Segment posture detection and lifting method and system, segment lifting device and industrial control computer

Abstract

The present disclosure relates to a method and system for detecting and lifting a pipe segment posture, a pipe segment lifting device and an industrial computer. The method for detecting and lifting a pipe segment posture comprises: obtaining the pipe segment quality information and pipe segment posture information of the pipe segment to be lifted; judging whether the pipe segment to be lifted meets the lifting conditions according to the pipe segment posture information and pipe segment quality information of the pipe segment to be lifted; planning the pipe segment lifting route when the pipe segment to be lifted meets the lifting conditions, and controlling the pipe segment lifting device to complete the pipe segment lifting according to the planned pipe segment lifting route through the pipe segment lifting control system. The present disclosure can realize automatic transportation of pipe segments and automatic detection of pipe segment placement posture, and improve the intelligent level of pipe segment lifting.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel boring machines, and in particular to a segment posture detection and lifting method and system, a segment lifting device and an industrial control computer. Background Art

[0002] Prefabricated shield segments are commonly used in shield tunnel construction to support and shape the tunnel. The shield segment is the innermost barrier of the tunnel, and is responsible for resisting soil pressure, groundwater pressure, and some special loads. Related technologies The lifting of segments in tunnels relies on manual operation. Due to the cumbersome operation process of shield segment lifting, the small space and complex environment in the tunnel, the long lifting distance of shield segments, and the limited vision of operators, shield segment lifting currently has low work efficiency and high requirements for workers' operating level and experience. At the same time, there are major safety hazards during manual observation operations. Summary of the invention

[0003] In view of at least one of the above technical problems, the present disclosure provides a segment posture detection and lifting method and system, a segment lifting device and an industrial computer, which can realize automatic transportation of segments and automatic detection of segment placement posture, thereby improving the intelligence level of segment lifting.

[0004] According to one aspect of the present disclosure, a method for detecting and hoisting a pipe segment posture is provided, comprising:

[0005] Obtaining the segment quality information and segment posture information of the segment to be hoisted;

[0006] According to the segment posture information and segment quality information of the segment to be hoisted, it is judged whether the segment to be hoisted meets the hoisting conditions;

[0007] When the pipe segment to be lifted meets the lifting conditions, the pipe segment lifting route is planned, and the pipe segment lifting control system controls the pipe segment lifting device to complete the pipe segment lifting according to the planned pipe segment lifting route.

[0008] In some embodiments of the present disclosure, the segment posture detection and lifting method further includes:

[0009] During the segment lifting process, the three-dimensional space of the segment lifting is displayed and monitored in real time.

[0010] In some embodiments of the present disclosure, the real-time display and monitoring of the three-dimensional space for segment lifting includes:

[0011] During the segment lifting process, the laser radar is used to obtain the segment lifting route point cloud data in real time;

[0012] Determine whether there is any obstacle intrusion on the segment lifting route, where obstacle intrusion includes people, objects or vehicles entering the working space;

[0013] In the event of an obstacle invading the line, re-plan the lifting route.

[0014] In some embodiments of the present disclosure, planning the segment lifting route includes:

[0015] Based on the real-time point cloud data of the segment lifting space, the position information of the segment to be lifted and the position information of the unloading point on the segment transport cart, the segment lifting route is planned according to predetermined rules, wherein the predetermined rules include at least one of safety rules, energy-saving rules and efficiency rules.

[0016] In some embodiments of the present disclosure, judging whether the segment to be hoisted meets the hoisting conditions according to the segment posture information and the segment quality information of the segment to be hoisted includes:

[0017] Whether the pipe segment meets the lifting conditions is determined by judging at least one of the following: whether the pipe segment unloading point is occupied, whether the pipe segment posture can be grasped, whether the pipe segment quality meets the requirements, and whether there are unavoidable obstacles in the lifting space.

[0018] In some embodiments of the present disclosure, obtaining the segment quality information of the segment to be hoisted includes:

[0019] Through segment lifting space scanning, obtain the spatial point cloud data of the segment to be lifted area;

[0020] Input the spatial point cloud data into the pre-trained shield segment point cloud deep learning prediction model to obtain the segment point cloud data and segment type;

[0021] Determine the segment quality information based on the segment point cloud data.

[0022] In some embodiments of the present disclosure, obtaining the segment posture information of the segment to be hoisted includes:

[0023] Input the segment point cloud data into the pre-trained shield segment positioning structure point cloud deep learning prediction model to obtain the segment positioning structure position information and segment positioning structure type;

[0024] The segment position information and segment placement posture information are determined based on the segment positioning structure position information and segment point cloud data.

[0025] In some embodiments of the present disclosure, when the segment to be hoisted meets the hoisting conditions, the segment posture detection and hoisting method further includes:

[0026] The segment lifting control system is used to control the segment lifting device to automatically locate and grab the segment to be lifted; the segment lifting route is planned, and the segment lifting device is controlled to complete the segment lifting according to the planned segment lifting route.

[0027] In some embodiments of the present disclosure, the controlling the segment lifting device to automatically position and grab the segment to be lifted includes:

[0028] According to the position information of the segment positioning structure, the segment lifting device is controlled to move to the position directly above the segment to be lifted, wherein the segment positioning structure is a segment positioning circular hole;

[0029] The segment lifting device is controlled to open the clamping claw mechanism, and the positioning pin of the segment lifting device is controlled to be aligned with the segment positioning structure to grab the segment to be lifted.

[0030] In some embodiments of the present disclosure, the obtaining of spatial point cloud data of the area to be hoisted of the segment by spatial scanning of the segment hoisting includes:

[0031] Control the segment lifting device to reset through the segment lifting control system;

[0032] Through segment lifting space scanning, the original space point cloud data of the segment to be lifted area is obtained;

[0033] Filter the original spatial point cloud data.

[0034] In some embodiments of the present disclosure, determining the segment quality information according to the segment point cloud data includes:

[0035] Perform cylinder fitting through the segment point cloud data to obtain the segment cylinder model parameters, wherein the segment cylinder model parameters include the segment radius and the segment axis direction vector;

[0036] Obtain the segment surface curvature according to the segment axis direction vector and the segment radius;

[0037] According to the curvature of the segment, search the point cloud data within the segment surface range, find the surface defect features, and filter the inherent grouting circular holes and positioning holes information of the segment. The surface defect features include the surface hole features, surface missing features and surface convex features.

[0038] Whether the segment quality meets the relevant requirements is judged based on the size of the surface defect characteristics, where the segment quality includes segment surface damage and segment surface flatness.

[0039] In some embodiments of the present disclosure, the segment positioning structure is a segment positioning circular hole, and the step of inputting the segment point cloud data into a pre-trained shield segment positioning structure point cloud deep learning prediction model to obtain the segment positioning structure position information includes:

[0040] Input the segment point cloud data into the pre-trained shield segment positioning structure point cloud deep learning prediction model to obtain the segment positioning circular hole point cloud data;

[0041] The point cloud data of the segment positioning circular hole is averaged to obtain the spatial coordinates of the center of the segment positioning circular hole.

[0042] In some embodiments of the present disclosure, determining the segment position information and the segment placement posture information according to the position information of the segment positioning structure and the segment point cloud data includes:

[0043] According to the segment point cloud data, segment support structure platform height information and segment thickness information, the number and height of the segments in the segment lifting area are calculated;

[0044] According to the geometric parameter information of the segment cylinder and the point cloud data information of the segment cylinder, combined with the spatial point cloud data of the segment lifting, the posture of the segment is calculated to determine whether the segment meets the lifting conditions.

