A robot motion real-time parameter identification and compensation control method and robot

Abstract

The application discloses a robot motion real-time parameter identification and compensation control method and a robot, and through real-time acquisition of robot motion information and real-time calculation of friction parameters more close to the current moment, the application continuously updates the friction parameters to realize a high-precision friction model, so that the predicted friction torque can be more close to the actual friction torque, a more accurate dynamic model can be constructed, and the response speed and control precision of the robot control system are improved; in addition, the compensation control method of the application calculates the feedforward compensation torque once after the real-time parameter identification is executed once, is used for the feedback system of the robot to realize the real-time compensation control of the robot motion, and the motion control precision of the robot can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of robot technology, in particular to a robot motion real-time parameter identification and compensation control method and a robot. BACKGROUND

[0002] With the increasing demand for robots, the application field of robots is also expanding, and robots have been widely used in polishing, polishing, assembly and other fields. To achieve high-precision motion control, the robot dynamics model must be considered for position control or torque control. By compensating for the dynamics characteristics of the robot, the system can quickly reach a steady state, reduce overshoot, and suppress motion vibration, thereby improving dynamic response speed and control accuracy.

[0003] The joint motion of the robot is transmitted to the mechanical arm through the RV reducer or the harmonic reducer of the servo motor, but due to the friction between the moving parts of the reducer, the friction force must be considered in the establishment of the robot dynamics equation. The friction parameters identified by the prior art are fixed, and identification needs to be made for different working conditions, which requires a lot of time, and due to the complex characteristics of the friction model, the accuracy of the calculated friction torque is not high. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a robot motion real-time parameter identification and compensation control method and a robot, which realizes a high-precision friction model by updating the friction parameters in real time, and the predicted friction torque is closer to the actual friction torque, which helps to build a more accurate dynamics model and improve the response speed and control accuracy of the robot control system.

[0005] To solve the above technical problems, the first aspect of the present application discloses a robot motion real-time parameter identification method, which comprises the following steps:

[0006] S1. Calculate the overall regression matrix and the friction regression matrix at the current time according to the trajectory information of the robot joint at the current k time : wherein, , , , , are the position, velocity and acceleration vectors of the robot joint, , ;

[0007] S2. Calculate the gain matrix at the current time, the formula is as follows:​​

[0008] ,

[0009] wherein, is the friction regression matrix calculated at the last time step, I is the identity matrix, is the transpose of , is the covariance matrix at the last time step;

[0010] S3. Calculate the estimated friction torque at the current time step , as follows:

[0011] ,

[0012] wherein, is the joint torque at the current time step, is the overall regression matrix at the current time step, is the inertia-based parameter;

[0013] S4. Calculate the estimation error from the friction parameter at the last time step:

[0014] ;

[0015] S5. Update the current friction parameter of the robot from the estimation error and the gain matrix :

[0016] ;

[0017] S6. Calculate the covariance matrix at the current time step, for the next time step calculation:

[0018] ;

[0019] S7. Calculate the compensated friction torque at the current time step from the current friction regression matrix and the current friction parameter :

[0020] .

[0021] As an optional embodiment, the sampled torque of each joint is calculated from the sampled current:

[0022] The sampled current is first filtered:

[0023] ,

[0024] wherein, is the collected current at the kth moment, is the filtered current at the k-1th moment, is the filtered current at the kth moment, is the filtering coefficient;

[0025] The joint torque at the kth moment is obtained by the following conversion formula :

[0026] ,

[0027] wherein, is the deceleration ratio, is the torque coefficient.

[0028] As another optional embodiment,

[0029] is the friction regression matrix at the current kth moment, wherein , is the velocity of the i th joint of the robot, is the sign function;

[0030] The overall regression matrix is calculated by the following formula:

[0031] ,

[0032] wherein, is the joint torque, is the minimum inertia parameter set containing the to-be-identified dynamic parameters of the basic inertia parameters and the friction parameters of each link, is the inertia regression matrix determined in advance according to the parameters of the robot itself, is the inertia basic parameter determined in advance according to the parameters of the robot itself, is the friction parameter at the current moment .

