Method for automatically splicing shield segments in whole process based on stereoscopic vision of unmanned aerial vehicle
Through drone stereo vision technology, the automatic identification and installation of shield pipe segments in the tunnel is solved, and the problem of difficult to ensure construction quality and safety by manual operation is achieved, precise assembly in narrow environments is achieved, and construction safety and quality are improved.
Patent Information
- Application Number
- CN202411120979.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-19
AI Technical Summary
In the prior art, manual grasping and installing shield pipe sheets by operating the assembly machine control handle cannot ensure construction quality and construction safety. Especially when the space in the tunnel is narrow and the viewing conditions are poor, quality and safety accidents are easily caused.
Using the full process automatic assembly method of shield pipe segment based on drone stereo vision, the drone is equipped with a camera and inertial navigation system to obtain the position data of the pipe segment and calculate the target motion amount of the assembly joints to achieve the accuracy of the pipe segment. Crawl and locate installation.
In an environment with narrow space and poor viewing conditions, the precise identification and installation of pipe segments is achieved, which reduces labor intensity and safety hazards, improves construction quality and safety, and avoids the occurrence of quality and safety accidents.
Smart Images

Figure CN118979748A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of shield tunnel construction, and in particular to a full-process automatic assembly method of shield tunnel segments based on unmanned aerial vehicle stereo vision. Background Art
[0002] The shield segment is the main assembly component of the shield construction and the innermost barrier of the tunnel. It is responsible for resisting soil pressure, groundwater pressure and some special loads. The shield segment is the permanent lining structure of the shield tunnel. The assembly quality of the shield segment is directly related to the overall quality and safety of the tunnel, and affects the waterproof performance and durability of the tunnel.
[0003] Although the number of shield tunnels in my country is growing rapidly, there is a shortage of experienced segment assemblers. At present, conventional segment assembly operations involve manually grabbing and installing shield segments by operating the assembly machine handle. Due to the narrow space and poor visibility in the tunnel, the labor intensity of the assemblers is high and the safety risks are high, so tunnel quality accidents and safety accidents occur frequently. Therefore, it is necessary to provide a full-process automatic assembly method for shield segments based on drone stereo vision, which can solve the problem in the prior art that manual operation of the assembly machine control handle to grab and install shield segments cannot guarantee construction quality and construction safety. Summary of the invention
[0004] The purpose of the present invention is to provide a method for fully automatic assembly of shield segments based on drone stereo vision, which can solve the problem in the prior art that manual operation of the assembly machine control handle to grab and install shield segments cannot ensure construction quality and construction safety.
[0005] The present invention is achieved in that:
[0006] A method for fully automatic assembly of shield segments based on drone stereo vision, comprising the following steps:
[0007] Step 1: Install the translation beam of the segment assembly machine in cantilever form inside the shield machine, take the intersection of the shield machine centerline and the vertical plane of the installation fixed end of the translation beam as the origin O0, and establish the global coordinate system X0Y0Z0;
[0008] Step 2: Find three non-collinear fixed points in the internal space of the shield machine, and obtain the vectors of the three fixed points in the global coordinate system X0Y0Z0;
[0009] Step 3: Determine the starting position of the drone, i.e., the first position, and establish the first drone local coordinate system X1Y1Z1 of the drone;
[0010] Step 4: The UAV moves from the starting position to the second position P2, and establishes the second UAV local coordinate system X2Y2Z2;
[0011] Step 5: The UAV moves from the second position P2 to the third position P3, establishes the third UAV local coordinate system X3Y3Z3, captures the pose data of the segment to be assembled on the segment feeder through the camera carried by the UAV, and calculates its first pose matrix K1;
[0012] Step 6: Using the first pose matrix K1, based on the kinematic inverse solution algorithm of the assembling machine, the target motion amount of each joint of the assembling machine is calculated, so that the assembling machine can grab the pipe segment to be assembled;
[0013] Step 7: The UAV moves from the third position P3 to the fourth position P4, establishes a fourth UAV local coordinate system X4Y4Z4, captures the pose data of the target position of the segment to be assembled through the camera carried by the UAV, and calculates its second pose matrix K2;
[0014] Step 8: Using the second posture matrix K2, based on the inverse kinematics algorithm of the assembling machine, calculate the target motion of each joint of the assembling machine, so that the assembling machine can position and install the pipe segment to be assembled.
