Optimization method of robot kinetic model and related equipment

By constructing momentum observer and optimizing dynamic model with residual vectors, the problem of insufficient recognition accuracy of robot dynamic model in the prior art is solved, the recognition and fitting ability of the model are improved, and more accurate motion trajectory prediction is achieved .

CN119057785APending Publication Date: 2024-12-03HUILING TECH ROBOTIC CO LTD

Patent Information

Application Number
CN202411260618.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-10
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

When describing the relationship between robot motion and the force and moment, the existing robot dynamics model has high computational efficiency but insufficient recognition accuracy, resulting in the excitation trajectory velocity and acceleration far below the motion limit, making it difficult to find the best balance point.

Method used

Based on the antisymmetric properties of generalized momentum method and dynamic model, a momentum observer is constructed, and the residual vector output by the momentum observer is transformed and optimized, and the dynamic model parameters are adjusted to improve the recognition accuracy.

Benefits of technology

By optimizing the dynamic model, the recognition accuracy of the robot's motion trajectory is improved, the fitting ability of the model is enhanced, making the predicted value of the dynamic model closer to the actual observed value, and the recognition ability of the system is improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119057785A_ABST
    Figure CN119057785A_ABST
Patent Text Reader

Abstract

The embodiment of the invention discloses a robot kinetic model optimization method and related equipment, and the method comprises the steps: constructing a momentum observer according to the generalized momentum of a generalized momentum method and the antisymmetric property of a kinetic model; based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint and the actual first velocity vector of the joint, obtaining a first residual vector output by the momentum observer; performing Laplace transformation on the first residual vector to obtain a second residual vector of each joint of the robot; and optimizing the kinetic model by using the second residual vector. The residual vector is the difference between the actual observation value and the model prediction value, and the dynamic model is optimized through the residual vector, so that the model parameters can be continuously adjusted, the model prediction value is closer to the actual observation value, and the identification precision of the model is improved.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to the technical field of robot dynamics models, and in particular to an optimization method for a robot dynamics model and related equipment. [Background technology]

[0002] The robot dynamics model takes into account multiple factors such as the robot's mass, inertia, joint friction, gravity, etc., and describes the relationship between the acceleration, velocity and position of each joint of the robot and the force and torque applied to the joint through a set of differential equations. Although the existing dynamics model has high computational efficiency, it may not accurately reflect all the characteristics of the system. The excitation trajectory obtained by using the existing dynamics model has inaccurate identification accuracy because the speed and acceleration are far below the robot's motion limit. This contradiction makes it difficult to find the best balance point in practical applications. [Summary of the invention]

[0003] In view of this, the present invention provides a method for optimizing a robot dynamics model and related equipment.

[0004] The specific technical solution of the first embodiment of the present invention is: a method for optimizing a robot dynamics model, wherein the dynamics model is a mathematical model of the relationship between the robot's motion and the forces and torques it is subjected to, and the method comprises: constructing a momentum observer according to the generalized momentum of the generalized momentum method and the antisymmetric properties of the dynamics model; obtaining a first residual vector output by the momentum observer based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint, and the actual first velocity vector of the joint; performing a Laplace transform on the first residual vector to obtain a second residual vector for each joint of the robot; and optimizing the dynamics model using the second residual vector.

[0005] Preferably, the first residual vector output by the momentum observer is obtained based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint and the actual first velocity vector of the joint, including: obtaining a third residual vector based on the first difference between the momentum observer and the generalized momentum, and the observer diagonal gain matrix; integrating the third residual vector to obtain a fourth residual vector; and obtaining the first residual vector based on the fourth residual vector, the actual first inertia matrix of the joint, the actual first velocity vector of the joint and the momentum observer.

[0006] Preferably, the obtaining of the first residual vector based on the fourth residual vector, the actual first inertia matrix of the joint, the actual first velocity vector of the joint and the momentum observer includes: substituting the actual first inertia matrix of the joint and the actual first velocity vector of the joint into the fourth residual vector to obtain a fifth residual vector; substituting the transposed matrix of the joint torque, friction, centrifugal force and Coriolis force and the second inertia matrix observed by the momentum observer into the fifth residual vector to obtain the first residual vector.

[0007] Preferably, the generalized momentum includes the resultant force actively applied to the joint, the torque applied to the joint by the external force, the friction force, the centrifugal force and the Coriolis force matrix and the velocity vector of the joint.

[0008] Preferably, the fourth residual vector is obtained in the following manner: obtaining a second difference between the momentum at the moment of movement and the momentum at the initial moment; obtaining a third difference between the first difference and the observation value of the momentum observer; and obtaining the fourth residual vector based on the third difference and the observer diagonal gain matrix.

