Full-automatic intelligent ultrasonic detection system and method for cast-in-place pile

Abstract

The application provides a full-automatic intelligent ultrasonic detection system and method for a cast-in-place pile. The detection system comprises a walking device, a mechanical arm, a camera and a control system. The walking device is provided with an automatic winding and unwinding device for a guide wire. The mechanical arm is arranged on the walking device, and an ultrasonic detection device is movably arranged in a terminal guide slot. The guide wire is connected to the detection device through an internal passage. The camera is arranged on the mechanical arm and collects image data of a pile top. The control system identifies pixel coordinates of a pile core and a sounding pipe opening according to the image data, converts the pixel coordinates into physical coordinates, compares a deviation of the pile core to determine whether the walking device is in place, controls the mechanical arm to align after the walking device is in place, drives the winding and unwinding device to lower the guide wire, and enables the detection device to vertically drop into the sounding pipe under the cooperation of the guide slot and gravity. The full-automatic intelligent ultrasonic detection system and method for the cast-in-place pile can solve the problems of high dependence on manual operation, low efficiency and insufficient detection data quality in ultrasonic detection of the cast-in-place pile.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of cast-in-place pile ultrasonic testing, in particular to a full-automatic intelligent ultrasonic testing system and method for cast-in-place piles. BACKGROUND

[0002] With the rapid development of economy, the application of concrete cast-in-place piles in engineering construction is becoming more and more widespread. As an important part of safety quality detection of engineering construction, after the completion of construction, the integrity of the concrete cast-in-place piles needs to be tested by ultrasonic waves. Especially in large-scale construction and bridge engineering, a large number of cast-in-place piles are usually constructed at the same time, so the number of piles that need to be tested by ultrasonic waves is large, and the length of the cast-in-place piles is long, and the position and number of each pile need to be accurately matched with the design drawings.

[0003] The traditional ultrasonic testing method usually requires the tester to check the drawings to confirm the position and number of the tested pile on the drawings. During the ultrasonic testing, two testers are usually required, one of whom lowers and raises the ultrasonic transducer in the sound measuring pipe of the cast-in-place pile, and the other operates the data collector to collect data. After the test is completed, the ultrasonic transducer wire needs to be recovered and wound. The test process is time-consuming, and the testers are very tired due to the weather and construction site factors. Therefore, the traditional detection method has the following disadvantages: low efficiency of manual operation, high labor intensity, poor standardization, easy confusion of pile position information and interference with data collection quality, which affects the authenticity and effectiveness of the detection data.

[0004] At present, there is no effective solution to the problem of high dependence on manual operation, low efficiency and insufficient quality of detection data in related art. SUMMARY

[0005] The full-automatic intelligent ultrasonic testing system and method for cast-in-place piles provided by the embodiments of the present application at least solve the problem of high dependence on manual operation, low efficiency and insufficient quality of detection data in cast-in-place pile ultrasonic testing.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0007] The application provides a full-automatic intelligent ultrasonic detection system for a cast-in-place pile, which comprises a walking device provided with an automatic winding and unwinding device, a mechanical arm arranged on the walking device, a wire channel arranged in the mechanical arm, a guide slot arranged at the end of the mechanical arm and connected with the wire channel, an ultrasonic detection device movably arranged in the guide slot, an end of the wire connected with the ultrasonic detection device through the wire channel, a camera arranged on the mechanical arm and used for collecting pile top image data of each cast-in-place pile in real time, and a control system arranged on the walking device and used for identifying pixel coordinates of a pile core and a sounding pipe opening embedded in each cast-in-place pile according to the pile top image data, determining whether the walking device reaches a target position according to a deviation between physical coordinates converted from the pixel coordinates of the pile core and theoretical coordinates of the pile core in a pile drawing, and driving the mechanical arm to move the ultrasonic detection device to physical coordinates converted from the pixel coordinates of the sounding pipe opening and driving the automatic winding and unwinding device to lower the wire when the walking device moves to the target position, so that the ultrasonic detection device vertically enters the inside of a corresponding sounding pipe under the cooperation of the guide slot and self weight until the bottom.

[0008] Preferably, the control system comprises an image processing module connected with the camera, used for converting the colorful pile top image data into corresponding gray image data and extracting gradient amplitudes, a feature extraction module connected with the image processing module, used for extracting a top surface target contour of each cast-in-place pile and quantized features of each connected domain in the top surface target contour based on each gray image data and the gradient amplitudes, and a target recognition module connected with the feature extraction module, used for identifying a pile body region and a sounding pipe opening region of the cast-in-place pile from each connected domain based on the quantized features and determining pixel coordinates of a pile core and a sounding pipe opening of the corresponding cast-in-place pile based on center coordinates of the pile body region and the sounding pipe opening region.

[0009] Preferably, the feature extraction module further comprises an edge detection module connected with the image processing module, used for extracting a top surface target contour of each cast-in-place pile based on each gray image data and outputting a binary edge image with a single-pixel width, a repair module connected with the edge detection module, used for performing edge morphological closing operation on the binary edge image to obtain an edge image with a continuous and complete contour, and a contour feature quantization module connected with the repair module, used for performing connected domain analysis and quantization processing on the edge image to extract quantized features of each connected domain, wherein the quantized features comprise size features, position features, morphological features and gray features.

[0010] Preferably, the target recognition module comprises: a pile body recognition module, connected with the feature extraction module, configured to filter, based on the quantified features, a region with the largest area and an average gray value greater than a first set threshold from each of the connected domains, as a pile body candidate region; and a pile body determination module, connected with the pile body recognition module, configured to filter, from the pile body candidate region, a region with a circularity and an equivalent diameter size both satisfying a set range, as a pile body region, and determine a centroid of the pile body region as a pixel coordinate of the corresponding pile center.

[0011] Preferably, the target recognition module further comprises: a pipe mouth recognition module, connected with the feature extraction module, configured to filter, based on the quantified features, a region with an area satisfying a limit range and an average gray value less than a second set threshold from each of the connected domains, as an initial candidate set of pipe mouths; and a pipe mouth determination module, connected with the pipe mouth recognition module, configured to filter, from the initial candidate set of pipe mouths, a region with a circularity, an equivalent diameter, a center position and each center connection pattern all satisfying a set range, as a sounding pipe mouth region, and determine a center coordinate of the sounding pipe mouth region as a pixel coordinate of the corresponding sounding pipe mouth.

[0012] Preferably, the control system further comprises: an information extraction module, configured to extract, based on an input bored pile drawing, a theoretical coordinate of a pile center of each bored pile in the construction coordinate system; a coordinate conversion module, connected with the information extraction module and the target recognition module, configured to convert the pixel coordinates of the pile center and the sounding pipe mouth into corresponding physical coordinates in the construction coordinate system based on a Beidou coordinate of the walking device, a conversion matrix between the camera and the walking device, and an end effector of the mechanical arm; a deviation calculation module, connected with the information extraction module and the coordinate conversion module, configured to calculate a planar deviation distance between the physical coordinate of the pile center and the theoretical coordinate of the pile center; and a pile position confirmation module, connected with the deviation calculation module, configured to compare the planar deviation distance with a set deviation threshold, and confirm that the walking device reaches a target position when the planar deviation distance is less than or equal to the set deviation threshold.

[0013] Preferably, the control system further comprises a multi-source data acquisition module for acquiring encoder data of the automatic winding and unwinding device, ultrasonic detection data of the ultrasonic detection device, video stream data of the camera, and running state data of the walking device, the mechanical arm and the automatic winding and unwinding device; a display module connected with the multi-source data acquisition module for visually displaying theoretical parameters of each cast-in-place pile, a lowering progress of the ultrasonic detection device, the ultrasonic detection data, a video screen captured by the camera and the running state data; wherein the lowering progress is calculated based on the encoder data and a pile length in the theoretical data of the cast-in-place pile.

[0014] Preferably, the automatic winding and unwinding device comprises a winding drum, a servo wire winder and a servo motor; the winding drum is rotatably arranged on the walking device; the servo wire winder is arranged on the walking device and is arranged in parallel with the winding drum; the wire is wound on the winding drum after being wound by the servo wire winder; the servo motor is arranged on the walking device and is coaxially fixedly connected with a central shaft of the winding drum; the servo motor and the servo wire winder are connected with the control system.

[0015] Preferably, the ultrasonic detection device is arranged in at least two and the number thereof corresponds to the number of the mechanical arm and the automatic winding and unwinding device one by one; each ultrasonic detection device is connected with one automatic winding and unwinding device through one wire passing through a wire channel of the corresponding mechanical arm, forming a detection assembly.

