Robot adaptive impedance control method and device, computer equipment, readable storage medium and program product

Abstract

The application relates to a robot adaptive impedance control method and device, computer equipment, a readable storage medium and a program product. The method comprises the following steps: a motion control step: controlling a motion robot to move according to teaching motion parameter information of the motion robot and a preset spring damping constraint, and detecting actual motion parameter information of the motion robot; determining a motion parameter error formed between the teaching motion parameter information and the actual motion parameter information; correcting a motion impedance influence parameter in the preset spring damping constraint according to a preset local kernel function and the motion parameter error; updating the actual motion parameter information based on the teaching motion parameter information and the corrected motion impedance influence parameter, and returning to the motion control step until the motion parameter error is smaller than a preset error threshold, so that target actual motion parameter information is obtained; and controlling the motion of the motion robot according to the target actual motion parameter information. The method can accurately control the motion impedance.

Description

Technical Field

[0001] This application relates to the field of automatic control technology, and in particular to a robot adaptive impedance control method, apparatus, computer equipment, computer-readable storage medium, and computer program product. Background Technology

[0002] Controlling the motion impedance of a robot is greatly influenced by the dynamic parameters of its interaction environment. For example, changes in environmental stiffness can cause disturbances from external forces, which in turn affect the robot's position and speed. Therefore, controlling the motion impedance of a robot is a critical issue.

[0003] In traditional technology, the method for controlling the motion impedance of a robot is generally achieved by teaching the robot. Specifically, the robot is taught in advance by manually dragging it to obtain the taught motion position and taught motion speed. Then, according to the taught motion position and taught motion speed, the actual motion position and actual motion speed of the robot are controlled separately so that the actual motion position and actual motion speed of the robot approach the taught motion position and taught motion speed of the device, thereby achieving control of the motion impedance of the robot.

[0004] However, current robot impedance control methods still suffer from inaccuracies. Summary of the Invention

[0005] Therefore, it is necessary to provide an accurate robot adaptive impedance control method, device, computer equipment, computer-readable storage medium, and computer program product to address the above-mentioned technical problems.

[0006] In a first aspect, this application provides a robot adaptive impedance control method, including:

[0007] Obtain teaching motion parameter information of the robot;

[0008] Motion control steps: Based on the taught motion parameters and preset spring damping constraints, control the motion of the robot and detect the actual motion parameters of the robot. The preset spring damping constraints are used to characterize the relative correlation between the taught motion parameters and the actual motion parameters of the robot.

[0009] Determine the motion parameter error between the taught motion parameter information and the actual motion parameter information;

[0010] Based on the preset local kernel function and motion parameter error, the motion impedance influence parameter in the preset spring damping constraint is corrected. The motion impedance influence parameter is used to adjust the relative correlation parameter between the taught motion parameter information and the actual motion parameter information of the motion robot.

[0011] Based on the taught motion parameter information and the corrected motion impedance influence parameters, the actual motion parameter information of the motion robot is updated, and the motion control steps are returned until the motion parameter error is less than the preset error threshold, and the target actual motion parameter information is obtained.

[0012] The motion impedance of the robot is controlled based on the actual motion parameters of the target.

[0013] In one embodiment, the taught motion parameter information includes taught position parameters and taught velocity parameters, and the actual motion parameter information includes actual position parameters and actual velocity parameters; determining the motion parameter error between the taught motion parameter information and the actual motion parameter information includes:

[0014] The position parameter error is generated based on the taught position parameters and the actual position parameters, and the speed parameter error is generated based on the taught speed parameters and the actual speed parameters;

[0015] The position parameter error and the velocity parameter error are combined to obtain the motion parameter error.

[0016] In one embodiment, the motion impedance influence parameters in the preset spring damping constraint are corrected based on the preset local kernel function and motion parameter error, including:

[0017] Obtain the preset local kernel function and the preset total kernel function. The preset local kernel function is used to characterize the kernel function corresponding to each motion parameter error, and the preset total kernel function is used to characterize the sum of the kernel functions corresponding to all motion parameter errors.

[0018] For any motion parameter error, the local proportion of the preset local kernel function in the preset total kernel function is obtained, and based on the local proportion and the motion parameter error, the motion impedance influence parameter corresponding to the motion parameter error in the preset spring damping constraint is corrected.

[0019] In one embodiment, the motion impedance influence parameters include feedforward parameters, stiffness parameters, and damping parameters; based on local proportions and motion parameter errors, the motion impedance influence parameters corresponding to the motion parameter errors in the preset spring damping constraint are corrected, including:

[0020] Based on the position parameter error and velocity parameter error, the combined parameter error is obtained;

[0021] The stiffness parameters are corrected based on local proportion, combined parameter error, and position parameter error.

[0022] Based on local proportion, combined parameter error, and velocity parameter error, the damping parameters are corrected.

[0023] The feedforward parameters are corrected based on the local ratio and combined parameter errors.

[0024] In one embodiment, a combined parameter error is obtained based on the position parameter error and the velocity parameter error, including:

[0025] Obtain the preset position weight coefficient and the preset velocity weight coefficient. The preset position weight coefficient is used to represent the weight of the position parameter error in all parameter errors, and the preset velocity weight coefficient is used to represent the weight of the velocity parameter error in all parameter errors. All parameter errors are the sum of the position parameter error and the velocity parameter error shown.

[0026] The position parameter error is updated based on the preset position weight coefficient, and the velocity parameter error is updated based on the preset velocity weight coefficient;

[0027] The updated position parameter error is fused with the updated velocity parameter error to obtain the combined parameter error.

[0028] In one embodiment, acquiring the taught motion parameters of the robot includes:

[0029] While controlling the end effector motion of the robot, detect the taught joint motion parameters of the robot in the joint coordinate system;

[0030] The coordinate system is transformed by the taught joint motion parameters to obtain the taught motion parameters of the robot in the end-effector coordinate system.

[0031] Secondly, this application also provides a robot adaptive impedance control device, comprising:

[0032] The teaching motion parameter acquisition module is used to acquire the teaching motion parameter information of the robot.

[0033] The actual motion parameter acquisition module is used for the motion control steps: controlling the motion of the robot based on the taught motion parameter information and the preset spring damping constraint, and detecting the actual motion parameter information of the robot. The preset spring damping constraint is used to characterize the relative correlation parameters between the taught motion parameter information and the actual motion parameter information of the robot.

