Handling system and method based on rocker-controlled cable-tightly coupled multi-manipulator collaboration

Abstract

This invention belongs to the field of robotics, specifically a handling system and method based on rocker-controlled cables tightly coupling multiple robotic arms. The control system comprises multiple multi-degree-of-freedom robotic arms, each of which is connected to multiple fixed points on the object being handled via multiple cables. A rocker is used to control the position and posture of the object being handled. The operating parameters of the robotic arms and rocker meet certain requirements. This system offers advantages such as precise detection and stable operation, and is suitable for robotics teaching and handling fragile objects.

Description

Technical Field

[0001] The present invention relates to a robot, and in particular to a handling system and method based on rocker-controlled cable-tightly coupled multi-robot collaboration. Background Art

[0002] In recent years, with the improvement of production efficiency, a single industrial robotic arm can no longer meet production needs. Robots have gradually developed from single-unit operations to clustered operations. Multiple robots are needed to collaborate to complete operations such as moving, pushing, pulling, and lifting the same object. This requires physical connections and internal force constraints between robots.

[0003] When the relative positions of the robots deviate, large internal forces are generated, which can damage the objects being transported and the robots themselves. Although experiments have shown that using torque-controlled manipulators can solve this problem, position-controlled manipulators are still widely used in the market, and torque-controlled solutions are not suitable for practical applications.

[0004] Currently in the laboratory, an academic team has conducted dynamic modeling of a handling system consisting of two robots after considering the uncertainty of the system and external interference. Based on the operation of the system, a force control method was constructed to eliminate the internal forces generated between the robots. However, the robots normally available on the market are all speed-controlled and position-controlled. If torque control is to be used, the controller needs to be replaced, which cannot be applied in daily life.

[0005] like Figure 1 As shown, one existing solution is to install a passive joint at the end of the robotic arm. This eliminates the need to replace the robot controller, simplifies the coordinated control of multiple robots, and eliminates the influence of internal forces. However, the passive joint itself is relatively complex and needs to be customized according to the size of the robotic arm, making it difficult to implement in practice.

[0006] Existing lifting operations require multiple cranes to tow and lift large objects. Cables are introduced to replace the end-of-robot support mechanism, shifting from rigid to flexible control and eliminating internal forces caused by positional errors. However, there is currently no method to directly control each robotic arm by controlling the position and posture of the object being transported. Instead, each robotic arm must be controlled individually, resulting in poor human-machine interaction. Furthermore, as a large-scale collaborative handling system, its redundant degrees of freedom have not yet been utilized for feedback optimization. Summary of the Invention

[0007] In order to solve the deficiencies in the above-mentioned prior art solutions, the present invention provides a handling system based on rocker-controlled cables tightly coupled with multi-robotic arm collaboration.

[0008] The purpose of the present invention is achieved through the following technical solutions:

[0009] A handling system based on rocker-controlled cable-tightly coupled multi-manipulator collaboration, comprising:

[0010] Multiple multi-degree-of-freedom robotic arms, the ends of which are connected to multiple fixed points of the transported objects via multiple cables;

[0011] A joystick, which is used to control the position and posture of the object being transported;

[0012] The working parameters of the robot arm and rocker meet the requirements.

[0013] , ;

[0014] J m is the Jacobian matrix between the robot arm joint angle and the position and posture of the robot end, Δθ i is the slight change in the angle of each joint of the robot arm, J0 is the Jacobian matrix of the relationship between the slight displacement of each robot end and the slight variable of the position and posture of the transported object, K j is the coefficient of variation between the joystick information and the velocity information of each degree of freedom of the transported object, X jn The information output when the joystick is in use, X j0 is the origin information of the joystick, t is the sampling time of each joystick, k is a constant, I is the unit matrix, H is the evaluation function, x e is the position matrix of the end of the robotic arm.

