A foundation pit monitoring and measuring system and a monitoring method

Abstract

The present invention relates to the technical field of foundation pit construction monitoring, and provides a foundation pit monitoring and measuring system and a monitoring method, which include an AGV intelligent vehicle-mounted system and a central processing system. The AGV intelligent vehicle-mounted system has an AGV intelligent vehicle, which can drive along the runway around the foundation pit according to the instructions of the central processing system. A camera module is provided on the vehicle for collecting foundation pit image information. The central processing system can receive and process the foundation pit image information collected by the camera module, perform deformation detection and calculation, obtain the deformation analysis result, and is communicatively connected to the AGV intelligent vehicle-mounted system, and navigates the running track of the AGV intelligent vehicle according to the three-dimensional model coordinate data of the foundation pit and the positioning system. The foundation pit monitoring and measuring system described in this application uses the AGV intelligent vehicle-mounted system to carry the camera module and drive along the guiding runway around the foundation pit, and combines the image recognition and analysis technology of the central processing system to realize the comprehensive and automatic monitoring of the foundation pit.

Description

Technical Field

[0001] The present invention relates to the technical field of foundation pit construction monitoring, and particularly relates to a foundation pit monitoring and measuring system and a monitoring method. Background Art

[0002] In the field of foundation pit monitoring, the safety and stability of the foundation pit are of crucial importance. Traditional monitoring methods usually rely on manual inspections and fixed-point monitoring. These methods are not only inefficient but also difficult to achieve real-time monitoring, and are significantly affected by personnel experience and subjective factors. Although the progress of sensor technology and image processing technology in recent years has promoted the development of foundation pit monitoring towards automation and intelligence, the existing technologies still face many challenges.

[0003] The traditional manual inspection method is less efficient and difficult to comprehensively reflect the overall deformation of the foundation pit. Although fixed-point monitoring can provide some data, its monitoring points are limited and it is impossible to achieve comprehensive monitoring of the entire foundation pit. In early projects, engineers often used tools such as level gauges and theodolites for fixed-point measurement. These methods are time-consuming and the results are easily affected by human factors.

[0004] Currently, displacement sensors, stress sensors, etc. are widely used in foundation pit monitoring, which can collect deformation data in real time and provide support for safety assessment. However, the deployment cost of sensors is usually high, and they may be affected by environmental factors (such as temperature, humidity, etc.), resulting in errors in the monitoring data. In addition, the monitoring range of sensors is limited and it is difficult to achieve comprehensive monitoring of large and complex foundation pit projects.

[0005] With the progress of image processing technology, foundation pit monitoring solutions based on image recognition have gradually emerged. Such technologies can provide new solutions for foundation pit monitoring by collecting image information in real time and applying it to deformation detection. For example, the invention patent with the publication number CN107119657B discloses a foundation pit monitoring method based on visual measurement. It sets multiple spikes or steel bars with crosshairs as detection points, and after obtaining the digital color image, it performs gray conversion. By removing noise, identifying multi-target areas, and calculating the center of the targets, the optimal coordinates of multiple target centers are obtained. A foundation pit deformation prediction model is established using a BP neural network to estimate and warn of possible structural deformations of the foundation pit. Although the above solution realizes the monitoring of foundation pit deformation to a certain extent, its monitoring process mainly relies on the image collection of a fixed camera module and subsequent software algorithms for processing. Although it realizes automation to a certain extent, it still lacks flexibility.

[0006] Therefore, in various existing monitoring technologies, recognition accuracy, stability, and monitoring cost are still the main limiting factors. There are still many problems and limitations in the existing foundation pit monitoring technologies, which are difficult to meet the increasing monitoring requirements of foundation pit projects. Summary of the Invention

[0007] In view of this, the present invention aims to solve the technical problems existing in the existing foundation pit monitoring technology, such as low monitoring efficiency, limited monitoring points, and great influence by personnel experience and subjective factors. A foundation pit monitoring and measurement system and a monitoring method are proposed. By providing a foundation pit monitoring and measurement system and method, an AGV intelligent vehicle system is used to carry a camera module to drive along the guiding runway around the foundation pit, and combined with the image recognition and analysis technology of the central processing system, comprehensive and automated monitoring of the foundation pit is realized.

[0008] To achieve the above object, the technical solution of the present invention is realized as follows:

[0009] A foundation pit monitoring and measurement system, comprising:

[0010] An AGV intelligent vehicle system, including an AGV intelligent vehicle, which can drive along the runway around the foundation pit according to the instructions of the central processing system. A camera module is arranged on the AGV intelligent vehicle, and the camera module can collect image information of the foundation pit;

[0011] A central processing system, which can receive the image information of the foundation pit collected by the camera module, and combined with the image recognition and analysis system, perform deformation detection and deformation calculation on the image information of the foundation pit to obtain the foundation pit deformation analysis result;

[0012] The central processing system is communicatively connected to the AGV intelligent vehicle system and can navigate the running track of the AGV intelligent vehicle according to the initial standard three-dimensional model coordinate data of the foundation pit and the positioning system.

[0013] Furthermore, a guiding runway is arranged around the foundation pit, and a guiding module and a driving module are integrated on the AGV intelligent vehicle. The position-type PID algorithm is adopted to drive the AGV intelligent vehicle to move along the foundation pit guiding.

[0014] Furthermore, a lifting device is arranged on the AGV intelligent vehicle. A first connecting frame is arranged at the upper end of the lifting device, a telescopic mechanism is arranged on the first connecting frame, the plane where the telescopic direction of the telescopic mechanism is located is perpendicular to the lifting direction of the lifting device, a first mounting frame is arranged at the cantilever end where the telescopic mechanism extends, and the camera module is fixed on the first mounting frame.

