Method for controlling multi-robot collaborative intelligent mobile cabin drilling and riveting system

Abstract

A method for controlling a multi-robot collaborative intelligent mobile cabin drilling and riveting system. The method comprises: generating a robot scanning trajectory according to key feature parameters of a mobile cabin; performing laser scanning on an angle aluminum line, so as to capture drilling positions of the mobile cabin; establishing a task allocation mechanism, establishing a drilling and riveting task tendency model according to the task allocation mechanism, and performing multi-robot automatic drilling and riveting task allocation on the basis of the drilling and riveting task tendency model; and establishing a drilling angle compensation value model, and an automatic drilling and riveting robot performing position compensation on the basis of the drilling angle compensation value model, thereby completing drilling and riveting. In the control method, automatic drilling and riveting of angle aluminum of a mobile cabin are collaboratively performed by means of ordered work division by a plurality of automatic drilling and riveting robots, thereby achieving the high efficiency, high quality and high degree of intelligence of an angle aluminum mounting operation of the mobile cabin.

Description

A control method for a multi-robot collaborative intelligent drilling and riveting system for a shelter Technical Field

[0001] The present invention relates to the field of automated production of shelters, and in particular to a control method for a multi-robot collaborative intelligent drilling and riveting system for shelters. Background Art

[0002] As the main highly mobile carrier for military equipment such as command and control systems or civilian facilities such as medical facilities, the demand for modular hospitals in the military and civilian markets is constantly increasing. Given the application scenarios of modular hospital products, they generally have the characteristics of high degree of customization, multiple varieties and small batches.

[0003] When assembling the front cabin angle aluminum to the main panel frame, the drilling and riveting process for installing the angle aluminum requires multiple steps, including drilling the bottom hole, applying glue, and riveting. Due to the lack of intelligent equipment, this process relies primarily on manual labor or automated machines. This labor-intensive process requires manual switching between drilling, gluing, and riveting tools, often requiring the coordination of multiple personnel. This results in low efficiency and significant waste of human resources. As a typical labor-intensive industry, with rising labor costs, shelter manufacturers are currently facing a widespread shortage of workers and difficulties in finding new hires.

[0004] Existing automated specialized machines, however, lack the necessary intelligent sensing, path planning, and task allocation algorithms. Currently, manual programming is required to obtain the coordinates of drilling and riveting points through teach-in programming. This arduous programming task can even exceed the machine's actual operating time, resulting in limited improvements in production efficiency. Furthermore, since the modular cabin assembly process often generates certain dimensional and angular errors, the end effectors of existing automated process equipment struggle to adapt to these errors, further reducing product quality and consistency. Furthermore, due to inherent rigidity flaws in automated process equipment such as automatic drilling and riveting robots, the axial forces generated during the drilling process can easily cause deformation of the process equipment, leading to deviations from the target point. Existing automated functional process equipment struggles to adapt and correct these deviations, resulting in hole position errors that often fail to meet riveting requirements, leading to interruptions in the automated riveting process and damage to the equipment or the cabin itself, further reducing product quality and production efficiency. Furthermore, modular cabins are typically over 6 meters long, making it difficult for a single piece of automated equipment to fully cover them.

[0005] Therefore, the above problems need to be solved urgently.

[0006] Summary of the Invention

[0007] Purpose of the invention: The purpose of the present invention is to provide a control method for a multi-robot collaborative intelligent drilling and riveting system for a square cabin with high efficiency.

[0008] Technical solution: To achieve the above objectives, the present invention discloses a control method for a multi-robot collaborative shelter intelligent drilling and riveting system, comprising the following steps:

[0009] (1) Generate the robot scanning trajectory based on the key characteristic parameters of the cabin;

[0010] (2) Laser scanning of the diagonal aluminum wires to capture the drilling locations of the shelters;

[0011] (3) Establish a task allocation mechanism, establish a drilling and riveting task tendency model based on the task allocation mechanism, and perform multi-robot automatic drilling and riveting task allocation based on the drilling and riveting task tendency model;

[0012] (4) A drilling angle compensation value model is established, and the automatic drilling and riveting robot performs position compensation based on the drilling angle compensation value model to complete drilling and riveting.

[0013] The specific method for generating the robot scanning trajectory according to the key characteristic parameters of the cabin in step (1) includes the following steps:

[0014] The key characteristic parameters of the shelter include the length l, the width w1, the projection width w2 of the chamfered part of the shelter top on the horizontal plane, the height h, and the chamfer angle θ. If the shelter is a right-angled shelter, w2 = 0.

[0015] Set the reference point of the cabin on the robot's ground track as the coordinate origin, and establish the coordinate axis through the right-hand system; store the cabin endpoints in two groups of arrays according to the right-angle cabin and the chamfered cabin respectively;

[0016] With d _scan Indicates the scanning safety distance. For the 1#, 2#, 3#, and 4# cabin automatic drilling and riveting robots, the corresponding scanning action safety points of each automatic drilling and riveting robot are numbered O1, O2, O3, and O4, and are parameterized as follows: O1 = (-d _scan ,d _scan ,h+w2*sinθ+d _scan ) O2=(w1+d _scan ,d _scan ,h+w2*sinθ+d _scan ) O3=(-d _scan ,l+d _scan ,h+w2*sinθ+d _scan ) O4=(w+d _scan ,l+d _scan ,h+w2*sinθ+d _scan )

[0017] The parametric storage of angle aluminum uses an array with a capacity of 8 to store the 8 faces of the chamfered square cabin, and an array with a capacity of 6 to store the 6 faces of the rectangular square cabin. The chamfered square cabin and the rectangular square cabin can be represented in the order from front to back, from top to bottom, and from left to right as follows:

[0018] FACE1: [face1, face2, face3, face4, face5, face6, face7, face8], FACE2: [face1, face2, face3, face4, face5, face6], each face of the cabin has 4 aluminum corners, which are marked as a set of arrays JL: [jl1, jl2, jl3, jl4] in the clockwise direction. All the aluminum corners of a chamfered cabin are represented as:

[0019] All the aluminum corner edges of a rectangular cabin are expressed as:

[0020] That is, jl1face1, jl1face2... are used to store all the aluminum corner edges of the cabin to be polished. Any aluminum corner edge is represented by jlifacek, where the values ​​of i and k are

[0021] Determine whether the input cabin key feature parameter w2 is 0. If w2=0, it means the cabin is a rectangular cabin with 6 faces, and the maximum value of k is 6; if w2≠0, it means the cabin has 8 faces, and the maximum value of k is 8;

[0022] The robot's scanning vector can be obtained by the coordinate difference of the endpoints of the cabin. The robot's scanning trajectory is generated by the robot driving the camera coordinate system along the scanning vector parallel to the corner aluminum to be scanned, starting from the scanning action safety point and traversing all the corner aluminum edges in sequence. The corner aluminum scanning trajectory corresponding to each robot is stored in a structure array with a capacity equal to the maximum number of corner aluminum edges. The maximum number of corner aluminum edges for a chamfered cabin is 32, and the maximum number of corner aluminum edges for a right-angled cabin is 24. The style of each structure element in the array is as follows:

[0023] Preferably, the endpoints of the cabins are stored in two arrays according to the rectangular cabins and the chamfered cabins, respectively. The rectangular cabins are represented by an array with a capacity of 8 as follows:

[0024] The chamfered cabin is represented as an array with a capacity of 12:

[0025] First, determine whether the input cabin key characteristic parameter w2 is 0. If it is 0, call the rectangular cabin array; if not, call the chamfered cabin.

[0026] Furthermore, the laser scanning of the diagonal aluminum wire in step (2) to capture the drilling position of the cabin specifically includes the following steps:

[0027] (2.1) Select n fixed points P within the reach of the N# (N∈[1,4]) automatic drilling and riveting robot, and obtain the corresponding coordinates P_robot and P_camera of the fixed point P in the camera coordinate system and the robot coordinate system;

[0028] (2.2) Through P_robot = T * P_camera, we get the coordinate transformation matrix T = P_robot * (P_camera) -1 ;

[0029] (2.3) The automatic drilling and riveting robot sends a string to the 3D vision camera, which parses the string and obtains the number of the automatic drilling and riveting robot;

[0030] (2.4) The 3D vision camera activates the line laser, and the automatic drilling and riveting robot scans the track, scanning the area where the angle aluminum is installed on the edge of the cabin;

[0031] (2.5) During the 3D vision camera scanning process, the angle aluminum and the cabin aluminum skin are distinguished based on the height difference of the reflected laser light, and a point cloud image of the angle aluminum is generated;

[0032] (2.6) Extract the boundary points and coordinates of the aluminum corner point cloud image and remove the noise in the point cloud image;

[0033] (2.7) Connect the extracted boundary points in clockwise order to obtain the aluminum angle contour image;

[0034] (2.8) The contour image of the aluminum angle edge is offset in the width direction by the set minimum distance Δ1 between the rivet and the edge of the aluminum angle edge in the length direction, and in the length direction by the minimum distance Δ2 between the rivet and the edge of the aluminum angle edge in the width direction;

[0035] (2.9) Calculate the intersection of the two offset lines as the starting point for drilling the angle aluminum hole;

[0036] (2.10) Create a stack for storing the drilling positions of angle aluminum;

[0037] (2.11) The coordinate transformation matrix T is used to convert the drilling starting point position in the camera coordinate system into the coordinate position of the automatic drilling and riveting robot and store it at the bottom of the stack;

[0038] (2.12) Determine whether the remaining angle aluminum length is less than the maximum distance of the drill bit 3*D. If so, go to (2.13); if not, go to (2.14);

[0039] (2.13) Based on the maximum distance 3*D of the drill bit, offset along the length direction to obtain the drilling position of the angle aluminum drill bit and convert it into the coordinates of the automatic drilling and riveting robot;

[0040] (2.14) Mark the current point processing method as row drilling and save the current point into the stack;

[0041] (2.15) Determine whether the set angle aluminum boundary distance has been reached. If not, go to (2.16). If yes, jump to (2.17).

[0042] (2.16) Using the minimum hole spacing D of the drilling row as an offset, the drilling starting point position in the camera coordinate system is converted into the coordinate position of the automatic drilling and riveting robot;

[0043] (2.17) Mark the current point processing method as single drilling, save the current point position into the stack, and jump to (2.15);

[0044] (2.18) The coordinates of each point in the stack are sent to the automatic drilling and riveting robot in a first-in-last-out order to be used as the location points for punching, thus ending the scanning process.

[0045] Furthermore, the task allocation mechanism in step (3) includes that each automatic drilling and riveting robot first processes the cabin angle aluminum that needs the most drilling and riveting, that is, first selects the angle aluminum with the largest number of holes to be drilled; each automatic drilling and riveting robot avoids interference and waiting with other robots, that is, first selects the path with the smallest interference area with other robots; each automatic drilling and riveting robot avoids cabin angle aluminum with high movement complexity, that is, first selects the angle aluminum with the least total joint movement and the shortest total movement time; each angle aluminum is completed by at most two automatic drilling and riveting robots.

[0046] Furthermore, the drilling and riveting task tendency model in step (3) is: F K =α1M t -α2T t -α3S t ,

[0047] Among them, α1 is the weight coefficient of the work task and its value range is [1,2]. Among them, M t is the number of holes to be drilled for the t-th angle aluminum, and is the length of the angle aluminum L t The ratio of the number of angle aluminum edges to the hole spacing dΔ is rounded up, and max is the total number of angle aluminum edges, where the maximum number of chamfered square cabins is 32, and the maximum number of right-angled square cabins is 24;

[0048] α2 is the security weight system value range is [0.5,1], T t The length R of the overlapped area between the current angle aluminum and other robots’ reachability t With the current angle aluminum length L t The ratio of T t =R t / L t ;

[0049] α3 is the efficiency weight coefficient, and its value range is [0.2, 0.5]. Among them A t The total time required for the current automatic drilling and riveting robot to complete the current angle aluminum operation: A t =A safe1_t +A sm_t +A safe2_t +A hd_t +A safe3_t +A dk_t +A hd2_t +A safe4_t +A tm_t +A wait_t ,

[0050] Among them A safe1_t A is the time it takes for the robot to move from the safety point to the scanning starting point. sm_t is the scanning time, A safe2_t A is the time from the scanning end point to the safe point, hd_t is the tool change time, A safe3_t A is the time from tool change point to safety point, dk_t is the punching time, A hd2_t For the time of changing glue rivet tool, A safe4_t A is the time from the tool change point to the safety point of the glue riveting module. tm_t A is the execution time of glue riveting, wait_t is the waiting time, and total is the total number of aluminum angle edges that the robot can currently determine to be reachable.

[0051] Preferably, the multi-robot automatic drilling and riveting task allocation in step (3) specifically includes the following steps:

[0052] (3.1) Initialization, set the automatic drilling and riveting robot serial number N = 1, set the number of times all angle aluminum to be drilled and riveted M t =1;

[0053] (3.2) Obtain the projection width w2 of the chamfered portion of the cabin roof on the horizontal plane;

[0054] (3.3) Check if w2 is 0. If so, jump to (3.4). If not, jump to (3.5).

[0055] (3.4) Set the maximum number of angle aluminum to 24 and the maximum number of planes to 6.

[0056] (3.5) Set the maximum number of angle aluminum to 32 and the maximum number of planes to 8.

[0057] (3.6) Check if N > 4. If not, jump to (3.7). If so, jump to (3.26).

[0058] (3.7) Set k = 1;

[0059] (3.8) Check if k > max_p. If not, jump to (3.9). If yes, jump to (3.23).

[0060] (3.9) Set i = 1

[0061] (3.10) Check if i>4? If not, jump to (3.11); if yes, jump to (3.22);

[0062] (3.11) Set t = (k-1)*max_p+i;

[0063] (3.12) Determine the accessibility of the N# automatic drilling and riveting robot at the t-th aluminum corner edge. If it is reachable, go to (3.13); if not, jump to (3.21).

[0064] (3.13) Determine the number of drilling and rivet marks on the current aluminum corner edge M t =2? If yes, jump to (3.14), if not, jump to (3.21);

[0065] (3.14) Substitute the current angle aluminum side length and punching spacing setting values ​​into INT(L t / dΔ);

[0066] (3.15) Calculate the current working task weight value of the angle aluminum edge

[0067] (3.16) Calculate the overlap length L of the reachable area between the current automatic drilling and riveting robot and the other automatic drilling and riveting robots during the operation of the current angle aluminum edge. t ;

[0068] (3.17) Calculate the current angle aluminum safety weight coefficient T t =R t / L t ;

[0069] (3.18) Calculate and estimate the operating time of the current automatic drilling and riveting robot on the current aluminum corner edge: A t =A safe1_t +A sm_t +A safe2_t +A hd_t +A safe3_t +A dk_t +A hd2_t +A safe4_t +A tm_t +A wait_t ;

[0070] (3.19) Calculate the current aluminum corner processing efficiency weight coefficient

[0071] (3.20) is substituted into the drilling and riveting task tendency model, F K =α1M t -α2T t -α3A t Calculate the drilling and riveting task tendency of the current aluminum angle edge;

[0072] (3.21)i=i+1, jump to (3.9);

[0073] (3.22) k = k + 1, jump to (3.8);

[0074] (3.23) Sort the drilling and riveting task tendency of the current automatic drilling and riveting robot for all aluminum corner edges from large to small;

[0075] (3.24) Take Int(max / 3) angle aluminum edges according to the sorting results, and set the number of times these angle aluminum edges are drilled and riveted to M. t =M t +1;

[0076] (3.25) N = N + 1, jump to (3.6);

[0077] (3.26) Set s = 1;

[0078] (3.27) Check if s > max_p. If yes, jump to (3.28). If no, jump to (3.33).

[0079] (3.28) Determine the t-th root angle aluminum M t >2, if not, jump to (3.32), if yes, jump to (3.29);

[0080] (3.29) Get the numbers of the two automatic drilling and riveting robots currently assigned to the drilling and riveting tasks for the current angle aluminum;

[0081] (3.30) Calculate the number of drilling tasks currently assigned to the two automatic drilling and riveting robots Σ(L z / dΔ);

[0082] (3.31) Compare the number of drilling tasks assigned to the two automatic drilling and riveting robots, and assign the drilling task to the automatic drilling and riveting robot with the smaller number of drilling tasks. If the number of tasks of the two automatic drilling and riveting robots is the same, the automatic drilling and riveting robot with the smaller number will be assigned to it by default;

[0083] (3.32) s = s + 1, jump to (3.27);

[0084] (3.33) Complete all automatic drilling and riveting robot task assignments, and send all drilling and riveting task assignments to the corresponding automatic drilling and riveting robots according to the stored automatic drilling and riveting robot numbers. The automatic drilling and riveting robots begin to perform automatic drilling and riveting operations in descending order according to the stored drilling and riveting task tendencies.

[0085] Furthermore, the establishment of the drilling angle compensation value model in step (4) specifically includes the following steps:

[0086] At each cabin corner aluminum facek, a numerical model is established between the compensation value and the distance between the drilling point of the cabin corner aluminum and the center of the automatic drilling and riveting robot system:

[0087] Where C is the compensation value of each automatic drilling and riveting robot system, t is the distance between the automatic drilling and riveting robot and the drilling center point, and x1, x2, and x3 are parameters to be determined;

[0088] By operating the automatic drilling and riveting robot to perform a drilling test on the corner aluminum at the jlifacek position, the distance of the drilling deviation from the set value and the distance between the automatic drilling and riveting robot and the drilling center point are obtained. The parameter acquisition optimization model of the numerical model of the distance between the drilling point of the corner aluminum of the square cabin and the center of the automatic drilling and riveting robot is established:

[0089] Substituting the data, we can get the optimal solution X(x1, x2, x3) for each set of compensation value coefficients, and then get the drilling angle compensation value model for the automatic drilling and riveting of the angle aluminum. Repeat until the compensation value models of all angle aluminums are obtained.

[0090] Preferably, the position compensation in step (4) specifically includes the following steps:

[0091] (4.1) Confirm the angle aluminum number;

[0092] (4.2) According to the angle aluminum number, call the compensation value model of the angle aluminum;

[0093] (4.3) Based on the scanning results, calculate the distance between each target drilling point and the center of the automatic drilling and riveting robot;

[0094] (4.4) Substitute the distance between each target drilling point and the center of the automatic drilling and riveting robot into the compensation value model to obtain and store the angle compensation values ​​of all drilling points;

[0095] (4.5) Compensate the drilling angle according to the compensation value obtained.

[0096] Furthermore, the intelligent drilling and riveting system for the container cabin includes an automatic drilling and riveting robot, a quick-change plate, a drilling end effector, a riveting and gluing end effector, a 3D vision camera, a tool changing table, a PLC electrical control cabinet, a robot base plate, a container cabin transfer vehicle and a robot ground rail. There are a total of four automatic drilling and riveting robots, numbered 1#, 2#, 3# and 4#.

[0097] Beneficial effects: Compared with the existing technology, the present invention has the following significant advantages: the present invention can be completely based on the production process requirements of the square cabin, and has the functions of drilling, gluing, and riveting. It also has a mechanism for identifying and scanning production trajectories and performing task allocation. It can adjust the processing position according to the square cabin assembly error and accurately position it. Through orderly division of labor, multiple automatic drilling and riveting robots can coordinate the automated drilling and riveting of the square cabin corner aluminum, thereby realizing efficient, high-quality, and highly intelligent installation of the square cabin corner aluminum. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] FIG1 is a structural diagram of the intelligent drilling and riveting system for a center cabin according to the present invention;

[0099] FIG2 is a structural diagram of the end effector of the automatic drilling and riveting system of the present invention;

[0100] FIG3 is a structural diagram of the end effector of the automatic drilling and riveting system of the present invention;

[0101] FIG4 is a flow chart showing the drilling positions of the Chinese capture module according to the present invention;

[0102] FIG5 is a flowchart of drilling and riveting task allocation in the present invention;

[0103] FIG6 is a flowchart of angle compensation of the automatic drilling and riveting robot in the present invention.

[0104] Automatic drilling and riveting robot 1, quick-change plate 2, drilling end effector 3, riveting and gluing end effector 4, 3D vision camera 5, tool changer 6, PLC electric control cabinet 7, fixed base plate 8, shelter transfer vehicle 9, automatic drilling and riveting robot ground rail 10, spindle 11, spindle fixing plate 12, drilling quick-change plate 13, drill 14, glue gun 15, rivet gun 16, glue cylinder 17, riveting cylinder 18, glue and riveting mechanism bracket 19, glue and riveting mechanism quick-change plate 20

[0105] Implementation Method

[0106] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0107] As shown in Figures 1, 2 and 3, the intelligent drilling and riveting system for a container cabin of the present invention includes an automatic drilling and riveting robot 1, a quick-change plate 2, a drilling end effector 3, a riveting and gluing end effector 4, a 3D vision camera 5, a tool changer 6, a PLC electrical control cabinet 7, a fixed base plate 8, a container cabin transfer vehicle 9 and an automatic drilling and riveting robot ground rail 10. The drilling end effector 3 includes a spindle 11, a spindle fixing plate 12, a drilling quick-change plate 13 and a drill row 14, wherein the drill row 14 is clamped to the spindle 11 by a pneumatic chuck inside the spindle, and can be driven by the spindle 11 to drill. One end of the spindle fixing plate 12 is screwed and fastened to the spindle 11, and the other end is screwed and fixed with a drilling quick-change plate 13; wherein the drill row 14 is a drill bit arranged in a row of four, which can process four holes at a time. The riveting and gluing end effector 4 includes a glue gun 15, a rivet gun 16, a gluing cylinder 17, a rivet cylinder 18, a gluing and rivet mechanism bracket 19, and a gluing and rivet mechanism quick-change plate 20. The rivet gun 15 is fastened to the rivet cylinder 18 by screws, the glue gun 15 is fixed to the gluing cylinder 17 by screws, the gluing cylinder 17 is fastened to the gluing and rivet mechanism bracket 19 by screws, and the other end of the gluing and rivet mechanism bracket 19 is screwed and fastened to the quick-change plate 20. The riveting and gluing end effector 4 can apply glue and rivet simultaneously. The automatic drilling and riveting robot ground rail 10 is fixed to the ground by anchor bolts. The fixed base plate 8 can slide along the automatic drilling and riveting robot ground rail 10. The automatic drilling and riveting robot 1 and the tool changer 6 are both screwed and fixed on the fixed base plate. A blocking block is installed at one end of the automatic drilling and riveting robot ground rail 10, which serves as the reference point for the entire system. The quick-change disk 2 and the 3D vision camera 5 are screwed and fixed to the end effector of the automatic drilling and riveting robot 1. The drilling end effector 3 and the riveting and gluing end effector 4 can be quickly connected to the quick-change disk 2. In actual use, one is fixed to the end of the automatic drilling and riveting robot 1 through the drilling quick-change disk 13, and the other is idle and placed on the tool changing table 6.

[0108] In actual operation, as shown in Figure 2, four automatic drilling and riveting robots are evenly distributed in the four corners of the automatic drilling and riveting area of ​​the shelter, and are numbered 1#, 2#, 3#, and 4# in a counterclockwise direction.

[0109] As shown in FIG4, FIG5 and FIG6, the present invention provides a multi-robot collaborative shelter intelligent drilling and riveting system control method, including the following steps:

[0110] (1) Generate the robot scanning trajectory based on the key characteristic parameters of the cabin;

[0111] The specific method for generating the robot scanning trajectory according to the key characteristic parameters of the shelter includes the following steps:

[0112] The key characteristic parameters of the shelter include the length l, the width w1, the projection width w2 of the chamfered part of the shelter top on the horizontal plane, the height h, and the chamfer angle θ. If the shelter is a right-angled shelter, w2 = 0.

[0113] Set the reference point of the cabin on the robot's ground track as the coordinate origin, and establish the coordinate axis through the right-hand system; store the cabin endpoints in two groups of arrays according to the right-angle cabin and the chamfered cabin respectively;

[0114] The endpoints of the cabins are stored in two arrays, one for the right-angle cabin and the other for the chamfered cabin. The right-angle cabin is represented by an array with a capacity of 8 as follows:

[0115] The chamfered cabin is represented as an array with a capacity of 12:

[0116] First, determine whether the input cabin key characteristic parameter w2 is 0. If it is 0, call the rectangular cabin array; if not, call the chamfered cabin.

[0117] With d _scan Indicates the scanning safety distance. For the 1#, 2#, 3#, and 4# cabin automatic drilling and riveting robots, the corresponding scanning action safety points of each automatic drilling and riveting robot are numbered O1, O2, O3, and O4, and are parameterized as follows: O1 = (-d _scan ,d _scan ,h+w2*sinθ+d _scan ) O2=(w1+d _scan ,d _scan ,h+w2*sinθ+d _scan ) O3=(-d _scan ,l+d _scan ,h+w2*sinθ+d _scan ) O4=(w+d _scan ,l+d _scan ,h+w2*sinθ+d _scan )

[0118] The parametric storage of angle aluminum uses an array with a capacity of 8 to store the 8 faces of the chamfered square cabin, and an array with a capacity of 6 to store the 6 faces of the rectangular square cabin. The chamfered square cabin and the rectangular square cabin can be represented in the order from front to back, from top to bottom, and from left to right as follows:

[0119] FACE1: [face1, face2, face3, face4, face5, face6, face7, face8], FACE2: [face1, face2, face3, face4, face5, face6], each face of the cabin has 4 aluminum corners, which are marked as a set of arrays JL: [jl1, jl2, jl3, jl4] in the clockwise direction. All the aluminum corners of a chamfered cabin are represented as:

[0120] All the aluminum corner edges of a rectangular cabin are expressed as:

[0121] That is, jl1face1, jl1face2... are used to store all the aluminum corner edges of the cabin to be polished. Any aluminum corner edge is represented by jlifacek, where the values ​​of i and k are

[0122] Determine whether the input cabin key feature parameter w2 is 0. If w2=0, it means the cabin is a rectangular cabin with 6 faces, and the maximum value of k is 6; if w2≠0, it means the cabin has 8 faces, and the maximum value of k is 8;

[0123] The robot's scanning vector can be obtained by the coordinate difference of the endpoints of the cabin. The robot's scanning trajectory is generated by the robot driving the camera coordinate system along the scanning vector parallel to the corner aluminum to be scanned, starting from the scanning action safety point and traversing all the corner aluminum edges in sequence. The corner aluminum scanning trajectory corresponding to each robot is stored in a structure array with a capacity equal to the maximum number of corner aluminum edges. The maximum number of corner aluminum edges for a chamfered cabin is 32, and the maximum number of corner aluminum edges for a right-angled cabin is 24. The style of each structure element in the array is as follows:

[0124] (2) Laser scanning of the diagonal aluminum wires to capture the drilling locations of the shelters;

[0125] The laser scanning of the diagonal aluminum wire to capture the drilling position of the shelter specifically includes the following steps:

[0126] (2.1) Select n fixed points P within the reach of the N# (N∈[1,4]) automatic drilling and riveting robot, and obtain the corresponding coordinates P_robot and P_camera of the fixed point P in the camera coordinate system and the robot coordinate system;

[0127] (2.2) Through P_robot = T * P_camera, we get the coordinate transformation matrix T = P_robot * (P_camera) -1 ;

[0128] (2.3) The automatic drilling and riveting robot sends a string to the 3D vision camera, which parses the string and obtains the number of the automatic drilling and riveting robot;

[0129] (2.4) The 3D vision camera activates the line laser, and the automatic drilling and riveting robot scans the track, scanning the area where the angle aluminum is installed on the edge of the cabin;

[0130] (2.5) During the 3D vision camera scanning process, the angle aluminum and the cabin aluminum skin are distinguished based on the height difference of the reflected laser light, and a point cloud image of the angle aluminum is generated;

[0131] (2.6) Extract the boundary points and coordinates of the aluminum corner point cloud image and remove the noise in the point cloud image;

[0132] (2.7) Connect the extracted boundary points in clockwise order to obtain the aluminum angle contour image;

[0133] (2.8) The contour image of the aluminum angle edge is offset in the width direction by the set minimum distance Δ1 between the rivet and the edge of the aluminum angle edge in the length direction, and in the length direction by the minimum distance Δ2 between the rivet and the edge of the aluminum angle edge in the width direction;

[0134] (2.9) Calculate the intersection of the two offset lines as the starting point for drilling the angle aluminum hole;

[0135] (2.10) Create a stack for storing the drilling positions of angle aluminum;

[0136] (2.11) The coordinate transformation matrix T is used to convert the drilling starting point position in the camera coordinate system into the coordinate position of the automatic drilling and riveting robot and store it at the bottom of the stack;

[0137] (2.12) Determine whether the remaining angle aluminum length is less than the maximum distance of the drill bit 3*D. If so, go to (2.13); if not, go to (2.14);

[0138] (2.13) Based on the maximum distance 3*D of the drill bit, offset along the length direction to obtain the drilling position of the angle aluminum drill bit and convert it into the coordinates of the automatic drilling and riveting robot;

[0139] (2.14) Mark the current point processing method as row drilling and save the current point into the stack;

[0140] (2.15) Determine whether the set angle aluminum boundary distance has been reached. If not, go to (2.16). If yes, jump to (2.17).

[0141] (2.16) Using the minimum hole spacing D of the drilling row as an offset, the drilling starting point position in the camera coordinate system is converted into the coordinate position of the automatic drilling and riveting robot;

[0142] (2.17) Mark the current point processing method as single drilling, save the current point position into the stack, and jump to (2.15);

[0143] (2.18) The coordinates of each point in the stack are sent to the automatic drilling and riveting robot in a first-in-last-out order to be used as the location points for punching, thus ending the scanning process.

[0144] (3) Establish a task allocation mechanism, establish a drilling and riveting task tendency model based on the task allocation mechanism, and perform multi-robot automatic drilling and riveting task allocation based on the drilling and riveting task tendency model;

[0145] The task allocation mechanism includes that each automatic drilling and riveting robot gives priority to processing the square cabin angle aluminum that requires the most drilling and riveting, that is, it gives priority to the angle aluminum with the largest number of holes to be drilled; each automatic drilling and riveting robot avoids interference and waiting with other robots, that is, it gives priority to the path with the smallest interference area with other robots; each automatic drilling and riveting robot avoids square cabin angle aluminum with high movement complexity, that is, it gives priority to the angle aluminum with the least total joint movement and the shortest total movement time; each angle aluminum is completed by a maximum of two automatic drilling and riveting robots.

[0146] The drilling and riveting task tendency model is: F K =α1M t -α2T t -α3S t ,

[0147] Among them, α1 is the weight coefficient of the work task and its value range is [1,2]. Among them, M t is the number of holes to be drilled for the t-th angle aluminum, and is the length of the angle aluminum L t The ratio of the number of angle aluminum edges to the hole spacing dΔ is rounded up, and max is the total number of angle aluminum edges, where the maximum number of chamfered square cabins is 32, and the maximum number of right-angled square cabins is 24;

[0148] α2 is the security weight system value range is [0.5,1], T t The length R of the overlapped area between the current angle aluminum and other robots’ reachability t With the current angle aluminum length L t The ratio of T t =R t / L t ;

[0149] α3 is the efficiency weight coefficient, and its value range is [0.2, 0.5]. Among them A t The total time required for the current automatic drilling and riveting robot to complete the current angle aluminum operation: A t =A safe1_t +A sm_t +A safe2_t +A hd_t +A safe3_t +A dk_t +A hd2_t +A safe4_t +A tm_t +A wait_t ,

[0150] Among them A safe1_t A is the time it takes for the robot to move from the safety point to the scanning starting point. sm_t is the scanning time, A safe2_t A is the time from the scanning end point to the safe point,hd_t is the tool change time, A safe3_t A is the time from tool change point to safety point, dk_t is the punching time, A hd2_t For the time of changing glue rivet tool, A safe4_t A is the time from the tool change point to the safety point of the glue riveting module. tm_t A is the execution time of glue riveting, wait_t is the waiting time, and total is the total number of aluminum angle edges that the robot can currently determine to be reachable.

[0151] The multi-robot automatic drilling and riveting task allocation specifically includes the following steps:

[0152] (3.1) Initialization, set the automatic drilling and riveting robot serial number N = 1, set the number of times all angle aluminum to be drilled and riveted M t =1;

[0153] (3.2) Obtain the projection width w2 of the chamfered portion of the cabin roof on the horizontal plane;

[0154] (3.3) Check if w2 is 0. If so, jump to (3.4). If not, jump to (3.5).

[0155] (3.4) Set the maximum number of angle aluminum to 24 and the maximum number of planes to 6.

[0156] (3.5) Set the maximum number of angle aluminum to 32 and the maximum number of planes to 8.

[0157] (3.6) Check if N > 4. If not, jump to (3.7). If so, jump to (3.26).

[0158] (3.7) Set k = 1;

[0159] (3.8) Check if k > max_p. If not, jump to (3.9). If yes, jump to (3.23).

[0160] (3.9) Set i = 1

[0161] (3.10) Check if i>4? If not, jump to (3.11); if yes, jump to (3.22);

[0162] (3.11) Set t = (k-1)*max_p+i;

[0163] (3.12) Determine the accessibility of the N# automatic drilling and riveting robot at the t-th aluminum corner edge. If it is reachable, go to (3.13); if not, jump to (3.21).

[0164] (3.13) Determine the number of drilling and rivet marks on the current aluminum corner edge M t =2? If yes, jump to (3.14), if not, jump to (3.21);

[0165] (3.14) Substitute the current angle aluminum side length and punching spacing setting values ​​into INT(L t / dΔ);

[0166] (3.15) Calculate the current working task weight value of the angle aluminum edge

[0167] (3.16) Calculate the overlap length L of the reachable area between the current automatic drilling and riveting robot and the other automatic drilling and riveting robots during the operation of the current angle aluminum edge. t ;

[0168] (3.17) Calculate the current angle aluminum safety weight coefficient T t =R t / L t ;

[0169] (3.18) Calculate and estimate the operating time of the current automatic drilling and riveting robot on the current aluminum corner edge: A t =A safe1_t +A sm_t +A safe2_t +A hd_t +A safe3_t +A dk_t +A hd2_t +A safe4_t +A tm_t +A wait_t ;

[0170] (3.19) Calculate the current aluminum corner processing efficiency weight coefficient

[0171] (3.20) is substituted into the drilling and riveting task tendency model, F K =α1M t -α2T t -α3A t Calculate the drilling and riveting task tendency of the current aluminum angle edge;

[0172] (3.21)i=i+1, jump to (3.9);

[0173] (3.22) k = k + 1, jump to (3.8);

[0174] (3.23) Sort the drilling and riveting task tendency of the current automatic drilling and riveting robot for all aluminum corner edges from large to small;

[0175] (3.24) Take Int(max / 3) angle aluminum edges according to the sorting results, and set the number of times these angle aluminum edges are drilled and riveted to M. t =M t +1;

[0176] (3.25) N = N + 1, jump to (3.6);

[0177] (3.26) Set s = 1;

[0178] (3.27) Check if s > max_p. If yes, jump to (3.28). If no, jump to (3.33).

