Teleoperation control method, system and robot device for a robot

Abstract

The application provides a teleoperation control method and system of a robot and a robot device, and relates to the technical field of robot application. Through adaptive adjustment of spring-damping gain and combination of feedforward compensation processing, different working conditions can be dynamically adapted, system stability, operation sensitivity and response speed are improved, tracking error is reduced, and operation accuracy and real-time performance are improved.

Description

Technical Field

[0001] This application relates to the field of robot application technology, and more specifically, to a remote operation control method, system, and robot device for a robot. Background Technology

[0002] A teleoperation system is a technical system in which an operator remotely controls a slave operating device (slave arm) to complete a task. It is widely used in nuclear industry, deep-sea operations, medical surgery, hazardous environment operations, and space robotics. During teleoperation, the operator not only needs to control the position and movement of the slave arm, but also needs to perceive the contact force information between the slave arm and the environment in real time to achieve a safe, precise, and natural operating experience.

[0003] Currently, teleoperation control methods for bilateral force feedback robots are widely used to achieve force and position information interaction between the master and slave arms, allowing operators to "feel" the forces exerted by the remote environment, thereby improving the immersion and safety of operation. The proportional-derivative (PD) force coupling method is a common teleoperation strategy, where the spring-damping coefficient is typically set to a fixed value to suppress system oscillations. However, existing methods have the following problems: when the slave arm collides or comes into contact with the environment, existing control methods often weaken the force feedback signal from the master arm due to excessively low damping settings, making it difficult for the operator to perceive the remote contact status in a timely and accurate manner, thus affecting operational safety and precision operation capabilities. Furthermore, while increasing the spring coefficient can improve the clarity of contact, excessive damping used to suppress system oscillations can reduce system response speed and sensitivity; conversely, reducing damping to improve interaction performance can easily lead to system instability. There is a clear contradiction between stability and interaction performance, making it difficult to simultaneously satisfy both.

[0004] There is currently no effective technical solution to the above problems. Summary of the Invention

[0005] The purpose of this application is to provide a remote operation control method, system, and robot device for robots, aiming to resolve the contradiction between stability and interactive performance caused by the fixed spring-damping coefficient in traditional teleoperation systems, ensuring that the operator can perceive the remote contact status in a timely and accurate manner, while maintaining system stability and response speed.

[0006] In a first aspect, this application provides a method for remotely controlling a robot, comprising: Based on the current pose of the robot's main arm end effector, determine the target pose of the robot's slave arm; Based on the target pose of the slave arm, slave arm following control is performed to obtain the current pose of the slave arm; Error processing is performed based on the current slave arm pose and the target slave arm pose to obtain the control error; Based on the control error, the spring-damping gain is adaptively adjusted in combination with a preset error threshold to obtain the spring-damping gain of the robot. Based on the spring-damping gain, and combined with the current speed of the robot's main arm and the corresponding preset feedforward coefficient of the robot, compensation processing is performed to obtain the desired output force at the end of the main arm; The joint torque vector is determined based on the expected output force at the end of the main arm, and the main arm is driven based on the joint torque vector.

[0007] Through the above technical solution, this application effectively solves the contradiction between stability and interactive performance caused by the fixed spring-damping coefficient in traditional teleoperation systems, ensuring that the operator can perceive the remote contact status in a timely and accurate manner, while maintaining system stability and response speed.

[0008] Optionally, the control error includes pose error and velocity error, the preset error threshold includes a velocity error threshold and a pose error threshold, and the step of adaptively adjusting the spring-damping gain of the robot based on the control error and the preset error threshold to obtain the spring-damping gain of the robot includes: If the speed error is not greater than the speed error threshold and the pose error is not less than the pose error threshold, obtain the preset spring-damping gain base of the robot. Based on the spring-damping gain base, and combined with the difference between the pose error and the pose error threshold, the spring-damping gain is adaptively adjusted to obtain the spring-damping gain of the robot.