[0015] The present invention also aims to provide a control method for a handling system based on rocker-controlled cables tightly coupled with multi-manipulator collaboration. This invention objective is achieved through the following technical solutions:

[0016] The control method of a handling system based on rocker-controlled cables tightly coupled with multi-manipulator collaboration includes the following steps:

[0017] (A1) Obtain the change ΔX in the position and posture of the transported object each time 0i , and the area S2 of the area enclosed by the lines connecting multiple cable fixing points on the transported object;

[0018] (A2) Obtain the Jacobian matrix J0, the cable length change ΔL and the small changes in the position and posture of the transported object ΔX 0i The Jacobian matrix J of the relationship between w , the area S1 of the area enclosed by the lines connecting the ends of each robot arm;

[0019] (A3) Obtain the change Δx in the end position of the robot arm each time e ;

[0020] (A4) Determine whether area S1 is greater than area S2 and whether i is less than m. If the results are yes and no, respectively, proceed to the next step. If the first result is no, interrupt the process. If the second result is yes, return to step (A2).

[0021] (A5) X of the object to be transported 0target Set to X now , let n=1, and change x ei Set to x en , and read the real-time position information of the joystick X jn And button information button, confirm that the control button is pressed;

[0022] (A6) Using the method of step (A2), obtain the area S1 of the area enclosed by the line connecting the ends of each robot arm;

[0023] (A7) According to the real-time position information of each joystick X jn , Jacobian matrix J m and the Jacobian matrix J w , get the angle Δθ of each rotation of the robot arm n ;

[0024] (A8) n is incremented by 1 to determine whether area S1 is greater than area S2. If yes, proceed to the next step. If no, stop the process and finally update the position and posture X of the object being transported. now =X now +ΔX n , return to step (A7).

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] 1. Solved the problem of multi-robot collaboration;

[0027] A handling system and control method were designed, which used multiple robotic arms to collaboratively carry large and heavy objects. The robotic arms and objects were connected by cables, which effectively solved the problem of rigid connection between the robotic arms and objects in the multi-robotic arm collaborative handling system, avoided the generation of internal forces, and thus protected the objects being transported.

[0028] In the control method, the kinematic relationship of the entire system is used to calculate the joint parameters of multiple (e.g., three) robotic arms using only the position of the object, thereby achieving the purpose of collaborative handling.

[0029] 2. Simple structure;

[0030] There is no need to design and manufacture a dedicated clamping mechanism, which to a certain extent enhances the versatility and flexibility of system operation and reduces structural complexity;

[0031] 3. Good human-computer interaction;

[0032] Using the output value of the joystick as the target position and posture of the transported object enhances human-machine interaction and real-time performance, allowing operators to control the system and complete the transport task more intuitively, making it better suited for situations where multiple robotic arms coordinate to transport large and fragile items.

[0033] 4. Good stability;

[0034] By utilizing the multiple redundant degrees of freedom of a cable-tightly coupled multi-manipulator collaborative handling system, an evaluation function for the sum of squares of cable tension is proposed. This function is used to optimize the system, reducing the tension between the cables and optimizing system stability. This system can be miniaturized and used in school experiments, academic research, and other occasions to assist in robotics teaching activities. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The disclosure of the present invention will become more easily understood with reference to the accompanying drawings. Those skilled in the art will readily appreciate that these drawings are merely used to illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention.

[0036] Figure 1 This is a simplified diagram of the structure of a mechanical arm with a passive joint installed in the prior art;

[0037] Figure 2 is a simplified structural diagram of a handling system according to the present invention;

[0038] Figure 3 is a simplified diagram of the state of the handling system according to the present invention;

[0039] Figure 4 It is a schematic diagram of mathematical modeling of the end of the robotic arm and the object;

[0040] Figure 5 This is a schematic diagram of the tiny displacement of the end of the robotic arm and the object;

[0041] Figure 6 This is a diagram of the joystick;

[0042] Figure 7 This is a schematic diagram of the tension on the cable when no evaluation function is used;

[0043] Figure 8 is a schematic diagram of the tension on the cable when the evaluation function is used;

[0044] Figure 9 This is a schematic diagram of transporting simulation results when no evaluation function is used;

[0045] Figure 10 This is a schematic diagram of transporting simulation results when using the evaluation function.