[0179] (3.28) Determine the t-th root angle aluminum M t >2, if not, jump to (3.32), if yes, jump to (3.29);

[0180] (3.29) Get the numbers of the two automatic drilling and riveting robots currently assigned to the drilling and riveting tasks for the current angle aluminum;

[0181] (3.30) Calculate the number of drilling tasks currently assigned to the two automatic drilling and riveting robots Σ(L z / dΔ);

[0182] (3.31) Compare the number of drilling tasks assigned to the two automatic drilling and riveting robots and assign the drilling task to the automatic drilling and riveting robot with the smaller number of drilling tasks. If the number of tasks of the two automatic drilling and riveting robots is equal, the drilling and riveting robot with the smaller number will be assigned to it by default.

[0183] (3.32) s = s + 1, jump to (3.27);

[0184] (3.33) Complete all automatic drilling and riveting robot task assignments, and send all drilling and riveting task assignments to the corresponding automatic drilling and riveting robots according to the stored automatic drilling and riveting robot numbers. The automatic drilling and riveting robots begin to perform automatic drilling and riveting operations in descending order according to the stored drilling and riveting task tendencies.

[0185] (4) A drilling angle compensation value model is established, and the automatic drilling and riveting robot performs position compensation based on the drilling angle compensation value model to complete drilling and riveting.

[0186] The establishment of the drilling angle compensation value model specifically includes the following steps:

[0187] At each cabin corner aluminum facek, a numerical model is established between the compensation value and the distance between the drilling point of the cabin corner aluminum and the center of the automatic drilling and riveting robot system:

[0188] Where C is the compensation value of each automatic drilling and riveting robot system, t is the distance between the automatic drilling and riveting robot and the drilling center point, and x1, x2, and x3 are parameters to be determined;

[0189] By operating the automatic drilling and riveting robot to perform a drilling test on the corner aluminum at the jlifacek position, the distance of the drilling deviation from the set value and the distance between the automatic drilling and riveting robot and the drilling center point are obtained. The parameter acquisition optimization model of the numerical model of the distance between the drilling point of the corner aluminum of the square cabin and the center of the automatic drilling and riveting robot is established:

[0190] Substituting the data, we can get the optimal solution X(x1, x2, x3) for each set of compensation value coefficients, and then get the drilling angle compensation value model for the automatic drilling and riveting of the angle aluminum. Repeat until the compensation value models of all angle aluminums are obtained.

[0191] The position compensation comprises the following steps:

[0192] (4.1) Confirm the angle aluminum number;

[0193] (4.2) According to the angle aluminum number, call the compensation value model of the angle aluminum;

[0194] (4.3) Based on the scanning results, calculate the distance between each target drilling point and the center of the automatic drilling and riveting robot;

[0195] (4.4) Substitute the distance between each target drilling point and the center of the automatic drilling and riveting robot into the compensation value model to obtain and store the angle compensation values ​​of all drilling points;

[0196] (4.5) Compensate the drilling angle according to the compensation value obtained.

[0197] The intelligent drilling and riveting system for the shelter includes an automatic drilling and riveting robot, a quick-change plate, a drilling end effector, a riveting and gluing end effector, a 3D vision camera, a tool changer, a PLC electrical control cabinet, a robot base plate, a shelter transfer vehicle and a robot ground rail. There are four automatic drilling and riveting robots in total, numbered 1#, 2#, 3# and 4#.

[0198] The multi-robot collaborative intelligent drilling and riveting system for shelters in the present invention has a specific task execution process including:

[0199] Input the key characteristic parameters of the angle aluminum into the PLC electric control cabinet 7, establish the key characteristic parameterization of the cabin, and generate the scanning trajectory based on the key characteristic parameterization of the cabin, generate the scanning trajectory of the angle aluminum of each automatic drilling and riveting robot and store it; perform automatic drilling and riveting task allocation for multiple automatic drilling and riveting robots, and obtain the drilling and riveting tasks to be performed by each automatic drilling and riveting robot; 1#, 2#, 3#, 4# automatic drilling and riveting robots 1 calculate the drilling and riveting task tendency model results of each angle aluminum edge, and start to execute the angle aluminum with the highest drilling and riveting task tendency The automatic drilling and riveting task is to drive the 3D vision camera 5 by the automatic drilling and riveting robot 1 to capture the drilling position of the cabin, store the captured cabin drilling points in a stack, compensate for the drilling points, and move the automatic drilling and riveting robot to the tool changer 6. Through the quick connection between the quick changer 3 and the drilling changer 13, the drilling end effector 3 is picked up. According to the principle of last in first out of the stack, the automatic drilling and riveting robot 1 drives the drilling end effector 3 to drill holes in the drilling points out of the stack in turn until all holes are drilled. The automatic drilling and riveting robot 1 drives the drilling end effector 3 to move to the tool changer to replace the glue coating and riveting end effector 4, and then driven by the automatic drilling and riveting robot, it moves back to the drilling point and performs the glue coating and riveting action by riveting the glue coating end effector 4. During the automatic gluing process, the gluing cylinder 17 drives the glue gun 15 to apply glue to the previous hole first, while the riveting cylinder 18 simultaneously drives the riveting gun 16 to rivet the next, already-glued hole. Both operate simultaneously, improving work efficiency. This process repeats until each automatic drilling and riveting robot 1 completes its assigned drilling and riveting task. The process ends, and the entire shelter completes all automatic drilling and riveting work. This ultimately implements a control method for a shelter intelligent drilling and riveting system that coordinates multiple automatic drilling and riveting robots, completing efficient, safe, and intelligent unmanned installation of shelter corner aluminum.

[0200] The present invention provides a control method for a multi-robot collaborative shelter intelligent drilling and riveting system. The above description is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art may make various improvements and modifications without departing from the principles of the present invention, and such improvements and modifications are also within the scope of protection of the present invention. Components not specified in this embodiment may be implemented using existing technologies.

Claims

1. A multi-robot collaborative shelter intelligent drilling and riveting system control method, characterized in that: The steps include: (1) Generate the robot scanning trajectory based on the key characteristic parameters of the shelter; The specific method of generating the robot scanning trajectory according to the key characteristic parameters of the shelter in step (1) comprises the following steps: The key characteristic parameters of the shelter include the shelter length l, the shelter width w1, the projection width w2 of the chamfered part of the shelter top on the horizontal plane, the shelter height h, and the shelter chamfer angle θ. If the shelter is a right-angle shelter, w2 = 0; The reference point of the cabin on the robot track is set as the coordinate origin, and the coordinate axis is established through the right-hand system; the cabin endpoints are stored in two groups of arrays according to the right-angle cabin and the chamfered cabin; With d _scan Indicates the scanning safety distance. The four automatic drilling and riveting robots in the shelter are numbered 1#, 2#, 3#, and 4#. The scanning action safety points corresponding to each automatic drilling and riveting robot are numbered O1, O2, O3, and O4, and are parameterized as follows: O1=(-d _scan ,d _scan ,h+w2*sinθ+d _scan ) O2=(w1+d _scan ,d _scan ,h+w2*sinθ+d _scan ) O3=(-d _scan ,l+d _scan ,h+w2*sinθ+d _scan ) O4=(w+d _scan ,l+d _scan ,h+w2*sinθ+d _scan ) The parameterized storage of the aluminum corners uses an array with a capacity of 8 to store the 8 faces of the chamfered cabin, and an array with a capacity of 6 to store the 6 faces of the rectangular cabin. The chamfered cabin and the rectangular cabin can be expressed in the order from front to back, from top to bottom, and from left to right as follows: FACE1: [face1, face2, face3, face4, face5, face6, face7, face8], FACE2: [face1, face2, face3, face4, face5, face6], each face of the cabin has 4 corner aluminum edges, marked as a set of arrays JL in the clockwise direction: [jl1.jl2, jl3, jl4], all the corner aluminum edges of a chamfered cabin are expressed as: All the aluminum corner edges of a rectangular cabin are expressed as: That is, jl1face1, jl1face2, etc. are used to store all the aluminum corners of the cabin to be polished. Any aluminum corner is represented by jlifacek, where the values ​​of i and k are Determine whether the key characteristic parameter w2 of the input shelter is 0. w2=0 indicates that the shelter is a rectangular shelter with 6 faces, and the maximum value of k is 6; w2≠0 indicates that the shelter has 8 faces, and the maximum value of k is 8; The robot's scanning vector can be obtained through the endpoint coordinate difference of the cabin. The robot scanning trajectory is generated when the robot drives the camera coordinate system along the scanning vector parallel to the corner aluminum edge to be scanned, starting from the scanning action safety point, and traverses and scans all the corner aluminum edges in sequence. The corner aluminum edge scanning trajectory corresponding to each robot is stored in a structure array. The capacity of the structure array is the maximum number of corner aluminum edges. The maximum number of corner aluminum edges of the chamfered cabin is 32, and the maximum number of corner aluminum edges of the right-angle cabin is 24. The style of each structure element of the array is: (2) Line laser scanning of the diagonal aluminum edge is performed to capture the drilling position of the shelter; (3) Establish a task allocation mechanism, establish a drilling and riveting task preference model based on the task allocation mechanism, and perform multi-robot automatic drilling and riveting task allocation based on the drilling and riveting task preference model; (4) A drilling angle compensation value model is established, and the automatic drilling and riveting robot performs position compensation based on the drilling angle compensation value model to complete drilling and riveting; The step (4) of establishing the drilling angle compensation value model specifically includes the following steps: At each cabin corner aluminum edge jlifacek, a numerical model of the compensation value and the distance between the drilling point of the cabin corner aluminum edge and the center of the automatic drilling and riveting robot is established: Where C is the compensation value of each automatic drilling and riveting robot system, t is the distance between the automatic drilling and riveting robot and the center point of the drilling hole, and x1, x2, and x3 are parameters to be determined; By operating the automatic drilling and riveting robot to perform a drilling test on the aluminum corner at the jlifacek position, the distance of the drilling deviation from the set value and the distance between the automatic drilling and riveting robot and the center point of the drilling are obtained, and a parameter acquisition optimization model of the numerical model of the distance between the drilling point of the aluminum corner of the square cabin and the center of the automatic drilling and riveting robot is established: Substituting the data, we can get the optimal solution X(x1,x2,x3) for each set of compensation value coefficients, and then get the drilling angle compensation value model for the automatic drilling and riveting of the aluminum angle edge. Repeat until the compensation value models of all aluminum angle edges are obtained.

2. The control method of a multi-robot collaborative shelter intelligent drilling and riveting system according to claim 1 is characterized in that: The endpoints of the shelters are stored in two groups of arrays according to the right-angle shelters and the chamfered shelters, respectively, wherein the right-angle shelter is represented by an array with a capacity of 8 as follows: The chamfered cabin is represented as an array with a capacity of 12: First, determine whether the input cabin key characteristic parameter w2 is 0. If it is 0, call the rectangular cabin array; if not, call the chamfered cabin array.

3. The control method of a multi-robot collaborative shelter intelligent drilling and riveting system according to claim 1 is characterized in that: In step (2), the linear laser scanning of the diagonal aluminum edge to capture the drilling position of the shelter specifically includes the following steps: (2.1) Select n fixed points P within the reachable range of the N#th automatic drilling and riveting robot, N∈[1,4], and obtain the corresponding coordinates P_robot, P_camera of the fixed point P in the camera coordinate system and the robot coordinate system respectively; (2.2) Through P_robot = T*P_camera, we get the coordinate transformation matrix T = P_robot*(P_camera) -1 ; (2.3) The automatic drilling and riveting robot sends a string to the 3D vision camera, and the 3D vision camera parses the string to obtain the number of the automatic drilling and riveting robot; (2.4) The 3D vision camera activates the line laser, and the automatic drilling and riveting robot scans the track and the area where the angle aluminum is installed on the edge of the cabin; (2.5) During the scanning process of the 3D vision camera, the aluminum corner edge and the aluminum skin of the cabin are distinguished based on the height difference of the reflected laser, and a point cloud image of the aluminum corner edge is generated; (2.6) Extract the boundary points and coordinates of the aluminum corner edge point cloud image, and remove the noise points in the point cloud image; (2.7) Connect the extracted boundary points in clockwise order to obtain the contour image of the aluminum corner edge; (2.8) The contour image of the angle aluminum edge is offset in the width direction according to the set minimum distance Δ1 between the rivet and the edge of the angle aluminum edge in the length direction, and in the length direction according to the minimum distance Δ2 between the rivet and the edge of the angle aluminum edge in the width direction; (2.9) Calculate the intersection of the two offset straight lines as the starting point for drilling holes on the aluminum corners; (2.10) Establish a stack for storing the drilling positions of the angle aluminum edges; (2.11) Through the coordinate transformation matrix T, the drilling starting point position in the camera coordinate system is converted into the coordinate position of the automatic drilling and riveting robot and stored at the bottom of the stack; (2.12) Determine whether the remaining aluminum angle edge length is less than the maximum drill bit spacing 3*D. If yes, go to (2.13); if not, go to (2.14); (2.13) Based on the maximum spacing of the drill bit 3*D, the drill bit is offset along the length direction to obtain the drilling position of the corner aluminum edge and convert it into the coordinates of the automatic drilling and riveting robot; (2.14) Mark the current point processing method as row drilling and store the current point into the stack; (2.15) Determine whether the set angle aluminum boundary distance has been reached, if not, go to (2.16), if yes, jump to (2.17); (2.16) The drilling start point position in the camera coordinate system is converted into the coordinate position of the automatic drilling and riveting robot by using the minimum hole spacing D of the row drilling; (2.17) Mark the current point processing method as single drilling, store the current point in the stack, and jump to (2.15); (2.18) The coordinates of each point in the stack are sent to the automatic drilling and riveting robot in a first-in-last-out order to be used as the location point for drilling holes, thus ending the scanning process.

4. The control method of a multi-robot collaborative shelter intelligent drilling and riveting system according to claim 1 is characterized in that: The task allocation mechanism in step (3) includes that each automatic drilling and riveting robot first processes the aluminum corner edge of the cabin that needs the most drilling and riveting, that is, first selects the aluminum corner edge with the largest number of holes to be drilled; each automatic drilling and riveting robot avoids interference and waiting with other robots, that is, first selects the path with the smallest interference area with other robots; each automatic drilling and riveting robot avoids the aluminum corner edge of the cabin with a high complexity of its own movement, that is, first selects the aluminum corner edge with the least total joint movement and the shortest total movement time; each aluminum corner edge is completed by at most two automatic drilling and riveting robots.

5. The control method of a multi-robot collaborative shelter intelligent drilling and riveting system according to claim 4 is characterized in that: The drilling and riveting task tendency model in step (3) is: K =α1M t -α2T t -α3S t , Where α1 is the weight coefficient of the work task, and its value range is [1,2]. Among them, M t is the number of holes to be drilled on the t-th aluminum corner edge, INT(L t / dΔ) is the length of the aluminum angle side L t and the ratio of the hole spacing dΔ to the integer, max is the total number of angle aluminum edges, where the maximum number of chamfered square cabins is 32, and the maximum number of right-angled square cabins is 24; α2 is the safety weight coefficient, and its value range is [0.5,1]. t is the safety weight value, which is the length R of the overlap area between the current aluminum corner edge and other robot reachability areas t The length of the current aluminum angle side L t The ratio of t =R t / L t ; α3 is the efficiency weight coefficient, and its value range is [0.2, 0.5]. Among them A t The total time required for the current automatic drilling and riveting robot to complete the current angle aluminum operation: A t =A safe1_t +A sm_t +A safe2_t +A hd_t +A safe3_t +A dk_t +A hd2_t +A safe4_t +A tm_t +A wait_t , Among them A safe1_t is the time it takes for the robot to move from the safety point to the scanning start point, A sm_t is the scanning time, A safe2_t A is the time from the scanning end point to the safe point. hd_t is the tool change time, A safe3_t A is the time from tool change point to safety point, dk_t is the punching time, A hd2_t The time for changing the glue rivet tool is A safe4_t A is the time from the tool change point to the safety point of the glue riveting module. tm_t A is the execution time of glue-coated riveting, wait_t is the waiting time, and total is the total number of aluminum angle edges that the robot can currently determine to be reachable.

6. The control method of a multi-robot collaborative shelter intelligent drilling and riveting system according to claim 5 is characterized in that: The multi-robot automatic drilling and riveting task allocation in step (3) specifically includes the following steps: (3.1) Initialization, set the automatic drilling and riveting robot serial number N = 1, set the number of times all aluminum corners need to be drilled and riveted M t =1; (3.2) Obtain the projection width w2 of the chamfered part of the cabin top on the horizontal plane; (3.3) Check if w2 is 0. If yes, jump to (3.4); if not, jump to (3.5); (3.4) Set the maximum number of aluminum corner edges max to 24 and the maximum number of planes max_p to 6; (3.5) Set the maximum number of aluminum corner edges max to 32 and the maximum number of planes max_p to 8; (3.6) Check if N>4? N is the number of the automatic drilling and riveting robot. If not, jump to (3.7). If yes, jump to (3.26). (3.7) Set k = 1; (3.8) Check if k>max_p? If not, jump to (3.9); if yes, jump to (3.23); (3.9) Set i = 1 (3.10) Check if i>4? If not, jump to (3.11); if yes, jump to (3.22); (3.11) Set t = (k-1)*max_p+i; (3.12) Determine the accessibility of the N# automatic drilling and riveting robot at the t-th root aluminum corner edge. If it is reachable, go to (3.13); if it is not reachable, jump to (3.21); (3.13) Determine the number of drilling and rivet marks M on the current aluminum corner edge t =2? If yes, jump to (3.14), if no, jump to (3.21)； (3.14) Substitute the current angle aluminum side length and punching spacing setting value into INT(L t / dΔ); (3.15) Calculate the current working task weight value of the angle aluminum edge (3.16) Calculate the overlap length L of the reachable area of ​​the current automatic drilling and riveting robot and other automatic drilling and riveting robots during the operation of the current corner aluminum edge. t ; (3.17) Calculate the current angle aluminum edge safety weight value T t =R t / L t ; (3.18) Calculate and estimate the operating time of the current automatic drilling and riveting robot on the current aluminum corner edge: A t =A safe1_t +A sm_t +A safe2_t +A hd_t +A safe3_t +A dk_t +A hd2_t +A safe4_t +A tm_t +A wait_t ; (3.19) Calculate the current aluminum corner processing efficiency weight value (3.20) is substituted into the drilling and riveting task tendency model, F K =α1M t -α2T t -α3A t Calculate the drilling and riveting task tendency of the current aluminum corner edge; (3.21)i=i+1, jump to (3.10); (3.22) k = k + 1, jump to (3.8); (3.23) Sort the drilling and riveting task inclination of the current automatic drilling and riveting robot for all aluminum corner edges from large to small; (3.24) Take Int(max / 3) angle aluminum edges according to the sorting results, and set the number of times these angle aluminum edges need to be drilled and riveted as M. t =M t +1; (3.25) N = N + 1, jump to (3.6); (3.26) Set s = 1; (3.27) Check if s>max_p? If yes, jump to (3.28), if not, jump to (3.33); (3.28) Determine the tth root angle aluminum edge M t >2, if not, jump to (3.32), if yes, jump to (3.29); (3.29) Obtain the numbers of the two automatic drilling and riveting robots that are currently assigned to the drilling and riveting tasks for the corner aluminum edge; (3.30) Calculate the number of drilling tasks currently assigned to the two automatic drilling and riveting robots Σ(L z / dΔ); (3.31) Compare the number of drilling tasks assigned to the two automatic drilling and riveting robots, and assign the drilling task to the automatic drilling and riveting robot with the smaller number of drilling tasks. If the number of tasks of the two automatic drilling and riveting robots is the same, the task will be assigned to the automatic drilling and riveting robot with the smaller number by default; (3.32) s = s + 1, jump to (3.27); (3.33) Complete all automatic drilling and riveting robot task assignments, and send all drilling and riveting task assignments to the corresponding automatic drilling and riveting robots according to the stored automatic drilling and riveting robot numbers. The automatic drilling and riveting robots begin to perform automatic drilling and riveting operations in a descending order according to the stored drilling and riveting task preferences.

7. The control method of a multi-robot collaborative shelter intelligent drilling and riveting system according to claim 1 is characterized in that: The position compensation in step (4) specifically includes the following steps: (4.1) Confirm the aluminum angle edge number; (4.2) Based on the aluminum angle edge number, call the drilling angle compensation value model for the automatic drilling and riveting of the aluminum angle edge; (4.3) Based on the scanning results, calculate the distance between each target drilling point and the center of the automatic drilling and riveting robot; (4.4) Substituting the distance between each target drilling point and the center of the automatic drilling and riveting robot into the compensation value model, obtaining the angle compensation values ​​of all drilling points and storing them; (4.5) Compensate the drilling angle according to the compensation value obtained.

8. The control method of a multi-robot collaborative shelter intelligent drilling and riveting system according to claim 1 is characterized in that: The intelligent drilling and riveting system for the container cabin includes an automatic drilling and riveting robot, a quick-change plate, a drilling end effector, a riveting and gluing end effector, a 3D vision camera, a tool changer, a PLC electrical control cabinet, a robot base plate, a container cabin transfer vehicle and a robot ground rail. There are a total of 4 automatic drilling and riveting robots.

Images