A control method and system for non-grasping mobile pushing tasks of a humanoid robot

Abstract

This embodiment provides a control method and system for a non-grasping mobile pushing operation task of a humanoid robot, including: establishing a system interaction force model based on the pose information of pairwise interactions among the humanoid robot, the target object, and the environment, and then respectively establishing the dynamic models of the humanoid robot and the target object; establishing a system velocity constraint model and a system force constraint model based on the velocity and force constraint conditions satisfied by the pairwise interactions among the humanoid robot, the target object, and the environment, and obtaining the robot joint angular acceleration constraint conditions according to the established system velocity constraint model, system force constraint model, and the dynamic models of the humanoid robot and the target object; obtaining the system motion boundary conditions based on the established robot joint angular acceleration constraint conditions, and performing a non-grasping mobile operation task according to the motion trajectory of the humanoid robot under the condition of satisfying the system motion boundary conditions.

Description

Technical Field

[0001] The present invention belongs to the technical field related to robot control, and in particular relates to a method and system for controlling a non-grasping mobile pushing task of a humanoid robot. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Non-grasping pushing involves applying nonholonomic constraints to a target object. The object's motion is not entirely determined by the robot, but rather influenced by its own dynamics and its interaction with the environment. The advantages of non-grasping pushing are that it can move difficult-to-grasp objects and can cleverly utilize environmental assistance to achieve complex tasks with fewer robot degrees of freedom.

[0004] Humanoid robots move by using their legs to "manipulate" the ground, thereby achieving reverse motion of the main body. The interaction between the robot and the ground is also non-holonomically constrained. The similarities between leg-based movement and non-grasping pushing allow the mechanisms of the two to be reused. In addition, the humanoid robot platform can decouple the motion of the limb ends from the main body, isolating the work task from the main body's movement. Humanoid robots use non-grasping pushing to manipulate target objects, which can reduce the system complexity required for mobile operations and greatly expand the application areas of humanoid robot platforms. Therefore, non-grasping mobile pushing methods for humanoid robots have important research significance and broad application needs.

[0005] However, there is currently little research on non-grasping mobile pushing of humanoid robots worldwide, and it is limited to specific robot platforms and specific movement methods; my country has made certain achievements in the field of humanoid robot motion control, but research on non-grasping pushing is still in its infancy, and research on the combination of the two is almost blank. Summary of the Invention

[0006] In order to solve the above problems, the present invention provides a humanoid robot non-grasping mobile pushing task control method and system, which can drive the humanoid robot to move an object from a starting point to a target point by pushing without grasping the object, and ensure the stability of the humanoid robot's movement.

[0007] To achieve the above objectives, a first aspect of the present invention provides a method for controlling a non-grasping mobile pushing task of a humanoid robot, comprising:

[0008] Step 1: Establish a system interaction force model based on the posture information of the two-way interaction between the humanoid robot, the target object, and the environment. Based on the established system interaction force model, establish the dynamic models of the humanoid robot and the target object respectively.

[0009] Step 2: Based on the speed and force constraints satisfied by the pairwise interactions between the humanoid robot, the target object, and the environment, a system speed constraint model and a system force constraint model are established. The robot joint angular acceleration constraint conditions are obtained according to the established system speed constraint model, system force constraint model, and the dynamic models of the humanoid robot and the target object.

[0010] Step 3: Based on the established robot joint angular acceleration constraints, the system motion boundary conditions are obtained. The motion trajectory of the humanoid robot performs non-grasping mobile operation tasks while satisfying the system motion boundary conditions.

[0011] A second aspect of the present invention provides a humanoid robot non-grasping mobile pushing task control system, comprising:

[0012] Dynamic model building module: Based on the posture information of the two-way interaction between the humanoid robot, the target object, and the environment, a system interaction force model is established. Based on the established system interaction force model, the dynamic models of the humanoid robot and the target object are respectively established.

[0013] Constraint establishment module: Based on the speed and force constraints satisfied by the pairwise interactions between the humanoid robot, the target object, and the environment, a system speed constraint model and a system force constraint model are established. The robot joint angular acceleration constraint conditions are obtained according to the established system speed constraint model, system force constraint model, and the dynamic models of the humanoid robot and the target object.