[0033] As another optional embodiment, before step S1, the method further comprises:

[0034] controlling the robot to run according to the excitation trajectory, collecting motion data and determining the to-be-identified friction parameter during the running initial value at the initial moment ;

[0035] wherein, the excitation trajectory is designed by five-term finite Fourier series, and the excitation trajectory is defined as:

[0036] ,

[0037] in, , and Let i represent the position, angular velocity, and angular acceleration of the i-th joint at time t, respectively. It is the joint angle at the initial moment; It is the fundamental angular frequency of the Fourier series excitation trajectory. and These are the coefficients of the Fourier series excitation trajectory.

[0038] As another optional implementation, prior to step S1, the method further includes:

[0039] The robot is controlled to run along an excitation trajectory. During the operation, motion data of each joint of the robot is collected, and the following set of linear regression equations is derived:

[0040] ,

[0041] Where m is the number of dynamic data sets of the robot joints acquired when the robot moves according to the excitation trajectory, each dynamic data set contains the input torque, angle, angular velocity, and angular acceleration of all joints of the robot, and m is greater than the minimum set of inertial parameters. The number of parameters in; It is the error vector of nm × 1 least squares, where n is the robot's degrees of freedom; It is the overall regression matrix corresponding to a single time point. It refers to the m time points corresponding to the robot running along the excitation trajectory for a period of time. The overall observation matrix formed by; Let m be the joint torques corresponding to m time points;

[0042] Solve for the minimum set of inertia parameters in the linear regression equation system. .

[0043] As another optional implementation, in solving the minimum set of inertia parameters in the linear regression equations... Subsequently, the method further includes:

[0044] Obtain the error vector ;

[0045] For the error vector The errors are rearranged to form an m × n error matrix R, with the following structure:

[0046] ,

[0047] in, denotes an error vector composed of m sets of torque errors of the i th joint among n joints;

[0048] A covariance matrix C is calculated, and the formula is as follows:

[0049] ,

[0050] Wherein, C is an n × n symmetric matrix, and cov is a covariance calculation function;

[0051] The linear regression equation set is normalized:

[0052] ,

[0053] Wherein, is a weighted matrix;

[0054] The minimum inertia parameter set is solved by using the least square method based on the following constraints :

[0055] ,

[0056] Wherein, n is the number of links, m i is the mass of link i, I i is the inertia tensor matrix of link i in the center of mass coordinate system, f ci is the Coulomb friction coefficient of link i, f vi is the viscous friction coefficient of link i.

[0057] The minimum inertia parameter set solved is determined as the minimum inertia parameter set At the initial value at the initial moment.

[0058] As another optional implementation, the minimum inertia parameter set solved in the linear regression equation set comprises:

[0059] The minimum inertia parameter set in the linear regression equation set is solved by using the least square method based on the following constraints :

[0060] ,

[0061] Wherein, n is the number of links, m i is the mass of link i, I i is the inertia tensor matrix of link i in the center of mass coordinate system, f ci is the Coulomb friction coefficient of link i, f vi is the viscous friction coefficient of link i.

[0062] As another optional embodiment, the optimization index of the excitation trajectory is the minimum condition number of the observation matrix, and the optimization objective function of the excitation trajectory is:

[0063] ,

[0064] In the formula, is the condition number of the observation matrix, 、 、 、 、 、 are upper and lower limits of the rotatable positions, velocities and accelerations of the joints of the robot respectively; 、 are the initial joint velocities and accelerations of the robot respectively, 、 are the final joint velocities and accelerations of the robot respectively.

[0065] The second aspect of the present application discloses a robot motion real-time compensation control method, the robot has a feedback control system, and the method comprises the following steps:

[0066] When the robot runs an arbitrary trajectory, the motion data of each joint of the robot in the current time period is collected, the robot motion real-time parameter identification method as described in the first aspect of the present application is used to calculate the compensation friction torque corresponding to the current time period 、 and , and then the feedforward compensation torque is calculated according to the following formula: :

[0067] ,

[0068] The feedforward compensation torque is provided to the feedback control system as the actual value of the friction torque of the current robot.