[0009] Through the above technical solution, this application effectively solves the problems of the traditional remote operation control method, which weakens the force feedback signal due to the low damping setting when the slave arm collides or comes into contact with the environment, and makes it difficult to simultaneously meet the system stability and interaction performance under a high spring-damping coefficient.

[0010] Optionally, the step of adaptively adjusting the spring-damping gain of the robot based on the control error and a preset error threshold to obtain the spring-damping gain further includes: When the pose error is less than the pose error threshold, or when the velocity error is greater than the velocity error threshold, the preset default gain is determined as the spring-damping gain of the robot.

[0011] Through the above technical solution, this strategy of dynamically switching the gain based on the error threshold, combined with the strategy of using a power function to increase the spring-damped gain under normal contact conditions, significantly improves the robustness and safety of the remote operating system, enabling operators to obtain stable and accurate force feedback under various complex working conditions, thereby improving the immersion of operation and the ability to perform precise tasks.

[0012] Optionally, the step of performing error processing based on the current slave arm pose and the target slave arm pose to obtain the control error includes: The pose error is obtained by calculating the current slave arm pose and the target slave arm pose. Based on the pose error, differential processing is performed to obtain the velocity error; The pose error and the velocity error are used as the control error.

[0013] Optionally, the step of performing compensation processing based on the spring-damping gain, combined with the current speed of the robot's main arm and the corresponding preset feedforward coefficient of the robot, to obtain the desired output force at the end of the main arm includes: The spring-damping gain is filtered to obtain the filtered spring-damping gain; Based on the filtered spring-damping gain, and combined with the current speed of the robot's main arm and the corresponding preset feedforward coefficient of the robot, compensation processing is performed to obtain the desired output force at the end of the main arm.

[0014] Optionally, the step of performing compensation processing based on the filtered spring-damping gain, combined with the current speed of the robot's main arm and the corresponding preset feedforward coefficient of the robot, to obtain the desired output force at the end of the main arm includes: Based on the filtered spring-damping gain, combined with the pose error and the preset spring coefficient, spring compensation processing is performed to obtain spring force information; Based on the filtered spring-damping gain, and combined with the speed error and the preset damping coefficient, damping compensation processing is performed to obtain damping force information; By combining the current speed of the robot's main arm with the robot's corresponding preset feedforward coefficient, feedforward compensation processing is performed to obtain feedforward force information; The desired output force at the end of the main boom is obtained by superimposing the spring force information, the damping force information, and the feedforward force information.

[0015] Optionally, determining the joint torque vector based on the desired output force at the end of the main boom, and driving the main boom based on the joint torque vector, includes: Obtain the current gravity compensation information of the main arm; Based on the expected output force at the end of the main arm, and combined with the gravity compensation information, gravity compensation processing is performed on the main arm to obtain the main arm joint torque vector; The main arm is driven according to the main arm joint torque vector.

[0016] Optionally, determining the target pose of the robot's slave arm based on the current pose of the robot's master arm end effector includes: Obtain the current pose of the robot's end effector; The current pose of the main arm end is low-pass filtered to obtain the target pose of the slave arm.

[0017] Secondly, this application provides a remote operation control system for a robot, comprising: The determination module is used to determine the target pose of the robot's slave arm based on the current pose of the robot's main arm end effector. The control module is used to perform slave arm following control based on the target pose of the slave arm to obtain the current pose of the slave arm; The processing module is used to perform error processing based on the current slave arm pose and the slave arm target pose to obtain the control error; The adjustment module is used to adaptively adjust the spring-damping gain based on the control error and a preset error threshold to obtain the spring-damping gain of the robot. The compensation module is used to perform compensation processing based on the spring-damping gain, combined with the current speed of the robot's main arm and the corresponding preset feedforward coefficient of the robot, to obtain the desired output force at the end of the main arm; The drive module is used to determine the joint torque vector based on the expected output force at the end of the main arm, and drive the main arm based on the joint torque vector.