[0014] Motion execution module: Based on the established robot joint angular acceleration constraints, the system motion boundary conditions are obtained. The motion trajectory of the humanoid robot performs non-grasping mobile operation tasks while meeting the system motion boundary conditions.

[0015] The third aspect of the present invention provides a computer device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, a method for controlling a non-grasping mobile pushing task of a humanoid robot is performed.

[0016] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, a method for controlling a non-grasping mobile pushing task of a humanoid robot is executed.

[0017] The beneficial effects of the present invention are:

[0018] Aiming at the non-grasping pushing task of a humanoid robot, the present invention provides a method for constructing an integrated dynamic model of "humanoid robot-target object-environment" and a multi-constraint characterization, mapping, and fusion algorithm, providing a general modeling framework for the non-grasping pushing task of a humanoid robot.

[0019] The present invention provides a motion trajectory generation method based on the phase plane method for the multivariable and multi-constraint model of the "legged robot-target object-environment" system, which can generate the task motion trajectory with the best time efficiency under the premise of satisfying the system constraints.

[0020] The present invention provides a method and system for moving a humanoid robot, which can control the humanoid robot to move objects that are difficult to grasp, and can cleverly use the help of the environment to achieve complex working tasks with fewer degrees of freedom of the humanoid robot, thereby broadening the application scenarios of the humanoid robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0022] Figure 1 Schematic diagram of system elements for a non-grasping mobile pushing task of a humanoid robot in the first embodiment of the present invention;

[0023] Figure 2 This is an example diagram of a force constraint structure in the first embodiment of the present invention;

[0024] Figure 3 This is a schematic diagram of phase plane trajectory planning in Example 1 of the present invention. DETAILED DESCRIPTION

[0025] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0026] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0027] Example 1

[0028] This embodiment proposes a method for controlling a non-grasping mobile pushing task of a humanoid robot, including:

[0029] Step 1: Establish a system interaction force model based on the posture information of the two-way interaction between the humanoid robot, the target object, and the environment. Based on the established system interaction force model, establish the dynamic models of the humanoid robot and the target object respectively.

[0030] Step 2: Based on the speed and force constraints satisfied by the pairwise interactions between the humanoid robot, the target object, and the environment, a system speed constraint model and a system force constraint model are established. The robot joint angular acceleration constraint conditions are obtained according to the established system speed constraint model, system force constraint model, and the dynamic models of the humanoid robot and the target object.

[0031] Step 3: Based on the established robot joint angular acceleration constraints, the system motion boundary conditions are obtained. The motion trajectory of the humanoid robot performs non-grasping mobile operation tasks while satisfying the system motion boundary conditions.

[0032] In step 1 of this embodiment, a system interaction force model is established based on the posture information of the two-way interaction between the humanoid robot, the target object, and the environment:

[0033] Use vectors to describe the interaction posture information between the humanoid robot, the target object, and the environment, and establish the system interaction posture representation composed of the humanoid robot, the target object, and the environment;

[0034] Vectors are used to describe the interaction forces between the humanoid robot, target object, and environment in the system interaction posture model, and a system interaction force representation consisting of the humanoid robot, target object, and environment is established.

[0035] Specifically, we first define the interaction model established by the two-way interaction between the humanoid robot, the target object (i.e., the manipulated object), and the environment, such as Figure 1 As shown, using vector e A1 、e A2 etc. describe the interaction point position and interaction point posture between the humanoid robot and the target object, e B1 、e B2 etc. describe the interaction point position and interaction point posture between the humanoid robot and the environment, e C1 、e C2 The position and pose of the interaction point between the target object and the environment are described using vectors like [ 1 ] . The position and pose elements in each vector are independent of each other. For example, the interaction between a humanoid robot's foot and the ground can be represented by a six-dimensional vector containing both a three-dimensional position and a three-dimensional pose. When the robot's rounded extremities interact with an object, the interaction between the two only requires a three-dimensional position vector.

