Mechanical arm tail end variable load dynamics self-diagnosis and compensation method in quick change mode

Abstract

The invention relates to a mechanical arm tail end variable load dynamics self-diagnosis and compensation method in a quick change form. The mechanical arm tail end variable load dynamics self-diagnosis and compensation method is technically characterized in that a dynamics model is established based on iterative least square online identification; load parameter identification under variable working conditions is carried out based on Jacobian matrix mapping; establishing a hybrid dynamic model including connecting rod rigid body dynamics, connecting rod error compensation dynamics and load dynamics; establishing a variable load dynamical model; when the load at the tail end of the mechanical arm is changed, robot joints are controlled by updating parameters in the hybrid dynamic model; on-line identification and compensation are carried out by collecting sensor data in real time. The method can effectively improve the accuracy and robustness of dynamic representation of the robot when facing various uncertain factors such as connecting rod deformation, variable load and measurement noise, improves the safety and reliability of physical man-machine interaction of the robot, and can be widely applied to the fields of industrial automation, medical operation assistance and rescue robots.

Description

Technical Field

[0001] The invention belongs to the technical field of robots, and relates to a multi-degree-of-freedom mechanical arm dynamics control method, in particular to a mechanical arm end variable load dynamics self-diagnosis and compensation method in a quick-change form. Background Art

[0002] The robot arm is a strongly coupled nonlinear multi-input multi-output system. The accuracy of the dynamic model parameters in the system directly affects the performance of human-machine physical safety interaction. Due to the existence of unknown parameters in the robot dynamics, it is a challenge to establish an accurate dynamic model. Therefore, system identification becomes one of the best choices for its accurate evaluation. However, due to various complex uncertainties, such as flexible links, variable loads, assembly gaps, and measurement noise, the robot dynamics model exhibits strong variability and unpredictability. These factors are mainly reflected in the dynamic characteristics of the robot system and the friction identification system, the accuracy of joint torque perception, and the effectiveness of tactile perception and interactive control.

[0003] At present, many researchers have conducted in-depth research on robot dynamics modeling and identification methods. For example, in order to solve the influence of nonlinear friction on high-precision motion systems, a feedforward compensator was designed based on the identified friction model to achieve nonlinear friction compensation; in order to solve the problem of identification of variable load dynamic parameters under multi-task loads, an adaptive dynamic parameter identification method combining a non-model-based adaptive control algorithm with a recursive least squares method was proposed to address the problem of poor control performance under complex variable load conditions; in order to solve the problem that some parameters identified in the least squares method do not have geometric or physical meanings, linear matrix inequality constraints and semi-definite optimization are used to ensure that the identified model parameters have actual physical meanings; in addition, many intelligent optimization algorithms have been widely used in the identification of robot dynamics parameters, which to a certain extent solves the problem that model parameters are physically infeasible, and lays a solid foundation for overcoming the difficulty of obtaining accurate robot dynamics models.

[0004] However, the existing research methods still have shortcomings. Existing researchers have performed offline identification of load dynamic parameters based on replanned excitation trajectories. However, since the motion space constraints of robots in actual loading work scenarios are more complex, there are usually no spatial conditions for the safe operation of loading excitation trajectories. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a method for self-diagnosis and compensation of variable load dynamics at the end of a robotic arm in a quick-change form, so as to solve the problems of lack of coupling of time-varying factors in the field of robot dynamics and difficulty in dynamic compensation modeling and online precise identification, so as to improve the accuracy of the dynamic model.

[0006] The present invention solves the existing technical problems by adopting the following technical solutions:

[0007] A method for self-diagnosis and compensation of variable load dynamics at the end of a robotic arm in a quick-change mode comprises the following steps:

[0008] Step 1: Establish a dynamic model based on iterative least squares online identification;

[0009] Step 2: Identify load parameters under variable working conditions based on Jacobian matrix mapping;

[0010] Step 3, establishing a hybrid dynamics model including connecting rod rigid body dynamics, connecting rod error compensation dynamics and load dynamics;

[0011] Step 4, establishing a variable load dynamics model;

[0012] Step 5: Determine the new load dynamic parameter identification method: When the load needs to be changed in the robot working scene, the load dynamic parameters are identified step by step;

[0013] Step 6: When the load at the end of the robot arm changes, the robot joints are controlled by updating the parameters in the hybrid dynamics model;

[0014] Step 7: Perform online identification and compensation by collecting sensor data in real time.

