Foundation pit monitoring and measuring system and monitoring method

Abstract

The invention relates to the technical field of foundation pit construction monitoring, and provides a foundation pit monitoring and measuring system and a monitoring method.The foundation pit monitoring and measuring system comprises an AGV intelligent vehicle-mounted system and a central processing system, the AGV intelligent vehicle-mounted system is provided with an AGV intelligent vehicle, and the AGV intelligent vehicle can be guided to run along runways around a foundation pit according to instructions of the central processing system; the vehicle is provided with a camera module used for collecting foundation pit image information. And the central processing system can receive and process the foundation pit image information collected by the camera module, perform deformation detection and calculation and obtain a deformation analysis result, is in communication connection with the AGV intelligent vehicle-mounted system, and navigates the moving track of the AGV intelligent vehicle according to the foundation pit three-dimensional model coordinate data and the positioning system. According to the foundation pit monitoring and measuring system, the AGV intelligent vehicle-mounted system is used for carrying the camera module to run along the guiding runway around the foundation pit, and comprehensive and automatic monitoring of the foundation pit is achieved in combination with the image recognition and analysis technology of the central processing system.

Description

Technical Field

[0001] The present invention relates to the technical field of foundation pit construction monitoring, and in particular 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 vital importance. Traditional monitoring methods usually rely on manual inspections and fixed-point monitoring, which 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 technology still faces many challenges.

[0003] Traditional manual inspection methods are inefficient and difficult to fully 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 levels and theodolites for fixed-point measurements. These methods are time-consuming and the results are easily affected by human factors.

[0004] At present, displacement sensors and stress sensors 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 and humidity), resulting in errors in monitoring data. In addition, the monitoring range of sensors is limited, making it difficult to achieve comprehensive monitoring of large and complex foundation pit projects.

[0005] With the advancement of image processing technology, foundation pit monitoring solutions based on image recognition have gradually emerged. This type of technology 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 patent publication number CN107119657B discloses a foundation pit monitoring method based on visual measurement, which sets multiple spikes or steel bars with crosshairs as detection points, and performs grayscale conversion after obtaining digital color images. The optimal coordinates of multiple target centers are obtained through noise removal, multi-target area recognition, and target center calculation. The BP neural network is used to establish a foundation pit deformation prediction model to estimate and warn possible structural deformation of the foundation pit. Although the above scheme has achieved the monitoring of foundation pit deformation to a certain extent, its monitoring process mainly relies on the fixed arrangement of camera modules to collect images and process them through subsequent software algorithms. Although automation has been achieved to a certain extent, it still lacks flexibility.

[0006] Therefore, among the various existing monitoring technologies, recognition accuracy, stability and monitoring cost are still the main constraints. The existing foundation pit monitoring technology still has many problems and limitations, and it is difficult to meet the growing monitoring needs 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 of personnel experience and subjective factors, and proposes a foundation pit monitoring and measurement system and a monitoring method. By providing a foundation pit monitoring and measurement system and a method, an AGV intelligent vehicle-mounted system is used to carry a camera module along a guide 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 achieved.

[0008] To achieve the above object, the technical solution of the present invention is achieved as follows: A foundation pit monitoring and measuring system, comprising: The AGV intelligent vehicle system includes an AGV intelligent vehicle, which can be guided along the runway around the foundation pit according to the instructions of the central processing system, and a camera module is set on the AGV intelligent vehicle, and the camera module can collect image information of the foundation pit; Central processing system: The central processing system can connect to the camera module to collect the image information of the foundation pit, and combine with the image recognition and analysis system to perform deformation detection and deformation calculation on the image information of the foundation pit to obtain the deformation analysis results of the foundation pit; The central processing system is communicatively connected with the AGV intelligent vehicle-mounted 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.

[0009] Furthermore, a guide track is arranged around the foundation pit, a guide module and a drive module are integrated on the AGV smart car, and a position PID algorithm is adopted to drive the AGV smart car to move along the foundation pit.

[0010] Furthermore, a lifting device is arranged on the AGV smart 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, a plane where the telescopic direction of the telescopic mechanism is located is arranged perpendicular to the lifting direction of the lifting device, a first mounting frame is arranged at the cantilever end extending from the telescopic mechanism, and the camera module is fixed on the first mounting frame.