[0069] The third aspect of the present application discloses a robot comprising a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to realize the steps in the method disclosed in the first aspect or the second aspect of the present application.

[0070] The fourth aspect of the present application discloses a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the steps in the method disclosed in the first aspect or the second aspect of the present application.

[0071] Compared with the prior art, the embodiments of the present application have the following beneficial effects: ​

[0072] Compared with the prior art, since the joint friction of the robot has a nonlinear characteristic and the joint motion speed changes greatly, the embodiment of the present application proposes an online friction parameter identification method, friction parameters more close to the current moment are calculated in real time by acquiring robot motion information in real time, and a high-precision friction model is realized by continuously updating the friction parameters, so that the predicted friction torque can be more close to the actual friction torque, which helps to build a more accurate dynamic model and improve the response speed and control precision of the robot control system; the compensation control method of the present application calculates the feedforward compensation torque once after performing the real-time parameter identification described above, which is used for the feedback system of the robot to realize real-time compensation control of robot motion, and can improve the motion control precision of the robot. BRIEF DESCRIPTION OF DRAWINGS

[0073] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0074] Figure 1 is a flowchart of a robot motion real-time parameter identification method disclosed by the embodiment of the present application;

[0075] Figure 2 is a partial flowchart of a robot motion real-time parameter identification method disclosed by the embodiment of the present application;

[0076] Figure 3 and Figure 4 is a specific parameter composition diagram of the inertia base parameter disclosed by the embodiment of the present application;

[0077] Figure 5 is a flowchart of a robot motion real-time compensation control method disclosed by the embodiment of the present application;

[0078] Figure 6 is a structural diagram of a robot disclosed by the embodiment of the present application. DETAILED DESCRIPTION

[0079] In order to make the personnel in the technical field better understand the present application scheme, the technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0080] Embodiment one

[0081] Referring to Figures 1-2 , the embodiments of the present application disclose a robot motion real-time parameter identification method, the method comprising the steps of:

[0082] S1. According to the trajectory information of the robot joint at the current k moment , , to calculate the overall regression matrix and the friction regression matrix at the current moment: wherein, , , are the position, velocity and acceleration vectors of the robot joint, is , is .

[0083] S2. Calculate the gain matrix at the current moment, the formula is as follows:

[0084] ,

[0085] In the formula, is the friction regression matrix calculated at the last moment, I is a unit diagonal matrix, is the transpose matrix of , is the covariance matrix at the last moment.

[0086] S3. Calculate the estimated friction torque at the current moment, the formula is as follows:

[0087] ,

[0088] In the formula, is the joint torque at the current moment, is the overall regression matrix at the current moment, is the inertia-based parameter.

[0089] In this step, because the robot joint torque includes the inertia torque and the friction torque, here subtracts the inertia torque part to obtain the estimated friction torque , which is used for subsequent steps to calculate the current new friction parameter .

[0090] S4. According to the friction parameter at the last moment, calculate the estimation error :

[0091] .

[0092] In this step, the estimation error here refers to the error of the friction torque calculated by the friction parameter at the last moment and the actual friction torque (the estimated friction torque at the current moment calculated in the above step S3, because the friction torque cannot be measured, it can only be calculated by step S3).

[0093] S5. Update the current friction parameter of the robot by the estimation error and the gain matrix .

[0094] .

[0095] S6. Calculate the covariance matrix at the current moment for reserving for the next moment calculation:

[0096] .

[0097] S7. Calculate the compensation friction torque at the current moment according to the current friction regression matrix and the current friction parameter :

[0098] .