[0015] The step 2 comprises the following sub-steps:
[0016] Step 2.1: Find three non-collinear fixed points in the shield machine, which are respectively recorded as the first fixed point A, the second fixed point B and the third fixed point C;
[0017] Step 2.2: Get the spatial coordinates of the three fixed points in the global coordinate system X0Y0Z0;
[0018] Step 2.3: Calculate the vector based on the spatial coordinates of the three fixed points and
[0019] The step 3 comprises the following sub-steps:
[0020] Step 3.1: The drone has its own inertial navigation system, which includes a gyroscope and an accelerometer, and a camera is mounted on the drone;
[0021] Step 3.2: Determine the starting position P1 of the UAV, i.e., the first position, and establish the first UAV local coordinate system X1Y1Z1 with the center of the inertial navigation system as the origin O1.
[0022] The step 4 comprises the following sub-steps:
[0023] Step 4.1: Use the gyroscope to measure the heading angle, pitch angle and roll angle of the drone at the second position P2, which are α1, β1 and γ1 respectively;
[0024] Step 4.2: Use the accelerometer to measure the spatial displacement of the drone and determine its position coordinates (x f1 ,y f1 ,z f1 ), so its pose matrix in the first drone local coordinate system X1Y1Z1 is expressed as:
[0025]
[0026] Step 4.3: Use the camera to obtain the coordinates of the three fixed points corresponding to points A', B', and C' in the local coordinate system X2Y2Z2 of the second drone, and calculate and
[0027] Step 4.4: Choose a fixed point and solve the pose transformation matrix from the second drone coordinate system X2Y2Z2 to the global coordinate system X0Y0Z0 When the first fixed point A is selected, the expression is:
[0028]
[0029] in, is the Euler transformation matrix; is the translation transformation vector; Satisfaction relationship
[0030]
[0031] therefore, It can be expressed as:
[0032]
[0033] Substitute the coordinates of the first fixed point A and its corresponding point A' into Obtain
[0034] Step 4.5: The pose transformation matrix from the second drone coordinate system X2Y2Z2 to the global coordinate system X0Y0Z0 is expressed as:
[0035]
[0036] The step 5 comprises the following sub-steps:
[0037] Step 5.1: Use the gyroscope to measure the heading angle, pitch angle and roll angle of the drone at the third position P3, which are α2, β2 and γ2 respectively;
[0038] Step 5.2: Use the accelerometer to measure the spatial displacement of the drone and determine its position coordinates (x f2 ,y f2 ,zf2 ), so its pose matrix in the first drone local coordinate system X1Y1Z1 is expressed as:
[0039]
[0040] Step 5.3: Establish the matrix transformation relationship from the local coordinate system X3Y3Z3 of the third drone to the local coordinate system X2Y2Z2 of the second drone Satisfy the following formula:
[0041]
[0042] therefore,
[0043] Step 5.4: The camera on the drone is used to capture the pose data of the segment to be assembled on the segment feeder, that is, the pose matrix T3 of the segment to be assembled in the local coordinate system X3Y3Z3 of the third drone:
[0044]
[0045] Where R3 is the attitude matrix of the segment to be assembled in the local coordinate system X3Y3Z3 of the third UAV; S3 is the position matrix of the segment to be assembled in the local coordinate system X3Y3Z3 of the third UAV; α3, β3 and γ3 are the attitude angles of the segment to be assembled in the local coordinate system X3Y3Z3 of the third UAV; (x f3 ,y f3 ,z f3 ) is the position coordinate of the segment to be assembled in the local coordinate system X3Y3Z3 of the third UAV;
[0046] Step 5.5: According to the pose matrix T3 of the segment to be assembled in the local coordinate system X3Y3Z3 of the third UAV, the first pose matrix K1 of the segment to be assembled in the global coordinate system X0Y0Z0 is calculated as follows:
[0047]
[0048] The step 7 comprises the following sub-steps:
[0049] Step 7.1: Use the gyroscope to measure the heading angle, pitch angle and roll angle of the UAV at the fourth position P4, which are α4, β4 and γ4 respectively;
[0050] Step 7.2: Use the accelerometer to measure the spatial displacement of the drone and determine its position coordinates (x f4 ,y f4 ,z f4 ), so its pose matrix in the first drone local coordinate system X1Y1Z1 is expressed as:
[0051]
[0052] Step 7.3: Establish the matrix transformation relationship from the fourth drone local coordinate system X4Y4Z4 to the second drone local coordinate system X2Y2Z2 Satisfy the following formula:
[0053]
[0054] therefore,
[0055] Step 7.4: The camera carried by UAV 4 captures the pose matrix T5 of the target position of the segment to be assembled in the local coordinate system X4Y4Z4 of the fourth UAV:
[0056]
[0057] Where R5 is the attitude matrix of the target position of the segment to be assembled in the fourth UAV local coordinate system X4Y4Z4; S5 is the position matrix of the target position of the segment to be assembled in the fourth UAV local coordinate system X4Y4Z4; α5, β5 and γ5 are the attitude angles of the target position of the segment to be assembled in the fourth UAV local coordinate system X4Y4Z4; (x f5 ,y f5 ,z f5 ) is the position coordinate of the target position of the segment to be assembled in the local coordinate system X4Y4Z4 of the fourth UAV;
[0058] Step 7.5: Calculate the second pose matrix K2 of the target position of the segment to be assembled in the global coordinate system X0Y0Z0. The calculation formula is:
[0059]
[0060] Compared with the prior art, the present invention has the following beneficial effects:
[0061] The present invention adopts a drone equipped with a camera, which can obtain the position of the pipe segment to be assembled on the feeding machine and the target position required for installation. It can be used in an environment with narrow space and poor visibility conditions, and can realize the recognition of key posture information of the pipe segment to be assembled, and calculate and convert it into the movement amount of each joint of the assembler, thereby ensuring the controllability and accuracy of the movement of each joint of the assembler, thereby realizing the precise grasping and positioning installation of the pipe segment to be assembled by the assembler, solving the problem that traditional tunnel construction must rely on manual operation of the assembler handle to complete the grasping and installation of the pipe segment to be assembled and the problem that it is easy to cause quality and safety accidents. BRIEF DESCRIPTION OF THE DRAWINGS
[0062] Figure 1It is a schematic diagram of steps 1 to 4 in the full-process automatic assembly method of shield segments based on drone stereo vision of the present invention;
[0063] Figure 2 It is a schematic diagram of step 5 and step 6 in the full-process automatic assembly method of shield segments based on drone stereo vision of the present invention;
[0064] Figure 3 It is a schematic diagram of steps 7 and 8 in the full-process automatic assembly method of shield segments based on drone stereo vision of the present invention.
[0065] In the figure, 1 is the translation beam, 2 is the shield machine, 3 is the slewing ring, 4 is the drone, 5 is the segment feeder, and 6 is the segments to be assembled. DETAILED DESCRIPTION
[0066] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.
[0067] A method for fully automatic assembly of shield segments based on drone stereo vision, comprising the following steps:
[0068] Please see attached Figure 1 , Step 1: Cantilever the translation beam 1 of the segment assembly machine inside the shield machine 2, and take the intersection of the central axis of the shield machine 2 and the vertical plane of the installation fixed end of the translation beam 1 as the origin O0, and establish the global coordinate system X0Y0Z0.