[0009] Preferably, the first residual vector is obtained using the following formula:

[0010]

[0011] in, is the derivative of the first residual vector, K0 is the observer diagonal gain matrix, τ ext is the torque applied to the joint by the external force.

[0012] Preferably, the second residual vector is obtained using the following formula:

[0013]

[0014] Among them, r i is the second residual vector of the i-th joint, K 0,i is the observer diagonal gain matrix of the i-th joint, s is the Laplace transform, τ ext,i is the torque applied by the external force to the i-th joint, T 0,i =1 / K 0,i , T 0,i is the time constant.

[0015] The specific technical solution of the second embodiment of the present invention is: a robot dynamic model optimization system, the system comprising: an observer construction module, a residual output module, a Laplace transform module and an optimization module; the observer construction module is used to construct a momentum observer according to the generalized momentum of the generalized momentum method and the antisymmetric properties of the dynamic model; the residual output module is used to obtain the first residual vector output by the momentum observer based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint and the actual first velocity vector of the joint; the Laplace transform module is used to perform a Laplace transform on the first residual vector to obtain a second residual vector for each joint of the robot; the optimization module is used to optimize the dynamic model using the second residual vector.

[0016] The specific technical solution of the third embodiment of the present invention is: a robot dynamics model optimization device, including a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method described in any one of the first embodiments of the present application.

[0017] The specific technical solution of the fourth embodiment of the present invention is: a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to perform the steps of the method described in any one of the first embodiments of the present application.

[0018] Implementing the embodiments of the present invention will have the following beneficial effects:

[0019] The present invention constructs a momentum observer based on the antisymmetric properties of the generalized momentum of the generalized momentum method and the dynamic model; obtains the first residual vector output by the momentum observer based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint and the actual first velocity vector of the joint; performs Laplace transformation on the first residual vector to obtain the second residual vector of each joint of the robot; and optimizes the dynamic model using the second residual vector. The residual vector is the difference between the actual observed value and the model predicted value. By optimizing the dynamic model through the residual vector, the model parameters can be continuously adjusted to make the model predicted value closer to the actual observed value, thereby improving the recognition accuracy of the model.

Brief Description of the Drawings

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1A flowchart of the steps of the optimization method for the robot dynamics model;

[0022] Figure 2 Flowchart of the steps for obtaining the residual vector based on the momentum observer and generalized momentum;

[0023] Figure 3 A flow chart of the steps for obtaining the residual vector based on the inertia matrix and the velocity vector;

[0024] Figure 4a This is the current simulation diagram of the first joint of the robot;

[0025] Figure 4b This is the current simulation diagram of the second joint of the robot;

[0026] Figure 4c This is the current simulation diagram of the third joint of the robot;

[0027] Figure 4d This is the current simulation diagram of the fourth joint of the robot;

[0028] Figure 5 is the block diagram of the momentum observer;

[0029] Figure 6 It is a schematic diagram of the structure of the optimization system of the robot dynamics model;

[0030] Figure 7 A diagram of the internal structure of a computer device;

[0031] Among them, 401 is an observer construction module; 402 is a residual output module; 403 is a Laplace transform module; and 404 is an optimization module. [Specific implementation method]

[0032] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0033] The terms "first", "second", etc. in the specification, claims and drawings of this application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or modules is not limited to the listed steps or modules, but optionally includes steps or modules that are not listed, or optionally includes other steps or modules that are inherent to these processes, methods, products or devices.

[0034] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0035] Since it was found in the experiment that the original excitation trajectory had inaccurate identification accuracy due to its speed and acceleration being far below the robot's motion limit, it was considered to improve the robot's dynamic model by optimizing the excitation trajectory.

[0036] The excitation trajectory is designed as a 6.25 second 30th-order Fourier series trajectory. The 30th-order Fourier series trajectory can mainly allow the robot to move at a high speed and high acceleration state, and is used to accurately represent the inertia.

[0037] There are two main purposes of torque calibration:

[0038] 1. Verify whether the optimized excitation trajectory reaches the expected torque peak;

[0039] 2. Whether the recognition accuracy of the optimized excitation trajectory is improved compared with the original excitation trajectory.

[0040] For 1., the robot can be actually run with the optimized excitation trajectory. After collecting the current and converting the torque, it can be determined whether the torque requirement is met.

[0041] For 2., the robot can be allowed to move to a random point, and the trajectory can be recorded. The joint torque can be calculated using the two sets of dynamic parameters before and after optimization, and compared with the actual joint torque to obtain the optimized experimental results.