[0034] The error determination module is used to determine the motion parameter error between the taught motion parameter information and the actual motion parameter information;

[0035] The influence parameter correction module is used to correct the motion impedance influence parameters in the preset spring damping constraint based on the preset local kernel function and motion parameter error. The motion impedance influence parameters are used to adjust the relative correlation parameters between the taught motion parameter information and the actual motion parameter information of the motion robot.

[0036] The motion parameter correction module is used to update the actual motion parameter information of the robot based on the taught motion parameter information and the corrected motion impedance influence parameters, and return to the motion control steps until the motion parameter error is less than the preset error threshold, and obtain the target actual motion parameter information.

[0037] The impedance control module is used to control the motion impedance of the robot based on the actual motion parameters of the target.

[0038] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0039] Obtain teaching motion parameter information of the robot;

[0040] Motion control steps: Based on the taught motion parameters and preset spring damping constraints, control the motion of the robot and detect the actual motion parameters of the robot. The preset spring damping constraints are used to characterize the relative correlation between the taught motion parameters and the actual motion parameters of the robot.

[0041] Determine the motion parameter error between the taught motion parameter information and the actual motion parameter information;

[0042] Based on the preset local kernel function and motion parameter error, the motion impedance influence parameter in the preset spring damping constraint is corrected. The motion impedance influence parameter is used to adjust the relative correlation parameter between the taught motion parameter information and the actual motion parameter information of the motion robot.

[0043] Based on the taught motion parameter information and the corrected motion impedance influence parameters, the actual motion parameter information of the motion robot is updated, and the motion control steps are returned until the motion parameter error is less than the preset error threshold, and the target actual motion parameter information is obtained.

[0044] The motion impedance of the robot is controlled based on the actual motion parameters of the target.

[0045] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0046] Obtain teaching motion parameter information of the robot;

[0047] Motion control steps: Based on the taught motion parameters and preset spring damping constraints, control the motion of the robot and detect the actual motion parameters of the robot. The preset spring damping constraints are used to characterize the relative correlation between the taught motion parameters and the actual motion parameters of the robot.

[0048] Determine the motion parameter error between the taught motion parameter information and the actual motion parameter information;

[0049] Based on the preset local kernel function and motion parameter error, the motion impedance influence parameter in the preset spring damping constraint is corrected. The motion impedance influence parameter is used to adjust the relative correlation parameter between the taught motion parameter information and the actual motion parameter information of the motion robot.

[0050] Based on the taught motion parameter information and the corrected motion impedance influence parameters, the actual motion parameter information of the motion robot is updated, and the motion control steps are returned until the motion parameter error is less than the preset error threshold, and the target actual motion parameter information is obtained.

[0051] The motion impedance of the robot is controlled based on the actual motion parameters of the target.

[0052] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0053] Obtain teaching motion parameter information of the robot;

[0054] Motion control steps: Based on the taught motion parameters and preset spring damping constraints, control the motion of the robot and detect the actual motion parameters of the robot. The preset spring damping constraints are used to characterize the relative correlation between the taught motion parameters and the actual motion parameters of the robot.

[0055] Determine the motion parameter error between the taught motion parameter information and the actual motion parameter information;

[0056] Based on the preset local kernel function and motion parameter error, the motion impedance influence parameter in the preset spring damping constraint is corrected. The motion impedance influence parameter is used to adjust the relative correlation parameter between the taught motion parameter information and the actual motion parameter information of the motion robot.

[0057] Based on the taught motion parameter information and the corrected motion impedance influence parameters, the actual motion parameter information of the motion robot is updated, and the motion control steps are returned until the motion parameter error is less than the preset error threshold, and the target actual motion parameter information is obtained.

[0058] The motion impedance of the robot is controlled based on the actual motion parameters of the target.

[0059] The aforementioned robot adaptive impedance control method, device, computer equipment, computer-readable storage medium, and computer program product differ from simply acquiring actual motion parameter information based on taught motion parameter information. This application provides a more accurate robot adaptive impedance control method, which involves detecting the actual motion parameter information of the robot when controlling its movement based on taught motion parameter information and preset spring damping constraints. By using a preset local kernel function and the motion parameter error formed between the taught and actual motion parameter information, the motion impedance influence parameter of the relative correlation parameter between the taught and actual motion parameter information of the robot is corrected. Then, based on the taught motion parameter information and the corrected motion impedance influence parameter, the actual motion parameter information of the robot is iteratively updated to obtain the target actual motion parameter information. Since the target actual motion parameter information is the actual motion parameter information when the motion parameter error is less than a preset error threshold, the motion impedance control of the robot based on the target actual motion parameter information is more accurate. Attached Figure Description

[0060] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0061] Figure 1 This is an application environment diagram of the robot adaptive impedance control method in one embodiment;

[0062] Figure 2 This is a flowchart illustrating a robot adaptive impedance control method in one embodiment;

[0063] Figure 3 This is a flowchart illustrating the robot adaptive impedance control method in another embodiment;

[0064] Figure 4 This is a schematic diagram of the environmental dynamics system and the spring damping system in one embodiment;

[0065] Figure 5This is a schematic diagram of a preset kernel function and a local ratio in one embodiment;

[0066] Figure 6 This is a schematic diagram illustrating how the stiffness, damping, and feedforward terms of the controller change over time as the number of iterations increases in one embodiment.

[0067] Figure 7 This is a schematic diagram illustrating how the position parameter error and velocity parameter error change over time as the number of iterations increases in one embodiment.

[0068] Figure 8 This is a structural block diagram of a robot adaptive impedance control device in one embodiment;

[0069] Figure 9 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0070] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application.

[0071] In related technologies, a motion robot is approximated as a flexible spring with adjustable stiffness, damping, and feedforward. The approximate spring force is used to meet the robotic arm's output force requirements to the environment. The method for controlling the motion impedance of a motion robot is generally achieved through teaching the robot. Specifically, the robot is pre-taught using a manual dragging method to obtain the taught motion position and speed. Then, based on the taught motion position and speed, the actual motion position and speed of the robot are controlled separately, so that the actual motion position and speed of the robot approach the taught motion position and speed of the device, thereby achieving control of the motion impedance of the robot.

[0072] However, while simple teaching can control the motion impedance of a robot so that its actual position and speed are close to the taught position and speed of the device, it cannot accurately control the robot's motion in some special cases.

[0073] For example, when a robot is pushed forcefully, its actual position and speed will not closely approximate the taught position and speed of the device due to the influence of the external force. The greater the external disturbance, the greater the deviation of the actual position from the taught position. In this case, the impedance control of the robot's motion is largely affected by the environmental contact stiffness (i.e., the external force). Therefore, a more accurate adaptive impedance control method for the robot is needed to control its motion, resulting in smaller motion errors when influenced by external forces.