[0015] Furthermore, the first mounting bracket includes a first connecting member, an adapter and a second connecting member, a first driving device is arranged on the first connecting member, one end of the adapter is connected to the first driving device, and the first driving device can drive the adapter to rotate in a horizontal direction, a second connecting member is arranged at an end of the adapter away from the first driving device, two connecting arms for connecting a camera module are arranged on the second connecting member, and a second driving device is arranged on one of the connecting arms, which can drive the camera module to rotate in a vertical direction.

[0016] Furthermore, the foundation pit monitoring and measurement system also includes a target system, and the target system includes a plurality of target devices, and the plurality of target devices are arranged in a matrix on the side wall of the foundation pit.

[0017] Furthermore, a three-dimensional modeling module and a graphic deformation monitoring module are set in the central processing system. The three-dimensional modeling module can perform image processing according to the image transmitted by the AGV intelligent vehicle system and the parameter information of the image to construct a measured three-dimensional model of the foundation pit; the graphic deformation monitoring module includes a preset graphic deformation monitoring model, and the graphic deformation monitoring model uses a three-dimensional convolutional neural network fused with an attention mechanism to detect deformed targets and obtain foundation pit deformation characteristic values ​​including lateral displacement, longitudinal displacement and tilt displacement.

[0018] Furthermore, the AGV intelligent vehicle-mounted system uses the camera module to take an oblique photograph of the foundation pit, processes the acquired image, obtains the characteristic density distribution value of the foundation pit, performs aerial triangulation on the pixel coordinates in the image according to the key structural feature points and the position and posture measurement system information of the camera module, obtains the three-dimensional position relative to the foundation pit when the image is taken and the three-dimensional coordinate relationship between the key structural feature points, generates sparse point cloud data corresponding to the foundation pit based on the three-dimensional position and the three-dimensional coordinate relationship, matches the key structural feature points in the sparse point cloud data with the image, generates dense point cloud data, uses spline curves to connect the characteristic pixel points, obtains the initial three-dimensional model network of the foundation pit, and connects the characteristic pixel points based on the initial three-dimensional model network. The three-dimensional map of the environment is constructed by connecting two triangles to form a triangular surface, and the target three-dimensional model grid is obtained. The target three-dimensional model grid is reconstructed and analyzed to obtain the initial three-dimensional model surface corresponding to the foundation pit. Based on the initial three-dimensional model network and the initial three-dimensional model surface of the foundation pit, a measured three-dimensional model of the foundation pit is constructed. The graphic deformation monitoring module can receive the measured three-dimensional model of the foundation pit output by the three-dimensional modeling module, obtain the three-dimensional structural information of the foundation pit monitoring, and identify the deformation area and deformation degree of the foundation pit by comparing the difference between the current three-dimensional model and the initial standard three-dimensional model in combination with the preset graphic deformation monitoring model. For the detected deformation area, the three-dimensional coordinates of the deformation area are accurately measured and calculated, and its lateral displacement, longitudinal displacement and inclination displacement are calculated and output.

[0019] Further, the central processing system further includes an early warning module. The foundation pit deformation characteristic values of the lateral displacement, longitudinal displacement, and tilt displacement output by the graphic deformation monitoring model are compared with the corresponding deformation thresholds. When at least one of the foundation pit deformation characteristic values is greater than the corresponding deformation threshold, the early warning mechanism is immediately activated to issue an early warning message.

[0020] Compared with the prior art, the foundation pit monitoring and measuring system of the present invention has the following advantages:

[0021] (1) The foundation pit monitoring and measuring system of the present invention, through the automated navigation operation of the AGV intelligent vehicle, maximally realizes the full coverage of the foundation pit area, avoids the blind areas existing in the traditional monitoring methods. At the same time, based on the integration of the image recognition and analysis technology, the analysis of the monitoring data is more scientific and systematic, can quickly reflect the deformation situation of the foundation pit, provides a reliable basis for safety early warning, can be quickly implemented in different construction environments, thereby reducing the monitoring cost and improving the timeliness of the data, and is helpful for the safety control of the foundation pit project.

[0022] (2) The foundation pit monitoring and measuring system of the present invention has high flexibility and adaptability, can be quickly implemented in different construction environments, reduces the monitoring cost, and significantly improves the safety control level of the foundation pit project.

[0023] Another object of the present invention is to propose a foundation pit monitoring method, which is applied to the foundation pit monitoring and measuring system as described above, and includes the following steps:

[0024] S1: Set a guiding runway around the foundation pit. The guiding runway is arranged at an equal distance from the side wall of the foundation pit along the circumferential direction of the foundation pit.

[0025] S2: The AGV intelligent vehicle in the AGV intelligent vehicle system can navigate to the guiding runway to check whether the lifting device, telescopic mechanism, and camera module are in normal working conditions.

[0026] S3: Start the AGV intelligent vehicle, and under the instruction of the central processing system, idle around the guiding runway once according to the Beidou satellite positioning module and the position type PID algorithm inside it for path verification and equipment status confirmation.

[0027] S4: After the path verification and equipment status confirmation, adjust the working positions of the lifting device, telescopic mechanism, and the first mounting rack in combination with the initial standard three-dimensional model coordinate data in the central processing system, and adjust the position and attitude measurement system information of the camera module.

[0028] S5: Start the AGV intelligent vehicle again. The camera module continuously collects the real-time image information of the foundation pit during the driving process and sends it to the central processing system through the communication system.

[0029] S6: The central processing system receives the image information, processes the image using image recognition and analysis technology, conducts 3D modeling processing on the foundation pit image through the 3D modeling module and the graphic deformation monitoring module, performs deformation analysis, identifies the deformation area and the degree of deformation, calculates the lateral displacement, longitudinal displacement, and tilt displacement of the deformation area, and outputs the deformation characteristic values.