[0042] After the excitation trajectory is optimized, the robot dynamics model can be optimized, and the collision detection based on the dynamics model will be more sensitive.

[0043] The residual vector usually refers to the residual vector in regression analysis, that is, the difference between the actual value of the dependent variable and the predicted value of the regression model. This residual vector is one of the important outputs of regression analysis and is used to measure the degree of fit and predictive ability of the model. For a dynamic model, if there is a difference between its predicted results (such as the future state or behavior of the system) and the actual observed results, these differences can be expressed in the form of a residual vector.

[0044] In the field of dynamics, dynamic models are used to describe the motion state of an object or system and how it changes over time. When the predictions of a dynamic model do not match the actual observations, the residual vector can be used to quantify the degree of this discrepancy. Specifically, the residual vector may represent the difference between the actual value of the system state variable and the model prediction, or the difference between the actual value of the system output (such as force, velocity, acceleration, etc.) and the model prediction.

[0045] See also Figure 1 , is a flowchart of a method for optimizing a robot dynamics model in the first embodiment of the present application, wherein the dynamics model is a mathematical model of the relationship between the robot motion and the forces and torques applied thereto, thereby improving the recognition accuracy of the model, and the method comprises:

[0046] Step 101, constructing a momentum observer according to the antisymmetric property of the generalized momentum of the generalized momentum method and the dynamic model;

[0047] Step 102: obtaining a first residual vector output by the momentum observer based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint, and the actual first velocity vector of the joint;

[0048] Step 103: Perform Laplace transform on the first residual vector to obtain a second residual vector of each joint of the robot;

[0049] Step 104: Optimize the dynamic model using the second residual vector.

[0050] Specifically, the dynamic model of the robot is:

[0051]

[0052] Where M(q) is the inertia matrix, is the centrifugal force and Coriolis force matrix, g(q) is the gravity vector, q, are joint position, velocity, and acceleration vectors, τ joint is the joint output torque, τ f For friction.

[0053] For the 4-axis collaborative robot 2442, if only joints 1, 2, and 4 are considered and the harmonic reducer is regarded as a rigid connection, the dynamic model is:

[0054]

[0055] The dynamic model has an antisymmetric property, that is, is an antisymmetric matrix, which leads to:

[0056]

[0057] When the robot is subjected to external force, the dynamic model becomes:

[0058]

[0059] Among them, τ ext is the torque applied to the joint by the external force, τ total is the resultant force actively applied to the joint.

[0060]

[0061] is the Jacobian matrix of the external moment, F ext is the vector of the external force at the joint, f ext 、m ext The amount of external force.

[0062] Define generalized momentum as:

[0063] Combining formulas The antisymmetric property (Formula 3) corresponds to have to:

[0064]

[0065] Where p is the generalized momentum, is the derivative of p, is the derivative of M(q), is the velocity vector of the axis, M(q) is the inertia matrix, is the acceleration vector, τ total is the resultant force actively applied to the joint, τ f is the friction force, is the centrifugal force and Coriolis force matrix, for The transpose of the matrix is ​​the swapping of rows and columns, τ ext is the torque applied to the joint by the external force.

[0066] In the specific embodiments, see Figure 2 In step 102, based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint and the actual first velocity vector of the joint, a first residual vector output by the momentum observer is obtained, including:

[0067] Step 201: obtaining a third residual vector based on a first difference between the momentum observer and the generalized momentum, and an observer diagonal gain matrix;

[0068] Step 202: Integrate the third residual vector to obtain a fourth residual vector;

[0069] Step 203: Obtain a first residual vector based on the fourth residual vector, the actual first inertia matrix of the joint, the actual first velocity vector of the joint, and the momentum observer.

[0070] Specifically, the momentum observer is designed as:

[0071]

[0072]

[0073] in, is the momentum observer, is the joint torque, is the friction force, is the transposed matrix of centrifugal force and Coriolis force, r is the observer output residual vector, is the derivative of r, p is the generalized momentum, is the derivative of p, K O is the observer diagonal gain matrix.

[0074] Integrating formula 8, we obtain:

[0075]

[0076] Among them, p(t) is the momentum at time t, and p(0) is the momentum at the initial moment.

[0077] The integrated residual vector is combined with the actual first inertia matrix of the joint, the actual first velocity vector of the joint and the momentum observer to obtain a first residual vector.

[0078] In the specific embodiments, see Figure 3 In step 203, obtaining a first residual vector based on the fourth residual vector, the actual first inertia matrix of the joint, the actual first velocity vector of the joint and the momentum observer includes:

[0079] Step 301, substituting the actual first inertia matrix of the joint and the actual first velocity vector of the joint into the fourth residual vector to obtain a fifth residual vector;

[0080] Step 302: Substitute the transposed matrices of the joint torque, friction force, centrifugal force and Coriolis force observed by the momentum observer and the second inertia matrix into the fifth residual vector to obtain the first residual vector.

[0081] Specifically, assuming that the robot is not moving in the initial state, the p(0) term is a zero vector, and Substituting formula (7) into the above formula, we can get the expression of the residual vector r:

[0082]

[0083] Observe equation (8) again, if Then formula (10) can be rewritten as:

[0084]

[0085] In a specific embodiment, the generalized momentum includes the resultant force actively applied to the joint, the torque applied to the joint by the external force, the friction force, the centrifugal force and the Coriolis force matrix and the velocity vector of the joint.

[0086] In a specific embodiment, the fourth residual vector is obtained in the following manner: obtaining a second difference between the momentum at the moment of movement and the momentum at the initial moment; obtaining a third difference between the first difference and the observation value of the momentum observer; and obtaining the fourth residual vector based on the third difference and the observer diagonal gain matrix.

[0087] In a specific embodiment, the first residual vector is obtained using the following formula:

[0088]

[0089] in, is the derivative of the first residual vector, K0 is the observer diagonal gain matrix, τ ext is the torque applied to the joint by the external force.

[0090] In a specific embodiment, the second residual vector is obtained using the following formula:

[0091]

[0092] Among them, r i is the second residual vector of the i-th joint, K 0,i is the observer diagonal gain matrix of the i-th joint, s is the Laplace transform, τ ext,i is the torque applied by the external force to the i-th joint, T 0,i =1 / K 0,i , T 0,i is the time constant.

[0093] In the specific embodiments, see Figure 4a , Figure 4b , Figure 4c and Figure 4d , is the current value (current identification value) of the collision test using the optimized dynamic model, and the current value (current original value) of the collision test using the original dynamic model. By comparing different current values, it can be seen that the current obtained using the optimized dynamic model is smoother and has a more obvious peak, which can more significantly reflect the collision process.

[0094] In the specific embodiment, it can be seen that formula (12) is a first-order inertia link. When K O →∞, r≈τ ext The block diagram of the momentum observer designed according to formula (10) is as follows: Figure 5 shown.

[0095] In the specific embodiments, see Figure 6 , which is a structural schematic diagram of an optimization system for a robot dynamics model provided in the second embodiment of the present application, the system comprises: an observer construction module 401, a residual output module 402, a Laplace transform module 403 and an optimization module 404; the observer construction module 401 is used to construct a momentum observer according to the generalized momentum of the generalized momentum method and the antisymmetric property of the dynamics model; the residual output module 402 is used to obtain the first residual vector output by the momentum observer based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint and the actual first velocity vector of the joint; the Laplace transform module 403 is used to perform a Laplace transform on the first residual vector to obtain a second residual vector of each joint of the robot; the optimization module 404 is used to optimize the dynamics model using the second residual vector.

[0096] The system in this embodiment constructs a momentum observer based on the antisymmetric properties of the generalized momentum of the generalized momentum method and the dynamic model; obtains the first residual vector output by the momentum observer based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint, and the actual first velocity vector of the joint; performs a Laplace transform on the first residual vector to obtain the second residual vector of each joint of the robot; and optimizes the dynamic model using the second residual vector. The residual vector is the difference between the actual observed value and the model predicted value. By optimizing the dynamic model through the residual vector, the model parameters can be continuously adjusted to make the model predicted value closer to the actual observed value, thereby improving the recognition accuracy of the model.

[0097] In a specific embodiment, the third embodiment of the present application provides an optimization device for a robot dynamic model, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method described in any one of the first embodiments of the present application. The device in this embodiment constructs a momentum observer according to the antisymmetric properties of the generalized momentum and dynamic model of the generalized momentum method; obtains the first residual vector output by the momentum observer based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint, and the actual first velocity vector of the joint; performs a Laplace transform on the first residual vector to obtain the second residual vector of each joint of the robot; and optimizes the dynamic model using the second residual vector. The residual vector is the difference between the actual observed value and the model predicted value. By optimizing the dynamic model through the residual vector, the model parameters can be continuously adjusted to make the model predicted value closer to the actual observed value, thereby improving the recognition accuracy of the model.