[0018] Thirdly, this application provides a robot device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; The processor, when executing a program stored in memory, implements the teleoperation control method of the robot described in any of the preceding descriptions.

[0019] As can be seen from the above, the remote operation control method, system and robot equipment of the robot provided in this application can dynamically adapt to different working conditions by adaptively adjusting the spring-damping gain and combining it with feedforward compensation processing, thereby improving system stability, operation sensitivity and response speed, while reducing tracking error and improving the accuracy and real-time performance of operation.

[0020] Other features and advantages of this application will be set forth in the following description and will be apparent in part from the description or may be learned by practicing embodiments of this application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings. Attached Figure Description

[0021] Figure 1 This is a flowchart of a remote operation control method for a robot provided in an embodiment of this application.

[0022] Figure 2 This is a schematic diagram of the teleoperation control structure provided in an embodiment of this application.

[0023] Figure 3 This is a schematic diagram of the structure of the remote operation control system for a robot provided in an embodiment of this application.

[0024] Figure 4 This is a schematic diagram of the structure of the robot device provided in the embodiments of this application.

[0025] Labeling Explanation: 201, Main Arm; 202, Slave Arm; 21, Determination Module; 22, Control Module; 23, Processing Module; 24, Adjustment Module; 25, Compensation Module; 26, Drive Module; 111, Processor; 112, Communication Interface; 113, Memory; 114, Communication Bus. Detailed Implementation

[0026] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0027] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0028] Please refer to Figure 1This application provides a remote operation control method, system, and robot device for robots, aiming to resolve the contradiction between stability and interactive performance caused by the fixed spring-damping coefficient in traditional teleoperation systems, ensuring that the operator can perceive the remote contact status in a timely and accurate manner, while maintaining system stability and response speed.

[0029] This application provides a method for remote control of a robot, including: Step S1: Determine the target pose of the robot's slave arm based on the current pose of the robot's main arm end effector. Step S2: Perform slave arm following control based on the slave arm target pose to obtain the current slave arm pose; Step S3: Perform error processing based on the current slave arm pose and the slave arm target pose to obtain the control error; Step S4: Based on the control error and combined with the preset error threshold, the spring-damping gain is adaptively adjusted to obtain the spring-damping gain of the robot. Step S5: Based on the spring-damping gain, and combined with the current speed of the robot's main arm and the robot's corresponding preset feedforward coefficient, compensation processing is performed to obtain the expected output force at the end of the main arm. Step S6: Determine the joint torque vector based on the expected output force at the end of the main arm, and drive the main arm according to the joint torque vector.

[0030] The current pose of the end effector refers to the position and orientation information of the robot's end effector in three-dimensional space. It is usually represented by a transformation matrix, which includes translational components (such as the three translational components in the x, y, and z directions) and rotational components (such as the three rotational components in the rx, ry, and rz directions). This current pose of the end effector is the input information that the system acquires in real time when the operator performs operations through the end effector.

[0031] Slave arm target pose: refers to the target position and orientation that the robot's slave arm needs to achieve when performing a task. This slave arm target pose is obtained by processing the current pose of the master arm's end effector and serves as a reference for the slave arm's following control.

[0032] Slave follower control: This refers to the motion control of the slave arm based on a given target pose (i.e., slave arm target pose), so that the end effector of the slave arm approaches or reaches the target pose as closely as possible. Slave follower control can be based on existing admittance control, which simulates mechanical impedance characteristics to make the slave arm exhibit compliance when interacting with the environment.

[0033] Specifically, such as Figure 2As shown, when operating the master arm 201, the current pose of the master arm's end effector is first obtained. For example, the target pose of the slave arm can be obtained by processing the current pose of the master arm's end effector with a low-pass filter. The slave arm 202 follows the target pose of the slave arm using an admittance control strategy and returns the current pose of the slave arm in real time.