[0045] In some embodiments of the present disclosure, the step of calculating the posture of the segment based on the normal vector information of the segment cylinder axis and the segment cylinder point cloud data information, combined with the segment lifting space point cloud data, to determine whether the segment meets the lifting conditions includes:

[0046] According to the spatial coordinate information of the center of the segment positioning hole and the normal vector information of the segment cylindrical axis, the central spatial coordinates of the segment and the horizontal placement angle of the segment are determined;

[0047] Obtain the vertical angle between the segment and the ground based on the segment axis normal vector to determine whether the segment can be grabbed by the hoist;

[0048] Based on the point cloud data information of the segment cylinder and the point cloud data of the segment lifting space, it is determined whether there are obstacles in the segment grabbing route.

[0049] According to another aspect of the present disclosure, an industrial computer is provided, comprising:

[0050] The segment posture detection module is configured to obtain segment quality information and segment posture information of the segment to be hoisted;

[0051] The segment lifting module is configured to determine whether the segment to be lifted meets the lifting conditions based on the segment posture information and segment quality information of the segment to be lifted; if the segment to be lifted meets the lifting conditions, plan the segment lifting route, and control the segment lifting device through the segment lifting control system to complete the segment lifting according to the planned segment lifting route.

[0052] In some embodiments of the present disclosure, the industrial computer is used to execute operations to implement the segment posture detection and lifting method as described in any of the above embodiments.

[0053] According to another aspect of the present disclosure, an industrial computer is provided, comprising:

[0054] A memory for storing instructions;

[0055] The processor is used to execute the instructions so that the industrial computer performs operations to implement the segment posture detection and lifting method as described in any of the above embodiments.

[0056] According to another aspect of the present disclosure, a segment lifting device is provided, comprising a laser radar and an industrial computer as described in any of the above embodiments.

[0057] According to another aspect of the present disclosure, a segment posture detection and lifting system is provided, comprising a segment lifting control system and a segment lifting device as described in any of the above embodiments.

[0058] According to another aspect of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, the segment posture detection and lifting method as described in any of the above embodiments is implemented.

[0059] The present invention can realize automatic transportation of pipe segments and automatic detection of pipe segment placement posture, thereby improving the intelligence level of pipe segment lifting. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0061] Figure 1 Schematic diagrams of some embodiments of the pipe segment lifting device disclosed in the present invention.

[0062] Figure 2 Schematic diagrams of some embodiments of the segment posture detection and lifting method disclosed in the present invention.

[0063] Figure 3 Schematic diagrams of other embodiments of the segment posture detection and lifting method disclosed in the present invention.

[0064] Figure 4 Schematic diagrams of some embodiments of the pipe segment structure disclosed in the present invention.

[0065] Figure 5 Schematic diagram of some embodiments of the intelligent posture detection method of pipe segments disclosed in the present invention.

[0066] Figure 6 Schematic diagram of the geometric characteristics of the tube segment cylinder in some embodiments of the present disclosure.

[0067] Figure 7 Schematic diagram of the placement angles of pipe segments in some embodiments of the present disclosure.

[0068] Figure 8 Schematic diagrams of some embodiments of the intelligent segment lifting method disclosed in the present invention.

[0069] Fig. 9 Schematic diagram of some embodiments of the industrial computer disclosed in the present invention.

[0070] Fig.10 It is a schematic diagram of the structure of some other embodiments of the industrial control computer disclosed in the present invention. DETAILED DESCRIPTION

[0071] The technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all of the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present disclosure and its application or use. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0072] Unless specifically stated otherwise, the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure.

[0073] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0074] Technologies, methods, and apparatus known to ordinary technicians in the relevant field may not be discussed in detail, but where appropriate, such technologies, methods, and apparatus should be considered part of the authorization specification.

[0075] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0076] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0077] The inventor discovered through research that the related technology sets a camera on the segment assembly machine to achieve obstacle avoidance path planning.

[0078] However, the related technology uses cameras and builds two-dimensional image models to determine obstacles on the assembly path, but cannot accurately obtain obstacle distance information. In addition, the image model has high requirements on the on-site lighting environment. The path planning of traditional methods has poor adaptability in complex assembly environments.

[0079] In view of at least one of the above technical problems, the present disclosure provides a segment posture detection and lifting method and system, a segment lifting device and an industrial computer. The present disclosure is described below through specific embodiments.

[0080] Figure 1 Schematic diagrams of some embodiments of the pipe segment lifting device disclosed in the present invention. Figure 1 As shown, the intelligent segment lifting device disclosed in the present invention includes a laser radar 1, an industrial computer 2, and a warning device 3.

[0081] The laser radar 1 may be an array laser radar, which is installed in the front and rear directions of the segment lifting device. The laser radar 1 is used to obtain point cloud data of the segment lifting operation space.

[0082] In some embodiments of the present disclosure, a planar array laser radar 1 is installed at each of the front and rear positions of the segment lifting device, and the window surface of the laser radar 1 is parallel to the horizontal ground when it is installed.

[0083] The warning device 3 is installed on the segment lifting device and is used to provide sound and light warnings.

[0084] The industrial computer 2 is placed in the electrical control cabinet for segment lifting, and is connected to the area array laser radar through Ethernet to obtain the segment lifting space point cloud data scanned by the area array laser radar.

[0085] In some embodiments of the present disclosure, the industrial computer 2 is installed in a control cabinet on the top of the segment lifting device.

[0086] In some embodiments of the present disclosure, the laser radar and the industrial computer are connected via Gigabit Ethernet, using a standard RJ45 Ethernet interface.

[0087] In some embodiments of the present disclosure, the industrial computer 2 is connected to the segment lifting control system via Ethernet to achieve segment lifting motion control.

[0088] In some embodiments of the present disclosure, the industrial computer 2 is configured to move the segment lifting device to the preset segment lifting position through the segment lifting control system, turn on the laser radar to collect spatial point cloud data of the segment to be lifted area, and then transmit the data to the industrial computer through Ethernet. After data processing, the industrial computer obtains the segment position and the segment lifting route, and transmits control instructions to the segment lifting control system through Ethernet to complete the automated operation of the segment lifting.

[0089] According to another aspect of the present disclosure, a segment posture detection and lifting system is provided, including a segment lifting control system and a segment lifting control system as described in any of the above embodiments (for example Figure 1 The pipe segment lifting device described in embodiment).

[0090] In some embodiments of the present disclosure, the industrial computer 2 is configured to execute the pipe segment posture detection and lifting method of the present disclosure, wherein the pipe segment posture detection and lifting method of the present disclosure includes the pipe segment intelligent posture detection method of the present disclosure and the pipe segment intelligent lifting method of the present disclosure.

[0091] In some embodiments of the present disclosure, the intelligent posture detection method of the pipe segment of the present disclosure is deployed in the industrial computer 2. The industrial computer 2 is configured to receive the acquired spatial point cloud data of the pipe segment to be hoisted, automatically detect the pipe segment after data processing, calculate and obtain the pipe segment placement posture information and the spatial coordinates of the pipe segment positioning structure, and comprehensively judge whether the pipe segment meets the hoisting requirements according to the intelligently identified pipe segment category and pipe segment quality.

[0092] In some embodiments of the present disclosure, the intelligent segment lifting method of the present disclosure is deployed in the industrial computer 2. The industrial computer 2 is configured to obtain the real-time point cloud data of the segment lifting space, combine the position information of the segment to be lifted and the position information of the unloading point on the segment transport trolley, automatically plan the segment lifting route according to the principles of safety, energy saving, efficiency, etc., control the segment lifting device to automatically move to the target position through the segment lifting control system, realize the automation of segment lifting, and at the same time, process the route point cloud data during the segment transportation process to realize the safety protection of segment transportation and dynamically display the real-time three-dimensional picture of the lifting space in the upper computer and other equipment.

