A virtual decomposition control method for flexible hydraulic manipulator considering nonlinear deformation
By performing multi-joint cutting and virtual power flow processing on the flexible hydraulic robot arm, combined with deformation curve and arc deformation theory, the problem of high coupling and deformation measurement costs of hydraulic robot arm is solved, and high-precision attitude control and low cost are achieved The deformation compensation improves the positioning accuracy and stability of the robot arm.
Patent Information
- Application Number
- CN202410404531.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-07
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-04-07
AI Technical Summary
The hydraulic actuator and joint of the flexible hydraulic robot arm form a closed chain structure, resulting in the linear actuator output force and the weight of the arm, affecting the precise positioning and motion control accuracy, and the existing methods require multiple sensors, resulting in high computational costs.
By cutting, decoupling joints, establishing a multi-module subsystem, introducing virtual power flow, combining the deformation curve and arc deformation theory of the arm frame, calculating the deformation variable and compensation amount in real time, real-time, and real-time, the closed-chain structure is realized by realizing the structure of the closed-chain structure by combining the deformation curve of the arm and the arc deformation theory. decoupling and deformation compensation.
It improves the positioning accuracy and motion stability of the robotic arm, reduces costs, and realizes low-cost deformation compensation and high-precision attitude control.
Smart Images

Figure CN118123833B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of flexible hydraulic mechanical arm control, and in particular to a flexible hydraulic mechanical arm virtual decomposition control method considering nonlinear deformation. Background Art
[0002] The flexible hydraulic robot arm is a highly flexible and adaptable robot arm system, which consists of multiple links and joints and is driven by hydraulics to achieve movement. Compared with traditional rigid robot arms, the flexible hydraulic robot arm has stronger adaptability to complex working environments.
[0003] To ensure stability and accuracy in complex environments, flexible hydraulic robotic arms also have higher requirements for precision. Through high-precision position control, flexible hydraulic robotic arms can achieve precise positioning and movement, thus demonstrating excellent performance. Accurate position control can not only improve production efficiency, but also reduce personal injury and error rates.
[0004] However, since the hydraulic actuator and joint of the flexible hydraulic manipulator form a closed chain structure, the linear actuator output force and the arm weight are coupled, so the manipulator may be affected by the coupling when performing precise positioning and motion control. At present, there are authorized hydraulic manipulator control methods, such as patent publication number CN114536338, entitled a control method for a hydraulic manipulator, but this control method does not make a detailed analysis of the closed chain structure unique to the hydraulic manipulator, and cannot achieve high-precision control of the hydraulic manipulator. Therefore, a control method for decoupling the closed chain joints is urgently needed.
[0005] At the same time, the deformation of the connecting rod and the joint may cause problems such as position error, posture deviation and motion instability, affecting the accuracy and performance of the robot arm. At present, there are authorized methods for determining the deformation of the boom, such as the patent publication number CN114777726, entitled Method, Device, Boom System and Engineering Machinery Equipment for Determining the Deformation of the Boom, but this method requires multiple displacement sensors, pressure sensors and inclination sensors, and the calculation cost is high. A low-cost deformation compensation method is urgently needed.
[0006] In order to solve the problem of high cost of coupling and arm deformation measurement of hydraulic manipulator, a virtual decomposition control method of flexible hydraulic manipulator considering nonlinear deformation is proposed. It aims to achieve precise posture control and motion tracking by sensing and analyzing the deformation of the manipulator and taking corresponding control strategies. Summary of the invention
[0007] To this end, the present invention discloses a virtual decomposition control method for a flexible hydraulic mechanical arm considering nonlinear deformation, the method comprising: firstly, cutting a multi-joint flexible hydraulic mechanical arm system, dividing the system into multi-module subsystems, realizing arm decoupling between multiple joints, and then performing secondary cutting on the rotation closed chain obtained by cutting, dividing the hydraulic cylinder and the arm into a drive chain and a non-drive chain, realizing decoupling of the closed chain structure, and ensuring the overall stability through virtual power flow between each subsystem; finally, calculating the deformation amount of each point on the arm during the movement from the arm deformation curve, and at the same time, according to the circular arc deformation theory, obtaining the relationship between the deformation amount and the rotation angle between adjacent coordinate systems, and calculating the position offset and compensation amount caused by the deformation of the flexible hydraulic mechanical arm, realizing real-time deformation compensation, and improving the terminal positioning accuracy.