[0069] The slewing ring 3 of the assembling machine is movably mounted on the translation beam 1. After the pipe segment 6 to be assembled is sucked by the suction cup, the slewing ring 3 can rotate around the translation beam 1, so that the pipe segment 6 to be assembled can be rotated to the vicinity of the target position in the space.
[0070] Please see attached Figure 1 , Step 2: Find three non-collinear fixed points in the internal space of the shield machine 2, and obtain the vectors of the three fixed points in the global coordinate system X0Y0Z0.
[0071] The step 2 comprises the following sub-steps:
[0072] Step 2.1: Find three non-collinear fixed points in the shield machine 2, which are respectively recorded as the first fixed point A, the second fixed point B and the third fixed point C.
[0073] Step 2.2: Get the spatial coordinates of the three fixed points in the global coordinate system X0Y0Z0.
[0074] Step 2.3: Calculate the vector based on the spatial coordinates of the three fixed points and
[0075] Please see attached Figure 1, Step 3: Determine the starting position of the UAV 4, that is, the first position, and establish the first UAV local coordinate system X1Y1Z1 of the UAV 4.
[0076] The drone 4 is small in size and can be used in tunnels with narrow space and poor visibility, reducing reliance on human operation, thereby ensuring safe and efficient construction while reducing the occurrence of safety accidents.
[0077] The step 3 comprises the following sub-steps:
[0078] Step 3.1: The drone 4 is equipped with an inertial navigation system, which includes a gyroscope and an accelerometer, and a camera is mounted on the drone 4.
[0079] The camera may be a stereoscopic vision sensor of the prior art, which can be used to capture the position image of the segment 6 to be assembled on the segment feeder 5 and the target position image of the segment 6 to be assembled, so as to determine the position information of the segment 6 to be assembled by extracting the target feature on the image signal, i.e., the segment 6 to be assembled. The target feature extraction of the image signal is a conventional operation method of image processing, which will not be described in detail here.
[0080] Cameras can be used to replace human eyes for measurement and judgment. Compared with human eye observation, they have higher accuracy and safety, greatly reducing the labor intensity of construction workers, safety hazards and accidents during the construction process.
[0081] Step 3.2: Determine the starting position P1 of the UAV 4, that is, the first position, and establish a first UAV local coordinate system X1Y1Z1 with the center of the inertial navigation system as the origin O1.
[0082] The starting position P1 can be adaptively selected according to actual construction needs, such as the shield machine propulsion control operation room.
[0083] The gyroscope is used to monitor the rotation angles of the drone 4 around its own axes (including the heading angle α, the pitch angle β and the roll angle γ), and the accelerometer is used to calculate the spatial displacement of the drone 4 to calculate the three-dimensional coordinates of the drone 4 in the first drone local coordinate system X1Y1Z1. The gyroscope and accelerometer are commonly used posture measurement devices in this field, and their working principles and working processes are not described here.
[0084] Please see attached Figure 1 , Step 4: UAV 4 moves from the starting position to the second position P2, and establishes a second UAV local coordinate system X2Y2Z2.
[0085] The second position P2 can be adaptively selected according to the positions of the three fixed points to ensure that when the drone 4 is located at the second position P2, the image capture range of the camera can cover the positions of the three fixed points.
[0086] The step 4 comprises the following sub-steps:
[0087] Step 4.1: Use the gyroscope to measure the heading angle, pitch angle and roll angle of the drone 4 at the second position P2, which are α1, β1 and γ1 respectively.
[0088] Step 4.2: Use the accelerometer to measure the spatial displacement of the drone 4 and determine its position coordinates (x f1 ,y f1 ,z f1 ), that is, the coordinates of the drone 4 in the first drone local coordinate system X1Y1Z1 when the drone 4 is at the second position P2, so its pose matrix in the first drone local coordinate system X1Y1Z1 is expressed as:
[0089]
[0090] The first UAV local coordinate system X1Y1Z1 is used to provide the posture information of the UAV 4 at different positions. The conversion of the second UAV local coordinate system X2Y2Z2 to the global coordinate system can be established through the three fixed points A, B, and C.