[0011] 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.

[0012] 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.

[0013] 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.

[0014] 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.

[0015] Furthermore, the central processing system also includes an early warning module, and the foundation pit deformation characteristic values ​​of the lateral displacement, longitudinal displacement and inclined 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 and the early warning information is issued.

[0016] Compared with the prior art, the foundation pit monitoring and measurement system of the present invention has the following advantages: (1) The foundation pit monitoring and measurement system described in the present invention, through the automated navigation operation of the AGV intelligent vehicle, achieves full coverage of the foundation pit area to the maximum extent possible, avoiding the blind spots existing in 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, which can quickly respond to the deformation of the foundation pit and provide a reliable basis for safety warning. It can be quickly implemented in different construction environments, thereby reducing the monitoring cost and improving the timeliness of the data, which is helpful for the safety management and control of the foundation pit project.

[0017] (2) The foundation pit monitoring and measurement system described in the present invention has high flexibility and adaptability, can be quickly implemented in different construction environments, reduces monitoring costs, and significantly improves the safety management and control level of foundation pit projects.

[0018] Another object of the present invention is to provide a foundation pit monitoring method, which is applied to the foundation pit monitoring and measurement system as described above, and comprises the following steps: S1: A guide runway is arranged around the foundation pit, and the guide runway is arranged equidistantly from the side wall of the foundation pit along the circumferential direction of the foundation pit; S2: The AGV in the AGV system can navigate to the guide track and check whether the lifting device, telescopic mechanism and camera module are in normal working condition; S3: Start the AGV smart car, and according to its internal Beidou satellite positioning module and position PID algorithm, it will idle along the guide runway for one circle under the instruction of the central processing system for path verification and equipment status confirmation; S4: After the path validation and equipment status are confirmed, the working positions of the lifting device, the telescopic mechanism, and the first mounting frame are adjusted in combination with the initial standard three-dimensional model coordinate data in the central processing system, and the position of the camera module and the posture measurement system information are adjusted; S5: The AGV smart car is started again. The camera module continuously collects real-time image information of the foundation pit during driving and sends it to the central processing system through the communication system. S6: The central processing system receives the image information, processes the image using the image recognition and analysis technology, performs 3D modeling and deformation analysis on the foundation pit image through the 3D 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 value; S7: The early warning module of the central processing system compares the deformation characteristic value with the deformation threshold value, and when at least one deformation characteristic value is greater than the corresponding deformation threshold value, the early warning mechanism is immediately activated.

[0019] Furthermore, in step S4, the shooting direction of the camera module is set perpendicular to the running direction of the AGV smart vehicle, and the camera module shoots the foundation pit at an angle. The upper edge of the camera module's lens covers the upper edge of the foundation pit and extends to the location of the guide runway, and the lower edge of the camera module's lens covers the lower edge of the foundation pit and extends to the bottom of the foundation pit.

[0020] The foundation pit monitoring and measurement system and monitoring method described in this application improve the efficiency and accuracy of monitoring work, and effectively avoid safety accidents caused by foundation pit deformation, providing important protection for the safety of personnel and property on the construction site, thereby improving the level of overall project management and emergency response capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings: Figure 1 It is a structural schematic diagram of the foundation pit monitoring and measuring system according to an embodiment of the present invention; Figure 2 It is a structural schematic diagram of image acquisition by the AGV intelligent vehicle-mounted system in the foundation pit monitoring and measurement system according to an embodiment of the present invention; Figure 3 It is a structural schematic diagram of an AGV intelligent vehicle running in a guide track in a foundation pit monitoring and measuring system according to an embodiment of the present invention; Figure 4 This is a structural schematic diagram of the camera module according to an embodiment of the present invention installed on a first mounting frame; Figure 5 for Figure 4 A schematic diagram of a side view of the structure shown in; Figure 6 It is a schematic diagram of the process of the foundation pit monitoring method according to an embodiment of the present invention; Description of reference numerals: 100. AGV intelligent vehicle system; 200. Central processing system; 300. Foundation pit; 400. Target system; 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. Guide runway; 9. Lens shoots the upper edge; 10. Lens shoots the lower edge. DETAILED DESCRIPTION

[0022] In order to make the technical means, objectives and effects of the present invention easy to understand, the embodiments of the present invention are described in detail below with reference to specific drawings.

