Steel pipe concrete structure lateral pressure applying device and lateral loading instability judging method

Abstract

The present application belongs to the technical field of concrete structure stress monitoring, and discloses a steel pipe concrete structure lateral pressure device and a lateral loading instability judgment method. The cloud data control center uses the ICP algorithm to register the real-time point cloud with the pre-stored initial three-dimensional model to realize deformation quantitative analysis and surface defect evaluation, and outputs a strain anomaly index; a real-time demodulation light signal generates a stress distribution matrix, which is uploaded to the cloud data control center, and a difference operation is performed between the stress distribution matrix and a preset reference stress field matrix to generate a stress offset matrix, and an abnormality judgment threshold is dynamically corrected; the strain anomaly index and the stress offset matrix are coupled to construct an abnormality probability model, and the three-dimensional coordinates of the abnormal area are output to complete the abnormal positioning of the local buckling instability of the structure. The present application can effectively prevent the local instability damage of the steel pipe concrete structure.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of stress monitoring of concrete structures, and particularly relates to a steel pipe concrete structure lateral pressure applying device and a lateral loading instability judging method. BACKGROUND

[0002] The coal pillarless mining technology has a crucial influence on the development of coal, and the automatic cutting and pressure releasing and roadway forming technology is widely applied and causes a large amount of research because it can optimize the stress structure of surrounding rock, reduce the gas gathering of the working face and improve the economic benefits. After the implementation of the technology, the rock stratum in the cutting area collapses rapidly to form a gangue bank. With the fracture and rotation of the rock mass not cut by the roof, the vertical force in the vertical direction promotes the further crushing, swelling and compaction of the gangue bank, which causes the continuous increase of the internal horizontal extrusion force. This force will inevitably make the gangue bank swell and deform towards the free space of the roadway, and continuously act on the support structure beside the gangue roadway, and the pressure often appears with the deformation and damage of the support structure beside the roadway. In order to ensure the safety, economy and efficiency of the self-forming roadway surrounding rock control, it is necessary to develop a support and roof and bank structure with high strength, high stiffness and convenient construction to simultaneously bear the vertical load of the roof and the lateral load of the collapsed rock mass in the goaf. The steel pipe concrete structure, as a composite component with the core concrete constrained by the steel pipe, can effectively solve this demand. Under the constraint of the steel pipe, the core concrete is in a three-way compression state, which not only significantly improves the compression bearing capacity and crack resistance of the concrete, but also suppresses the local buckling of the steel pipe. This synergistic effect enables the CFST to fully utilize the performance advantages of the two materials, and has the characteristics of high bearing capacity and high stiffness. When applied to the support beside the roadway, the steel pipe concrete structure can not only effectively support the roof, but also resist the lateral pressure from the collapsed rock mass in the goaf. Therefore, in-depth research on the deformation behavior and control method of the structure under the action of lateral pressure has important theoretical and practical significance for promoting the progress of the coal pillarless roadway support technology.

[0003] At present, the determination of the resistance of the steel pipe concrete structure to lateral pressure mainly adopts the three-point bending determination method and the four-point bending determination method. However, in the actual application, the lateral extrusion load of the gangue presents a non-uniform distribution of "small on the top and large on the bottom", and at the same time bears the vertical load of the roof surrounding rock, which is easy to cause the local instability and damage of the steel pipe concrete structure. Moreover, the lateral pressure may cause the hoop stress of the steel pipe, which further affects the constraint effect on the core concrete. Higher lateral pressure may reduce the constraint effect of the steel pipe, thereby easily affecting the ultimate bearing capacity of the structure. SUMMARY

[0004] To overcome the problems existing in related technologies, the present invention discloses a lateral loading device and a method for judging lateral loading instability of steel-concrete composite structures. The purpose of this invention is to propose a lateral loading device and an intelligent monitoring and instability judgment method for steel-concrete composite structures. The assembly and disassembly operation is simple, and it can be combined with various types of testing machines to complete vertical and lateral joint loading tests. Through intelligent monitoring data system analysis, it provides real-time anomaly location of the steel-concrete composite structure, assists in optimizing the parameters of the steel-concrete composite structure, further strengthens the easily unstable areas of the steel-concrete composite structure, and prevents local instability and failure of the steel-concrete composite structure.

[0005] The technical solution is as follows: A method for judging the instability of steel-concrete composite structures under lateral loading, comprising the following steps:

[0006] S1 utilizes a 3D laser scanning device to scan the steel-concrete composite structure in real time, collecting point cloud data on the surface of the structure. This data is then uploaded to a cloud-based data control center via a wireless transmitter and receiver. The cloud-based data control center uses the ICP algorithm to register the real-time point cloud data with a pre-stored initial 3D model, enabling deformation quantification analysis and surface defect assessment, and outputting a strain anomaly index. ;

[0007] S2 is a distributed fiber optic stress sensor array that monitors points in a dot matrix arrangement, and demodulates optical signals in real time to generate a stress distribution matrix. Upload the stress distribution matrix to the cloud data control center. With the preset reference stress field matrix Perform difference operations to generate the stress offset matrix. Dynamically adjust the anomaly detection threshold ;

[0008] S3, Coupled Strain Anomaly Index With stress offset matrix An anomaly probability model is constructed, and the local buckling instability algorithm of a convolutional neural network (CNN) is used to output the three-dimensional coordinates of the abnormal region, thereby completing the local buckling instability anomaly localization of the structure.

[0009] In step S1, the cloud-based data control center adopts a hierarchical data processing architecture. After preprocessing, the raw point cloud enters the core analysis module for further processing, outputting the strain anomaly index. The core analysis modules include a raw point cloud preprocessing and data extraction module, an ICP registration algorithm module, and a strain anomaly index modeling module.

