A computer vision technology-based bored pile forming quality monitoring system

Abstract

This invention discloses a drilling pile quality monitoring system based on computer vision technology. After the initial concrete pouring is normal, the system calculates the required height for lifting the tremie pipe based on the real-time volume of subsequent concrete grout pouring, and automatically completes the tremie pipe lifting. Simultaneously, four monitoring pipes are arranged at different heights around the pile. The ends of these pipes are transparent observation chambers containing miniature cameras, LED unidirectional light sources, and conical reflectors. Based on computer vision technology, the system automatically identifies potential abnormal events during underwater concrete pouring around the grouting pipes, adjusts the tremie pipe lifting speed and burial depth, and issues alarms to alert on-site technicians, allowing for appropriate remedial measures to prevent pile breakage, interlayer formation, and other pile quality defects caused by excessive lifting speed or insufficient tremie pipe burial depth. This invention significantly improves the automation level of drilling pile grouting operations, reduces manual operation, and thus greatly accelerates construction speed and pile formation efficiency.

Description

Technical Field

[0001] This invention relates to the field of pile foundation construction technology, and in particular to a pile quality monitoring system and construction method for bored cast-in-place piles based on computer vision technology. Background Technology

[0002] Drilled cast-in-place piles are a type of underground engineering project with high quality requirements, numerous construction procedures, and the need to be completed continuously within a short period. Due to their relatively low construction noise and vibration, and less stringent requirements on the geological environment of the foundation, they have been widely used in road, building, and bridge engineering. However, factors such as complex geological conditions at the construction site, varying levels of competence among construction management personnel, unstable quality of materials and equipment, and the difficulty in controlling the quality of underwater concrete pouring make it difficult to guarantee construction quality, significantly impacting pile quality and bearing capacity. If construction failures such as borehole collapse, borehole enlargement, tremie pipe leakage, pipe blockage, or pile breakage occur and are not detected in time, rework after pile completion is labor-intensive, time-consuming, and detrimental to project schedule and cost control. Therefore, monitoring and controlling the quality of drilled cast-in-place piles has always been a challenge in engineering construction and a key focus for all participating parties.

[0003] For example, Chinese patent document CN214401796U discloses a quality control device for bored cast-in-place piles. It proposes using ultrasonic principles, with ultrasonic testing heads installed along the outer wall of a guide tube. With the aid of a computer imaging system, the distribution of concrete density can be viewed on a display screen, allowing workers to adjust the vibration intensity and add concrete in real time. However, this testing method has significant shortcomings. First, the principle of ultrasonic non-destructive testing technology is based on the ultrasonic testing head emitting and absorbing ultrasonic waves. This requires driving a guide tube before pouring, placing a reinforcing cage inside, pouring concrete, and vertically installing the ultrasonic testing head on the outer wall of the guide tube. After pouring, the guide tube is pulled out of the borehole. In practice, if the pile size and length are large and the pouring time is long, the concrete setting progress along the pile height will differ, resulting in strength differences. This may lead to difficulties in pulling the tube, or further increase the risk of pile breakage due to external force. Furthermore, ultrasonic testing results can only reveal density differences within the pile, but cannot reveal the detailed causes of these differences, hindering on-site personnel from taking targeted remedial measures.

[0004] For example, Chinese patent document CN115511808A discloses an underwater concrete quality inspection method based on convolutional neural networks. It proposes using an industrial camera to acquire image datasets of underwater concrete during the pouring process, and then employing artificial intelligence technologies such as convolutional neural networks to correct, enhance, and semantically segment the image data to ultimately obtain the concrete aggregate ratio. This allows for understanding the segregation of concrete in the borehole, providing a basis for the appropriate depth and lifting speed of the guide pipe. The approach is relatively clear and provides a new technical approach for the quality control of bored pile devices. However, the document does not mention the implementation method and equipment for acquiring underwater concrete pouring images, and therefore lacks feasibility at present.