[0034] Subsequently, error processing is performed based on the current slave arm pose and the slave arm target pose to obtain the control error, which includes pose error and velocity error. Specifically, the pose error can be obtained by multiplying the slave arm target pose by the inverse matrix of the current slave arm pose, i.e. ,in, For pose error, From the target pose of the arm, This represents the current slave arm pose. The velocity error is obtained by differential processing of the pose error.

[0035] When the arm 202 makes contact with the object, if the speed error is not greater than a preset speed error threshold and the pose error is not less than a preset pose error threshold, the robot obtains its corresponding preset spring-damping gain base. Based on the spring-damping gain base and the difference between the pose error and the pose error threshold, a power function is used to... Adaptive adjustment of the spring-damping gain is performed, where, for Directional spring-damping gain, The direction can be any one of the six directions: x, y, z, rx, ry, and rz. This is the base of the spring-damped gain. for Positional error in direction, for The pose error threshold in the direction. This adjustment mechanism causes the pose error to gradually increase after the follower arm comes into contact with the environment due to compliance, while the velocity after contact is usually small. At this time, the spring-damping gain will increase in a power function form, thereby improving the damping of the system and enhancing the clarity of force feedback.

[0036] When the pose error is less than the pose error threshold, or the velocity error is greater than the velocity error threshold, the preset default gain (e.g., 1.0) is determined as the spring-damped gain. This mechanism ensures that the system remains stable when the slave arm 202 is not in contact with the environment or when violent movements occur, avoiding oscillations caused by excessive gain.

[0037] To further improve stability, the adaptively adjusted spring-damping gain is subjected to low-pass filtering to obtain the filtered spring-damping gain.

[0038] Next, based on the filtered spring-damping gain, and combined with the current speed of the main boom and the preset feedforward coefficient, compensation processing is performed to obtain the desired output force at the end of the main boom. Through compensation processing, especially feedforward compensation, the response speed and accuracy of force feedback can be effectively improved.

[0039] Finally, the current gravity compensation information of the main arm 201 is obtained. Based on the expected output force at the end of the main arm, gravity compensation processing is performed on the main arm 201 in conjunction with the gravity compensation information to obtain the joint torque vector of the main arm. The main arm 201 is driven according to this joint torque vector, thereby accurately feeding back the contact force between the distal arm 202 and the environment to the operator, enabling the operator to clearly perceive the interaction force between the instrument and the contact object (such as biological tissue).

[0040] Through the above method, this application can adaptively increase the spring-damping gain when the slave arm 202 contacts the environment, enhancing the clarity of force feedback and enabling the operator to more accurately perceive the distal contact state. Simultaneously, during non-contact or vigorous movement, the spring-damping gain is maintained at its default value to ensure system stability. Compared to the traditional method of fixing the spring-damping coefficient, this application effectively resolves the contradiction between stability and interactive performance by dynamically adjusting the gain. For example, when a doctor is performing a fine cut, the pose error increases while the speed error is relatively small when the slave arm contacts the tissue; the system automatically increases the gain so that the doctor can clearly feel the cutting resistance and avoid excessive force. Conversely, when the doctor moves the master arm 201 rapidly, the speed error is large; the system sets the gain to its default value to avoid oscillations caused by high gain and ensure smooth operation. This adaptive adjustment mechanism enables the teleoperation system to provide excellent performance in different operating scenarios, significantly improving operational safety and precision operation capabilities.

[0041] In some implementations, the control error includes pose error and velocity error, and the preset error threshold includes a velocity error threshold and a pose error threshold. Based on the control error and in conjunction with the preset error threshold, the spring-damping gain is adaptively adjusted to obtain the robot's spring-damping gain, including: Under the condition that the speed error is not greater than the speed error threshold and the pose error is not less than the pose error threshold, obtain the preset spring-damping gain base of the robot. Based on the base of the spring-damping gain, and combined with the difference between the pose error and the pose error threshold, the spring-damping gain is adaptively adjusted to obtain the robot's spring-damping gain.