[0008] A virtual decomposition control method for a flexible hydraulic mechanical arm considering nonlinear deformation, characterized by comprising the following steps:
[0009] Step 1: To address the problems of strong coupling, weak stiffness and inaccurate positioning of the flexible hydraulic manipulator, the flexible hydraulic manipulator system is cut to decouple the multiple joints and establish a kinematic transfer model of the multi-joint subsystem. At the same time, the rotating closed chain joints are cut again to obtain the driven open chain and the non-driven open chain, and the transfer models are established respectively to obtain the kinematic transfer matrix U of the system. The end load force is then reversely transferred to obtain the resultant force / torque vector F required for each sub-coordinate system. r size;
[0010] Step 2: Through virtual power flow (VPF), dynamic interaction between subsystems is achieved while ensuring the virtual stability of the overall system;
[0011] Step 3: Measure the inlet and outlet pressures p through the pressure sensors at both ends of the hydraulic cylinder. a and p b , find out the magnitude of the hydraulic cylinder's external output force F, simplify the gravity center of each subsystem and apply it at the center of mass, and get the joint driving torque τ;
[0012] Step 4, simplify each boom section into a cantilever beam model with combined compression and bending deformation, the posture of which is related to the joint angle θ, and obtain the deformation curve equation of each boom, where the joint angle θ is measured by an angle encoder;
[0013] Step 5, calculating the deformation amount of each point on the boom according to the joint driving torque and the boom deformation curve equation of step 3 and step 4;
[0014] Step 6, according to the arm deformation in step 5, the linear distance between the origins of the adjacent coordinate systems after the arm deformation is calculated in combination with the arc deformation curve equation, and the end coordinates of the multi-joint robot arm are obtained to achieve kinematic compensation;
[0015] Step 7: modify the kinematic transfer matrix U of step 1 according to the arm deformation obtained in step 5 and the end coordinates obtained in step 6, and then modify the dynamics required force F r Compensation is performed to ultimately achieve real-time compensation of the end position;
[0016] Furthermore, the kinematic transfer matrix U described in step 1 is specifically as follows, taking the coordinate system {A} and the coordinate system {B} as an example:
[0017]
[0018] in, A R B is the rotation matrix from coordinate system {A} to coordinate system {B}, ( A r AB ×) is a skew-symmetric matrix operator, r x , r y and r z Respectively represent the components of the distance between the origin of coordinate system {A} and the origin of coordinate system {B} on the x, y, and z axes of coordinate system {A}.
[0019] A virtual decomposition control method for a flexible hydraulic manipulator considering nonlinear deformation is characterized in that the resultant force vectors of each sub-coordinate system described in step 1, taking subsystem {A} as an example:
[0020]
[0021] Among them C A ( A w)∈R 6×6 is the matrix of Coriolis terms and centrifugal terms, G A ∈R 6 is the gravity term of the rigid body, M A ∈R 6×6 is the mass matrix of the subsystem, A F is the force vector of subsystem {A}.
[0022] Furthermore, the virtual power flow described in step 2 ensures the overall virtual stability:
[0023]
[0024] where we select a single non-negative adjoint function v α for:
[0025]
[0026] in, α V r is the velocity vector required for the sub-coordinate system {α}.
[0027] Let the non-negative adjoint functions of the rotation segment and the prism segment be:
[0028]
[0029] Combining the dynamic equations of the rotating section and the prismatic section, the time derivative is:
[0030]
[0031] Therefore, it indicates the virtual stability of the j-th robot structure and thus the virtual stability of the entire robot.