[0010] Furthermore, the point cloud preprocessing and data extraction module includes an outlier filtering stage, a data downsampling stage, and a feature enhancement stage;

[0011] The outlier filtering stage adopts a statistical filtering algorithm, and sets a field radius , is a point cloud density standard deviation, and points exceeding the range are removed;

[0012] The data downsampling stage is filtered according to an octree voxel, and the voxel size is adaptively adjusted according to the structural characteristics, and the expression is:

[0013] ;

[0014] In the formula, is the voxel size, is the maximum component size, The value of is 6-8;

[0015] The feature enhancement stage extracts the FPFH feature description factor, and samples the FPFH feature description factor, and the expression is:

[0016] ;

[0017] In the formula, is the normal angle, is the projection angle, is the distance ratio, is the fast point feature histogram of the point , is each point in the point cloud, is the neighborhood point set, is the total neighborhood point, is the total neighborhood point, is a weight function, which is used to measure the influence of point on point , which is usually related to the distance between them.

[0018] Further, the ICP registration algorithm module adopts a hierarchical ICP registration strategy to obtain a registered point cloud displacement field for a concrete-filled steel tube structure, and the hierarchical ICP registration strategy includes a coarse registration stage, a fine registration stage and a deformation analysis stage.

[0019] The coarse registration stage establishes an initial correspondence relationship by a RANSAC algorithm, and solves a rigid body deformation matrix:

[0020] ;

[0021] In the formula, is a rotation matrix, is a translation vector, are corresponding points of real-time point cloud and initial model respectively, is a rigid body transformation matrix, representing the transformation relationship of real-time point cloud to initial model, ​This represents the number of iterations for the RANSAC algorithm. More iterations increase the probability of finding the optimal transformation matrix, but also increase computation time. The choice of the number of iterations needs to be adjusted based on the noise level of the point cloud and the proportion of outliers.

[0022] The fine registration stage optimizes the objective function by introducing an adaptive weighting factor:

[0023] ;

[0024] In the formula, The weights are determined by the point cloud curvature similarity. Let be the objective function, representing the sum of weighted distances for all corresponding point pairs. This is a regularization parameter used to balance registration error and weighting factors;

[0025] The deformation analysis stage applies the above relationships obtained to the original point cloud data to calculate the displacement field of the registered point cloud:

[0026] ;

[0027] In the formula, The first in the original point cloud The position coordinates of each point Let be the displacement field, representing each point in space. The displacement vector, In the registered target point cloud, the first point cloud is compared with the first point cloud in the original point cloud. The coordinates of each point;

[0028] The strain anomaly index modeling module is derived from the multi-dimensional strain anomaly index defined by the registered displacement field calculation. :

[0029] ;

[0030] In the formula, For the maximum local deformation, For displacement gradient modulus, The change in curvature All are weighting coefficients. For the feature length or the scale of a local region, This represents the average displacement gradient modulus.

[0031] In step S2, fiber optic stress sensors are installed in a 54mm×54mm array along the circumferential and longitudinal directions of the steel-concrete composite structure to collect, analyze, and upload the data to the cloud data control center; this includes a data analysis system and dynamic threshold correction.

[0032] The data analysis system uses Hooke's law to calculate fiber strain in the collected data. with stress The discrete stress values are reconstructed into a matrix according to the spatial position, and a column coordinate system is established with the bottom surface of the column as the reference surface Therefore, the matrix element represents the stress value at position , and the stress distribution matrix is generated

[0033] ;

[0034] In the formula, is the radial layering, is the annular partition, and each 10° is a zone; is the height layer, and each 54mm is a layer;

[0035] The stress distribution matrix and the preset reference stress field matrix are subjected to difference operation in the data control center in the cloud, and a stress offset matrix is generated:

[0036] ;

[0037] The dynamic threshold correction dynamically adjusts the threshold value of the anomaly determination based on the statistical characteristics :

[0038] ;

[0039] In the formula, is the mean value, is the mean value of the change amount of the stress offset matrix, is an empirical coefficient.

[0040] In step S3, the coupling strain anomaly index and the stress offset matrix are constructed to build an anomaly probability model, including:

[0041] The strain anomaly index is linearly related to , and the anomaly probability model is:

[0042] ;

[0043] ;

[0044] In the formula, is the anomaly probability value, is a Sigmoid function, and the output probability value is ​​​is the Frobenius norm, quantifying the overall offset strength; the Sigmoid function maps the output to the interval [0, 1], is the negative exponential function with base e, used to convert the input x to a probability value;

[0045] determine as an exception, trigger the local buckling instability positioning algorithm based on convolutional neural network (CNN), output the three-dimensional coordinates of the abnormal area, and complete the structure local buckling instability positioning.

[0046] Another object of the present application is to provide a steel pipe concrete structure lateral pressure device, which implements the steel pipe concrete structure lateral loading instability judgment method, and the device comprises:

[0047] An external frame module is used for fixing and adjusting the steel pipe concrete structure lateral load applying device.

[0048] A lateral loading module is used for intelligently and real-timely regulating and controlling the lateral pressure.

[0049] A monitoring abnormal positioning module is used for real-timely monitoring the deformation characteristics of the steel pipe concrete structure during the lateral compression process of the steel pipe concrete structure, collecting data to construct a multi-source data fusion criterion, dynamically correcting the abnormal judgment threshold, realizing the structure abnormal positioning, and predicting the local instability prone area of the steel pipe concrete structure.

[0050] Further, the external frame module is provided with an external frame, which is a truss structure assembled by a transverse connecting beam and a vertical steel support column, the inner side of the vertical steel support column is welded with a sliding groove, and screw holes are reserved on both sides of the sliding groove, and the bottom is anchored to the bottom plate through foundation bolts of a fixed support;

[0051] The lateral loading module is arranged inside the external frame and is composed of an upper steel plate clamp and a lower steel plate clamp, the upper steel plate clamp and the lower steel plate clamp are symmetrically arranged and fixed in the screw holes through nuts, and positioning holes are arranged in the middle of the upper steel plate clamp and the lower steel plate clamp; an extension hydraulic rod is placed in the positioning hole, an arc-shaped steel plate is welded at the top end of the extension hydraulic rod and is in contact with the steel pipe concrete structure, a high-pressure oil pipe is reserved at the bottom of the extension hydraulic rod, a pressure gauge and an intelligent gate valve are installed on the high-pressure oil pipe, a small hydraulic pump is connected, the intelligent gate valve is programmed to control the pressure value of each node, and is used for simulating the lateral stress distribution of the goaf, and the intelligent gate valve is automatically closed when the preset pressure value is reached.

[0052] Further, the monitoring abnormal positioning module comprises a monitoring data collection module and an abnormal positioning module.

[0053] The monitoring data collection module scans the three-dimensional topographic data of each monitoring area of the steel pipe concrete structure through a three-dimensional laser scanning device in real time, and uploads the data to a cloud data control center through a wireless signal receiving transmitter; at the same time, the stress distribution is obtained in real time through a fiber optic stress sensing network, and the data is transmitted to the cloud data control center through a wireless signal receiving transmitter;

[0054] The abnormal positioning module stores and analyzes the data of the cloud data control center, and is used for predicting local buckling instability of the structure, and simultaneously transmits the data to a data imaging display platform through wireless transmission to monitor the deformation of the steel pipe concrete structure.

[0055] Further, the intelligent gate valve is programmed to control the pressure value of each node, which is used for simulating the lateral stress distribution of the goaf, including:

[0056] According to the target pressure value of each node input by the user or preset by the program, a small hydraulic pump is started to supply pressure to the system, the current node pressure value is obtained in real time through a pressure gauge, it is judged whether the current node pressure reaches the preset value, if not, the pressure supply is continued, if yes, the intelligent gate valve of the node is closed, it is checked whether all nodes reach the preset pressure, if not, the monitoring and control are continued, if yes, the small hydraulic pump is stopped, and the application is ended.

[0057] In combination with all the above technical solutions, the present application has the following beneficial effects:

[0058] Firstly, the device includes an external frame module, a lateral loading module, and a monitoring abnormal positioning module, the external frame module includes a fixed support, a vertical support column, a transverse connecting beam, a sliding groove, and a foundation bolt, the lateral loading module includes a steel plate, an extension hydraulic rod, a high-pressure oil pipe, a pressure gauge, an intelligent gate valve, and a small hydraulic pump, and the monitoring abnormal positioning module includes a three-dimensional laser scanning device, a fiber optic pressure sensor, a wireless signal receiving transmitter, and a data imaging display platform.

[0059] The lateral load applying device is composed of an external frame module and a lateral loading module, precise simulation of non-uniform lateral load is realized through intelligent valve control, and a dynamic change process is adapted in combination with an intelligent regulation and control mechanism. The monitoring module transmits signals to the cloud master control system through a wireless signal receiving transmitter through a three-dimensional laser scanning device and a pressure sensor. The abnormal positioning module judges the stability of the steel pipe concrete pier column by multi-source data fusion processing on the uploaded cloud database, locates the position of the local unstable pier column, and takes measures to reinforce and support the unstable part in time. The device is simple to assemble and disassemble, and can be combined with multiple types of test machines for axial and lateral double-direction combined loading test. The present application can simulate the real lateral load distribution of the gob side gangue extrusion through adjustable non-uniform pressure simulation, and can obtain the research on the steel pipe concrete structure under the action of real lateral load, which can effectively prevent the local instability and damage of the steel pipe concrete structure.

[0060] Secondly, through the intelligent loading and cloud early warning analysis technology, researchers can deeply study the response mechanism, failure mode and instability criterion of steel pipe concrete under complex real lateral load, accelerate the research and development and verification process of new, efficient and economic gob-side entry retaining support technology, and promote the improvement of related test standards. The stability of the steel pipe concrete structure of the gob-side entry retaining is studied, which provides an effective solution to the structure prone to instability for the fields of tunnel engineering and underground space development. The technology can be widely applied in tunnel support, mine roadway protection and underground building design optimization, improve the safety and service life of the structure, reduce the disaster control cost, and has significant market promotion value. Especially in the field of coal mining, the use of concrete-filled steel pipe support can effectively improve the stability and safety of the roadway and reduce the occurrence of roof collapse accidents, which meets the needs of coal mine safety production. Considering the huge market demand for mine roadway support, the application prospect of the technology is broad. In addition, the technology can also be applied to the stability monitoring and reinforcement of bridges, high-rise buildings and other structures, further expanding its commercial value.