[0015] Furthermore, the implementation method of step 1 includes the following steps:

[0016] Step 1.1, derive the rigid body dynamics model of the robot arm link by Newton-Euler method;

[0017] Step 1.2, linearize the rigid body dynamics model;

[0018] Step 1.3: Recursively update parameter estimates using iterative least squares.

[0019] Furthermore, the method for performing linearization processing in step 1.2 is: extracting steady-state inertia parameter constants from the rigid body dynamics model of the robot arm connecting rod, and establishing a linearized reconstruction model formula of the rigid body dynamics of the robot arm connecting rod.

[0020] Further, the implementation method of step 2 includes the following steps:

[0021] Step 2.1, determine the coupling relationship and decoupling method of the load dynamics model;

[0022] Step 2.2: Identify the load parameters under variable working conditions through Jacobi matrix mapping.

[0023] Furthermore, the hybrid dynamics model τ established in step 3 efor:

[0024]

[0025] Among them, τ e ' is the connecting rod error compensation dynamic model, τ load It represents the additional joint torque generated when the end effector of the robot arm runs with load, τ link represents the joint torque, Δτ link represents the error in the rigid body dynamics model of the connecting rod, q, and Represent the joint position, velocity and acceleration respectively, represents the joint inertia moment, Corresponding to the centrifugal torque and Coriolis torque of the joint; G(q) represents the gravity torque of the joint; τ f represents the joint friction torque, Y d represents the observation matrix, Θ d represents the unidentified parameter, ΔΘd represents the parameter of the error compensation dynamics model, Φ b represents the minimum inertia parameter set of the load, E b Represents Φ b The corresponding full-rank observation matrix.

[0026] Furthermore, the motion state of the robot arm in the variable load dynamics model includes:

[0027] Under no-load conditions, the robot joint moves from the zero position to the target position;

[0028] Under loading conditions, all robot joints are kept in position servo mode while gravity and inertia parameters are determined;

[0029] Under loading, the robot joints move from the zero position to the target position.

[0030] Furthermore, the implementation method of step 6 is: by starting the online identification program, using the recursive least squares method to identify the load gravity parameters under static conditions, decoupling and updating these parameters, and then judging whether the identification process is completed through the set convergence index, and finally applying these updated parameters in the hybrid dynamics model.

[0031] Furthermore, the step 7 uses the recursive least square method to estimate and correct the unknown parameters in the model, thereby realizing the compensation function of online identification and compensation of the dynamic parameters of the robot arm.

[0032] Furthermore, the robotic arm is a 6-axis robotic arm.

[0033] The advantages and positive effects of the present invention are:

[0034] 1. The present invention realizes comprehensive robot arm dynamic modeling and online compensation functions by establishing a hybrid dynamics model including connecting rod rigid body dynamics, connecting rod error compensation dynamics and load dynamics. It uses an improved recursive least squares method to perform real-time online identification and compensation of the robot arm dynamic parameters, which significantly improves the control accuracy and robustness of the robot arm when facing uncertain factors such as connecting rod deformation, changing load and measurement noise. The experimental verification on a 6-DOF robot arm platform demonstrates the effectiveness of its method under unloading and loading conditions, and the theoretical analysis is supported by experimental data.

[0035] 2. The present invention effectively improves the accuracy and robustness of the robot's dynamic characterization when facing multiple uncertainties such as connecting rod deformation, changing load and measurement noise, and provides important theoretical and practical contributions to the safety and reliability of robot physical human-machine interaction, especially in applications that require precise control, such as industrial automation, medical surgical assistance and rescue robots. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is an overall flow chart of the variable load dynamics self-diagnosis and compensation method of the robot end of the present invention;

[0037] Figure 2 It is a schematic diagram of the principle of online identification and compensation of the rigid body dynamics model of the robot arm connecting rod of the present invention;

[0038] Figure 3 It is a flow chart of step-by-step online identification of load dynamic parameters under variable load disturbance of the present invention;

[0039] Figure 4 It is a schematic diagram of a simplified model of a variable load at the end of a quick-change mechanical arm of the present invention;

[0040] Figure 5 This is a simplified model schematic diagram of the 6-DOF robotic arm of the present invention. DETAILED DESCRIPTION

[0041] The embodiments of the present invention are further described in detail below with reference to the accompanying drawings.