[0036] On this basis, the interaction configuration vector between the humanoid robot, the target object, and the environment is:

[0037] The interaction posture configuration between the humanoid robot and the target object is:

[0038] Interaction posture configuration between the humanoid robot and the surrounding environment:

[0039] Interaction pose configuration between the target object and the environment:

[0040] Interaction pose configuration within the humanoid robot-target object-environment system:

[0041] If some limbs of the humanoid robot are in a swinging state and do not interact with any object or environment, in this embodiment, the end position is represented by the vector e D1 、e D2 etc., then the end pose configuration of the humanoid robot's swinging limb is defined as:

[0042] According to the established system interaction force model, the dynamic models of the humanoid robot and the target object are established respectively as follows:

[0043] In step 1 of this embodiment, the dynamic models of the humanoid robot and the target object are respectively established based on the established system interaction force model:

[0044] Establish a vector representation of the humanoid robot based on the degrees of freedom of the humanoid robot body and the limb joint angles;

[0045] Establishing a humanoid robot dynamics model in an inertial coordinate system according to the inertia matrix of the humanoid robot, the centripetal force matrix of the humanoid robot, the gravity vector of the humanoid robot, the driving joint selection matrix of the humanoid robot, the joint driving torque vector of the humanoid robot, the Jacobian matrix of the interactive posture representation of the system, the interactive force selection matrix, and the established vector representation of the humanoid robot;

[0046] The dynamic model of the target object in the inertial coordinate system is established based on the center of mass position, posture angle, object inertia matrix, object gravity vector, and the interactive posture of the target object, the humanoid robot, and the environment.

[0047] Specifically, define the vector f A1 Description A1 The force and torque acting on the humanoid robot at the interaction point, f A1 With e A1 Should have the same dimensions. Similarly, the interaction force and interaction torque vectors generated at other interactions can be defined, and the interaction force configuration of the system can be defined as follows:

[0048] By e A The interaction force generated at and acting on the humanoid robot is configured as follows:

[0049] By e A The interaction force configuration acting on the target object is:

[0050] By e B The interaction force configuration acting on the humanoid robot is:

[0051] By e c The interaction force configuration acting on the target object is:

[0052] Interaction force configuration within the humanoid robot-target object-environment system:

[0053] In this description, the symbol 0 is defined m×n Represents a zero matrix of dimension m×n, symbol I n×n Represents the identity matrix of dimension n×n.

[0054] In this embodiment, the configuration vector for positioning the humanoid robot is q, which can be decomposed into:

[0055] q=[p b T θ T ] T (1)

[0056] in, Represents the i degrees of freedom of the robot body, including its position and attitude angle, i≤6, are the n joint angles on the robot’s limbs. The dynamic model of the humanoid robot in the inertial reference frame is:

[0057]

[0058] in, is the inertia matrix of the humanoid robot, For the Coriolis force / centripetal force matrix of the humanoid robot, is the gravity vector of the humanoid robot, S τ =[0 n×i I n×n ] is the matrix for driving joints of humanoid robots, For humanoid robot joint drive torque vector, From inertial coordinate system to The Jacobian matrix of Interaction force selection matrix.

[0059] In this embodiment, the definition Represents the j degrees of freedom of the target object itself, including its center of mass position and attitude angle, i≤6. Considering the object as a rigid body, its dynamic model in the inertial reference frame can be written as:

[0060]

[0061] in, is the object inertia matrix, is the object's gravity vector, From inertial coordinate system to The Jacobian matrix of Select the matrix for the interaction force.

[0062] Since the variables in e are linearly independent, J r and J o It should be a full row rank matrix. If the variables in e do not satisfy the linear independence condition, the corresponding J r and J o Perform SVD decomposition to obtain the reduced dimension e and then perform modeling.

[0063] Plan a suitable motion trajectory for θ so that p b and p o From the initial posture state to the desired state, while keeping the interaction configuration in e unchanged.

