Search

CN-121973214-A - Identification method for six-degree-of-freedom robot end load parameters

CN121973214ACN 121973214 ACN121973214 ACN 121973214ACN-121973214-A

Abstract

The invention relates to the technical field of industrial robot load identification, in particular to an identification method for six-degree-of-freedom robot tail end load parameters, which comprises the steps of establishing a six-degree-of-freedom serial robot dynamics model, reconstructing a load minimum parameter set, and determining that identification can be completed only by exciting a third joint, a fifth joint and a sixth joint; the method comprises the steps of designing four groups of symmetrical excitation tracks combined by Fourier series and penta polynomials, performing discrete sampling to generate an observation matrix, optimizing a load parameter set and a regression matrix, reducing the condition number of the observation matrix, obtaining a load minimum parameter set through differential calculation of no-load and loaded identification results, and solving the mass, the barycenter coordinates, the moment of inertia and the product of inertia of a load based on the parameter set. The method solves the problem of regression matrix morbidity by reconstructing the minimum parameter set, reduces the number of excitation joints, does not need to rely on a robot body dynamics model, and has the advantages of high identification precision, strong stability, low experimental complexity and high efficiency.

Inventors

  • HUANG JIE
  • WANG YONGSHANG
  • FENG LIANGYOU
  • YUAN YE

Assignees

  • 无锡信捷电气股份有限公司

Dates

Publication Date
20260505
Application Date
20260204

Claims (10)

  1. 1. The identification method for the six-degree-of-freedom robot end load parameters is characterized by comprising the following steps of: Establishing a six-degree-of-freedom serial robot dynamics model, constructing a minimum parameter set, a load minimum parameter set and a corresponding regression matrix of a robot without friction based on a linear superposition principle, determining parameters of all the minimum parameter sets by only driving a third joint, a fifth joint and a sixth joint through analyzing the relevance of the joints and the load parameters, and obtaining a new regression matrix and a load parameter set after eliminating zero items of the regression matrix; Designing excitation tracks for a third joint, a fifth joint and a sixth joint, wherein the excitation tracks adopt a form of combination of Fourier series and a fifth polynomial, four groups of mutually symmetrical excitation sequences are constructed, and discrete sampling is performed in a preset time interval to generate an observation matrix; Optimizing a load parameter set and a regression matrix, solving the integral of a specific item in the minimum parameter set and the regression matrix to other items, deleting redundant columns of the regression matrix, and obtaining a simplified load parameter set and a new regression matrix; Acquiring pose, speed and acceleration information of a third joint, a fifth joint and a sixth joint of the robot in no-load and loaded states, and obtaining a load minimum parameter set through differential calculation of two identification results; And (5) solving the mass, the barycenter coordinates, the moment of inertia and the product of inertia of the load through a preset formula based on the minimum load parameter set.
  2. 2. The method for identifying six degree of freedom robot end load parameters according to claim 1, wherein in step (1), the six degree of freedom serial robot dynamics model is: Wherein the method comprises the steps of For a positive definite inertia matrix, the inertia matrix is determined, For the centrifugal force and the coriolis force matrix, The force vector of the gravity is used to determine, Is the friction moment vector of the joint, Is a driving moment vector of the joint, As the angle vector of the joint, As a vector of the angular velocity of the joint, Is the angular acceleration vector of the joint; The load dynamics equation is: 。
  3. 3. The method of claim 2, wherein in step (1), when eliminating the zero term of the regression matrix, the method comprises the steps of New regression matrix Satisfies the following conditions Corresponding load parameter set Wherein , 。
  4. 4. The method for identifying six-degree-of-freedom robot end load parameters according to claim 1, wherein in step (2), the expression of the excitation trajectory is: Wherein the method comprises the steps of , 、 As the amplitude parameter of the fourier term, As an order of the fourier series, In order to be of an angular frequency, For the order of the harmonic sequence numbers, Is a fifth order polynomial term for ensuring the continuity of the track boundary.
  5. 5. The method of claim 4, wherein the four mutually symmetrical excitation sequences are generated by applying the third and fifth amplitude parameters to the four mutually symmetrical excitation sequences Is designed by adopting positive and negative symmetry, respectively )、( )、( )、( ) The amplitude parameter of the sixth axis trajectory remains unchanged.
  6. 6. The method of claim 1, wherein in step (2), the observation matrix is used for identifying the six-degree-of-freedom robot end load parameters Wherein M is the number of sampling points, Is the end of the preset time interval.
  7. 7. The method according to claim 1, wherein in the step (3), the specific term is a fourth term of a minimum parameter set Simplified load parameter set The first two terms of (2) satisfy the following formula: the simplified identification equation is as follows: Wherein, the , 。
  8. 8. The method for identifying a six degree of freedom robot end load parameter according to claim 7, wherein in step (5), the solution formula of the load mass is: 。
  9. 9. the method for identifying a six-degree-of-freedom robot end load parameter according to claim 1, wherein in the step (5), the solution formula of the load centroid coordinates is: Wherein the method comprises the steps of Is a structural parameter of a connecting rod at the tail end of the robot.
  10. 10. The method for identifying a six-degree-of-freedom robot end load parameter according to claim 1 or 9, wherein in the step (5), the solution formula of the load moment of inertia and the product of inertia is: 。

Description

Identification method for six-degree-of-freedom robot end load parameters Technical Field The invention relates to the technical field of industrial robot load identification, in particular to an identification method for six-degree-of-freedom robot tail end load parameters, which is suitable for industrial robot operation scenes such as welding, assembly, polishing, carrying and the like, in which tail end tools or workpieces need to be frequently replaced, and can accurately identify key kinetic parameters such as mass, mass center position, rotational inertia and the like of the tail end load. Background Industrial robots are often required to mount different types of tools or jigs on end effectors and to handle workpieces of various qualities and shapes during actual work in many operations such as welding, assembly, grinding, and handling in modern industrial production. The parameters such as the mass, the gravity center position, the moment of inertia and the like of the tail end load directly determine the dynamic characteristics of the robot, so that the accuracy of the load parameters has a critical influence on the motion control precision, the dynamic performance and the energy consumption efficiency of the robot. The existing six-degree-of-freedom mechanical arm load parameter identification technology is generally used for parameter estimation based on a robot dynamics model. The basic idea is to construct a series robot dynamics equation by measuring joint angles, angular velocities, angular accelerations and joint driving moments of the mechanical arm in the motion process: (1) Wherein, the For a positive definite inertia matrix, the inertia matrix is determined,For the centrifugal force and the coriolis force matrix,The force vector of the gravity is used to determine,Is the friction moment vector of the joint,Is a driving moment vector of the joint,As the angle vector of the joint,As a vector of the angular velocity of the joint,Is the angular acceleration vector of the joint. Considering the complexity of the joint friction model, the joint friction moment can be expressed as follows, when the joint friction model is usually modeled only by coulomb friction and viscous friction during identification:(2) Wherein, the Is a jointIs used for the production of a high-density polyethylene,Is a jointIs a viscous coefficient of friction; Is a joint So the robot dynamics equation can be rewritten as:(3) Wherein, the Kinetic parameters expressed as connecting rod i: (4) Wherein, the Is a connecting rodIs used for the quality of the (a),Respectively connecting rodsAt the position ofThe centroid in the coordinate system is the centroid,Is a connecting rod atMoment of inertia and product of inertia in the coordinate system. Based on the linear relationship between the kinetic parameters and the joint moment, equation (3) can be rewritten as follows: (5) theoretically if the moment is fed back in real time Regression matrixThe robot dynamics parameters can be identified by a weighted least square method, which is known. However, in the actual robot model, there is strong coupling and linear dependence between different parameters, that is, a disease state matrix occurs when constructing the least squares equation, which means that all parameters cannot be identified. Therefore, the minimum parameter set needs to be constructed, the regression matrix is simplified, and the regression matrix can be obtained after finishing: (6) Wherein, the After being recombinedX 1 of the column vectors of the inertial parameters,And representing the coefficient matrix corresponding to the recombined inertia parameters. For a load dynamics model of the robot, the load can be regarded as a part of the tail end connecting rod, a coordinate system of the load is established at the center of the tail end flange, and a robot load moment expression can be obtained by the formula (6) according to the linear superposition principle: (7) the simplification is as follows: (8) In order to obtain a unique solution for parameter identification, the jacobian matrix needs to be full of rank, so that the traditional method generally needs to drive a third joint, a fourth joint, a fifth joint and a sixth joint to finish load parameter identification. At present, a plurality of six-degree-of-freedom robot end load parameter identification schemes exist in the prior art, but the defects of the schemes are all present. The robot load identification method disclosed in a certain patent determines the priority of load dynamics identification and sequentially identifies the load dynamics, but the method needs multiple times of operation and takes long time; the robot load identification and compliance control research disclosed in a paper reduces model errors by calibrating the relation between current and conversion moment through static teaching points, finally obtains load dynamics parameters through least square identification, but t