[0030] S7: The warning module of the central processing system compares the deformation characteristic values with the deformation thresholds. When at least one deformation characteristic value is greater than the corresponding deformation threshold, the warning mechanism is immediately activated.

[0031] Further, in step S4, the shooting direction of the camera module is set perpendicular to the running direction of the AGV intelligent vehicle, and the camera module takes an inclined shot of the foundation pit. The upper edge of the lens of the camera module covers the upper edge of the foundation pit and extends to the position where the guiding runway is located, and the lower edge of the lens of the camera module covers the lower edge of the foundation pit and extends to the bottom surface of the foundation pit.

[0032] The foundation pit monitoring and measuring system and the monitoring method described in this application improve the efficiency and accuracy of the monitoring work, effectively avoid safety accidents caused by the deformation of the foundation pit, provide an important guarantee for the safety of personnel and property at the construction site, and thus improve the level of the entire project management and the emergency response ability. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0034] Figure 1 is a schematic structural diagram of the foundation pit monitoring and measuring system described in the embodiment of the present invention;

[0035] Figure 2 is a schematic structural diagram of the AGV intelligent vehicle system for image acquisition in the foundation pit monitoring and measuring system described in the embodiment of the present invention;

[0036] Figure 3 is a schematic structural diagram of the AGV intelligent vehicle running in the guiding runway in the foundation pit monitoring and measuring system described in the embodiment of the present invention;

[0037] Figure 4 is a schematic structural diagram of the camera module installed on the first mounting rack described in the embodiment of the present invention;

[0038] Figure 5 is Figure 4 a side view structural diagram of the structure shown in

[0039] Figure 6 is a schematic flow diagram of the foundation pit monitoring method described in the embodiment of the present invention;

[0040] Description of the reference numerals:

[0041] 100, AGV intelligent vehicle system; 200, central processing system; 300, foundation pit; 1, AGV intelligent vehicle; 2, camera module; 3, lifting device; 4, first connecting frame; 5, telescopic mechanism; 6, first mounting frame; 61, first connecting piece; 62, adapter; 63, second connecting piece; 64, first driving device; 65, second driving device; 7, target device; 8, guiding runway. Detailed implementation manners

[0042] In order to make the technical means, objectives and effects of the present invention easy to understand, the embodiments of the present invention will be described in detail below with reference to the specific drawings.

[0043] It should be noted that all the terms indicating directions and positions in the present invention, such as: "upper", "lower", "left", "right", "front", "rear", "vertical", "horizontal", "inner", "outer", "top", "bottom", "lateral", "longitudinal", "center", etc., are only used to explain the relative position relationship and connection situation between components in a specific state (as shown in the drawings), and are only for the convenience of describing the present invention, rather than requiring the present invention to be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation to the present invention. In addition, the descriptions involving "first", "second", etc. in the present invention are only for descriptive purposes, and cannot be construed as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features.

[0044] In the description of the present invention, unless otherwise clearly defined and limited, the terms "installation", "connection", and "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection; it can be directly connected, or indirectly connected through an intermediate medium, and it can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0045] In the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic descriptions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0046] Such as Figures 1 - 6As shown in the figure, the present application discloses a foundation pit monitoring and measuring system, including:

[0047] An AGV intelligent vehicle system 100, including an AGV intelligent vehicle 1, which can travel along the runway around the foundation pit 300 according to the instructions of the central processing system. A camera module 2 is arranged on the AGV intelligent vehicle 1, and the camera module 2 can collect image information of the foundation pit 300.

[0048] A central processing system 200, which can receive the image information of the foundation pit 300 collected by the camera module 2, combine with the image recognition and analysis system, perform deformation detection and deformation calculation on the image information of the foundation pit 300, and obtain the foundation pit deformation analysis result.

[0049] The central processing system 200 is communicatively connected to the AGV intelligent vehicle system 100, and can navigate the running track of the AGV intelligent vehicle 1 according to the initial standard three-dimensional model coordinate data of the foundation pit 300 and the positioning system.

[0050] The present application designs a foundation pit monitoring and measuring system integrating the central processing system 200 and the AGV intelligent vehicle system 100, which can independently and efficiently execute the monitoring and measuring tasks of the foundation pit 300. The AGV intelligent vehicle 1 in the AGV intelligent vehicle system 100 moves along the set runway according to the instructions of the central processing system. The camera module 2 equipped on the vehicle continuously collects real-time image information of the foundation pit 300 during operation. The collected image information is sent to the central processing system 200 through the communication system. The central processing system 200 uses an advanced image recognition and analysis system to process the image data, perform deformation detection and deformation calculation, so as to obtain the deformation analysis result of the foundation pit 300. In order to improve the accuracy and comprehensiveness of monitoring, the central processing system 200 can also combine the coordinate data of the initial standard three-dimensional model of the foundation pit 300 and the positioning system of the AGV intelligent vehicle 1 to perform real-time navigation on the running track of the AGV intelligent vehicle 1, ensuring that the camera module 2 in the AGV vehicle system can cover every corner of the foundation pit 300 for monitoring. During this process, the AGV vehicle system can flexibly adjust the operation mode to adapt to different foundation pit conditions, improving the flexibility and efficiency of monitoring.

[0051] The foundation pit monitoring and measuring system disclosed in the present application realizes the full coverage of the foundation pit area to the greatest extent through the automated navigation operation of the AGV intelligent vehicle, avoiding the blind areas existing in the traditional monitoring methods. At the same time, based on the integration of image recognition and analysis technology, the analysis of monitoring data is more scientific and systematic, can quickly reflect the deformation of the foundation pit, provide a reliable basis for safety warning, can be quickly implemented in different construction environments, thereby reducing the monitoring cost and improving the real-time nature of data, and contributing to the safety control of foundation pit projects.