[0023] It should be noted that all the terms used in the present invention for directional and positional indications, such as "upper", "lower", "left", "right", "front", "back", "vertical", "horizontal", "inside", "outside", "top", "lower", "lateral", "longitudinal", "center", etc., are only used to explain the relative positional relationship and connection status between the components in a certain state (as shown in the accompanying 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, and therefore cannot be understood as limiting the present invention. In addition, the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes, and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features.

[0024] In the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" 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 a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0025] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0026] like Figures 1 to 6 As shown, the present application discloses a foundation pit monitoring and measuring system, comprising: The AGV intelligent vehicle system 100 includes an AGV intelligent vehicle 1, which can be guided along the runway around the foundation pit 300 according to the instructions of the central processing system, and a camera module 2 is set on the AGV intelligent vehicle 1, and the camera module 2 can collect image information of the foundation pit 300; The central processing system 200 can be connected to the camera module 2 to collect image information of the foundation pit 300, and in combination with the image recognition and analysis system, perform deformation detection and deformation calculation on the image information of the foundation pit 300 to obtain the deformation analysis result of the foundation pit; The central processing system 200 is in communication connection with 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.

[0027] The present application designs a foundation pit monitoring and measurement system that integrates a central processing system 200 and an AGV intelligent vehicle-mounted system 100, which can autonomously and efficiently perform monitoring and measurement tasks of a foundation pit 300. The AGV intelligent vehicle 1 in the AGV intelligent vehicle-mounted system 100 moves along a 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 a 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, and thus obtain deformation analysis results 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 smart car 1 to perform real-time navigation on the running trajectory of the AGV smart car 1, ensuring that the camera module 2 in the AGV vehicle-mounted system can cover every corner of the foundation pit 300 for monitoring. During this process, the AGV vehicle-mounted system can flexibly adjust the operating mode to adapt to different foundation pit conditions, thereby improving the flexibility and efficiency of monitoring.

[0028] The foundation pit monitoring and measurement system disclosed in the present application, through the automated navigation operation of the AGV smart vehicle, achieves comprehensive coverage of the foundation pit area to the maximum extent, avoiding the blind spots existing in traditional monitoring methods. At the same time, based on the integration of image recognition and analysis technology, the analysis of monitoring data is made more scientific and systematic, which can quickly respond to the deformation of the foundation pit, provide a reliable basis for safety warning, and can be quickly implemented in different construction environments, thereby reducing monitoring costs and improving the real-time nature of data, which is helpful for the safety management and control of foundation pit projects.

[0029] In the example of the present application, a guide runway 8 is set 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. As a preferred example of the present application, a flat road around the foundation pit 300 is selected as the guide runway of the AGV smart car 1. Preferably, the guide runway 8 is equidistantly set at a distance of 2 to 3 meters from the side wall of the foundation pit. A guide module and a drive module are set inside the AGV smart car 1. Combined with the position PID algorithm, the driving direction and speed of the vehicle can be accurately controlled to ensure that the vehicle travels stably along the guide runway, and the internal Beidou satellite positioning module is used to provide the AGV smart car with high-precision positioning information, so that it can travel along the guide runway.

[0030] This setting optimizes the navigation task of the AGV smart car 1, ensures the driving accuracy and stability of the AGV smart car, and enables the AGV smart car to efficiently and accurately drive along the preset route when performing foundation pit monitoring and measurement tasks without human intervention, greatly reducing the operation complexity and labor costs, providing a solid foundation for subsequent image recognition analysis and deformation calculation, and improving the efficiency and accuracy of foundation pit monitoring and measurement.