[0064] In step 2 of this embodiment, the robot joint angular acceleration constraint condition is obtained based on the established system velocity constraint model, system force constraint model, and the dynamic model of the humanoid robot and the target object, which is:

[0065] The relationship between the robot body acceleration and the target object acceleration is established based on the Jacobian matrix corresponding to the robot body motion, the Jacobian matrix corresponding to the robot joint motion, and the system velocity constraint model;

[0066] Establishing a relationship representation that can describe the interaction force and the motion of the humanoid robot based on the Jacobian matrix corresponding to the robot body motion, the Jacobian matrix corresponding to the robot joint motion, the vector representation of the humanoid robot, and the humanoid robot dynamics model;

[0067] Based on the relationship between the interaction force and the motion of the humanoid robot and the dynamic model of the target object, the acceleration of the degree of freedom of the selected robot body is obtained respectively. Acceleration of the selected target object's degree of freedom

[0068] Based on the obtained Substitute it into the relationship representation between the interaction force and the humanoid robot's motion, and merge it with the interaction force constraint representation between the target object and the environment;

[0069] The fused representation is solved to obtain the robot joint angular acceleration constraints.

[0070] Specifically, to maintain the humanoid robot and the target object A The interaction configuration at , the speed between the two should have the following relationship:

[0071]

[0072] in, and To select the matrix, specifically:

[0073]

[0074]

[0075] Formula (4) can make the humanoid robot and the manipulated object in a fixed state at the interaction point. In addition, in some non-grasping mobile operation modes, relative sliding can also occur between the humanoid robot and the manipulated object. However, in the sliding situation, the humanoid robot cannot fully control the movement of the manipulated object. The research focus of this embodiment is to accurately control the movement of the manipulated object based on the non-grasping operation mode, so only the e A Fixed interactive operation mode.

[0076] Similarly, the interaction configuration between the robot and the environment should also be stable, that is, the robot should be stable in e B The velocity at should be 0:

[0077]

[0078] in, The selection matrix is ​​defined as follows:

[0079]

[0080] When an object is moved in a non-grasping manner, it should slide relative to the environment in some dimension. For example, when a box is pushed on a horizontal surface, its interaction with the ground will limit its vertical velocity and pitch angle to 0, while the horizontal velocity is uncertain. Define the matrix S c Used to select restricted / unrestricted degrees of freedom at the interaction:

[0081]

[0082] The elements in the matrix are defined as:

[0083]

[0084] Then e C The velocity constraint expression at can be written as:

[0085]

[0086] in,

[0087]

[0088] Combining equations (4), (7) and (11), the speed constraint of the system can be written as:

[0089]

[0090] in,

[0091]

[0092] In this embodiment, when the friction coefficient at each interaction point in the humanoid robot-object-environment system is known, the constraint rules for the interaction force are formulated as follows:

[0093] [P1] At an interaction with no relative motion, the interaction force should be within the friction cone;

[0094] [P2] The center of the interaction force between two interacting objects should be located within the contact area between them;

[0095] [P3] If relative sliding occurs at the interaction point, the interaction force at that point should be located at the edge of the friction cone.

[0096] According to [P1] and [P2], the feasible domain of interaction force at each interaction point can be written:

[0097]

[0098]

[0099]

[0100] Then the feasible domain of the system interaction force can be expressed as:

[0101]

[0102] Considering only the Coulomb friction model, based on [P3], it can be obtained that the interaction force at the sliding interaction between the object and the environment should satisfy:

[0103]

[0104] Where v represents the relative velocity between the target object and the environment, μ represents the friction coefficient at the interaction, and f n is the normal component of the interaction force, f v is the interaction force component along the v direction.

[0105] Using S given by equations (9) and (10) c Matrix definition, e C The interaction force constraint at can be written as:

[0106]

[0107] Formula (18) can be written in the following matrix form:

[0108] (I c×c -S c )Tf C =0 c×1 (19)

[0109] by Figure 2 Taking the plane model shown as an example, based on [P1] we can get:

[0110]

[0111]

[0112] The bottom of the manipulated object should always be attached to the ground. Based on [P2], we can get:

[0113]

[0114] The force constraint of the system can be summarized as follows:

[0115]

[0116] The manipulated object should slide forward along the ground. Based on [P3], we can get:

[0117]

[0118] In this case, the constraints in formula (19) can be specified as follows:

[0119]

[0120] The matrix J in equations (4), (7), and (13) is r Written as follows:

[0121] J r =[J b J θ] (26)

[0122] in, is the Jacobian matrix corresponding to the robot body motion, is the Jacobian matrix corresponding to the robot joint motion.

[0123] Combining formulas (1), (13), and (26), we get:

[0124]

[0125] Taking the time differential of equation (27) we can get:

[0126]

[0127] Formula (28) gives the relationship between the acceleration of the robot body and the acceleration of the manipulated object. The next task is to relate it to the dynamics of the robot and the object and the interaction force.

[0128] Considering equations (1) and (26), equation (2) can be written as:

[0129]

[0130] There is no torque variable in the top i rows of Equation (29), so it can be used to express the relationship between the interaction force and the robot motion:

[0131]

[0132] Define M b =M i×i , M θ =M i×n , C b =C i×i , C θ =C i×n and g b =g i×1 , formula (30) It can be written as follows:

[0133]

[0134] Multiply equation (31) by Can be obtained for:

[0135]

[0136] Similarly, It can be multiplied by formula (3) on the left Multiply to get:

[0137]

[0138] Substituting equations (32) and (33) into equation (28), we can obtain:

[0139]

[0140] in,

[0141] Φ=S r (J θ -J b M b -1 M θ )

[0142]

[0143]

[0144]

[0145]

[0146]

[0147] in, is the interaction acceleration caused by the humanoid robot's joint angular velocity, is the interaction acceleration caused by the rotational velocity of the humanoid robot body, the link, and the object; for e A and e C The force at The force on the target object, is the object's acceleration, is the interaction point acceleration; for e A and e B The force at The humanoid robot body is subjected to force. is the robot body acceleration, The interaction acceleration is caused by the acceleration of the humanoid robot itself.

[0148] Combining the mathematical relationship about the interaction force vector in formula (19) with formula (34) yields:

[0149]

[0150] in,

[0151]

[0152] Υ is a square matrix of (a + b + c)×(a + b + c). If Υ is invertible, multiply the left side of Equation (36) by Υ -1 The system force f can be obtained as follows:

[0153]

[0154] If Υ is not invertible and its rank is assumed to be r (r < a + b + c), it can be transformed into a matrix of r×(a + b + c) through methods such as SVD decomposition:

[0155]

[0156] where U, Π, and V * are the unitary matrix, diagonal matrix, and conjugate transpose of the unitary matrix obtained from SVD respectively, is a zero matrix.

[0157] Multiply the left side of Equation (39) by U -1 We can get:[[ID=2.)]]

[0158]

[0159] To solve for f in Equation (40), (a + b + c - r) constraints need to be added to f manually:

[0160]

[0161] where, and satisfy:

[0162] det(pγ′ T E T T )≠0 (42)

[0163] Combining Equation (40) and Equation (41) can obtain f:

[0164]

[0165] Write Equation (38) and Equation (43) in the following unified form:

[0166]

[0167] Combining Equation (16)(41) and Equation (44), we can get:

[0168]

[0169] Regarding the velocity term and gravity term on the left side of Equation (45) as the existing motion state of the system, Equation (45) gives the mathematical constraint expression for the joint angular acceleration of the robot at the current moment.

[0170] ​In step 3 of this embodiment, the robot joint angular acceleration constraint condition is mapped into the system space to obtain the constraint representation in the system space;

[0171] By selecting a certain degree of freedom for planning, the remaining degrees of freedom are pre-constrained or established in relation to the motion of the selected degree of freedom, and then the position, velocity, and acceleration representation of the selected degree of freedom are established;

[0172] The position, velocity, and acceleration representations of the selected degrees of freedom are substituted into the constraint representations in the system space to solve the feasible domain of the acceleration of the selected degrees of freedom. The feasible domain of the acceleration of the selected degrees of freedom is satisfied at any time when the system moves from the initial state position to the target state position.

[0173] Specifically, the principle of motion planning for the system is to find a suitable robot joint motion trajectory to satisfy: in all joint configurations and speed states corresponding to the entire process, the robot's joint angular acceleration The constraints in formula (45) are never violated.

[0174] However, it is not convenient and intuitive to perform motion planning in the robot joint space {θ}. Therefore, the task of this step is to map the constraints in Eq. (45) to the system workspace {p b ,p o ,e D}.

[0175] If the robot has no redundant joint degrees of freedom, {θ} and {p b ,p o ,e D If the robot has redundant joint degrees of freedom, a certain workspace configuration will correspond to multiple joint space configurations. In the study of the embodiment, it is assumed that additional constraints are added to the inverse kinematics solution of the system to make the inverse kinematics solution unique (for example, adding constraints to minimize the instantaneous kinematic energy of the robot), so that {p b ,p o ,e D} Solve for the unique {θ}. For the above robot without redundant joint degrees of freedom and with redundant joint degrees of freedom, a unified definition is given: For J θ The inverse matrix or generalized inverse matrix of , then:

[0176]

[0177]

[0178] where p and J p is defined as:

[0179]

[0180] Substituting equations (46) and (47) into equation (45), we can obtain the constraint expression in the system workspace:

[0181]

[0182] in:

[0183]

[0184]

[0185] How to plan the system trajectory so that the system can move from the initial configuration p to the initial configuration p under the premise of satisfying the constraints in formula (49) S Move to the target configuration p G , its essence is based on the changes in the system's motion trajectory Find something feasible However, it is very difficult to plan all the a+b+d degrees of freedom in p at the same time. This step solves this problem by using the system dimensionality reduction method, that is, selecting a certain degree of freedom in p for motion planning, and pre-constraining the remaining degrees of freedom or establishing mathematical connections with the motion of the selected degree of freedom. Based on this idea, p, and Written as:

[0186]

[0187] Among them, p, and are the position, velocity and acceleration values ​​of the selected degree of freedom respectively; h(p) is a vector whose internal elements are all functions of p; p pre , and Artificially set motion patterns for all degrees of freedom except the selected one.

[0188] Substituting formula (52) into formula (49) yields:

[0189]

[0190] in,

[0191]

[0192] It can be solved by (53) The feasible domain of , the steps are as follows:

[0193] Step (1) constrain the system force Written in the form of mathematical equations or inequalities, as shown in equations (20)(21)(22)(41);

[0194] Step (2) is based on the formula Write f as Expressions of

[0195] Step (3) Combine the mathematical expressions obtained in steps (1) and (2) to solve feasible domain.

[0196] like The feasible region does not exist, which means that for a given h(p), p pre 、 and In the current system state p and , the system is unstable.

[0197] like The feasible domain exists, which can be written as:

[0198]

[0199] The motion planning problem of the system can be expressed as: given the initial state of the system Plan a suitable p(t) to make the system move to the target state And at any time t, the constraints in formula (55) are satisfied.

[0200] This step will Plan the motion trajectory of the system in the phase plane. For any point on the phase plane, the acceleration range corresponding to the point can be solved by equation (55), and the range is expressed as the cone feasible region in the tangent direction of the point. The phase plane trajectory passing through the point must fall within the cone feasible region. Point cannot be obtained If the effective range of the point is not within the effective range, then there is no cone-shaped feasible region corresponding to the point in the phase plane, and it is impossible to plan an effective system trajectory at this point. The robot's workspace boundary and speed boundary can also be expressed in the form of regional boundaries. In the phase plane.

[0201] like Figure 3 As shown, the steps of motion trajectory planning in the phase plane are as follows:

[0202] Step 1: From the point Start solving According to the obtained Move forward to the new phase plane point and repeat this process until p = p G (α is the scaling factor, 0≤α≤1);

[0203] Step 2: From the point Start solving According to the obtained Move back to the new phase plane point and repeat this process until p = p S (β is the scaling factor, 0≤β≤1);

[0204] Step 3: Find the intersection of the forward and reverse trajectories obtained in the above two steps

[0205] Step 4: Check the results of the above three steps The trajectory is checked to see if it meets the system's workspace and velocity range requirements. If so, the trajectory is a valid trajectory for the task. If not, change the values ​​of α and β and start again from Step 1.

[0206] After obtaining a valid phase plane trajectory through the above steps, the robot's motion can be controlled accordingly to perform the desired non-grasping mobile task.

[0207] Example 2

[0208] This embodiment provides a humanoid robot non-grasping mobile pushing task control system, which includes:

[0209] Dynamic model building module: Based on the posture information of the two-way interaction between the humanoid robot, the target object, and the environment, a system interaction force model is established. Based on the established system interaction force model, the dynamic models of the humanoid robot and the target object are respectively established.