[0091] Step 4.3: Use the camera to obtain the coordinates of the three fixed points corresponding to points A', B', and C' in the local coordinate system X2Y2Z2 of the second drone, and calculate and
[0092] Step 4.4: Select one of the three fixed points A, B, and C to solve the pose transformation matrix from the second drone coordinate system X2Y2Z2 to the global coordinate system X0Y0Z0 Taking the selection of the first fixed point A as an example, it can be expressed as:
[0093]
[0094] in, is the Euler transformation matrix; is the translation transformation vector; Satisfaction relationship
[0095]
[0096] therefore, It can be expressed as:
[0097]
[0098] Substitute the coordinates of the first fixed point A and its corresponding point A' into You can get
[0099] Step 4.5: The pose transformation matrix from the second drone coordinate system X2Y2Z2 to the global coordinate system X0Y0Z0 is expressed as:
[0100]
[0101] Please see attached Figure 2 , Step 5: The UAV 4 moves from the second position P2 to the third position P3, establishes the third UAV local coordinate system X3Y3Z3, captures the pose data of the pipe segment 6 to be assembled on the film feeder 5 through the camera carried by the UAV 4, and calculates its first pose matrix K1.
[0102] The third position P3 is selected so that the camera carried by the drone 4 can clearly and completely capture the posture data of the pipe segment 6 to be assembled on the film feeder 5. The third position P3 can be adaptively selected according to the position of the film feeder 5 and the placement position of the pipe segment 6 to be assembled on the film feeder 5.
[0103] The step 5 comprises the following sub-steps:
[0104] Step 5.1: Use the gyroscope to measure the heading angle, pitch angle and roll angle of the drone 4 at the third position P3, which are α2, β2 and γ2 respectively.
[0105] Step 5.2: Use the accelerometer to measure the spatial displacement of the drone 4 and determine its position coordinates (x f2 ,y f2 ,z f2 ), that is, the coordinates of the drone 4 in the first drone local coordinate system X1Y1Z1 when the drone 4 is at the third position P3, so its pose matrix in the first drone local coordinate system X1Y1Z1 is expressed as:
[0106]
[0107] The transformation of the local coordinate system X3Y3Z3 of the third UAV to the global coordinate system can be established through the three fixed points A, B, and C.
[0108] Step 5.3: Establish the matrix transformation relationship from the local coordinate system X3Y3Z3 of the third drone to the local coordinate system X2Y2Z2 of the second drone Satisfy the following formula:
[0109]
[0110] therefore,
[0111] Step 5.4: The position and posture data of the segment 6 to be assembled on the segment feeder 5 are captured by the camera carried by the drone 4, that is, the position and posture matrix T3 of the segment 6 to be assembled in the local coordinate system X3Y3Z3 of the third drone:
[0112]
[0113] Wherein, R3 is the attitude matrix of the segment 6 to be assembled in the local coordinate system X3Y3Z3 of the third UAV; S3 is the position matrix of the segment 6 to be assembled in the local coordinate system X3Y3Z3 of the third UAV; α3, β3 and γ3 are the attitude angles (i.e., heading angle, pitch angle and roll angle) of the segment 6 to be assembled in the local coordinate system X3Y3Z3 of the third UAV; is the position coordinate of the segment 6 to be assembled in the local coordinate system X3Y3Z3 of the third UAV;
[0114] Step 5.5: According to the pose matrix T3 of the segment 6 to be assembled in the local coordinate system X3Y3Z3 of the third UAV, the first pose matrix K1 of the segment 6 to be assembled in the global coordinate system X0Y0Z0 is calculated as follows:
[0115]
[0116] Step 6: Using the first pose matrix K1, based on the inverse solution algorithm of the assembling machine's kinematics, calculate the target motion amount of each joint of the assembling machine, so that the assembling machine can grab the pipe segment 6 to be assembled.
[0117] The inverse solution algorithm of the kinematics of the assembler is a conventional technical means in this field, such as the inverse solution algorithm of the kinematics of the assembler disclosed in invention patent applications 202311182397.6, 202310791597.5, 202311389209.7, and 202311385602.9, which will not be repeated here.
[0118] In step 6, after calculating the target movement amount of each joint of the assembling machine, the target movement amount of each joint is sent to the control system of the assembling machine, and the control system controls the corresponding joints of the assembling machine to perform corresponding movements, thereby enabling the assembling machine to accurately grasp the pipe segment 6 to be assembled.
[0119] Step 7: The UAV 4 moves from the third position P3 to the fourth position P4, establishes a fourth UAV local coordinate system X4Y4Z4, captures the pose data of the target position of the segment 6 to be assembled through the camera carried by the UAV 4, and calculates its second pose matrix K2.
[0120] The fourth position P4 is preferably such that the target position of the to-be-assembled pipe segment 6 can be clearly and completely seen, and the fourth position P4 can be selected according to the adaptability of the position where the to-be-assembled pipe segment 6 is required to be assembled.
[0121] The step 7 comprises the following sub-steps:
[0122] Step 7.1: Use the gyroscope to measure the heading angle, pitch angle and roll angle of the drone 4 at the fourth position P4, which are α4, β4 and γ4 respectively.
[0123] Step 7.2: Use the accelerometer to measure the spatial displacement of the drone 4 and determine its position coordinates (x f4 ,y f4 ,z f4 ), that is, the coordinates of the drone 4 in the first drone local coordinate system X1Y1Z1 when the drone 4 is at the fourth position P4, so its pose matrix in the first drone local coordinate system X1Y1Z1 is expressed as:
[0124]
[0125] The transformation of the local coordinate system X4Y4Z4 of the fourth UAV to the global coordinate system can be established through the three fixed points A, B, and C.
[0126] Step 7.3: Establish the matrix transformation relationship from the fourth drone local coordinate system X4Y4Z4 to the second drone local coordinate system X2Y2Z2 Satisfy the following formula:
[0127]
[0128] therefore,
[0129] Step 7.4: The camera carried by the drone 4 captures the pose matrix T5 of the target position of the segment 6 to be assembled in the fourth drone local coordinate system X4Y4Z4:
[0130]
[0131] Wherein, R5 is the attitude matrix of the target position of the segment 6 to be assembled in the fourth UAV local coordinate system X4Y4Z4; S5 is the position matrix of the target position of the segment 6 to be assembled in the fourth UAV local coordinate system X4Y4Z4; α5, β5 and γ5 are the attitude angles of the target position of the segment 6 to be assembled in the fourth UAV local coordinate system X4Y4Z4; (x f5 ,y f5 ,z f5 ) is the position coordinate of the target position of the segment 6 to be assembled in the fourth UAV local coordinate system X4Y4Z4;
[0132] Step 7.5: Calculate the second pose matrix K2 of the target position of the segment 6 to be assembled in the global coordinate system X0Y0Z0, and the calculation formula is:
[0133]
[0134] Step 8: Using the second posture matrix K2, based on the inverse solution algorithm of the assembling machine's kinematics, the target motion amount of each joint of the assembling machine is calculated, so that the assembling machine can position and install the pipe segment 6 to be assembled to the target position.
[0135] In step 8, after calculating the target movement amount of each joint of the assembling machine, the target movement amount of each joint is sent to the control system of the assembling machine, and the control system controls the corresponding joints of the assembling machine to perform corresponding movements, so that the assembling machine can accurately position and install the pipe segment 6 to be assembled to the target position.