[0005] The construction of bored piles is quite concealed, and the quality of pile formation is affected by many factors such as technical conditions, personnel capabilities, and management level. In actual construction, quality problems such as broken piles and mud inclusions are easy to occur, which are difficult to detect on the ground in time. These problems affect the actual bearing capacity, endanger the safety of the superstructure, and require a large amount of rework later, which will inevitably increase construction costs, delay the construction period, and have adverse social impacts. Summary of the Invention

[0006] This invention provides an artificial intelligence-based drilling and grouting pile quality monitoring system and construction method. It realizes the function of acquiring real-time underwater concrete grouting image data, and based on artificial intelligence, it can automatically judge the current underwater concrete grouting quality, automatically adjust the tremie pipe burial depth and lifting speed, significantly improve the automation level of drilling and grouting pile grouting operations, reduce manual operation links, and thus improve construction speed and pile formation efficiency.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] A pile quality monitoring system for bored cast-in-place piles based on computer vision technology includes a grouting conduit, with several monitoring pipes symmetrically arranged around the grouting conduit and the monitoring pipes fixed around the grouting conduit by supports.

[0009] The lower end of the monitoring tube is a high-hardness transparent observation chamber, and the other end is an independent power supply and signal enhancement module and a data jack. The components at both ends are connected by wires inside the monitoring tube. A data cable is inserted into the data jack at the top and connected to the data processing and control system terminal. The data processing and control system terminal is connected to the control system of the lifting equipment. The lifting equipment is connected to the support through slings. The transparent observation chamber consists of a high-definition camera, a one-way light-emitting LED ring light strip, and a conical reflector.

[0010] In operation, the unidirectional LED ring light strip illuminates the concrete mud around the transparent observation chamber, and the high-definition camera, together with the conical reflector, collects underwater concrete image data and sends it to the data processing and control system terminal.

[0011] Further technology of the present invention:

[0012] Preferably, the grouting conduit is fixed by connecting segmented pipe sections through threaded interfaces, and the monitoring pipe is assembled from a first monitoring pipe section and a second monitoring pipe section. Each pipe section is fixed by connecting threaded interfaces, and the threaded interfaces of the grouting conduit and the monitoring pipe are at the same height.

[0013] Preferably, one end of the first monitoring tube is a high-hardness transparent observation chamber, and the other end is an independent power supply and signal enhancement module and a data jack;

[0014] The second monitoring tube section is the same as the first monitoring tube section at one end, with an independent power supply and signal enhancement module and a data jack, and the other end is a data plug;

[0015] The components at both ends of the first and second monitoring pipe sections are connected by wires.

[0016] When the first monitoring pipe section and the second monitoring pipe section are tightly connected together through a threaded interface, the plug is embedded in the data jack, which serves to transmit data between the pipe sections.

[0017] Preferably, the signal enhancement module, data jack, data plug and the inner wall of the monitoring tube are filled with sealing resin.

[0018] Preferably, the support is divided into a first support fixed on the lowest segment of the pipe section, a second support on the middle pipe section, and a third support at the top.

[0019] The first and second supports are reserved with circular limiting grooves for installing the monitoring tube. In addition to limiting the position, the circular limiting grooves also have a flexible buffer layer on the contact surface between the groove and the monitoring tube. The limiting support structure is located directly below the first support. The transparent observation chamber of the first monitoring tube section passes through the first support and is inserted into the limiting support structure.

[0020] Preferably, the third support is equipped with a drive motor, and the lower end of the third support is installed in a groove. The top of the second monitoring tube section is inserted into the groove and connected to the drive motor. During operation, the system controls the drive motor to drive the monitoring tube to rotate. The contact surface between the limiting support structure and the transparent observation chamber is also provided with a flexible silicone strip. When the monitoring tube rotates, it rubs against the outer wall of the transparent observation chamber.

[0021] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0022] 1. This invention is based on the experience gained in the practice of bored pile engineering. It has an original design of a detection tube installed around the grouting pipe, which realizes the function of acquiring real-time high-definition images of underwater concrete during the grouting process. Through the form of self-rotation friction, it realizes the self-cleaning function of the outer surface of the observation chamber, effectively solving the problems of blurred vision caused by the high viscosity of concrete slurry and wall-protecting mud adhering to the wall, and ensuring the reliability and timeliness of image data acquisition.

[0023] 2. This invention adopts an image recognition technology approach based on artificial intelligence to obtain real-time images of underwater concrete pouring. Compared with the single density distribution information obtained by ultrasonic detection, it can more accurately distinguish the types of abnormal events that affect the quality of pile formation, such as hole collapse, isolated boulders, and seepage of mud and groundwater, providing a basis for on-site technicians to take targeted remedial measures. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram illustrating the working process of a borehole pile quality monitoring system based on computer vision technology, as described in this invention.

[0026] Figure 2 This is a schematic diagram of the installation structure of the guide tube and monitoring tube of a borehole cast-in-place pile quality monitoring system based on computer vision technology according to the present invention.

[0027] Figure 3 This is a schematic diagram of the detailed structure of the first monitoring pipe section of a borehole pile quality monitoring system based on computer vision technology according to the present invention.

[0028] Figure 4 This is a schematic diagram of the detailed structure of the second monitoring pipe section of a borehole pile quality monitoring system based on computer vision technology according to the present invention.

[0029] Figure 5 This is a detailed structural diagram of the third support at the top of a borehole pile quality monitoring system based on computer vision technology, as described in this invention.

[0030] Figure 6 This is a schematic diagram of the structure of the first support at the bottom of a borehole pile quality monitoring system based on computer vision technology, as described in this invention.

[0031] Figure 7This is a detailed structural diagram of the second support in the middle of a borehole pile quality monitoring system based on computer vision technology, as described in this invention.

[0032] Figure 8 This is a detailed structural diagram of the first support in a borehole pile quality monitoring system based on computer vision technology, as described in this invention.

[0033] Figure 9 This is a supplementary side view of the transparent observation chamber structure at the bottom of the monitoring tube of a borehole pile quality monitoring system based on computer vision technology in this invention.

[0034] Figure 10 This is a supplementary schematic cross-section of the monitoring pipe-guide pipe-support structure of a borehole cast-in-place pile quality monitoring system based on computer vision technology in this invention.

[0035] Numbering in the diagram: 1. Conduit, 2. Monitoring tube, 3. Support, 4. Concrete slurry, 5. Mud, 6. Lifting equipment, 7. Data processing and control system terminal, 8. Threaded interface, 1-1 Segmented pipe section, 2-1 First monitoring pipe section, 2-2 Second monitoring pipe section, 2-3 Independent power supply and signal enhancement module, 2-4 Wire, 2-5 Sealing resin, 2-6 Data port, 2-7 Data plug, 3-1 First support, 3-2 Second support, 3-3 Third support, 3-4 Limiting groove, 3-5 Flexible buffer layer, 3-6 Limiting support structure, 2-1-1 Transparent observation chamber, 2-1-2 High-definition camera, 2-1-3 Unidirectional LED ring light strip, 2-1-4 Conical reflector, 3-3-1 Drive motor, 3-3-2 Mounting groove, 3-6-1 Flexible silicone strip. Detailed implementation method:

[0036] To make the technical means, creative features, achieved objectives, and effects of this invention readily understandable, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and are not intended to limit the invention.

[0037] like Figure 1 As shown, a borehole pile quality monitoring system based on computer vision technology is described. The grouting conduit 1 is composed of segmented pipe sections 1-1 screwed together and fixed by threaded interfaces 8. Four monitoring pipes 2 are symmetrically arranged around the grouting conduit 1. The monitoring pipes 2 are fixed around the grouting conduit 1 by supports 3.

[0038] like Figure 2As shown, the four monitoring pipes 2 are assembled from the first monitoring pipe section 2-1 and the second monitoring pipe section 2-2. The first monitoring pipe section 2-1 and the second monitoring pipe section 2-2 have different optional length specifications, which can be adapted according to the length specifications of the segmented pipe section 1-1, so that the threaded interface of the grouting conduit 1 and the monitoring pipe 2 are at the same height, so as to facilitate the installation and inspection of the pipe section before construction and the disassembly of the pipe section during the lifting process of the conduit during construction.

[0039] like Figure 3-4 As shown, one end of the first monitoring tube section 2-1 is a high-hardness transparent observation chamber 2-1-1, and the other end is an independent power supply and signal enhancement module 2-3 and a data jack 2-6.

[0040] The second monitoring tube section 2-2 is identical to the first monitoring tube section 2-1 at one end, equipped with an independent power supply and signal enhancement module 2-3 and a data jack 2-6, and a data plug 2-7 at the other end. The components at both ends of the first and second monitoring tube sections 2-1 are connected by wires 2-4. When the first and second monitoring tube sections 2-1 and 2-2 are tightly connected together via threaded interfaces 8, the plug 2-7 can be reliably and securely embedded in the data jack 2-6, facilitating data transmission between the tube sections. A data cable is inserted into the top data jack 2-6 and connected to the data processing and control system terminal 7. The data processing and control system terminal 7 is connected to the control system of the lifting equipment 6. The lifting equipment is connected to the support 3 via slings. Specifically, the space between the independent power supply and signal enhancement module 2-3, the data jack 2-6, the data plug 2-7, and the inner wall of the monitoring tube 2 is filled with sealing resin 2-5 for waterproofing and fixation.

