Cooperative motion control method for quadruped robot with mechanical arm

By using single rigid body dynamics model and overall kinematics model, the collaborative motion control of quadruple-leg robots is simplified, the complexity and cost of the multi-rigid body whole-body dynamics model are solved, and the stability of the centroid of the quadruple-leg robot and the mechanical arm are realized. High-precision control at the end.

CN118963119APending Publication Date: 2024-11-15CHINA ACAD OF SPACE TECH HANGZHOU CENT
View PDF 0 Cites 4 Cited by

Patent Information

Application Number
CN202411008263.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

In the existing collaborative motion control method of four-legged robots with robotic arms, the multi-rigid body whole-body dynamic model is complex in construction and difficult to model, requires a lot of known information and large amount of computing, and most robotic arms on the market only provide position control mode. , high price and difficult to develop the entire machine.

Method used

Using single rigid body dynamics model and overall kinematics model, design the model prediction matrix for n times in the future, construct variable weight cost functions and constraints, and convert them into quadratic planning problems, and calculate the foot-end force and machinery of the four-legged robot body The angular velocity of the arm joint simplifies the control model and reduces cost and difficulty.

Benefits of technology

The stability of the center of mass of the four-legged robot and the high-precision and high-rootability coordinated motion control of the end effector of the robot arm are realized, reducing the complexity and cost of the coordinated control of the foot arm, and avoiding the loss of center of mass caused by flexible movement of the robot arm Stability problem.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118963119A_ABST
    Figure CN118963119A_ABST
Patent Text Reader

Abstract

The invention provides a cooperative motion control method for a quadruped robot with a mechanical arm. The control method comprises the steps that S1, a single rigid body dynamic model of a quadruped robot body and an overall kinematics model of the quadruped robot with the mechanical arm are constructed; s2, designing a model prediction matrix of n moments in the future for the single rigid body dynamic model and the overall kinematics model; s3, constructing a variable weight cost function, a quadruped robot body foot end stress condition and a constraint condition for a mechanical arm joint rotating speed; s4, the variable weight cost function and the constraint condition are converted into a quadratic programming problem, and the acting force of the ground on the foot ends of the quadruped robot body and the expected joint angular velocity of the mechanical arm are calculated; and S5, the expected torque of a single-leg joint motor of the quadruped robot body and expected position instructions of all joints of a mechanical arm are calculated. According to the method, a simpler single-rigid-body quadruped robot dynamics model and an overall kinematics model are used, and simplification of a control model is achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application belongs to the field of robotics technology, and in particular relates to a method for controlling the coordinated motion of a quadruped robot with a robotic arm. Background Art

[0002] Compared with traditional wheeled and tracked mobile robots, quadruped robots have better movement ability and higher maneuverability in complex environments, but they do not have the ability to grasp objects. The combination of the robotic arm and the quadruped robot body can effectively improve the flexibility of the quadruped robot, simulate the activities of human arms, and realize grasping, carrying and other complex movements. This flexibility makes the application of quadruped robots in warehousing, logistics and medical fields more extensive. At present, most quadruped robots with robotic arms can only control the quadruped robot and the robotic arm as independent hardware units, without fully considering the mutual influence between the quadruped robot and the robotic arm, especially the influence of the interaction force. The method of independently controlling each unit cannot give full play to the six-degree-of-freedom motion advantage of the legged robot body, and it is very likely to cause the problem of instability of the overall center of gravity of the robot due to the activity of the robotic arm, which will greatly weaken the flexible workspace, accessibility and accuracy of the end of the robotic arm, and even cause the robot to fall in severe cases. Therefore, it is very necessary to optimize the overall control of the quadruped robot body and the robotic arm to achieve high-precision, high-robustness and high-stability collaborative motion control of the end effector of the robotic arm.