[0098] In a specific embodiment, the fourth embodiment of the present application provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the processor executes the steps of the method described in any one of the first embodiments of the present application. The storage medium in this embodiment constructs a momentum observer according to the antisymmetric properties of the generalized momentum and dynamic model of the generalized momentum method; based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint and the actual first velocity vector of the joint, the first residual vector output by the momentum observer is obtained; the first residual vector is Laplace transformed to obtain the second residual vector of each joint of the robot; and the dynamic model is optimized using the second residual vector. The residual vector is the difference between the actual observation value and the model prediction value. By optimizing the dynamic model by the residual vector, the model parameters can be continuously adjusted to make the model prediction value closer to the actual observation value, thereby improving the recognition accuracy of the model.

[0099] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0100] Figure 7 The internal structure of a computer device in one embodiment is shown. The computer device can be a terminal or a server. Figure 7The computer device includes a processor, a memory, etc. connected via a system bus. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor can implement the method in this embodiment. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor can implement the method in this embodiment. Those skilled in the art can understand that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0101] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

[0102] The above description is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for optimizing a robot dynamics model, wherein the dynamics model is a mathematical model of the relationship between the robot motion and the forces and moments applied thereto, characterized in that: The method comprises: According to the antisymmetric properties of the generalized momentum and dynamics model of the generalized momentum method, a momentum observer is constructed; Based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint and the actual first velocity vector of the joint, obtaining a first residual vector output by the momentum observer; Performing Laplace transform on the first residual vector to obtain a second residual vector of each joint of the robot; The dynamic model is optimized using the second residual vector.

2. The method for optimizing the robot dynamics model according to claim 1, characterized in that: The step of obtaining a first residual vector output by the momentum observer based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint, and the actual first velocity vector of the joint comprises: Obtaining a third residual vector based on the first difference between the momentum observer and the generalized momentum, and an observer diagonal gain matrix; Integrating the third residual vector to obtain a fourth residual vector; A first residual vector is obtained based on the fourth residual vector, a first actual inertia matrix of the joint, a first actual velocity vector of the joint, and the momentum observer.

3. The method for optimizing the robot dynamics model according to claim 1, characterized in that: The obtaining of the first residual vector based on the fourth residual vector, the actual first inertia matrix of the joint, the actual first velocity vector of the joint and the momentum observer comprises: Substituting the actual first inertia matrix of the joint and the actual first velocity vector of the joint into the fourth residual vector to obtain a fifth residual vector; The transposed matrices of the joint torque, friction force, centrifugal force and Coriolis force observed by the momentum observer and the second inertia matrix are substituted into the fifth residual vector to obtain the first residual vector.

4. The method for optimizing the robot dynamics model according to claim 1, characterized in that: The generalized momentum includes the resultant force actively applied to the joint, the torque applied to the joint by the external force, the friction force, the centrifugal force and the Coriolis force matrix and the velocity vector of the joint.

5. The method for optimizing the robot dynamics model according to claim 2, characterized in that: The fourth residual vector is obtained in the following manner: Obtaining a second difference between the momentum at the moving moment and the momentum at the initial moment; Obtaining a third difference between the first difference and the observation value of the momentum observer; The fourth residual vector is obtained according to the third difference and the observer diagonal gain matrix.

6. The method for optimizing the robot dynamics model according to claim 1, characterized in that: The first residual vector is obtained by using the following formula: in, is the derivative of the first residual vector, K0 is the observer diagonal gain matrix, τ ext is the torque applied to the joint by the external force.

7. The method for optimizing the robot dynamics model according to claim 1, characterized in that: The second residual vector is obtained by using the following formula: Among them, r i is the second residual vector of the i-th joint, K 0,i is the observer diagonal gain matrix of the i-th joint, s is the Laplace transform, τ ext,i is the torque applied by the external force to the i-th joint, T 0,i =1 / K 0,i , T 0,i is the time constant.

8. A robot dynamics model optimization system, characterized in that: The system comprises: an observer construction module, a residual output module, a Laplace transform module and an optimization module; The observer construction module is used to construct a momentum observer according to the antisymmetric properties of the generalized momentum of the generalized momentum method and the dynamic model; The residual output module is used to obtain a first residual vector output by the momentum observer based on the momentum observer, the observer diagonal gain matrix, the actual first inertia matrix of the joint and the actual first velocity vector of the joint; The Laplace transform module is used to perform Laplace transform on the first residual vector to obtain a second residual vector of each joint of the robot; The optimization module is used to optimize the dynamic model using the second residual vector.

9. A device for optimizing a robot dynamics model, comprising a memory and a processor, characterized in that: The memory stores a computer program, and when the computer program is executed by the processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 7.

Citation Information

Patent Citations

  • Kinetic parameter identification and collision detection method for six-joint robot

    CN111267105A

  • Mechanical arm external torque estimation method

    CN117532623A

Cited By

  • Dynamic model residual error compensation robot collision detection method and system considering frequency distribution difference

    CN119704187A