[0016] Another aspect of the present application provides a full-automatic intelligent ultrasonic detection method for a bored pile, comprising the following steps: setting a route by a control system to control a walking device to move, and collecting pile top image data of each bored pile to be tested in real time; wherein the walking device is provided with an automatic winding and unwinding device, and the automatic winding and unwinding device is wound with a guide wire; based on the pile top image data, pixel coordinates of pile cores and sounding pipe openings of each bored pile are identified in real time, and the pixel coordinates are converted into corresponding physical coordinates; based on the deviation between the physical coordinates corresponding to the pile cores and the theoretical coordinates of the pile cores in a bored pile drawing, it is determined whether the walking device reaches a target position; in the case of reaching the target position, a mechanical arm is controlled to move the ultrasonic detection device to the physical coordinates corresponding to the sounding pipe opening; wherein the mechanical arm is arranged on the walking device; a wire channel is arranged in the mechanical arm, and a guide slot communicating with the wire channel is arranged at the end of the mechanical arm; an ultrasonic detection device is movably arranged in the guide slot; the end of the guide wire passes through the wire channel and is connected with the ultrasonic detection device; in the case of reaching the physical coordinates corresponding to the sounding pipe opening, the automatic winding and unwinding device is controlled to lower the guide wire, so that the ultrasonic detection device vertically enters the inside of the corresponding sounding pipe under the synergistic action of the guide slot and the self-weight until the bottom is touched.

[0017] The above technical solution of the present application has the following beneficial effects compared with the prior art:

[0018] The embodiment of the application provides a full-automatic intelligent ultrasonic detection system and method for a cast-in-place pile, which realizes real-time collection of pile top image data of the cast-in-place pile through a control system combined with a camera on a mechanical arm, automatically identifies pixel coordinates of a pile core and a sounding pipe opening according to the pile top image data and converts the pixel coordinates into physical coordinates, compares the physical coordinates with theoretical coordinates of a drawing to determine whether a walking device is in place, completely replaces a link of manual checking of the drawing, eliminates confusion of pile position numbers, and greatly improves a positioning standardization degree. After the walking device moves to a target position, the control system controls the mechanical arm, so that an ultrasonic detection device originally arranged in the mechanical arm is accurately corresponding to the physical coordinates of the sounding pipe opening under the guidance of a guide groove, meanwhile, an automatic winding and unwinding device is driven to lower a guide wire, so that the ultrasonic detection device vertically enters a corresponding sounding pipe under the cooperation of constraint of the guide groove and self weight until the ultrasonic detection device touches the bottom, manual lowering of a transducer is not needed in the whole process, one on-site operation detection personnel is saved, and labor intensity is obviously reduced. In addition, the automatic winding and unwinding device arranged in the application can automatically complete lowering and recycling of the guide wire, replaces complicated manual recycling of the guide wire, improves detection efficiency, realizes automatic implementation of a whole detection flow, can reduce data collection interference caused by human intervention, guarantees perpendicularity and stability of ultrasonic detection, greatly improves authenticity and effectiveness of detection data, is suitable for application scenarios of centralized construction of a large number of long piles, and solves problems of high manual dependence, low efficiency and insufficient detection data quality in a traditional method. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the application or the technical solutions in the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the application, and for those skilled in the art, other embodiments can be obtained from these drawings without creative labor.

[0020] Figure 1 FIG. 1 is a structural schematic diagram of a full-automatic intelligent ultrasonic detection system for a cast-in-place pile according to an embodiment of the application.

[0021] Figure 2 FIG. 2 is a structural block diagram of a control system of the full-automatic intelligent ultrasonic detection system for the cast-in-place pile according to the embodiment of the application.

[0022] Figure 3 FIG. 3 is a flow schematic diagram of a full-automatic intelligent ultrasonic detection method for the cast-in-place pile according to the embodiment of the application.

[0023] Reference signs in the drawings:

[0024] 100. Walking equipment; 101. Automatic deployment and retraction equipment; 102. Positioning device; 200. Robotic arm; 201. Guide slot; 300. Camera; 400. Ultrasonic testing equipment; 500. Acoustic logging tube; 600. Cast-in-place pile; 700. Ground. Detailed Implementation

[0025] Embodiments of the present invention will now be described in more detail with reference to the accompanying drawings. While some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0026] To address the problems of high reliance on manual labor, low efficiency, and insufficient data quality in ultrasonic testing of cast-in-place piles, this invention provides a fully automated intelligent ultrasonic testing system and method for cast-in-place piles.

[0027] Among them, such as Figure 1 As shown, the embodiment of the present invention provides a fully automatic intelligent ultrasonic testing system for cast-in-place piles, comprising: a walking device 100, equipped with an automatic retraction device 101; the automatic retraction device 101 is wound with a wire; a robotic arm 200, mounted on the walking device 100; the robotic arm 200 has a wire routing channel inside, and a guide slot 201 connected to the wire routing channel is provided at the end of the robotic arm 200; an ultrasonic testing device 400 is movably mounted in the guide slot 201; the end of the wire is connected to the ultrasonic testing device 400 through the wire routing channel; a camera 300, mounted on the robotic arm 200, is used to acquire real-time image data of the pile top of each cast-in-place pile 600; and a control system, mounted on the walking device 100. The backup device 100 is used to identify the pixel coordinates of the core of each cast-in-place pile 600 and the pixel coordinates of the sonic logging pipe inlet pre-embedded in the cast-in-place pile based on the image data of the pile top. It determines whether the walking device 100 has reached the target position based on the deviation between the physical coordinates obtained by converting the pixel coordinates of the core and the theoretical coordinates of the core in the cast-in-place pile drawings. When the walking device 100 moves to the target position, the robotic arm 200 is controlled to move the ultrasonic testing device 400 to the physical coordinates obtained by converting the pixel coordinates of the sonic logging pipe inlet, and the automatic take-up and release device 101 is driven to lower the wire, so that the ultrasonic testing device 400, under the combined action of the guide slot 201 and its own weight, vertically enters the interior of the corresponding sonic logging pipe 500 until it touches the bottom.

[0028] Specifically, the cast-in-place pile 600 is buried below the ground 700, the pile body of the cast-in-place pile 600 is formed by pouring concrete, and the top of the pile is slightly higher than the ground 700; the sounding pipe 500 is pre-buried in the pile body along the length direction of the pile body of the cast-in-place pile 600, and the detection pipe opening at the top end is higher than the top of the pile of the cast-in-place pile 600, forming a detection channel for the ultrasonic detection equipment 400 to enter.

[0029] The walking device 100 is a bearing and moving platform with automatic walking ability, and is provided with core components such as automatic receiving and releasing device 101, mechanical arm 200, control system, etc., can realize automatic movement according to the instruction of the control system, and accurately arrive at the position of the target cast-in-place pile, providing stable operation reference for subsequent ultrasonic detection operation.

[0030] Further, the walking device 100 can be a wheeled automatic walking device 100 (such as AGV, Automated Guided Vehicle, automatic guided vehicle, AMR, Autonomous Mobile Robot, self-moving robot), tracked automatic walking device 100 (such as tracked robot chassis, tracked AGV) or multi-legged bionic walking device 100 (such as four-legged robot dog, six-legged bionic robot).

[0031] The automatic receiving and releasing device 101 is arranged on the walking device 100, can accurately control the length and speed of the wire release according to the operation requirement of the ultrasonic detection equipment 400, realize the stable release and safe recovery of the ultrasonic detection equipment 400 after operation, and ensure the neatness of the wire storage and avoid the entanglement and disorder affecting the operation.

[0032] The automatic receiving and releasing device 101 can be an electric winch (realizing wire release by rotating the drum driven by the motor, and the rotating speed and stroke can be accurately controlled), an automatic wire collector (with automatic wire arrangement function, suitable for small and medium-sized wire release), a hoist (heavy load type release device, suitable for large diameter and long stroke wire release requirement, suitable for deep pile detection scene), a servo control release reel (with position sensor, can feedback the wire release length in real time, and realize closed loop control with the control system).

[0033] The mechanical arm 200 is arranged on the walking device 100, has multi-degree-of-freedom motion ability, is internally integrated with a wire laying channel, and has a mechanical execution mechanism with a guide slot 201 at the tail end, can adjust the spatial position and attitude of the ultrasonic detection equipment 400 and the camera 300 according to the instruction of the control system, realize accurate alignment of the ultrasonic detection equipment 400 and the sounding pipe opening of the cast-in-place pile, and provide a stable shooting angle for the camera 300.

[0034] The mechanical arm 200 generally comprises a base, an arm rod assembly, and an effector. The base is the fixed foundation of the mechanical arm 200, which is usually rigidly fixed with the walking device 100 by bolts, welding or custom flanges. One end of the arm rod assembly (consisting of multiple arm rods connected by joints) is hinged to the base, and the other end is hinged to the effector. The effector is the terminal output component of the mechanical arm 200, which is provided with a guide slot 201.

[0035] The control system controls the rotation angle, rotation speed and movement sequence of each joint of the mechanical arm 200 by sending electrical signal instructions to the driving mechanism (such as a servo driver) of each joint. Each joint driving mechanism moves the corresponding arm rod assembly according to the instructions, adjusts the spatial coordinates of the effector (matches the physical coordinates of the sounding pipe opening) through the cooperation of multiple joints, and at the same time, the position sensor (such as an encoder) built-in the mechanical arm 200 real-time feedbacks the actual position of each joint to the control system, forming a closed loop control, ensuring that the effector drives the ultrasonic detection device 400 to move accurately to the target position, avoiding deviation.