[0003] The most influential method for cooperative motion control of a quadruped robot with a manipulator is the Whole-Body Control (WBC) optimization method based on a multi-rigid-body whole-body dynamics model proposed by the Swiss Federal Institute of Technology (ETH). This method first constructs a multi-rigid-body whole-body dynamics model of the manipulator and the quadruped robot body, and then uses the WBC optimization control method to control the torque of the 6 joint motors of the manipulator body and the 12 joint motors of the quadruped robot body. This method constructs a complete multi-rigid-body dynamics model of the manipulator and the quadruped robot, calculates the output torque of 18 motors through the WBC controller, and then uses the torque control mode of each motor to achieve cooperative motion control of the quadruped robot with a manipulator.

[0004] The WBC control method of a quadruped robot with a robotic arm requires the construction of a complete multi-rigid body full-body dynamics model. The quadruped robot with a robotic arm is a super-redundant composite robot with 18 degrees of freedom. The use of multi-rigid body dynamics modeling requires the knowledge of the mass, moment of inertia, length of each rod, relationship between each rod, etc. of the quadruped robot body and each active rod of the robotic arm. Therefore, its overall multi-rigid body full-body dynamics model is complex to construct, difficult to model, requires a lot of known information, and requires a large amount of calculation. In addition, this method requires the use of a robotic arm system that provides a torque control mode. However, most robotic arms on the market can only provide a position control mode, and the price of robotic arms that provide a torque control mode is high, and the development of the whole machine is difficult. Summary of the invention

[0005] The main purpose of the present invention is to provide a collaborative motion control method for a quadruped robot with a mechanical arm to solve the problems in the prior art that the multi-rigid body full-body dynamics model is complex in structure, difficult to model, requires a lot of known information and has a large amount of calculation.

[0006] In order to solve the above problems, according to one aspect of the present invention, a collaborative motion control method of a quadruped robot with a robotic arm is provided, wherein the quadruped robot with a robotic arm comprises a robotic arm and a quadruped robot body, wherein the control method comprises: step S1, constructing a single rigid body dynamics model of the quadruped robot body and an overall kinematic model of the quadruped robot with the robotic arm; step S2, designing a model prediction matrix for the single rigid body dynamics model and the overall kinematic model for the next n moments; step S3, constructing a variable weight cost function, force conditions at the foot end of the quadruped robot body and constraints on the rotation speed of the robotic arm joints; step S4, converting the variable weight cost function and the constraints into a quadratic programming problem, and calculating the force exerted by the ground on the foot end of the quadruped robot body and the expected joint angular velocity of the robotic arm; step S5, calculating the expected torque of the motor of the single leg joint of the quadruped robot body and the expected position instructions of each joint of the robotic arm.

[0007] Furthermore, the above step S1 includes: using a six-dimensional force sensor to detect the force and torque of the robotic arm on the quadruped robot body, constructing a single rigid body dynamics model of the quadruped robot body and an overall kinematic model of the quadruped robot with the robotic arm.

[0008] Furthermore, the single rigid body dynamics model of the quadruped robot body and the overall kinematics model of the quadruped robot with a mechanical arm include:

[0009]

[0010] M0 and G0 are:

[0011]

[0012] in, are the position, Euler angle, linear velocity and angular velocity vector of the quadruped robot’s center of mass in the world system, are the position, Euler angle, linear velocity and angular velocity vector of the robot end effector in the world system, To control the cycle, is a 3×3 identity matrix, is a 3×3 zero matrix, is a 3×6 zero matrix, The rotation matrix of the simplified angular velocity of the quadruped robot body to the Euler angular velocity, is the rotation matrix from the angular velocity of the end effector of the robot arm to the Euler angular velocity, is the mass of the quadruped robot, is the moment of inertia of the quadruped robot body in the world system, is the position vector from the four legs of the quadruped robot to the center of mass of the quadruped robot, are the rotation matrices from the center-of-mass coordinate system of the quadruped robot to the world system and from the base coordinate system to the center-of-mass coordinate system of the quadruped robot, are the position Jacobian matrix and rotation Jacobian matrix of the robot end effector to the robot base, respectively. is the translation vector from the six-dimensional force sensor coordinate system to the quadruped robot body center of mass coordinate system, is the force exerted by the manipulator on the quadruped robot in the six-dimensional force sensor coordinate system, is the rotation matrix from the six-dimensional force sensor coordinate system to the quadruped robot body center of mass coordinate system, is the gravitational acceleration vector, are the position and angular velocity of the robot end effector in the robot base coordinate system, is the translation vector from the manipulator base coordinate system to the quadruped robot body center of mass coordinate system, is the reaction force of the four feet of the quadruped robot body on the ground, and the robotic arm is selected from a six-degree-of-freedom robotic arm. is the angular velocity vector of the six-degree-of-freedom robot joint.