[0074] The robot adaptive impedance control method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, the motion robot 102 communicates with the controller 104 via a network. A data storage system can store the data that the controller 104 needs to process. The data storage system can be integrated into the controller 104 or placed in the cloud or on another network server.

[0075] The controller 104 acquires the taught motion parameter information of the motion robot 102 uploaded by the user, or acquires the taught motion parameter information of the motion robot 102 pre-stored in the controller 104 from the storage module, in order to execute the motion control steps: control the motion robot 102 to move according to the taught motion parameter information and the preset spring damping constraint, and detect the actual motion parameter information of the motion robot 102, wherein the preset spring damping constraint is used to characterize the relative correlation parameter between the taught motion parameter information and the actual motion parameter information of the motion robot 102.

[0076] Furthermore, the controller 104 determines the motion parameter error between the taught motion parameter information and the actual motion parameter information. Based on the preset local kernel function and the motion parameter error, it corrects the motion impedance influence parameter in the preset spring damping constraint. The motion impedance influence parameter is used to adjust the relative correlation parameter between the taught motion parameter information and the actual motion parameter information of the motion robot 102. Based on the taught motion parameter information and the corrected motion impedance influence parameter, the controller updates the actual motion parameter information of the motion robot 102 and returns to the motion control step until the motion parameter error is less than the preset error threshold, thereby obtaining the target actual motion parameter information. Based on the target actual motion parameter information, the controller controls the motion impedance of the motion robot 102.

[0077] In one exemplary embodiment, such as Figure 2 As shown, an adaptive impedance control method for a robot is provided, which is applied to... Figure 1 The following explanation uses controller 104 as an example. Wherein:

[0078] S100 acquires the teaching motion parameter information of the motion robot.

[0079] Among them, robot teaching is a basic programming method. It involves guiding the robot and recording a series of guiding operations and movement commands, which are then used as teaching motion parameter information so that the robot can move automatically based on the teaching motion parameter information.

[0080] The taught motion parameters can be acquired in real time or retrieved from the controller's storage module. Retrieving them from the controller's storage module means storing the taught motion parameters in the controller's storage module and then retrieving them later to enable the robot to move according to those parameters. Furthermore, the taught motion parameters can originate not only from user-generated manual instructions but also from images captured by imaging devices.

[0081] In one embodiment, obtaining the teaching motion parameter information of the motion robot includes: the user manually guides the motion robot and records a series of guidance operations and motion commands, and when the robot needs to be controlled, the motion parameter information of the motion robot at this time is used as the teaching motion parameter information, and the teaching motion parameter information is uploaded to the controller in real time to execute the motion control of the motion robot.

[0082] In another embodiment, obtaining the teaching motion parameter information of the motion robot further includes: the user manually guides the motion robot and records a series of guidance operations and motion commands, and stores the motion parameter information of the motion robot at this time as teaching motion parameter information in the storage module of the controller. When the controller responds to the robot motion control request later, the controller extracts the teaching motion parameter information from the storage module to perform motion control of the motion robot based on the teaching motion parameter information.

[0083] In another embodiment, taking an optical camera as the shooting device, the acquisition of the teaching motion parameter information of the motion robot further includes: the user manually guides the motion robot, the optical camera captures images of the process of the user manually guiding the motion robot, the motion parameter information of the motion robot is extracted frame by frame based on the process images, and the motion parameter information of the motion robot is sent to the controller so that the controller can acquire the teaching motion parameter information of the motion robot.

[0084] S200, Motion Control Steps: Based on the taught motion parameter information and preset spring damping constraints, control the motion of the robot and detect the actual motion parameter information of the robot.

[0085] Among them, the preset spring damping constraint is used to characterize the relative correlation parameters between the taught motion parameters and the actual motion parameters of the robot.

[0086] Specifically, the motion robot is approximated as a flexible spring with adjustable parameters, using an approximate spring force to meet the robotic arm's output force requirements to the environment. In other words, the motion robot approximates a spring-damped system that interacts with the unknown environment.

[0087] When a robot interacts with an unknown environment, two control systems can be established, namely a spring-damped system and an environmental dynamics system. The spring-damped system can be regarded as the robot's control system, while the environmental dynamics system describes the model of the environment and outputs environmental forces.

[0088] The input to a spring-damped system is taught motion parameters, and the output is actual motion parameters. The system must satisfy preset spring-damped constraints, which are second-order dynamical system parameters characterizing the relative correlation between the taught and actual motion parameters of the robot. The robot's actual motion parameters are influenced by these preset spring-damped constraints, which are also related to the forces exerted on the system by the environment.

[0089] When a robot is subjected to forces in its environment, even if its movement is controlled based on taught motion parameters and preset spring damping constraints, the actual motion parameters will deviate from the taught parameters due to the forces acting on them. Therefore, the actual motion parameters need to be corrected. This correction requires first acquiring the robot's actual motion parameters at the current moment, i.e., the uncorrected actual motion parameters.

[0090] S300, determine the motion parameter error between the taught motion parameter information and the actual motion parameter information.

[0091] Specifically, the motion parameter error between the taught motion parameter information and the actual motion parameter information is determined, that is, the difference between the taught motion parameter information and the actual motion parameter information is obtained, and the difference between the taught motion parameter information and the actual motion parameter information is determined as the motion parameter error.

[0092] Furthermore, the teaching motion of a robot is a motion trajectory, not a point. Therefore, the teaching motion parameter information is the motion parameter information corresponding to several points in the trajectory. Thus, it can be determined that there is more than one motion parameter error. As time goes by, the teaching motion parameter information and the actual motion parameter information of the robot change, and the corresponding motion parameter errors are also different.

[0093] S400 corrects the motion impedance influence parameters in the preset spring damping constraint based on the preset local kernel function and motion parameter error.

[0094] Among them, the motion impedance influence parameter is used to adjust the relative correlation between the taught motion parameters and the actual motion parameters of the robot. The preset local kernel function is a function composed of Gaussian functions, usually represented by w(s). The preset local kernel function is a local function whose independent variable s is in the range of [0, 1]. It can achieve dimensionality reduction of data, that is, compress the computation of trajectory learning through the Gaussian kernel function.