[0210] Constraint establishment module: Based on the speed and force constraints satisfied by the pairwise interactions between the humanoid robot, the target object, and the environment, a system speed constraint model and a system force constraint model are established. The robot joint angular acceleration constraint conditions are obtained according to the established system speed constraint model, system force constraint model, and the dynamic models of the humanoid robot and the target object.

[0211] Motion execution module: Based on the established robot joint angular acceleration constraints, the system motion boundary conditions are obtained. According to the motion trajectory of the humanoid robot, non-grasping mobile operation tasks are performed when the system motion boundary conditions are met.

[0212] Example 3

[0213] The purpose of this embodiment is to provide a computing device, including: a processor, a memory and a bus, the memory storing machine-readable instructions executable by the processor, when the computer device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, a method for controlling a non-grasping mobile pushing task of a humanoid robot is performed.

[0214] Example 4

[0215] The purpose of this embodiment is to provide a computer-readable storage medium.

[0216] A computer-readable storage medium is characterized in that a computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, a method for controlling a non-grasping mobile pushing task of a humanoid robot is executed.

[0217] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage and optical storage, etc.) containing computer-usable program code.

[0218] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0219] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A control method for non-grasping mobile pushing tasks of a humanoid robot, characterized in that, Including: Step 1: Establish a system interaction force model based on the pose information of pairwise interactions among the humanoid robot, the target object, and the environment, and establish the dynamic models of the humanoid robot and the target object according to the established system interaction force model. In the said Step 1, establishing the dynamic models of the humanoid robot and the target object according to the established system interaction force model specifically includes: Establish a vector representation of the humanoid robot based on the degrees of freedom of the humanoid robot body and the limb joint angles. Establish a dynamic model of the humanoid robot in the inertial coordinate system according to the inertia matrix of the humanoid robot, the centripetal force matrix of the humanoid robot, the gravity vector of the humanoid robot, the selected matrix of the humanoid robot drive joints, the joint drive torque vector of the humanoid robot, the Jacobian matrix of the said system interaction pose representation, the interaction force selection matrix, and the established vector representation of the humanoid robot. Establish a dynamic model of the target object in the inertial coordinate system according to the centroid position, attitude angle, object inertia matrix, object gravity vector of the target object, and the interaction pose representation between the target object, the humanoid robot, and the environment. Step 2: Establish a system velocity constraint model and a system force constraint model based on the velocity and force constraint conditions satisfied by the pairwise interactions among the humanoid robot, the target object, and the environment, and obtain the robot joint angular acceleration constraint conditions according to the established system velocity constraint model, system force constraint model, and the dynamic models of the humanoid robot and the target object. Step 3: Obtain the system motion boundary conditions based on the established robot joint angular acceleration constraint conditions, and when the motion trajectory of the humanoid robot satisfies the system motion boundary conditions, execute the non-grasping mobile operation task.

2. The control method for non-grasping mobile pushing operation tasks of a humanoid robot according to claim 1, characterized in that, In the said Step 1, establishing the system interaction force model based on the pose information of pairwise interactions among the humanoid robot, the target object, and the environment specifically includes: Use vectors to describe the pose information of pairwise interactions among the humanoid robot, the target object, and the environment, and establish a system interaction pose representation formed by the humanoid robot, the target object, and the environment. Use vectors to describe the interaction force situations among the humanoid robot, the target object, and the environment in the system interaction pose model, and establish a system interaction force representation formed by the humanoid robot, the target object, and the environment.

3. The control method for a non-grasping mobile pushing task of a humanoid robot according to claim 1, characterized in that In the said Step 2, for the contact and sliding relative motion forms among the robot, the target object, and the environment, obtain the expression of the relative velocity relationship among the various parts, that is, the velocity constraint model; the representation form of the interaction force among the various parts in its friction cone, that is, the interaction force constraint model.