[0013] Furthermore, in the above step S2, the model prediction matrix for the next n moments includes the relationship between the controlled variable and the control variable in the next n steps.

[0014] Furthermore, the model constructed in step S1 is simplified as follows:

[0015]

[0016] in, xi +1 is the controlled variable, ui is the controlled variable; the simplified model is iterated n times, and the following model prediction matrix is ​​designed:

[0017]

[0018] Where n is the calculation step length of model predictive control,

[0019] The prediction model matrix can be simplified as:

[0020]

[0021] in,

[0022] Furthermore, the process of constructing the variable weight cost function includes minimizing the sum of the norm of the difference between the desired pose of the end effector of the manipulator and the current pose of the end effector of the manipulator, the desired center of mass pose of the quadruped robot body and the initial center of mass pose of the quadruped robot body, and the norm of the control quantity:

[0023]

[0024] in, is the desired pose vector, and are adjustable control weights, all of which are diagonal positive definite matrices. The part of the Q weight related to the difference between the expected center of mass pose of the quadruped robot body and the initial center of mass pose of the quadruped robot body is a variable weight setting. The weight of the posture error of the end effector of the manipulator is set much larger than the weight of the center of mass pose error of the quadruped robot body. When the difference between the expected center of mass pose of the quadruped robot body and the initial center of mass pose is small, the Q weight is set to be smaller; when the difference between the expected center of mass pose of the quadruped robot body and the initial center of mass pose of the quadruped robot body is large, the Q weight is set to gradually increase.

[0025] Furthermore, the reaction forces f1, f2, f3, and f4 of the four feet of the quadruped robot body on the ground are subjected to the friction cone, and the angular velocity of the robot arm joint is Construct constraints at maximum speed.

[0026] Furthermore, in the above step S4, converting the variable weight cost function and the constraint conditions into a quadratic programming problem includes:

[0027]

[0028] H and g are:

[0029]

[0030] Among them, H and g are constants, and U is the control variable.

[0031] Furthermore, the quadratic programming solver is used to calculate U, where U is the n-step control variable u i vector, take the first step length u i As the control quantity, we get the reaction forces f1, f2, f3, f4 of the four feet of the quadruped robot body on the ground and the angular velocity vector of the six-degree-of-freedom manipulator joints.

[0032] Furthermore, the Jacobian matrix of each leg of the quadruped robot body and f1, f2, f3, f4 are used to calculate the expected torque of the motor of the single leg joint of the quadruped robot body; the expected angular velocity of each joint of the robotic arm Integrate and calculate the expected position instructions of each joint of the robot arm.

[0033] The scheme of the present invention is applied, and its purpose is to provide a foot-arm collaborative control method and system for a quadruped robot with a robotic arm based on a single rigid body dynamics model, which uses a simpler single rigid body quadruped robot dynamics model and an overall kinematic model to simplify the control model, generate desired position instructions for each joint of the robotic arm, and stabilize the center of mass of the quadruped robot body, so as to solve the problems of complex whole-body dynamics of the existing foot-arm collaborative control methods, such as the need to construct a complex control model, the cost problem caused by the need to use an expensive robotic arm with a torque control mode, and the problem of instability of the center of mass of the quadruped robot body caused by the flexible movement of the robotic arm. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The following will further explain the above characteristics, technical features, advantages and implementation methods of the present application in a clear and understandable manner through the description of the preferred embodiments and in combination with the accompanying drawings. The following drawings are only intended to illustrate and explain the present application, and do not limit the scope of the present application. Among them:

[0035] Figure 1 A structural diagram of a quadruped robot with a mechanical arm provided by an embodiment of the present application is shown;

[0036] Figure 2 A flow chart of a method for coordinated motion control of a quadruped robot with a mechanical arm provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0037] In order to have a clearer understanding of the technical features, purposes and effects of the present application, the specific implementation methods of the present application are now described with reference to the accompanying drawings.

