Mechanical arm joint parameter identification method and computer equipment

Abstract

The invention provides a mechanical arm joint parameter identification method and computer equipment, and relates to the technical field of robot control. According to the method, the improved Stribeck friction model is constructed for each solid joint of the target mechanical arm, and the joint friction force sampling values of the corresponding solid joints in the respective motor excitation rotating speed constant-speed sections of the motor rotating speed excitation tracks are obtained, and then according to the joint friction force sampling values of the solid joints at different motor excitation rotating speeds, the joint friction force sampling values of the solid joints at different motor excitation rotating speeds are obtained; carrying out model parameter iteration identification on the improved Stribeck friction model of the solid joint by utilizing a nonlinear least square method to obtain an actual friction model parameter of the solid joint, which meets a friction parameter distribution constraint piece; therefore, the joint performance parameter accurate identification function considering the joint nonlinear friction condition is achieved for different solid joints of the same mechanical arm, and the finally identified joint performance parameters can effectively reflect the real state of the joints of the mechanical arm.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of robot control, in particular to a mechanical arm joint parameter identification method and a computer device. BACKGROUND

[0002] With the continuous development of science and technology, robot technology has been widely valued by various industries due to its great research value and application value. The collaborative robot which can realize human-robot collaborative work function is an important research direction of robot technology today. In the actual use process of the collaborative robot, the feedforward control technology based on the mechanical arm dynamics model is usually used to realize the effective compromise of the control system tracking accuracy, the control system steady-state error and the control system dynamic quality. However, the control technology has a relatively high requirement on the accuracy of the mechanical arm dynamics model, and the joint performance parameters of the mechanical arm need to be effectively matched with the actual running state of the corresponding mechanical arm.

[0003] However, it is worth noting that the existing parameter identification scheme for the joint performance parameters of the mechanical arm mainly simplifies all the joints of the mechanical arm into a whole model for joint performance parameter identification. The joint design differences (including joint type differences, joint reducer differences, joint motor drive board differences, etc.) between different joints of the mechanical arm, the influence of the nonlinear friction conditions and the joint flexibility conditions of different joints of the mechanical arm in the actual motion process on the final parameter identification result are not considered in essence, resulting in that the final parameter identification result is too idealistic and cannot reflect the real performance parameters of the joints of the mechanical arm. SUMMARY

[0004] Therefore, the purpose of the present application is to provide a mechanical arm joint parameter identification method and a computer device, which can realize accurate joint performance parameter identification considering the joint nonlinear friction conditions for different entity joints of the same mechanical arm, so as to ensure that the joint performance parameters identified for the joints of the mechanical arm can effectively reflect the real state of the joints of the mechanical arm.

[0005] In order to achieve the above purpose, the technical solutions adopted by the embodiments of the present application are as follows: In a first aspect, the present application provides a mechanical arm joint parameter identification method, which comprises: For each entity joint of a target mechanical arm, an improved Stribeck friction model of the entity joint is constructed, and joint friction force sample values of the entity joint in a plurality of motor speed excitation trajectory uniform speed segments are obtained, wherein the motor excitation speeds corresponding to different uniform speed segments are different from each other; According to the sampled values of the joint friction force of the physical joint under different motor exciting rotating speeds, the model parameter iterative identification is performed on the improved Stribeck friction model of the physical joint by using a nonlinear least square method, and actual friction model parameters of the physical joint satisfying the friction parameter distribution constraint are obtained.

[0006] In an optional embodiment, the improved Stribeck friction model of a single physical joint is expressed as follows: ; The friction parameter distribution constraint of the improved Stribeck friction model of the physical joint is expressed as follows: ; wherein, is used to represent the motor rotating speed, is used to represent the joint friction force of the physical joint under the motor rotating speed , is used to represent the Coulomb friction force of the physical joint, is used to represent the static friction force of the physical joint, is used to represent the Stribeck characteristic speed of the physical joint, is used to represent the speed attenuation exponential factor of the physical joint, is used to represent the viscous friction coefficient of the physical joint, is used to represent the hyperbolic tangent function.

[0007] In an optional embodiment, the step of obtaining the actual friction model parameters of the physical joint satisfying the friction parameter distribution constraint according to the sampled values of the joint friction force of the physical joint under different motor exciting rotating speeds and by using the nonlinear least square method to perform the model parameter iterative identification on the improved Stribeck friction model of the physical joint comprises: For each iterative identification operation, the reference model parameters, the reference motor exciting rotating speed, the reference damping factor and the reference gradient of the current iterative identification operation are determined; According to the reference model parameters, the reference motor exciting rotating speed and the sampled values of the joint friction force matching the reference motor exciting rotating speed, the joint friction force residual is calculated based on the improved Stribeck friction model, and the target friction force residual of the current iterative identification operation is obtained; The target Jacobian matrix of the target friction force residual with respect to the reference model parameters is calculated, and the actual gradient of the current iterative identification operation is calculated according to the target Jacobian matrix and the target friction force residual; Calculating an adaptive gain ratio of this iterative identification operation according to the reference damping factor, the reference gradient, and the actual gradient, and determining an actual damping factor of this iterative identification operation according to the adaptive gain ratio and the reference damping factor; Substituting the actual damping factor and the actual gradient into the model parameter update equation associated with the improved Stribeck friction model to solve the model parameters, thereby obtaining the friction model parameters to be output that satisfy the friction parameter distribution constraint in this iterative identification operation; Based on the friction model parameters to be output, detecting whether the iterative identification operation satisfies the iteration termination condition; If it is detected whether the iterative identification operation satisfies the iteration termination condition, the friction model parameters to be output are directly used as the actual friction model parameters of the physical joint; otherwise, the next iterative identification operation is performed, wherein the reference model parameters of the next iterative identification operation are the friction model parameters to be output of the current iterative identification operation, the reference damping factor of the next iterative identification operation is the actual damping factor of the current iterative identification operation, and the reference gradient of the next iterative identification operation is the actual gradient of the current iterative identification operation.

[0008] In an optional embodiment, the step of calculating the joint friction force residual based on the improved Stribeck friction model according to the reference model parameters, the reference motor excitation speed, and the joint friction force sampling value matching the reference motor excitation speed to obtain the target friction force residual for this iterative identification operation includes: Substituting the reference model parameters and the reference motor excitation speed into the improved Stribeck friction model to predict the joint friction force, thereby obtaining a predicted value of the joint friction force for this iterative identification operation; A subtraction operation is performed on the joint friction force sample value that matches the reference motor excitation speed and the joint friction force prediction value to obtain the target friction force residual of this iterative identification operation.

[0009] In an alternative embodiment, a single solid joint is The adaptive gain ratio during the iterative identification operation is calculated using the following formula: ; in, Used to indicate that the entity joint is in Adaptive gain ratio of the iterative identification operation; Used to indicate that the entity joint is in The reference damping factor of the iterative identification operation is used to represent the damping factor of the entity joint in the first The actual damping factor of the iterative identification operation; a parameter update step length for representing the model parameter update equation; a reference model parameter for representing the entity joint in the first iteration identification operation, and for representing the to-be-output friction model parameter of the entity joint in the first iteration identification operation; a reference model parameter for representing the entity joint in the first iteration identification operation; a reference model parameter for representing the entity joint in the first iteration identification operation; a reference gradient for representing the entity joint in the first iteration identification operation; a reference gradient for representing the entity joint in the first iteration identification operation; a target friction force residual for representing the entity joint in the first iteration identification operation; a target Jacobian matrix for representing the target friction force residual of the entity joint in the first iteration identification operation with respect to the reference model parameter.

[0010] In an optional embodiment, the actual damping factor of a single entity joint in the first iteration identification operation is calculated by using the following function expression: wherein, an adaptive gain ratio for representing the entity joint in the first iteration identification operation; a reference damping factor for representing the entity joint in the first iteration identification operation, and for representing the actual damping factor of the entity joint in the first iteration identification operation; an actual damping factor for representing the entity joint in the first iteration identification operation; a scaling factor for representing the scaling factor.

[0011] In an optional embodiment, the model parameter update equation of a single entity joint in the first iteration identification operation is represented by using the following equation: wherein, a to-be-solved Jacobian matrix for representing the target friction force residual of the entity joint in the first iteration identification operation with respect to the to-be-output friction model parameter; ​Used to indicate that the entity joint is in The actual damping factor of the iterative identification operation; A parameter update step size for representing the model parameter update equation; Used to indicate that the entity joint is in The reference model parameters of the iterative identification operation; Used to indicate that the entity joint is in The actual gradient of the iterative identification operation; Used to represent the identity matrix.

[0012] In an optional embodiment, the identification method further includes: For each physical joint of the target robotic arm, a two-stiffness spring system model is constructed when the physical joint satisfies the flexible joint dynamics principle. The motor output shaft angular velocity and motor electromagnetic torque of the physical joint are sampled when the physical joint runs according to the motor position excitation trajectory. According to the sampled motor output shaft angular velocity and motor electromagnetic torque, a recursive least squares method with a variable forgetting factor is used to perform parameter iterative identification on the intermediate parameters involved in the differential equation of the two-stiffness spring system model, and in each intermediate parameter iterative identification process, a forgetting factor correction function involving different intermediate parameter identification residual states is called to update the forgetting factor to obtain the target intermediate parameters that meet the parameter identification termination condition; According to the numerical conversion relationship between the dynamic model parameters and the intermediate parameters in the corresponding differential equation of the two-stiffness spring system model, the actual dynamic model parameters of the physical joint that match the target intermediate parameters are calculated.

[0013] In an alternative embodiment, a single solid joint is The forgetting factor actually used in the iterative identification process of the intermediate parameters is expressed by the following forgetting factor correction function: ; in, , Used to indicate that the entity joint is in The forgetting factor actually used in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The intermediate parameter identification residual in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The actual observed function value corresponding to the least squares standard form of the two-stiffness spring system model during the iterative identification of the intermediate parameters, Used to indicate that the entity joint is in a predicted value of an observation function corresponding to the least square criterion in the secondary intermediate parameter iterative identification process, for representing the entity joint at the a coefficient matrix of an observation function corresponding to the least square criterion in the secondary intermediate parameter iterative identification process, for representing the entity joint at the an intermediate parameter identification result of the secondary intermediate parameter iterative identification operation, for representing an intermediate parameter identification residual threshold value, for representing the entity joint at the a mathematical expectation value between all known intermediate parameter identification residuals greater than in the secondary intermediate parameter iterative identification process, for representing the entity joint at the a standard deviation between the all known intermediate parameter identification residuals in the secondary intermediate parameter iterative identification process.

[0014] In a second aspect, the present application provides a computer device, comprising a processor and a memory, the memory storing a computer program capable of being executed by the processor, and the processor is capable of executing the computer program to implement the mechanical arm joint parameter identification method in any one of the preceding embodiments.

[0015] In this case, the beneficial effects of the embodiments of the present application can include the following: The present application constructs an improved Stribeck friction model for each entity joint of the target mechanical arm, and obtains joint friction force sampling values of the corresponding entity joint in the motor excitation speed uniform speed segment of each motor speed excitation trajectory. Then, according to the joint friction force sampling values of the entity joint under different motor excitation speeds, the improved Stribeck friction model of the entity joint is iteratively identified by using a nonlinear least square method to obtain the actual friction model parameters of the entity joint that satisfy the friction parameter distribution constraint (i.e., the joint performance parameters matched with the joint nonlinear friction condition), so as to realize the accurate identification of the joint performance parameters considering the joint nonlinear friction condition for different entity joints of the same mechanical arm, and make the finally identified joint performance parameters effectively reflect the real state of the mechanical arm joint.

[0016] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the following preferred embodiments are specifically described below, and the accompanying drawings are referred to for detailed description. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some of the embodiments of the present application, and therefore should not be regarded as a limitation on the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0018] Figure 1 The device composition schematic diagram of the computer device provided by the embodiments of the present application is shown in the figure. Figure 2 The flowchart of the method for identifying the joint parameters of the mechanical arm provided by the embodiments of the present application is shown in the figure. Figure 3 The flowchart of the method for identifying the joint parameters of the mechanical arm provided by the embodiments of the present application is shown in the figure. Figure 2 The flowchart of the sub-steps included in step S220 in the method for identifying the joint parameters of the mechanical arm provided by the embodiments of the present application is shown in the figure. Figure 4 The flowchart of the method for identifying the joint parameters of the mechanical arm provided by the embodiments of the present application is shown in the figure.

[0019] Figure: 10-computer device; 11-memory; 12-processor; 13-communication unit. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the embodiments of the present application more clear, the following will combine 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 some of the embodiments of the present application, but not all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0021] Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. 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.

[0022] It should be noted that: similar numbers and letters represent similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings.

[0023] In the description of the application, it needs to be understood that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship commonly understood by those skilled in the art, or the orientation or positional relationship commonly understood by those skilled in the art, or the orientation or positional relationship commonly understood by those skilled in the art, only for the convenience of describing the application and simplifying the description, and not indicating or implying that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the application.

[0024] In the description of the application, it also needs to be explained that, unless otherwise explicitly specified and limited, the terms "set", "install", "connect", "connect" should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the application can be understood according to the specific circumstances.

[0025] In addition, in the description of the application, it can be understood that the relationship terms such as "first" and "second" and the like are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations. Moreover, the terms "include", "contain" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or includes elements inherent to such process, method, article or device. Without more limitation, the element defined by the sentence "including a…" does not exclude the presence of other identical elements in the process, method, article or device including the element. For those skilled in the art, the specific meaning of the above terms in the application can be understood according to the specific circumstances.

[0026] Some embodiments of the application will be described in detail below with reference to the accompanying drawings. The following examples and features in the examples can be combined with each other without conflict.

[0027] Please refer to Figure 1 , Figure 1is a device composition schematic diagram of a computer device 10 provided by an embodiment of the present application. In an embodiment of the present application, the computer device 10 can be in communication connection with a to-be-identified mechanical arm (i.e., a target mechanical arm) that needs to achieve the effect of mechanical arm joint performance parameter identification, for controlling the to-be-identified mechanical arm to cooperate with the computer device 10 to perform accurate joint performance parameter identification when the to-be-identified mechanical arm is in an idle state, so as to ensure that the joint performance parameters respectively identified for each mechanical arm joint (i.e., a physical joint) of the to-be-identified mechanical arm can effectively reflect the real state of the mechanical arm joint (including the joint nonlinear friction condition of the corresponding mechanical arm joint in the actual motion process, and / or the joint flexibility condition of the mechanical arm joint in the actual motion process, etc.). Wherein, the to-be-identified mechanical arm can be, but is not limited to: a six-degree-of-freedom collaborative mechanical arm, a seven-degree-of-freedom collaborative mechanical arm, etc.; the computer device 10 can be a computing device independent of the to-be-identified mechanical arm, or can be a hardware module device integrated with the to-be-identified mechanical arm, wherein the computing device can be, but is not limited to: a personal computer, a cloud server, a notebook computer, a tablet computer, etc.

[0028] In an embodiment of the present application, the computer device 10 can include a memory 11, a processor 12 and a communication unit 13. Wherein, the memory 11, the processor 12 and the communication unit 13 are directly or indirectly electrically connected to each other to realize the transmission or interaction of data. For example, the memory 11, the processor 12 and the communication unit 13 can be electrically connected to each other through one or more communication buses or signal lines.

[0029] In an embodiment of the present application, the memory 11 can be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc. Wherein, the memory 11 is used to store a computer program, and the processor 12 can execute the computer program accordingly after receiving an execution instruction.

[0030] In the embodiments of the present application, the processor 12 can be an integrated circuit chip with signal processing capability. The processor 12 can be a general processor, including at least one of a central processing unit (CPU), a graphics processing unit (GPU), and a network processor (NP), a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, a discrete gate or transistor logic device, a discrete hardware component, etc. The general processor can be a microprocessor or the processor can also be any conventional processor, etc., which can implement or execute the disclosed methods, steps and logic block diagrams in the embodiments of the present application.

[0031] In the embodiments of the present application, the communication unit 13 is configured to establish a communication connection between the computer device 10 and other electronic devices through a network, and transceive data through the network, wherein the network includes a wired communication network and a wireless communication network. For example, when the to-be-identified robot arm is in an idle state, the computer device 10 can send one or more motor motion excitation trajectories to any one of the joints of the to-be-identified robot arm through the communication unit 13 to obtain motor motion data (for example, motor output shaft angular velocity, motor speed, motor electromagnetic torque, etc.) of the joint in the process of running according to the corresponding motor motion excitation trajectory, so as to realize accurate identification of joint performance parameters considering the nonlinear friction condition and / or the flexibility condition of the joint based on the obtained motor motion data, so that the finally identified joint performance parameters can effectively reflect the real state of the robot arm joint.

[0032] In the embodiments of the present application, the computer device 10 can pre-store a specific computer program related to the robot arm joint parameter identification function at the memory 11, and drive the to-be-identified robot arm to cooperate to realize accurate identification of joint performance parameters for different robot arm joints by driving the processor 12 to execute the specific computer program correspondingly when the to-be-identified robot arm is in an idle state, so that the finally identified joint performance parameters for different physical joints (i.e. robot arm joints) of the same robot arm can effectively reflect the real state of the robot arm joint, including the nonlinear friction condition of the corresponding robot arm joint in the actual motion process, and / or the flexibility condition of the robot arm joint in the actual motion process.

[0033] It can be understood that, Figure 1 It can be understood that, Figure 1more or fewer components than those shown in the figures, or have a different configuration of the components present in the figures. For example, the location of some components in the software realization can be different. Figure 1 Figure 1 The components shown in the figures can be realized in hardware, software, or a combination thereof.

[0034] In the present application, in order to ensure that the computer device 10 can drive the to-be-recognized mechanical arm to cooperate to realize the accurate identification of the joint performance parameters of different joints of the mechanical arm, and make the joint performance parameters finally identified for different physical joints (i.e., joints of the mechanical arm) of the same mechanical arm effectively reflect the real state of the joint of the mechanical arm, the embodiment of the present application realizes the foregoing purpose by providing a mechanical arm joint parameter identification method. The mechanical arm joint parameter identification method provided by the present application is described in detail below.

[0035] Please refer to Figure 2 , Figure 2 is one of the flowcharts of the mechanical arm joint parameter identification method provided by the embodiment of the present application. In the embodiment of the present application, Figure 2 The mechanical arm joint parameter identification method shown in the figure can include steps S210-S220 to realize the accurate identification of the joint performance parameters considering the nonlinear friction condition of the joint for different physical joints of the same mechanical arm, so that the joint performance parameters finally identified can effectively reflect the joint nonlinear friction condition of the corresponding joint of the mechanical arm in the actual motion process.

[0036] Step S210: For each physical joint of the target mechanical arm, an improved Stribeck friction model of the physical joint is constructed, and joint friction force sample values of the physical joint in a plurality of uniform speed segments of motor speed excitation trajectories are obtained, wherein the motor excitation speeds corresponding to different uniform speed segments are different from each other.

[0037] ​In the embodiment, the target robot arm is a robot arm to be identified for realizing the joint performance parameter identification effect of the robot arm; the uniform speed segment trajectories of the plurality of motor speed excitation trajectories corresponding to the single entity joint of the target robot arm can ensure that the entity joint rotates at a uniform speed in the speed mode according to different motor excitation speeds, wherein the single motor speed excitation trajectory can be divided into a forward rotation T-shaped speed curve and a reverse rotation T-shaped speed curve, so as to drive the corresponding entity joint motor to move from a starting rotation position to a certain specific rotation position according to the forward rotation T-shaped speed curve, and then move from the specific rotation position to the starting rotation position according to the reverse rotation T-shaped speed curve, wherein the uniform speed segment trajectories of the forward rotation T-shaped speed curve and the reverse rotation T-shaped speed curve included in the same motor speed excitation trajectory correspond to the same motor excitation speed, and each entity joint is in a torque balance state when moving according to the uniform speed segment trajectory of the single motor speed excitation trajectory, at which time the motor electromagnetic torque of the entity joint is approximately equal to the actual friction torque of the entity joint, the joint friction sampling value of the entity joint at the motor excitation speed corresponding to the uniform speed segment trajectory can be obtained by sampling the DQ axis current of the entity joint in the corresponding uniform speed segment trajectory and performing the conversion operation between the acting torque and the acting force, so as to effectively monitor the actual joint friction conditions of each robot arm joint of the target robot arm at different motor excitation speeds.

[0038] In the embodiment, for different entity joints of the same target robot arm, the Stribeck friction model (which is used to describe the nonlinear dynamic model of the friction force changing with the speed) can be used to effectively represent the actual friction force characteristics of the corresponding entity joint in the process of changing the joint speed, and by introducing the Tanh function to replace the Sign function in the original Stribeck friction model, the improved Stribeck friction model in the application can effectively improve the joint friction characteristic description defect problem of the original Stribeck friction model due to the zero-crossing discontinuity of the Sign function. At this time, for each entity joint of the target robot arm, the improved Stribeck friction model of the entity joint can be represented by the following formula: ; wherein, is used to represent the motor speed, is used to represent the joint friction force of the entity joint at the motor speed , is used to represent the Coulomb friction force of the entity joint, is used to represent the static friction force of the entity joint, is used to represent the Stribeck characteristic speed of the entity joint, a velocity decay exponent factor for representing the entity joint, a viscous friction coefficient for representing the entity joint, a hyperbolic tangent function. At this time, the joint performance parameter of the entity joint matching the joint nonlinear friction condition is the actual friction model parameter of the corresponding improved Stribeck friction model (which can be represented by , and the friction parameter distribution constraint of the corresponding improved Stribeck friction model is represented by .

[0039] Step S220: According to the joint friction force sample values of the entity joint at different motor excitation speeds, the model parameter iterative identification of the improved Stribeck friction model of the entity joint is performed by using the nonlinear least squares method, so as to obtain the actual friction model parameter of the entity joint satisfying the friction parameter distribution constraint.

[0040] In the embodiment, when the respective joint friction force sample values of the single entity joint of the target robot arm at multiple motor excitation speeds are obtained, the obtained multiple joint friction force sample values can be preprocessed by using the Chebyshev filter to eliminate outliers and high-frequency noise parts, so as to obtain the joint friction force sample values of the entity joint after preprocessing at different motor excitation speeds.

[0041] Then, the nonlinear least squares method based on the gradient descent method, the Newton method, the Gauss-Newton method or the Levenberg-Marquardt (LM) method can be used to perform the model parameter (i.e. ) iterative identification of the improved Stribeck friction model of the entity joint based on the joint friction force sample values of the entity joint after preprocessing at different motor excitation speeds, so as to obtain the actual friction model parameter of the entity joint satisfying the friction parameter distribution constraint and the minimization effect of the target optimization function, complete the joint performance parameter accurate identification function of different entity joints under the same target robot arm, and make the finally identified joint performance parameter effectively reflect the joint nonlinear friction condition of the corresponding robot arm joint in the actual motion process. For the improved Stribeck friction model of a single entity joint, the target optimization function of the improved Stribeck friction model in the application process of the nonlinear least squares method can be represented by , wherein is used to represent the target friction force residual of the entity joint in the th iterative identification operation, is used to represent the total number of iterative identification operations.

[0042] Optionally, please refer to Figure 3 , Figure 3 isFigure 2 The flowchart of the sub-steps included in step S220 is shown in FIG. 22. In the embodiment of the present application, the computer device 10 can use the nonlinear least square method based on the LM method to perform model parameter iterative identification on the improved Stribeck friction model of a single entity joint, and at this time, the step S220 can include sub-step S221 to sub-step S227, so as to effectively improve the convergence speed of the friction model parameter identification through the adaptive dynamic adjustment operation of the damping factor while improving the accuracy of the friction model parameter identification by using the nonlinear least square method based on the LM method.

[0043] In the sub-step S221, for each iterative identification operation, the reference model parameter, the reference motor excitation speed, the reference damping factor and the reference gradient of the current iterative identification operation are determined.

[0044] In the embodiment, the reference model parameter used in the first iterative identification operation is the pre-configured initial friction model parameter The reference damping factor used in the first iterative identification operation is the pre-configured initial damping factor The reference gradient used in the first iterative identification operation is the pre-configured initial gradient, and the reference motor excitation speed used in the first iterative identification operation is any one of the motor excitation speeds corresponding to the entity joint in the plurality of motor excitation speeds at step S210. For any one of the iterative identification operations other than the first iterative identification operation, the reference model parameter used in the current iterative identification operation is the to-be-output friction model parameter identified in the last iterative identification operation, the reference damping factor used in the current iterative identification operation is the actual damping factor adaptively adjusted in the last iterative identification operation, the reference gradient used in the current iterative identification operation is the actual gradient calculated in the last iterative identification operation, and the reference motor excitation speed used in the current iterative identification operation is any one of the motor excitation speeds corresponding to the entity joint in the plurality of motor excitation speeds at step S210. The reference motor excitation speeds used in different iterative identification operations can be the same or different.

[0045] In the sub-step S222, the joint friction force residual error of the current iterative identification operation is obtained by calculating the joint friction force residual error based on the improved Stribeck friction model according to the reference model parameter, the reference motor excitation speed and the joint friction force sample value matched with the reference motor excitation speed.

[0046] In the embodiment, for each iterative identification operation, the step of calculating the target friction force residual error of the current iterative identification operation can include: Substitute the reference model parameters and reference motor excitation speed of this iterative identification operation into the corresponding improved Stribeck friction model to predict the joint friction force, and obtain the predicted value of the joint friction force of this iterative identification operation; The joint friction force sample value that matches the reference motor excitation speed of this iterative identification operation is subtracted from the joint friction force prediction value of this iterative identification operation to obtain the target friction force residual of this iterative identification operation.

[0047] Sub-step S223 , calculating the target Jacobian matrix of the target friction residual relative to the reference model parameters, and calculating the actual gradient of this iterative identification operation based on the target Jacobian matrix and the target friction residual.

[0048] In this embodiment, when determining the When the target friction residual and reference model parameters of the iterative identification operation are obtained, the first derivative matrix of the target friction residual relative to the reference model parameters can be obtained. The target Jacobian matrix of the iterative identification operation. The actual gradient of the iterative identification operation can be obtained by ” is calculated, where Used to indicate that the entity joint is in The actual gradient of the iterative identification operation, Used to indicate that the entity joint is in The target friction residual of the iterative identification operation is, Used to indicate that the entity joint is in The target Jacobian matrix of the target friction residual relative to the reference model parameters for the iterative identification operation.

[0049] Sub-step S224 , calculating the adaptive gain ratio of this iterative identification operation according to the reference damping factor, the reference gradient, and the actual gradient, and determining the actual damping factor of this iterative identification operation according to the adaptive gain ratio and the reference damping factor.

[0050] In this embodiment, when determining the When the reference damping factor, reference gradient and actual gradient of the iterative identification operation are obtained, the following formula can be used to calculate the corresponding entity joint in the first Adaptive gain ratio during the iterative identification operation: ; in, Used to indicate that the entity joint is in Adaptive gain ratio of the iterative identification operation; Used to indicate that the entity joint is in The reference damping factor of the iterative identification operation is used to represent the damping factor of the entity joint in the first The actual damping factor of the iterative identification operation; A parameter update step size for representing the model parameter update equation; Used to indicate that the entity joint is in The reference model parameters of the iterative identification operation are used to represent the entity joint in the The friction model parameters to be outputted from the iterative identification operation; Used to indicate that the entity joint is in The reference model parameters of the iterative identification operation; Used to indicate that the entity joint is in The actual gradient of the iterative identification operation is used to represent the entity joint in the The reference gradient of the iterative identification operation; Used to indicate that the entity joint is in The actual gradient of the iterative identification operation; Used to indicate that the entity joint is in The target friction residual of the iterative identification operation is, Used to indicate that the entity joint is in The target Jacobian matrix of the target friction residual relative to the reference model parameters for the iterative identification operation.

[0051] Then, the computer device 10 will be based on a single entity joint in the first The adaptive gain ratio and reference damping factor of the iterative identification operation can be calculated using the following formula for the corresponding entity joint in the first The actual damping factor during the iterative identification operation: ; in, Used to indicate that the entity joint is in Adaptive gain ratio of the iterative identification operation; Used to indicate that the entity joint is in The reference damping factor of the iterative identification operation is used to represent the damping factor of the entity joint in the first The actual damping factor of the iterative identification operation; Used to indicate that the entity joint is in The actual damping factor of the iterative identification operation; Used to indicate the scaling factor, and its value range is 2~10.

[0052] In sub-step S225 , the actual damping factor and the actual gradient are substituted into the model parameter update equation associated with the improved Stribeck friction model to solve the model parameters, thereby obtaining the friction model parameters to be output that satisfy the friction parameter distribution constraint of this iterative identification operation.

[0053] In this embodiment, a single entity joint The model parameter update equation during the iterative identification operation is expressed as follows: ; in, Used to indicate that the entity joint is in The Jacobian matrix to be solved for the target friction force residual of the iterative identification operation relative to the friction model parameters to be output; Used to indicate that the entity joint is in The actual damping factor of the iterative identification operation; A parameter update step size for representing the model parameter update equation; Used to indicate that the entity joint is in The reference model parameters of the iterative identification operation; Used to indicate that the entity joint is in The actual gradient of the iterative identification operation; Used to represent the identity matrix.

[0054] Thus, when determining a single entity joint in the When the actual damping factor and actual gradient of the iterative identification operation are obtained, the actual damping factor and actual gradient of the iterative identification operation can be substituted into the model parameter update equation corresponding to the physical joint to solve the physical joint in the first The Jacobian matrix to be solved for the iterative identification operation is then The target friction force residual of the iterative identification operation and the solved Jacobian matrix to be solved are used to further solve the friction model parameters to be output that meet the friction parameter distribution constraints for the purpose of minimizing the above target optimization function.

[0055] Sub-step S226 , based on the friction model parameters to be output, detecting whether the iterative identification operation meets the iteration termination condition.

[0056] In this embodiment, when a certain entity joint is determined in the first After the output friction model parameters of the first iterative identification operation, the friction model parameters of the first The specific numerical distribution of the friction model parameters to be output and / or the actual gradient calculated during the first iterative identification operation is detected. The secondary iteration identification operation operates whether the current iteration satisfies an iteration termination condition to determine whether the iteration identification operation is to be terminated. The iteration termination condition can include any one or more of a combination of “ ”, ” and “ ”, wherein is used to represent a step conversion coefficient, is used to represent a preset gradient threshold value, is used to represent a gradient change amplitude threshold value, is used to represent an L2 norm, is used to represent a positive infinite norm.

[0057] In this embodiment, if it is detected that the current iteration identification operation satisfies the iteration termination condition, it indicates that the to-be-output friction model parameter of the current iteration identification operation can stably and accurately reflect the joint nonlinear friction condition of the corresponding joint of the robot arm in the actual movement process. At this time, sub-step S227 is executed. If it is detected that the current iteration identification operation does not satisfy the iteration termination condition, it indicates that the friction model parameter iteration identification operation needs to be continued. At this time, the process jumps to sub-step S221 to continue to execute the next iteration identification operation until the to-be-output friction model parameter that satisfies the iteration termination condition is determined. For the next iteration identification operation of the current iteration identification operation, the reference model parameter of the next iteration identification operation is the to-be-output friction model parameter of the current iteration identification operation, the reference damping factor of the next iteration identification operation is the actual damping factor of the current iteration identification operation, and the reference gradient of the next iteration identification operation is the actual gradient of the current iteration identification operation.

[0058] In sub-step S227, the to-be-output friction model parameter is directly used as the actual friction model parameter of the entity joint.

[0059] Thus, the application can execute the above-mentioned sub-steps S221 to S227 for each joint of the same robot arm to be identified to improve the friction model parameter identification accuracy by using the nonlinear least square method based on the LM method while effectively improving the friction model parameter identification convergence speed by the damping factor adaptive dynamic adjustment operation, so as to quickly and accurately achieve the joint performance parameter accurate identification function considering the joint nonlinear friction condition for different joints of the robot arm, so that the identified joint performance parameter (i.e., the actual friction model parameter) can effectively reflect the joint nonlinear friction condition of the joint of the robot arm in the actual movement process.

[0060] The application can implement the accurate identification function of the joint performance parameters considering the nonlinear friction condition of the joint for different physical joints of the same mechanical arm by performing the above steps S210-S220, so that the finally identified joint performance parameters can effectively reflect the joint nonlinear friction condition of the corresponding mechanical arm joint in the actual movement process.

[0061] Optionally, please refer to Figure 4 , Figure 4 is a second flowchart of the mechanical arm joint parameter identification method provided by the embodiment of the application. In the embodiment of the application, compared with the mechanical arm joint parameter identification method shown in Figure 2 , the mechanical arm joint parameter identification method shown in Figure 4 may further include steps S310-S330 to implement the accurate identification function of the joint performance parameters considering the joint flexibility condition for different physical joints of the same mechanical arm, and by introducing the forgetting factor adaptive correction function based on the Gaussian naive Bayes classifier, the convergence speed and accuracy of the joint performance parameter identification are improved, so that the finally identified joint performance parameters can effectively reflect the joint flexibility condition of the corresponding mechanical arm joint in the actual movement process.

[0062] In step S310, for each physical joint of the target mechanical arm, a two-stiffness spring system model of the physical joint when satisfying the flexible joint dynamics principle is constructed, and the motor output shaft angular velocity and motor electromagnetic torque of the physical joint when the motor position excitation trajectory is run are sampled.

[0063] In the embodiment, since each mechanical arm joint of the target mechanical arm can include a joint driving motor and a harmonic reducer, an elastic action appears between the joint motor and the joint connecting rod of the corresponding mechanical arm joint, at this time, the mechanical arm joint (i.e. the physical joint) can be equivalent to a two-stiffness spring system model when satisfying the flexible joint dynamics principle, wherein the joint motor output shaft and the joint connecting rod (load) are equivalent to rigid bodies respectively, and the intermediate transmission mechanism with elastic action is equivalent to a torsional spring, at this time, the motion differential equation of the two-stiffness spring system model of a single physical joint can be expressed by the following Laplace form equation set: ; Among them, is used to represent the motor electromagnetic torque of the corresponding physical joint, is used to represent the transmission torque of the corresponding physical joint, is used to represent the equivalent moment of inertia of the connecting rod end of the corresponding physical joint, is used to represent the equivalent moment of inertia of the motor end of the corresponding physical joint, is used to represent the motor output shaft angular velocity of the corresponding physical joint, a link angular velocity of the corresponding physical joint, an external load torque, a Laplace operator, a motor output shaft angular displacement of the corresponding physical joint, a link angular displacement of the corresponding physical joint, an equivalent spring torsional stiffness.

[0064] At this time, the transfer function of the two-stiffness spring system model about the motor output shaft angular velocity and the motor electromagnetic torque may be expressed as: ; wherein, an equivalent moment of inertia containing the external load of the joint motor. For the above transfer function, the motor output shaft angular velocity and the motor electromagnetic torque can be directly measured or obtained through measurement conversion, and thus in the process of the motion of the robot arm (which implements regular motion by placing the corresponding physical joint in the position mode through the motor position excitation trajectory), the motor output shaft angular velocity and the motor electromagnetic torque of the corresponding physical joint when running according to the motor position excitation trajectory can be sampled online, and the dynamic model parameters of the corresponding physical joint matched with the two-stiffness spring system model can be identified by using the parameter identification technology (which can be expressed by “ ”).

[0065] Step S320, according to the sampled motor output shaft angular velocity and motor electromagnetic torque, the intermediate parameters involved in the difference equation of the two-stiffness spring system model are iteratively identified by using the recursive least square method with variable forgetting factors, and the forgetting factor updating is performed by calling the forgetting factor correction function involving the residual state of the identification of different intermediate parameters in each intermediate parameter iterative identification process, so as to obtain the target intermediate parameters satisfying the parameter identification termination condition.

[0066] In the embodiment, the transfer function of a single physical joint can be converted into a corresponding Z-domain discrete transfer function by using a zero-order holder, and the corresponding Z-domain discrete transfer function can be converted into a difference equation of the corresponding two-stiffness spring system model by using the Z-transform lag theorem, at this time, the difference equation of the two-stiffness spring system model can be expressed as follows: ; wherein, an motor output shaft angular velocity of the corresponding physical joint at the th data sampling period, an motor electromagnetic torque of the corresponding physical joint at the th data sampling period, a data sampling period, Used to indicate the system resonant frequency.

[0067] On this basis, the least squares standard form of the difference equation of the above two-stiffness spring system model in the application of recursive least squares method is " ”, where the intermediate parameter of the least squares standard form is , the observation function result matrix of the least squares standard form is , the observation function coefficient matrix of the least squares standard form is The aforementioned intermediate parameters Respective and dynamic model parameters The numerical conversion relationship between them is as follows: .