[0032] Furthermore, the hydraulic cylinder described in step 3 outputs a force F i The size is:
[0033] F=p a A a -p b A b
[0034] Among them, p a and p b are the pressures at the inlet and outlet of the hydraulic cylinder measured by the pressure sensor, A a and A b are the cross-sectional areas of the rod chamber and rodless chamber of the hydraulic cylinder respectively.
[0035] Furthermore, in step 3, the center of gravity is simplified to the center of mass, and the joint driving torque τ is obtained as:
[0036] τ=(p a A a -p b A b )r
[0037] Where r is the length of the lever arm.
[0038] Furthermore, the arm deformation curve equation related to the joint angle in step 4 is:
[0039]
[0040] Where x refers to any distance on the boom, w is the deformation of the corresponding point on the boom, ρ is the equivalent density of the boom, and A is the equivalent cross-sectional area. The real-time deformation of the boom can be calculated by combining the joint driving torque and deformation curve equation obtained in steps 3 and 4.
[0041] Furthermore, the arc deformation curve mentioned in step 6 is:
[0042]
[0043] Among them, αi is the arc angle extracted from the matrix, and the radius of curvature is:
[0044]
[0045] The geometric cosine theorem gives:
[0046]
[0047] in, is the arm length, L i is the distance between adjacent coordinate systems after the arm is deformed, and θ is measured by the angle encoder fixed at the head end of each arm. The position coordinates of the end of the flexible manipulator relative to the base can be obtained by obtaining the length of each arm after deformation:
[0048]
[0049] Furthermore, in step 7, the obtained arm deformation and end coordinates are used to modify the kinematic transfer matrix U, and then the dynamics required force F r Compensation is performed to achieve real-time compensation of the end position. Take the coordinate system {A} and the coordinate system {B} as an example:
[0050]
[0051] After compensating for the dynamic force / torque:
[0052] A F r = B F n + A U B B F r
[0053] in, B F n is the net force of the {B} subsystem itself, and the load force / torque of the end coordinate system {T} becomes:
[0054]
[0055] in, is the end angle, The top slope is represented as follows:
[0056]
[0057] Beneficial effects of the present invention:
[0058] (1) The present invention proposes a virtual decomposition control method, which obtains a multi-module subsystem by cutting the arm frame and joints of a multi-joint robotic arm, thereby reducing the coupling between the robotic arm rods and joints and improving the positioning accuracy; at the same time, a virtual power flow is introduced to ensure overall stability.
[0059] (2) The present invention proposes a virtual decomposition control method for a multi-joint hydraulic robotic arm taking deformation compensation into consideration, which uses the arm deformation curve equation and arc deformation theory to calculate the deformation amount and offset angle in real time, and further improves the end control accuracy based on the virtual decomposition control.
[0060] (3) The flexible hydraulic manipulator arm deformation compensation method proposed in the present invention measures the joint rotation angle through the angle encoder at the manipulator arm joint, and then obtains the deformation amount at each position in combination with the arm frame deformation equation, thereby realizing arm frame deformation compensation under low-cost conditions. BRIEF DESCRIPTION OF THE DRAWINGS
[0061] Figure 1 Virtual decomposition flow chart for flexible hydraulic manipulator;
[0062] Figure 2 This is the deformation compensation flow chart of the flexible hydraulic manipulator;
[0063] Figure 3 This is a schematic diagram of the arc deformation of the flexible hydraulic manipulator;
[0064] Figure 4 Schematic diagram of the motion deformation of the flexible hydraulic robotic arm. Specific implementation methods
[0065] In order to better understand the present invention, the content of the present invention is further explained below in conjunction with the drawings and embodiments, but the content of the present invention is not limited to the following embodiments.
[0066] Example 1
[0067] See also Figure 1 This embodiment proposes a virtual decomposition control method for a flexible hydraulic mechanical arm considering nonlinear deformation, and the specific implementation steps are as follows:
[0068] Step 1: To address the problems of strong coupling, weak stiffness and inaccurate positioning of the flexible hydraulic manipulator, the flexible hydraulic manipulator system is cut to decouple the multiple joints and establish a kinematic transfer model of the multi-joint subsystem. At the same time, the rotating closed chain joints are cut again to obtain the driven open chain and the non-driven open chain, and the transfer models are established respectively to obtain the kinematic transfer matrix U of the system. The end load force is then reversely transferred to obtain the resultant force / torque vector F required for each sub-coordinate system. r size;
[0069] The kinematic transfer matrix U is specifically as follows, taking the coordinate system {A} and the coordinate system {B} as an example:
[0070]
[0071] in, A R B is the rotation matrix from coordinate system {A} to coordinate system {B}, ( A r AB ×) is a skew-symmetric matrix operator, expressed as:
[0072]
[0073] Among them, r x ,t y and r z They are the components of the distance between the origin of coordinate system {A} and the origin of coordinate system {B} on the x, y, and z axes of coordinate system {A}, respectively.
[0074] The resultant force / torque vector of each sub-coordinate system, taking the {A} subsystem as an example:
[0075]
[0076] Among them C A ( A w)∈R 6×6 is the matrix of Coriolis terms and centrifugal terms, G A ∈R 6 is the gravity term of the rigid body, M A ∈R 6×6 is the mass matrix of the subsystem. In addition, the force at the closed chain driving point {T} T F∈R 6 for:
[0077]
[0078] in, and They are the driving forces for open chains 1 and 2 respectively:
[0079]
[0080] Keep α1+α2=1, and T η=[ T η fx T η fy 0 0 0 T η mz ] T is the internal force vector, T η fx , T η fy and T η mzare the components in the x, y and z directions respectively.
[0081] Step 2, through virtual power flow (VPF), dynamic interaction between subsystems is achieved while ensuring the virtual stability of the overall system.
[0082] The virtual stability described is equivalent to the virtual stability of each robot module:
[0083]
[0084] where we select a single non-negative adjoint function v α for:
[0085]
[0086] in, α V r is the required velocity direction of the coordinate system {α}.
[0087] Let the non-negative adjoint functions of the rotation segment and the prismatic segment be:
[0088]
[0089] Combining the dynamic equations of the rotating section and the prismatic section, the time derivative is:
[0090]
[0091] Therefore, it implies the virtual stability of the j-th robot structure and thus the virtual stability of the entire robot.
[0092] Step 3: Measure the inlet and outlet pressures p by installing pressure sensors at both ends of the hydraulic cylinder. a and p b , find the magnitude of the external output force F of the hydraulic cylinder, simplify the center of gravity to be applied at the center of mass, and obtain the joint driving torque τ;
[0093] The magnitude of the external output force F of the described hydraulic cylinder is:
[0094] F=p a A a -p b A b
[0095] Among them, p a and p b A is the pressure at both ends of the hydraulic cylinder measured by the pressure sensor. a and A b It is the cross-sectional area of the rod chamber and the rodless chamber of the hydraulic cylinder.
[0096] Simplifying the center of gravity to the center of mass, the joint driving torque τ of each joint manipulator is obtained as:
[0097] τ1=(p a1 A a1 -p b1 A b1 )r1
[0098] τ2=(p a2 A a2 -p b2 A b2 )r2
[0099] where r1, r2 and r3 are the arm lengths of each arm respectively.
[0100] Step 4, simplify each arm segment into a cantilever beam model with combined compression and bending deformation, whose posture is related to the joint angle θ, and calculate the deformation curve equation of each arm, where the joint angle θ is measured by the angle encoder.
[0101] The deformation curve equation of the two arms related to the joint angle is:
[0102]
[0103] Where x refers to any distance on the boom, w is the deformation variable of the corresponding point on the boom, ρ is the equivalent density of the boom, and A is the equivalent cross-sectional area.
