A neural dynamics posture adjustment method based on the acceleration layer of a three-wheeled omnidirectional mobile manipulator

Abstract

The present invention relates to a neurodynamic posture adjustment method for the acceleration layer of a three-wheeled omnidirectional mobile manipulator. The method comprises the following steps: obtaining state parameters of the three-wheeled omnidirectional mobile manipulator and inputting the state parameters into a pre-constructed acceleration layer posture adjustment model; solving the acceleration layer posture adjustment model; and simultaneously adjusting the mobile platform and the manipulator to a desired posture based on the calculation results of the acceleration layer posture adjustment model; the acceleration layer posture adjustment model is established based on performance indicators described on the acceleration layer, in combination with the kinematic equations of the mobile platform acceleration layer and the physical limits of the three-wheeled omnidirectional mobile manipulator; the performance indicators described on the acceleration layer are derived using neurodynamic design formulas based on the idea of ​​zeroing the error between the current posture of the mobile manipulator and the desired posture. The present invention can efficiently achieve automatic posture adjustment of the three-wheeled omnidirectional mobile manipulator at both the acceleration layer and the velocity layer, facilitating the execution of operational tasks in different postures.

Description

Technical Field

[0001] The present invention relates to the field of motion planning of a three-wheeled omnidirectional mobile robotic arm, and in particular to a neurodynamic posture adjustment method of an acceleration layer of a three-wheeled omnidirectional mobile robotic arm. Background Art

[0002] Due to its flexibility and high maneuverability, three-wheeled omnidirectional mobile robotic arms have been applied in numerous fields, including industry, medicine, and the military. After completing a given task, a three-wheeled omnidirectional mobile robotic arm often needs to adjust its current posture to a specified or desired posture to continue the task. Generally, posture adjustment of a three-wheeled omnidirectional mobile robotic arm can be achieved manually: first adjusting the mobile platform, then the robotic arm; or first adjusting the robotic arm, then the mobile platform. Obviously, this method is cumbersome and time-consuming: it requires multiple measurements of the mobile platform's position and orientation angle, as well as the robotic arm's joint angles, and it cannot adjust both the mobile platform and the robotic arm simultaneously.

[0003] To rapidly adjust the posture of a three-wheeled, omnidirectional mobile manipulator, several solutions have been proposed. For example, Chinese patent application CN202111142636.6 discloses a state adjustment method for a wheeled mobile robot based on gradient descent. This method, based on the gradient descent formula, designs a new performance indicator described at the velocity level. Currently, existing technologies are all researched and designed at the velocity level, which limits their applicability. These solutions cannot be directly applied to a three-wheeled, omnidirectional mobile manipulator controlled by acceleration, nor can they further consider and address acceleration and torque limits. Summary of the Invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a new method for automatically adjusting the posture of a three-wheeled omnidirectional mobile robotic arm, which is easy to operate, has low workload, and standardized operation, and can efficiently realize automatic adjustment of the posture of the three-wheeled omnidirectional mobile robotic arm both on the acceleration layer and on the speed layer.

[0005] The purpose of the present invention can be achieved by the following technical solutions:

[0006] As a first aspect of the present invention, a method for adjusting the posture of a three-wheeled omnidirectional mobile manipulator in an acceleration layer is provided. The three-wheeled omnidirectional mobile manipulator comprises a mobile platform having a plurality of omnidirectional drive wheels and a manipulator disposed on the mobile platform. The method comprises the following steps:

[0007] Obtain the state parameters of the three-wheeled omnidirectional mobile manipulator and input the state parameters into the pre-built acceleration layer attitude adjustment model;

[0008] Solve the acceleration layer attitude adjustment model to obtain the rotation angle, angular velocity and angular acceleration of the mobile platform's drive wheels, as well as the angle, velocity and acceleration of the robotic arm's joints;

[0009] According to the calculation results, the mobile platform and the robotic arm are adjusted simultaneously to reach the desired posture;

[0010] The acceleration layer posture adjustment model is established based on the performance indicators described on the acceleration layer, combined with the kinematic equations of the mobile platform acceleration layer and the physical limits of the three-wheeled omnidirectional mobile robotic arm; the performance indicators described on the acceleration layer are derived using the neurodynamic design formula based on the idea of ​​zeroing the error between the current posture of the mobile robotic arm and the desired posture.

[0011] Furthermore, the performance indicators described on the acceleration layer are described as follows:

[0012]

[0013] Among them, ||·||2 represents the two-norm of the vector, M∈R (3+n)×(3+n) is the coefficient matrix; is the coefficient matrix, e is the augmented matrix, which represents the error between the current posture and the expected posture, and represents the joint acceleration vector and joint velocity vector of the three-wheeled omnidirectional mobile manipulator, and Represent the acceleration and velocity of the robot arm joints, and They represent the rotational angular acceleration and rotational angular velocity of the omnidirectional drive wheels of the mobile platform, n represents the number of degrees of freedom of the robotic arm installed on the mobile platform, and 3 represents the number of omnidirectional drive wheels of the mobile platform.

[0014] Furthermore, the coefficient matrix middle:

[0015] A=[H,0;0,I n ]∈R (n+3)×(n+3)

[0016] B=[I2,0,0;0,cosφ,0;0,0,I n ]∈R (3+n)×(3+n)

[0017]

[0018] in, represents the time derivative of A, I n ∈R n×n represents the identity matrix, φ∈R represents the orientation angle of the mobile platform, k iand λ i (i=1,2,3) are the design parameters for achieving posture adjustment, H∈R 3×3 is the structural parameter coefficient matrix of the mobile platform;

[0019] The augmented matrix e is expressed as follows:

[0020]

[0021] Among them, p xy Indicates the coordinates of the mobile platform along the XY axis on the horizontal ground, p xyd ,φ d and θ d They represent the desired coordinates and orientation angles of the mobile platform along the X and Y axes on the horizontal ground, as well as the desired angles of the robotic arm joints.

[0022] Furthermore, the mobile platform structural parameter coefficient matrix H is expressed as follows:

[0023]

[0024] Where r>0∈R represents the radius of the omnidirectional drive wheel of the mobile platform, l>0∈R represents the distance from the center point of the mobile platform to the omnidirectional drive wheel, and φ∈R represents the heading angle of the mobile platform.

[0025] Furthermore, the acceleration layer attitude adjustment model is constrained by the acceleration layer kinematic equations of the mobile platform, including the rotation angle limit, rotation angular velocity limit and rotation angular acceleration limit of the omnidirectional drive wheels and the angle limit, velocity limit and acceleration limit of the robotic arm joints.

[0026] Furthermore, the specific steps of solving the acceleration layer attitude adjustment model are:

[0027] Convert the acceleration layer attitude adjustment model into a quadratic optimization problem;

[0028] The solution of quadratic optimization problem is equivalently converted into the solution of piecewise linear projection equation;

[0029] A numerical algorithm is used to obtain the optimal solution to the quadratic optimization problem, including the rotation angle, rotation angular velocity, rotation angular acceleration of the omnidirectional drive wheels of the mobile platform, as well as the angle, velocity and acceleration of the robotic arm joints.

[0030] Furthermore, the acceleration layer attitude adjustment model is transformed into the following quadratic optimization problem:

[0031] Minimize: x T Qx / 2+p T x

[0032] Constraints: x- ≤x≤x +

[0033] Among them, the superscript T represents the transpose of a matrix or vector, represents the decision variable of the quadratic optimization problem; x ± represents the limit of x, and, η>0∈R and γ>0∈R represent the limit conversion parameters, ε>0∈R 3+n Indicates the limit conversion margin; Q = M T M∈R (3+n)×(3+n) , M=BA∈R (3+n)×(3+n) , p=M T q∈R 3+n , and are the augmented angle vector, velocity vector, and acceleration vector of the three-wheeled omnidirectional mobile manipulator respectively; accordingly, and Respectively represent u, and Furthermore, the piecewise linear projection equation is as follows:

[0034] F Ω (x-(Qx+p))-x=0∈R 3+n

[0035] Among them, F Ω (·) represents the piecewise linear projection operator;

[0036] Furthermore, the piecewise linear projection equation is solved using the following numerical algorithm:

[0037] e(x k )=x k -F Ω (x k -(Qx k +p))

[0038] x k+1 =F Ω (x k -α(x k )β(x k ))

[0039] β(x k )=Q T e(x k )+Qx k +p

[0040]

[0041] Here, the superscript k represents the number of iterations and k = 0, 1, 2, ... .

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] 1) This invention provides a novel method for simultaneously and automatically adjusting a three-wheeled omnidirectional mobile platform and a robotic arm between different postures on an acceleration layer. This method uses a neurodynamic design formula to derive a performance metric described on the acceleration layer. Based on minimizing this performance metric, an acceleration layer posture adjustment model is established. This novel solution can be directly applied to both acceleration-controlled and velocity-controlled three-wheeled omnidirectional robotic arms, and can further consider and address the acceleration and torque limits of the three-wheeled omnidirectional robotic arm.

[0044] 2) This invention transforms the acceleration layer posture adjustment model into a quadratic optimization problem. Using a numerical algorithm, iterative calculations are performed to obtain the optimal solution, namely, the rotation angle, angular velocity, and angular acceleration of the mobile platform's omnidirectional drive wheels, as well as the angles, velocities, and accelerations of the robotic arm's joints. Based on these calculations, the three-wheeled omnidirectional mobile robotic arm controller drives the mobile platform's three omnidirectional wheels and the robotic arm's joints in real time, enabling the three-wheeled omnidirectional mobile robotic arm to quickly and accurately adjust from its current posture to the desired one. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of a method for adjusting the acceleration layer posture of a three-wheeled omnidirectional mobile robotic arm based on neurodynamics in the present invention. DETAILED DESCRIPTION

[0046] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0047] Example 1

[0048] As one of the embodiments of the present invention, this embodiment provides a method for adjusting the acceleration layer posture of a three-wheeled omnidirectional mobile manipulator based on neurodynamics. The three-wheeled omnidirectional mobile manipulator includes a mobile platform with multiple omnidirectional drive wheels and a manipulator disposed on the mobile platform. First, the state parameters of the three-wheeled omnidirectional mobile manipulator are obtained, and the state parameters are input into a pre-constructed acceleration layer posture adjustment model; the acceleration layer posture adjustment model is solved to obtain the rotation angle, rotation angular velocity and rotation angular acceleration of the mobile platform drive wheels and the angle, velocity and acceleration of the manipulator joints; the mobile platform and the manipulator are adjusted simultaneously to reach the desired posture according to the calculation results; the acceleration layer posture adjustment model is established based on the performance indicators described on the acceleration layer in combination with the kinematic equations of the mobile platform acceleration layer and the physical limits of the three-wheeled omnidirectional mobile manipulator; the performance indicators described on the acceleration layer are derived using the neurodynamic design formula based on the idea of ​​zeroing the error between the current posture of the mobile manipulator and the desired posture. The specific steps are as follows: Figure 1 As shown:

[0049] Step 1: Based on the idea of ​​zeroing the error between the current and desired postures of the mobile manipulator, a performance indicator described at the acceleration level is derived using a neurodynamic design formula.

[0050] Step 2: Based on minimizing the performance index, an acceleration layer attitude adjustment model is established; the scheme is constrained by the acceleration layer kinematic equations of the mobile platform, the rotation angle limit, rotation angular velocity limit, rotation angular acceleration limit of the omnidirectional drive wheel, and the angle limit, velocity limit, and acceleration limit of the robot arm joint;

[0051] Step 3: Convert the acceleration layer attitude adjustment model into a quadratic optimization problem;

[0052] Step 4: Use numerical algorithms to perform iterative calculations to obtain the solution to the quadratic optimization problem;

[0053] Step 5: The controller of the three-wheeled omnidirectional mobile manipulator adjusts the mobile platform and the manipulator simultaneously according to the calculation results so that they accurately reach the desired posture (i.e., the initial posture specified when performing different operation tasks).

[0054] The performance indicators of the acceleration layer derived from the above neurodynamic design formula are described as follows:

[0055]

[0056] Among them, ||·||2 represents the vector's two-norm, and represents the joint acceleration vector and joint velocity vector of the three-wheeled omnidirectional mobile manipulator, and Represent the acceleration and velocity of the robot arm joints, and They represent the rotational angular acceleration and rotational angular velocity of the omnidirectional driving wheels of the mobile platform, n represents the number of degrees of freedom of the manipulator installed on the mobile platform, and 3 represents the number of omnidirectional driving wheels of the mobile platform; M = BA∈R (3+n)×(3+n) is the coefficient matrix; is the coefficient matrix, and e is the augmented matrix, which represents the error between the current posture and the desired posture. n ]∈R (3 +n)×(3+n) , A=[H,0;0,I n ]∈R (n+3)×(n+3) ,H∈R 3×3 is the structural parameter coefficient matrix of the mobile platform, and H is expressed as follows:

[0057]

[0058] Among them, r>0∈R represents the radius of the omnidirectional driving wheel of the mobile platform, l>0∈R represents the distance from the center point of the mobile platform to the omnidirectional driving wheel, φ∈R represents the heading angle of the mobile platform, and I n ∈R n×n Represents the identity matrix. represents the time derivative of A, And k i and λ i (i=1,2,3) are the design parameters for achieving attitude adjustment; e=[λ1k1(p xy -p xyd );λ2k2(sin(φ)-sin(φ d ));λ3k3(θ-θ d )] and p xy represents the coordinates of the mobile platform along the XY axis on the horizontal ground (i.e., the position where the robotic arm is fixed on the mobile platform), φ∈R represents the orientation angle of the mobile platform, and p xyd ,φ d and θ d They represent the desired coordinates and orientation angles of the mobile platform along the XY axis on the horizontal ground, and the desired angles of the robotic arm joints. Then, correspondingly, the augmented vector F d =[p xyd ;φ d θ d ]∈R 3 +n represents the desired pose of the three-wheeled omnidirectional mobile manipulator (i.e., the initial pose when performing different manipulation tasks).

[0059] On the basis of minimizing the above performance indicators, combined with the kinematic equations of the mobile platform acceleration layer and the physical limits of the three-wheeled omnidirectional mobile manipulator, the corresponding three-wheeled omnidirectional mobile manipulator acceleration layer attitude adjustment model is established as follows:

[0060] Minimize:

[0061] Constraints:

[0062]

[0063]

[0064]

[0065] θ - ≤θ≤θ + (7)

[0066]

[0067]

[0068] Among them, the equality constraint The kinematic equations representing the acceleration layer of the mobile platform, H∈R 3×3 is the structural parameter coefficient matrix of the mobile platform, is the derivative of H. and are the driving wheel rotation angle, driving wheel rotation angular velocity and driving wheel rotation angular acceleration, respectively. and Respectively represent their corresponding limits. θ、 and Represent the manipulator joint angle, robot joint velocity and robot joint acceleration respectively, θ ± 、 and Represent their corresponding limits respectively.

[0069] To facilitate the calculation description, we further define The above acceleration layer attitude adjustment model (2)-(9) can be transformed into the following quadratic optimization problem:

[0070] Minimize: x T Qx / 2+p T x (10)

[0071] Constraints: x - ≤x≤x + (11)

[0072] Among them, the superscript Trepresents the transpose of a matrix or vector, represents the decision variable of the quadratic optimization problem; x ± represents the limit of x, and, η>0∈R and γ>0∈R represent the limit conversion parameters, ε>0∈R 3+n Indicates the limit conversion margin; Q = M T M∈R (3+n)×(3+n) , M=BA∈R (3+n)×(3+n) , p=M T q∈R 3+n ,

[0073] The quadratic optimization problem converted from the above-mentioned acceleration layer attitude adjustment model is iteratively calculated using a numerical algorithm. The specific operation is to convert the solution of the quadratic optimization problem into the solution of a piecewise linear projection equation, so that the numerical algorithm is used to obtain the optimal solution of the optimization problem, namely the rotation angle, rotation angular velocity, rotation angular acceleration of the omnidirectional drive wheels of the mobile platform, and the angle, velocity and acceleration of the robotic arm joints.

[0074] Specifically, solving the above quadratic optimization problems (10)-(11) can be achieved by solving the following piecewise linear projection equation:

[0075] F Ω (x-(Qx+p))-x=0∈R 3+n (12)

[0076] Among them, F Ω (·) denotes the piecewise linear projection operator.

[0077] For the piecewise linear projection equation (12), the following numerical algorithm can be used to solve it:

[0078] e(x k )=x k -F Ω (x k -(Qx k +p)) (13)

[0079] x k+1 =F Ω (x k -α(x k )β(x k )) (14)

[0080] β(x k )=Q T e(x k )+Qx k +p (15)

[0081]

[0082] Where, the superscript k represents the number of iterations and k=0,1,2,…. Given an initial value x 0 ∈R 3+n , through the iterative calculation of numerical algorithms (13)-(16), the numerical solution of the piecewise linear projection equation (12) can be obtained, thereby obtaining the optimal solution of the quadratic optimization problem (10)-(11), that is, the optimal solution of the acceleration layer attitude adjustment model (2)-(9) of the three-wheeled omnidirectional mobile manipulator, including the rotation angle, rotation angular velocity, rotation angular acceleration of the mobile platform driving wheel and the angle, velocity and acceleration of the manipulator joint.

[0083] According to the calculation results, the three-wheeled omnidirectional mobile robot arm controller drives the three omnidirectional wheels of the mobile platform and the various joints of the robot arm in real time, enabling the three-wheeled omnidirectional mobile robot arm to quickly and accurately adjust from the current posture to the desired posture (as the starting posture for different operation tasks), completing efficient automatic adjustment between different postures.

[0084] The above describes in detail the preferred embodiments of the present invention. It should be understood that those skilled in the art can make numerous modifications and variations based on the concepts of the present invention without inventive effort. Therefore, any technical solutions that can be derived by those skilled in the art through logical analysis, reasoning, or limited experimentation based on the concepts of the present invention and the prior art should be within the scope of protection defined by the claims.

Claims

1. A method for adjusting the posture of a three-wheeled omnidirectional mobile manipulator in an acceleration layer using neural dynamics, wherein the three-wheeled omnidirectional mobile manipulator comprises a mobile platform having a plurality of omnidirectional drive wheels and a manipulator disposed on the mobile platform, characterized in that: The method steps include: Obtain the state parameters of the three-wheeled omnidirectional mobile manipulator and input the state parameters into the pre-built acceleration layer attitude adjustment model; Solve the acceleration layer attitude adjustment model to obtain the rotation angle, angular velocity and angular acceleration of the mobile platform's drive wheels, as well as the angle, velocity and acceleration of the robotic arm's joints; According to the calculation results, the mobile platform and the robotic arm are adjusted simultaneously to reach the desired posture; The acceleration layer attitude adjustment model is established based on the performance indicators described on the acceleration layer, combined with the kinematic equations of the mobile platform acceleration layer and the physical limits of the three-wheeled omnidirectional mobile manipulator; the performance indicators described on the acceleration layer are derived using neurodynamic design formulas based on the idea of ​​zeroing the error between the current attitude of the mobile manipulator and the desired attitude; the performance indicators described on the acceleration layer are described as follows: in, represents the two-norm of the vector, is the coefficient matrix; , is the coefficient matrix, is the augmented matrix, which represents the error between the current posture and the desired posture, and represents the joint acceleration vector and joint velocity vector of the three-wheeled omnidirectional mobile manipulator, and Represent the acceleration and velocity of the robot arm joints, and They represent the rotational angular acceleration and rotational angular velocity of the omnidirectional driving wheels of the mobile platform, represents the number of degrees of freedom of the robotic arm mounted on the mobile platform, Indicates the number of omnidirectional drive wheels of the mobile platform; The coefficient matrix middle: in, express The time derivative of represents the identity matrix, represents the heading angle of the mobile platform, and To achieve the design parameters of posture adjustment, is the structural parameter coefficient matrix of the mobile platform; The augmented matrix It is expressed as follows: in, Indicates the coordinates of the mobile platform along the XY axis on the horizontal ground. , and They represent the desired coordinates and orientation angles of the mobile platform along the X and Y axes on the horizontal ground, as well as the desired angles of the robotic arm joints.

2. The neural dynamic posture adjustment method of the acceleration layer of a three-wheeled omnidirectional mobile manipulator according to claim 1 is characterized in that: The mobile platform structural parameter coefficient matrix The expression is as follows: in, represents the radius of the omnidirectional drive wheel of the mobile platform, Indicates the distance from the center point of the mobile platform to the omnidirectional drive wheel. Indicates the heading angle of the mobile platform.

3. The neural dynamic posture adjustment method of the acceleration layer of a three-wheeled omnidirectional mobile manipulator according to claim 1 is characterized in that: The acceleration layer attitude adjustment model is constrained by the acceleration layer kinematic equations of the mobile platform, including the rotation angle limit, rotation angular velocity limit and rotation angular acceleration limit of the omnidirectional drive wheel and the angle limit, velocity limit and acceleration limit of the robot arm joint.

4. The method for adjusting the posture of a three-wheeled omnidirectional mobile manipulator using a neural dynamics acceleration layer according to claim 1, characterized in that: The specific steps of solving the acceleration layer attitude adjustment model are: Convert the acceleration layer attitude adjustment model into a quadratic optimization problem; The solution of quadratic optimization problem is equivalently converted into the solution of piecewise linear projection equation; A numerical algorithm is used to obtain the optimal solution to the quadratic optimization problem, including the rotation angle, rotation angular velocity, rotation angular acceleration of the omnidirectional drive wheels of the mobile platform, as well as the angle, velocity and acceleration of the robotic arm joints.

5. The method for adjusting the posture of a three-wheeled omnidirectional mobile manipulator using a neural dynamics acceleration layer according to claim 4, characterized in that: The acceleration layer attitude adjustment model is transformed into the following quadratic optimization problem: Minimize: Constraints: Among them, the superscript represents the transpose of a matrix or vector, Represents the decision variables of the quadratic optimization problem; express The limit, and , , and represents the limit conversion parameter, Indicates the limit conversion margin; , , , ; 、 and are the augmented angle vector, velocity vector, and acceleration vector of the three-wheeled omnidirectional mobile manipulator respectively; accordingly, 、 and Respectively 、 and limit.

6. The method for adjusting the posture of a three-wheeled omnidirectional mobile manipulator using a neural dynamics acceleration layer according to claim 5, characterized in that: The piecewise linear projection equation is specifically as follows: in, represents the piecewise linear projection operator.

7. The method for adjusting the posture of a three-wheeled omnidirectional mobile manipulator using a neural dynamics acceleration layer according to claim 6, characterized in that: The piecewise linear projection equation is solved using the following numerical algorithm: Among them, the superscript represents the number of iterations and .

Images