[0136] Similar to step 6, the inverse solution algorithm of the kinematics of the assembler is a conventional technical means in this field, such as the inverse solution algorithm of the kinematics of the assembler disclosed in invention patent applications 202311182397.6, 202310791597.5, 202311389209.7, and 202311385602.9, which will not be repeated here.
[0137] The above are only preferred embodiments of the present invention and are not intended to limit the scope of protection of the invention. Therefore, any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. A method for fully automatic assembly of shield segments based on drone stereo vision, characterized by: The following steps are involved: Step 1: The translation beam (1) of the segment assembly machine is cantilevered and installed inside the shield machine (2), and the intersection of the central axis of the shield machine (2) and the vertical plane of the installation fixed end of the translation beam (1) is taken as the origin O0 to establish a global coordinate system X0Y0Z0; Step 2: Find three non-collinear fixed points in the internal space of the shield machine (2), and obtain the vectors of the three fixed points in the global coordinate system X0Y0Z0; Step 3: Determine the starting position of the drone (4), i.e., the first position, and establish a first drone local coordinate system X1Y1Z1 of the drone (4); Step 4: The drone (4) moves from the starting position to the second position P2, and establishes a second drone local coordinate system X2Y2Z2; Step 5: The drone (4) moves from the second position P2 to the third position P3, establishes a third drone local coordinate system X3Y3Z3, captures the position and posture data of the segment (6) to be assembled on the segment feeder (5) through the camera carried by the drone (4), and calculates its first position matrix K1; Step 6: Using the first pose matrix K1, based on the kinematic inverse algorithm of the assembling machine, the target motion amount of each joint of the assembling machine is calculated, so that the assembling machine can grasp the pipe segment (6) to be assembled; Step 7: The drone (4) moves from the third position P3 to the fourth position P4, establishes a fourth drone local coordinate system X4Y4Z4, captures the pose data of the target position of the segment (6) to be assembled by a camera carried by the drone (4), and calculates its second pose matrix K2; Step 8: Using the second posture matrix K2, based on the inverse solution algorithm of the assembly machine kinematics, the target motion amount of each joint of the assembly machine is calculated, so that the assembly machine can position and install the pipe segment (6) to be assembled.
2. The method for fully automatic assembly of shield segments based on drone stereoscopic vision according to claim 1 is characterized by: The step 2 comprises the following sub-steps: Step 2.1: Find three non-collinear fixed points in the shield machine (2), which are respectively recorded as the first fixed point A, the second fixed point B and the third fixed point C; Step 2.2: Get the spatial coordinates of the three fixed points in the global coordinate system X0Y0Z0; Step 2.3: Calculate the vector based on the spatial coordinates of the three fixed points and 3. The method for fully automatic assembly of shield segments based on drone stereoscopic vision according to claim 1 is characterized by: The step 3 comprises the following sub-steps: Step 3.1: The drone (4) is equipped with an inertial navigation system, which includes a gyroscope and an accelerometer, and a camera is mounted on the drone (4); Step 3.2: Determine the starting position P1 of the UAV (4), i.e., the first position, and establish the first UAV local coordinate system X1Y1Z1 with the center of the inertial navigation system as the origin O1.