[0061] Thirdly, the present application can accurately identify the "unstable position" of the steel pipe concrete pier column under a specific load in advance in the laboratory, providing effective data support for further improving the stability design of the structure, so that targeted reinforcement and support can be carried out on these key parts in actual engineering, changing passive rescue to active prevention, greatly reducing the later roadway maintenance cost and engineering failure risk. Through the intelligent lateral loading and cloud early warning analysis technology, the present application effectively simulates complex load environment and accurately locates the structure instability area, provides reliable early warning for the predictability of steel pipe concrete structure instability, and is closer to actual engineering application, providing a more reliable basis for the optimization selection design and maintenance of steel pipe concrete structure. BRIEF DESCRIPTION OF DRAWINGS

[0062] The accompanying drawings, which are incorporated into and form a part of the specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the disclosure;

[0063] Figure 1 is a schematic diagram of a lateral pressing device for a steel pipe concrete structure provided by an embodiment of the present application;

[0064] Figure 2 is a sectional view of a lateral pressing device for a steel pipe concrete structure provided by an embodiment of the present application;

[0065] Figure 3 is a programming control flow chart of an intelligent gate valve of a lateral pressing device for a steel pipe concrete structure provided by an embodiment of the present application;

[0066] Figure 4 is a connection schematic diagram of a fiber optic pressure sensor of a lateral pressing device for a steel pipe concrete structure provided by an embodiment of the present application;

[0067] Figure 5 is a flow chart of an abnormal positioning algorithm of a lateral pressing device for a steel pipe concrete structure provided by an embodiment of the present application;

[0068] In the figure: 1, external frame; 101, transverse connecting beam; 102, vertical steel support column; 103, fixed support; 104, anchor bolt; 105, sliding groove; 106, threaded hole; 2, pressure testing machine; 3, steel pipe concrete structure; 4, upper steel plate clamp; 5, lower steel plate clamp; 6, nut; 7, positioning hole; 8, telescopic hydraulic rod; 9, arc-shaped steel plate; 10, high-pressure oil pipe; 11, pressure gauge; 12, intelligent gate valve; 13, small hydraulic pump; 14, three-dimensional laser scanning device; 15, wireless signal receiver transmitter; 16, fiber optic stress sensor; 17, data imaging display platform. DETAILED DESCRIPTION

[0069] In order to make the above objectives, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application are described in detail below in conjunction with the accompanying drawings. In the following description, a large number of specific details are set forth in order to facilitate a full understanding of the present application. However, the present application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the spirit of the present application, so the present application is not limited to the specific embodiments disclosed below.

[0070] The application has the following innovative points: the lateral pressure device can more realistically simulate actual working conditions, provide more reliable and more actual working condition close mechanical property data, and directly guide the optimization of support structure design. The intelligent monitoring system can monitor the structure state in real time and give early warning, and the instability judgment method can more accurately evaluate the stability of the structure, thereby improving the safety and reliability of the steel pipe concrete structure in the roadway support. Compared with the traditional monitoring and pressure scheme, the method has significant improvement in accuracy, intelligence and practicality, provides a reference for engineering design and construction, provides more reliable technical support for the application of the steel pipe concrete structure in the coal mine roadway support, effectively prevents and controls the instability and damage of the roadway, and improves the safety production level of the coal mine.

[0071] Embodiment 1, as shown in the application, the embodiment of the application provides a lateral pressure device for a steel pipe concrete structure, which specifically comprises an external frame module, a lateral loading module and a monitoring abnormal positioning module. Figures 1-4

[0072] The external frame module is used for fixing and adjusting the lateral load applying device of the steel pipe concrete structure.

[0073] The lateral loading module is used for intelligent real-time regulation and control of the lateral pressure.

[0074] The monitoring abnormal positioning module is used for real-time monitoring of the deformation characteristics of the steel pipe concrete structure during the lateral pressure process of the steel pipe concrete structure, collecting data to construct a multi-source data fusion criterion, dynamically correcting the abnormal judgment threshold, realizing structure abnormal positioning, and predicting the local instability prone area of the steel pipe concrete structure.

[0075] Illustratively, the external frame module is provided with an external frame 1, the external frame 1 is a truss structure assembled by a transverse connecting beam 101 and a vertical steel support column 102, the inner side of the vertical steel support column 102 is welded with a sliding groove 105, both sides of the sliding groove 105 are provided with reserved threaded holes 106, and the bottom is anchored to the bottom plate through anchor bolts 104 and fixed supports 103.

[0076] ​The lateral loading module is arranged inside the outer frame 1 and is composed of an upper steel plate clamp 4 and a lower steel plate clamp 5, the upper steel plate clamp 4 and the lower steel plate clamp 5 are symmetrically arranged and are fixed at threaded holes 106 by nuts 6 to realize multi-interval and multi-stage positioning, and positioning holes 7 are arranged in the middle of the upper steel plate clamp 4 and the lower steel plate clamp 5; a telescopic hydraulic rod 8 is arranged in the positioning hole 7, an arc-shaped steel plate 9 is welded at the top end of the telescopic hydraulic rod 8 and is in contact with the steel pipe concrete structure, a high-pressure oil pipe 10 is arranged at the bottom of the telescopic hydraulic rod 8, a pressure gauge 11 and an intelligent gate valve 12 are arranged on the high-pressure oil pipe 10 and are connected with a small-sized hydraulic pump 13, the intelligent gate valve 12 is programmed to control the pressure value of each node, the lateral stress distribution of the goaf can be accurately simulated, and the intelligent gate valve is automatically closed when the preset pressure value is reached.

[0077] As Figure 3 The intelligent gate valve programming control flowchart of the steel pipe concrete structure lateral pressure applying device is provided by the embodiment of the application, the intelligent gate valve is programmed to control the pressure value of each node, and is used for simulating the lateral stress distribution of the goaf, and comprises the following steps:

[0078] According to the target pressure value of each node input by a user or preset by a program, the small-sized hydraulic pump is started to supply pressure to the system, the pressure value of the current node is monitored in real time through the pressure gauge, it is judged whether the pressure value of the current node reaches the preset value, if not, the pressure supply is continued, if yes, the intelligent gate valve of the node is closed, it is checked whether all nodes reach the preset pressure, if not, the monitoring and control are continued, and if yes, the small-sized hydraulic pump is stopped to end the application.