[0099] The method of the application can be independently calculated in the robot control system, or can be run on a PC or other host computer. Optionally, the interval between two moments can be set to 1-2 ms, that is, the calculation is performed once every 1-2 ms. In fact, in the verification experiment, the actual calculation time of each time reached about 800 microseconds, that is, about 0.8 ms, and the result can be calculated.

[0100] Since the joint friction of the robot has nonlinear characteristics and the joint motion speed changes greatly, the embodiment of the application proposes an online friction parameter identification method, which obtains the motion information of the robot in real time and calculates the friction parameter more suitable for the current moment in real time, so as to realize a high-precision friction model by continuously updating the friction parameter, so that the predicted friction torque can be more close to the actual friction torque, which is helpful to build a more accurate dynamic model and improve the response speed and control precision of the robot control system.

[0101] In an optional embodiment, the sampling torque of each joint is calculated by sampling current:

[0102] The sampling current is first subjected to filtering processing:

[0103] ,

[0104] wherein, ​​is the collected current at the kth moment, is the filtered current at the k-1th moment, is the filtered current at the kth moment, is the filtered current at the kth moment,

[0105] The joint torque at the kth moment is obtained through the following conversion formula :

[0106] ,

[0107] wherein, is the deceleration ratio, is the torque coefficient.

[0108] In yet another optional embodiment, the joint position signal emitted by the sensor is processed through a first-order low-pass filter to obtain position information q, and then the position information q is twice differentiated to obtain angular velocity and angular acceleration, respectively. The sensor can be an existing position sensor, which outputs current or voltage signals.

[0109] In yet another optional embodiment, is the friction regression matrix at the current kth moment, wherein , is the velocity of the i th joint of the robot, is a sign function;

[0110] The overall regression matrix is calculated through the following formula:

[0111] ,

[0112] wherein, is the joint torque, is the minimum inertia parameter set containing the basic inertia parameters and friction parameters of each link to be identified, is the inertia regression matrix determined in advance according to the parameters of the robot itself, is the inertia basic parameter determined in advance according to the parameters of the robot itself, is the friction parameter at the current moment . Wherein, , is the Coulomb friction coefficient, is the viscous friction coefficient, is the friction force bias; Figures 3-4 represents 40 component parameters of the inertia basic parameter , wherein c is the joint mass center position, B is the motor inertia, m is the joint mass, a is the link length, d is the joint coordinate system offset, and I is the inertia moment.

[0113] In this embodiment, the dynamics model of the robot considering friction is established based on Newton-Euler method:

[0114] ,

[0115] In the formula: , , are the position, velocity and acceleration vectors of the joints of the robot, is the inertia matrix of the robot arm, is the velocity term matrix related to centrifugal force and Coriolis force, is the gravity term, is the joint torque vector; wherein, represents the friction force / torque, and the friction of joint i is modeled using the Coulomb viscous friction model:

[0116] ,

[0117] In the formula, is the Coulomb friction coefficient, is the viscous friction coefficient, is the friction force bias, is the sign function.

[0118] The dynamics model of the robot considering friction described above is linearized, and the full-rank least regression matrix expression and the corresponding least inertia parameter set are obtained by using QR decomposition parameter reorganization method, i.e. the above formula .

[0119] In another optional embodiment, before step S1, the method further comprises:

[0120] controlling the robot to run according to an excitation trajectory, collecting motion data and determining the friction parameters to be identified during the running process initial value at the initial time ;

[0121] wherein the excitation trajectory is designed by five-term finite Fourier series, and the excitation trajectory is defined as:

[0122] ,

[0123] wherein, , and respectively represent the position, angular velocity and angular acceleration of the i-th joint at time t, is the joint angle at the initial time; is the base angular frequency of the Fourier series excitation trajectory, and is a coefficient of the Fourier series excitation trajectory.