[0036] The ultrasonic detection device 400 is the core detection component that can move vertically along the sounding pipe, detects the quality of the pile body by using the principle of ultrasonic reflection, and is connected with the wire and is dragged and signal-transmitted by the wire. The ultrasonic detection device 400 can emit ultrasonic signals and receive the reflection signals of the gap between the pile body concrete and the sounding pipe and the defects (such as cavities, cracks, and mud inclusions) of the pile body during the movement in the sounding pipe, and transmits the detection signals to the control system through the wire to provide original data for the quality evaluation of the cast-in-place pile.

[0037] Further, the ultrasonic detection device 400 can be an ultrasonic pile measuring transducer (divided into single emission and single reception, single emission and double reception, etc., suitable for different diameter sounding pipes), an integrated ultrasonic detection probe (integrating the emission and reception units, small in size and convenient to move in the sounding pipe), and a multi-channel ultrasonic detection device 400 (which can simultaneously detect multiple sounding pipes, improving the detection efficiency).

[0038] The camera 300 can be an industrial color camera 300, a CCD (Charge-Coupled Device, Charge-Coupled Device) vision camera 300, etc. The camera 300 can be directly arranged at the end of the effector or mounted on the effector by means of a support.

[0039] The lens of the camera 300 faces vertically to the ground 700, which is used for vertically downward shooting of the pile top area in a constant light source (such as a ring-shaped LED fill light) environment during the walking of the walking device 100, to real-time collect the pile top image data of the cast-in-place pile 600. The pile top image data mainly includes the contour shape and boundary range of the pile top, the flatness of the pile top surface, the position, number, shape of the pre-buried sounding pipe opening, the identification or features of the center area of the pile top, and the auxiliary environmental information of the periphery of the pile top.

[0040] The control system at least includes: a core processor, a drive control module, a communication module, a storage module and a sensor feedback unit. The core processor can be an industrial CPU (Central Processing Unit), a PLC (Programmable Logic Controller), responsible for instruction operation and logic control. The drive control module can be a servo driver, a motor controller, connected to and controlling the driving mechanism of the walking device 100, the mechanical arm 200 and the automatic winding and unwinding device 101. The communication module, such as Ethernet, CAN (Controller Area Network) bus, wireless communication unit, realizes data transmission between components. The storage module is used to store the pile drawing, detection data, image data, etc.

[0041] The walking device 100 is also provided with a positioning device 102, such as a GPS (Global Positioning System) / Beidou positioning unit, an inertial navigation module, to assist the walking device 100 in positioning. The positioning device 102 is electrically connected to the core processor and realizes bidirectional data interaction, to transmit the spatial position coordinates and motion attitude data of the walking device 100 to the core processor in real time, to provide a reference positioning basis for the core processor to combine the camera 300 intrinsic parameter, the conversion matrix between the camera 300 and the walking device 100 and the mechanical arm 200, to accurately convert the pile core and the sounding pipe mouth pixel coordinates obtained by image recognition into physical coordinates in the construction coordinate system.

[0042] The control system sends a shooting instruction to the camera 300, and the camera 300 continuously collects pile top image data during the movement of the walking device 100 and transmits it to the control system in real time. The core processor of the control system pre-processes the collected raw images (including denoising, contrast enhancement, distortion correction, etc.), to eliminate the influence of environmental interference such as construction dust, strong light and shadow on the image, and to improve the image clarity and quality. Then, based on the pre-processed image, through a preset algorithm (such as an edge detection algorithm to identify the pile top contour and the sounding pipe mouth edge, a Hough circle transformation algorithm to locate the center of the sounding pipe mouth, and a barycentric algorithm to calculate the barycenter of the pile top contour, which is the pile core pixel coordinates), the pile core pixel coordinates and the sounding pipe mouth pixel coordinates of each cast-in-place pile 600 are accurately identified, and the identification result is fed back to the core processor, to realize automatic and accurate positioning of the pile core and the sounding pipe mouth, replace manual positioning, improve positioning efficiency and accuracy, and avoid subjective errors of manual positioning.

[0043] The core processor triggers the drawing input flow of the control system, and imports the bored pile drawing into the control system. The core processor of the control system autonomously identifies and extracts key engineering data such as the pile core theoretical coordinates in the drawing, and then stores them in the storage module. After completing data storage, the core processor calls the above-mentioned key data in the storage module, converts the pile core pixel coordinates obtained by image recognition into actual physical coordinates in the construction coordinate system through a preset pixel coordinate and physical coordinate conversion algorithm (combined with the camera 300 internal parameter, the Beidou coordinates of the walking device 100, and the conversion matrix between the camera 300 and the walking device 100 and the mechanical arm 200). Then, the core processor calculates the deviation value of the actual physical coordinates and the pile core theoretical coordinates in the drawing. If the deviation value is within the preset threshold range, it is determined that the walking device 100 has reached the target position; if the deviation value exceeds the threshold, the core processor sends a position adjustment instruction to the driving control module of the walking device 100 to drive the walking device 100 to move accurately until the deviation value meets the preset requirements.

[0044] When the walking device 100 reaches the target position, the core processor converts the pixel coordinates of the sounding pipe mouth into corresponding physical coordinates, combines the current posture data of the mechanical arm 200, and plans the motion trajectory of the mechanical arm 200 through the kinematics forward and inverse solution algorithm. Then send the trajectory instruction to the driving control module of the mechanical arm 200 to control the coordinated motion of each joint of the mechanical arm 200, drive the end of the ultrasonic detection device 400 to move to the physical coordinates of the sounding pipe mouth, realize the accurate alignment of the detection device and the sounding pipe mouth, so as to ensure the accurate and smooth action of the mechanical arm 200, realize the automatic alignment of the ultrasonic detection device 400 and the sounding pipe mouth, avoid the tediousness and error of manual alignment, and improve the operation efficiency.

[0045] The core processor sends a lowering instruction to the driving module of the automatic winding and unwinding device 101 to drive the automatic winding and unwinding device 101 to lower the lead wire at a preset speed; the lead wire drives the ultrasonic detection device 400 to move, and at this time the guide slot 201 at the end of the mechanical arm 200 limits the lateral displacement of the detection device, and cooperates with the self-weight of the detection device to make the detection device enter the sounding pipe inside along the vertical direction, avoiding tilting and jamming. When the bottom touch sensor at the bottom of the ultrasonic detection device 400 detects the signal of contacting the bottom of the sounding pipe (or the lowering length of the lead wire reaches the preset sounding pipe depth threshold), the touch signal is fed back to the core processor, and the core processor sends a stop lowering instruction to stop the automatic winding and unwinding device 101 from working, ensuring that the detection device reaches the specified detection starting position, and providing a guarantee for the accuracy of subsequent pile body quality detection.

[0046] Therefore, the full-automatic intelligent ultrasonic detection system for cast-in-place piles provided by the embodiment of the application completely replaces the link of manually checking the drawing, eliminates the confusion of pile position numbers, and greatly improves the standardization degree of positioning.

[0047] After the walking device 100 moves to the target position, the control system controls the mechanical arm 200, so that the ultrasonic detection device 400 originally arranged in the inside of the mechanical arm 200 accurately corresponds to the physical coordinates of the sounding pipe opening under the guidance of the executor guide slot 201, and drives the automatic winding and unwinding device 101 to lower the guide wire, so that the ultrasonic detection device 400 vertically enters the corresponding sounding pipe under the joint action of the constraint of the guide slot 201 and the self-weight until the bottom is touched, and the transducer does not need to be manually lowered throughout the process, one on-site operation detection personnel is saved, and the labor intensity is significantly reduced.

[0048] Secondly, the automatic winding and unwinding device 101 arranged by the application can automatically complete the lowering and recycling of the guide wire, replace the cumbersome operation of manually recycling the guide wire, improve the detection efficiency, and realize the automatic implementation of the whole detection process, so as to reduce the data acquisition interference caused by human intervention, ensure the perpendicularity and stability of ultrasonic detection, greatly improve the authenticity and effectiveness of the detection data, and at the same time adapt to the application scene of centralized construction of a large number of long piles, solve the problems of high degree of manual dependence, low efficiency and insufficient detection data quality of the traditional method.

[0049] Further, as shown in Figure 2 The core processor of the control system preferably comprises an image processing module, a feature extraction module and a target recognition module.

[0050] The image processing module is connected with the camera 300, and is used to convert the colorful pile top image data into corresponding gray image data and extract gradient amplitudes.

[0051] The feature extraction module is connected with the image processing module, and is used to extract the top surface target contour of each cast-in-place pile 600 and the quantized features of each connected domain in the top surface target contour based on each gray image data and the gradient amplitudes.

[0052] The target recognition module is connected with the feature extraction module, and is used to recognize the pile body region and the sounding pipe opening region of the cast-in-place pile 600 from each connected domain based on the quantized features, and determine the pixel coordinates of the pile core and the sounding pipe opening of the corresponding cast-in-place pile 600 based on the center coordinates of the pile body region and the sounding pipe opening region.