[0031] As a preferred example of the present application, a lifting device 3 is provided on the AGV smart car 1, 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 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. In the example of the present application, by providing a lifting device 3 and a telescopic mechanism 5 on the AGV smart car 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, and to easily cope with various complex construction environments. When the AGV smart car 1 navigates according to the actual situation of the environment, the camera module 2 can be quickly adjusted to the best observation position and angle according to actual needs, whether it is monitoring the upper edge or the depth of the foundation pit, it can obtain clear and accurate image information, and achieve full coverage of the foundation pit 300, thereby providing accurate data support. The freedom and adaptability of this combination not only enables the AGV smart car 1 to better cope with diverse working environments, but also ensures that monitoring tasks can be carried out smoothly during the complex and dynamically changing foundation pit construction process.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] As a preferred example of the present application, the foundation pit monitoring and measurement system further includes a target system 400, and the target system 400 includes a plurality of target devices 7, and the plurality of target devices 7 are arranged in a matrix on the side wall of the foundation pit 300. In the example of the present application, the target system 400 is further introduced and combined with the image recognition technology to monitor the deformation of the foundation pit. The target device 7 is a circular or square stainless steel sheet with a diameter or side length of about 50mm to 100mm and a thickness of about 2mm to 5mm. It is fixed on the side wall of the foundation pit 300 by embedded parts or high-performance epoxy resin adhesives. In daily dynamic monitoring, the camera module 2 can use the plurality of target devices 7 of the target system 400 as key structural feature points of the image, and by comparing the three-dimensional modeling position changes of the target device 7, the deformation information of the foundation pit is extracted by using the image recognition technology, so as to further realize the accurate monitoring of the deformation of the foundation pit and reduce the error interference in the overall monitoring process.

[0036] This setting achieves comprehensive monitoring of foundation pit deformation with wide coverage and high precision by introducing target devices 7 arranged intermittently in a matrix shape, so that even small deformations can be captured quickly and accurately, making the image recognition process simpler and more effective. In subsequent real-time monitoring, the deformation of the foundation pit can be analyzed quickly and accurately, providing intuitive and reliable deformation data support for construction personnel. In the example of the present application, the target device 7 needs to be regularly maintained during specific use to avoid damage, oil or mud contamination. If the target device 7 needs to be replaced, the foundation pit image after the target device 7 is replaced needs to be photographed and modeled again, and the three-dimensional modeling image data of the foundation pit deformation needs to be updated.

[0037] As a preferred example of the present 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 image transmitted by the AGV intelligent vehicle system 100 and the image parameter information 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) fused with an attention mechanism (SE, Squeeze-and-Excitatio) to detect deformation targets and obtain foundation pit deformation feature values ​​including lateral displacement, longitudinal displacement and tilt displacement. In the example of the present application, the central processing system 200 realizes accurate 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-mounted system 100 takes an image of the foundation pit 300, the shooting direction of the camera module 2 on the AGV intelligent vehicle-mounted system 100 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 through the lifting device 3, the telescopic mechanism 5 and the first mounting frame 6, and processes the acquired image to obtain the characteristic density distribution value of the foundation pit 300. According to the key structural feature points (such as the foundation pit turning point, significant markers, layered interfaces, target device 7, etc.) and the position and posture measurement system information of the camera module 2, the pixel coordinates in the image are processed by aerial triangulation to obtain the relative position of the foundation pit 300 when the image is taken. 0 and the three-dimensional coordinate relationship between the key structural feature points, based on the three-dimensional position and the three-dimensional coordinate relationship, generate sparse point cloud data corresponding to the foundation pit 300, match the key structural feature points in the sparse point cloud data with the image, generate dense point cloud data, use spline curves to connect the feature pixel points, obtain the initial three-dimensional model network of the foundation pit 300, based on the initial three-dimensional model network, connect the feature pixel points in pairs to form triangular faces, obtain the target three-dimensional model grid, reconstruct and analyze the target three-dimensional model grid, obtain the initial three-dimensional model surface corresponding to the foundation pit, and construct the measured 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 measured three-dimensional model of the foundation pit 300 output by the three-dimensional modeling module, obtain the three-dimensional structural information monitored by the foundation pit 300, 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, accurately measure and calculate the three-dimensional coordinates of the detected deformation area, calculate its lateral displacement, longitudinal displacement and tilt displacement and output them.