[0068] In the application process of the recursive least squares method based on the above-mentioned least squares standard formula, it is necessary to detect whether the intermediate parameters to be output meet the parameter identification termination conditions (which may be but are not limited to: "the torque difference between the theoretical motor electromagnetic torque and the actual motor electromagnetic torque calculated using the dynamic model parameters corresponding to the intermediate parameters to be output is less than the preset torque error threshold", "whether the dynamic model parameters corresponding to the intermediate parameters to be output are consistent with the joint design ideas of the corresponding physical joint", etc.) when the intermediate parameters to be output do not meet the parameter identification termination conditions, start the next intermediate parameter iterative identification operation, otherwise directly use the intermediate parameters to be output that meet the parameter identification termination conditions as the target intermediate parameters for characterizing the joint flexibility state of the corresponding physical joint.

[0069] In this process, the For the first intermediate parameter iterative identification operation (where each intermediate parameter iterative identification operation corresponds to a data sampling cycle), The intermediate parameters to be output from the iterative identification operation of the intermediate parameters The calculation is done using the following function group: .

[0070] in, Used to indicate that the entity joint is in The intermediate parameter identification result of the intermediate parameter iterative identification operation (i.e. the intermediate parameter to be output), Used to indicate that the entity joint is in The forgetting factor actually used in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The covariance matrix calculated during the iterative identification of the intermediate parameters is Used to indicate that the entity joint is in The gain vector matrix calculated during the iterative identification of the intermediate parameters is Used to indicate that the entity joint is in The observation function coefficient matrix corresponding to the least squares standard form in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The covariance matrix calculated during the iterative identification of the intermediate parameters.

[0071] For a single entity joint, The forgetting factor actually used in the iterative identification process of the intermediate parameters For example, the Gaussian naive Bayes classifier can be introduced to divide different intermediate parameter identification residual states, and different forgetting factor correction functions can be configured for different intermediate parameter identification residual states, so that During the iterative identification of the intermediate parameters, the appropriate forgetting factor is dynamically adjusted to accurately identify the intermediate parameters, thereby improving the convergence speed and accuracy of the joint performance parameter identification. The forgetting factor actually used in the iterative identification process of the intermediate parameters is expressed by the following forgetting factor correction function: ; in, , Used to indicate that the entity joint is in The forgetting factor actually used in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The intermediate parameter identification residual in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The actual observed function value corresponding to the least squares standard form of the two-stiffness spring system model during the iterative identification of the intermediate parameters, Used to indicate that the entity joint is in The predicted value of the observation function corresponding to the least squares standard form in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The observation function coefficient matrix corresponding to the least squares standard form in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The intermediate parameter identification results of the intermediate parameter iterative identification operation, Used to represent the residual threshold of intermediate parameter identification, Used to indicate that the entity joint is in The intermediate parameter is greater than a mathematical expectation value between all known intermediate parameter identification residuals, for representing a standard deviation between the all known intermediate parameter identification residuals in the first intermediate parameter iteration identification process of the entity joint.

[0072] Step S330, according to the numerical conversion relationship between the dynamic model parameters in the corresponding difference equation and the intermediate parameters of the two stiffness spring system model, the actual dynamic model parameters of the entity joint matched with the target intermediate parameters are calculated.

[0073] Therefore, the present application can realize the accurate identification function of the joint performance parameters considering the joint flexibility of different entity joints of the same mechanical arm by executing the above steps S310~S330, and the convergence speed and accuracy of the joint performance parameter identification are improved by introducing the forgetting factor adaptive correction function based on the Gaussian naive Bayes classifier, so that the finally identified joint performance parameters can effectively reflect the joint flexibility of the corresponding mechanical arm joint in the actual motion process.

[0074] In the embodiments provided by the present application, it should be understood that the disclosed apparatus and method can also be implemented by other ways. The apparatus embodiments described above are only schematic, and for example, the flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architectures, functions and operation of the apparatus, method and computer program product according to the embodiments of the present application. In this regard, each block in the flowcharts and block diagrams can represent a module, a segment or a portion of code, which comprises one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the accompanying drawings. For example, two blocks shown in succession can in fact be executed substantially concurrently or in the reverse order, depending on the functionality involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented by dedicated hardware-based systems, which perform the specified functions or acts, or can be implemented by a combination of dedicated hardware-based systems and computer instructions.

[0075] In addition, each function module in each embodiment of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part. Each function provided by the present application, if realized in the form of a software function module and sold or used as an independent product, can be stored in a storage medium. Based on this understanding, the technical solutions of the present application or the parts that essentially contribute to the prior art or the parts of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a computer device (for example, a notebook computer, a collaborative robot arm, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing readable storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various storage program codes.

[0076] The above is only various embodiments of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for identifying joint parameters of a robotic arm, characterized in that: The identification method includes: For each physical joint of the target robotic arm, an improved Stribeck friction model is constructed for the physical joint, and the joint friction force sampling values ​​of the physical joint are obtained in each uniform speed segment of multiple motor speed excitation trajectories, where the motor excitation speeds corresponding to different uniform speed segments are different. According to the joint friction force sampling values ​​of the physical joint at different motor excitation speeds, the nonlinear least squares method is used to iteratively identify the model parameters of the improved Stribeck friction model of the physical joint, and the actual friction model parameters of the physical joint that meet the friction parameter distribution constraints are obtained.

2. The method according to claim 1, characterized in that The improved Stribeck friction model of a single solid joint is expressed as follows: ； The friction parameter distribution constraint of the improved Stribeck friction model of the solid joint is expressed as follows: ； in, Used to indicate the motor speed. Used to indicate the motor speed of the entity joint The friction force of the joints below Used to represent the Coulomb friction force of the joint of the entity, Used to represent the static friction of the entity joint, Used to represent the Stribeck characteristic velocity of the entity joint, Used to represent the velocity attenuation exponential factor of the entity joint, Used to represent the viscous friction coefficient of the entity joint, Used to represent the hyperbolic tangent function.