[0079] According to the target pressure value of each node input by a user or preset by a program, the small-sized hydraulic pump 13 is started to supply pressure to the system, the pressure value of the current node is monitored in real time through the pressure gauge 11, it is judged whether the pressure value of the current node reaches the preset value, if not, the pressure supply is continued (the intelligent gate valve 12 is kept open), if yes, the intelligent gate valve 12 of the node is closed, it is checked whether all nodes reach the preset pressure, if not, the monitoring and control are continued, and if yes, the small-sized hydraulic pump 13 is stopped to end the application.

[0080] The monitoring and abnormity positioning module comprises a monitoring data collection module and an abnormity positioning module.

[0081] The monitoring data collection module scans the three-dimensional topographic data of each monitoring area of the steel pipe concrete structure 3 in real time through a three-dimensional laser scanning device 14, and uploads the data to a cloud data control center through a wireless signal receiver 15; meanwhile, the stress distribution is obtained in real time through a fiber stress sensor 16 network, and the data is transmitted to the cloud data control center through the wireless signal receiver 15;

[0082] The anomaly location module stores and analyzes data from the cloud data control center to predict local buckling instability of the structure. Simultaneously, it wirelessly transmits the data to the data imaging display platform 17 to monitor the deformation of the steel-concrete composite structure.

[0083] For example, all data monitored by the steel-concrete composite structure 3 is uploaded to the cloud data control center for multi-source data fusion processing. Abnormal and unstable areas of the steel-concrete composite pier data are identified, located, and subsequently processed. An abnormal location model of the steel-concrete composite structure is constructed, and reinforcement processing is carried out in advance for areas prone to local instability.

[0084] Example 2, the present invention provides a method for judging the instability of steel-concrete composite structures under lateral loading (method for locating anomalies in lateral loading monitoring of steel-concrete composite structures), the method comprising:

[0085] S1, using a 3D laser scanning device 14, real-time scanning is performed on the key monitoring areas deployed in the steel-concrete composite structure 3. Point cloud data is collected on the surface of the key monitoring areas of the steel-concrete composite structure. The point cloud data is uploaded to the cloud data control center in real time via a wireless signal receiver and transmitter 15. The cloud data control center uses the ICP algorithm to register the real-time point cloud with the pre-stored initial 3D model, realizing deformation quantification analysis and surface defect assessment, and outputting the strain anomaly index. ;

[0086] Specifically, a 3D laser scanning device is used to scan the key monitoring areas (upper, middle, and lower) of the steel-concrete composite structure in real time, acquiring surface point cloud data at a frequency of ≥10Hz. This data is then uploaded to the cloud-based data control center via a wireless signal receiver and transmitter. The cloud-based system uses the ICP algorithm to register the real-time point cloud data with a pre-stored initial 3D model, enabling deformation quantification analysis and surface defect assessment, and outputting the strain anomaly index I. d ;

[0087] S2 is a distributed fiber optic stress sensor array that monitors points in a dot matrix arrangement, and demodulates optical signals in real time to generate a stress distribution matrix. Upload the stress distribution matrix to the cloud data control center. With the preset reference stress field matrix Perform difference operations to generate the stress offset matrix. Dynamically adjust the anomaly detection threshold ;

[0088] S3, Coupled Strain Anomaly Index With stress offset matrix , construct an abnormal probability model, and use a local buckling instability positioning algorithm of a convolutional neural network (CNN) to output three-dimensional coordinates of an abnormal region, thereby completing abnormal positioning of local buckling instability of the structure.

[0089] For example, assuming that the strain anomaly index is linearly related to Δ S , the abnormal probability model is:

[0090] ;

[0091] ;

[0092] In the formula, is an abnormal probability value, is a Sigmoid function, and an output probability value is outputted; is a Frobenius norm, which quantifies the overall offset strength; the Sigmoid function maps the output to the interval [0, 1], is a negative exponential function with a natural constant e as the base, which is used to convert the input x into a probability value;

[0093] If it is determined to be abnormal, a local buckling instability positioning algorithm based on a convolutional neural network (CNN) is triggered to output three-dimensional coordinates of an abnormal region, thereby completing abnormal positioning of local buckling instability of the structure. Fusion of point clouds and stress data can improve the accuracy of abnormal detection.

[0094] For example, step S1 specifically comprises: arranging a three-dimensional laser scanning device 14 at a key position of the concrete-filled steel tube structure 3, collecting point cloud data on the surface of a key monitoring area of the concrete-filled steel tube structure, uploading the point cloud data to a cloud data control center in real time through a wireless signal receiver, and using a layered data processing architecture by the cloud data control center. After preprocessing, the raw point cloud enters the core analysis module. Processing is performed, and a strain anomaly index is outputted; the strain anomaly index mainly includes a raw point cloud preprocessing and data extraction module, an ICP registration algorithm module, and a strain anomaly index modeling module;

[0095] The point cloud preprocessing and data extraction module includes an outlier filtering stage, a data downsampling stage, and a feature enhancement stage.

[0096] The outlier filtering stage uses a statistical filtering algorithm, sets a field radius ( , and removes points outside the range ; the data downsampling stage uses an octree voxel filtering algorithm, and the voxel size is adaptively adjusted according to the structure characteristics:

[0097] ;

[0098] In the formula, For voxel size, For the maximum component size, The value ranges from 6 to 8;

[0099] The feature enhancement stage extracts FPFH feature descriptors and samples them, with the expression being:

[0100] ;

[0101] In the formula, The angle between the normal and the direction of motion. The projection angle is... For distance ratio, For point Fast point feature histogram, For each point in the point cloud, For the neighborhood point set, For the total neighborhood points, is a weighting function used to measure the weight of points. Point The impact is usually related to the distance between them.

[0102] The ICP registration algorithm module is designed for the characteristics of steel-concrete composite structures and adopts a layered ICP registration strategy, which mainly includes a coarse registration stage, a fine registration stage, and a deformation analysis stage.

[0103] The coarse registration stage establishes an initial correspondence using the RANSAC algorithm and solves for the rigid body deformation matrix.

[0104] ;

[0105] In the formula, Let be a rotation matrix. It is a translation vector. These are the corresponding points in the real-time point cloud and the initial model, respectively. Let be the rigid body transformation matrix, representing the transformation relationship from the real-time point cloud to the initial model. This represents the number of iterations for the RANSAC algorithm. More iterations increase the probability of finding the optimal transformation matrix, but also increase computation time. The choice of the number of iterations needs to be adjusted based on the noise level of the point cloud and the proportion of outliers.

[0106] The fine registration stage optimizes the objective function by introducing an adaptive weighting factor:

[0107] ;

[0108] In the formula, The weight is determined by the point cloud curvature similarity, is the target function, representing the sum of the weighted distances of all corresponding point pairs, is a regularization parameter used to balance the registration error and the weight factor;

[0109] The deformation analysis stage applies the above relationship to the original point cloud data to calculate the post-registration point cloud displacement field:

[0110] ;

[0111] In the formula, is the position coordinate of the i-th point in the original point cloud, is the displacement field, representing the displacement vector of each point in space, is the position coordinate of the corresponding point in the target point cloud after registration corresponding to the i-th point in the original point cloud; The strain anomaly index modeling module calculates the multi-dimensional strain anomaly index from the post-registration displacement field:

[0112]

[0113] ;

[0114] In the formula, is the maximum local deformation, is the displacement gradient modulus, is the curvature change amount, are all weight coefficients, which are calibrated according to experiments; is the characteristic length or the scale of the local area, is the average displacement gradient modulus.

[0115] Exemplarily, in step S2, the optical fiber stress sensors are installed on the steel pipe concrete structure along the ring direction and the longitudinal direction of the pier column with a point array of 54mm×54mm, data is collected, and the cloud data control center is analyzed and uploaded. Mainly includes information collection system, data analysis system and dynamic threshold correction.

[0116] The data analysis system realizes the conversion of optical fiber strain and stress through Hooke's law, reconstructs the discrete point stress value into a matrix according to the spatial position, takes the bottom surface of the pier column as the reference surface, establishes the column coordinate system Therefore, the matrix element represents the stress value at position , and the stress distribution matrix is generated as follows:

[0117] ​​​ ;

[0118] In the formula, It is radially layered. The zones are divided into circumferential zones, each zone measuring 10°. The height is divided into layers, each layer being 54mm thick.

[0119] Upload the stress distribution matrix to the cloud data control center. With the preset reference stress field matrix Perform a difference operation to generate the stress offset matrix:

[0120] ;

[0121] The dynamic threshold correction is based on The statistical characteristics dynamically adjust the threshold for anomaly detection. :

[0122] ;

[0123] In the formula, The mean, This represents the mean of the changes in the stress offset matrix. This is an empirical coefficient.

[0124] For example, in step S3 above, the local buckling instability anomaly of the structure is located, the three-dimensional coordinates of the anomaly region are output, and the strain anomaly index is assumed. and The linear correlation and anomaly probability model are as follows:

[0125] ;

[0126] ;

[0127] In the formula, This represents the probability value of anomalies. This is the Sigmoid function, which outputs probability values. The Frobenius norm is used to quantize the overall offset strength; the Sigmoid function maps the output to the [0,1] interval. For the natural constant A negative exponential function with base 0.5 is used to convert the input... Convert to probability values;

[0128] If an anomaly is detected, a local buckling instability positioning algorithm based on a convolutional neural network (CNN) is triggered, outputting the three-dimensional coordinates of the anomaly region to complete the local buckling instability anomaly localization. The fusion of point cloud and stress data can improve the anomaly detection accuracy.

[0129] In step S3, the vulnerable area is reinforced and supported in advance according to the abnormal positioning model;

[0130] In step S3, the data transmission cloud data control center, the cloud service platform database is constructed by summarizing and analyzing, and the reinforcement scheme such as 'local hoop reinforcement' or 'grouting reinforcement' is generated through wireless network transmission to the imaging display platform. The database is applied to the scene to provide reliable early warning for the abnormal instability of the on-site roadside steel pipe concrete structure support, and ensure the safety of the on-site construction and the stability of the support.

[0131] Another exemplary, Figure 5 The abnormal positioning algorithm flowchart of the steel pipe concrete structure is provided by the embodiment of the application.

[0132] Embodiment 3, a test method of a steel pipe concrete structure lateral pressure device, the method is operated by the steel pipe concrete structure lateral pressure device, and the method comprises the following steps:

[0133] Step 1, assemble the lateral loading test device, which comprises an external frame module, a lateral loading module and a monitoring abnormal positioning module. According to the overall size of the existing pressure testing machine and the size of the test piece in the laboratory, the device is fixed and adjusted to realize vertical and lateral combined loading.

[0134] Step 2, make three steel pipe concrete structure test pieces with a size (diameter D x wall thickness t x height L) of 108 mm x 5 mm x 1200 mm. First, weld the steel plate to the bottom of the steel pipe, and pay attention to geometric alignment during welding. Then, pour the prepared core concrete into the opening of the steel pipe. The concrete used is of a design strength of C40. During grouting, use a vibrating rod to vibrate the inside of the steel pipe, and at the same time, shake the test piece vigorously to ensure that the concrete inside the steel pipe is uniform and dense. After the test piece is grouted, it is naturally cured together with the concrete test block for 28 days. During the curing process, water is continuously sprayed on the surface of the concrete to ensure a certain humidity, so that the concrete reaches the preset strength. After curing, fill the test piece with high-strength mortar to level the surface of the concrete that is lower than the surface of the steel pipe. For the test piece whose concrete surface is higher than the surface of the steel pipe, polish it to the same level as the steel pipe and the concrete surface to ensure that both can be stressed simultaneously during the loading process.

[0135] Step 3, place the test piece into the pressure testing machine and fix it. Install the lateral loading device on one side of the test piece, and install the optical fiber stress sensor 16 along the ring and longitudinal direction of the pier column with a 54mm x 54mm dot matrix.

[0136] Step 4, divide the surface of the other side of the test piece into three monitoring areas according to the upper, middle and lower positions. Deploy one three-dimensional laser scanning device 14 in each area. The three-dimensional laser scanning device 14 is 0.5-1m away from the surface of the pier column, and the elevation angle is ≤30° to avoid point cloud distortion.

[0137] Step five, start the pressure testing machine and lateral loading device to load, record the load, displacement, strain and other conditions of the steel pipe concrete structure sample during loading, and upload the corresponding monitoring data to the cloud data control center.

[0138] The above merely describes preferred specific embodiments of the present application, but the protection scope of the present application is not limited thereto, and any modification, equivalent replacement and improvement made by any person skilled in the art within the technical range disclosed by the present application and within the spirit and principle of the present application shall be encompassed within the protection scope of the present application.

Claims

1. A method for judging the instability of steel-concrete composite structures under lateral loading, characterized in that, The method includes the following steps: S1 utilizes a 3D laser scanning device to scan the steel-concrete composite structure in real time, collecting point cloud data on the surface of the structure. This data is then uploaded to a cloud-based data control center via a wireless transmitter and receiver. The cloud-based data control center uses the ICP algorithm to register the real-time point cloud data with a pre-stored initial 3D model, enabling deformation quantification analysis and surface defect assessment, and outputting a strain anomaly index. ; S2 is a distributed fiber optic stress sensor array that monitors points in a dot matrix arrangement, and demodulates optical signals in real time to generate a stress distribution matrix. Upload the stress distribution matrix to the cloud data control center. With the preset reference stress field matrix Perform difference operations to generate the stress offset matrix. Dynamically adjust the anomaly detection threshold ; S3, Coupled Strain Anomaly Index With stress offset matrix An anomaly probability model is constructed, and the local buckling instability algorithm of a convolutional neural network (CNN) is used to output the three-dimensional coordinates of the abnormal region, thereby completing the local buckling instability anomaly localization of the structure.

2. The method for judging the instability of steel-concrete composite structures under lateral loading according to claim 1, characterized in that, In step S1, the cloud-based data control center adopts a hierarchical data processing architecture. After preprocessing, the raw point cloud enters the core analysis module for further processing, outputting the strain anomaly index. The core analysis modules include a raw point cloud preprocessing and data extraction module, an ICP registration algorithm module, and a strain anomaly index modeling module.

3. The method for judging the lateral loading instability of steel-concrete composite structures according to claim 2, characterized in that, The point cloud preprocessing and data extraction module includes an outlier filtering stage, a data downsampling stage, and a feature enhancement stage. The outlier filtering stage employs a statistical filtering algorithm, with a set neighborhood radius. , The standard deviation of point cloud density is used to remove values ​​exceeding a certain threshold. Points within a range; The data downsampling stage uses octree voxel filtering, and the voxel size is adaptively adjusted according to structural features, expressed as follows: ； In the formula, For voxel size, For the maximum component size, The value ranges from 6 to 8; The feature enhancement stage extracts FPFH feature descriptors and samples them, with the expression being: ； In the formula, The angle between the normal and the direction of motion. The projection angle is... For distance ratio, For point Fast point feature histogram, For each point in the point cloud, For the neighborhood point set, For the total neighborhood points, is a weighting function used to measure the weight of points. Point The impact.

4. The method for judging the instability of steel-concrete composite structures under lateral loading according to claim 2, characterized in that, The ICP registration algorithm module is designed for concrete-filled steel tube structures. It employs a layered ICP registration strategy to obtain the displacement field of the registered point cloud. The layered ICP registration strategy includes a coarse registration stage, a fine registration stage, and a deformation analysis stage. The coarse registration stage establishes an initial correspondence using the RANSAC algorithm and solves for the rigid body deformation matrix. ； In the formula, For rotation matrix, It is a translation vector. These are the corresponding points in the real-time point cloud and the initial model, respectively. Let be the rigid body transformation matrix, representing the transformation relationship from the real-time point cloud to the initial model. This represents the number of iterations of the RANSAC algorithm. In the fine registration stage, the objective function is optimized by introducing an adaptive weighting factor: ； In the formula, The weights are determined by the point cloud curvature similarity. Let be the objective function. For regularization parameters; The relationships obtained in the deformation analysis stage are applied to the original point cloud data to calculate the displacement field of the registered point cloud: ； In the formula, The first in the original point cloud The position coordinates of each point Let be the displacement field, representing each point in space. The displacement vector, In the registered target point cloud, the first point cloud is compared with the first point cloud in the original point cloud. The coordinates of each point; The strain anomaly index modeling module originates from the registered displacement field and calculates and defines a multi-dimensional strain anomaly index. : ； In the formula, For the maximum local deformation, For displacement gradient modulus, The change in curvature All are weighting coefficients. For the feature length or the scale of a local region, This represents the average displacement gradient modulus.