[0052] In the example of this application, a guiding runway 8 is arranged around the foundation pit 300. A guiding module and a driving module are integrated on the AGV intelligent vehicle 1. The positional PID algorithm is adopted to drive the AGV intelligent vehicle 1 to move along the guidance of the foundation pit 300. As a preferred example of this application, a flat road around the foundation pit 300 is selected as the guiding runway for the AGV intelligent vehicle 1. Preferably, the guiding runway 8 is equidistantly arranged at a position 2-3 m away from the side wall of the foundation pit. A guiding module and a driving module are arranged inside the AGV intelligent vehicle 1. Combining with the positional PID algorithm, the driving direction and speed of the vehicle can be accurately controlled to ensure that the vehicle runs stably along the guiding runway. Relying on the internal Beidou satellite positioning module, high-precision positioning information is provided for the AGV intelligent vehicle, enabling it to run along the guiding runway.

[0053] This setting optimizes the navigation task of the AGV intelligent vehicle 1, ensuring the driving accuracy and stability of the AGV intelligent vehicle. When the AGV intelligent vehicle performs the foundation pit monitoring and measurement task, it can efficiently and accurately run along the preset route without manual intervention, greatly reducing the operation complexity and labor cost, providing a solid foundation for subsequent image recognition analysis and deformation calculation, and improving the efficiency and accuracy of foundation pit monitoring and measurement.

[0054] As a preferred example of this application, a lifting device 3 is arranged on the AGV intelligent vehicle 1. A first connecting frame 4 is arranged at the upper end of the lifting device 3. A telescopic mechanism 5 is arranged on the first connecting frame 4. The plane where the telescopic direction of the telescopic mechanism 5 is located is perpendicular to the lifting direction of the lifting device 3. A first mounting frame 6 is arranged at the cantilever end where the telescopic mechanism 5 extends. The camera module 2 is fixed on the first mounting frame 6. In the example of this application, by arranging the lifting device 3 and the telescopic mechanism 5 on the AGV intelligent vehicle 1, the lifting device 3 is used to adjust the vertical height of the camera module 2, and the telescopic mechanism 5 can be telescoped in the horizontal direction. The combination of the two enables the camera module 2 to be flexibly adjusted in three-dimensional space to easily cope with various complex construction environments. When the AGV intelligent vehicle 1 navigates according to the actual situation of the environment, the camera module 2 can quickly adjust to the best observation position and angle according to actual needs. Whether monitoring the upper edge or the depth of the foundation pit, clear and accurate image information can be obtained, realizing the full coverage of the foundation pit 300, thereby providing accurate data support. The freedom and adaptability of this combination enable the AGV intelligent vehicle 1 not only to better cope with diverse working environments but also to ensure the smooth progress of the monitoring task during the complex and dynamically changing foundation pit construction process.

[0055] Through the combination of the lifting device 3 and the telescopic mechanism 5, the camera module 2 can be accurately adjusted to the optimal observation position and angle, ensuring full coverage of any foundation pit 300 structure, enabling it to quickly respond to various construction conditions, improving monitoring efficiency and flexibility in practical applications.

[0056] As a preferred example of the present application, the first mounting frame 6 includes a first connecting member 61, an adapter 62, and a second connecting member 63. A first driving device 64 is arranged on the first connecting member 61. One end of the adapter 62 is connected to the first driving device 64, and the first driving device 64 can drive the adapter 62 to rotate in the horizontal direction. A second connecting member 63 is arranged at one end of the adapter 62 away from the first driving device 64. Two connecting arms for connecting the camera module 2 are arranged on the second connecting member 63, and a second driving device 65 is arranged on one of the connecting arms, which can drive the camera module 2 to rotate in the vertical direction. In the example of the present application, by designing the structure of the first mounting frame 6 including the first connecting member 61, the adapter 62, the second connecting member 63, and the first driving device 64 and the second driving device 65, it is ensured that the camera module 2 has flexible multi-directional adjustment capabilities during the actual detection process, so that the camera module 2 can not only freely adjust the shooting angle in the horizontal direction, but also achieve accurate shooting angle setting in the vertical direction. During the actual detection process, there is no need to move the entire AGV smart car 1 or its main structure. The shooting angle of the camera module 2 can be quickly and efficiently adjusted through the cooperation of the first drive device 64 and the second drive device 65, so that the AGV smart car 1 can perform multi-angle shooting of suspected deformed areas when monitoring foundation pit deformation, providing necessary image data for subsequent three-dimensional modeling of foundation pit deformation characteristic value verification, thereby improving monitoring efficiency and accuracy and providing intuitive and reliable data support for construction personnel.

[0057] This design can significantly improve the flexibility and practicality of the camera module 2 during the detection process. Without changing the overall transfer structure, the camera module 2 can achieve rapid horizontal and vertical angle settings through a multi-degree-of-freedom adjustment mechanism, thereby achieving multi-angle shooting and obtaining comprehensive and accurate image information, providing intuitive and reliable data support for construction personnel, and laying a solid foundation for subsequent decision-making analysis.

[0058] As a preferred example of the present application, the foundation pit monitoring and measuring system further includes a target system, and the target system includes a number of target devices 7, and multiple target devices 7 are arranged in a matrix on the side wall of the foundation pit 300. In the example of the present application, by further introducing the target system and combining image recognition technology for foundation pit deformation monitoring, the target device 7 is a circular or square stainless steel sheet, with a diameter or side length of about 50 mm to 100 mm and a thickness of about 2 mm to 5 mm, and it is fixed on the side wall of the foundation pit 300 through embedded parts or high-performance epoxy resin adhesives. During daily dynamic monitoring, the camera module 2 can use multiple target devices 7 of the target system as key structural feature points of the image. By comparing the position changes of the three-dimensional modeling of the target device 7, the deformation information of the foundation pit is extracted by using image recognition technology, further realizing the accurate monitoring of the foundation pit deformation and reducing the error interference in the overall monitoring process.