[0038] The technical solution provided by this application solves the problem of complex construction and difficult modeling of multi-rigid body full-body dynamics models.

[0039] like Figure 1As shown, the quadruped robot with a robotic arm includes a robotic arm and a quadruped robot body, and the four sets of robot bodies include four legs, and one end of each of the four legs away from the quadruped robot body is called a foot end.

[0040] According to a typical implementation of the present application, Figure 2 As shown, a method for cooperative motion control of a quadruped robot with a robotic arm is provided, wherein the quadruped robot with a robotic arm comprises a robotic arm and a quadruped robot body, wherein the control method comprises: step S1, constructing a single rigid body dynamics model of the quadruped robot body and an overall kinematics model of the quadruped robot with the robotic arm; step S2, designing a model prediction matrix for the single rigid body dynamics model and the overall kinematics model for the next n moments; step S3, constructing a variable weight cost function, force conditions at the foot end of the quadruped robot body and constraints on the rotation speed of the robotic arm joints; step S4, converting the variable weight cost function and the constraints into a quadratic programming problem, and calculating the force acting on the foot end of the quadruped robot body by the ground and the expected joint angular velocity of the robotic arm; step S5, calculating the expected torque of the motor of the single leg joint of the quadruped robot body and the expected position instructions of each joint of the robotic arm.

[0041] In one embodiment, the above step S1 includes: using a six-dimensional force sensor to detect the force and torque of the mechanical arm on the quadruped robot body, and constructing a single rigid body dynamics model of the quadruped robot body and an overall kinematics model of the quadruped robot with the mechanical arm. The application of the six-dimensional force sensor decouples the mechanical arm and the quadruped robot dynamics, and a simple and practical single rigid body dynamics model of the quadruped robot body can be established separately; using the Jacobian matrix of the mechanical arm, the relationship between the mechanical arm end effector posture and the foot end force of the quadruped robot body, the mechanical arm joint angular velocity and the overall kinematics can be constructed.

[0042] Furthermore, the single rigid body dynamics model of the quadruped robot body and the overall kinematics model of the quadruped robot with a mechanical arm include:

[0043]

[0044] M0 and G0 are:

[0045]

[0046] in, are the position, Euler angle, linear velocity and angular velocity vector of the quadruped robot’s center of mass in the world system, are the position, Euler angle, linear velocity and angular velocity vector of the robot end effector in the world system, To control the cycle, is a 3×3 identity matrix, is a 3×3 zero matrix, is a 3×6 zero matrix, The rotation matrix of the simplified angular velocity of the quadruped robot body to the Euler angular velocity, is the rotation matrix from the angular velocity of the end effector of the robot arm to the Euler angular velocity, is the mass of the quadruped robot, is the moment of inertia of the quadruped robot body in the world system, is the position vector from the four legs of the quadruped robot to the center of mass of the quadruped robot, are the rotation matrices from the center-of-mass coordinate system of the quadruped robot to the world system and from the base coordinate system to the center-of-mass coordinate system of the quadruped robot, are the position Jacobian matrix and rotation Jacobian matrix of the robot end effector to the robot base, respectively. is the translation vector from the six-dimensional force sensor coordinate system to the quadruped robot body center of mass coordinate system, is the force exerted by the manipulator on the quadruped robot in the six-dimensional force sensor coordinate system, is the rotation matrix from the six-dimensional force sensor coordinate system to the quadruped robot body center of mass coordinate system, is the gravitational acceleration vector, are the position and angular velocity of the robot end effector in the robot base coordinate system, is the translation vector from the manipulator base coordinate system to the quadruped robot body center of mass coordinate system, is the reaction force of the four feet of the quadruped robot body on the ground, and the robotic arm is selected from a six-degree-of-freedom robotic arm. is the angular velocity vector of the six-degree-of-freedom robot joint.