[0093] In the segment hoisting of the above embodiment of the present disclosure, if Figure 1 As shown, a preset position of the segment lifting device is set between the segment transport trolley 5 and the segment waiting lifting area 4. After the segment 6 lifting operation is started, the segment lifting device first travels to the preset position of the segment lifting device, and then starts the laser radar scanning workflow to obtain real-time spatial point cloud data of the segment waiting lifting area.

[0094] In some embodiments of the present disclosure, the industrial computer 2 can be configured to, after acquiring the spatial point cloud data of the segment to be hoisted, first, filter the noise data through a straight-through filtering algorithm, and perform sparse processing on the spatial point cloud data through voxel filtering to obtain the processed spatial point cloud data; secondly, input the collected spatial point cloud data into a pre-trained shield segment point cloud deep learning prediction model to obtain the segment point cloud data and the segment type; thirdly, input the predicted segment point cloud data into a pre-trained shield segment positioning structure point cloud deep learning prediction model to obtain the position information and type of the segment positioning structure, and calculate the segment position and placement posture information according to the position information of the segment positioning structure and the segment point cloud data; finally, obtain information such as segment surface damage and segment surface flatness according to the segment point cloud data to obtain the segment point cloud data. The quality of the segment is determined by judging whether the lifting conditions are met based on the acquired segment placement posture and positioning structure position information, segment transport trolley unloading point position information, segment lifting space point cloud data, and segment quality information. If the lifting conditions are not met, a light and sound alarm will be emitted through the warning device. If the lifting conditions are met, the segment lifting route will be automatically planned in accordance with the principles of safety, efficiency, and energy saving. The segment lifting control system will control the segment lifting device to automatically complete the segment lifting according to the planned route. At the same time, the segment route point cloud data will be collected and processed in real time during the segment lifting process, and the lifting space screen will be viewed in real time on the host computer, handheld terminal and other devices. If anomalies such as line invasion occur, the planned route will be adjusted in real time according to the actual situation to automatically avoid obstacles or terminate the segment lifting process, thereby realizing intelligent segment lifting operations.

[0095] The above-mentioned embodiments of the present disclosure can quickly and accurately obtain the segment posture, segment type, segment quality, and segment positioning structure position information from the working space, automatically and in real time plan the segment lifting route according to the principles of safety, efficiency, and energy saving, realize the intelligent positioning and automatic lifting of the segments, improve the efficiency of segment lifting, and at the same time be able to view the real-time three-dimensional picture of the lifting space in real time, effectively solving the problems of accurate, stable, safe, and reliable lifting and transportation of the segments, and greatly reducing the labor intensity of on-site professional operators.

[0096] In view of the technical problems that the related technology uses cameras and establishes two-dimensional image models to judge obstacles on the assembly path, it is impossible to accurately obtain the obstacle distance information, and the image model has high requirements on the on-site lighting environment. The path planning of the traditional method has poor adaptability in a complex assembly environment. The above-mentioned embodiment of the present invention is based on the technical solution of laser three-dimensional point cloud, has low requirements on the on-site environment, can accurately identify the three-dimensional information of obstacles, and uses a path planning algorithm based on deep reinforcement learning to adapt to more complex scenes.

[0097] The following describes the segment posture detection and lifting method disclosed in the present invention, as well as the structure and function of the industrial computer disclosed in the present invention through specific embodiments.

[0098] Figure 2 Schematic diagram of some embodiments of the method for detecting and lifting the pipe segment posture disclosed in the present invention. Preferably, Figure 2 The embodiment can be executed by the industrial computer disclosed in the present invention or the intelligent lifting device for pipe segments disclosed in the present invention. Figure 2 As shown, Figure 2 The method of the embodiment may include steps 21 to 23, wherein:

[0099] Step 21, obtaining the segment quality information and segment posture information of the segment to be hoisted.

[0100] In some embodiments of the present disclosure, in step 21, the step of obtaining the segment quality information of the segment to be lifted may include: obtaining the spatial point cloud data of the segment to be lifted area through segment lifting space scanning; inputting the spatial point cloud data into a pre-trained shield segment point cloud deep learning prediction model to obtain the segment point cloud data and the segment type; and determining the segment quality information based on the segment point cloud data.

[0101] In some embodiments of the present disclosure, the segment quality information may include whether the segment size meets the requirements, the flatness of the segment surface, the damaged protrusions on the segment surface, the damage on the segment surface, and the flatness of the segment surface.

[0102] In some embodiments of the present disclosure, in step 21, the step of obtaining the segment posture information of the segment to be lifted may include: inputting the segment point cloud data into a pre-trained shield segment positioning structure point cloud deep learning prediction model to obtain the segment positioning structure position information and the segment positioning structure type; determining the segment position information and the segment placement posture information based on the segment positioning structure position information and the segment point cloud data.

[0103] In some embodiments of the present disclosure, determining the segment position information and segment placement posture information based on the segment positioning structure position information and segment point cloud data may include: performing cylinder fitting through the acquired segment point cloud data, judging the vertical inclination angle of the segment placement through the cylinder axis normal vector, and judging the horizontal angle of the segment surface by identifying the positioning structure information and the segment normal vector information.

[0104] Step 22, judging whether the segment to be lifted meets the lifting conditions according to the segment posture information and segment quality information of the segment to be lifted.

[0105] In some embodiments of the present disclosure, step 22 may include: judging whether the segment unloading point is occupied, whether the segment posture can be grasped, whether the segment quality meets the requirements, and whether there are unavoidable obstacles in the lifting space to determine whether the segment meets the lifting conditions.

[0106] Step 23, when the pipe segment to be lifted meets the lifting conditions, plan the pipe segment lifting route, and control the pipe segment lifting device through the pipe segment lifting control system to complete the pipe segment lifting according to the planned pipe segment lifting route.

[0107] Figure 3 Schematic diagrams of other embodiments of the method for detecting and lifting pipe segments disclosed in the present invention. Preferably, Figure 3 The embodiment can be executed by the industrial computer disclosed in the present invention or the intelligent lifting device for pipe segments disclosed in the present invention. Figure 3 As shown, Figure 3 The method of the embodiment may include steps 100 and 200, wherein:

[0108] Step 100, segment intelligent posture detection method.

[0109] In some embodiments of the present disclosure, step 100 may include: Figure 2 Step 21 of the embodiment.

[0110] Step 200, intelligent lifting method for pipe segments.

[0111] In some embodiments of the present disclosure, step 200 may include: Figure 2 Step 22-Step 23 of the embodiment.

[0112] The above embodiments of the present disclosure are intended to establish a segment lifting space coordinate system for reference, such as Figure 1 As shown, the origin of the three-dimensional space coordinate system is set at the boundary position of the segment lifting area, the forward direction of the segment lifting device along the segment lifting area is the positive direction of the X-axis, the viewing angle is toward the segment transport trolley, the left-hand direction is the positive direction of the Y-axis, and the vertical upward direction is the positive direction of the Z-axis; a preset position of the segment lifting device is set between the segment transport trolley and the segment lifting area, and after the segment lifting is started, the segment lifting device automatically runs to the preset position of the segment lifting device, and then executes the subsequent operation process; according to the coordinates of the laser radar installation position in the segment lifting space, the point cloud data collected by the laser radar is converted from the laser radar space coordinates to the segment lifting space coordinates, and then the data calculation and processing are performed.

[0113] In some embodiments of the present disclosure, let P = (x, y, z) represent a spatial coordinate point in the segment lifting spatial coordinate system, Pcd = {P1, P2, ..., Pn} represent a point cloud data set, and Seg = {P1, P2, ..., Pn} represent a segment point cloud data set.

[0114] Figure 4 Schematic diagrams of some embodiments of the pipe segment structure disclosed in the present invention. Figure 4 As shown, the segment includes two segment positioning circular holes 6-1.

[0115] Figure 5Schematic diagram of some embodiments of the intelligent posture detection method of pipe segments disclosed in the present invention. Preferably, Figure 5 The embodiment can be executed by the industrial computer disclosed in the present invention or the intelligent lifting device for pipe segments disclosed in the present invention. Figure 5 As shown, the intelligent posture detection method of the pipe segment disclosed in the present invention (for example Figure 2 Step 21 of the embodiment or Figure 3 Step 100 of the embodiment may include at least one of steps 110 to 170, wherein:

[0116] Step 110: training a deep learning prediction model for shield segment point cloud and a deep learning prediction model for shield segment positioning structure point cloud.

[0117] In some embodiments of the present disclosure, step 110 may include at least one of steps 111 to 113, wherein:

[0118] Step 111, collecting segment point cloud data in different scenarios on site, using annotation tools to create segment training data sets according to segment types, and using annotation tools to create segment positioning structure training data sets according to segment positioning structures.

[0119] Step 112, design a point cloud residual multilayer perceptron deep learning network, import the data set annotated in (1) to train the shield segment point cloud deep learning prediction model and the shield segment positioning structure point cloud deep learning prediction model respectively.

[0120] Step 113, saving the trained shield segment point cloud deep learning prediction model and shield segment positioning structure point cloud deep learning prediction model files.

[0121] Step 120, the segment lifting device is reset. After the segment lifting is started, the segment lifting control system resets the segment lifting device to the preset position of the segment lifting device, and the grabbing device rotates and resets to the initial setting state.

[0122] Step 130, segment lifting space scanning. After the segment lifting device is reset, the segment lifting control system sends a space scanning command to the industrial computer. The industrial computer turns on the laser radar to scan the segment to-be-lifted area, obtains a high-density point cloud data set Pcd1 of the segment to-be-lifted area, and transmits it to the industrial computer via Ethernet.

[0123] Step 140: filtering and processing the original point cloud data.

[0124] In some embodiments of the present disclosure, step 140 may include at least one of steps 141 to 142, wherein:

[0125] Step 141, through the straight-through filtering method, filter the point cloud data outside the working space of the pipe segment lifting area, remove the interference data, reduce the amount of point cloud data to be processed, and generate a filtered point cloud data set Pcd2.

[0126] Step 142, downsampling is achieved by improving the voxel grid method. Voxel filtering can achieve point cloud data sparseness without destroying the geometric structure of the point cloud itself, thereby improving the point cloud data processing speed. A three-dimensional voxel grid is created by inputting the point cloud data Pcd2, and then in each voxel, the center of gravity of all points in the voxel is used to approximate the other points in the voxel. All points in the voxel are finally represented by a center of gravity point, and a filtered sparse point cloud data set Pcd3 is generated.

[0127] Step 150, shield segment identification and quality inspection.

[0128] In some embodiments of the present disclosure, step 150 may include at least one of steps 151 to 153, wherein:

[0129] Step 151, input the spatial point cloud data into a pre-trained shield segment point cloud deep learning prediction model to obtain the segment point cloud data and segment type.

[0130] In some embodiments of the present disclosure, step 151 may include: inputting the filtered point cloud data Pcd3 into the shield segment point cloud deep learning prediction model trained in S0, obtaining the segment point cloud data set {Seg1, Seg2,…, SegN}, and obtaining the segment type (standard segment, adjacent segment, capped segment) information.

[0131] Step 152, performing cylinder fitting through the segment point cloud data to obtain the segment cylinder model parameters.

[0132] In some embodiments of the present disclosure, the segment cylinder model parameters may include parameters such as the segment radius, the segment axis direction vector, and the three-dimensional coordinates of a point on the axis.

[0133] In some embodiments of the present disclosure, step 152 may include: using an improved random sampling consistency algorithm to fit the segment point cloud Seg obtained by the segment point cloud deep learning prediction model to a cylinder, randomly selecting a sample subset from the sample, using the minimum variance estimation algorithm to calculate the cylinder model parameters for this subset, and then calculating the deviation between all samples and the cylinder model, and then using a pre-set threshold to compare the deviation. When the deviation is less than the threshold, the sample point belongs to the sample point inside the cylinder model, referred to as the inner point, otherwise it is a sample point outside the cylinder model, referred to as the outer point. Record the current number of inner points and then repeat this process. For each repetition, the current best cylinder model parameters are recorded, that is, the number of inner points is the largest, and the corresponding number of inner points at this time is the best number of inner points. At the end of each iteration, an iteration end judgment factor is calculated based on the expected error rate, the best number of inner points, the total number of samples, and the current number of iterations. After the iteration ends, the best model parameters are the final estimated values ​​of the cylinder model parameters. The above algorithm is used to obtain the segment cylinder axis direction vector (a, b, c), cylinder radius r and other information, such as Figure 6 shown. Figure 6 Schematic diagram of the geometric characteristics of the tube segment cylinder in some embodiments of the present disclosure.

[0134] Step 153, the segment point cloud Seg obtained by the segment point cloud deep learning prediction model is used to establish a point cloud space topological relationship using an improved dynamic grid k-neighborhood algorithm, detect segment surface damage and segment surface flatness, and determine whether the segment quality meets the requirements.

[0135] In some embodiments of the present disclosure, step 153 may include: obtaining the surface curvature of the segment according to the segment axis direction vector and the segment radius information obtained in step 153; searching the point cloud data within the segment surface range according to the improved dynamic grid k-neighborhood algorithm in combination with the segment curvature, finding the surface defect features, and providing a basis for detecting segment damage and flatness, wherein the surface defect features include surface hole features, surface missing features, and surface protrusion features; and filtering the information such as the inherent grouting circular holes and positioning holes of the segment; judging whether the segment quality meets the relevant requirements according to the size of the surface defect features, wherein the segment quality includes segment surface damage and segment surface flatness.

[0136] Step 160: Positioning circular hole target detection.

[0137] In some embodiments of the present disclosure, step 160 may include at least one of steps 161 to 162, wherein:

[0138] Step 161, input the segment point cloud data obtained in step 150 into the shield segment positioning structure point cloud deep learning prediction model trained in step 110 to obtain the segment positioning circular hole point cloud data.

[0139] In some embodiments of the present disclosure, the circular hole point cloud data is the circular boundary points, and the three-dimensional coordinate point data information on the circumference is fitted.

[0140] Step 162, performing mean processing on the segment positioning circular hole point cloud data to obtain the spatial coordinates P1 and P2 of the center of the segment positioning circular hole.

[0141] In some embodiments of the present disclosure, the circular characteristics are taken into consideration, and the mean processing is based on taking the maximum and minimum points in any axial direction on the circumference of a positioning circular hole, obtaining the mean, and then calculating the center point of the two mean points to filter out interference data during circular fitting.

[0142] Step 170: segment posture detection.

[0143] In some embodiments of the present disclosure, step 170 may include at least one of steps 171 to 172, wherein:

[0144] Step 171, based on the segment point cloud data set {Seg1, Seg2, ..., SegN} obtained in step 150, combined with the segment support structure platform height information and segment thickness information, calculate the segment stack number and stack height information in the segment lifting area. At the same time, based on the fitted segment cylinder axis normal vector geometric parameter information and segment cylinder point cloud data information, combined with the segment lifting space point cloud data, calculate the segment posture to determine whether the segment meets the lifting conditions.