[0046] In the accompanying drawings, 11-robotic arm, 111-end, 21-rocker, 31-cable, 41-transported object, 411-fixed point. DETAILED DESCRIPTION

[0047] Figure 2-Figure 10 The following description describes optional embodiments of the present invention to teach those skilled in the art how to implement and reproduce the present invention. In order to teach the technical solution of the present invention, some conventional aspects have been simplified or omitted. Those skilled in the art will understand that variations or substitutions derived from these embodiments will be within the scope of the present invention. Those skilled in the art will understand that the following features can be combined in various ways to form multiple variations of the present invention. Thus, the present invention is not limited to the following optional embodiments, but is limited only by the claims and their equivalents.

[0048] Example 1.

[0049] The embodiment of the present invention is a transport system based on rocker-controlled cables tightly coupled with multi-manipulator collaboration, such as Figure 2 Shown, including:

[0050] A plurality of multi-degree-of-freedom robotic arms 11 , wherein the distal ends 111 of the robotic arms 11 are respectively connected to a plurality of fixed points 411 of the transported object 41 via a plurality of cables 31 .

[0051] The rocker 21 is used to control the position and posture of the transported object 41 .

[0052] The working parameters of the robot arm 11 and the rocker 21 meet the following requirements:

[0053] , .

[0054] J m is the Jacobian matrix between the joint angles of the robot arm 11 and the position and posture of the end 111 of the robot arm 11, Δθ i is the slight change in the angle of each joint of the robot arm 11, J0 is the Jacobian matrix of the relationship between the slight displacement of the end 111 of each robot arm 11 and the slight variable of the position and posture of the transported object 41, K j is the coefficient of variation between the information of the rocker 21 and the velocity information of each degree of freedom of the transported object 41, X jn is the information output by the joystick 21 when in use, X j0 is the origin information of the joystick 21, t is the sampling time of each joystick 21, k is a constant, I is the unit matrix, H is the evaluation function, xe is the position matrix of the end 111 of the robot arm 11.

[0055] In order to optimize stability, the evaluation function reduces the tension between the individual cables 31 .

[0056] , ,

[0057] .

[0058] N is the number of cables 31, K a is a constant, τ min is the minimum tension in the cables 31, τ th is the reference value, J w It is the Jacobian matrix of the small variables of the position and posture of the transported object 41 and the small variables of the cable 31 vector, τ is the vector of the tension of the cable 31, and F0 is the external force vector acting on the coordinate system of the transported object 41 including gravity, which is obtained by the force sensor installed at the end 111 of each robot arm 11.

[0059] When the constant k is a positive number, the evaluation function is maximized; when the constant k is a negative number, the evaluation function is minimized.

[0060] In order to obtain the Jacobian matrices J0 and J at different positions w , which can be obtained by:

[0061] The length L of the cable 31 is taken as a constant. Total differential, obtain the small displacement Δx of the end 111 of each robot 11 ei The small change ΔX in the position and posture of the object 41 0i The Jacobian matrix J of the relationship between 0。

[0062] The position of the end 111 of the robot arm 11 is taken as a constant. The total differential gives the length change ΔL of the cable 31 and the change ΔX in the position and posture of the transported object 41. 0i The Jacobian matrix J of the relationship between w。

[0063] x ej is the vector of the position of the end 111 of the robot 11 at the jth position, x i is the fixed point 411 vector at position i, L ij is the vector from the j-th robotic arm 11 to the i-th fixed point 411 .

[0064] The control method of the transport system of this embodiment includes the following steps:

[0065] (A1) Obtaining the change ΔX in the position and posture of the conveyed object 41 each time 0i =(X 0target -X 0intital ) / m, and the area S2 of the area enclosed by the lines connecting the multiple fixed points of the cables 31 on the transported object 41, where m is the number of movements of the transported object 41 between the initial and the target.

[0066] X 0intital is the initial position and posture of the object 41, X 0target These are the target position and posture of the conveyed object 41 .

[0067] (A2) Obtaining the small displacement Δx of the end 111 of each robot arm 11 ei The small change ΔX in the position and posture of the object 41 0i The Jacobian matrix J0 of the relationship between the length change ΔL of the cable 31 and the small change ΔX of the position and posture of the transported object 41 0i The Jacobian matrix J of the relationship between w , each cable 31 tension The area S1 of the area enclosed by the line connecting the ends 111 of each robot arm 11, and F0 is the torque acting on the object 41 to be transported.

[0068] (A3) Obtain the change Δx of the position of the end 111 of the robot arm 11 each time e ;

[0069] , and according to the Jacobian matrix J of the robot 11 itself m Get the rotation angle Δθ of the robot arm 11 i , i plus 1; k is a constant, H is the evaluation function, ΔX n is the position and posture of the transported object 41, including the position Δx of the object n and the object's posture Δφ n。

[0070] (A4) Determine whether area S1 is greater than area S2 and whether i is less than m. If the results are yes and no, respectively, proceed to the next step. If the first result is no, interrupt the process. If the second result is yes, return to step (A2).

[0071] Through steps (A1) to (A4), the transported object 41 is changed from an inclined state to a horizontal state, as shown in FIG. Figure 3 As shown in Step 1.

[0072] (A5) X of the object 41 to be transported 0target Set to X now , let n=1, and change x ei Set to x en, and read the real-time position information X of the joystick 21 jn And button information button, confirm that the control button is pressed.

[0073] (A6) Obtaining the small displacement Δx of the end 111 of each robot arm 11 en The small change ΔX in the position and posture of the object 41 0n The Jacobian matrix J0 of the relationship between the length change ΔL of the cable 31 and the small change ΔX of the position and posture of the transported object 41 0n The Jacobian matrix J of the relationship between w , each cable 31 tension The area S of the area enclosed by the line connecting the ends 111 of each robot arm 11 1。

[0074] (A7) According to the real-time position information X of the joystick 21 each time jn、 Jacobian matrix J m and the Jacobian matrix J w , obtain the angle Δθ of each rotation of the robot arm 11 n , ΔX n is the position and posture of the object 41 each time it moves, and the displacement Δx of the end 111 of the robot arm 11 each time en ;

[0075] , .

[0076] (A8) n is incremented by 1 to determine whether area S1 is greater than area S2. If yes, proceed to the next step. If no, stop the process and finally update the position and posture X of the object 41. now =X now +ΔX n , return to step (A7).

[0077] Through steps (A5) to (A8), the horizontal object 41 is gradually moved to the target position. Figure 3 As shown in Step 2.

[0078] Example 2.

[0079] An application example of the handling system and method for tightly coupled multi-robotic arm collaboration based on rocker-controlled cables in Example 1 of the present invention.

[0080] In this application example, if Figure 2 As shown, there are three robotic arms 11 and three fixed points 411 on the transported object 41 , which are evenly distributed on the circumference of the transported object 41 . The end 111 of each robotic arm 11 is connected to two fixed points 411 via two cables 31 .

[0081] like Figure 2 As shown, three fixed points 411 are taken on the transported object 41, and the order is set as i (i=1,2,3). Two cables 31 are fixed to each point, and the other ends of the cables 31 are fixed to the ends 111 of each robot 11, and the order is set as j (j=1,2,3). Then the vector from the end 111 of the j-th robot 11 to the i-th fixed point 411 can be expressed by L ij ∈R 3×1 This handling system can prevent excessive internal forces caused by position errors and can also prevent swaying to a certain extent during the handling process.

[0082] like Figure 4 As shown, the system is kinematically modeled. The position of the transported object 41 is represented by x0∈R 3×1 Represented (based on the coordinates in the world coordinate system), the pose vector represented by the Euler angle is set to φ0∈R 3×1 , according to the geometric relationship, we get the following.

[0083] ,

[0084] , R mj , R0 are and The rotation matrix of .

[0085] Since the length of the cable 31 is known, it can be obtained according to the above formula.