[0047] The single rigid body dynamic model constructed above is relatively simple compared to the whole-body dynamic modeling and requires less known information. It only needs to consider the equivalent mass, equivalent moment of inertia, the distance from the center of mass to the four foot ends, the position and velocity of the center of mass of the quadruped robot body, and does not need to know the mass, moment of inertia and position relationship of each joint and connecting rod of the quadruped robot. The linear velocity and angular velocity of the end effector of the manipulator in the world system are obtained through the dynamic relationship between the Jacobian matrix of the manipulator and the center of mass of the quadruped robot body, and are integrated to obtain the position and posture of the end effector of the manipulator in the world system, that is, the kinematics of the center of mass of the quadruped robot and the end effector of the manipulator in the world system.

[0048] The controlled quantities in this model are the position, Euler angle, linear velocity and angular velocity x of the center of mass of the quadruped robot in the inertial system at time i+1 (i.e. the next time) c ,Θ c , ω c , as well as the position, Euler angle, linear velocity and angular velocity x of the end effector of the robot arm in the inertial systeme ,Θ e , ω e ; Control input i (i.e. current moment) the four feet of the quadruped robot exert forces f1, f2, f3, f4 on the ground and the angular velocities of the six joints of the robotic arm

[0049] In order to further improve the robustness and control accuracy of the controlled quantity, preferably in the above step S2, the model prediction matrix for the next n moments includes the relationship between the controlled quantity and the control quantity in the next n steps.

[0050] In one embodiment, the model constructed in step S1 is simplified as follows:

[0051]

[0052] in, x i+1 is the controlled quantity, u i To control the amount; perform n iterations on the simplified model and design the following model prediction matrix:

[0053]

[0054] Among them, n is the calculation step of model predictive control, and the prediction model matrix can be simplified as:

[0055]

[0056] in,

[0057] In one embodiment, the process of constructing the variable weight cost function preferably includes: minimizing the sum of the norm of the difference between the desired pose of the end effector of the manipulator and the current pose of the end effector of the manipulator, the desired center of mass pose of the quadruped robot body and the initial center of mass pose of the quadruped robot body, and the norm of the control quantity:

[0058]

[0059] in, is the desired pose vector, and The Q weight is a variable weight setting for the difference between the expected center of mass pose of the quadruped robot and the initial center of mass pose of the quadruped robot. The weight of the end effector pose error is much larger than the weight of the center of mass pose error of the quadruped robot to further ensure the high-precision pose tracking effect of the end effector. When the difference between the expected center of mass pose of the quadruped robot and the initial center of mass pose is small, the Q weight is set to be small; when the difference between the expected center of mass pose of the quadruped robot and the initial center of mass pose of the quadruped robot is large, the Q weight is set to gradually increase to further ensure the stability of the center of mass pose of the quadruped robot and prevent the center of mass instability problem.

[0060] The variable weight cost function designed in the present application can ensure the high-precision posture tracking effect of the end effector of the manipulator under the premise that the center of gravity of the quadruped robot body is stable: the weight of the posture error of the end effector of the manipulator is much larger than the weight of the posture error of the center of gravity of the quadruped robot body, so as to ensure the high-precision posture tracking effect of the end effector of the manipulator; when the posture error of the center of gravity of the quadruped robot body increases, the weight of the posture error of the center of gravity of the quadruped robot body gradually increases to ensure the stability of the center of gravity of the quadruped robot body.

[0061] In one embodiment, it is preferred that the reaction forces f1, f2, f3, and f4 of the four foot ends of the quadruped robot body on the ground are subjected to the friction cone, and the angular velocity of the robot arm joint is Construct constraints at maximum speed.

[0062] In one embodiment, preferably in the above step S4, converting the variable weight cost function and the constraint condition into a quadratic programming problem includes:

[0063]

[0064] H and g are:

[0065]

[0066] Among them, H and g are constants, and U is the control variable.