[0042] The present invention focuses on the online accurate identification and compensation method of dynamic models, aiming to solve the problems of lack of coupling of time-varying factors in the field of robot dynamics and difficulty in dynamic compensation modeling and online accurate identification. The present invention is mainly used in a robotic arm system under variable load working conditions, the system mainly includes human-computer interaction control, a teaching operation platform, a 6-axis robotic arm and a control system; the robotic arm is a 6-axis robotic arm, which has a higher degree of freedom and flexibility, can perform more complex actions and tasks, and has more accurate movement in three-dimensional space, suitable for precision assembly and complex operations, and can adapt to more working environments. It is especially important in automated production lines and can improve production efficiency and product quality.

[0043] Based on the above variable load manipulator system, the present invention proposes a method for self-diagnosis and compensation of variable load dynamics at the end of the manipulator in a quick-change form, such as Figures 1 to 5 As shown, the following steps are included:

[0044] Step 1: Establish a dynamic model based on iterative least squares online identification. Compared with traditional offline identification, this dynamic model reduces the identification error.

[0045] In this step, the robot's dynamic equations are first linearized to construct a linear state equation; then, during the operation of the robot, new joint state data are continuously collected and the parameter estimates are recursively updated using the iterative least squares method. In order to improve the model's ability to track the latest dynamic characteristics, a forgetting factor is introduced to adjust the weights of new and old data; finally, the predicted torque is calculated using the identified dynamic parameters and compared with the actual torque to obtain the model residual, and a compensatory control strategy is designed to reduce the impact of the identification error.

[0046] The specific implementation method of this step is:

[0047] Step 1.1: Use the Newton-Euler method to derive the following rigid body dynamics model of the robot arm link:

[0048]

[0049] Among them, q, and Respectively represent joint position, velocity and acceleration; τ link represents the joint torque; represents the joint inertia moment; Corresponding to the centrifugal torque and Coriolis torque of the joint; G(q) represents the gravity torque of the joint; τ f represents the joint friction torque; f c and f v They represent Coulomb friction coefficient and viscous friction coefficient respectively; sign() represents the sign function.

[0050] Step 1.2: Linearize the dynamic equations, that is, linearize the rigid body dynamics model.

[0051] In order to simplify the parameter identification process, the steady inertia parameter constants are extracted from the rigid body dynamics model of the robot arm link, and the linearized reconstruction model of the rigid body dynamics of the robot arm link is established, which is expressed as:

[0052]

[0053] In the formula represents the observation matrix formed by combining the transformed motion state information; Θ represents the inertial parameter constant; Θ = [Θ 1 Θ 2 …Θ i ], where Θ i is the inertia parameter constant of joint i.

[0054]

[0055] in, and represents the inertia tensor matrix of link i; and represents the mass moment of connecting rod i; m i represents the mass of connecting rod i; represent the Coulomb friction coefficient and viscous friction coefficient of connecting rod i respectively.

[0056] Since the observation matrix Y is usually not full rank, the complexity of identification increases, so QR decomposition or parameter reorganization method can be used to eliminate some inertial parameters and obtain the minimum inertial parameter set Θ b .

[0057]

[0058] Where Θ b represents the minimum inertia parameter set, Y b represents the full-rank observation matrix.

[0059] Step 1.3: Recursively update parameter estimates using iterative least squares.

[0060] The most widely used method for offline system identification and unknown parameter fitting is based on the least squares method. The least squares method estimates parameters by minimizing the sum of squares of prediction errors, thereby obtaining statistically optimal parameter fitting results. The process of identifying the minimum inertia parameter set based on the least squares method can be described by the following formula:

[0061]

[0062] Step 2: Identify load parameters under variable working conditions based on Jacobian matrix mapping.

[0063] Assuming that the linear regression reconstruction of the robot arm dynamics model and the extraction of the minimum inertia parameter set have been completed at the current moment, the expression of the joint torque expressed by the unknown parameters and the observation matrix is ​​obtained:

[0064] τ d =Y d ·Θ d

[0065] Where τ d represents the joint torque; Θ d Indicates an unrecognized parameter; Y d Represents the observation matrix. The above are all parameters in the recursive identification process.

[0066] According to the principle of recursive least squares method with forgetting factor, by introducing the forgetting factor Z N] , increasing the weight ratio of the newly sampled data and reducing the weight ratio of the old data, we can derive the parameter identification method of the recursive least squares method with forgetting factor.