[0104] Step 5, calculating the deformation amount of each position of the boom according to the joint driving torque and the boom deformation curve equation obtained in steps 3 and 4.
[0105] Step 6: According to the arm deformation obtained in step 5, the straight-line distance between adjacent coordinate systems after deformation can be calculated in combination with the arc deformation curve equation, thereby obtaining the end coordinates of the multi-joint robot arm;
[0106] The arc deformation curve is mentioned as:
[0107]
[0108] Among them, α i is the arc angle extracted from the matrix, and the radius of curvature is:
[0109]
[0110] The geometric cosine theorem gives:
[0111]
[0112] in, is the arm length, L iis the distance between adjacent coordinate systems after the arm is deformed, and U is measured by the angle encoder fixed at the head end of each arm. The position coordinates of the end of the flexible robotic arm relative to the base can be obtained by obtaining the length of each arm after deformation, which is:
[0113] X=L1cos(θ1)+v1sin(θ1)+L2cos(U1+U2)+v2sin(θ1+θ2)
[0114] +L3cos(θ1+θ2+θ3)+v2sin(θ1+θ2+θ3)
[0115] Y=L1sin(θ1)+v1cos(θ1)+L2sin(θ1+θ2)
[0116] +v2cos(θ1+θ2)+L3sin(θ1+θ2+θ3)+v2cos(θ1+θ2+θ3)
[0117] Step 7, based on the arm deformation obtained in step 5 and the end coordinates obtained in step 6, the kinematic transfer matrix U of step 1 is modified, and then the dynamic force F is calculated. r Compensation can be carried out and finally the real-time compensation of the end position can be realized.
[0118] Substitute the obtained arm deformation into the kinematic transfer matrix U to obtain the real-time changing U, thus realizing kinematic deformation compensation:
[0119]
[0120] Force / torque F for dynamics r After compensation:
[0121] A F r = B F n + A U B B F r
[0122] in, B F n is the net force of the {B} subsystem itself, and the load force / torque of the end coordinate system {T} becomes:
[0123]
[0124] in, is the end angle, The top slope is represented as follows:
[0125]
[0126] Therefore, the kinematics and dynamics of the virtual decomposition control method are modified according to the arm deformation, and the arm deformation compensation is realized.
Claims
1. A virtual decomposition control method for a flexible hydraulic manipulator considering nonlinear deformation, characterized in that: The following steps are involved: Step 1: To address the problems of strong coupling, weak stiffness and inaccurate positioning of the flexible hydraulic manipulator, the multi-joint flexible hydraulic manipulator system is cut to decouple the multiple joints and establish a kinematic transfer model of the multi-joint subsystem. At the same time, the rotating closed chain joints are cut again to obtain the driving open chain and the driven open chain, and the transfer models are established respectively to obtain the kinematic transfer matrix U of the system. The end load force is then reversely transferred to obtain the resultant force vector F required for each subsystem. r size; Step 2: Through the virtual power flow VPF, dynamic interaction is achieved between the subsystems while ensuring the virtual stability of the overall system; Step 3: Measure the inlet and outlet pressures p through the pressure sensors at both ends of the hydraulic cylinder. a and p b , find out the magnitude of the hydraulic cylinder's external output force F, simplify the gravity center of each subsystem and apply it at the center of mass, and get the joint driving torque τ; Step 4, simplify each arm section into a cantilever beam model with combined compression and bending deformation, whose posture is related to the joint angle θ, and find the arm deformation curve equation of each arm section, where the joint angle is measured by an angle encoder; Step 5, calculating the deformation amount of each point on the boom according to the joint driving torque and the boom deformation curve equation of step 3 and step 4; Step 6, according to the arm deformation in step 5, the linear distance between the origins of the adjacent coordinate systems after the arm deformation is calculated in combination with the arc deformation curve equation, and the end coordinates of the multi-joint robot arm are obtained to achieve kinematic compensation; Step 7: modify the kinematic transfer matrix U of step 1 according to the arm deformation obtained in step 5 and the end coordinates obtained in step 6, and then modify the dynamics required force F r Compensation is performed and finally the real-time compensation of the end position is achieved.
2. The method for virtual decomposition control of a flexible hydraulic manipulator considering nonlinear deformation according to claim 1 is characterized in that: The kinematic transfer matrix U described in step 1 is as follows, taking adjacent subsystems A and B as an example: in, A R B is the rotation matrix from subsystem A to subsystem B, ( A r AB ×) is a skew-symmetric matrix operator, r x , r y and r z They represent the components of the distance between the origin of subsystem A and the origin of subsystem B on the x, y, and z axes of subsystem A respectively.
3. The method for virtual decomposition control of a flexible hydraulic manipulator considering nonlinear deformation according to claim 1 is characterized in that: The resultant force vectors of each subsystem described in step 1, taking subsystem A as an example: Among them C A ( A w)∈R 6×6 is the matrix of Coriolis terms and centrifugal terms, G A ∈R 6 is the gravity term of the rigid body, M A ∈R 6×6 is the mass matrix of the system, A F is the force vector of subsystem A.
4. The method for virtual decomposition control of a flexible hydraulic mechanical arm considering nonlinear deformation according to claim 1, characterized in that: The virtual power flow described in step 2 ensures overall virtual stability: where v α is an individual non-negative adjoint function, v pj and v ptj are the non-negative adjoint functions of the rotation segment and the prismatic segment, respectively.
5. The method for virtual decomposition control of a flexible hydraulic manipulator considering nonlinear deformation according to claim 1, characterized in that: The magnitude of the external output force F of the hydraulic cylinder described in step 3 is: F=p a A a -p b A b Among them, p a and p b A is the pressure at both ends of the hydraulic cylinder measured by the pressure sensor. a and A b It is the cross-sectional area of the rod cavity and the rodless cavity.
6. The method for virtual decomposition control of a flexible hydraulic manipulator considering nonlinear deformation according to claim 1, characterized in that: In step 3, the center of gravity is simplified to the center of mass, and the joint driving torque τ is obtained as: τ6(p a A a -p b A b )r Where r is the moment arm of the arm.
7. The method for virtual decomposition control of a flexible hydraulic mechanical arm considering nonlinear deformation according to claim 1, characterized in that: The arm deformation curve equation described in step 4 is: Where x refers to any distance on the boom, w is the deformation at the corresponding point on the boom, ρ is the equivalent density of the boom, and S is the equivalent cross-sectional area.
8. The method for virtual decomposition control of a flexible hydraulic mechanical arm considering nonlinear deformation according to claim 1, characterized in that: In step 5, the real-time deformation amount of the arm can be calculated by combining the joint driving torque and deformation curve equation obtained in steps 3 and 4.
9. The method for virtual decomposition control of a flexible hydraulic mechanical arm considering nonlinear deformation according to claim 1, characterized in that: The arc deformation curve in step 6 is: in, is the length of the boom itself, L i is the distance between adjacent coordinate systems after the arm is deformed, and θ is measured by the angle encoder fixed at the head end of each arm. After calculating the deformation of each arm, the position coordinates of the end of the flexible robotic arm relative to the base can be calculated to achieve kinematic compensation. The position coordinates after compensation are:
10. The method for virtual decomposition control of a flexible hydraulic mechanical arm considering nonlinear deformation according to claim 1, characterized in that: The arm deformation obtained in step 5 and the end coordinates obtained in step 6 are used to modify the kinematic transfer matrix U, thereby compensating for the force required for dynamics, and finally achieving real-time compensation of the end position; taking subsystem A and subsystem B as an example: After compensating for the dynamic force: A F r = B F n + A U B B F r in, B F n is the net force of subsystem B itself, and the load force / torque of the terminal subsystem T becomes: in, is the end angle, The top slope is represented as follows:
