Robot friction torque online prediction method and device, and computer readable storage medium

Abstract

The invention discloses a robot friction torque online prediction method and device and a computer readable storage medium, for a to-be-identified joint type series industrial robot, according to the embodiment of the invention, a robot kinetic model considering friction is established through a Newton-Euler method, and after the kinetic model is linearized, the friction torque of the to-be-identified joint type series industrial robot is predicted. Designing an excitation track by using five finite Fourier series, collecting joint motion data of the robot under the excitation track to establish a regression equation set, solving the regression equation set to obtain an integral kinetic parameter to be identified, and calculating the friction torque of the robot on line according to the integral kinetic parameter to be identified. The friction torque of the robot can be predicted on line.

Description

Technical Field

[0001] The present invention relates to the field of robot technology, and in particular to a method and device for online prediction of robot friction torque, and a computer-readable storage medium. Background Art

[0002] With the development of science and technology, industrial robots have been widely used in many industries, greatly improving productivity. In the fields of 3C manufacturing, human-machine collaboration, and medical services, the application of robotic arms has become more extensive and in-depth. At the same time, people have higher and higher performance requirements for industrial robots, requiring them to achieve more complex functions, which has led to a continuous increase in the requirements for control accuracy, collision detection and other capabilities of robotic arms. Accurate dynamic models are the cornerstone of high-precision trajectory control and external force perception of robots. Usually, the dynamic parameters of robots include the friction parameters of robot joints. However, the joint friction of robots has nonlinear characteristics, and it is difficult to accurately identify the friction torque. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a method and device for online prediction of the friction torque of a robot, and a computer-readable storage medium, which can realize online prediction of the friction torque of the robot.

[0004] In order to solve the above technical problems, the first aspect of the present invention discloses a robot friction torque online prediction method, which is applied to an articulated serial industrial robot. The method comprises the following steps:

[0005] S1. Establish the dynamic model of the robot considering friction by Newton-Euler method:

[0006]

[0007] Where: q, are the position, velocity, and acceleration vectors of the robot joints, M(q) is the inertia matrix of the robot arm, is the velocity term matrix related to centrifugal force and Coriolis force, g(q) is the gravity term, τ is the joint torque vector, represents friction / torque;

[0008] S2. Linearize the kinetic model to obtain the following kinetic expression:

[0009]

[0010] in, is the regression matrix, β b is the minimum inertial parameter set of the dynamic parameters to be identified; for is the regression matrix related to the inertial basis parameters, Regression matrix related to friction / torque; parameter set to be identified β b It is divided into two subsets, one of which is related to the inertial basic parameter, denoted as β ib , and the other is related to friction / torque, denoted as β f , β f =[f ci f vi b i ] T ;

[0011] S3. Use Fourier series to design the excitation trajectory, which is defined as:

[0012]

[0013]

[0014]

[0015] Among them, q i (t) and They represent the position, angular velocity and angular acceleration of the joint i at time t, respectively, and q i0 is the joint angle at the initial moment; ω f is the fundamental frequency of the Fourier series excitation trajectory, a k and b k are the coefficients of the Fourier series excitation trajectory;

[0016] S4. Control the robot to run according to the excitation trajectory, collect the motion data of each joint of the robot, and combine the linear regression equations:

[0017]

[0018]

[0019]

[0020] Wherein, n is the number of dynamic data sets of the robot joints acquired when the robot moves according to the excitation trajectory, and each of the dynamic data sets includes the input torque, joint angle, joint angular velocity and joint angular acceleration of all joints of the robot;

[0021] S5. Solve the linear regression equations to determine the parameter set β to be identified b ;

[0022] S6. Identification parameter set β obtained by calculation b , calculate the friction torque τ at each moment f :

[0023]

[0024] S7. The friction torque τ f Split into the friction observation matrix Φ for each joint i fi and the corresponding friction parameter β fi The product of

[0025] τ fi =Φ fi β fi ;

[0026] S8. Collect the friction torque τ of each joint i at the current k moment fi (k) and joint motion information q i (k) Perform online friction estimation:

[0027] S81. Calculate the gain coefficient K of joint i i (k):

[0028]

[0029] Among them, Rv=0.999 is the observation noise variance, and P is the covariance matrix;

[0030] S82. Calculate the covariance matrix P of joint i i (k):

[0031] P i (k) = (1-K i (k)Φ fi (k))P i (k-1)+R w ;

[0032] in, is the covariance matrix of process noise;

[0033] S83. Calculate the friction parameters of joint i:

[0034] β fi (k) = β fi (k-1)+K i (k)(τ fi (k)-Φ fi (k)β fi (k-1))

[0035] S84. Use the updated friction parameter to obtain the friction torque of joint i at this moment:

[0036] τ fi (k) = Φ fi(k)β fi .

[0037] As an optional method, the linear regression equations are solved to determine the parameter set β to be identified. b ,include:

[0038] The least squares identification method under physical consistency constraints is used to solve the linear regression equations to determine the parameter set β to be identified. b .

[0039] The second aspect of the present invention discloses an online prediction device for robot friction torque, comprising a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps in the method disclosed in the first aspect of the present invention.

[0040] A third aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the method disclosed in the first aspect of the present invention.

[0041] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0042] Compared with the prior art, for an articulated serial industrial robot to be identified, an embodiment of the present invention establishes a dynamic model of the robot taking friction into consideration through the Newton-Euler method, and then linearizes the dynamic model, uses five finite Fourier series to design an excitation trajectory, collects the joint motion data of the robot under the excitation trajectory to establish a regression equation group, solves the regression equation group to obtain the overall dynamic parameters to be identified, and calculates the friction torque of the robot online based on the overall dynamic parameters to be identified, thereby realizing online prediction of the friction torque of the robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0044] Figure 1 It is a flow chart of a method for online prediction of friction torque of a robot disclosed in an embodiment of the present invention;

[0045] Figure 2 It is another flow chart of a method for online prediction of friction torque of a robot disclosed in an embodiment of the present invention;

[0046] Figure 3It is a structural schematic diagram of a robot friction torque online prediction device disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0047] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] Embodiment 1

[0049] See also Figures 1-2 The embodiment of the present invention discloses a method for online prediction of friction torque of a robot, comprising the following steps:

[0050] S1. Establish the dynamic model of the robot considering friction by Newton-Euler method:

[0051]

[0052] Where: q, are the position, velocity, and acceleration vectors of the robot joints, M(q) is the inertia matrix of the robot arm, is the velocity term matrix related to centrifugal force and Coriolis force, g(q) is the gravity term, τ is the joint torque vector, Represents friction / torque.

[0053] S2. Linearize the kinetic model and obtain the following kinetic expression:

[0054]

[0055] in, is the regression matrix, β b It is the minimum inertia parameter set including the basic inertia parameters and friction parameters of each connecting rod waiting for the dynamic parameters to be identified.

[0056] It can be expressed as is the regression matrix related to the inertial basis parameters, The regression matrix related to friction / torque, Parameter set to be identified β b It can be divided into two subsets, one of which is related to the inertial basic parameter, which can be recorded as β ib , and the other is related to friction / torque and can be recorded as β f ,Right now where β f=[f ci f vi b i ] T , f c is the Coulomb friction coefficient, f v is the viscous friction coefficient, b is the friction bias; β ib These include center of mass location, motor inertia, mass, link length, joint coordinate system offset, and moment of inertia.

[0057] S3. Use Fourier series to design the excitation trajectory, which is defined as:

[0058]

[0059]

[0060]

[0061] Among them, q i (t) and They represent the position, angular velocity and angular acceleration of the joint i at time t, respectively, and q i0 is the joint angle at the initial moment, which can also be called the initial offset of the joint. f is the fundamental frequency of the Fourier series excitation trajectory, a k and b k are the coefficients of the Fourier series excitation trajectory and are the parameters to be determined.

[0062] The purpose of the excitation trajectory is to generate as much data as possible by driving the robot to move, which helps to accurately identify the parameters of the dynamic model. Fourier series is an ideal choice in the design of excitation trajectories, especially for the parameter identification of robots with periodic motion. The use of Fourier series helps to generate smooth, low-noise, dynamic-rich trajectories, while allowing parameters to be adjusted through optimization to meet specific needs.

[0063] S4. Control the serial robot to run according to the excitation trajectory and collect joint motion data. Assuming that n sets of joint motion data are collected, the combined linear regression equation group is:

[0064]

[0065]

[0066]

[0067] Among them, n is the number of dynamic data groups of the robot joints obtained when the robot moves according to the excitation trajectory, and each of the dynamic data groups contains the input torque, joint angle, joint angular velocity and joint angular acceleration of all joints of the robot.

[0068] S5. Solve the linear regression equations to determine the parameter set β to be identified b ; The parameter set to be identified β b As fixed parameters of the robot for subsequent position control;

[0069] S6. Identification parameter set β obtained by calculation b , calculate the friction torque τ at each moment f :

[0070]

[0071] S7. The friction torque τ f Split into the friction observation matrix Φ for each joint i fi and the corresponding friction parameter β fi The product of

[0072] τ fi =Φ fi β fi ;

[0073] S8. Collect the friction torque τ of each joint i at the current k moment fi (k) and joint motion information q i (k) Perform online friction estimation:

[0074] S81. Calculate the gain coefficient K of joint i i (k):

[0075]

[0076] Among them, Rv=0.999 is the observation noise variance, and P is the covariance matrix;

[0077] S82. Calculate the covariance matrix P of joint i i (k):

[0078] P i (k) = (1-K i (k)Φ fi (k))P i (k-1)+R w ;

[0079] in, is the covariance matrix of process noise;

[0080] S83. Calculate the friction parameters of joint i:

[0081] β fi (k) = β fi (k-1)+K i (k)(τ fi (k)-Φ fi (k)β fi (k-1))

[0082] S84. Use the updated friction parameter to obtain the friction torque of joint i at this moment:

[0083] τ fi (k) = Φ fi (k)β fi .

[0084] Optionally, the least squares identification method under physical consistency constraints can be used to solve the problem. Solve the parameter set to be identified β b After that, the friction parameter β can be separated f The friction resistance is brought into the dynamic model for subsequent calculation of the motion trajectory to achieve subsequent power control.

[0085] Compared with the prior art, for an articulated serial industrial robot to be identified, an embodiment of the present invention establishes a dynamic model of the robot taking friction into consideration through the Newton-Euler method, and then linearizes the dynamic model, uses five finite Fourier series to design an excitation trajectory, collects the joint motion data of the robot under the excitation trajectory to establish a regression equation group, solves the regression equation group to obtain the overall dynamic parameters to be identified, and calculates the friction torque of the robot online based on the overall dynamic parameters to be identified, thereby realizing online prediction of the friction torque of the robot.

[0086] Embodiment 2

[0087] See also Figure 3 , Figure 3 It is a structural schematic diagram of a robot friction torque online prediction device disclosed in an embodiment of the present invention, including a memory 201, a processor 202 and a computer program stored in the memory, and the processor executes the computer program to implement the steps in the method disclosed in Example 1.

[0088] Embodiment 3

[0089] An embodiment of the present invention discloses a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the method disclosed in the first embodiment are implemented.

[0090] The contents disclosed in the embodiments of the present invention only disclose the preferred embodiments of the present invention, which are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A robot friction torque online prediction method, characterized in that: The method comprises the following steps: S1. Establish the dynamic model of the robot considering friction through the Newton-Euler method: Where: q, are the position, velocity, and acceleration vectors of the robot joints, M(q) is the inertia matrix of the robot arm, is the velocity term matrix related to centrifugal force and Coriolis force, g(q) is the gravity term, τ is the joint torque vector, represents friction / torque; S2. Linearize the kinetic model to obtain the following kinetic expression: in, is the regression matrix, β b is the minimum inertial parameter set of the dynamic parameters to be identified; for is the regression matrix related to the inertial basis parameters, Regression matrix related to friction / torque; parameter set to be identified β b It is divided into two subsets, one of which is related to the inertial basic parameter, denoted as β ib , and the other is related to friction / torque, denoted as β f , β f =[f ci f vi b i ] T ; S3. Use Fourier series to design the excitation trajectory, which is defined as: Among them, q i (t) and They represent the position, angular velocity and angular acceleration of the joint i at time t, respectively, and q i0 is the joint angle at the initial moment; ω f is the fundamental frequency of the Fourier series excitation trajectory, a k and b k are the coefficients of the Fourier series excitation trajectory; S4. Control the robot to run according to the excitation trajectory, collect the motion data of each joint of the robot, and combine the linear regression equations: Wherein, n is the number of dynamic data sets of the robot joints acquired when the robot moves according to the excitation trajectory, and each of the dynamic data sets includes the input torque, joint angle, joint angular velocity and joint angular acceleration of all joints of the robot; S5. Solve the linear regression equations to determine the parameter set β to be identified b ; S6. Identification parameter set β obtained by calculation b , calculate the friction torque τ at each moment f : S7. The friction torque τ f Split into the friction observation matrix Φ for each joint i fi and the corresponding friction parameter β fi The product of: t fi =Φ fi b fi ; S8. Collect the friction torque τ of each joint i at the current k moment fi (k) and joint motion information q i (k) Perform an online estimate of friction, where: S81. Calculate the gain coefficient K of joint i i (k): Among them, Rv=0.999 is the observation noise variance, and P is the covariance matrix; S82. Calculate the covariance matrix P of joint i i (k): P i (k)=(1-K i (k)Φ fi (k))P i (k-1)+R w ; in, is the covariance matrix of process noise; S83. Calculate the friction parameters of joint i: b fi (k)=β fi (k-1)+K i (k)(t) fi (k)-Φ fi (k)b fi (k-1)) S84. Use the updated friction parameter to obtain the friction torque of joint i at this moment: t fi (k)=Φ fi (k)b fi 。 2. The robot friction torque online prediction method according to claim 1, characterized in that: The linear regression equations are solved to determine the parameter set β to be identified. b ,include: The least squares identification method under physical consistency constraints is used to solve the linear regression equations to determine the parameter set β to be identified. b .

3. A robot friction torque online prediction device, comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to claim 1.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method as claimed in claim 1 are implemented.

Images