[0053] Specifically, the camera 300 of the embodiment of the present application is preferably an industrial color camera 300, and the directly photographed pile top image data is RGB (Red Green Blue) image data. The image processing module is used for pre-processing the pile top image data photographed by the camera 300, and at least includes an image contrast enhancement process, preferably using adaptive histogram equalization (CLAHE, Contrast Limited Adaptive Histogram Equalization) to enhance the contrast of the concrete, the pipe mouth and the soil body, and the processing steps preferably include the following steps:

[0054] (1) Input and conversion. Obtain a high-resolution pile top surface RGB image, and convert it into a grayscale image according to the standard luminance perception coefficients (0.299, 0.587, 0.114). The perception coefficients conform to the ITU-R BT.601 standard, can most accurately reflect the perception of the human eye to brightness, and are beneficial to subsequent brightness-based segmentation.

[0055] (2) Local contrast adaptive enhancement. Apply the CLAHE algorithm to the grayscale image. Divide the image into MxN local blocks (preferably M=N=8), perform equalization on the histogram of each block after contrast limiting (Clip Limit 2.0-3.0), and eliminate block effects through bilinear interpolation to obtain an enhanced image.

[0056] (3) Morphological noise suppression. In order to eliminate isolated noise points and small texture interference that may be introduced in the enhanced image, a circular structure element with a radius of 2 pixels is used to perform morphological opening operation on the enhanced image to obtain an optimized grayscale image.

[0057] (4) Gradient feature extraction. Calculate the pixel gradient amplitude map of the optimized grayscale image. The Sobel operator is used to calculate the gradients in the x and y directions, and the comprehensive gradient amplitude is obtained by squaring and taking the square root (or sum of absolute values). This step can significantly strengthen the input features of subsequent edge detection.

[0058] (5) Output. The pre-processing module finally outputs the optimized grayscale image and its corresponding gradient amplitude map for use by the subsequent recognition module.

[0059] The above steps provided by the embodiment of the present application precisely match the brightness perception characteristics of the human eye to complete image grayscale, effectively improve the area differentiation degree of the concrete, the pipe mouth and the soil body and eliminate block effects, realize the suppression of isolated noise points and small texture interference, significantly strengthen the image edge features, and finally output the optimized grayscale image and the gradient amplitude map, which greatly improves the accuracy and reliability of subsequent pile body and pipe mouth area recognition.

[0060] Furthermore, the feature extraction module preferably includes: an edge detection module, connected to the image processing module, used to extract the top surface target contour of each grouting pile 600 based on each grayscale image data, and output a binary edge image with a single pixel width; a repair module, connected to the edge detection module, used to perform edge morphological closing operations on the binary edge image to obtain an edge image with a continuous and complete contour; and a contour feature quantization module, connected to the repair module, used to perform connected component analysis and quantization processing on the edge image, and extract the quantization features of each connected component; the quantization features include: size features, position features, morphological features, and grayscale features.

[0061] Specifically, the edge detection module preferably employs the Canny edge detection algorithm. Canny edge detection is performed on the preprocessed optimized grayscale image to extract the salient contours of all targets on the top surface of the cast-in-place pile. The specific implementation steps include: Gaussian smoothing. Standard deviation is used. A 5×5 Gaussian kernel is used to convolve the optimized grayscale image to suppress high-frequency noise, resulting in a smooth image. Gradient calculation: The Sobel operator is used to calculate the gradients of the smoothed image in the x and y directions, obtaining the gradient magnitude map and direction map. Non-maximum suppression: The gradient magnitude map is traversed, retaining only the pixels with the largest local magnitudes in the gradient direction, and setting the magnitudes of other pixels to zero, resulting in a refined edge response map. Double-threshold hysteresis connection: A high threshold is set to 20% of the maximum magnitude of the edge response map, and a low threshold is set to 40%. Pixels in the edge response map with magnitudes higher than the high threshold are marked as strong edges, and pixels with magnitudes between the low and high thresholds and connected to the neighborhood of strong edges are marked as weak edges and connected, finally outputting a binary edge image with a single pixel width.

[0062] The edge detection module provided in the embodiments of the present invention effectively suppresses high-frequency noise in the image by standardizing the Canny edge detection process, refines the edges to a single pixel width, and accurately filters and connects strong and weak edges of the target on the top surface of the grouting pile through dual threshold hysteresis connection. The output binary edge image can extract the significant contours of all targets with high precision, providing a reliable edge basis for subsequent connected component quantization feature extraction.

[0063] Furthermore, the repair module preferably employs edge morphological closing operations to process edge contours in binary edge images that are broken due to insignificant local grayscale changes. The specific implementation process includes: structuring element definition: Define a circular structuring element with a radius of 2 pixels. Dilation and erosion: Dilate the binary edge image using the circular structuring element to moderately thicken the edges and connect adjacent breakpoints, obtaining an intermediate image. Then, erode the intermediate image using the same circular structuring element to roughly restore the edges to their original thickness, while maintaining the connection of already connected breakpoints, ultimately obtaining the connected edge image.

[0064] The above repair module provided by the embodiment of the application performs morphological closing operation on the binary edge image through a circular structure element with a radius of 2 pixels, effectively connects the broken edge contour due to insignificant local gray scale change, and outputs the edge image with continuous and complete contour while roughly restoring the original thickness of the edge, thereby providing complete edge basis for subsequent connected domain analysis and quantitative feature extraction.

[0065] Further, the contour feature quantization module preferably adopts contour closure and region marking to perform connected domain analysis on the edge image, and finds all connected regions formed by white pixels. For each connected domain, the minimum circumscribed rectangle, area and contour point set thereof are calculated. At this time, the concrete pile body, ultrasonic detection pipe mouth and surrounding soil are usually segmented into independent closed or approximately closed contour regions, i.e. connected domains. Then, through quantization processing, the quantitative features of each connected domain are extracted by calculation. The quantitative features include size features (pixel area, equivalent diameter, minimum circumscribed rectangle size), position features (centroid pixel coordinates, circumscribed rectangle center coordinates), shape features (circularity, eccentricity, compactness) and gray scale features (average gray scale value, gray scale variance, gray scale extreme value and gradient mean value).

[0066] The above contour feature quantization module provided by the embodiment of the application performs connected domain analysis on the edge image through contour closure and region marking, realizes independent connected domain segmentation of the concrete pile body, ultrasonic detection pipe mouth and surrounding soil, calculates the minimum circumscribed rectangle, area and contour point set of each connected domain, and extracts four types of quantitative features, i.e. size, position, shape and gray scale, thereby providing structured feature data support for accurate recognition of the pile body and pipe mouth by the subsequent target recognition module.

[0067] The above feature extraction module provided by the embodiment of the application cooperates the edge detection, repair and contour feature quantization sub-modules, accurately extracts the single-pixel binary edge of the cast-in-place pile top surface target through the Canny algorithm and effectively suppresses noise, connects the broken contour through morphological closing operation to obtain a continuous and complete edge image, and finally realizes independent segmentation of the concrete pile body, ultrasonic detection pipe mouth and surrounding soil through connected domain analysis, and extracts four types of core quantitative features, i.e. size, position, shape and gray scale, thereby providing complete and reliable structured feature data basis for the subsequent target recognition module to accurately distinguish and locate the pile body and ultrasonic detection pipe mouth region.

[0068] Further, the target recognition module preferably comprises: a pile body recognition module, connected with the feature extraction module, configured to filter, based on the quantified features, a region with the largest area and an average gray value greater than a first set threshold from each connected domain, to determine the region as a pile body candidate region; and a pile body determination module, connected with the pile body recognition module, configured to filter, from the pile body candidate region, a region with a circularity and an equivalent diameter size both satisfying a set range, to determine the region as a pile body region, and to determine a centroid of the pile body region as a pixel coordinate of a corresponding pile core.

[0069] Further, the pile body recognition module filters the concrete pile body from all connected domains, and specifically comprises the following steps: based on the average gray value of each connected domain in the quantified features, and in combination with the gray characteristic difference between the concrete pile body and the surrounding soil and the sounding pipe mouth, a high threshold (preferably 80% to 90% of the average gray value of the concrete pile body based on sample statistics) is preset, and a region set with an average gray value greater than the preset high threshold is filtered from all connected domains; since the concrete pile body usually appears as a high gray connected domain with the largest area in the pile top image, the soil region is mostly low gray, and the sounding pipe mouth is a small area low gray connected domain, therefore, a set number of regions with the largest area are further selected from the region set as the pile body candidate region.

[0070] The pile body determination module first calculates the circularity of the pile body candidate region, and if the value is greater than 0.8, it is determined that the pile cross section is circular. The equivalent diameter is calculated and converted into a physical size through a pre-labeled pixel-millimeter ratio to verify whether it falls within the allowable error range of the designed pile diameter (for example, ±5%). If the verification is passed, the pile body candidate region is confirmed as the concrete pile body region, and its centroid (Cx_p, Cy_p) is recorded as the pixel coordinate of the pile core.

[0071] The pile body recognition module and the pile body determination module provided by the embodiments of the present application work cooperatively to efficiently filter the pile body candidate region and effectively exclude interference and false targets, accurately confirm the concrete pile body region after double verification, and accurately record the centroid as the pixel coordinate of the pile core, thereby providing high-precision data support for subsequent positioning and coordinate conversion of the walking equipment 100.

[0072] Further, the target recognition module preferably further comprises: a pipe mouth recognition module, connected with the feature extraction module, configured to filter, based on the quantified features, a region with an average gray value less than a second set threshold and an area satisfying a limit range from each connected domain, to determine the region as a pipe mouth initial candidate set; and a pipe mouth determination module, connected with the pipe mouth recognition module, configured to filter, from the pipe mouth initial candidate set, a region with a circularity, an equivalent diameter, a center position, and each center connection pattern all satisfying a set range, to determine the region as a sounding pipe mouth region, and to determine a center coordinate of the sounding pipe mouth region as a pixel coordinate of a corresponding sounding pipe mouth.

[0073] The tube opening recognition module, for all connected domains output by the contour feature quantization module, based on the gray level characteristic difference between the acoustic measuring tube opening and the concrete pile body and the surrounding soil, first filters out the area with an average gray level value lower than a preset low threshold (preferably 30%-50% of the average gray level value of the concrete pile body); then, combined with the common tube opening size of the 50mm-60mm specification ultrasonic detection tube, the pixel-millimeter conversion ratio calibrated by the camera 300 is used to obtain the pixel area range [A_tube_min, A_tube_max], and the area falling within the range is further filtered out; finally, the area that meets the gray level and area conditions at the same time constitutes the initial candidate set of the acoustic measuring tube opening.

[0074] The determination step of the tube opening determination module preferably includes: circular fine detection, for each acoustic measuring tube opening initial candidate area, a Hough circle transformation is performed on the original gray level image in the pixel neighborhood range thereof to detect a circular contour; in the transformation process, the pixel radius range of the circle needs to be strictly limited to , wherein is the minimum pixel radius corresponding to the 50mm specification ultrasonic detection tube opening radius, is the maximum pixel radius corresponding to the 60mm specification ultrasonic detection tube opening radius, both of which are obtained by the pixel-millimeter conversion ratio calibrated by the camera 300. Comprehensive verification is performed on each circular contour (denoted as the ith circle, with parameters , wherein is the horizontal coordinate of the center, is the vertical coordinate of the center, is the radius of the circle, all of which are pixel coordinates) detected by the Hough circle transformation, and the following three verifications are performed in turn:

[0075] 1. Size verification: the pixel radius of the circle is converted into a physical radius through the pixel-millimeter conversion ratio, and then a physical diameter is calculated to verify whether the physical diameter falls within the common tube opening size range of 50mm-60mm ultrasonic detection tube. 2. Position verification: after the pixel boundary range of the identified pile body candidate area is converted into a physical range, it is verified whether the physical coordinates of the center of the ith circle are located inside the pile body physical range, or the straight-line distance from the center of the circle to the physical boundary of the pile body is not more than 50mm. 3. Topological verification: when the number of acoustic measuring tubes designed for the pile to be detected 600 is greater than or equal to 3, the physical coordinates of the centers of all circular contours that pass the size verification and the position verification are extracted to form a geometric point set, and it is verified whether the physical side length and the physical angle of the geometric figure formed by the point set are consistent with the size and angle requirements of the preset acoustic measuring tube arrangement form (such as square, equilateral triangle, etc.) in the design drawing of the cast-in-place pile.

[0076] Finally, based on the results of the above three verifications, the pixel coordinates of the centers of all acoustic measuring tube openings that pass the verification completely are filtered out and arranged into a set wherein n is the final confirmed number of sounding pipe openings, is the center pixel coordinate of the kth sounding pipe opening, so as to obtain the final pixel coordinates of each sounding pipe opening.

[0077] The above pipe opening identification module and pipe opening determination module provided by the embodiments of the present application, through cooperative operation, efficiently screen out the initial candidate set of sounding pipe openings and effectively eliminate irrelevant interference areas such as soil and concrete protrusions, and after accurate verification, determine the specific position of the sounding pipe opening, and accurately record the core pixel coordinates, thereby providing key positioning basis for subsequent accurate alignment and automatic detection of the ultrasonic detection device.

[0078] Further, after obtaining the pixel coordinates of the pile core and the sounding pipe opening of each cast-in-place pile 600, the target recognition module of the embodiments of the present application further creates a mask with the same size as the image, deducts the pile body candidate area and the small circle area of each verified pipe opening, and then performs connected domain analysis on the remaining pixels, determines the first N areas (usually N = 1 or 2) with the largest area as the surrounding soil area, and completes scene understanding. Finally, the pixel coordinates of the pile core and the sounding pipe opening and the pile body boundary contour are uniformly output, thereby providing absolute spatial reference for subsequent walking device navigation and transducer centering.

[0079] Further, as shown in Figure 2 Further, the control system implemented by the embodiments of the present application preferably further comprises: an information extraction module configured to extract, based on the input cast-in-place pile drawing, the theoretical coordinates of the pile core of each cast-in-place pile in the construction coordinate system in the cast-in-place pile drawing; a coordinate conversion module connected with the information extraction module and the target recognition module, and configured to convert the pixel coordinates of the pile core into corresponding physical coordinates in the construction coordinate system based on the Beidou coordinates of the walking device 100 and the conversion matrix between the camera 300 and the end effector of the walking device 100 and the mechanical arm; a deviation calculation module connected with the information extraction module and the coordinate conversion module, and configured to calculate the planar deviation distance between the physical coordinates of the pile core and the theoretical coordinates of the pile core; and a pile position confirmation module connected with the deviation calculation module, and configured to compare the planar deviation distance with a set deviation threshold, and confirm that the walking device 100 reaches the target position when the planar deviation distance is less than or equal to the set deviation threshold.

[0080] Further, the information extraction module receives the cast-in-place pile drawing (such as a pile foundation CAD design drawing, supporting mainstream engineering drawing formats such as DXF and DWG), automatically identifies and locates target layers such as the pile position number layer and the pile core design coordinate layer in the drawing based on a preset drawing layer definition rule and attribute field mapping relationship, reads the unique pile position number of the current cast-in-place pile 600 to be detected and the corresponding theoretical planar coordinates (X, Y) of the cast-in-place pile 600 to be detected in the design construction coordinate system from the attribute database associated with the layer, and outputs the theoretical planar coordinates (X, Y) of the pile core of the cast-in-place pile 600 to be detected. , ); The module has a coordinate validity verification function, which ensures data accuracy by comparing the consistency of the extracted coordinates with the preset coordinate range of the project, and combining the detection task list issued by the system to complete the accurate matching of the to-be-detected pile number and the theoretical coordinates. Finally, the structured pile number and theoretical plane coordinate data are output for subsequent coordinate conversion module and deviation calculation module to call.

[0081] The coordinate conversion module is also connected with the positioning device 102. After receiving the pile core pixel coordinates, the pixel coordinates are first corrected and normalized based on the pre-calibrated camera 300 intrinsic matrix and distortion coefficient to convert them into coordinates in the camera 300 coordinate system. Then, through the conversion matrix between the camera 300 and the walking device 100 and the actuator, the coordinate mapping from the camera 300 coordinate system to the walking device 100 coordinate system is realized; finally, combined with the Beidou coordinates of the walking device 100 sent by the positioning device 102 (i.e. the pose of the walking device 100 in the construction coordinate system), the pile core coordinates in the walking device 100 coordinate system are converted into physical coordinates in the construction coordinate system through coordinate system transformation, completing the coordinate conversion from pixel dimension to physical construction dimension.

[0082] The conversion matrix between the camera 300 and the walking device 100 and the actuator is a 4x4 homogeneous transformation matrix that describes the rigid transformation relationship between two coordinate systems. Its core function is to establish accurate mapping between the camera 300 coordinate system and the walking device 100 coordinate system and the actuator coordinate system, and realize coordinate cross-system conversion. Among them, the conversion matrix between the camera 300 and the walking device 100 is used to associate the relative pose of the camera 300 and the walking device 100 carrier. It can be calculated by static calibration method, i.e. fixing the standard calibration board at a known construction coordinate position, controlling the walking device 100 to stop and taking pictures of the calibration board by the camera 300, combining the physical size of the calibration board and the image pixel coordinates, and solving the rotation and translation parameters of the camera 300 relative to the walking device 100 coordinate system, to obtain the conversion matrix between the camera 300 and the walking device 100.

[0083] The conversion matrix between the camera 300 and the actuator (i.e. the hand-eye transformation matrix) is used to associate the relative pose of the camera 300 and the actuator (such as the ultrasonic detection probe driving mechanism). It can be calculated by hand-eye calibration method (such as Tsai-Lenz algorithm), i.e. moving the calibration target driven by the actuator at different poses, and the camera 300 synchronously takes pictures of the target image. Through multiple sets of pose data, the relative transformation relationship between the camera 300 and the actuator end is solved, and the calibrated parameters are stored in the system as fixed parameters for subsequent coordinate conversion.

[0084] The calculation formula of the plane deviation distance between the physical coordinates of the pile core and the theoretical coordinates of the pile core is:

[0085] .

[0086] .

[0087] .

[0088] wherein, , is the theoretical coordinate component of the pile core in the construction coordinate system; , is the visual physical coordinate component of the pile core pixel coordinate after conversion by the coordinate conversion module in the construction coordinate system; , is the position deviation component of the pile core visual physical coordinate and the theoretical coordinate in the X-axis and Y-axis of the construction coordinate system; is the comprehensive plane deviation distance between the pile core visual physical coordinate and the theoretical coordinate.

[0089] The pile position confirmation module realizes pile position confirmation in a threshold judgment manner, and the specific logic is as follows: first, the allowed positioning tolerance threshold is set according to the engineering specification (preferably 100 mm), then the pile core comprehensive plane deviation distance output by the deviation calculation module is compared with the positioning tolerance threshold ; if , it is determined that the walking device 100 has accurately arrived at the target pile position, and the module immediately outputs a "pile position confirmation success" signal, which will serve as a trigger instruction to start the subsequent full-automatic ultrasonic detection process.

[0090] The control system provided by the embodiment of the application realizes the extraction of the theoretical pile position of the cast-in-place pile, the conversion of the pixel coordinate to the construction physical coordinate, the calculation of the pile position deviation, and the automatic confirmation of the target pile position of the walking device 100, thereby providing high-precision and high-reliability pile positioning support for the smooth start of the subsequent full-automatic ultrasonic detection process.

[0091] Further, the control system of the application further preferably comprises: a multi-source data acquisition module for acquiring encoder data of the automatic winding and unwinding device 101, ultrasonic detection data of the ultrasonic detection device 400, video stream data of the camera 300, and running state data of the walking device 100, the mechanical arm 200 and the automatic winding and unwinding device 101; a display module connected with the multi-source data acquisition module, for visually displaying the theoretical parameters of each cast-in-place pile 600, the lowering progress of the ultrasonic detection device 400, the ultrasonic detection data, the video screen shot by the camera 300 and the running state data; wherein the lowering progress is calculated based on the encoder data and the pile length in the theoretical data of the cast-in-place pile 600.

[0092] Further, the encoder data of the automatic winding and unwinding device 101. Wherein, the encoder is integrated into the winding drum drive motor shaft or cable tension detection shaft end of the automatic winding and unwinding device 101, the multi-source data acquisition module collects the pulse count or absolute position data of the encoder through the pulse output interface or digital communication interface (such as RS485) of the encoder in real time; The module has built-in device transmission ratio and winding drum diameter parameters, which can synchronously associate the collected data to the cable unwinding length calculation, and provide a basis for subsequent unwinding progress derivation.

[0093] The ultrasonic detection data of the ultrasonic detection device 400 includes: the acoustic time domain waveform and acoustic parameters of the ultrasonic host. The multi-source data acquisition module connects with the ultrasonic host through Ethernet or RS232 / 485 special communication interface, reads the acoustic time domain waveform digital quantization data (i.e. voltage signal time sequence of ultrasonic echo) generated by the ultrasonic host in real time according to the preset device communication protocol, and the acoustic parameters (including acoustic time, amplitude, main frequency, acoustic attenuation coefficient, etc.) calculated by the host built-in algorithm; During the collection process, the module performs real-time caching and format standardization on the data to ensure data integrity.

[0094] The industrial camera 300 is connected with the multi-source data acquisition module through GigE Vision or USB3.0 special industrial interface, and the module collects the original video stream data (supports RAW, JPEG, etc.) output by the camera 300 in real time according to the camera 300 communication protocol (such as GigE Vision protocol); During the collection process, the module can perform real-time decoding or frame caching processing on the video stream according to the system requirements, guarantee the continuity and clarity of the video picture, and meet the visual presentation requirements of the display module.

[0095] The running state data of the walking device 100, the mechanical arm 200 and the automatic winding and unwinding device 101 includes the power, temperature and servo state signals of each device. Wherein, the power data is obtained by: connecting the battery management system (BMS) of the walking device 100, the power management module of the mechanical arm 200 and the automatic winding and unwinding device 101 through RS485 / Modbus communication interface to output the remaining power, bus voltage, working current and other data of the device, and the multi-source data acquisition module reads in real time through the corresponding communication interface.

[0096] The temperature data is that the temperature sensors (such as PT100 thermal resistance, digital temperature sensor DS18B20) are installed on the motors, controllers and drive units of the three types of devices, and the analog or digital signals output by the sensors are connected to the analog input channel or digital communication interface of the acquisition module, and the module converts the signals to temperature values in real time.

[0097] The servo status signals are acquired as follows: the drive motor of the walking device 100, the joint servo motor of the robotic arm 200, and the drum servo motor of the automatic take-up and untake-down device 101 will output digital status signals such as ready, running, fault, and limit. These signals are connected to the multi-source data acquisition module through the switch input interface. The module obtains the servo status of the device in real time by detecting the high and low levels of the signals.

[0098] The lowering progress is calculated based on the encoder data of the automatic lowering and retracting device 101 and the pile length in the theoretical data of the cast-in-place pile. The display module interface adopts a structured layout with different functional areas as follows: Static information area, which persistently displays the number of the currently inspected pile, design coordinates, geometric parameters, and inspection tube layout diagram, intuitively presenting the theoretical parameters of the cast-in-place pile; Dynamic progress area, which displays the current depth H of the transducer and the progress according to the formula in real time and animation using a virtual ruler and progress bar. The system calculates the percentage of the detection progress P (where L is the theoretical pile length of the cast-in-place pile). When the transducer touches the bottom, the percentage of progress P automatically returns to zero, dynamically displaying the lowering progress of the ultrasonic testing equipment 400. The data visualization area displays the real-time received waveform window and the acoustic parameter-depth growth curve window side by side. The curve extends and is drawn in real time as the transducer is raised, visually presenting the ultrasonic testing data. The equipment status area uses color icons and an instrument panel to display the total power of the device, the servo status of each axis, and the status of the alignment mechanism, centrally displaying the operating status data. The scene video area embeds real-time video footage captured by the camera 300, and can overlay the identified augmented reality (AR) marker boxes of the pile body and pipe opening, clearly presenting the video footage from the camera 300.

[0099] The multi-source data acquisition module and display module of this invention work together to achieve comprehensive acquisition and integrated visualization of key data throughout the entire detection system process, providing core support for real-time monitoring, progress tracking and status control of the detection process, and greatly improving the human-computer interaction efficiency and overall controllability of the detection operation.

[0100] Furthermore, the control system in this embodiment of the invention also includes an alarm module. During the operation of the detection system, the background monitoring thread continuously compares real-time data with rule thresholds: if the torque of any servo motor exceeds 150% of its rated value, a "stuck" alarm is triggered, and the system automatically stops. When the alarm is triggered, a warning window pops up on the interface, a log with timestamps and complete context data is recorded, and the video module is activated to capture images of the scene.

[0101] Furthermore, the control system synchronizes and stores all monitoring data (depth, progress, waveform, status, video frames, alarms) with high precision along a unified timeline. Users can click on alarm records or abnormal curve points to retrieve the depth, waveform screenshots, and on-site video footage at that moment with a single click, enabling full-chain traceability.

[0102] Further, the automatic winding and unwinding device 101 of the embodiment of the present application comprises a winding drum, a servo wire winder and a servo motor. The winding drum is rotatably arranged on the walking device 100. The servo wire winder is arranged on the walking device 100 and is arranged in parallel with the winding drum. The wire is wound on the winding drum after being wound by the servo wire winder. The servo motor is arranged on the walking device 100 and is fixedly connected with the center shaft of the winding drum. The servo motor and the servo wire winder are connected with the control system.

[0103] Specifically, the winding drum is made of high-strength lightweight alloy material and is rotatably supported on the special rack of the walking device 100 through high-precision deep groove ball bearings. The surface of the winding drum is processed with a spiral groove matching the specification of the wire for constraining the winding track of the wire. The servo wire winder is a precision wire winding structure of a screw rod sliding table type, the end of the sliding table is equipped with a wear-resistant ceramic wire winding guide wheel, and the whole is fixed on the rack of the walking device 100 in parallel with the axis of the winding drum, maintaining a predetermined distance with the winding drum to ensure the winding accuracy. The signal transmission line of the ultrasonic detection probe and the power integrated wire are guided by the guide wheel of the servo wire winder and orderly wound after winding, and then wound in the spiral groove of the winding drum. The servo motor is selected from high-response permanent magnet synchronous servo motors, which are coaxially fixedly connected with the center shaft of the winding drum through a rigid coupling, and the output torque can directly drive the winding drum to rotate. The servo motor and the servo wire winder are in bidirectional communication with the control system through a digital pulse command interface and an encoder feedback interface. The control system can realize motion synchronization control based on the encoder feedback data of the two, ensuring the accurate matching of the wire winding action and the winding speed during the winding and unwinding of the wire.

[0104] The automatic winding and unwinding device 101 provided by the embodiment of the present application realizes the orderly winding and unwinding and accurate length control of the detection wire, effectively avoids the problems of wire winding, knotting or disorder, and guarantees the synchronous and smooth lowering and lifting of the ultrasonic detection probe with the wire, accurately responds to the action instructions of the control system, and provides reliable equipment support for the real-time calculation of the lowering progress of the ultrasonic detection device 400 and the automation of the whole process of the detection.

[0105] Further, the ultrasonic detection device 400 is arranged at least twice, and the number corresponds to the number of the mechanical arm 200 and the automatic winding and unwinding device 101. Each ultrasonic detection device 400 is connected with an automatic winding and unwinding device 101 through a wire after passing through the wire channel of the corresponding mechanical arm 200, forming a group of detection components.

[0106] Specifically, the number of the ultrasonic detection device 400 matches the designed number of the detected cast-in-place pile sounding pipe 500, and is arranged at least twice. During the detection process, each ultrasonic detection device 400 cooperates with the transmitting probe and the receiving probe to realize the quality detection of the cast-in-place pile by utilizing the propagation characteristics of ultrasonic waves in the concrete medium.

[0107] The number of ultrasonic detection devices 400 is in strict one-to-one correspondence with the number of mechanical arms 200 and automatic winding and unwinding devices 101 arranged on the walking device 100; each ultrasonic detection device 400 is independently connected with a special wear-resistant integrated wire, the wire passes through a smooth wear-resistant wiring channel (the inner wall of the channel is embedded with a polytetrafluoroethylene wear-resistant bushing to reduce wire wear and friction resistance) in the corresponding mechanical arm 200 in sequence, and is finally wound and connected with the reel spiral groove of an automatic winding and unwinding device 101 to form an independent detection assembly; the servo motor and servo wire winder of each detection assembly and the joint servo drive unit of the corresponding mechanical arm 200 are independently connected to the control system to support independent action control of a single assembly and synchronous cooperative control of multiple assemblies.

[0108] The above detection system provided by the embodiment of the application realizes independent precise control of each detection assembly and synchronous cooperative operation of multiple assemblies, ensures that each ultrasonic detection device 400 can be accurately corresponded with a target sounding pipe, completely avoids mutual interference between wires, greatly improves the operation efficiency, alignment accuracy and overall fault tolerance of multi-channel ultrasonic detection, and provides a core structural guarantee for synchronized and batched intelligent detection of cast-in-place piles 600.

[0109] As shown in Figure 3 The embodiment of the application provides a full-automatic intelligent ultrasonic detection method for a cast-in-place pile, which can be applied to the full-automatic intelligent ultrasonic detection system for a cast-in-place pile provided in the above embodiment and includes the following steps S1 to S5.

[0110] In step S1, the control system sets a route to control the walking device 100 to move and real-time collects pile top image data of each cast-in-place pile 600 to be tested; the walking device 100 is provided with an automatic winding and unwinding device 101, and the automatic winding and unwinding device 101 is wound with a wire.

[0111] In step S2, based on the pile top image data, pixel coordinates of pile cores and sounding pipe openings of each cast-in-place pile 600 are identified in real time, and the pixel coordinates are converted into corresponding physical coordinates.

[0112] In step S3, whether the walking device 100 reaches a target position is determined based on a deviation between the physical coordinates corresponding to the pile core and theoretical coordinates of the pile core in a pile drawing.

[0113] Step S4, in the case of reaching the target position, the mechanical arm 200 is controlled to move the ultrasonic detection device 400 to the physical coordinates corresponding to the sounding pipe opening; wherein the mechanical arm 200 is arranged on the walking device 100; the inside of the mechanical arm 200 is provided with a wire channel, and the end of the mechanical arm 200 is provided with a guide clamping groove 201 connected with the wire channel; the ultrasonic detection device 400 is movably arranged in the guide clamping groove 201; and the end of the wire is connected with the ultrasonic detection device 400 through the wire channel.

[0114] Step S5, in the case of reaching the physical coordinates corresponding to the sounding pipe opening, the automatic winding and unwinding device 101 is controlled to lower the wire, so that the ultrasonic detection device 400 vertically enters the inside of the corresponding sounding pipe 500 under the synergistic action of the guide clamping groove 201 and the self weight until the bottom is touched.

[0115] The above steps of the full-automatic intelligent ultrasonic detection method of the cast-in-place pile provided by the embodiment of the present application rely on the hardware architecture and control logic of the foregoing detection system embodiment, and the same features are referred to the foregoing detection system embodiment, effectively solve the technical problems in the conventional ultrasonic detection of the cast-in-place pile, such as large deviation of manual pile positioning and sounding pipe opening, insufficient positioning accuracy of the mechanical arm 200, and easy tilting and blocking of the detection probe during lowering, realize full-process automation of the ultrasonic detection of the cast-in-place pile, such as path planning and pile top image acquisition of the walking device 100, pile core and sounding pipe opening coordinate identification and conversion, automatic pile position confirmation, precise positioning of the mechanical arm 200, and vertical and stable lowering of the detection probe, and greatly improve the positioning accuracy, operation efficiency and process stability of the detection operation, thereby providing core method support for the intelligentization and unmanned landing of the ultrasonic detection of the cast-in-place pile.

[0116] Further, the ultrasonic detection device 400 of the embodiment of the present application takes the ultrasonic transducer as an example, and the configuration process includes: in the standby state, the ultrasonic transducer is preset in the guide clamping groove 201 in the end effector inside the mechanical arm 200. After receiving the “pile position confirmation success” signal, the signal triggers the subsequent automatic ultrasonic detection process. The mechanical arm 200 first moves, accurately aligns the clamping groove outlet with the sounding pipe opening through rotation and extension according to the physical coordinates of each sounding pipe opening. Then, through the release wire action of the servo winch mechanism, the transducer vertically enters the sounding pipe along the preset trajectory from the clamping groove. After the detection is completed, it is pulled back to the clamping groove. The specific implementation steps include:

[0117] (1) Parameterized motion planning. The acquisition device receives the pile body parameters (pile length, diameter, pipe number) from the image recognition system and the data monitoring system. According to the pile length L, the optimal lowering speed V_down (V_down=k / L+b, wherein k and b are efficiency coefficients) is calculated according to the efficiency principle, and the lifting speed V_up is constrained to be less than or equal to 0.5 m / s. At the same time, an S-shaped speed curve is generated to ensure the smooth start and stop of the device.

[0118] (2) Precise depth and synchronous lowering. The transducer is located at the pipe opening under the guidance of the alignment device. Control the N sets of servo winch mechanisms to start synchronously to lower the transducer at a speed of V_down. The high-resolution encoder feeds back the depth value Hi in real time. By comparing the motor driving current (torque) and the depth value Hi in real time, the pile bottom is judged: when all the motor currents jump up and Hi≈L, it is judged that the synchronous bottom is touched, the lowering is stopped, and the pile bottom depth reference Hb is recorded.

[0119] (3) Synchronous lifting and adaptive detection. Start the ultrasonic emission / receiving instrument. The controller drives all servo motors to synchronously lift the transducer at a speed of V_up in the electronic gear synchronous mode. Depth-data binding: the depth information Hi fed back by the encoder and the acoustic parameters (sound time, amplitude) collected by the ultrasonic instrument are bound and stored in real time one by one. Speed adaptation: if the real-time monitored received waveform signal-to-noise ratio is lower than the threshold, V_up is automatically reduced by one step (such as from 0.5 m / s to 0.3 m / s) to obtain more intensive and stable data.

[0120] (4) Neat recovery and state reset. When the transducer is lifted to the pipe opening, the photoelectric sensor is triggered to switch to low speed.

[0121] After the transducer completely leaves the pipe opening, the alignment device is removed, and the automatic winding and unwinding device 101 continues to wind back the cable, and the servo wire arrangement device integrated in the winding drum shaft works synchronously to ensure that the cable is tightly and neatly arranged. The device is reset to the standby state, ready to move to the next pile position for detection.

[0122] In summary, the technical effects that can be achieved by the full-automatic intelligent ultrasonic detection system and method for cast-in-place piles provided by the embodiments of the application include:

[0123] (1) Full-process automatic intelligent detection, realizing industrial upgrading, avoiding human intervention, improving quality and efficiency, and facilitating supervision.

[0124] (2) The pile position is automatically confirmed by the Beidou positioning system, solving the problem of complex, time-consuming, labor-intensive and easy-to-mix pile position and number confirmation in traditional cast-in-place pile ultrasonic detection.

[0125] (3) The number and position of the ultrasonic detection pipes pre-set on the cast-in-place pile are identified by the image recognition system, the ultrasonic transducer is aligned with the ultrasonic detection pipe, and the ultrasonic transducer is lowered and lifted.

[0126] (4) The problem of many devices and high labor intensity in traditional ultrasonic detection is solved by the automatic walking intelligent system.

[0127] (5) The automatic acquisition system solves the problems of traditional detection, such as the out-of-sync and unstable speed of the ultrasonic transducer, which affects the quality of data acquisition, and the traditional hand-cranking reel easily causes the wires to be tangled and affects the detection efficiency.