[0095] Specifically, the parameters in the preset spring-damped constraint include not only actual motion parameters, taught motion parameters, and environmental input forces, but also motion impedance influence parameters. These motion impedance influence parameters are used to adjust the relative correlation between the taught and actual motion parameters of the robot. The purpose of this application is to adjust the motion impedance influence parameters so that the motion parameter errors gradually converge and decrease, and the actual and taught motion parameters gradually become more consistent.

[0096] The spring-damped system is essentially an error-driven negative feedback system. That is, based on the motion parameter error obtained from the spring-damped system, combined with a preset local kernel function, the motion impedance influence parameters in the preset spring-damped constraint are corrected by feedback.

[0097] The purpose of feedback correction is to influence the motion parameter error so that the motion parameter error is within the accuracy error range of the robot's motion. That is, when the motion impedance influence parameter is the optimal motion impedance influence parameter, the motion parameter error is less than the preset error threshold. Generally, the preset error threshold is set to 0.

[0098] S500 updates the actual motion parameter information of the robot based on the taught motion parameter information and the corrected motion impedance influence parameters, and returns to the motion control steps until the motion parameter error is less than the preset error threshold, thus obtaining the target actual motion parameter information.

[0099] Specifically, the preset spring damping constraint includes motion impedance influence parameters. Therefore, correcting the motion impedance influence parameters means updating the preset spring damping constraint. Then, based on the updated preset spring damping constraint, the motion robot is controlled again to update the actual motion parameter information of the motion robot.

[0100] After updating the actual motion parameters of the robot, the actual motion parameters may not be close to the taught motion parameters. The error between the actual and taught motion parameters may still be greater than the preset error threshold. Therefore, it is necessary to determine the error between the updated actual motion parameters and the taught motion parameters and compare the error with the preset error threshold.

[0101] When the motion parameter error is less than the preset error threshold, the actual motion parameter information at this time is taken as the target actual motion parameter information. When the motion parameter error is greater than or equal to the preset error threshold, the actual motion parameter information of the robot still needs to be corrected. That is, the motion impedance influence parameters in the preset spring damping constraint are corrected according to the preset local kernel function and the motion parameter error. This process is repeated iteratively to obtain the actual motion parameter information when the motion parameter error is less than the preset error threshold, and the actual motion parameter information when the motion parameter error is less than the preset error threshold is determined as the target actual motion parameter information.

[0102] The S600 controls the motion impedance of the robot based on the actual motion parameters of the target.

[0103] Among them, motion impedance characterizes the resistance tendency of a robot during movement. Depending on the characteristics of the environment, whether flexible or rigid, the robotic arm is expected to perform different compliant movements. Specifically, if the object being manipulated is flexible, a lower mechanical impedance is required to dissipate the interaction forces generated by contact; while if the robotic arm is to manipulate an object with a rigid surface, a higher impedance should be specified to ensure good stability of the object during operation.

[0104] Specifically, the target actual motion parameter information is the motion parameter information when the motion parameter error is less than the preset error threshold. In other words, the actual motion parameter information at this time is close to the taught motion parameter information. Therefore, the motion impedance of the robot can be controlled according to the target actual motion parameter information, so that even under the action of environmental forces, the robot can keep the error between the actual motion parameter information and the taught motion parameter information within the normal threshold range, thereby improving the accuracy of the robot's motion impedance control.

[0105] In the aforementioned adaptive impedance control method for robots, unlike simply obtaining actual motion parameter information based on taught motion parameter information, this application provides a more accurate adaptive impedance control method for robots. Specifically, when controlling the robot's motion based on taught motion parameter information and preset spring-damped constraints, the actual motion parameter information of the robot is detected. By using a preset local kernel function and the motion parameter error between the taught and actual motion parameter information, the motion impedance influence parameter of the relative correlation between the taught and actual motion parameter information is corrected. Then, based on the taught motion parameter information and the corrected motion impedance influence parameter, the actual motion parameter information of the robot is iteratively updated to obtain the target actual motion parameter information. Since the target actual motion parameter information is the actual motion parameter information when the motion parameter error is less than a preset error threshold, the motion control of the robot based on the target actual motion parameter information is more accurate.

[0106] In one exemplary embodiment, such as Figure 3 As shown, the teaching motion parameter information includes teaching position parameters and teaching velocity parameters, and the actual motion parameter information includes actual position parameters and actual velocity parameters. S300 includes:

[0107] S320 generates position parameter error based on the taught position parameters and the actual position parameters, and generates speed parameter error based on the taught speed parameters and the actual speed parameters.

[0108] S340 combines the position parameter error and the velocity parameter error to obtain the motion parameter error.

[0109] Specifically, motion parameter information includes position parameters and velocity parameters. Therefore, motion parameter errors include position parameter errors and velocity parameter errors.

[0110] The specific methods for obtaining the two parameter errors include: generating position parameter error based on the taught position parameters and the actual position parameters, and generating velocity parameter error based on the taught velocity parameters and the actual velocity parameters. More specifically, the difference between the taught position parameters and the actual position parameters is determined as the position parameter error, and the difference between the taught velocity parameters and the actual velocity parameters is determined as the velocity parameter error. Further, the position parameter error and the velocity parameter error are combined to obtain the motion parameter error; that is, the motion parameter error includes both position parameter error and velocity parameter error.

[0111] In one embodiment, when the actual motion parameter information with an error less than a preset error threshold is obtained, the actual motion parameter information is fed back and corrected. In essence, the actual position parameter and the actual velocity parameter are corrected separately so that the position parameter error is less than a preset position error threshold and the velocity parameter error is less than a preset velocity error threshold. Generally, both the preset position error threshold and the preset velocity error threshold are set to 0.

[0112] In this embodiment, by setting motion parameter information including position parameters and velocity parameters, the position and velocity of the motion parameter information can be comprehensively corrected. Unlike teaching only the position parameters or velocity parameters, the motion of the robot can be controlled more accurately based on the position parameter error and velocity parameter error.

[0113] In an exemplary embodiment, the motion impedance influence parameters in the preset spring damping constraint are corrected based on the preset local kernel function and motion parameter error, including:

[0114] Obtain the preset local kernel function and the preset total kernel function. For any motion parameter error, obtain the local proportion of the preset local kernel function in the preset total kernel function. Based on the local proportion and the motion parameter error, correct the motion impedance influence parameter corresponding to the motion parameter error in the preset spring damping constraint.

[0115] The preset local kernel function is used to characterize the kernel function corresponding to each motion parameter error, and describes a local function composed of Gaussian functions. The preset global kernel function is used to characterize the sum of the kernel functions corresponding to all motion parameter errors.

[0116] Specifically, each actual position parameter corresponds to a local kernel function, and the preset total kernel function is the sum of the local kernel functions corresponding to all actual position parameters. For example, let the local kernel function corresponding to the actual position parameter s be... Where i refers to the i-th position in the trajectory of the robot's motion, and the preset total kernel function at this time is... .

[0117] Since each actual position parameter corresponds to a motion parameter error, in order to simply describe the preset local kernel function, the preset local kernel function can be used as the kernel function representing each motion parameter error, and the preset total kernel function can be used as the sum of the kernel functions representing all motion parameter errors.

[0118] For any motion parameter error, the motion impedance influence parameter corresponding to the motion parameter error in the preset spring damping constraint is corrected by a discrete learning method.

[0119] Specifically, this involves obtaining the local proportion of a preset local kernel function within a preset overall kernel function. Let the preset local kernel function be... The default total kernel function is The local proportion of the preset local kernel function in the preset total kernel function is then determined. , where k is the kth iteration.

[0120] Furthermore, during the k-th iteration, based on the local proportion and motion parameter error corresponding to the k-th iteration, the motion impedance influence parameter corresponding to the motion parameter error in the preset spring damping constraint is corrected. That is, based on the local proportion of the motion parameter error using the Gaussian kernel function, a finite number of discrete motion parameter errors are controlled, thereby learning the motion impedance influence parameter corresponding to the motion parameter error in the preset spring damping constraint. In other words, let the motion parameter error be E. s，k In the k-th iteration, the corrected motion impedance influence parameter = the original motion impedance influence parameter + g s，k f(E) s，k ), where f(E) s，k ) is a function expression related to motion parameter error when the actual position parameter is s during the k-th iteration. The function expression includes at least one of motion parameter error and higher-order motion parameter error.

[0121] Furthermore, this application gradually reduces motion parameter errors by adjusting the motion impedance influence parameter. However, the motion impedance influence parameter is not a single value, but a continuous function described using a kernel function. The purpose is to ensure that the motion impedance influence parameter changes over time during a dynamic motion process. To achieve this, in each iteration, a kernel function with local scope is used to describe the stochastic process, thereby making the functional shape of the motion impedance influence parameter change with the actual motion parameter information. Furthermore, by controlling a finite number of discrete motion impedance influence parameters, the final motion impedance influence parameter in the current iteration is learned.

[0122] In other words, since the corrected motion impedance influence parameter is actually a discrete motion impedance influence parameter, further processing is required. This involves multiplying the discrete motion impedance influence parameter corresponding to each motion parameter error during this iteration by the local proportion, and then superimposing and updating the result to obtain the final motion impedance influence parameter. Specifically, let the corrected motion impedance influence parameter be... Then the expression can be the final motion impedance influence parameter = Where i refers to the i-th position in the trajectory of the robot's movement, and N is the total number of positions. It is actually a local proportion in the current iteration process.

[0123] In one embodiment, the motion parameter error includes the position parameter error e s，k Error with speed parameters In this case, the corrected motion impedance influence parameter = the original motion impedance influence parameter + g s，k f(e) s，k , ).

[0124] In another embodiment, the correction of the motion impedance influence parameter corresponding to the motion parameter error in the preset spring damping constraint is also related to the learning rate Q during the iteration process. k Regarding the learning rate Q during the iteration process... k Local scaling and motion parameter errors are used to correct the parameters affecting motion impedance. Among these, the learning rate Q... k It is a constant value used to control learning convergence. For example, let the motion parameter error be E. s，k In the k-th iteration, the corrected motion impedance influence parameter = the original motion impedance influence parameter + Q k g s，k f(E) s，k ).

[0125] Furthermore, the motion parameter error includes the position parameter error e s，k Error with speed parameters In this case, the corrected motion impedance influence parameter = the original motion impedance influence parameter + Q k g s，k f(e) s，k , ).

[0126] In this embodiment, by presetting the local proportion of the local kernel function in the presetting total kernel function, the discrete motion parameter error is controlled, thereby correcting the motion impedance influence parameter and making the corrected result more accurate.

[0127] In an exemplary embodiment, the motion impedance influence parameters include feedforward parameters, stiffness parameters, and damping parameters; based on local proportions and motion parameter errors, the motion impedance influence parameters corresponding to the motion parameter errors in the preset spring damping constraint are corrected, including:

[0128] Based on the position parameter error and velocity parameter error, the combined parameter error is obtained. Based on the local proportion, combined parameter error and position parameter error, the stiffness parameter is corrected. Based on the local proportion, combined parameter error and velocity parameter error, the damping parameter is corrected. Based on the local proportion and combined parameter error, the feedforward parameter is corrected.

[0129] Specifically, in the k-th iteration, the corrected motion impedance influence parameter = the original motion impedance influence parameter + Q k g s，k f(e) s，k , ), where f(E) s，k ) is a function expression related to motion parameter error when the actual position parameter is s during the k-th iteration. The function expression includes at least one of motion parameter error and higher-order motion parameter error.

[0130] The motion parameter errors in this application include not only position parameter errors or velocity parameter errors, but also position parameter errors e. s，k Error with speed parameters The combined parameter error. In other words, the combined parameter error. =E(e s，k , ).

[0131] Higher-order motion parameter errors refer to the product of two parameter errors, including but not limited to the product of combined parameter errors and position parameter errors, as well as the product of combined parameter errors and velocity parameter errors.

[0132] Therefore, the motion impedance influence parameters are discretized and corrected based on at least one of the following: position parameter error, velocity parameter error, product of combined parameter error and position parameter error, and product of combined parameter error and velocity parameter error.

[0133] Furthermore, the parameters affecting the motion impedance in a spring-damped system include three basic parameters: the feedforward parameter v, the stiffness parameter, and so on. and damping parameters Therefore, the preset spring damping constraint of the spring damping system includes .

[0134] The following describes the method for discretizing and correcting each motion impedance influence parameter, assuming the discrete motion impedance influence parameter is... By controlling multiple discrete motion impedance influence parameters, the final motion impedance influence parameters in the current iteration process can be learned.

[0135] For discrete stiffness parameters Correction: Based on local scale g s，k Combined parameter error and position parameter error For stiffness parameters Make corrections. More specifically, based on the combined parameter error. Error with position parameters The product of these factors yields the higher-order motion parameter errors, which are then based on the local proportional g. s，k Errors with higher-order motion parameters, affecting stiffness parameters The stiffness parameters are then corrected to obtain the discrete stiffness parameters. .