[0124] In yet another optional embodiment, before the step S1, the method further comprises:

[0125] controlling the robot to run according to the excitation trajectory, collecting motion data of each joint of the robot (collecting position information q of each joint of the robot and driving current of each joint; the sampling data needs to be smoothed and filtered first), and combining the following linear regression equation set:

[0126] ,

[0127] wherein m is the number of the dynamic data sets of the joints of the robot obtained when the robot moves according to the excitation trajectory, each of the dynamic data sets contains input torque, angle, angular velocity and angular acceleration of all joints of the robot, and m is greater than the number of parameters in the minimum inertia parameter set ; is an nm × 1 least square error vector, and n is the degree of freedom of the robot; is the overall regression matrix corresponding to a single time point, is an overall observation matrix composed of the regression matrices corresponding to m time points obtained when the robot runs according to the excitation trajectory for a period of time; is the joint torque corresponding to the m time points; solving the minimum inertia parameter set

[0128] in the linear regression equation set.

[0129] In yet another optional embodiment, after solving the minimum inertia parameter set in the linear regression equation set, the method further comprises:

[0130] obtaining the error vector ;

[0131] rearranging the error vector to form an m × n error matrix R, which has the following structure:

[0132] ,

[0133] wherein represents an error vector composed of m sets of torque errors of the i th joint among the n joints;

[0134] calculating the covariance matrix C, which has the following formula:

[0135] , ​

[0136] where C is an n x n symmetric matrix, cov is a covariance computation function;

[0137] normalizing the system of linear regression equations:

[0138] ,

[0139] where, is a weighting matrix;

[0140] solving the minimum inertia parameter set based on the following constraints :

[0141] ,

[0142] where n is the number of links, m i is the mass of link i, I i is the inertia tensor matrix of link i in the center of mass coordinate system, f ci is the Coulomb friction coefficient of link i, f vi is the viscous friction coefficient of link i;

[0143] determining the minimum inertia parameter set obtained by solving to be the minimum inertia parameter set at the initial time.

[0144] on the basis of having solved the minimum inertia parameter set , considering that the measurement noise among the joint torques will interfere with each other when m groups of joint motion data are collected, in order to reduce the influence of abnormal data on the identification result, the error vector is rearranged in the embodiment, and then the minimum inertia parameter set is solved again to obtain the initial value of the final minimum inertia parameter set , which can provide an initial parameter value accurate to the characteristics of the robot itself in the subsequent operation of the robot along an arbitrary trajectory, and improve the accuracy of the subsequent calculation of the friction parameter .

[0145] In yet another optional embodiment, the solving of the minimum inertia parameter set in the system of linear regression equations comprises:

[0146] solving the minimum inertia parameter set in the system of linear regression equations based on the following constraints by least squares:

[0147] ,

[0148] where n is the number of links, m iis the mass of the link i, I i is the inertia tensor matrix of the link i in the center of mass coordinate system, f ci is the Coulomb friction coefficient of the link i, f vi is the viscous friction coefficient of the link i.

[0149] For this step, the least square method can be directly used to solve the minimum inertia parameter set β b However, the unconstrained least square method cannot guarantee the physical meaning of the minimum inertia parameter set, for example, the mass of the link solved may be negative, which is contrary to the actual physical meaning. Therefore, the physical meaning of the inertia parameter needs to be considered in the solving process. For the link i, its mass m i and the inertia tensor matrix I i in the center of mass coordinate system must satisfy the following physical constraints: That is, the mass m i of the link must be positive, and the inertia tensor matrix I i in the center of mass coordinate system must be a positive definite matrix. This embodiment considers the physical meaning of the inertia parameter in the solving process, avoiding being contrary to the actual physical meaning.

[0150] In yet another optional embodiment, the optimization index of the excitation trajectory is to minimize the condition number of the observation matrix, and the optimization objective function of the excitation trajectory is:

[0151] ,

[0152] wherein, is the condition number of the observation matrix, , , , , , are the upper and lower limits of the rotatable positions, velocities and accelerations of each joint of the robot, respectively; , are the initial joint velocities and accelerations of the robot at the initial time, , are the joint velocities and accelerations of the robot at the terminal time .