[0128] (6) The data terminal system can set different ultrasonic transducer quantities, ultrasonic wave speeds, lowering and lifting speeds according to the diameter, length, ultrasonic detection tube quantity, and concrete strength of the cast-in-place pile 600, and can automatically judge the data acquisition quality, thereby solving the problem that the detection quality is subject to the experience of the detection personnel.

[0129] It should be noted that the term "comprising" and variations of the term "comprising", used in the embodiments of the present application, are intended to cover a non-exclusive inclusion. The term "based on" is "based at least in part on". The term "one embodiment" means "at least one embodiment". The term "another embodiment" means "at least one additional embodiment". The term "some embodiments" means "at least some embodiments". The modification of "one", "multiple" in the embodiments of the present application is illustrative and not restrictive, and those skilled in the art should understand that it should be understood as "one or more" unless the context clearly indicates otherwise.

[0130] The steps described in the method embodiments provided by the embodiments of the present application can be executed in different orders and / or in parallel. In addition, the method embodiments can include additional steps and / or omit the execution of the steps shown. The scope of protection of the present application is not limited in this respect.

[0131] The word "embodiment" in the present specification means that the specific features, structures or characteristics described in connection with the embodiments can be included in at least one embodiment of the present application. The presence of this phrase in various places in the specification does not necessarily mean the same embodiment, nor does it mean independence or alternatives to other embodiments. Each embodiment in the present specification is described in a relevant manner, and the same or similar parts between each embodiment are cross-referenced. In particular, for device, apparatus, system embodiments, since they are basically similar to method embodiments, the description is relatively simple, and the relevant parts are referred to the part of the method embodiment.

[0132] The above-described embodiments only express several embodiments of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of protection. It should be noted that for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of protection of the present application. Therefore, the scope of protection of the present application should be subject to the appended claims.

Claims

1. A fully automated intelligent ultrasonic testing system for cast-in-place piles, characterized in that, include: The walking device is equipped with an automatic retraction and deployment device; the automatic retraction and deployment device is wound with a wire. A robotic arm is mounted on the walking device; the robotic arm has a wiring channel inside, and the end of the robotic arm has a guide slot that communicates with the wiring channel; an ultrasonic testing device is movably mounted in the guide slot; the end of the wire is connected to the ultrasonic testing device through the wiring channel. A camera, mounted on the robotic arm, is used to collect real-time image data of the top of each cast-in-place pile. The control system, installed on the walking device, is used to identify the pixel coordinates of the core of each of the cast-in-place piles and the pixel coordinates of the sonic logging pipes embedded in the cast-in-place piles based on the image data of the pile top, and to determine whether the walking device has reached the target position based on the deviation between the physical coordinates obtained by converting the pixel coordinates of the core and the theoretical coordinates of the core in the drawing of the cast-in-place pile. When the walking device moves to the target position, the robotic arm is manipulated to move the ultrasonic testing device to the physical coordinates obtained by converting the pixel coordinates of the acoustic tube opening, and the automatic take-up and put-down device is driven to lower the wire, so that the ultrasonic testing device, under the combined action of the guide slot and its own weight, vertically enters the interior of the corresponding acoustic tube until it touches the bottom. The control system includes: An image processing module, connected to the camera, is used to convert the color image data of the pile top into corresponding grayscale image data and extract the gradient magnitude. The feature extraction module, connected to the image processing module, is used to extract the top surface target contour of each of the cast-in-place piles and the quantized features of each connected component in the top surface target contour based on each of the grayscale image data and the gradient magnitude. The target recognition module, connected to the feature extraction module, is used to identify the pile body region and the sonic logging pipe region of the cast-in-place pile from each of the connected domains based on the quantized features, and to determine the corresponding pixel coordinates of the pile core and the sonic logging pipe region based on the center coordinates of the pile body region and the sonic logging pipe region. The information extraction module is used to extract the theoretical coordinates of the pile core of each pile in the construction coordinate system based on the input pile drawings. The coordinate transformation module, connected to the information extraction module and the target recognition module, is used to convert the pixel coordinates of the pile center and the sonic logging pipe opening into the corresponding physical coordinates under the construction coordinate system based on the Beidou coordinates of the walking device and the transformation matrix between the camera and the walking device and the end effector of the robotic arm. The deviation calculation module, connected to the information extraction module and the coordinate transformation module, is used to calculate the planar deviation distance between the physical coordinates of the pile core and the theoretical coordinates of the pile core; The pile position confirmation module is connected to the deviation calculation module and is used to compare the plane deviation distance with a set deviation threshold. When the plane deviation distance is less than or equal to the set deviation threshold, the module confirms that the walking device has reached the target position.

2. The detection system according to claim 1, characterized in that, The feature extraction module further includes: An edge detection module, connected to the image processing module, is used to extract the top surface target contour of each of the cast-in-place piles based on the grayscale image data, and output a binary edge image with a single pixel width. The repair module, connected to the edge detection module, is used to perform edge morphological closing operations on the binary edge image to obtain an edge image with continuous and complete contours. The contour feature quantization module, connected to the repair module, is used to perform connected component analysis and quantization processing on the edge image and extract the quantization features of each connected component; the quantization features include: size features, position features, morphological features and grayscale features.

3. The detection system according to claim 2, characterized in that, The target recognition module includes: The pile body recognition module is connected to the feature extraction module and is used to select the region with the largest average gray value and the largest area from each of the connected components based on the quantized features, and determine it as the pile body candidate region. The pile body determination module, connected to the pile body recognition module, is used to filter out areas from the pile body candidate areas where both the roundness and equivalent diameter dimensions meet the set range, determine them as pile body areas, and determine the centroid of the pile body area as the pixel coordinates of the corresponding pile center.

4. The detection system according to claim 3, characterized in that, The target recognition module further includes: The pipe opening identification module is connected to the feature extraction module and is used to select regions with an average gray value less than a second set threshold and an area that meets the restriction range from each of the connected components based on the quantized features, and determine them as the initial candidate set of pipe openings. The pipe opening determination module, connected to the pipe opening recognition module, is used to filter areas from the initial candidate set of pipe openings where the circularity, equivalent diameter, center position, and the connection of each center all meet the set range, and determine them as the acoustic pipe opening area, and determine the center coordinates of the acoustic pipe opening area as the pixel coordinates of the corresponding acoustic pipe opening.

5. The detection system according to claim 1, characterized in that, The control system further includes: The multi-source data acquisition module is used to acquire encoder data of the automatic take-up and take-down device, ultrasonic detection data of the ultrasonic detection device, video stream data of the camera, and operating status data of the walking device, the robotic arm and the automatic take-up and take-down device. The display module, connected to the multi-source data acquisition module, is used to visually display the theoretical parameters of each of the cast-in-place piles, the lowering progress of the ultrasonic testing equipment, the ultrasonic testing data, the video footage captured by the camera, and the operating status data; wherein, the lowering progress is calculated based on the encoder data and the pile length in the theoretical data of the cast-in-place pile.

6. The detection system according to claim 1, characterized in that, The automatic take-up and unwinding device includes: a reel, a servo cable tray, and a servo motor; The drum is rotatably mounted on the traveling device; the servo cable guide is mounted on the traveling device and is parallel to the drum; the wire is wound onto the drum after being guided by the servo cable guide; the servo motor is mounted on the traveling device, and its output end is coaxially and fixedly connected to the central axis of the drum; both the servo motor and the servo cable guide are connected to the control system.

7. The detection system according to claim 1, characterized in that, At least two ultrasonic testing devices are provided, and their number corresponds one-to-one with the number of the robotic arm and the automatic take-up and take-down device. Each of the ultrasonic testing devices is connected to an automatic take-up and take-down device via a wire that passes through the wiring channel of the corresponding robotic arm, forming a set of testing components.

8. A fully automated intelligent ultrasonic testing method for cast-in-place piles, applied to the testing system according to any one of claims 1 to 7, characterized in that, Includes the following steps: The control system sets the route to control the movement of the walking equipment and collects real-time image data of the top of each cast-in-place pile to be tested; wherein, the walking equipment is equipped with an automatic retraction device, and the automatic retraction device is wound with a wire. Based on the pile top image data, the pixel coordinates of the pile core and sonic logging pipe opening of each of the cast-in-place piles are identified in real time, and the pixel coordinates are converted into corresponding physical coordinates. Based on the deviation between the physical coordinates of the pile core and the theoretical coordinates of the pile core in the pile drawing, it is determined whether the walking equipment has reached the target position. Upon reaching the target location, the robotic arm is manipulated to move the ultrasonic testing device to the physical coordinates corresponding to the acoustic probe opening. The robotic arm is mounted on a traveling device. An internal wiring channel is provided within the robotic arm, and a guide slot connected to the wiring channel is located at the end of the robotic arm. The ultrasonic testing device is movably mounted within the guide slot. The end of the lead wire is connected to the ultrasonic testing device via the wiring channel. Upon reaching the physical coordinates corresponding to the acoustic tube opening, the automatic take-up and take-down device is operated to lower the wire, so that the ultrasonic testing device, under the combined action of the guide slot and its own weight, vertically enters the interior of the corresponding acoustic tube until it touches the bottom.

Images