[0086] ;

[0087] x ej is the vector of the position of the end 111 of the robot 11 at the jth position, x i is the fixed point 411 vector at position i, L ij is the vector from the j-th robotic arm 11 to the i-th fixed point 411 .

[0088] like Figure 5 As shown, the small displacement Δx of the end 111 of each robot arm 11 e1 , Δx e2 , Δx e3 The relationship between the small variables Δx0 and ΔΦ0 of the position and posture of the conveyed object 41.

[0089] Will Total differential, get the Jacobian matrix J0∈R 6×9 .

[0090] , .

[0091] The position of the end 111 of the robot arm 11 is taken as a constant, and the length of the cable 31 is taken as a variable.

[0092] Will Total differential, get the Jacobian matrix J w .

[0093] ,

[0094] .

[0095] like Figure 6 As shown, the tilt of each axis of the three-degree-of-freedom rocker 21 determines the speed of each degree of freedom of the transported object 41. When controlling the position of the transported object 41, the front, back, left, right and rotation changes of the rocker 21 are used as X j , Y j , Z j In addition, when controlling the posture of the object 41, the front, back, left, right and rotation changes of the rocker 21 are used as Roll j (around the Z axis), Pitch j (around the Y axis), Yaw j Speed ​​in the direction (around the X axis).

[0096] The position and posture change instructions of the object 41 obtained from the joystick 21 can be obtained by the following formula:

[0097] , K j =(K jX ,K jY ,K jZ ,K φ ) is the coefficient of variation between the rocker 21 information and the velocity information of each degree of freedom of the transported object 41, X jn =(X jn ,Y jn ,Z jn ,φ jn ) T The information output by the joystick 21 when in use; j0 =(X j0 ,Y j0 ,Z j0 ,φ j0 ) T is the origin information of the joystick 21, ΔX n =(X xn ,Y yn ,Z zn ,φ n ) T is the minute change of the transported object 41 at each sampling point under speed control; and t is the sampling time for each time.

[0098] In this handling system, there are 12 redundant degrees of freedom, three of which relate to the position of the end 111 of the robot arm 11, and the remaining nine relate to the posture of the robot arm 11. Assuming the evaluation function for the tension of the cable 31 is H, the optimal motion of the end 111 of the robot arm 11 in this system can be derived from the following equation:

[0099] , .

[0100] J0 + ∈R 9×6 When the constant k is a positive number, the evaluation function can be maximized; when it is a negative number, the evaluation function is minimized.

[0101] After the target position and posture are achieved within each sampling time t, a Calculate the optimal solution of .

[0102] In order to keep the tension of the cable 31 below a certain value, a penalty function is introduced into the term of the minimum tension in the evaluation function H, K a As a constant to maintain the minimum tension of the cable 31, when the tension value is close to τ min and lower than the reference value τ th When the tension value is greater than τ min And when the value becomes larger, the sum of the squares of the tension decreases.

[0103] , .

[0104] When the tension τ i When it is lower than the reference value, it is possible to prevent the position of the end 111 of the robot arm 11 from changing drastically.

[0105] exist In the example, when the constant k=0, the evaluation function H is not used, and the calculated result is the minimum norm solution; when k≠0, the evaluation function H that reduces the sum of the squares of the tension of the cable 31 is used to calculate the optimal solution for the position of the end 111 of the robot arm 11.

[0106] Assume that τ th = 4.3N, τ min = 3.9N,K a = 1, compare the two results, such as Figure 7 (k=0) and Figure 8 According to the test results, when the sum of the squares of the tension of the cable 31 is used as the evaluation function, the value of the tension of the cable 31 will become smaller and the system will become stable.

[0107] By inputting the position and posture of the transported object 41 through the joystick 21, the joint parameters of the three robotic arms 11 can be calculated, thereby achieving the purpose of collaborative transport.

[0108] ,

[0109] .