[0145] In some embodiments of the present disclosure, the segment cylinder axis normal vector geometric parameter information may be determined in step 152, such as Figure 6 The information shown is the segment cylinder axis direction vector (a, b, c), cylinder radius r and other information.

[0146] In some embodiments of the present disclosure, in step 171, the step of calculating the number of segment stacks and stack height information of the segment in the segment lifting area based on the acquired segment point cloud data set, the segment lifting area platform height information, and the segment thickness information may include: according to the number of acquired segment point cloud data sets, the number of segment stacks can be known (the segment stacks intelligently obtain the top segment point cloud data, and the segments below are blocked and no information can be obtained); according to the height information of the segment support structure platform in the segment lifting area and the center height information of the surface segment point cloud data, the height of the stack is obtained by the difference between the two, and the thickness of each stack is judged according to the segment thickness information.

[0147] In some embodiments of the present disclosure, in step 171, the step of calculating the posture of the segment based on the normal vector information of the segment cylinder axis and the point cloud data information of the segment cylinder, combined with the segment lifting space point cloud data, to determine whether the segment meets the lifting conditions may include: determining the center space coordinates and the horizontal placement angle of the segment based on the space coordinate information of the center of the segment positioning hole and the normal vector information of the segment cylinder axis; obtaining the vertical angle between the segment and the ground based on the normal vector of the segment axis to determine whether the segment can be grabbed by the hoist; determining whether there are obstacles in the segment grabbing route based on the point cloud data information of the segment cylinder, combined with the segment lifting space point cloud data, to prevent obstacles vertically above the segment from affecting the normal lifting of the segment.

[0148] In some embodiments of the present disclosure, the segment posture mainly includes the vertical inclination angle between the segment and the ground and the horizontal placement angle of the segment.

[0149] In some embodiments of the present disclosure, when the vertical angle between the segment and the ground is too large, that is, when the inclination angle of the segment is large, the height difference of the curved surface of the segment is too large, resulting in the inability of the hanger to grab the segment; when the horizontal angle of the segment exceeds the rotation angle of the hanger, it cannot be grabbed.

[0150] Step 172, based on the spatial coordinate information of the center of the segment positioning hole obtained in S5 and the normal vector information of the segment cylindrical axis, the following is accurately determined according to formula (1): Figure 7 The center spatial coordinates P0 of the pipe segment and the horizontal placement angle θ of the pipe segment are shown:

[0151]

[0152] In formula (1), P1 and P2 are the spatial coordinates of the positioning holes, such as Figure 7 shown. Figure 7 Schematic diagram of the placement angles of pipe segments in some embodiments of the present disclosure.

[0153] In other embodiments of the present disclosure, the intelligent posture detection technology for pipe segments can use a binocular camera to obtain pipe segment and pipe segment positioning structure information through machine learning methods. Compared with point cloud data, the position information obtained by image recognition has lower accuracy and is easily affected by ambient lighting, and cannot fully obtain information on more complex geometric structure components.

[0154] Figure 8 Schematic diagram of some embodiments of the intelligent lifting method of pipe segments disclosed in the present invention. Preferably, Figure 8 The embodiment can be executed by the industrial computer disclosed in the present invention or the intelligent lifting device for pipe segments disclosed in the present invention. Figure 5 As shown, the intelligent lifting method of pipe segments disclosed in the present invention (for example Figure 2 Steps 22-23 of the embodiment or Figure 3Step 200 of the embodiment may include at least one of steps 210 to 240, wherein:

[0155] The present invention needs to perform coordinate conversion before path planning, converting the point cloud data from each laser radar independent coordinate system into the segment lifting space coordinate system, and processing the point cloud data in the segment lifting space coordinate system to obtain the motion path of the segment lifting device.

[0156] Step 210: training of segment lifting path planning model.

[0157] In some embodiments of the present disclosure, step 210 may include at least one of steps 211 to 212, wherein:

[0158] Step 211, design a deep reinforcement learning improved Dueling deep Q network for segment lifting path planning. Initialize the experience pool and environment, implement network learning according to control strategies such as safety, efficiency, and energy saving, and automatically plan routes and avoid obstacles. Input the state s sequence, obtain the Q value Q(s,a) of the corresponding action; select the action a' with the largest Q value and execute it; after execution, the environment changes, and the reward r of the environment can be obtained; use the reward r to update Q(s,a') and continue learning until a better network model is obtained.

[0159] In some embodiments of the present disclosure, the safety strategy refers to avoiding obstacles and maintaining a safe distance when grabbing, transporting, and releasing the pipe segments; the efficiency strategy refers to finding the shortest transportation path on the principle of meeting safety; and the energy-saving strategy refers to reducing energy-consuming actions such as crane lifting as much as possible when planning the lifting route.

[0160] Step 212, save the trained deep reinforcement learning improved Dueling deep Q network model to obtain the segment lifting path planning model.

[0161] Step 220: Automatic positioning and grabbing of the pipe segment.

[0162] In some embodiments of the present disclosure, step 220 may include at least one of steps 221 to 223, wherein:

[0163] Step 221, if the intelligent segment posture detection method determines that the segment placement posture does not meet the lifting conditions or the segment lifting path planning model determines that the segment lifting device cannot be operated, the industrial computer sends a command to the segment lifting device control system, and the control system issues an alarm to prompt relevant personnel to handle the fault and terminate the segment lifting. If the segment meets the lifting conditions, the segment lifting safe operation space is automatically calculated based on the collected point cloud data of the segment lifting space, and the best driving route is planned to avoid obstacles and achieve safe and reliable segment lifting.

[0164] In some embodiments of the present disclosure, the segment lifting path planning model determines that the conditions that cannot be met for the operation of the segment lifting device may include: safety, efficiency, and energy saving. When the model determines that the crane cannot safely grab the segment (segment position problem), cannot safely release the segment (segment release point is occupied), and the segment lifting obstacles cannot be avoided, the segment lifting device cannot be operated.

[0165] Step 222, based on the X and Y information of the spatial coordinates P0 of the lifting center of the segment to be grabbed, the optimal driving route is calculated through the segment lifting path planning model, and the industrial computer sends a movement instruction to the segment lifting control system, and the segment lifting control system moves the segment crane to the top of the segment to be lifted.

[0166] Step 223, based on the calculated Z coordinate information in the spatial coordinate P0 of the center of the segment, the segment lifting control system lowers the hoist to 90 cm above the segment to be lifted and opens the segment hoist claw mechanism, rotates the hoist based on the calculated normal vector information of the segment cylinder axis and the segment positioning hole position information (P1, P2 and θ), and grabs the segment after adjusting the positioning pin on the segment hoist to align with the segment positioning hole.

[0167] Step 230, segment lifting path planning. After the segment is successfully lifted, the host computer obtains the lifting space point cloud data through the laser radar, combines the unloading point location information of the segment transport trolley, and automatically plans the segment lifting device operation route through the segment lifting path planning model, automatically avoids obstacles in the segment lifting space, and selects the optimal operation route based on the principles of safety, efficiency, and energy saving.

[0168] In some embodiments of the present disclosure, step 230 may include: when the lifting route is automatically planned, the height of the lifting equipment will be automatically adjusted according to the principles of safety, efficiency, energy saving, etc.

[0169] Step 240, safety protection and monitoring of pipe segment lifting. During pipe segment lifting, the host computer obtains the pipe segment lifting route point cloud data in real time through the laser radar, determines whether there are obstacles invading the pipe segment lifting route, and replans the lifting route through the pipe segment lifting path planning model when the line is invading. If the pipe segment lifting conditions cannot be met, an alarm is issued through the warning device and the pipe segment lifting is suspended until the obstacle is removed. At the same time, a three-dimensional spatial model of pipe segment lifting is established based on the real-time spatial point cloud data, and is displayed in real time on the host computer, handheld terminal and other devices.

[0170] In some embodiments of the present disclosure, line intrusion refers to personnel, engineering machinery vehicles, and various objects that intrude into the work space during operation. During laser scanning in the above embodiments of the present disclosure, it is determined whether the relevant objects obstruct the lifting route and a corresponding judgment is made.

[0171] In some embodiments of the present disclosure, the pipe segment lifting condition in step 240 means that the pipe segment has been grabbed by the crane during the lifting process, which only involves path planning. Pipe segment lifting condition: There is no lifting route that can ensure safety in the lifting space, and obstacles cannot be avoided.

[0172] In other embodiments of the present disclosure, segment lifting path planning may adopt traditional algorithms and basic deep learning algorithms, but these algorithms have poor robustness, cannot respond quickly to complex environmental changes, have slow convergence speeds, and have poor application effects in actual scenarios.

[0173] The above-mentioned embodiments of the present disclosure are based on constructing a deep reinforcement learning model of lidar point cloud, which can realize automatic planning of the pipe segment lifting path.

[0174] The above-mentioned embodiments of the present disclosure are based on constructing a laser radar point cloud deep learning model, which can realize the detection of the placement posture of shield segments.

[0175] The above-mentioned embodiment of the present disclosure can realize the real-time three-dimensional spatial display and safety monitoring of the pipe segment lifting by arranging a laser radar on the pipe segment lifting device.

[0176] The above-mentioned embodiments of the present disclosure can realize the quality inspection of the pipe segments to be hoisted based on the laser radar point cloud data processing.

[0177] The segment posture detection and lifting method of the above-mentioned embodiment of the present invention can automatically, in real time and comprehensively scan the point cloud data of the segment lifting space objects, obtain the spatial position of the segment to be lifted, the segment type, the segment quality, the segment placement posture, the segment positioning structure position and other information, and automatically plan the driving route of the segment lifting device in combination with the lifting space point cloud data. During the segment lifting process, it can protect and dynamically adjust the lifting route in real time according to the acquired route point cloud data, and display the three-dimensional picture of the lifting space in real time, thereby improving the segment lifting efficiency, reducing labor intensity, and improving the intelligence level of segment lifting.

[0178] The present disclosure identifies pipe segments through a deep learning method of point cloud data, and can realize automatic path planning and obstacle avoidance operation for pipe segment transportation. However, the image monitoring and protection system in another related technology cannot obtain the three-dimensional spatial information of the pipe segment transportation route, relies on two-dimensional images with a large misjudgment rate and cannot identify obstacles such as on-site ropes that cannot be effectively represented by images.

[0179] In view of the technical problem that in another related technology, after the intelligent camera obtains the segment image, it is necessary to manually mark and confirm the position of the shield segment, and then calculate the position of the segment positioning hole through data processing to grab the segment, the present disclosure uses point cloud data deep learning technology to directly process point cloud data to identify the segment and obtain the coordinate information of the segment positioning hole. The present disclosure embodiment can realize automatic segment positioning, and the laser radar collects point cloud data that is disturbed by the on-site environment, which can realize segment lifting path planning and automatic protection.

[0180] With regard to another related technology that uses image recognition to obtain segment shape information, it is necessary to rely on the QR code information pre-set on the shield segment to obtain the specific type of the segment, and fails to achieve automatic identification. The segment transportation runs on a fixed height route. The above-mentioned embodiment of the present disclosure can realize automatic adjustment of the segment lifting height and automatic obstacle avoidance protection.

[0181] Fig. 9 Schematic diagram of some embodiments of the industrial control computer disclosed in the present invention. Fig. 9 As shown, the industrial computer (for example Figure 1 The industrial computer 2) of the embodiment may include a segment posture detection module 91 and a segment lifting module 92, wherein:

[0182] The segment posture detection module 91 is configured to obtain segment quality information and segment posture information of the segment to be hoisted.

[0183] In some embodiments of the present disclosure, the segment posture detection module 91 can be configured to obtain spatial point cloud data of the segment to be lifted in the segment lifting area through segment lifting space scanning; input the spatial point cloud data into a pre-trained shield segment point cloud deep learning prediction model to obtain segment point cloud data and segment type; and determine segment quality information based on the segment point cloud data.

[0184] In some embodiments of the present disclosure, the segment posture detection module 91 can be configured to control the segment lifting device to reset through the segment lifting control system when the spatial point cloud data of the segment to be lifted area is obtained through the segment lifting space scanning; to obtain the original spatial point cloud data of the segment to be lifted area through the segment lifting space scanning; and to filter the original spatial point cloud data.

[0185] In some embodiments of the present disclosure, the segment posture detection module 91 can be configured to perform cylinder fitting through the segment point cloud data to obtain the segment cylinder model parameters, wherein the segment cylinder model parameters include the segment radius and the segment axis direction vector; obtain the segment surface curvature according to the segment axis direction vector and the segment radius; search for point cloud data within the segment surface range according to the segment curvature, find the surface defect features, and filter the inherent grouting circular holes and positioning hole information of the segment, wherein the surface defect features include the surface hole features, the surface missing features, and the surface protrusion features; judge whether the segment quality meets the relevant requirements according to the size of the surface defect features, wherein the segment quality includes the damage on the segment surface and the smoothness of the segment surface.

[0186] In some embodiments of the present disclosure, the segment posture detection module 91 can be configured to input the segment point cloud data into a pre-trained shield segment positioning structure point cloud deep learning prediction model to obtain the segment positioning structure position information and the segment positioning structure type; determine the segment position information and segment placement posture information based on the segment positioning structure position information and the segment point cloud data.

[0187] In some embodiments of the present disclosure, the segment positioning structure is a segment positioning circular hole.

[0188] In some embodiments of the present disclosure, the segment posture detection module 91 can be configured to input the segment point cloud data into a pre-trained shield segment positioning structure point cloud deep learning prediction model to obtain the segment positioning structure position information, and to input the segment point cloud data into a pre-trained shield segment positioning structure point cloud deep learning prediction model to obtain the segment positioning circular hole point cloud data; perform mean processing on the segment positioning circular hole point cloud data to obtain the spatial coordinates of the center of the segment positioning circular hole.

[0189] In some embodiments of the present disclosure, the segment posture detection module 91 can be configured to determine the segment position information and the segment placement posture information based on the position information of the segment positioning structure and the segment point cloud data, and then calculate the number of stacks and stack height information of the segments in the segment lifting area based on the segment point cloud data, the segment support structure platform height information and the segment thickness information; calculate the segment posture based on the segment cylinder geometric parameter information and the segment cylinder point cloud data information, combined with the segment lifting space point cloud data, to determine whether the segment meets the lifting conditions.

[0190] In some embodiments of the present disclosure, in some embodiments of the present disclosure, the segment posture detection module 91 can be configured to determine the center space coordinates and the horizontal placement angle of the segment according to the space coordinate information of the center of the segment positioning hole and the normal vector information of the segment cylinder axis, based on the segment cylinder axis normal vector information and the segment cylinder point cloud data information, to determine whether the segment meets the lifting conditions; obtain the vertical angle between the segment and the ground according to the segment axis normal vector to determine whether the segment can be grabbed by the hoist; and determine whether there are obstacles in the segment grabbing route according to the segment cylinder point cloud data information and the segment lifting space point cloud data.

[0191] The segment lifting module 92 is configured to determine whether the segment to be lifted meets the lifting conditions based on the segment posture information and segment quality information of the segment to be lifted; if the segment to be lifted meets the lifting conditions, plan the segment lifting route, and control the segment lifting device through the segment lifting control system to complete the segment lifting according to the planned segment lifting route.

[0192] In some embodiments of the present disclosure, the segment lifting module 92 can be configured to control the segment lifting device to automatically position and grab the segment to be lifted through the segment lifting control system when the segment to be lifted meets the lifting conditions; plan the segment lifting route, and control the segment lifting device to complete the segment lifting according to the planned segment lifting route.

[0193] In some embodiments of the present disclosure, the segment lifting module 92 can be configured to control the segment lifting device to move to just above the segment to be lifted according to the position information of the segment positioning structure when the segment lifting device is controlled to automatically position and grasp the segment to be lifted, wherein the segment positioning structure is a segment positioning circular hole; control the segment lifting device to open the clamping mechanism, and control the positioning pin of the segment lifting device to align with the segment positioning structure to grab the segment to be lifted.

[0194] In some embodiments of the present disclosure, the segment lifting module 92 can be configured to determine whether the segment meets the lifting conditions by judging at least one of whether the segment unloading point is occupied, whether the segment posture can be grasped, whether the segment quality meets the requirements, and whether there are unavoidable obstacles in the lifting space.

[0195] In some embodiments of the present disclosure, the segment lifting module 92 can be configured to plan the segment lifting route according to predetermined rules based on the real-time point cloud data of the segment lifting space, the position information of the segment to be lifted, and the position information of the unloading point on the segment transport cart when planning the segment lifting route, wherein the predetermined rules include at least one of safety rules, energy-saving rules and efficiency rules.

[0196] In some embodiments of the present disclosure, the segment lifting module 92 may also be configured to display and monitor the segment lifting three-dimensional space in real time during the segment lifting process.

[0197] In some embodiments of the present disclosure, the segment lifting module 92 can be configured to acquire segment lifting route point cloud data in real time through laser radar during the segment lifting process; determine whether there is any obstacle intrusion on the segment lifting route, where obstacle intrusion includes people, objects or vehicles entering the working space; and re-plan the lifting route in the event of obstacle intrusion.

[0198] In some embodiments of the present disclosure, the industrial computer can be used to implement any of the above embodiments (for example Figure 2-Figure 8 The operation of the segment posture detection and lifting method described in any embodiment).

[0199] The industrial computer of the above-mentioned embodiment of the present invention can realize intelligent lifting of segments and detection of the posture of segments, and is committed to solving the problems of automatic lifting of segments and detection of the posture of segments in shield segment lifting. It improves the traditional operation mode of segment lifting that relies on manual remote control operation, can automatically transport segments, intelligently detect the placement posture and quality of segments, improves the work efficiency and safety protection of segment lifting, improves the work process of tunnel excavation workers, and realizes the automation and intelligence of segment lifting.

[0200] Fig.10 Schematic diagram of the structure of some other embodiments of the industrial control computer disclosed in the present invention. Fig.10 As shown, the industrial computer includes a memory 11 and a processor 12 .

[0201] The memory 11 is used to store instructions. The processor 12 is coupled to the memory 11. The processor 12 is configured to execute the above-mentioned embodiments (for example, Figure 2-Figure 8 Any embodiment) involves a method.

[0202] like Fig.10 As shown, the industrial computer also includes a communication interface 13 for information exchange with other devices. At the same time, the industrial computer also includes a bus 14, through which the processor 12, the communication interface 13, and the memory 11 communicate with each other.

[0203] The memory 11 may include a high-speed RAM memory, and may also include a non-volatile memory, such as at least one disk memory. The memory 11 may also be a memory array. The memory 11 may also be divided into blocks, and the blocks may be combined into virtual volumes according to certain rules.

[0204] In addition, the processor 12 may be a central processing unit (CPU), or may be an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present disclosure.

[0205] According to another aspect of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, any of the above embodiments (for example Figure 2-Figure 8 The segment posture detection and lifting method as described in any embodiment).

[0206] The computer-readable storage medium of the present disclosure may be implemented as a non-transitory computer-readable storage medium.

[0207] Compared with the related art, the present invention adopts an intelligent segment lifting system based on deep reinforcement learning, which solves the problem of segment lifting according to a pre-set route in the related art, and can automatically plan the driving route according to the on-site environment and automatically avoid obstacles during segment lifting to achieve automatic protection, thus fully realizing unattended automatic operation of segment lifting. The present invention adopts segment intelligent posture detection and segment positioning structure intelligent recognition technology based on deep learning of point cloud data, which solves the problem that the complex three-dimensional feature information of the segment structure cannot be obtained in the related segment recognition technology, and provides technical support for the realization of automatic segment lifting. Therefore, the present invention discloses an intelligent segment lifting system and posture detection method to realize automatic segment lifting, which can improve the efficiency of segment lifting, reduce the energy consumption of segment lifting, improve the working environment of construction workers, and provide technical reserves for intelligent segment lifting.

[0208] In actual use, the segment intelligent posture detection technology and segment intelligent lifting system disclosed in the present invention are applicable to but not limited to segment production, segment quality inspection, segment transportation, segment lifting, segment assembly and other fields involving segment application; the combination of the segment intelligent lifting device and laser radar disclosed in the present invention may be improved according to the usage scenario, including but not limited to crane combination, mechanical arm combination, mobile robot combination; the segment shape and the positioning structure on the segment may change according to the actual usage scenario. The segment recognition algorithm in the present invention can recognize segments of any geometric shape, positioning structures of any geometric features, and segment information of any geometric features.

[0209] Those skilled in the art will appreciate that the embodiments of the present disclosure may be provided as methods, devices, or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present disclosure may take the form of a computer program product implemented on one or more computer-usable non-transient storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0210] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0211] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0212] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0213] The industrial computer, segment posture detection module and segment lifting module described above can be implemented as a general-purpose processor, programmable logic controller (PLC), digital signal processor (DSP), application-specific integrated circuit (ASIC), field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component or any appropriate combination thereof for performing the functions described in this application.

[0214] Those skilled in the art will appreciate that all or part of the steps of the above-described embodiments of the present invention may be accomplished by hardware, and the hardware may be implemented as a general-purpose processor, a programmable logic controller (PLC), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, or any appropriate combination thereof for executing the methods described in the present application.

[0215] So far, the present disclosure has been described in detail. In order to avoid obscuring the concept of the present disclosure, some details known in the art are not described. Based on the above description, those skilled in the art can fully understand how to implement the technical solution disclosed here.

[0216] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware, or may be accomplished by instructing the relevant hardware through a program, and the program may be stored in a non-transitory computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0217] The description of the present disclosure is given for the purpose of illustration and description, and is not intended to be exhaustive or to limit the present disclosure to the disclosed form. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present disclosure, and to enable those of ordinary skill in the art to understand the present disclosure and thereby design various embodiments with various modifications suitable for specific uses.

Claims

1. A method for detecting and lifting a pipe segment, comprising: Obtaining the segment quality information and segment posture information of the segment to be hoisted; According to the segment posture information and segment quality information of the segment to be hoisted, it is judged whether the segment to be hoisted meets the hoisting conditions; When the pipe segment to be hoisted meets the hoisting conditions, the pipe segment hoisting route is planned, and the pipe segment hoisting control system controls the pipe segment hoisting device to complete the pipe segment hoisting according to the planned pipe segment hoisting route; Wherein, the obtaining of the segment quality information of the segment to be hoisted includes: Through segment lifting space scanning, obtain the spatial point cloud data of the segment to be lifted area; Input the spatial point cloud data into the pre-trained shield segment point cloud deep learning prediction model to obtain the segment point cloud data and segment type; Determine the segment quality information based on the segment point cloud data; Wherein, the obtaining of the segment posture information of the segment to be hoisted includes: Input the segment point cloud data into the pre-trained shield segment positioning structure point cloud deep learning prediction model to obtain the segment positioning structure position information and segment positioning structure type; The segment position information and segment placement posture information are determined based on the segment positioning structure position information and segment point cloud data.