[0136] For discrete damping parameters Correction: Based on local scale g s，k Combined parameter error and speed parameter error Regarding damping parameters Make corrections. More specifically, based on the combined parameter error. Error with speed parameters The product of these factors yields the higher-order motion parameter errors, which are then based on the local proportional g. s，k Errors with higher-order motion parameters, affecting damping parameters The stiffness parameters are then corrected to obtain the discrete stiffness parameters. .

[0137] For feedforward parameters Correction: Based on local scale g s，k Error with combined parameters The feedforward parameter v is corrected to obtain the corrected discrete feedforward parameter. .

[0138] By using the above correction methods, each motion impedance parameter is discretized and corrected. If the error of the motion parameter is larger, the local kernel function to be adjusted can be increased.

[0139] Furthermore, during the iteration process, since higher-order motion parameter errors can be used to correct stiffness and damping parameters, stiffness and damping parameters can be used as the priority functions for adaptation.

[0140] In this embodiment, based on motion parameter errors, higher-order motion parameter errors, etc., various motion impedance influencing parameters are corrected accurately and effectively. Moreover, the motion impedance influencing parameters include feedforward parameters, stiffness parameters, and damping parameters, which can comprehensively correct the actual motion parameter information of the robot and improve the accuracy of the correction.

[0141] In an exemplary embodiment, a combined parameter error is obtained based on the position parameter error and the velocity parameter error, including:

[0142] Obtain preset position weight coefficients and preset velocity weight coefficients; update position parameter errors based on preset position weight coefficients and update velocity parameter errors based on preset velocity weight coefficients; fuse the updated position parameter errors and updated velocity parameter errors to obtain combined parameter errors.

[0143] The preset position weight coefficient represents the weight of position parameter error in all parameter errors, and the preset velocity weight coefficient represents the weight of velocity parameter error in all parameter errors. All parameter errors are the sum of the position parameter error and the velocity parameter error.

[0144] Specifically, combined parameter error , where e s，k For position parameter error, This represents the speed parameter error.

[0145] Obtain the preset position weight coefficient and preset velocity weight coefficient. The preset position weight coefficient is used to characterize the weight of the position parameter error in all parameter errors, that is, the proportion of the position parameter error in the sum of the position parameter error and the velocity parameter error. The preset velocity weight coefficient is used to characterize the weight of the velocity parameter error in all parameter errors, that is, the proportion of the velocity parameter error in the sum of the position parameter error and the velocity parameter error.

[0146] The position parameter error is updated based on the preset position weight coefficient, and the velocity parameter error is updated based on the preset velocity weight coefficient. The updated position parameter error and the updated velocity parameter error are then fused to obtain the combined parameter error.

[0147] The updated position parameter error and the updated velocity parameter error are fused to obtain the combined parameter error, which includes summing the updated position parameter error and the updated velocity parameter error to obtain the combined parameter error.

[0148] For example, let the position parameter error e s，k The preset position weight coefficient is Speed ​​parameter error The preset speed weighting coefficient is Then based on the preset position weight coefficient Update the position parameter error to obtain the updated position parameter error. And based on the preset speed weighting coefficient The updated speed parameter error is obtained as follows: Then based on the updated position parameter error Error with updated velocity parameters Generate combined parameter error Generally speaking, =1, which means the combined parameter error .

[0149] In this embodiment, by weighted summing of position parameter errors and velocity parameter errors, the position parameter errors and velocity parameter errors are accurately fused to obtain combined parameter errors. Based on the combination of combined parameter errors and other motion parameter errors, higher-order motion parameter errors are generated. Then, through higher-order motion parameter errors, the motion impedance influence parameters of the robot are more accurately corrected.

[0150] In one exemplary embodiment, obtaining the taught motion parameter information of the motion robot includes:

[0151] While controlling the end effector motion of the robot, the teaching joint motion parameters of the robot in the joint coordinate system are detected, and the coordinate system is transformed to obtain the teaching motion parameters of the robot in the end effector coordinate system.

[0152] The end effector of the motion robot is a force control unit equipped with a grinding head. This model enables the controller to be flexible, output force to the environment, and achieve active and flexible interaction with the environment.

[0153] Specifically, taking user teaching as an example, the user manually drags the end effector of the robot for teaching. The robot records the teaching joint motion parameters in the joint coordinate system at the current time based on the joint encoder. The teaching joint motion parameters include the teaching joint position parameters and the teaching joint velocity parameters, resulting in a sequence of teaching joint motion parameters, denoted as . , These are the taught joint position parameters at the current time. These are the taught joint velocity parameters at the current time, t=1, 2, ... .

[0154] Since the required information is the teaching motion parameters of the robot's end effector, a coordinate system transformation is necessary to convert the teaching joint motion parameters in the joint coordinate system to the teaching motion parameters of the robot's end effector in the end effector coordinate system.

[0155] In this embodiment, by detecting the joint encoder and recording the taught joint motion parameters of the joint in the joint coordinate system at the current time, and converting them into taught motion parameters of the robot in the end effector coordinate system, the teaching motion parameters are accurately obtained, thereby improving the accuracy of subsequent robot motion control.

[0156] In one exemplary embodiment, such as Figure 4 As shown, the environmental dynamics component (the transmission relationship between position and force) in an environmental dynamics system is typically a complex nonlinear function, as assumed to be:

[0157]

[0158] in, Characterizing the locational characteristics of the environment, i.e. These are the actual position parameters. These are actual speed parameters. This represents the dynamic parameters in the environmental model. It originates from the output of the previous spring-damped system. It is the force that the environmental dynamics system outputs to the environment.

[0159] In other words, the spring-damped system and the environmental dynamics system are equivalent to a second-order dynamics system consisting of a spring block and a damping block, with position as input and force as output.

[0160] For a spring-damped system, its motion characteristics are controlled by parameters affecting its kinematic impedance. These parameters include the feedforward parameter v and the stiffness parameter K. S and damping parameter K d It can be described as:

[0161]

[0162] in, It is the force that the environment inputs to the module. It is a position parameter error. It's a speed parameter error. , , For teaching position parameters, This refers to the teaching speed parameter.

[0163] The above formula can be simplified to:

[0164]

[0165] in, Characterizes multiple parameters affecting motion impedance.

[0166] That is, the purpose of this application is to control , making Approaching In this case, Approaching 0, where, It is characterized as the parameter affecting the optimal motion impedance.