[0110] The control method of the transport system of this embodiment includes the following steps:

[0111] (A1) Obtaining the change ΔX in the position and posture of the conveyed object 41 each time 0i =(X 0target -X 0intital ) / m, and the area S2 of the area enclosed by the lines connecting the multiple cable 31 fixing points 411 on the transported object 41, where m is the number of movements of the transported object 41 between the initial and target positions.

[0112] X 0intital is the initial position and posture of the object 41, X 0target These are the target position and posture of the conveyed object 41 .

[0113] (A2) Obtaining the small displacement Δx of the end 111 of each robot arm 11 ei The small change ΔX in the position and posture of the object 41 0i The Jacobian matrix J0 of the relationship between the length change ΔL of the cable 31 and the small change ΔX of the position and posture of the transported object 41 0i The Jacobian matrix J of the relationship between w , each cable 31 tension The area S1 of the area enclosed by the line connecting the ends 111 of each robot arm 11, and F0 is the torque acting on the object 41 to be transported.

[0114] (A3) Obtain the change Δx of the position of the end 111 of the robot arm 11 each time e ;

[0115] , and according to the Jacobian matrix J of the robot 11 itself m Get the rotation angle Δθ of the robot arm 11 i , i plus 1; k is a constant, H is the evaluation function, ΔX n is the position and posture of the transported object 41, including the position Δx of the object n and the object's posture Δφ n。

[0116] (A4) Determine whether area S1 is greater than area S2 and whether i is less than m. If the results are yes and no, respectively, proceed to the next step. If the first result is no, interrupt the process. If the second result is yes, return to step (A2).

[0117] Through steps (A1) to (A4), the transported object 41 is changed from an inclined state to a horizontal state, as shown in FIG. Figure 3 As shown in Step 1.

[0118] (A5) X of the object 41 to be transported 0target Set to X now , let n=1, and change x ei Set to x en , and read the real-time position information X of the joystick 21 jn And button information button, confirm that the control button is pressed.

[0119] (A6) Obtaining the small displacement Δx of the end 111 of each robot arm 11 en The small change ΔX in the position and posture of the object 41 0n The Jacobian matrix J0 of the relationship between the length change ΔL of the cable 31 and the small change ΔX of the position and posture of the transported object 41 0n The Jacobian matrix J of the relationship between w , each cable 31 tension The area S of the area enclosed by the line connecting the ends 111 of each robot arm 11 1。

[0120] (A7) According to the real-time position information X of the joystick 21 each time jn、 Jacobian matrix J m and the Jacobian matrix J w , obtain the angle Δθ of each rotation of the robot arm 11 n , ΔX n is the position and posture of the object 41 each time it moves, and the displacement Δx of the end 111 of the robot arm 11 each time en。

[0121] ,

[0122] .

[0123] (A8) n is incremented by 1 to determine whether area S1 is greater than area S2. If yes, proceed to the next step. If no, stop the process and finally update the position and posture X of the object 41. now =X now +ΔX n , return to step (A7).

[0124] Through steps (A5) to (A8), the horizontal object 41 is gradually moved to the target position. Figure 3 As shown in Step 2.

[0125] Verification example.

[0126] The joystick 21 is a THRUSTMASTER 3-DOF TCA joystick 21; the robotic arm 11 is a Yaskawa Electric 6-DOF MOTOMAN-HP3J robotic arm 11; the initial position of each robotic arm 11 is measured using the HAS-U1 produced by DITECT (measurement accuracy ≤ 0.3mm); the 6-axis force sensor at the end 111 of each robotic arm 11 is the PFS055YA501U6 produced by LEPTRINO (resolution ±0.05N); the IMU sensor on the transported object 41 is the 3DM-GX5-25 produced by LORD; the transported object 41 is a 2.5 kg cylinder (φ165 mm x 200 mm).