3. The identification method according to claim 1, characterized in that: The step of performing iterative model parameter identification on the improved Stribeck friction model of the physical joint using a nonlinear least squares method based on the joint friction force sampling values ​​of the physical joint at different motor excitation speeds to obtain actual friction model parameters of the physical joint that meet the friction parameter distribution constraint includes: For each iterative identification operation, determining reference model parameters, reference motor excitation speed, reference damping factor, and reference gradient for this iterative identification operation; Calculating the joint friction force residual based on the improved Stribeck friction model according to the reference model parameters, the reference motor excitation speed, and the joint friction force sampling value that matches the reference motor excitation speed, to obtain the target friction force residual for this iterative identification operation; Calculating a target Jacobian matrix of the target friction residual relative to the reference model parameters, and calculating an actual gradient of this iterative identification operation based on the target Jacobian matrix and the target friction residual; Calculating an adaptive gain ratio of this iterative identification operation according to the reference damping factor, the reference gradient, and the actual gradient, and determining an actual damping factor of this iterative identification operation according to the adaptive gain ratio and the reference damping factor; Substituting the actual damping factor and the actual gradient into the model parameter update equation associated with the improved Stribeck friction model to solve the model parameters, thereby obtaining the friction model parameters to be output that satisfy the friction parameter distribution constraint in this iterative identification operation; Based on the friction model parameters to be output, detecting whether the iterative identification operation satisfies the iteration termination condition; If it is detected whether the iterative identification operation satisfies the iteration termination condition, the friction model parameters to be output are directly used as the actual friction model parameters of the physical joint; otherwise, the next iterative identification operation is performed, wherein the reference model parameters of the next iterative identification operation are the friction model parameters to be output of the current iterative identification operation, the reference damping factor of the next iterative identification operation is the actual damping factor of the current iterative identification operation, and the reference gradient of the next iterative identification operation is the actual gradient of the current iterative identification operation.

4. The identification method according to claim 3, characterized in that: The step of calculating the joint friction force residual based on the improved Stribeck friction model according to the reference model parameters, the reference motor excitation speed, and the joint friction force sampling value matching the reference motor excitation speed to obtain the target friction force residual of this iterative identification operation includes: Substituting the reference model parameters and the reference motor excitation speed into the improved Stribeck friction model to predict the joint friction force, thereby obtaining a predicted value of the joint friction force for this iterative identification operation; A subtraction operation is performed on the joint friction force sample value that matches the reference motor excitation speed and the joint friction force prediction value to obtain the target friction force residual of this iterative identification operation.

5. The identification method according to claim 3, characterized in that: A single entity joint The adaptive gain ratio during the iterative identification operation is calculated using the following formula: ； in, Used to indicate that the entity joint is in Adaptive gain ratio of the iterative identification operation; Used to indicate that the entity joint is in The reference damping factor of the iterative identification operation is used to represent the damping factor of the entity joint in the first The actual damping factor of the iterative identification operation; A parameter update step size for representing the model parameter update equation; Used to indicate that the entity joint is in The reference model parameters of the iterative identification operation are used to represent the entity joint in the The friction model parameters to be outputted from the iterative identification operation; Used to indicate that the entity joint is in The reference model parameters of the iterative identification operation; Used to indicate that the entity joint is in The actual gradient of the iterative identification operation is used to represent the entity joint in the The reference gradient of the iterative identification operation; Used to indicate that the entity joint is in The actual gradient of the iterative identification operation; Used to indicate that the entity joint is in The target friction residual of the iterative identification operation is, Used to indicate that the entity joint is in The target Jacobian matrix of the target friction residual relative to the reference model parameters for the iterative identification operation.

6. The identification method according to claim 3, characterized in that: A single entity joint The actual damping factor during the iterative identification operation is calculated using the following function expression: ； in, Used to indicate that the entity joint is in Adaptive gain ratio of the iterative identification operation; Used to indicate that the entity joint is in The reference damping factor of the iterative identification operation is used to represent the damping factor of the entity joint in the first The actual damping factor of the iterative identification operation; Used to indicate that the entity joint is in The actual damping factor of the iterative identification operation; Used to indicate the scaling factor.

7. The identification method according to claim 3, characterized in that: A single entity joint The model parameter update equation during the iterative identification operation is expressed as follows: ； in, Used to indicate that the entity joint is in The Jacobian matrix to be solved for the target friction force residual of the iterative identification operation relative to the friction model parameters to be output; Used to indicate that the entity joint is in The actual damping factor of the iterative identification operation; A parameter update step size for representing the model parameter update equation; Used to indicate that the entity joint is in The reference model parameters of the iterative identification operation; Used to indicate that the entity joint is in The actual gradient of the iterative identification operation; Used to represent the identity matrix.

8. The identification method according to any one of claims 1 to 7, characterized in that: The identification method further includes: For each physical joint of the target robotic arm, a two-stiffness spring system model is constructed when the physical joint satisfies the flexible joint dynamics principle. The motor output shaft angular velocity and motor electromagnetic torque of the physical joint are sampled when the physical joint runs according to the motor position excitation trajectory. According to the sampled motor output shaft angular velocity and motor electromagnetic torque, a recursive least squares method with a variable forgetting factor is used to perform parameter iterative identification on the intermediate parameters involved in the differential equation of the two-stiffness spring system model, and in each intermediate parameter iterative identification process, a forgetting factor correction function involving different intermediate parameter identification residual states is called to update the forgetting factor to obtain the target intermediate parameters that meet the parameter identification termination condition; According to the numerical conversion relationship between the dynamic model parameters and the intermediate parameters in the corresponding differential equation of the two-stiffness spring system model, the actual dynamic model parameters of the physical joint that match the target intermediate parameters are calculated.

9. The identification method according to claim 8, characterized in that: A single entity joint The forgetting factor actually used in the iterative identification process of the intermediate parameters is expressed by the following forgetting factor correction function: ； in, , Used to indicate that the entity joint is in The forgetting factor actually used in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The intermediate parameter identification residual in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The actual observed function value corresponding to the least squares standard form of the two-stiffness spring system model during the iterative identification of the intermediate parameters, Used to indicate that the entity joint is in The predicted value of the observation function corresponding to the least squares standard form in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The observation function coefficient matrix corresponding to the least squares standard form in the iterative identification process of the intermediate parameters, Used to indicate that the entity joint is in The intermediate parameter identification results of the intermediate parameter iterative identification operation, Used to represent the residual threshold of intermediate parameter identification, Used to indicate that the entity joint is in The intermediate parameter is greater than The mathematical expectation value between all known intermediate parameter identification residuals, Used to indicate that the entity joint is in The standard deviation between all known intermediate parameter identification residuals in the intermediate parameter iterative identification process.

10. A computer device, characterized in that: It includes a processor and a memory, the memory stores a computer program that can be executed by the processor, and the processor can execute the computer program to implement the robot arm joint parameter identification method described in any one of claims 1 to 9.

Images