[0167] The specific models of the three modules are explained below:

[0168] 1. Teaching Model:

[0169] Taking user teaching as an example, the main purpose of teaching is to obtain teaching motion parameter information, such as teaching position parameters and teaching speed parameters. The user manually drags the end effector to move it, while the robotic arm records the teaching joint motion parameter information of the joints in the joint coordinate system based on the joint encoder, ultimately obtaining a trajectory sequence. However, since what needs to be obtained is the teaching motion parameter information of the robot's end effector, a coordinate system transformation is required to convert the teaching joint motion parameter information of the joints in the joint coordinate system to the teaching motion parameter information of the robot's end effector in the end effector coordinate system.

[0170] 2. Admittance control model based on spring-damped system:

[0171] The admittance control model is also known as the preset spring damping constraint. The preset spring damping constraint includes multiple kinematic impedance influence parameters, including but not limited to the feedforward parameter v and the stiffness parameter. and damping parameters The preset spring damping constraint is as follows:

[0172]

[0173] 3. Iterative correction model:

[0174] Multiple discrete motion impedance influence parameters are learned separately using a discrete learning method. The learning method is as follows:

[0175]

[0176] Where k represents the k-th iteration, It is the learning rate, a constant value used to control learning convergence. It is the local proportion of the preset local kernel function within the preset global kernel function. , It is a local kernel function pre-defined as consisting of Gaussian functions. The independent variable s is the normalized actual position parameter, with an interval in [0, 1]. Let N = 20, as... Figure 5 The image shows a preset kernel function in one embodiment. (Left image) and local scale g s，k (See diagram on the right).

[0177] It is the combined parameter error, that is, the weighted sum of the position parameter error and the velocity parameter error. Generally speaking, , For position parameter error, This represents the speed parameter error.

[0178] Furthermore, by controlling a finite number of discrete motion impedance influencing parameters, such as , as well as This allows us to learn the final motion impedance parameters during the current iteration.

[0179] For example, taking stiffness parameters as an example, Through this Then, perform a final feedback correction on the stiffness parameters in the preset spring-damped constraint. Repeat this step to correct all motion impedance-affecting parameters.

[0180] Then, based on the taught motion parameter information and the corrected motion impedance influence parameters, the actual motion parameter information of the motion robot is updated to achieve iteration until the motion parameter error is less than the preset error threshold, and the target actual motion parameter information is obtained.

[0181] The motion of the robot is controlled based on the actual motion parameters of the target.

[0182] In one exemplary embodiment, the result of motion control of the robot is as follows:

[0183] 1. For example Figure 6 As shown, the stiffness, damping, and feedforward terms of the controller change with time as the number of iterations increases.

[0184] 2. For example Figure 7 As shown, with the increase of the number of iterations, the position parameter error and velocity parameter error gradually decrease, and the motion accuracy gradually improves.

[0185] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0186] Based on the same inventive concept, this application also provides a robot adaptive impedance control device for implementing the robot adaptive impedance control method described above. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more robot adaptive impedance control device embodiments provided below can be found in the limitations of the robot adaptive impedance control method described above, and will not be repeated here.

[0187] In one exemplary embodiment, such as Figure 8 As shown, a robot adaptive impedance control device is provided, comprising: a teaching motion parameter acquisition module 100, an actual motion parameter acquisition module 200, an error determination module 300, an influence parameter correction module 400, a motion parameter correction module 500, and an impedance control module 600, wherein:

[0188] The teaching motion parameter acquisition module 100 is used to acquire the teaching motion parameter information of the motion robot;

[0189] The actual motion parameter acquisition module 200 is used for the motion control steps: controlling the motion of the robot based on the taught motion parameter information and the preset spring damping constraint, and detecting the actual motion parameter information of the robot. The preset spring damping constraint is used to characterize the relative correlation parameters between the taught motion parameter information and the actual motion parameter information of the robot.

[0190] The error determination module 300 is used to determine the motion parameter error between the taught motion parameter information and the actual motion parameter information.

[0191] The influence parameter correction module 400 is used to correct the motion impedance influence parameters in the preset spring damping constraint according to the preset local kernel function and motion parameter error. The motion impedance influence parameters are used to adjust the relative correlation parameters between the taught motion parameter information and the actual motion parameter information of the motion robot.

[0192] The motion parameter correction module 500 is used to update the actual motion parameter information of the motion robot based on the taught motion parameter information and the corrected motion impedance influence parameters, and return to the motion control steps until the motion parameter error is less than the preset error threshold, and obtain the target actual motion parameter information.

[0193] The impedance control module 600 is used to control the motion impedance of the robot based on the actual motion parameters of the target.

[0194] In one embodiment, the taught motion parameter information includes taught position parameters and taught velocity parameters, and the actual motion parameter information includes actual position parameters and actual velocity parameters; the error determination module 300 is further configured to generate a position parameter error based on the taught position parameters and the actual position parameters, and to generate a velocity parameter error based on the taught velocity parameters and the actual velocity parameters; the position parameter error and the velocity parameter error are combined to obtain the motion parameter error.

[0195] In one embodiment, the influence parameter correction module 400 is further configured to obtain a preset local kernel function and a preset total kernel function, wherein the preset local kernel function is used to characterize the kernel function corresponding to each motion parameter error, and the preset total kernel function is used to characterize the sum of the kernel functions corresponding to all motion parameter errors; for any motion parameter error, the local proportion of the preset local kernel function in the preset total kernel function is obtained, and based on the local proportion and the motion parameter error, the motion impedance influence parameter corresponding to the motion parameter error in the preset spring damping constraint is corrected.

[0196] In one embodiment, the motion impedance influence parameters include feedforward parameters, stiffness parameters, and damping parameters; the influence parameter correction module 400 is further configured to obtain a combined parameter error based on the position parameter error and the velocity parameter error; correct the stiffness parameter based on the local proportion, the combined parameter error, and the position parameter error; correct the damping parameter based on the local proportion, the combined parameter error, and the velocity parameter error; and correct the feedforward parameter based on the local proportion and the combined parameter error.

[0197] In one embodiment, the influence parameter correction module 400 is further configured to obtain a preset position weight coefficient and a preset velocity weight coefficient, wherein the preset position weight coefficient is used to characterize the weight of the position parameter error in all parameter errors, and the preset velocity weight coefficient is used to characterize the weight of the velocity parameter error in all parameter errors, and all parameter errors are the sum of the position parameter error and the velocity parameter error; the position parameter error is updated based on the preset position weight coefficient, and the velocity parameter error is updated based on the preset velocity weight coefficient; the updated position parameter error and the updated velocity parameter error are fused to obtain a combined parameter error.