[0067] In one embodiment, it is preferred that the quadratic programming solver is used to calculate U, where U is the vector of n step-length control quantities ui, and the first step-length ui is taken as the control quantity to obtain the reaction forces f1, f2, f3, f4 of the four foot ends of the quadruped robot body on the ground and the angular velocity vector of the six-degree-of-freedom manipulator joints

[0068]

[0069] The model predictive controller designed in this application (establishment of dynamic and kinematic models, design of variable weight cost function, design of constraint conditions, and implementation of quadratic programming) realizes torque control of the joint motors of the quadruped robot body and position control of the joint motors of the robotic arm. This method makes it possible to use a cheaper robotic arm with only joint motor position control mode to realize the coordinated motion control of a quadruped robot with a robotic arm, which can effectively reduce the cost and difficulty of realizing the coordinated motion control of the foot and arm.

[0070] In one embodiment, the Jacobian matrix of each leg of the quadruped robot body and f1, f2, f3, f4 are preferably used to calculate the desired torque of the single leg joint motor of the quadruped robot body; the desired angular velocity of each joint of the robotic arm is calculated. Integrate and calculate the expected position instructions of each joint of the robot arm.

[0071] It should be understood that although this specification is described according to various embodiments, not every embodiment contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

[0072] The above description is only an illustrative embodiment of the present application and is not intended to limit the scope of the present application. Any equivalent changes, modifications and combinations made by any technician in the field without departing from the concept and principle of the present application shall fall within the scope of protection of the present application.

Claims

1. A method for controlling the coordinated motion of a quadruped robot with a mechanical arm, wherein the quadruped robot with a mechanical arm comprises a mechanical arm and a quadruped robot body, wherein: The control method comprises: Step S1, constructing a single rigid body dynamics model of the quadruped robot body and an overall kinematics model of the quadruped robot with a mechanical arm; Step S2, designing a model prediction matrix for the single rigid body dynamics model and the overall kinematics model at n moments in the future; Step S3, constructing a variable weight cost function, force conditions at the foot ends of the quadruped robot body, and constraints on the rotation speed of the robot arm joints; Step S4, converting the variable weight cost function and the constraint conditions into a quadratic programming problem, and calculating the ground force acting on the foot end of the quadruped robot body and the expected joint angular velocity of the robot arm; Step S5, calculating the expected torque of the single-leg joint motor of the quadruped robot body and the expected position instructions of each joint of the robotic arm.

2. The method for controlling the coordinated motion of a quadruped robot with a mechanical arm according to claim 1, characterized in that: The step S1 comprises: A six-dimensional force sensor is used to detect the force and torque of the mechanical arm on the quadruped robot body, and a single rigid body dynamics model of the quadruped robot body and an overall kinematic model of the quadruped robot with the mechanical arm are constructed.

3. The method for controlling the coordinated motion of a quadruped robot with a mechanical arm according to claim 2, characterized in that: The single rigid body dynamics model of the quadruped robot body and the overall kinematics model of the quadruped robot with a mechanical arm include: M0 and G0 are: in, are the position, Euler angle, linear velocity and angular velocity vector of the center of mass of the quadruped robot in the world system, are the position, Euler angle, linear velocity and angular velocity vector of the end effector of the robot arm in the world system, To control the cycle, is a 3×3 identity matrix, is a 3×3 zero matrix, is a 3×6 zero matrix, is the rotation matrix of the simplified angular velocity of the quadruped robot body to the Euler angular velocity, is the rotation matrix from the angular velocity of the end effector of the manipulator to the Euler angular velocity, is the mass of the quadruped robot body, is the moment of inertia of the quadruped robot body in the world system, is the position vector from the four foot ends of the quadruped robot body to the mass center of the quadruped robot body, are the rotation matrices from the mass center coordinate system of the quadruped robot body to the world system and from the base coordinate system to the mass center coordinate system of the quadruped robot body, are the position Jacobian matrix and rotation Jacobian matrix of the robot end effector to the robot base, respectively. is the translation vector from the six-dimensional force sensor coordinate system to the quadruped robot body center of mass coordinate system, is the force exerted by the manipulator on the quadruped robot body in the six-dimensional force sensor coordinate system, is the rotation matrix from the six-dimensional force sensor coordinate system to the quadruped robot body center of mass coordinate system, is the gravitational acceleration vector, are the position and angular velocity of the end effector of the robot arm in the coordinate system of the robot arm base, is the translation vector from the manipulator base coordinate system to the quadruped robot body mass center coordinate system, is the reaction force of the four foot ends of the quadruped robot body on the ground, and the mechanical arm is selected from a six-degree-of-freedom mechanical arm, is the angular velocity vector of the six-degree-of-freedom robot joint.