[0038] In the example of the present 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 response between channels. The graphic deformation monitoring model includes an initial standard three-dimensional model and a deformation model monitored subsequently. The initial standard three-dimensional model is established based on the building coordinates and foundation pit coordinates of the building location and foundation pit location matched by the construction site. The deformation model monitored subsequently matches the building coordinates with the initial standard three-dimensional model. By adopting a three-dimensional convolutional neural network (3D CNN), the foundation pit deformation features are extracted, including relevant information on lateral displacement, longitudinal displacement and tilt displacement, and a foundation pit deformation data sample is established 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 activation function selects the ReLU function. After the fully connected layer, a Softmax classifier is used to classify the convolution features to accurately determine the type and degree of deformation of the foundation pit. In the process of training the three-dimensional convolutional neural network, the labeled foundation pit deformation database is used for network training. By adopting data enhancement techniques such as rotation, scaling, translation, and simulation of different lighting conditions, occlusion conditions, and deformation conditions, the diversity of training data is increased, and the generalization ability of the network and the ability to recognize complex deformation conditions are improved. The trained network is applied to the new foundation pit three-dimensional model for deformation detection. By comparing with the results of manual annotation, especially the recognition accuracy of lateral displacement, longitudinal displacement and tilt displacement, the accuracy and reliability of the algorithm are evaluated. According to the detection results and feedback, the network structure and parameters are continuously optimized to improve the algorithm's recognition accuracy and robustness for foundation pit deformation.

[0039] The central processing system 200 of the present application realizes comprehensive and accurate monitoring of foundation pit deformation by integrating two modules: three-dimensional modeling and graphic deformation monitoring. The three-dimensional modeling module can quickly construct a measurement 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 of foundation pit deformation characteristics, and can accurately judge the deformation type and degree of the foundation pit.

[0040] As a preferred example of the present application, the central processing system 200 also includes an early warning module, and 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 and early warning information is issued. 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 threshold, the early warning module of the central processing system 200 sends early warning information such as abnormal prompt information to the user terminal.

[0041] This setting integrates an early warning module in the central processing system 200, monitors the deformation characteristic value of the foundation pit, intelligently compares the deformation threshold, immediately triggers the early warning mechanism and sends the early warning information to the user terminal, making the entire monitoring and early warning process more automated and intelligent, reducing the burden of manual monitoring, and 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 safe value of the horizontal displacement, vertical displacement, and inclination angle of the foundation pit 300.

[0042] The present application also discloses a foundation pit monitoring method, comprising the following steps: S1: a guide runway 8 is arranged around the foundation pit 300, and the guide runway 8 is arranged equidistant from the side wall of the foundation pit 300 along the circumferential direction of the foundation pit 300; S2: The AGV smart vehicle 1 in the AGV smart vehicle system 100 is able to navigate to the guide runway 8 and check that the lifting device 3, the telescopic mechanism 5 and the camera module 2 are in normal working condition; S3: Start the AGV smart car 1, and according to the Beidou satellite positioning module and the position PID algorithm inside it, it will idle along the guide runway 8 for one circle under the instruction of the central processing system for path verification and equipment status confirmation; S4: After the path validation and device status are confirmed, the working positions of the lifting device 3, the telescopic mechanism 5, and the first mounting frame 6 are adjusted in combination with the initial standard three-dimensional model coordinate data in the central processing system 200, and the position and posture measurement system information of the camera module 2 are adjusted; S5: The AGV smart car 1 is started again, and the camera module 2 continuously collects real-time image information of the foundation pit 300 during driving, 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 the image recognition and analysis technology, performs 3D modeling and deformation analysis on the foundation pit image through the 3D 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 value; S7: The early warning module of the central processing system 200 compares the deformation characteristic value with the deformation threshold value, and immediately starts the early warning mechanism when at least one deformation characteristic value is greater than the corresponding deformation threshold value.

[0043] The present application discloses a foundation pit monitoring method. By setting a guide runway, a stable driving path is provided for an AGV smart car, ensuring the smooth progress of the entire monitoring process. Before the camera module takes images for deformation recognition, the equipment status check and path verification steps of the AGV smart car before driving can timely discover and correct potential problems, further enhancing the reliability and safety of monitoring. After adjusting the information of the position and posture 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 capture the changes in the foundation pit state in a timely manner. The combination of the image recognition and analysis technology of the central processing system and the three-dimensional modeling module ensures that the three-dimensional model of the foundation pit can accurately reflect the current state of the foundation pit, so that engineering personnel can continuously obtain relevant data. The graphic deformation monitoring module based on the three-dimensional convolutional neural network can not only efficiently analyze the image, but also accurately identify the deformation area and degree. This process greatly improves the accuracy and efficiency of deformation detection. Combined with the early warning module in the central processing system, the deformation characteristic value is compared with the set threshold. Once the deformation characteristic value exceeds the safety threshold, the early warning mechanism can be activated in the first time to ensure that engineering management personnel can quickly take necessary measures. The foundation pit monitoring method described in this application realizes comprehensive, accurate and automated monitoring of 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 monitoring work, but also provides strong technical guarantees for safe construction of foundation pits. In this way, the safety of personnel and property at the construction site is also effectively maintained, avoiding potential safety accidents caused by foundation pit deformation, and improving the overall engineering management level and emergency response capabilities. In the example of this application, the AGV smart car 1 in the AGV smart vehicle system 100 performs periodic multiple detections according to the instructions of the central processing system 200, and increases the monitoring frequency when the early warning mechanism is activated.