4. The control method for a non-grasping mobile pushing task of a humanoid robot according to claim 1, characterized in that, In the said Step 2, obtain the robot joint angular acceleration constraint conditions according to the established system velocity constraint model, system force constraint model, and the dynamic models of the humanoid robot and the target object, specifically including: Establish the relationship representation between the robot body acceleration and the target object acceleration according to the Jacobian matrix corresponding to the robot body motion, the Jacobian matrix corresponding to the robot joint motion, and the system velocity constraint model. Based on the Jacobian matrix corresponding to the movement of the robot body, the Jacobian matrix corresponding to the joint movement of the robot, the vector representation of the humanoid robot, and the dynamic model of the humanoid robot, establish a relationship representation that can describe the interaction force and the movement of the humanoid robot; Based on the relationship representation between the interaction force and the movement of the humanoid robot and the dynamic model of the target object, obtain the acceleration of the robot body and the acceleration of the target object respectively; Substitute the obtained acceleration of the robot body and the acceleration of the target object into the relationship representation between the interaction force and the movement of the humanoid robot, and fuse it with the interaction force constraint representation between the target object and the environment; Solve the fused representation to obtain the joint angular acceleration constraint condition of the robot.

5. The control method for a non-grasping mobile pushing task of a humanoid robot according to claim 1, characterized in that, In step 3, map the joint angular acceleration constraint condition of the robot into the system space to obtain the constraint representation in the system space; By selecting a certain degree of freedom for planning, pre-constrain the remaining degrees of freedom or establish a representation connection with the movement of the selected degree of freedom, and then establish the representation of the position, velocity, and acceleration of the selected degree of freedom; Substitute the established representation of the position, velocity, and acceleration of the selected degree of freedom into the constraint representation in the system space, solve the feasible region of the acceleration of the selected degree of freedom, and when moving from the initial state position of the system to the target state position, satisfy the feasible region of the acceleration of the selected degree of freedom at any time.

6. The control method for a non-grasping mobile pushing task of a humanoid robot according to claim 5, wherein In step 3, in the phase plane, solve the feasible region of the acceleration of the selected degree of freedom from the initial state of the system until moving forward to the target state position of the system; Solve the feasible region of the acceleration of the selected degree of freedom from the initial state of the system until moving backward to the target state position of the system; Obtain the intersection position according to the above forward movement and backward movement, and from the initial state position of the system to the intersection position and then to the system target position, perform a non-grasping movement operation task under the condition of satisfying the constraints in the system space.

7. A non-grasping mobile pushing and transporting task control system for a humanoid robot, characterized in that, Include: Dynamic model establishment module: Establish a system interaction force model based on the pose information of pairwise interactions between the humanoid robot, the target object, and the environment, and establish the dynamic models of the humanoid robot and the target object according to the established system interaction force model; among them, establishing the dynamic models of the humanoid robot and the target object according to the established system interaction force model specifically is: Based on the degrees of freedom of the humanoid robot body and the limb joint angles, establish the vector representation of the humanoid robot; According to the inertia matrix of the humanoid robot, the centripetal force matrix of the humanoid robot, the gravity vector of the humanoid robot, the selected matrix of the driving joints of the humanoid robot, the joint driving torque vector of the humanoid robot, the Jacobian matrix of the system interaction pose representation, the interaction force selection matrix, and the established vector representation of the humanoid robot, establish the dynamic model of the humanoid robot in the inertial coordinate system; According to the centroid position, attitude angle, object inertia matrix, object gravity vector of the target object, and the interaction pose representation between the target object and the humanoid robot and the environment, establish the dynamic model of the target object in the inertial coordinate system; Constraint establishment module: Based on the velocity and force constraint conditions satisfied by the pairwise interactions among the humanoid robot, the target object, and the environment, establish a system velocity constraint model and a system force constraint model, and obtain the robot joint angular acceleration constraint conditions according to the established system velocity constraint model, system force constraint model, and the dynamic models of the humanoid robot and the target object; Motion execution module: Based on the established robot joint angular acceleration constraint conditions, obtain the system motion boundary conditions, and when the motion trajectory of the humanoid robot satisfies the system motion boundary conditions, execute the non-grasping mobile operation task.

8. A computer device, characterized in that, Including: A processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device runs, the processor communicates with the memory through the bus. When the machine-readable instructions are executed by the processor, it executes a method for controlling a non-grasping mobile pushing operation task of a humanoid robot according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium. When the computer program is run by the processor, it executes a method for controlling a non-grasping mobile pushing operation task of a humanoid robot according to any one of claims 1 to 6.

Images