[0041] like Figure 3 As shown, the transparent observation chamber 2-1-1 consists of a high-definition camera 2-1-2, a one-way emitting LED ring light strip 2-1-3, and a conical reflector 2-1-4. In operation, the one-way emitting LED ring light strip 2-1-3 illuminates the concrete mud surrounding the transparent observation chamber 2-1-1. The high-definition camera 2-1-2, in conjunction with the conical reflector 2-1-4, can collect underwater concrete image data and upload it to the data processing and control system terminal 7.

[0042] like Figure 5 , Figure 6 , Figure 7 As shown, the support 3 is divided into a first support 3-1 fixed on the lowest segment of the pipe section, a second support 3-2 on the middle pipe section, and a third support 3-3 at the top.

[0043] like Figure 6 , Figure 7 , Figure 8 , Figure 9 and Figure 10As shown, the first support 3-1 and the second support 3-2 have four circular limiting grooves 3-4 reserved for installing the monitoring tube 2. In addition to their limiting function, the circular limiting grooves 3-4 also have a flexible buffer layer 3-5 on the contact surface between the groove and the monitoring tube, allowing the monitoring tube 2 to rotate around its own axis within the limiting grooves 3-4 and reducing frictional resistance. The first support 3-1 is further characterized by having a limiting support structure 3-6 directly below it, in addition to the limiting grooves 3-4, ensuring smooth rotation of the monitoring tube 2 around its axis and bearing its own weight to prevent it from falling. The contact surface between the limiting support structure 3-6 and the transparent observation chamber 2-1-1 is also provided with a flexible silicone strip 3-6-1. When the monitoring tube 2 rotates, it rubs against the outer wall of the transparent observation chamber 2-1-1, immediately cleaning any residual slurry adhering to the outer surface of the observation chamber during the lifting process, ensuring a clear view for the camera.

[0044] like Figure 5 As shown, the third support 3-3 is characterized in that the four circular limiting grooves are replaced with mounting grooves 3-3-2 with internal threads, connected to the drive motor 3-3-1, which can drive the monitoring tube 2 to rotate around its axis. During operation, the system controls the drive motor 3-3-1 to drive the monitoring tube 2 to rotate at a certain speed. Friction with the flexible silicone strip 3-6-1 on the fixed welded support 3-1 achieves the self-cleaning function of the outer wall of the transparent observation chamber 2-1-1, ensuring a clear view for the camera.

[0045] The construction method of a borehole pile quality monitoring system based on computer vision technology in this example is carried out according to the following steps:

[0046] Step 1: Preparations before grouting, including drilling, hole cleaning, equipment positioning, and installation and hoisting of the reinforcing cage.

[0047] Step 2: Following the method described in claim 2, select the appropriate length specifications for the end monitoring pipe 2-1 and the extension relay pipe 2-2, assemble the grouting conduit 1 and monitoring pipe 2 on the ground, and power them on for inspection. Debug the video and control signals.

[0048] Step 3: Lower the grouting pipe 1 and monitoring pipe 2, calculate the grouting volume according to the specifications, and complete the first grouting.

[0049] Step 4: The drive motor 3-3-1 rotates the monitoring tube 2. Simultaneously, the high-definition camera 2-1-2 and the unidirectional LED ring light strip 2-1-3 in the end monitoring tube 2-1 are powered on to acquire and record real-time image data of the grouting situation near the outlet of the grouting pipe 1. This process determines changes in the concrete aggregate and mortar mix ratio around the transparent observation chamber 2-1-1, identifying abnormal events affecting pile quality such as borehole collapse, isolated boulders, and seepage of mud and groundwater. If any of these abnormalities occur, the lifting speed and burial depth of the grouting pipe will be adjusted immediately, and on-site technicians will be alerted to take appropriate remedial measures.

[0050] The aforementioned anomaly event identification specifically includes:

[0051] Step 1: After distortion correction and cropping of the video images acquired by the high-definition camera (2-1-2), the acquired images are optimized and augmented using a generative adversarial network (DE-GAN, etc.), and after annotation, they are divided into training set, validation set and test set.