[0042] Control error is a key indicator measuring the deviation between the actual state and the target state of the slave arm. Pose error represents the difference between the current position and orientation of the slave arm and the target position and orientation (which can be quantified using Euclidean distance). Velocity error reflects the difference between the current motion velocity of the slave arm and the target motion velocity, and can be obtained by time differentiation of the pose error or by directly measuring the difference between the end-effector velocity and the target velocity. This error information forms the basis for subsequent adaptive adjustment of the spring-damping gain, ensuring that the system can respond accurately to different types of deviations. Preset error thresholds serve as reference standards for determining whether specific gain adjustments are needed to the current system state. The velocity error threshold defines the acceptable range of slave arm motion velocity deviation; for example, when the velocity error is below this threshold, it may mean that the slave arm is in a relatively stable contact or slow motion state. The pose error threshold defines the acceptable range of slave arm position and orientation deviation; for example, when the pose error exceeds this threshold, it may indicate that the slave arm has made contact with the environment or has a large positional deviation. These thresholds can be preset based on the specific application scenario of the robot system, operational accuracy requirements, and empirical data. When the velocity error is no greater than the velocity error threshold and the pose error is no less than the pose error threshold, the preset spring-damping gain base of the robot is obtained. This step defines the conditions for triggering a specific gain adjustment strategy. When the velocity error of the slave arm is at an acceptable low level (no greater than the velocity error threshold), but its pose error is relatively large (no less than the pose error threshold), this usually indicates that the slave arm has made contact with the environment and is experiencing a certain reaction force, but its motion speed has been effectively suppressed. In this specific situation, the system needs to obtain a preset spring-damping gain base. This spring-damping gain base can be a constant value pre-stored in the system memory, for example, it can be preset according to the specific application scenario of the robot system, the operational accuracy requirements, and empirical data; this spring-damping gain base serves as the benchmark for subsequent adaptive gain adjustment, ensuring that the gain adjustment can start from a reasonable and stable starting point when contact occurs. Based on the spring-damping gain base, combined with the difference between the pose error and the pose error threshold, the spring-damping gain is adaptively adjusted to obtain the robot's spring-damping gain. After obtaining the base of the spring-damped gain, the gain is further adjusted using the difference between the pose error and the pose error threshold. For example, The base of the spring-damped gain As a base, the difference between the pose error and the pose error threshold serves as an exponent, causing the spring-damping gain to increase exponentially as the pose error exceeds the pose error threshold. This adaptive adjustment mechanism allows the spring-damping gain to dynamically change according to the severity of the actual contact situation (reflected by the pose error), thereby effectively enhancing the operator's perception of the remote contact force while ensuring system stability.

[0043] This solution accurately identifies the contact state of the slave arm (low speed and large pose error) and dynamically adjusts the spring-damping gain in this state. This significantly improves the clarity and response speed of force feedback when the slave arm is in contact with the environment, while avoiding system instability or sluggish response caused by high gain in non-contact states. This conditional and adaptive gain adjustment strategy enables the system to significantly enhance the operator's perception of the remote environment while maintaining stability, thus effectively balancing the contradiction between the stability and interactive performance of the teleoperation system.

[0044] Through the above technical solution, this application effectively solves the problems of weakened force feedback signals due to excessively low damping settings and the difficulty in simultaneously satisfying stability and interaction performance under high spring-damping coefficients in traditional teleoperation control methods when the slave arm collides or comes into contact with the environment. Specifically, this solution accurately identifies the contact state of the slave arm, i.e., the velocity error is not greater than the velocity error threshold and the pose error is not less than the pose error threshold, thereby triggering targeted adaptive adjustment of the spring-damping gain. This adjustment mechanism uses a preset spring-damping gain base and combines it with the difference between the pose error and the pose error threshold for gain calculation, so that when the slave arm comes into contact with the environment, the spring-damping gain can be dynamically and non-linearly increased according to the degree of contact (the magnitude of the pose error). This refined adaptive adjustment enables the system to quickly and effectively improve the clarity and response speed of force feedback when the slave arm comes into contact with the environment. Operators can perceive the contact state and force of the remote environment more accurately and timely, thereby significantly improving the immersion, safety, and precision operation capabilities. Meanwhile, since the gain adjustment is triggered based on specific conditions, unnecessary gain increases in the non-contact state are avoided, thereby effectively suppressing system oscillations and maintaining system stability. Therefore, this solution significantly optimizes the interactive performance of the teleoperation system without sacrificing system stability, achieving an effective balance between stability and interactive performance.