[0127] 1. Object Level Experiment (SETP1)

[0128] Table 1 shows the target and experimental poses of the object 41 under the minimum norm solution (k = 0) and using the evaluation function (k = -0.5). The results in this table show that both the minimum norm solution and the control method proposed in this invention can transform the object 41 from its initial tilted state to a horizontal position suitable for handling. When k = 0, the maximum position error is 2.5 mm, and the maximum attitude error is 0.01 rad. When k = -0.5, the maximum position error is 4.7 mm, and the maximum attitude error is 0.05 rad. This is due to the initial position and attitude errors of the end 111 of the robot arm 11. When the evaluation function is used, continuous iterative calculations are required, and errors accumulate.

[0129] Table 1: Object pose results of Setp1 horizontal experiment.

[0130] .

[0131] Table 2 shows the tension in cable 31 when using the minimum norm solution and the evaluation function in the Setp1 level experiment. The results in the table show that, in both simulation and experimental values, the tension in cable 31 using the evaluation function is lower than when using the minimum norm solution without the evaluation function. This demonstrates the effectiveness of the control method and evaluation function of the present invention, resulting in a more optimized and stable system.

[0132] Table 2: Simulation and experimental results of cable tension in the Setp1 horizontal experiment.

[0133] .

[0134] 2. Position Control Experiment (SETP2)

[0135] Table 3 shows the data of the position control experiment of the transported object 41 using the rocker 21. According to the experimental results, when k=0, the maximum position error is 3.4mm; when k=-0.5, it is 3.8mm. The error is slightly increased, which is consistent with the reason in the horizontal experiment. It is considered to be caused by the initial position and posture errors of the end 111 of the robot arm 11.

[0136] Table 3: Object pose results of Setp2 position control experiment.

[0137] .

[0138] Table 4 shows the cable tension when using the minimum norm solution and evaluation function in the Setp2 position control experiment. The results in this table show that, in both simulation and experimental values, the tension in cable 31 using the evaluation function is lower than when using the minimum norm solution without the evaluation function. This demonstrates the effectiveness of the control method and evaluation function of the present invention when position-controlling the transported object 41 using the rocker 21, resulting in a more optimized and stable system.

[0139] Table 4: Simulation and experimental results of cable tension control at Setp2 position.

[0140] .

[0141] 3. Posture Control Experiment (SETP2)

[0142] like Figure 9 As shown, during the operation without using the evaluation function (minimum norm solution), the movement of the robot arm 11 is relatively simple, close to a singular posture, the movement becomes unstable, the position where the system finally stops is quite different from the position obtained by simulation, and the transported object 41 also loses its horizontal posture.

[0143] like Figure 10 As shown, when using the evaluation function (minimum norm solution), the object's motion stabilizes, and its final resting position is consistent with the simulation results. This demonstrates that using the evaluation function as feedback optimization is essential when controlling the posture of the transported object 41 using the joystick 21.

[0144] Table 5 shows the experimental results of using the joystick 21 to control the posture of the transported object 41 in Setp2. According to the results, when k = 0, the actual posture of the transported object 41 is significantly different from the simulation value and the target value, verifying that the posture of the transported object 41 cannot be controlled without using the evaluation function of redundant degrees of freedom. When k = -0.5, the maximum position deviation is 3.8mm, and the pitch error is reduced from 0.17rad to 0.02rad, which shows the effectiveness of this method.

[0145] Table 5: Object pose results of Setp2 posture control experiment.

[0146] .

[0147] Table 6 shows the tension in cable 31 when using the minimum norm solution and evaluation function in the posture control experiment using the joystick 21 at Setp2. The results in the table show that in both simulation and experimental values, the tension in cable 31 using the evaluation function is lower than that obtained using the minimum norm solution without the evaluation function. This demonstrates the effectiveness of the control method and evaluation function of the present invention when using the joystick 21 to positionally control the transported object 41, resulting in a more optimized and stable system.

[0148] Table 6 Simulation and experimental results of cable tension for Setp2 posture control.

[0149] .

[0150] In summary, the experimental results verify the feasibility of the control method proposed in the present invention. It is possible to calculate the joint parameters of the three robotic arms 11 by simply controlling the rocker 21 and thus controlling the position of the transported object 41, thereby achieving the purpose of collaborative transport. After using the evaluation function calculated from the redundant degrees of freedom, the tension between the cables 31 of the system is reduced, thereby optimizing the stability of the system.

Claims

1. A control method implemented by a handling system based on rocker-controlled cables tightly coupled with multi-manipulator collaboration, the handling system comprising: Multiple multi-degree-of-freedom robotic arms, the ends of which are connected to multiple fixed points of the transported objects via multiple cables; A joystick, which is used to control the position and posture of the object being transported; The working parameters of the robot arm and rocker meet the requirements. ， ； J m is the Jacobian matrix between the robot arm joint angle and the position and posture of the robot end, Δθ i is the slight change in the angle of each joint of the robot arm, J0 is the Jacobian matrix of the relationship between the slight displacement of each robot end and the slight variable of the position and posture of the transported object, , K j is the coefficient of variation between the joystick information and the velocity information of each degree of freedom of the transported object, X jn The information output when the joystick is in use, X j0 is the origin information of the joystick, t is the sampling time of each joystick, k is a constant, I is the unit matrix, H is the evaluation function, x e is the position matrix of the end of the robotic arm; It is characterized in that the control method comprises the following steps: (A1) Obtain the change ΔX in the position and posture of the transported object each time 0i , and the area S2 of the area enclosed by the lines connecting multiple cable fixing points on the transported object; (A2) Obtain the Jacobian matrix J0, the cable length change ΔL and the small changes in the position and posture of the transported object ΔX 0i The Jacobian matrix J of the relationship between w , the area S1 of the area enclosed by the lines connecting the ends of each robot arm; (A3) Obtain the change Δx in the end position of the robot arm each time e ; (A4) Determine whether area S1 is greater than area S2 and whether i is less than m. If the results are yes and no, respectively, proceed to the next step. If the first result is no, interrupt the process. If the second result is yes, return to step (A2). (A5) X of the object to be transported 0target Set to X now , let n=1, and change x ei Set to x en , and read the real-time position information of the joystick X jn And button information button, confirm that the control button is pressed; (A6) Using the method of step (A2), obtain the area S1 of the area enclosed by the line connecting the ends of each robot arm; (A7) According to the real-time position information of each joystick X jn , Jacobian matrix J m and the Jacobian matrix J w , get the angle Δθ of each rotation of the robot arm n ; (A8) n is incremented by 1 to determine whether area S1 is greater than area S2. If yes, proceed to the next step. If no, stop the process and finally update the position and posture X of the object being transported. now =X now +ΔX n , return to step (A7).

2. The control method according to claim 1, characterized in that: Jacobian matrices J0 and J w The way to obtain is: Taking the cable length L as a constant, Full differential to obtain the small displacement Δx at the end of each robotic arm ei The slight change ΔX in the position and posture of the object being transported 0i The Jacobian matrix J0 of the relationship between them; The position of the end of the robot arm is taken as a constant. Total differential, we get the change in cable length ΔL and the change in position and posture of the transported object ΔX 0i The Jacobian matrix J of the relationship between w ; x ej is the vector of the end position of the robotic arm at position j, x i is the fixed point vector at position i, L ij is the vector from the j-th robotic arm to the i-th fixed point.

3. The control method according to claim 1, characterized in that: The torque F0 is obtained using the force sensors at the end of each robotic arm.

4. The control method according to claim 1, wherein: The evaluation function reduces the tension between the cables; ， ， ； N is the number of cables, K a is a constant, τ min is the minimum tension in the cables, τ th is the reference value, J w is the Jacobian matrix of the small variables of the position and posture of the transported object and the small variables of the cable vector, τ is the cable tension vector, and F0 is the external force vector acting on the coordinate system of the transported object including gravity.

5. The control method according to claim 4, characterized in that: When the constant k is a positive number, the evaluation function is maximized; when the constant k is a negative number, the evaluation function is minimized.

6. The control method according to claim 1, characterized in that: There are three robotic arms and three fixed points, which are evenly distributed on the circumference of the transported object. The end of each robotic arm is connected to two fixed points via two cables.

Images