[0052] Step 2: Extract and segment the R, G, B three-channel color information and gray-level co-occurrence matrix statistics of the mud and aggregate regions of the image as key feature information. After training the model using Support Vector Machine (SVM) and obtaining the model parameters, perform semantic segmentation and label output (aggregate region, slurry / mud region, etc.) on the acquired underwater concrete pouring image.

[0053] Step 3: Based on the segmented output image, obtain and record the changes in the proportions of concrete aggregate, grout / mud, and other components during the pouring process.

[0054] Step 4: Input the processed image data into a deep learning recognition network (YOLOv5, etc.) to identify the current perfusion quality and determine the type of abnormal output event.

[0055] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The descriptions in the foregoing invention and specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A computer vision technology-based bored pile piling quality monitoring system, characterized in that: It includes a grouting conduit (1), and several monitoring pipes (2) are symmetrically arranged around the grouting conduit. The monitoring pipes (2) are fixed around the grouting conduit (1) by supports (3). The lower end of the monitoring tube (2) is a high-hardness transparent observation chamber (2-1-1), and the other end is an independent power supply and signal enhancement module (2-3) and a data jack (2-6). The components at both ends are connected by wires (2-4) inside the monitoring tube (2). A data cable is inserted into the data jack (2-6) at the top and connected to the data processing and control system terminal (7). The data processing and control system terminal (7) is connected to the control system of the lifting equipment (6). The lifting equipment is connected to the support (3) by slings. The transparent observation chamber (2-1-1) consists of a high-definition camera (2-1-2), a unidirectional LED ring light strip (2-1-3), and a conical reflector (2-1-4). A one-way emitting LED ring light strip (2-1-3) illuminates the concrete mud around the transparent observation chamber (2-1-1), and a high-definition camera (2-1-2) works with a conical reflector (2-1-4) to collect underwater concrete image data to the data processing and control system terminal (7). The grouting conduit (1) is fixed by connecting the segmented pipe sections (1-1) through the threaded interface (8). The monitoring pipe (2) is assembled from the first monitoring pipe section (2-1) and the second monitoring pipe section (2-2). Each pipe section is fixed by connecting the threaded interface (8). The threaded interfaces of the grouting conduit (1) and the monitoring pipe (2) are at the same height. The support (3) is divided into a first support (3-1) fixed on the lowest segment of the pipe section, a second support (3-2) on the middle pipe section and a third support (3-3) at the top. The first support (3-1) and the second support (3-2) are reserved with circular limiting grooves (3-4) for installing the monitoring tube (2). In addition to limiting the position, the circular limiting groove (3-4) is also provided with a flexible buffer layer (3-5) on the contact surface between the groove and the monitoring tube. The limiting support structure (3-6) is located directly below the first support (3-1). The transparent observation chamber (2-1-1) of the first monitoring tube section (2-1) passes through the first support (3-1) and is connected to the limiting support structure (3-6). The third support (3-3) is equipped with a drive motor (3-3-1), and the lower end of the third support (3-3) is equipped with a mounting groove (3-3-2). The top of the second monitoring tube section (2-2) is inserted into the mounting groove (3-3-2) and connected to the drive motor (3-3-1). When working, the system controls the drive motor (3-3-1) to drive the monitoring tube (2) to rotate. The contact surface between the limiting support structure (3-6) and the transparent observation chamber (2-1-1) is also provided with a flexible silicone strip (3-6-1). When the monitoring tube (2) rotates, it rubs against the outer wall of the transparent observation chamber (2-1-1).

2. The borehole pile quality monitoring system based on computer vision technology as described in claim 1, characterized in that: The first monitoring tube section (2-1) has a high-hardness transparent observation chamber (2-1-1) at one end and an independent power supply and signal enhancement module (2-3) and data jack (2-6) at the other end. The second monitoring tube section (2-2) is the same as the first monitoring tube section (2-1) at one end, and is equipped with an independent power supply and signal enhancement module (2-3) and a data jack (2-6), and the other end is a data plug (2-7). The components at both ends of the first monitoring tube section (2-1) and the second monitoring tube section (2-2) are connected by a wire (2-4); When the first monitoring tube section (2-1) and the second monitoring tube section (2-2) are tightly connected together through the threaded interface (8), the plug (2-7) is embedded in the data jack (2-6) to transmit data between the tube sections.

3. The computer vision technology-based bored pile quality monitoring system according to claim 2, wherein: The signal enhancement module (2-3), data jack (2-6), data plug (2-7) and the inner wall of the monitoring tube (2) are filled with sealing resin glue (2-5).

Images