[0067]

[0068] Where D N ] represents the gain matrix; Z N W represents the forgetting factor; N ] represents the covariance matrix.

[0069] Since it is difficult to accurately characterize the dynamic characteristics of the robot arm connecting rod through offline parameter identification, a connecting rod compensation torque representation model related to the robot arm motion state information is established considering the characterization error of the dynamic characteristics of the robot arm connecting rod.

[0070]

[0071] Where Δτ link represents the error in the rigid body dynamics model of the connecting rod; ΔΘd represents the parameters of the error compensation dynamics model.

[0072]

[0073] Furthermore, a rigid body dynamics model with connecting rod and a connecting rod error compensation dynamics model τ were established. e 'The hybrid dynamics model of the manipulator:

[0074]

[0075] Where τ link is obtained during the offline recognition process in step 1.

[0076] Step 3: Establish a hybrid dynamics model including connecting rod rigid body dynamics, connecting rod error compensation dynamics and load dynamics. The hybrid dynamics model can improve the accuracy of the dynamics model under multivariable uncertain disturbances.

[0077] In this step, a compensating dynamic model is designed according to the deformation and modeling error of the connecting rod. The Jacobian matrix is ​​then used to map the load dynamics from the end-effector dynamics to the joint space, and a decoupled load dynamics model is established. A framework for characterizing the dynamic characteristics of the robotic arm is proposed, including the rigid body dynamics of the connecting rod, the error compensation dynamics of the connecting rod, and the load dynamics.

[0078] When the robot moves with a load, the force / torque W acting on the end effector of the robot can be mapped to the joint space using the Jacobian matrix J(q). In this way, the additional load torque τload generated by each joint can be calculated:

[0079]

[0080] Where τ load It represents the additional joint torque generated when the end effector of the robot arm runs with load; J T (q) represents the Jacobian matrix R n×6 ; F load and N load They represent the force and torque vectors acting on the end effector of the robot arm respectively.

[0081] According to dynamic analysis, the force and torque expressions of the load acting on the end effector of the robot arm are:

[0082]

[0083] In the formula represents the angular velocity of the end effector; represents the angular acceleration of the end effector; represents the linear velocity of the end effector; Represents the linear acceleration of the end effector; m l Indicates the load mass; Indicates the load center of mass position; Indicates that the load center of mass is at X l ,Y l ,Z l Position on the axis; I l is the inertia tensor matrix at the load center of mass; Respectively represent the load around X l ,Y l ,Z l Mass moment of inertia of the coordinate axes; Respectively represent the load at X l Yl , Y l Z l , Z l X l Moment of inertia on a plane.

[0084] The linearized reorganized expressions of force and torque acting on the end effector of the robot arm can be simplified as follows:

[0085]

[0086] In the formula They respectively represent the forces and moments generated by the robot load that are independent of the kinetic inertia parameters of the load; Φ represents the kinetic inertia constant of the load.

[0087] At this point, the linearized reconstruction model of the dynamic load of the robot arm can be obtained:

[0088]

[0089] In the formula represents the load observation matrix.

[0090] Usually, in the observation matrix E, some columns are completely composed of zeros or are linearly correlated, which leads to the existence of parameters in the corresponding load moment of inertia parameter set Φ that cannot be completely independently identified. Similarly, the QR decomposition method can obtain the minimum inertia parameter set of the load. In this case, the load dynamic model can be expressed as:

[0091]

[0092] Where Φ b represents the minimum inertia parameter set of the load, E b Represents Φ b The corresponding full-rank observation matrix.

[0093] In addition to the gravity load, the inertia load also affects the motion process. Usually, the robot arm manufacturer only provides an interface for correcting the dynamic inertia parameters of the load gravity term, while ignoring the coupling relationship between the load gravity term and the inertia term. However, in high-precision robot arm control, the influence of the load inertia term also needs to be considered. This paper proposes a new load dynamic parameter identification method, that is, when the load needs to be changed in the robot arm working scenario, the load dynamic parameters are identified step by step.

[0094] In order to facilitate parameter identification, it is best to set the parameter to be identified Φ b The inertial parameters related to the gravity term are separated and collected into a single vector Φ g , Similarly, the observation matrix representing the load gravity term can be extracted and expressed as vector E g Therefore, the following relationship holds:

[0095] G load (q) = E g (q)·Φ g

[0096] It is not difficult to see that when the robot arm is not subjected to external force, all joints are in position servo mode, that is, τ ext =0, When , the measurement value of the joint torque sensor is equal to the torque provided by the link gravity term and the load gravity term.