[0044] 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 smart car 1, and the camera module 2 shoots the foundation pit 300 at an angle, the upper edge 9 of the camera module 2 covers the upper edge of the foundation pit 300 and extends to the position of the guide runway 8, and the lower edge 10 of the camera module 2 covers the lower edge of the foundation pit 300 and extends to the bottom of the foundation pit.

[0045] This setting adopts an inclined shooting method, which eliminates the need for repeated shooting and can save the cost of using image acquisition devices. At the same time, it expands the monitoring range and improves the accuracy and detail of monitoring. It enables the central processing system to more accurately construct a three-dimensional model of the foundation pit based on richer image information through three-dimensional modeling and graphic deformation monitoring modules, and analyze its deformation, thereby achieving all-round and precise monitoring of foundation pit deformation, providing strong technical support for the safe construction of foundation pit projects, greatly improving the efficiency and accuracy of monitoring, and reducing the difficulty and risk of manual monitoring.

[0046] 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, the first driving device 64 or the second driving device 65 of the first mounting frame 6 is adjusted, and the camera module 2 is adjusted to swing at least a preset angle in the horizontal direction and at least a preset angle in the vertical direction, and the shooting direction of the camera module 2 is adjusted, and the detection of steps S5 to S7 is performed again. This setting not only improves the sensitivity of monitoring by introducing an intelligent adjustment and verification mechanism, taking images from multiple angles and performing image deformation analysis, but also can provide accurate deformation data and early warning information to engineering personnel in a timely manner when potential deformation risks are discovered.

[0047] The foundation pit monitoring and measurement system described in the present application integrates the AGV intelligent vehicle system and the central processing system. The AGV intelligent vehicle, as the core component, can automatically travel along the preset guide track without human intervention, greatly reducing labor costs and the risk of human errors, and ensuring the smooth progress of the monitoring process. The camera module it carries can quickly collect image information of the foundation pit according to monitoring instructions, and quickly transmit the data to the central processing system through an efficient communication system, providing data guarantee for accurate monitoring. The central processing system uses advanced image recognition technology to conduct in-depth analysis of the received data, realizing comprehensive, accurate and automated foundation pit monitoring, significantly improving the efficiency and accuracy of foundation pit deformation monitoring, and can timely identify the deformation of the foundation pit and perform accurate deformation calculations, thereby providing scientific and well-founded safety assessments for engineering personnel. In addition, the central processing system combines the initial standard three-dimensional model coordinate data to optimize the running trajectory of the AGV smart car, ensuring the comprehensiveness and completeness of monitoring, so that the monitoring performance is 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 identifies potential risks in advance by calculating the deformation characteristic values, greatly improving the timeliness and accuracy of the early warning. The foundation pit monitoring and measurement system and monitoring method described in this application improve the efficiency and accuracy of monitoring work, and effectively avoid safety accidents caused by foundation pit deformation, providing important guarantees for the safety of personnel and property on the construction site, thereby improving the level of the entire project management and emergency response capabilities.

[0048] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A foundation pit monitoring and measurement system, characterized in that: include: An AGV intelligent vehicle-mounted system (100) includes an AGV intelligent vehicle (1), wherein the AGV intelligent vehicle (1) is capable of traveling along a track around a foundation pit (300) according to instructions from a central processing system, and a camera module (2) is provided on the AGV intelligent vehicle (1), wherein the camera module (2) is capable of collecting image information of the foundation pit (300); The central processing system (200) is capable of connecting to the camera module (2) to collect image information of the foundation pit (300), and combining 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 foundation pit deformation analysis result; The central processing system (200) is in communication with the AGV intelligent vehicle system (100) and is capable of navigating the running trajectory of the AGV intelligent vehicle (1) based on the initial standard three-dimensional model coordinate data of the foundation pit (300) and the positioning system.

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

3. The foundation pit monitoring and measurement system according to claim 1 or 2, characterized in that: A lifting device (3) is provided on the AGV intelligent vehicle (1), 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), a 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 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 measurement system according to claim 3, characterized in that: 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 provided 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 a horizontal direction. A second connecting member (63) is provided at one end of the adapter (62) away from the first driving device (64). Two connecting arms for connecting to the camera module (2) are provided on the second connecting member (63). A second driving device (65) is provided on one of the connecting arms to drive the camera module (2) to rotate in a vertical direction.

5. The foundation pit monitoring and measurement system according to claim 1, characterized in that: The foundation pit monitoring and measurement system further comprises a target system (400), wherein the target system (400) comprises a plurality of target devices (7), wherein the plurality of target devices (7) are arranged in a matrix on the side wall of the foundation pit (300).

6. The foundation pit monitoring and measurement system according to claim 1, characterized in that: The central processing system (200) is provided with a three-dimensional modeling module and a graphic deformation monitoring module. The three-dimensional modeling module is capable of performing image processing based on the image transmitted by the AGV intelligent vehicle-mounted 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 detect deformation targets and obtain foundation pit deformation characteristic values ​​including lateral displacement, longitudinal displacement and tilt displacement.

7. The foundation pit monitoring and measurement system according to claim 6, characterized in that: The AGV intelligent vehicle-mounted system (100) uses the camera module (2) to perform tilted shooting of the foundation pit (300), processes the acquired image, obtains the characteristic density distribution value of the foundation pit (300), performs aerial triangulation on the pixel coordinates in the image according to the key structural characteristic points and the position and posture measurement system information of the camera module (2), obtains the three-dimensional position relative to the foundation pit (300) when the image is shot and the three-dimensional coordinate relationship between the key structural characteristic points, generates 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 characteristic 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 an initial three-dimensional model network of the foundation pit (300), and based on the initial three-dimensional model network, The characteristic pixel points are connected in pairs to form triangular faces to obtain a target three-dimensional model grid, and the target three-dimensional model grid is reconstructed and analyzed to obtain an 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 measured three-dimensional model of the foundation pit (300) is constructed. The graphic deformation monitoring module can receive the measured 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, 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 a 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.

8. The foundation pit monitoring and measurement system according to claim 7, characterized in that: The central processing system (200) further comprises an early warning module, wherein 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 corresponding deformation thresholds, and 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.

9. A foundation pit monitoring method, applied to the foundation pit monitoring and measurement system according to any one of claims 1 to 8, characterized in that: The steps include: S1: a guide runway (8) is provided around the foundation pit (300), wherein the guide runway (8) is provided equidistantly 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) is able to navigate to the guide runway (8) and check whether the lifting device (3), the telescopic mechanism (5) and the camera module (2) are in normal working condition; S3: Start the AGV intelligent vehicle (1), and according to the Beidou satellite positioning module and position PID algorithm inside it, it will idle along the guide runway (8) under the instruction of the central processing system for one circle, which is used for path verification and equipment status confirmation; S4: After the path validation and the device status are confirmed, the working positions of the lifting device (3), the telescopic mechanism (5), and the first mounting frame (6) are adjusted in combination with the initial standard three-dimensional model coordinate data in the central processing system (200), and the position and attitude measurement system information of the camera module (2) are adjusted; S5: The AGV intelligent vehicle (1) is started again, and the camera module (2) continuously collects real-time image information of the foundation pit (300) during driving, and sends the real-time image information 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 on the foundation pit image and performs deformation analysis 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 value; S7: The early warning module of the central processing system (200) compares the deformation characteristic value with the deformation threshold value, and immediately activates the early warning mechanism when at least one deformation characteristic value is greater than the corresponding deformation threshold value.

10. The foundation pit monitoring method according to claim 9, characterized in that: 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) at an angle, the upper edge (9) of the camera module (2) covers the upper edge of the foundation pit (300) and extends to the position of the guide runway (8), and the lower edge (10) 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