4. The method for controlling the coordinated motion of a quadruped robot with a mechanical arm according to claim 3, characterized in that: In step S2, the model prediction matrix for the next n moments includes the relationship between the controlled variable and the control variable in the next n steps.

5. The method for controlling the coordinated motion of a quadruped robot with a mechanical arm according to claim 4, characterized in that: The model constructed in step S1 is simplified as follows: in, xi +1 is the controlled variable, ui is the control variable; The simplified model is iterated n times and the following model prediction matrix is ​​designed: Where n is the calculation step length of model predictive control, The prediction model matrix can be simplified as: in, 6. The method for controlling the coordinated motion of a quadruped robot with a mechanical arm according to claim 1, wherein in step S3, the process of constructing a variable weight cost function comprises: Minimize the sum of the norm of the difference between the desired pose of the end effector of the manipulator and the current pose of the end effector of the manipulator, the desired center of mass pose of the quadruped robot body and the initial center of mass pose of the quadruped robot body, and the norm of the control quantity: in, is the desired pose vector, and are adjustable control weights, all of which are diagonal positive definite matrices. The Q weight is a variable weight setting for the difference between the expected center of mass posture of the quadruped robot body and the initial center of mass posture of the quadruped robot body. The weight of the posture error of the manipulator end effector is set much larger than the weight of the center of mass posture error of the quadruped robot body. When the difference between the expected center of mass posture of the quadruped robot body and the initial center of mass posture is small, the Q weight is set to be small; when the difference between the expected center of mass posture of the quadruped robot body and the initial center of mass posture of the quadruped robot body is large, the Q weight is set to gradually increase.

7. The method for controlling the coordinated motion of a quadruped robot with a mechanical arm according to claim 6, characterized in that: The reaction forces f1, f2, f3, and f4 of the four feet of the quadruped robot body on the ground are based on the friction cone as the force condition, and the angular velocity of the robot arm joint is The constraint condition is constructed at a maximum rotation speed.

8. The method for controlling the coordinated motion of a quadruped robot with a mechanical arm according to claim 1, characterized in that: In step S4, converting the variable weight cost function and the constraint condition into a quadratic programming problem includes: H and g are: Among them, H and g are constants, and U is the control variable.

9. The method for controlling the coordinated motion of a quadruped robot with a mechanical arm according to claim 8, characterized in that: The quadratic programming solver is used to calculate U, where U is the vector of n step-length control quantities ui. The first step-length ui is taken as the control quantity to obtain the reaction forces f1, f2, f3, f4 of the four foot ends of the quadruped robot body on the ground and the angular velocity vector of the six-degree-of-freedom manipulator joints 10. According to the method for controlling the coordinated motion of a quadruped robot with a mechanical arm in claim 1, the step S5 comprises: using the Jacobian matrix of each leg of the quadruped robot body and f1, f2, f3, f4 to calculate the desired torque of the motor of the single leg joint of the quadruped robot body; calculating the desired angular velocity of each joint of the mechanical arm The desired position instructions of each joint of the robotic arm are calculated by integration.

Citation Information

Cited By

  • Quadruped single-arm robot cooperative motion control method and system

    CN119820576A

  • Multi-humanoid robot distributed collaborative carrying method based on model predictive control

    CN119847023A

  • Multi-arm mobile robot cooperative control framework and multi-arm mobile robot system

    CN119858171A

  • Operation device based on robot dog and mechanical arm and control method thereof

    CN120134284A