5. The method for judging the instability of steel-concrete composite structures under lateral loading according to claim 1, characterized in that, In step S2, fiber optic stress sensors are installed in a 54mm×54mm array along the circumferential and longitudinal directions of the steel-concrete composite structure to collect data, analyze it, and upload it to the cloud data control center. This includes data analysis systems and dynamic threshold correction; The data analysis system uses Hooke's law to calculate fiber strain in the collected data. With stress The transformation reconstructs the discrete stress values ​​into a matrix according to their spatial location, and establishes a column coordinate system with the bottom surface of the pier as the reference plane. Therefore, matrix elements Indicates the location Stress value at Generate stress distribution matrix for: ； In the formula, It is radially layered. The zones are divided into circumferential zones, each zone measuring 10°. The height is divided into layers, each layer being 54mm thick. Upload the stress distribution matrix to the cloud data control center. With the preset reference stress field matrix Perform a difference operation to generate the stress offset matrix: ； The dynamic threshold correction is based on The statistical characteristics dynamically adjust the threshold for anomaly detection. : ； In the formula, The mean, This represents the mean of the changes in the stress offset matrix. This is an empirical coefficient.

6. The method for judging the lateral loading instability of steel-concrete composite structures according to claim 1, characterized in that, In step S3, the coupled strain anomaly index With stress offset matrix The construction of anomaly probability models includes: Strain Anomaly Index and The linear correlation and anomaly probability model are as follows: ； ； In the formula, This represents the probability value of anomalies. This is the Sigmoid function, which outputs probability values. The Frobenius norm is used to quantify the overall offset intensity. These are all weighting coefficients; the Sigmoid function maps the output to the [0,1] interval. For the natural constant A negative exponential function with base 0.5 is used to convert the input... Convert to probability values; If an anomaly is detected, a local buckling instability positioning algorithm based on a convolutional neural network (CNN) is triggered, outputting the three-dimensional coordinates of the abnormal region to complete the local buckling instability anomaly localization of the structure.

7. A lateral pressure application device for a steel-concrete composite structure, characterized in that, The device implements the method for determining lateral loading instability of steel-concrete composite structures as described in any one of claims 1-6. The device includes: External frame module, used to fix and adjust the lateral load application device of the steel-concrete composite structure; Lateral loading module for intelligent real-time control of lateral pressure; The anomaly monitoring and localization module is used to monitor the deformation characteristics of steel-concrete composite structures in real time during lateral compression. It collects data to construct multi-source data fusion criteria, dynamically corrects the anomaly judgment threshold, realizes the location of structural anomalies, and predicts the local unstable areas of the steel-concrete composite structure.

8. The lateral pressure application device for steel-concrete composite structures according to claim 7, characterized in that, The external frame module is provided with an external frame (1), which is a truss structure assembled from a transverse connecting beam (101) and a vertical steel support column (102). The inner side of the vertical steel support column (102) is welded with a sliding groove (105), and threaded holes (106) are reserved on both sides of the sliding groove (105). The bottom is anchored to the base plate with an anchor bolt (104) to the fixed support (103). The lateral loading module is set inside the outer frame (1) and consists of an upper steel plate clamp (4) and a lower steel plate clamp (5). The upper steel plate clamp (4) and the lower steel plate clamp (5) are symmetrically arranged and fixed at the threaded hole (106) by a nut (6). A positioning hole (7) is provided in the middle of the upper steel plate clamp (4) and the lower steel plate clamp (5). A telescopic hydraulic rod (8) is placed in the positioning hole (7). An arc-shaped steel plate (9) is welded to the top of the telescopic hydraulic rod (8) and is in close contact with the steel pipe concrete structure. A high-pressure oil pipe (10) is left at the bottom of the telescopic hydraulic rod (8). A pressure gauge (11) and an intelligent gate valve (12) are installed on the high-pressure oil pipe (10) and connected to a small hydraulic pump (13). The pressure value of each node is controlled by programming on the intelligent gate valve (12) to simulate the lateral force distribution of the goaf. When the preset pressure value is reached, the intelligent gate valve (12) automatically closes.

9. The lateral pressure application device for steel-concrete composite structures according to claim 7, characterized in that, The monitoring anomaly location module includes: a monitoring data collection module and an anomaly location module; The monitoring data collection module scans the three-dimensional topography data of each monitoring area of ​​the steel pipe concrete structure (3) in real time through a three-dimensional laser scanning device (14), and uploads it to the cloud data control center through a wireless signal receiver (15); at the same time, it obtains the stress distribution in real time through the fiber optic stress sensing (16) network, and transmits the data to the cloud data control center through the wireless signal receiver (15). The abnormal location module stores and analyzes data from the cloud data control center for predicting local buckling instability of the structure. At the same time, it transmits the data wirelessly to the data imaging display platform (17) to monitor the deformation of the steel pipe concrete structure.

10. The lateral pressure application device for steel-concrete composite structures according to claim 8, characterized in that, The intelligent gate valve (12) controls the pressure values ​​of each node through programming to simulate the lateral force distribution in the goaf, including: According to the user input or the program preset the target pressure value of each node, start the small hydraulic pump (13) to start supplying pressure to the system, monitor the pressure value of the current node in real time through the pressure gauge (11), and determine whether the pressure of the current node has reached the preset value. If it has not reached the preset value, continue to supply pressure. If it has reached the preset value, close the intelligent gate valve (12) of the node, check whether all nodes have reached the preset pressure. If there are still nodes that have not reached the preset pressure, continue to monitor and control. If all nodes have reached the preset pressure, stop the small hydraulic pump (13) and end the application.

Images