[0045] In some implementations, the spring-damping gain of the robot is adaptively adjusted based on the control error and a preset error threshold to obtain the spring-damping gain. This further includes: When the pose error is less than the pose error threshold, or when the velocity error is greater than the velocity error threshold, the preset default gain is determined as the robot's spring-damping gain.

[0046] The pose error threshold is a pre-set upper limit for allowable pose error. When the pose error is less than this threshold, it usually means that the slave arm is very close to or in a relatively stable contact state. In this case, continuing to perform complex adaptive gain adjustments may cause the system to become oversensitive or generate unnecessary oscillations.

[0047] The speed error threshold is a pre-set upper limit for allowable speed error. When the speed error exceeds this threshold, it usually means that the slave arm is experiencing rapid motion changes, shocks, or unstable states, such as sudden collisions or loss of contact. In this situation, adaptive gain adjustment may not respond in time and could even exacerbate system instability.

[0048] The preset default gain is a pre-set fixed gain value, specifically set to 1.0 in this application. This default gain serves as a backup gain in a safe or stable mode, replacing adaptive gain calculations under specific abnormal or stable conditions to ensure the predictability and stability of system behavior. Setting it to 1.0 means that in these specific conditions, the spring-damping coefficient will directly use its original set value without additional amplification or reduction, thus providing a baseline, relatively conservative force feedback effect.

[0049] This application optimizes the adaptive adjustment process of the spring-damping gain by introducing a conditional judgment mechanism. During teleoperation, the pose and velocity errors of the slave arm are continuously monitored. When the pose error of the slave arm is less than a preset pose error threshold, it indicates that the slave arm is very close to the target pose or in a stable contact state. If an exponentially increasing adaptive gain is continued at this time, it may lead to overly sensitive force feedback and cause system oscillation. Simultaneously, when the velocity error of the slave arm exceeds a preset velocity error threshold, it indicates that the slave arm may have encountered a sudden impact or is in a state of violent motion. To address these specific situations, this application's solution optimizes the spring-damping gain... The default gain is directly set to 1.0. This mechanism ensures a rapid switch to a safe and stable gain mode when the slave arm is in a stable state (small pose error) or an unstable state (large velocity error). In this way, the proposed solution complements the strategy of using a power function to increase the spring-damped gain when the velocity error is not greater than the velocity error threshold and the pose error is not less than the pose error threshold. Together, they construct a more robust and intelligent adaptive adjustment strategy for the spring-damped gain, thereby optimizing the interactive performance of teleoperation while ensuring system stability.

[0050] Through the above technical solution, this strategy of dynamically switching the gain based on the error threshold, combined with the strategy of using a power function to increase the spring-damped gain under normal contact conditions, significantly improves the robustness and safety of the remote operating system, enabling operators to obtain stable and accurate force feedback under various complex working conditions, thereby improving the immersion of operation and the ability to perform precise tasks.

[0051] In some implementations, error processing is performed based on the current slave arm pose and the slave arm target pose to obtain the control error, including: The pose error is calculated by using the current slave arm pose and the target slave arm pose. Based on the pose error, differential processing is performed to obtain the velocity error; The pose error and velocity error are used as control errors.