[0198] In one embodiment, the teaching motion parameter acquisition module 100 is further configured to detect the teaching joint motion parameter information of the motion robot in the joint coordinate system while controlling the end effector motion of the motion robot; and to perform coordinate system transformation on the teaching joint motion parameter information to obtain the teaching motion parameter information of the motion robot in the end effector coordinate system.

[0199] The modules in the aforementioned robot adaptive impedance control device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0200] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 9 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores data such as teaching motion parameters. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a robot adaptive impedance control method.

[0201] Those skilled in the art will understand that Figure 9 The structure shown is a block diagram of a partial structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0202] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0203] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0204] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0205] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0206] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0207] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A robot adaptive impedance control method, characterized in that, The method includes: Obtain teaching motion parameter information of the robot; Motion control steps: Based on the taught motion parameter information and the preset spring damping constraint, control the motion of the motion robot and detect the actual motion parameter information of the motion robot. The motion robot is used as a spring damping system that satisfies the preset spring damping constraint, which is used to characterize the relative correlation parameter between the taught motion parameter information and the actual motion parameter information of the motion robot. Determine the motion parameter error between the taught motion parameter information and the actual motion parameter information; Based on the preset local kernel function and the error of the motion parameters, the motion impedance influence parameter in the preset spring damping constraint is corrected, wherein the motion impedance influence parameter is used to adjust the relative correlation parameter between the taught motion parameter information and the actual motion parameter information of the motion robot; Based on the taught motion parameter information and the corrected motion impedance influence parameters, the actual motion parameter information of the motion robot is updated, and the motion control step is returned until the motion parameter error is less than the preset error threshold, and the target actual motion parameter information is obtained. Based on the actual motion parameters of the target, the motion impedance of the motion robot is controlled. The step of correcting the motion impedance influence parameters in the preset spring damping constraint based on the error between the preset local kernel function and the motion parameters includes: Obtain a preset local kernel function and a preset total kernel function, wherein the preset local kernel function is used to characterize the kernel function corresponding to each of the motion parameter errors, and the preset total kernel function is used to characterize the sum of the kernel functions corresponding to all the motion parameter errors; For any of the aforementioned motion parameter errors, the local proportion of the preset local kernel function in the preset total kernel function is obtained, and based on the local proportion and the motion parameter error, the motion impedance influence parameter corresponding to the motion parameter error in the preset spring damping constraint is corrected.

2. The method according to claim 1, characterized in that, The teaching motion parameter information includes teaching position parameters and teaching speed parameters, and the actual motion parameter information includes actual position parameters and actual speed parameters; determining the motion parameter error between the teaching motion parameter information and the actual motion parameter information includes: A position parameter error is generated based on the taught position parameters and the actual position parameters, and a speed parameter error is generated based on the taught speed parameters and the actual speed parameters; The position parameter error and the velocity parameter error are combined to obtain the motion parameter error.

3. The method according to claim 2, characterized in that, The motion impedance influence parameters include feedforward parameters, stiffness parameters, and damping parameters; the correction of the motion impedance influence parameters corresponding to the motion parameter errors in the preset spring damping constraint based on the local proportion and the motion parameter errors includes: Based on the position parameter error and the velocity parameter error, a combined parameter error is obtained; The stiffness parameter is corrected based on the local proportion, the combined parameter error, and the position parameter error. The damping parameters are corrected based on the local proportion, the combined parameter error, and the velocity parameter error. The feedforward parameters are corrected based on the local ratio and the error of the combined parameters.

4. The method according to claim 3, characterized in that, The method of obtaining the combined parameter error based on the position parameter error and the velocity parameter error includes: Obtain a preset position weight coefficient and a preset velocity weight coefficient, wherein the preset position weight coefficient is used to characterize the weight of the position parameter error in all parameter errors, and the preset velocity weight coefficient is used to characterize the weight of the velocity parameter error in all parameter errors, wherein all parameter errors are the sum of the position parameter error and the velocity parameter error. The position parameter error is updated based on the preset position weight coefficient, and the velocity parameter error is updated based on the preset velocity weight coefficient. The updated position parameter error is fused with the updated velocity parameter error to obtain the combined parameter error.

5. The method according to claim 1, characterized in that, The acquisition of the teaching motion parameter information of the motion robot includes: While controlling the end effector motion of the robot, the taught joint motion parameters of the robot in the joint coordinate system are detected; The teaching joint motion parameter information is transformed into a coordinate system to obtain the teaching motion parameter information of the robot in the end-effector coordinate system.

6. A robot adaptive impedance control device, characterized in that, The device includes: The teaching motion parameter acquisition module is used to acquire the teaching motion parameter information of the robot. The actual motion parameter acquisition module is used in the motion control steps: controlling the motion of the motion robot according to the taught motion parameter information and the preset spring damping constraint, and detecting the actual motion parameter information of the motion robot, wherein the preset spring damping constraint is used to characterize the relative correlation parameter between the taught motion parameter information and the actual motion parameter information of the motion robot; An error determination module is used to determine the motion parameter error between the taught motion parameter information and the actual motion parameter information; The influence parameter correction module is used to correct the motion impedance influence parameters in the preset spring damping constraint based on the preset local kernel function and the motion parameter error. The motion robot is used as a spring damping system that satisfies the preset spring damping constraint. The motion impedance influence parameters are used to adjust the relative correlation parameters between the taught motion parameter information and the actual motion parameter information of the motion robot. The motion parameter correction module is used to update the actual motion parameter information of the motion robot based on the taught motion parameter information and the corrected motion impedance influence parameters, and return to the motion control step until the motion parameter error is less than a preset error threshold, thereby obtaining the target actual motion parameter information. An impedance control module is used to control the motion impedance of the motion robot based on the actual motion parameter information of the target. The influence parameter correction module is further used to obtain a preset local kernel function and a preset total kernel function. The preset local kernel function is used to characterize the kernel function corresponding to each motion parameter error, and the preset total kernel function is used to characterize the sum of the kernel functions corresponding to all motion parameter errors. For any motion parameter error, the local proportion of the preset local kernel function in the preset total kernel function is obtained, and based on the local proportion and the motion parameter error, the motion impedance influence parameter corresponding to the motion parameter error in the preset spring damping constraint is corrected.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

9. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images