[0097] τ load =G link (q)+G load (q)

[0098] In the formula, τ load is the measured value of the joint torque sensor under load change conditions, G link (q) is the joint torque generated by the link gravity term, G load (q) is the joint torque generated by the load gravity term.

[0099] The dynamic parameters of the connecting rod have been determined in the previous article. The rigid body dynamics model of the robot arm connecting rod can be used to calculate the gravity term G of the connecting rod. link (q). This allows the load gravity term load G load (q) is decoupled from the joint torque. Therefore, in the case of static equilibrium without external forces, the recursive least squares method can be used to identify the dynamic parameters related to the load-gravity term, thereby obtaining a more accurate estimate of the load-gravity term parameters.

[0100] Indicator C 1 and C 2 The load gravity term and inertia term are defined.

[0101]

[0102] in, and The trigger threshold indicating the completion of the recognition process is selected by weighing the accuracy and time of the recognition process.

[0103] Finally, the hybrid dynamics model of the manipulator includes the connecting rod rigid body dynamics model, the connecting rod error compensation dynamics model and the load dynamics model, which can be expressed as:

[0104]

[0105] Step 4: Establish a variable load dynamics model. The motion state of the robot arm in the variable load dynamics model includes the following contents:

[0106] Under no-load conditions, the robot joint moves from the zero position to the target position;

[0107] Under loading conditions, all robot joints are kept in position servo mode while gravity and inertia parameters are determined;

[0108] Under loading, the robot joints move from the zero position to the target position.

[0109] Step 5: Determine the new load dynamic parameter identification method:

[0110] The specific implementation method of this step is: it is adopted when the load needs to be changed in the robot working scenario. The method first starts the online identification program, and then uses the recursive least squares method (RLS) to identify the gravity parameters of the load under static conditions. These parameters are then decoupled and updated to more accurately reflect the new load characteristics. After the update process is completed, the set convergence indicators are used to determine whether the identification process has been completed. These indicators are usually set based on the determinant of the observation matrix. Once it is confirmed that the identification process is complete, the updated parameters will be applied to the hybrid dynamics model to achieve accurate prediction and effective control of the robot joint torque. This process ensures that the robot can maintain the accuracy and robustness of its dynamic performance when facing load changes.

[0111] Step 6: When the load at the end of the robot arm changes, the robot joints are controlled by updating the parameters in the hybrid dynamics model.

[0112] The specific implementation method of this step is as follows: when a load change is detected, the robot will start the online identification program and use the recursive least squares method (RLS) to identify the gravity parameters of the load in real time under static conditions. After these parameters are decoupled and updated, the preset convergence index will be used to determine whether the identification process has been completed. Once confirmed, the updated parameters will be applied to the hybrid dynamics model for compensation in the control algorithm. This process includes a comprehensive consideration of the rigid body dynamics of the connecting rod, the dynamics of the connecting rod error compensation, and the load dynamics, ensuring that the robot can reduce the prediction error and improve the accuracy and reliability of the operation through the compensation control strategy when facing load changes. In addition, in order to adapt to possible further changes or disturbances, the online identification and compensation algorithm will continue to run to ensure that the robot maintains optimal performance throughout the operation.

[0113] Step 7: Online identification and compensation of the dynamic parameters of the robot arm are performed by collecting sensor data in real time.

[0114] This step is implemented by using recursive least squares (RLS) to collect and process sensor data from the robot joints, such as torque, position, velocity, and acceleration, in real time. This approach allows the system to recursively update parameter estimates in the dynamic model as new data arrives, without having to reprocess the entire data set. By comparing the predicted and actual torques, the system can calculate model residuals and design compensating control strategies to reduce the impact of these errors. In addition, a closed-loop feedback mechanism ensures that the control strategy can be adjusted based on the residuals fed back in real time to improve the accuracy and reliability of the robot's operation. The entire process involves the online update of the link dynamics parameters and the load dynamics parameters, as well as the application of a hybrid dynamics model that combines the rigid body dynamics of the link, the link error compensation dynamics, and the load dynamics. Through continuous monitoring and adjustment, the robot is able to adapt to changing dynamic conditions, thereby maintaining the accuracy and stability of its operation in the face of model imperfections and external disturbances.

[0115] It should be emphasized that the embodiments described in the present invention are illustrative rather than restrictive. Therefore, the present invention includes but is not limited to the embodiments described in the specific implementation manner. Any other implementation manners derived by those skilled in the art based on the technical solution of the present invention also fall within the scope of protection of the present invention.

Claims

1. A method for self-diagnosis and compensation of variable load dynamics at the end of a robotic arm in a quick-change mode, characterized in that: The following steps are involved: Step 1: Establish a dynamic model based on iterative least squares online identification; Step 2: Identify load parameters under variable working conditions based on Jacobian matrix mapping; Step 3, establishing a hybrid dynamics model including connecting rod rigid body dynamics, connecting rod error compensation dynamics and load dynamics; Step 4, establishing a variable load dynamics model; Step 5: Determine the new load dynamic parameter identification method: When the load needs to be changed in the robot working scene, the load dynamic parameters are identified step by step; Step 6: When the load at the end of the robot arm changes, the robot joints are controlled by updating the parameters in the hybrid dynamics model; Step 7: Perform online identification and compensation by collecting sensor data in real time.

2. The method for self-diagnosis and compensation of variable load dynamics at the end of a robot arm in a quick-change mode according to claim 1, characterized in that: The implementation method of step 1 comprises the following steps: Step 1.1, derive the rigid body dynamics model of the robot arm link by Newton-Euler method; Step 1.2, linearize the rigid body dynamics model; Step 1.3: Recursively update parameter estimates using iterative least squares.

3. The method for self-diagnosis and compensation of variable load dynamics at the end of a robot arm in a quick-change mode according to claim 2, characterized in that: The method for performing linearization processing in step 1.2 is: extracting steady-state inertia parameter constants from the rigid body dynamics model of the robot arm connecting rod, and establishing a linearized reconstruction model formula of the rigid body dynamics of the robot arm connecting rod.

4. The method for self-diagnosis and compensation of variable load dynamics at the end of a robot arm in a quick-change mode according to claim 1, characterized in that: The implementation method of step 2 comprises the following steps: Step 2.1, determine the coupling relationship and decoupling method of the load dynamics model; Step 2.2: Identify the load parameters under variable working conditions through Jacobi matrix mapping.

5. The method for self-diagnosis and compensation of variable load dynamics at the end of a robot arm in a quick-change mode according to claim 1, characterized in that: The hybrid dynamics model τ established in step 3 e for: Among them, τ e ' is the connecting rod error compensation dynamics model, τ load It represents the additional joint torque generated when the end effector of the robot arm runs with load, τ link represents the joint torque, Δτ link represents the error in the rigid body dynamics model of the connecting rod, q, and Represent the joint position, velocity and acceleration respectively, represents the joint inertia moment, Corresponding to the centrifugal torque and Coriolis torque of the joint; G(q) represents the gravity torque of the joint; τ f represents the joint friction torque, Y d represents the observation matrix, Θ d represents the unidentified parameter, ΔΘd represents the parameter of the error compensation dynamics model, Φ b represents the minimum inertia parameter set of the load, E b Represents Φ b The corresponding full-rank measurement matrix.

6. The method for self-diagnosis and compensation of variable load dynamics at the end of a robot arm in a quick-change mode according to claim 1, characterized in that: The motion states of the robot arm in the variable load dynamics model include: Under no-load conditions, the robot joint moves from the zero position to the target position; Under loading conditions, all robot joints are kept in position servo mode while gravity and inertia parameters are determined; Under loading, the robot joints move from the zero position to the target position.

7. The method for self-diagnosis and compensation of variable load dynamics at the end of a robot arm in a quick-change mode according to claim 1, characterized in that: The implementation method of step 6 is: by starting the online identification program, using the recursive least squares method to identify the load gravity parameters under static conditions, decoupling and updating these parameters, and then judging whether the identification process is completed through the set convergence index, and finally applying these updated parameters in the hybrid dynamics model.

8. The method for self-diagnosis and compensation of variable load dynamics at the end of a robot arm in a quick-change mode according to claim 1, characterized in that: The step 7 uses the recursive least square method to estimate and correct the unknown parameters in the model, thereby realizing the compensation function of online identification and compensation of the dynamic parameters of the robot arm.

9. The method for self-diagnosis and compensation of variable load dynamics at the end of a robot arm in a quick-change mode according to any one of claims 1 to 8, characterized in that: The robotic arm is a 6-axis robotic arm.

Images