[0059] This setting realizes the comprehensive monitoring of the foundation pit deformation with a wide coverage range and high precision by introducing the target devices 7 arranged discontinuously in a matrix, so that even minor deformations can be quickly and accurately captured, making the image recognition process more simple and effective. In subsequent real-time monitoring, the deformation situation of the foundation pit can be quickly and accurately analyzed, providing intuitive and reliable deformation data support for construction personnel. In the example of the present application, during the specific use of the target device 7, the target device 7 needs to be regularly maintained to avoid damage, oil stain or mud pollution. If the target device 7 needs to be replaced, the foundation pit image after replacing the target device 7 needs to be photographed and modeled again, and the three-dimensional modeling image data of the foundation pit deformation needs to be updated.

[0060] As a preferred example of this application, the image recognition and analysis system in the central processing system 200 includes a three-dimensional modeling module and a graphic deformation monitoring module. The three-dimensional modeling module can perform image processing based on the images transmitted by the AGV intelligent vehicle system 100 and the parameter information of the images to construct a measurement three-dimensional model of the foundation pit 300. The graphic deformation monitoring module includes a preset graphic deformation monitoring model, which uses a three-dimensional convolutional neural network (3D CNN) integrated with an attention mechanism (SE, Squeeze-and-Excitatio) for deformation target detection to obtain foundation pit deformation eigenvalue including lateral displacement, longitudinal displacement, and tilt displacement. In the example of this application, the central processing system 200 realizes precise monitoring and deformation analysis of the foundation pit by integrating the three-dimensional modeling module and the graphic deformation monitoring module. When the AGV intelligent vehicle system 100 captures images of the foundation pit 300, the shooting direction of the camera module 2 on the AGV intelligent vehicle system 100 is set perpendicular to the running direction of the AGV intelligent vehicle 1, and the camera module 2 is tilted to shoot the foundation pit 300 through the lifting device 3, the telescopic mechanism 5, and the first mounting bracket 6. The acquired images are processed to obtain the characteristic density distribution value of the foundation pit 300. According to the key structural feature points (such as foundation pit inflection points, significant markers, stratification interfaces, target devices 7, etc.) and the position and attitude measurement system information of the camera module 2, the aerial triangulation processing is performed on the pixel point coordinates in the image to obtain the three-dimensional position relative to the foundation pit 300 and the three-dimensional coordinate relationship between the key structural feature points during image shooting. Based on the three-dimensional position and the three-dimensional coordinate relationship, sparse point cloud data corresponding to the foundation pit 300 is generated. The key structural feature points in the sparse point cloud data are matched with the images to generate dense point cloud data. The characteristic pixel points are connected by spline curves to obtain the initial three-dimensional model network of the foundation pit 300. Based on the initial three-dimensional model network, the characteristic pixel points are connected pairwise to form triangular faces to obtain the target three-dimensional model grid. The target three-dimensional model grid is reconstructed and analyzed to obtain the initial three-dimensional model surface corresponding to the foundation pit. Based on the initial three-dimensional model network and the initial three-dimensional model surface of the foundation pit 300, a measurement three-dimensional model of the foundation pit 300 is constructed. The graphic deformation monitoring module can receive the measurement three-dimensional model of the foundation pit 300 output by the three-dimensional modeling module, obtain the three-dimensional structure information of the foundation pit 300 monitoring, identify the deformation area and deformation degree of the foundation pit by comparing the difference between the current three-dimensional model and the initial standard three-dimensional model, and combine the preset graphic deformation monitoring model. For the detected deformation area, the three-dimensional coordinates of the deformation area are accurately measured and calculated, and its lateral displacement, longitudinal displacement, and tilt displacement are calculated and output.

[0061] In the example of this application, the SE attention mechanism is integrated into the 3D CNN to enhance the network's sensitivity to important features. The SE mechanism improves the network's ability to identify deformation defects by recalibrating the feature responses between channels. The graphic deformation monitoring model includes an initial standard three-dimensional model and a subsequently monitored deformed model. The initial standard three-dimensional model is established based on the building coordinates and foundation pit coordinates of the construction location and foundation pit location matched by the construction site. The subsequently monitored deformed model matches the building coordinates with the initial standard three-dimensional model. By using a three-dimensional convolutional neural network (3D CNN), the deformation characteristics of the foundation pit are extracted, including information related to lateral displacement, longitudinal displacement, and tilt displacement, to establish a foundation pit deformation data sample for network training. The three-dimensional convolutional neural network (3D CNN) includes an input layer, a 3D convolutional layer, a pooling layer, an activation function, and a fully connected layer. The ReLU function is selected as the activation function, and a Softmax classifier is used after the fully connected layer to classify the convolutional features to accurately judge the deformation type and degree of the foundation pit. During the training process of the three-dimensional convolutional neural network, a labeled foundation pit deformation database is used for network training. By using data augmentation techniques such as rotation, scaling, translation, and simulating different lighting conditions, occlusion situations, and deformation situations, the diversity of training data is increased, the generalization ability of the network and the recognition ability for complex deformation situations are improved. The trained network is applied to the new foundation pit three-dimensional model for deformation detection, and the accuracy and reliability of the algorithm are evaluated by comparing with the manually labeled results, especially the recognition accuracy of lateral displacement, longitudinal displacement, and tilt displacement. According to the detection results and feedback, the network structure and parameters are continuously optimized to improve the recognition accuracy and robustness of the algorithm for foundation pit deformation.

[0062] The central processing system 200 of this application realizes comprehensive and accurate monitoring of foundation pit deformation by integrating two major modules of three-dimensional modeling and graphic deformation monitoring. The three-dimensional modeling module can quickly construct a measured three-dimensional model of the foundation pit based on the images taken by the AGV intelligent vehicle system, providing accurate basic data for subsequent deformation analysis; the graphic deformation monitoring module uses advanced 3D CNN and attention mechanism to improve the sensitivity and recognition ability to the deformation characteristics of the foundation pit, and can accurately judge the deformation type and degree of the foundation pit.

[0063] As a preferred example of the present application, the central processing system 200 further includes an early warning module. The foundation pit deformation characteristic values of the lateral displacement, longitudinal displacement, and tilt displacement output by the graphic deformation monitoring model are compared with the corresponding deformation thresholds. When at least one of the foundation pit deformation characteristic values is greater than the corresponding deformation threshold, the early warning mechanism is immediately activated to issue an early warning message. In the example of the present application, when the foundation pit deformation characteristic values of the lateral displacement, longitudinal displacement, or tilt displacement output by the graphic deformation monitoring model of the central processing system 200 are greater than the corresponding deformation thresholds, the early warning module of the central processing system 200 issues an early warning message such as an abnormal prompt message to the user terminal.

[0064] This setting integrates an early warning module in the central processing system 200. By monitoring the foundation pit deformation characteristic values, intelligently comparing the deformation thresholds, immediately triggering the early warning mechanism, and sending early warning messages to the user terminal, the entire monitoring and early warning process becomes more automated and intelligent, reducing the burden of manual monitoring, and at the same time improving the accuracy and reliability of monitoring. In the example of the present application, the deformation threshold is a preset empirical value. Preferably, the deformation threshold is 0.8 times the maximum safety values of the horizontal displacement, vertical displacement, and tilt angle of the foundation pit 300.

[0065] The present application also discloses a foundation pit monitoring method, including the following steps:

[0066] S1: Set a guiding runway 8 around the foundation pit 300. The guiding runway 8 is arranged at an equal distance from the side wall of the foundation pit 300 along the circumferential direction of the foundation pit 300;

[0067] S2: The AGV intelligent vehicle 1 in the AGV intelligent vehicle system 100 can navigate to the guiding runway 8 to check whether the lifting device 3, the telescopic mechanism 5, and the camera module 2 are in normal working conditions;

[0068] S3: Start the AGV intelligent vehicle 1. According to the Beidou satellite positioning module and the position type PID algorithm inside it, under the instruction of the central processing system, it idles along the guiding runway 8 for one week for path verification and equipment status confirmation;

[0069] S4: After the path verification and equipment status confirmation, adjust the working positions of the lifting device 3, the telescopic mechanism 5, and the first mounting bracket 6 in combination with the initial standard three-dimensional model coordinate data in the central processing system 200, and adjust the position and attitude measurement system information of the camera module 2;

[0070] S5: Start the AGV intelligent vehicle 1 again. The camera module 2 continuously collects real-time image information of the foundation pit 300 during the driving process and sends it to the central processing system 200 through the communication system;

[0071] S6: The central processing system 200 receives the image information, processes the image using image recognition and analysis techniques, performs three-dimensional modeling processing on the foundation pit image through the three-dimensional modeling module and the graphic deformation monitoring module, and conducts deformation analysis, identifies the deformation area and the degree of deformation, calculates the lateral displacement, longitudinal displacement and tilt displacement of the deformation area, and outputs the deformation characteristic values;

[0072] S7: The warning module of the central processing system 200 compares the deformation characteristic values with the deformation thresholds. When at least one deformation characteristic value is greater than the corresponding deformation threshold, the warning mechanism is immediately activated.

[0073] This application discloses a foundation pit monitoring method. By setting up the guiding runway, a stable driving path is provided for the AGV intelligent vehicle, ensuring the smooth progress of the entire monitoring process. Before the camera module takes pictures for deformation recognition, the equipment status check and path verification steps before the AGV intelligent vehicle travels can timely detect and correct potential problems, further enhancing the reliability and safety of the monitoring. After adjusting the information of the position and attitude of the camera module, a more accurate basis is provided for subsequent image acquisition. During the image acquisition process, the camera module continuously records the monitoring image information of the foundation pit. The use of this design enables the monitoring to timely capture the changes in the state of the foundation pit. The combination of the image recognition and analysis technology and the three-dimensional modeling module of the central processing system ensures that the three-dimensional model of the foundation pit can accurately reflect the current state of the foundation pit, enabling engineering personnel to continuously obtain relevant data. The graphic deformation monitoring module based on the three-dimensional convolutional neural network can not only efficiently analyze images, but also accurately identify the deformation area and degree. This process greatly improves the accuracy and efficiency of deformation detection. Combining the comparison of the deformation characteristic values with the set thresholds by the warning module in the central processing system can activate the warning mechanism immediately when the deformation characteristic values exceed the safety thresholds, ensuring that engineering managers can quickly take necessary measures. The foundation pit monitoring method described in this application realizes the comprehensive, accurate and automatic monitoring of the foundation pit deformation through a series of carefully designed steps. This series of comprehensive monitoring processes effectively reduces the workload of manual monitoring, not only improves the efficiency and accuracy of the monitoring work, but also provides a strong technical guarantee for the safe construction operation of the foundation pit. In this way, the personal safety and property safety at the construction site are effectively maintained, avoiding potential safety accidents caused by foundation pit deformation, and improving the overall project management level and emergency response ability. In the example of this application, the AGV intelligent vehicle 1 in the AGV intelligent vehicle system 100 performs periodic multiple detections according to the instructions of the central processing system 200, and increases the monitoring frequency when the warning mechanism is activated.

[0074] As a preferred example of the present application, in step S4, the shooting direction of the camera module 2 is set perpendicular to the running direction of the AGV intelligent vehicle 1, and the camera module 2 shoots the foundation pit 300 obliquely. The upper edge of the lens of the camera module 2 covers the upper edge of the foundation pit 300 and extends to the position where the guiding runway 8 is located, and the lower edge of the lens of the camera module 2 covers the lower edge of the foundation pit 300 and extends to the bottom surface of the foundation pit.

[0075] This setting adopts an oblique shooting method, which does not require repeated shooting, can save the cost of using the image acquisition device, expand the monitoring range, improve the accuracy and meticulousness of monitoring, enable the central processing system to more accurately construct the three-dimensional model of the foundation pit and analyze its deformation situation based on richer image information through the three-dimensional modeling and graphic deformation monitoring module, realizes the all-round and accurate monitoring of the foundation pit deformation, provides a strong technical guarantee for the safe construction of the foundation pit project, greatly improves the efficiency and accuracy of monitoring, and reduces the difficulty and risk of manual monitoring.

[0076] As a preferred example of the present application, in step S7, when the deformation characteristic value calculated by the central processing system 200 through the graphic deformation monitoring module is greater than the corresponding deformation threshold, by adjusting the first driving device 64 or the second driving device 65 of the first mounting rack 6, the shooting direction of the camera module 2 is adjusted by swinging at least a preset angle in the horizontal direction and at least a preset angle in the vertical direction, and the detection of steps S5 to S7 is performed again. This setting introduces an intelligent adjustment and verification mechanism, adopts multi-angle shooting of images and performs image deformation analysis, which not only improves the sensitivity of monitoring, but also can provide accurate deformation data and early warning information for engineering personnel in a timely manner when potential deformation risks are found.

[0077] The foundation pit monitoring and measurement system described in this application integrates an AGV intelligent vehicle system and a central processing system. As the core component, the AGV intelligent vehicle can automatically drive along a preset guiding runway without manual intervention, greatly reducing labor costs and the risk of human error, and ensuring the smooth progress of the monitoring process. The camera module carried on it quickly collects image information of the foundation pit according to the monitoring instructions, and transmits the data to the central processing system quickly through an efficient communication system, providing data guarantee for accurate monitoring. The central processing system then uses advanced image recognition technology to deeply analyze the received data, realizing the comprehensiveness, accuracy and automation of foundation pit monitoring, significantly improving the efficiency and accuracy of foundation pit deformation monitoring, being able to timely identify the deformation situation of the foundation pit and conduct accurate deformation calculation, so as to provide scientific and well-founded safety assessment for engineering personnel. In addition, the central processing system combines the initial standard three-dimensional model coordinate data to optimize the running track of the AGV intelligent vehicle, ensuring the comprehensiveness and non-omission of monitoring, making the monitoring performance still efficient and stable in complex environments. By combining the application of three-dimensional modeling and deformation monitoring modules, it not only intuitively displays the state changes of the foundation pit, but also calculates the deformation characteristic values to identify potential risks in advance, greatly improving the timeliness and accuracy of early warning. The foundation pit monitoring and measurement system and monitoring method described in this application improve the efficiency and accuracy of monitoring work, effectively avoid safety accidents caused by foundation pit deformation, provide important guarantee for the safety of construction site personnel and property, and thus improve the level of the entire project management and emergency response ability.

[0078] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A foundation pit monitoring and measuring system, characterized in that, include: An AGV intelligent vehicle-mounted system (100) comprises an AGV intelligent vehicle (1), wherein the AGV intelligent vehicle (1) is capable of driving along a track around a foundation pit (300) according to instructions from a central processing system, and a camera module (2) is arranged on the AGV intelligent vehicle (1), wherein the camera module (2) is capable of collecting image information of the foundation pit (300); A central processing system (200), the central processing system (200) being able to connect to the camera module (2) to collect image information of the foundation pit (300), and in combination with the image recognition and analysis system, to perform deformation detection and deformation calculation on the image information of the foundation pit (300), thereby obtaining a deformation analysis result of the foundation pit; The central processing system (200) is in communication connection with the AGV intelligent vehicle system (100), and is capable of navigating the running track of the AGV intelligent vehicle (1) according to the initial standard three-dimensional model coordinate data of the foundation pit (300) and the positioning system; The central processing system (200) is provided with a three-dimensional modeling module and a graphic deformation monitoring module. The three-dimensional modeling module can perform image processing according to the image transmitted by the AGV intelligent vehicle system (100) and the parameter information of the image to construct a measurement three-dimensional model of the foundation pit (300). The graphic deformation monitoring module includes a preset graphic deformation monitoring model. The graphic deformation monitoring model uses a three-dimensional convolutional neural network fused with an attention mechanism to perform deformation target detection to obtain a foundation pit deformation feature value including a lateral displacement, a longitudinal displacement and an inclined displacement.

2. The foundation pit monitoring and measuring system according to claim 1, characterized in that, A guide track (8) is arranged around the foundation pit (300), a guide module and a drive module are integrated on the AGV smart car (1), and a position PID algorithm is used to drive the AGV smart car (1) to move along the foundation pit (300).

3. The foundation pit monitoring and measuring system according to claim 1 or 2, characterized in that, The AGV smart vehicle (1) is provided with a lifting device (3), a first connecting frame (4) is provided at the upper end of the lifting device (3), a telescopic mechanism (5) is provided on the first connecting frame (4), the plane where the telescopic direction of the telescopic mechanism (5) is located is arranged perpendicular to the lifting direction of the lifting device (3), a first mounting frame (6) is provided at the cantilever end extending from the telescopic mechanism (5), and the camera module (2) is fixed on the first mounting frame (6).