[0153] In yet another optional embodiment, the initial value of the covariance matrix P at the initial time is set as a diagonal matrix of a x a:

[0154] ,

[0155] wherein, a is equal to the number of friction parameters of all joints, Take 10^6. For example, there are 3 friction parameters in each joint, and there are 18 parameters in a 6-DOF robot arm, so It is an 18x18 diagonal matrix, and each value is .

[0156] Example two

[0157] Referring to Figure 5 , the embodiment of the application discloses a robot motion real-time compensation control method, the robot has a feedback control system, and the method comprises the steps of:

[0158] S100. When the robot runs any trajectory, the motion data of each joint of the robot in the current time period is collected, the compensation friction torque corresponding to the current time period is calculated according to the robot motion real-time parameter identification method in any one of claims 1-7 , the overall regression matrix and the inertia base parameter , and the feedforward compensation torque is calculated according to the following formula :

[0159] .

[0160] S200. The feedforward compensation torque is provided as the actual value of the friction torque of the current robot to the feedback control system.

[0161] The embodiment of the application calculates the feedforward compensation torque once after performing a parameter identification, which is used for the feedback system of the robot to realize the robot motion real-time compensation control, improves the robot motion control precision, and can be used in robot mobile handling, polishing and polishing, and heavy load hydraulic industrial scenes.

[0162] Example three

[0163] Referring to Figure 6 , the embodiment of the application discloses a robot, comprising a memory 201, a processor 202 and a computer program stored on the memory, characterized in that the processor executes the computer program to realize the steps of the method as claimed in embodiment one or embodiment two.

[0164] Example four

[0165] The embodiment of the application discloses a computer readable storage medium, which stores a computer program, and the computer program is characterized in that when the computer program is executed by a processor, the steps of the method as claimed in embodiment one or embodiment two are realized.

[0166] The embodiments disclosed in the content of the present application are only the preferred embodiments of the present application, and are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand; the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced equivalently; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for real-time identification of robot motion parameters, characterized in that, The method includes the following steps: S1. Based on the trajectory information of the robot joints at the current time k. , , To calculate the overall regression matrix at the current time. and friction regression matrix :in, , , These are the position, velocity, and acceleration vectors of the robot's joints, respectively. for , for ; S2. Calculate the gain matrix at the current time step. The formula is as follows: ， In the formula, Let I be the friction regression matrix calculated at the previous time step, and let I be the unit diagonal matrix. for The transpose of the matrix, It is the covariance matrix of the previous time step; S3. Calculate the estimated frictional torque at the current moment. The formula is as follows: ， In the formula, The joint torque at the current moment, The overall regression matrix at the current moment, These are the fundamental parameters of inertia; S4. Based on the friction parameters from the previous moment. To calculate the estimation error : ； S5. Due to estimation error and gain matrix Update the robot's current friction parameters : ； S6. Calculate the covariance matrix at the current time. This is used to reserve a time for calculation. ； S7. Based on the current friction regression matrix and current friction parameters Calculate the compensating friction torque at the current moment. : 。 2. The method for real-time robot motion parameter identification according to claim 1, characterized in that, The sampling torque of each joint is calculated using the sampling current: The sampled current is first filtered: ， in, It is the current sampled at time k. It is the filter current at time k-1. It is the filter current at time k. These are the filter coefficients; The joint torque at time k can then be obtained using the following conversion formula. : ， in, For the reduction ratio, This is the torque coefficient.

3. The method for real-time robot motion parameter identification according to claim 1, characterized in that, Let be the friction regression matrix at time k, where , Let be the velocity of the i-th joint of the robot. It is a symbolic function; Overall Regression Matrix It is calculated using the following formula: ， in, For joint torque, It is the smallest set of inertial parameters to be identified, containing the basic inertial parameters and friction parameters of each link. The inertial regression matrix is ​​determined in advance based on the robot's own parameters. These are the inertial baseline parameters determined in advance based on the robot's own parameters. Friction parameters at the current moment .

4. The method for real-time robot motion parameter identification according to claim 1, characterized in that, Prior to step S1, the method further includes: The robot is controlled to move along an excitation trajectory, and motion data is collected during the operation to determine the friction parameters to be identified. Initial value at the initial moment ; The excitation trajectory is designed using a five-term finite Fourier series and is defined as follows: ， in, , and Let i represent the position, angular velocity, and angular acceleration of the i-th joint at time t, respectively. It is the joint angle at the initial moment; It is the fundamental angular frequency of the Fourier series excitation trajectory. and These are the coefficients of the Fourier series excitation trajectory.

5. The method for real-time robot motion parameter identification according to claim 1, characterized in that, Prior to step S1, the method further includes: The robot is controlled to run along an excitation trajectory. During the operation, motion data of each joint of the robot is collected, and the following set of linear regression equations is derived: ， Where m is the number of dynamic data sets of the robot joints acquired when the robot moves according to the excitation trajectory, each dynamic data set contains the input torque, angle, angular velocity, and angular acceleration of all joints of the robot, and m is greater than the minimum set of inertial parameters. The number of parameters in; It is the error vector of nm × 1 least squares, where n is the robot's degrees of freedom; It is the overall regression matrix corresponding to a single time point. It refers to the m time points corresponding to the robot running along the excitation trajectory for a period of time. The overall observation matrix formed by; Let m be the joint torques corresponding to m time points; Solve for the minimum set of inertia parameters in the linear regression equation system. .

6. The method for real-time robot motion parameter identification according to claim 5, characterized in that, The minimum set of inertial parameters in solving the linear regression equation system Subsequently, the method further includes: Obtain the error vector ; For the error vector The errors are rearranged to form an m × n error matrix R, with the following structure: ， in, This represents the error vector consisting of m sets of torque errors from the i-th joint out of n joints; The covariance matrix C is calculated using the following formula: ， Where C is an n × n symmetric matrix, and cov is the covariance calculation function; Normalize the linear regression equations: ， in, It is a weighted matrix; The minimum inertial parameter set is solved using the least squares method based on the following constraints. : ， Where n is the number of links, m i Let I be the mass of link i. i Let f be the inertia tensor matrix in the coordinate system of the center of mass of link i. ci f is the Coulomb friction coefficient of connecting rod i. vi It is the viscous friction coefficient of connecting rod i; The minimum set of inertial parameters obtained by the solution is determined to be the minimum set of inertial parameters. The initial value at the initial moment.

7. The method for real-time robot motion parameter identification according to claim 5, characterized in that, The solution to the minimum set of inertia parameters in the linear regression equation system. ,include: The minimum set of inertia parameters in the linear regression equations is solved using the least squares method based on the following constraints. : ， Where n is the number of links, m i Let I be the mass of link i. i Let f be the inertia tensor matrix in the coordinate system of the center of mass of link i. ci f is the Coulomb friction coefficient of connecting rod i. vi It is the viscous friction coefficient of connecting rod i.

8. The method for real-time robot motion parameter identification according to claim 4, characterized in that, Minimizing the condition number of the observation matrix is ​​used as the optimization index for the excitation trajectory, and the objective function for optimizing the excitation trajectory is: ， In the formula, The condition number of the observation matrix. , , , , , These represent the upper and lower limits of the rotatable positions, velocities, and accelerations of each joint of the robot. , These are the initial joint velocities and accelerations of the robot. , The termination time is respectively The speed and acceleration of each joint of the robot at that time.

9. A method for real-time motion compensation control of a robot, wherein the robot has a feedback control system, characterized in that, The method includes: When the robot runs an arbitrary trajectory, motion data of each joint of the robot in the current time period is collected, and the compensation friction torque corresponding to the current time period is calculated according to the robot motion real-time parameter identification method as described in any one of claims 1 to 8. , and Then calculate the feedforward compensation torque according to the following formula. : ， The feedforward compensation torque The actual value of the frictional torque of the robot is provided to the feedback control system.

10. A robot 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 as claimed in any one of claims 1-9.

Images