2. The segment posture detection and lifting method according to claim 1, further comprising: During the segment lifting process, the three-dimensional space of the segment lifting is displayed and monitored in real time.

3. The method for detecting and lifting the pipe segment posture according to claim 2, wherein: The real-time display and monitoring of the three-dimensional space for segment lifting includes: During the segment lifting process, the laser radar is used to obtain the segment lifting route point cloud data in real time; Determine whether there is any obstacle intrusion on the segment lifting route, where obstacle intrusion includes people, objects or vehicles entering the working space; In the event of an obstacle invading the line, re-plan the lifting route.

4. The segment posture detection and lifting method according to any one of claims 1 to 3, wherein: The planned segment lifting route includes: Based on the real-time point cloud data of the segment lifting space, the position information of the segment to be lifted and the position information of the unloading point on the segment transport cart, the segment lifting route is planned according to predetermined rules, wherein the predetermined rules include at least one of safety rules, energy-saving rules and efficiency rules.

5. The segment posture detection and lifting method according to any one of claims 1 to 3, wherein: The determining whether the segment to be hoisted meets the hoisting conditions according to the segment posture information and the segment quality information of the segment to be hoisted includes: Whether the pipe segment to be lifted meets the lifting conditions is determined by judging at least one of whether the posture of the pipe segment can be grasped and whether the quality of the pipe segment meets the requirements.

6. The segment posture detection and lifting method according to any one of claims 1 to 3, further comprising: Whether the pipe segment to be lifted meets the lifting conditions is determined by judging at least one of whether the pipe segment unloading point is occupied and whether there are unavoidable obstacles in the lifting space.

7. The segment posture detection and lifting method according to any one of claims 1 to 3, wherein: When the pipe segment to be hoisted meets the hoisting conditions, the pipe segment posture detection and hoisting method further includes: The segment lifting control system is used to control the segment lifting device to automatically locate and grab the segment to be lifted; the segment lifting route is planned, and the segment lifting device is controlled to complete the segment lifting according to the planned segment lifting route.

8. The method for detecting and lifting the pipe segment posture according to claim 7, wherein: The control segment lifting device to automatically position and grab the segment to be lifted includes: According to the position information of the segment positioning structure, the segment lifting device is controlled to move to the position directly above the segment to be lifted, wherein the segment positioning structure is a segment positioning circular hole; The segment lifting device is controlled to open the clamping claw mechanism, and the positioning pin of the segment lifting device is controlled to be aligned with the segment positioning structure to grab the segment to be lifted.

9. The segment posture detection and lifting method according to any one of claims 1 to 3, wherein: The obtaining of spatial point cloud data of the area to be hoisted of the segment by means of segment hoisting spatial scanning comprises: Control the segment lifting device to reset through the segment lifting control system; Through segment lifting space scanning, the original space point cloud data of the segment to be lifted area is obtained; Filter the original spatial point cloud data.

10. The segment posture detection and lifting method according to any one of claims 1 to 3, wherein: The segment quality information determined based on the segment point cloud data includes: Perform cylinder fitting through the segment point cloud data to obtain the segment cylinder model parameters, wherein the segment cylinder model parameters include the segment radius and the segment axis direction vector; Obtain the segment surface curvature according to the segment axis direction vector and the segment radius; According to the curvature of the segment, search the point cloud data within the segment surface range, find the surface defect features, and filter the inherent grouting circular holes and positioning holes information of the segment. The surface defect features include the surface hole features, surface missing features and surface convex features. Whether the segment quality meets the relevant requirements is judged based on the size of the surface defect characteristics, where the segment quality includes segment surface damage and segment surface flatness.

11. The method for detecting and lifting the pipe segment posture according to any one of claims 1 to 3, wherein: The segment positioning structure is a segment positioning circular hole, and the step of inputting the segment point cloud data into a pre-trained shield segment positioning structure point cloud deep learning prediction model to obtain the segment positioning structure position information includes: Input the segment point cloud data into the pre-trained shield segment positioning structure point cloud deep learning prediction model to obtain the segment positioning circular hole point cloud data; The point cloud data of the segment positioning circular hole is averaged to obtain the spatial coordinates of the center of the segment positioning circular hole.

12. The method for detecting and lifting the pipe segment posture according to claim 11, wherein: Determining the segment position information and the segment placement posture information according to the segment positioning structure position information and the segment point cloud data includes: According to the segment point cloud data, segment support structure platform height information and segment thickness information, the number and height of the segments in the segment lifting area are calculated; According to the geometric parameter information of the segment cylinder and the point cloud data information of the segment cylinder, combined with the spatial point cloud data of the segment lifting, the posture of the segment is calculated to determine whether the segment meets the lifting conditions.

13. The method for detecting and lifting the pipe segment posture according to claim 12, wherein: The step of calculating the posture of the segment based on the normal vector information of the segment cylinder axis and the segment cylinder point cloud data information, combined with the segment lifting space point cloud data, to determine whether the segment meets the lifting conditions includes: According to the spatial coordinate information of the center of the segment positioning hole and the normal vector information of the segment cylindrical axis, the central spatial coordinates of the segment and the horizontal placement angle of the segment are determined; Obtain the vertical angle between the segment and the ground based on the segment axis normal vector to determine whether the segment can be grabbed by the hoist; Based on the point cloud data information of the segment cylinder and the point cloud data of the segment lifting space, it is determined whether there are obstacles in the segment grabbing route.

14. An industrial computer, comprising: The segment posture detection module is configured to obtain segment quality information and segment posture information of the segment to be hoisted; The segment lifting module is configured to determine whether the segment to be lifted meets the lifting conditions according to the segment posture information and the segment quality information of the segment to be lifted; When the pipe segment to be hoisted meets the hoisting conditions, the pipe segment hoisting route is planned, and the pipe segment hoisting control system controls the pipe segment hoisting device to complete the pipe segment hoisting according to the planned pipe segment hoisting route; The segment posture detection module is configured to obtain the segment quality information of the segment to be hoisted by scanning the segment hoisting space to obtain the spatial point cloud data of the segment to be hoisted area; input the spatial point cloud data into the pre-trained shield segment point cloud deep learning prediction model to obtain the segment point cloud data and the segment type; determine the segment quality information according to the segment point cloud data; Among them, the segment posture detection module is configured to input the segment point cloud data into the pre-trained shield segment positioning structure point cloud deep learning prediction model after obtaining the segment posture information of the segment to be lifted, so as to obtain the segment positioning structure position information and segment positioning structure type; determine the segment position information and segment placement posture information according to the segment positioning structure position information and segment point cloud data.

15. The industrial computer according to claim 14, wherein the industrial computer is used to execute the operations of the segment posture detection and lifting method according to any one of claims 2 to 13.

16. An industrial computer, comprising: A memory for storing instructions; The processor is used to execute the instructions so that the industrial computer performs the operations of the segment posture detection and lifting method as described in any one of claims 1 to 13.

17. A segment lifting device, comprising a laser radar and an industrial computer as described in any one of claims 14 to 16.

18. A segment posture detection and lifting system, comprising a segment lifting control system and the segment lifting device according to claim 17.

19. A computer-readable storage medium, wherein: The computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, the segment posture detection and lifting method according to any one of claims 1 to 13 is implemented.

Images