4. The method for fully automatic assembly of shield segments based on drone stereo vision according to claim 1 is characterized by: The step 4 comprises the following sub-steps: Step 4.1: Using a gyroscope to measure the heading angle, pitch angle and roll angle of the drone (4) at the second position P2, which are α1, β1 and γ1 respectively; Step 4.2: Use the accelerometer to measure the spatial displacement of the drone (4) and determine its position coordinates (x f1 ,y f1 ,z f1 ), so its pose matrix in the first drone local coordinate system X1Y1Z1 is expressed as: Step 4.3: Use the camera to obtain the coordinates of the three fixed points corresponding to points A', B', and C' in the local coordinate system X2Y2Z2 of the second drone, and calculate and Step 4.4: Choose a fixed point and solve the pose transformation matrix from the second drone coordinate system X2Y2Z2 to the global coordinate system X0Y0Z0 When the first fixed point A is selected, the expression is: in, is the Euler transformation matrix; is the translation transformation vector; Satisfaction relationship therefore, It can be expressed as: Substitute the coordinates of the first fixed point A and its corresponding point A' into Obtain Step 4.5: The pose transformation matrix from the second drone coordinate system X2Y2Z2 to the global coordinate system X0Y0Z0 is expressed as:
5. The method for fully automatic assembly of shield segments based on drone stereo vision according to claim 1 is characterized by: The step 5 comprises the following sub-steps: Step 5.1: Using a gyroscope to measure the heading angle, pitch angle and roll angle of the drone (4) at the third position P3, which are α2, β2 and γ2 respectively; Step 5.2: Use the accelerometer to measure the spatial displacement of the drone (4) and determine its position coordinates as (x f2 ,y f2 ,z f2 ), so its pose matrix in the first drone local coordinate system X1Y1Z1 is expressed as: Step 5.3: Establish the matrix transformation relationship from the local coordinate system X3Y3Z3 of the third drone to the local coordinate system X2Y2Z2 of the second drone Satisfy the following formula: therefore, Step 5.4: The position and posture data of the segment to be assembled (6) on the segment feeder (5) are captured by the camera carried by the drone (4), that is, the position and posture matrix T3 of the segment to be assembled (6) in the local coordinate system X3Y3Z3 of the third drone: Wherein, R3 is the attitude matrix of the segment to be assembled (6) in the local coordinate system X3Y3Z3 of the third UAV; S3 is the position matrix of the segment to be assembled (6) in the local coordinate system X3Y3Z3 of the third UAV; α3, β3 and γ3 are the attitude angles of the segment to be assembled (6) in the local coordinate system X3Y3Z3 of the third UAV; (x f3 ,y f3 ,z f3 ) is the position coordinate of the segment to be assembled (6) in the local coordinate system X3Y3Z3 of the third unmanned aerial vehicle; Step 5.5: According to the pose matrix T3 of the segment to be assembled (6) in the local coordinate system X3Y3Z3 of the third UAV, the first pose matrix K1 of the segment to be assembled (6) in the global coordinate system X0Y0Z0 is calculated as follows:
6. The method for fully automatic assembly of shield segments based on drone stereoscopic vision according to claim 1 is characterized by: The step 7 comprises the following sub-steps: Step 7.1: Using a gyroscope to measure the heading angle, pitch angle and roll angle of the drone (4) at the fourth position P4, which are α4, β4 and γ4 respectively; Step 7.2: Use the accelerometer to measure the spatial displacement of the drone (4) and determine its position coordinates (x f4 ,y f4 ,z f4 ), so its pose matrix in the first drone local coordinate system X1Y1Z1 is expressed as: Step 7.3: Establish the matrix transformation relationship from the fourth drone local coordinate system X4Y4Z4 to the second drone local coordinate system X2Y2Z2 Satisfy the following formula: therefore, Step 7.4: The camera carried by the drone 4 captures the pose matrix T5 of the target position of the segment to be assembled (6) in the fourth drone local coordinate system X4Y4Z4: Wherein, R5 is the attitude matrix of the target position of the segment to be assembled (6) in the fourth UAV local coordinate system X4Y4Z4; S5 is the position matrix of the target position of the segment to be assembled (6) in the fourth UAV local coordinate system X4Y4Z4; α5, β5 and γ5 are the attitude angles of the target position of the segment to be assembled (6) in the fourth UAV local coordinate system X4Y4Z4; (x f5 ,y f5 ,z f5 ) are the position coordinates of the target position of the segment (6) to be assembled in the fourth unmanned aerial vehicle local coordinate system X4Y4Z4; Step 7.5: Calculate the second pose matrix K2 of the target position of the segment to be assembled (6) in the global coordinate system X0Y0Z0, and the calculation formula is:
