utilizes a posture adjustment module comprised of an adsorption force analysis unit, an adsorption force adjustment unit, and an adsorption point distribution optimization unit. Together, these components enable the wall-climbing robot to achieve stable adhesion and dynamic posture adjustment in complex working environments. The adsorption force analysis unit extracts the actual pressure FPac and the ideal required pressure FPid from the adsorption points within the environmental feature set to assess deviations. When insufficient adsorption pressure is detected, the adsorption force adjustment unit calculates the adsorption force coefficient FP and dynamically adjusts the rotor motor power to optimize the adsorption pressure. Subsequently, the adsorption point distribution optimization unit combines the adjusted adsorption force coefficient FP, the robot's center of gravity position, and its working position to dynamically calculate the number of adsorption points N and optimize their distribution, ensuring uniform adsorption force and stable robot posture. Through this module, the wall-climbing robot can adapt in real-time to multiple environmental disturbances such as wind, high humidity, salt spray concentration, and surface tilt on complex surfaces like bridge piers. The implementation of this module achieved the following goals and effects: First, through dynamic adsorption force adjustment and optimization, the robot can maintain stable attachment in complex environments, significantly reducing the risk of detachment due to insufficient adsorption force; Second, through optimization of adsorption point distribution, the robot's center of gravity balance and posture stability were improved, enabling it to move safely and efficiently on complex surfaces.

[0102] Example 4

[0103] This embodiment is an explanation based on Embodiment 3. Please refer to it. Figure 1 Specifically: the crack detection module includes a crack identification unit and a crack volume analysis unit;

[0104] After adjusting the suction force of the wall-climbing robot, the crack identification unit starts the wall-climbing robot to move from bottom to top on the pier of the coastal bridge. During the movement, the imaging equipment installed on the wall-climbing robot collects crack data on the surface of the pier of the coastal bridge.

[0105] Imaging equipment includes visible light cameras, thermal imaging cameras, and multispectral imaging equipment;

[0106] The crack data includes multispectral difference values ​​G(x,y), surface temperature distribution T(x,y), and images of the crack region;

[0107] The edge detection algorithm is used to calculate the pixel intensity gradient ▽I(x,y) of the crack area image. Then, based on the multispectral difference value G(x,y) and the surface temperature distribution T(x,y), the feature response of the crack area is calculated to form the crack boundary B(x,y). The crack location of the seaside bridge pier is identified and located. After the crack location is identified, the climbing robot automatically stops moving.

[0108] The crack boundary B(x, y) is calculated and output using the following algorithm formula;

[0109] B(x, y)=argmax[▽I(x, y)+k·G(x, y)·cos(λ·T(x, y))];

[0110] In the formula, argmax represents the upper limit function, (x, y) represents the coordinates in the two-dimensional image of the coastal bridge pier, k represents the multispectral feature enhancement factor, which is used to adjust the intensity of the multispectral difference value and to optimize the recognition effect, and λ represents the temperature difference sensitivity coefficient, which reflects the influence weight of temperature distribution on crack recognition and depends on the thermal properties of the material.

[0111] The significance of the formula lies in: accurately identifying the boundary location of cracks, avoiding misjudgments due to surface stains or reflections, and supplementing the limitations of single-spectral images with multispectral and temperature difference information to improve the recognition accuracy.

[0112] The crack volume analysis unit is used to capture the depth matrix D(x,y) of the crack area through the laser scanning technology of the wall-climbing robot after the cracks of the coastal bridge piers are identified and located. The three-dimensional data of the crack area is supplemented by multi-time image reconstruction technology, and the total crack volume Vc is calculated by integration and discrete summation.

[0113] The total crack volume Vc is calculated and output using the following algorithm formula;

[0114]

[0115] In the formula, The crack boundary range is represented by the crack boundary B(x, y), dx represents a small change in the x-direction, dy represents a small change in the y-direction, and dxdy represents the area of ​​the small matrix region.

[0116] In this embodiment, the system comprises a crack identification unit and a crack volume analysis unit through a crack detection module, comprehensively realizing accurate identification and 3D modeling of cracks in bridge piers. The crack identification unit initiates detection after adjusting the adhesion force of the climbing robot, moving from the bottom to the top of the pier. It uses imaging equipment (including a visible light camera, a thermal imaging camera, and a multispectral imaging device) onboard the robot to collect crack data in real time. Combined with an edge detection algorithm, it calculates the pixel intensity gradient ▽I(x,y) of the crack area image and calculates the crack feature response based on the multispectral difference value G(x,y) and the surface temperature distribution T(x,y), generating the crack boundary B(x,y). After crack identification is complete, the robot automatically stops moving to avoid duplicate detection or omissions. The crack volume analysis unit, based on the identification results, captures the crack depth matrix D(x,y) using laser scanning technology, generates 3D data of the crack area using multi-temporal image reconstruction technology, and finally calculates the total crack volume Vc through integration and discrete summation. This module achieves the following goals and effects through advanced imaging and data processing technologies: First, by fusing multispectral and thermal imaging data, it significantly improves the accuracy of crack identification and avoids misjudgments caused by surface stains, light reflection, or complex textures; Second, by combining laser scanning and 3D modeling technologies, it can efficiently and comprehensively capture the depth and volume information of cracks, providing accurate crack parameters for subsequent repairs.

[0117] Example 5

[0118] This embodiment is an explanation based on Embodiment 4. Please refer to it. Figure 1 Specifically: the intelligent spraying and detection module for repair materials includes a repair material volume analysis unit and a repair integrity analysis unit;

[0119] The repair material volume analysis unit calculates and outputs the repair material volume Vm based on the obtained total crack volume Vc, adjusts the output parameters of the spraying system in real time, and then uses SLAM synchronous positioning and mapping technology to dynamically update the spraying trajectory and optimize the spraying path based on the three-dimensional data of the crack area to ensure uniform coverage of the crack area.

[0120] The repair material volume Vm is calculated and output using the following algorithm formula;

[0121] Vm=π·Vc·R+α·sin(ξ)·ΔP;

[0122] In the formula, π represents the mathematical constant pi, with a value of 3.14; R represents the coverage ratio of the repair material, where R > 1 indicates complete material coverage; α represents the additional material compensation coefficient, used to compensate for the loss of sprayed material, such as the decrease in adhesion caused by wind speed or surface roughness; sin(ξ) represents the spraying angle correction term, where ξ represents the angle between the surface where the crack is located and the spraying equipment, which is obtained directly through the spraying system; and ΔP represents the spraying pressure difference, which is obtained directly through the spraying system.

[0123] The repair integrity analysis unit scans the repaired area with a thermal imaging camera after crack repair, records the actual temperature distribution Tr(x,y) after repair, and calculates the difference between the actual temperature and the ideal temperature to obtain the repair integrity error Er.

[0124] The repair integrity error Er is calculated and output using the following algorithm formula;

[0125]

[0126] In the formula, A represents the area of ​​the crack region. The crack outline boundary is calculated by the crack detection module, and the total area is obtained. Ts(x,y) represents the ideal temperature of the repair area.

[0127] In this embodiment, the system comprises a repair material volume analysis unit and a repair integrity analysis unit through an intelligent spraying and detection module for repair materials. This module comprehensively realizes accurate calculation of repair materials, dynamic spraying, and real-time feedback of repair effects. The repair material volume analysis unit calculates the required material volume Vm based on the total crack volume Vc output by the crack detection module. Simultaneously, it uses SLAM (Simultaneous Localization and Mapping) technology to update the spraying path in real time, combining the three-dimensional data of the crack area to ensure uniform material coverage of the crack area. To address material loss during spraying, this unit dynamically adjusts the output parameters of the spraying system to achieve optimal material utilization. The repair integrity analysis unit scans the repaired area using a thermal imaging camera, records the actual temperature distribution Tr(x,y), and calculates the difference between it and the ideal temperature distribution Ts(x,y), outputting the repair integrity error Er to evaluate the repair effect in real time. When Er > 0, the system generates a respray warning to ensure complete crack repair. The implementation of this module achieved the following goals and effects: First, through precise calculation and dynamic spraying path optimization, the utilization rate of repair materials was significantly improved, avoiding material waste and omissions in coverage; second, through repair integrity analysis, closed-loop monitoring of repair effects was achieved, effectively ensuring the integrity and durability of the repaired area; finally, through intelligent spraying adjustment and feedback, the system can adapt to the crack repair needs in complex environments, ensuring the quality of the operation.

[0128] Example 6

[0129] This embodiment is an explanation based on Embodiment 2. Please refer to it. Figure 1 Specifically, the repair error assessment content of the repair analysis module is as follows;

[0130] When the repair integrity error Er > 0, it indicates that the repair is incomplete, and a repair spraying adjustment warning is generated, which prompts the engineer to perform repair operations through the ground control system;

[0131] When the repair integrity error Er ≤ 0, it indicates that the repair quality is qualified and no warning needs to be generated.

[0132] In this embodiment, the system establishes an intelligent monitoring and feedback mechanism for repair quality based on real-time evaluation of the repair integrity error Er through the repair analysis module. The system judges the repair effect based on the error Er output by the repair integrity analysis unit: when Er > 0, it indicates incomplete repair, and the system automatically generates a warning for respray adjustment, prompting the ground control system to notify engineers to perform respray operations; when Er ≤ 0, it indicates that the repair quality is acceptable, and no further processing is required. This module uses repair integrity error as the core evaluation indicator, combined with thermal imaging data, boundary calculations from the crack detection module, and the crack area A, to comprehensively analyze the repair effect, providing intelligent closed-loop control for repair operations. The implementation of this module achieves the following goals and effects: First, by accurately judging the repair status, it avoids incomplete or substandard repairs, ensuring the integrity and durability of crack repairs; second, the automatically generated warning information significantly improves repair efficiency and reduces the workload of manual inspection; finally, through real-time monitoring and dynamic adjustment of the repair results, it improves the system's operational intelligence level.

[0133] 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.

Claims

1. An automatic detection system for cracks in bridge piers based on a wall-climbing robot, characterized in that: It includes a working environment detection module, a detection posture adjustment module, a crack detection module, a repair material intelligent spraying detection module, and a repair analysis module; The working environment detection module is used to install a detection equipment group on the wall-climbing robot to collect working environment data on the surface of the coastal bridge piers and transmit the collected working environment data to the ground control system of the wall-climbing robot. The working environment data is then preprocessed and features are extracted to obtain an environmental feature set. The detection posture adjustment module evaluates the deviation between the actual pressure FPac and the ideal required pressure FPid at the adsorption point of the wall-climbing robot. When the adsorption state is found to be abnormal, the module calculates and outputs the adsorption force coefficient FP based on the environmental feature set, and then calculates and outputs the number of adsorption points N based on the adsorption force coefficient FP. The crack detection module collects crack data of the coastal bridge piers through imaging equipment, extracts the crack boundary B(x,y) through image processing algorithm to identify and locate the cracks in the coastal bridge piers, and collects the depth distribution data of the crack area through integrated scanning technology after location, and calculates and outputs the total crack volume Vc. The intelligent spraying and detection module for repair materials calculates the volume of repair material Vm in real time based on the total crack volume Vc, monitors the repair effect through imaging, and outputs the repair integrity error Er. The repair analysis module evaluates the repair error based on the output of the repair integrity error Er, analyzes the repair effect before and after the repair, and generates early warning information based on the repair error evaluation results. The intelligent spraying and detection module for repair materials includes a repair material volume analysis unit and a repair integrity analysis unit. The repair material volume analysis unit calculates and outputs the repair material volume Vm based on the obtained total crack volume Vc, adjusts the output parameters of the spraying system in real time, and then uses SLAM synchronous positioning and mapping technology to dynamically update the spraying trajectory and optimize the spraying path based on the three-dimensional data of the crack area. The volume Vm of the repair material is calculated and output using the following algorithm formula; ； In the formula, R represents pi (π), with a value of 3.14, and R represents the coverage ratio of the repair material. This indicates the additional material compensation coefficient. This indicates the spraying angle correction option. The angle between the surface where the crack is located and the spraying equipment is represented, and △P represents the spraying pressure difference. The repair integrity analysis unit scans the repaired area with a thermal imaging camera after the crack is repaired, records the actual temperature distribution Tr(x,y) after repair, and calculates the difference between the actual temperature and the ideal temperature to obtain the repair integrity error Er. The repair integrity error Er is calculated and output using the following algorithm formula; ； In the formula, A represents the area of ​​the crack region, and Ts(x, y) represents the ideal temperature of the repair area; The specific repair error assessment content of the repair analysis module is as follows; When the repair integrity error Er > 0, it indicates that the repair is incomplete, and a repair spraying adjustment warning is generated, which prompts the engineer to perform repair operations through the ground control system; When the repair integrity error Er ≤ 0, it indicates that the repair quality is qualified and no warning needs to be generated.

2. The automatic detection system for bridge pier cracks based on a wall-climbing robot according to claim 1, characterized in that: The working environment detection module includes a data acquisition unit, a preprocessing unit, and a feature extraction unit; The data acquisition unit is used to install a detection equipment group on the wall-climbing robot to collect real-time working environment data of the wall-climbing robot when it is working on the surface of the bridge piers on the coast. The testing equipment group includes an ultrasonic anemometer, a laser roughness measuring instrument, an infrared humidity sensor, a conductivity sensor, a barometer, and a strain gauge pressure sensor. The operational environment data includes wind speed v and wind direction angle. Surface roughness coefficient Cr, surface humidity H, salt concentration S, pressure change rate EPR, and actual adsorption point pressure FPac; The preprocessing unit is used to integrate the collected working environment data into the central processing unit of the wall-climbing robot, connect the ground control system to the central processing unit of the wall-climbing robot via a wireless network, and transmit the real-time collected working environment data to the ground control system. The collected operational environment data is preprocessed in the ground control system. The preprocessing methods include data cleaning and normalization to unify the dimensions of the operational environment data.

3. The automatic detection system for bridge pier cracks based on a wall-climbing robot according to claim 2, characterized in that: The feature extraction unit preprocesses the operational environment data to extract parameters such as wind speed v and wind direction angle. The surface roughness coefficient Cr, surface humidity H, and salt concentration S are used to extract environmental feature sets. The environmental feature set includes wind shear force (WF), surface adhesion factor (SAF), pressure change rate (EPR), and actual adsorption point pressure (FPac). The wind shear force WF is obtained by extracting wind speed v and wind direction angle from the working environment data. The specific algorithm formula for the calculation and extraction is as follows: In the formula This represents the air density, the initial setting value, and cos represents the cosine function. The surface adhesion factor (SAF) is calculated and extracted by considering the surface roughness coefficient (Cr), surface humidity (H), and salt concentration (S). The specific algorithm formula is as follows: In the formula, k represents the humidity sensitivity coefficient, and e represents the exponential function.

4. The automatic detection system for bridge pier cracks based on a wall-climbing robot according to claim 1, characterized in that: The detection attitude adjustment module includes an adsorption force analysis unit, an adsorption force adjustment unit, and an adsorption point distribution optimization unit. The adsorption force analysis unit extracts the actual pressure FPac of the adsorption point in the environmental feature set and compares it with the ideal demand pressure FPid initially set by the user to evaluate the deviation and analyze the adsorption force of the current wall-climbing robot during operation. The specific evaluation content is as follows. When the actual adsorption pressure FPac - the ideal required pressure FPid ≥ F1, the adsorption state is considered normal and no adjustment is needed. When the actual adsorption pressure FPac - the ideal required pressure FPid < F1, the adsorption state is determined to be abnormal and the adsorption state needs to be adjusted. F1 represents the allowable deviation value, which is initially set by the user.

5. The automatic detection system for bridge pier cracks based on a wall-climbing robot according to claim 4, characterized in that: The adsorption force adjustment unit is used to trigger when an abnormal adsorption state is detected. It calculates and outputs the adsorption force coefficient FP by extracting the wind shear force WF and surface adhesion factor SAF from the environmental feature set and combining them with the pressure change rate EPR. It then dynamically adjusts the rotor motor power to optimize the adsorption pressure of the wall-climbing robot. The adsorption force coefficient FP is calculated and output using the following algorithm formula; ； In the formula, Indicates the adjustment factor. This represents the adhesion efficiency coefficient. Represents the dynamic environmental impact coefficient. Indicates the inclination angle of the surface of the pier column of a coastal bridge. This represents the inverse proportional term of the surface adhesion factor; The adsorption point distribution optimization unit dynamically adjusts the adsorption force of each adsorption point of the wall-climbing robot by using the adsorption force coefficient FP. Then, it calculates and outputs the number of adsorption points N by combining the center of gravity and working position of the wall-climbing robot, and dynamically distributes the adsorption points. The number of adsorption points N is calculated and output using the following algorithm formula; ； In the formula, Ftotal represents the load-bearing capacity of the wall-climbing robot, m represents the total number of adhesion points, and FP i d represents the adjusted adsorption force coefficient at the i-th adsorption point. i This represents the weight correction value resulting from the distance of the i-th adsorption point relative to the robot's center of gravity.

6. The automatic detection system for bridge pier cracks based on a wall-climbing robot according to claim 5, characterized in that: The crack detection module includes a crack identification unit and a crack volume analysis unit; The crack identification unit adjusts the suction force of the wall-climbing robot and then starts the wall-climbing robot to move from bottom to top on the pier of the coastal bridge. During the movement, the imaging equipment installed on the wall-climbing robot collects crack data on the surface of the pier of the coastal bridge. The imaging equipment includes a visible light camera, a thermal imaging camera, and a multispectral imaging device; The crack data includes multispectral difference values ​​G(x,y), surface temperature distribution T(x,y), and crack region images; The edge detection algorithm is used to calculate the pixel intensity gradient ▽I(x,y) of the crack area image. Then, based on the multispectral difference value G(x,y) and the surface temperature distribution T(x,y), the feature response of the crack area is calculated to form the crack boundary B(x,y). The crack location of the seaside bridge pier is identified and located. After the crack location is identified, the wall-climbing robot automatically stops moving. The crack boundary B(x, y) is calculated and output using the following algorithm formula; ； In the formula, argmax represents the upper bound function, (x, y) represents the coordinates in the two-dimensional image of the coastal bridge pier, and k represents the multispectral feature enhancement factor. This represents the temperature difference sensitivity coefficient.

7. The automatic detection system for bridge pier cracks based on a wall-climbing robot according to claim 6, characterized in that: The crack volume analysis unit is used to capture the depth matrix D(x,y) of the crack area through the laser scanning technology of the wall-climbing robot after identifying and locating the cracks in the piers of the coastal bridge. It then uses multi-time image reconstruction technology to supplement the three-dimensional data of the crack area and calculates the total crack volume Vc through integration and discrete summation. The total crack volume Vc is calculated and output using the following algorithm formula; ； In the formula, dx represents the extent of the crack boundary, dy represents the small change in the x-direction, and dxdy represents the small change in the y-direction. dxdy represents the area of ​​the small matrix region.

Images