4. The foundation pit monitoring and measuring system according to claim 3, wherein, The first mounting frame (6) comprises a first connecting member (61), an adapter (62) and a second connecting member (63); a first driving device (64) is arranged on the first connecting member (61); one end of the adapter (62) is connected to the first driving device (64), and the first driving device (64) is capable of driving the adapter (62) to rotate in a horizontal direction; a second connecting member (63) is arranged at one end of the adapter (62) away from the first driving device (64); two connecting arms for connecting to a camera module (2) are arranged on the second connecting member (63); a second driving device (65) is arranged on one of the connecting arms, and is capable of driving the camera module (2) to rotate in a vertical direction.

5. The foundation pit monitoring and measuring system according to claim 1, characterized in that The foundation pit monitoring and measuring system further includes a target system, and the target system includes a plurality of target devices (7), and the multiple target devices (7) are arranged in a matrix on the side wall of the foundation pit (300).

6. The foundation pit monitoring and measuring system according to claim 1, characterized in that The AGV intelligent vehicle system (100) tilts and photographs the foundation pit (300) with the camera module (2), processes the acquired images, obtains the characteristic density distribution value of the foundation pit (300), performs space resection on the pixel coordinates in the images according to the key structural feature points and the position and attitude measurement system information of the camera module (2), obtains the three-dimensional position relative to the foundation pit (300) and the three-dimensional coordinate relationship between the key structural feature points when the images are taken, generates the sparse point cloud data corresponding to the foundation pit (300) based on the three-dimensional position and the three-dimensional coordinate relationship, matches the key structural feature points in the sparse point cloud data with the images, generates the dense point cloud data, connects the characteristic pixel points with spline curves to obtain the initial three-dimensional model network of the foundation pit (300), based on the initial three-dimensional model network, connects the characteristic pixel points in pairs to form triangular faces to obtain the target three-dimensional model network, performs reconstruction analysis on the target three-dimensional model network to obtain the initial three-dimensional model surface corresponding to the foundation pit, constructs the measurement three-dimensional model of the foundation pit (300) based on the initial three-dimensional model network and the initial three-dimensional model surface of the foundation pit (300), the graphic deformation monitoring module can receive the measurement three-dimensional model of the foundation pit (300) output by the three-dimensional modeling module, obtain the three-dimensional structural information of the foundation pit (300) monitoring, identify the deformation area and deformation degree of the foundation pit by comparing the difference between the current three-dimensional model and the initial standard three-dimensional model, and combine the preset graphic deformation monitoring model. For the detected deformation area, accurately measure and calculate the three-dimensional coordinates of the deformation area, calculate its lateral displacement, longitudinal displacement and tilt displacement and output them.

7. The foundation pit monitoring and measurement system according to claim 6, wherein The central processing system (200) further includes an early warning module, which compares the foundation pit deformation characteristic values of the lateral displacement, longitudinal displacement and tilt displacement output by the graphic deformation monitoring model with the corresponding deformation thresholds. When at least one of the foundation pit deformation characteristic values is greater than the corresponding deformation threshold, the early warning mechanism is immediately activated and early warning information is issued.

8. A foundation pit monitoring method, which is applied to the foundation pit monitoring and measuring system according to any one of claims 1 to 7, characterized in that, It includes the following steps: S1: A guiding runway (8) is arranged around the foundation pit (300), and the guiding runway (8) is arranged at an equal distance from the side wall of the foundation pit (300) along the circumferential direction of the foundation pit (300); S2: The AGV intelligent vehicle (1) in the AGV intelligent vehicle system (100) can navigate to the guiding runway (8) to check that the lifting device (3), the telescopic mechanism (5) and the camera module (2) are in normal working states; S3: Start the AGV intelligent vehicle (1), and according to the Beidou satellite positioning module and the position type PID algorithm inside it, idle one week along the guiding runway (8) under the instruction of the central processing system for path verification and equipment status confirmation; S4: After path verification and equipment status confirmation, adjust the working positions of the lifting device (3), telescopic mechanism (5), and first mounting bracket (6) in combination with the initial standard three-dimensional model coordinate data in the central processing system (200), and adjust the position and attitude measurement system information of the camera module (2). S5: Start the AGV intelligent vehicle (1) again. During the driving process, the camera module (2) continuously collects real-time image information of the foundation pit (300) and sends it to the central processing system (200) through the communication system. S6: The central processing system (200) receives the image information, processes the image using image recognition and analysis technology, performs three-dimensional modeling processing and deformation analysis on the foundation pit image through the three-dimensional modeling module and the graphic deformation monitoring module, identifies the deformation area and deformation degree, calculates the lateral displacement, longitudinal displacement, and tilt displacement of the deformation area, and outputs the deformation characteristic values. S7: The warning module of the central processing system (200) compares the deformation characteristic values with the deformation thresholds. When at least one deformation characteristic value is greater than the corresponding deformation threshold, the warning mechanism is immediately activated.

9. The foundation pit monitoring method according to claim 8, wherein In step S4, the shooting direction of the camera module (2) is set perpendicular to the running direction of the AGV intelligent vehicle (1), and the camera module (2) takes an inclined shot of the foundation pit (300). The upper edge of the lens of the camera module (2) covers the upper edge of the foundation pit (300) and extends to the position where the guiding runway (8) is located, and the lower edge of the lens of the camera module (2) covers the lower edge of the foundation pit (300) and extends to the bottom surface of the foundation pit.

Images