[0052] Specifically, the error processing involves calculating the pose error using the current slave arm pose and the target slave arm pose; performing differential processing based on the pose error to obtain the velocity error; and using the pose error and velocity error as control errors. The current slave arm pose refers to the actual spatial position and orientation of the robot's slave arm at a given moment. This pose information can be acquired in real time by the slave arm's own sensor system, for example, through joint encoders combined with forward kinematics calculations, or through measurement and estimation using external vision systems, inertial measurement units (IMUs), and other sensors. Error calculation refers to quantifying the difference between the current slave arm pose and the target slave arm pose using mathematical methods. Its purpose is to accurately reflect the degree of deviation between the slave arm and the target, providing a basis for subsequent control decisions. For example, multiplying the target slave arm pose by the inverse matrix of the current slave arm pose yields a relative transformation matrix representing the transition from the current slave arm pose to the target pose; the translation and rotation components of this matrix constitute the pose error. Another approach is to extract the translation and rotation vectors representing the target pose and the current pose of the slave arm, respectively. Then, the translation vectors are directly subtracted, and the rotation vectors are calculated using the corresponding difference calculation, thus obtaining the linear and angular pose error components. The pose error *e* is the result of error calculation; it is a multi-dimensional vector (e.g., x, y, z, rx, ry, rz directions) or matrix containing the deviation information of the slave arm in spatial position and attitude, serving as the basis for feedback adjustments in the teleoperation control system. For example, the pose error *e* can be represented as a 6-dimensional vector, where the first three dimensions represent the error in spatial position, and the last three dimensions represent the error in spatial attitude. Differential processing is a mathematical operation used to approximate the rate of change of a signal over time, i.e., the derivative. In this context, it is used to extract velocity information from continuous or discrete pose error sequences, capturing the dynamic trend of the deviation between the slave arm and the target, and providing velocity feedback to the control system. One implementation method is to use the first-order difference method, which involves subtracting the pose error of the previous moment from the pose error of the current moment, and then dividing by the sampling time interval. Another approach is to employ a more complex numerical differentiation algorithm, or to combine it with a low-pass filter for filtered differentiation, to reduce the impact of noise on velocity error calculation. Velocity error, obtained after differential processing, represents the relative velocity deviation between the slave arm and the target in terms of spatial position and attitude; it reflects the dynamic performance of the slave arm in following the target's pose. For example, velocity error can be a 6-dimensional vector, where the first three dimensions represent linear velocity error and the last three dimensions represent angular velocity error. Control error refers to the comprehensive error signal formed by integrating pose error and velocity error. It provides the teleoperation control system with comprehensive information about the static and dynamic deviations between the slave arm and the target.

[0053] This solution precisely constructs the control error through explicit steps. First, it calculates the pose error directly using the current slave arm pose and the target slave arm pose. This process ensures the accuracy of the positional deviation, avoids error accumulation that might be introduced by indirect methods, and provides a reliable basis for subsequent processing. Then, differential processing is performed based on the obtained pose error to obtain the velocity error. This step utilizes the dynamic changes in the pose error to derive velocity information, rather than relying solely on direct velocity measurement. This simplifies system implementation, improves real-time performance, and effectively captures changes in the slave arm's motion state. Finally, the pose and velocity errors are integrated into a complete control error. This integration enhances the integrity of the control error, providing multi-dimensional and comprehensive data support for subsequent adaptive adjustment of the spring-damping gain. In this way, this solution provides more accurate and real-time control error, significantly improving the teleoperation system's perception accuracy and response efficiency to the contact status of the remote environment, and providing operators with a more natural and safer force feedback experience.

[0054] In some implementations, compensation is performed based on the spring-damping gain, combined with the robot's current arm speed and the robot's corresponding preset feedforward coefficient, to obtain the desired output force at the end of the arm, including: The spring-damping gain is filtered to obtain the filtered spring-damping gain. Based on the filtered spring-damping gain, and combined with the robot's current main arm speed and the robot's corresponding preset feedforward coefficient, compensation processing is performed to obtain the desired output force at the end of the main arm.

[0055] Specifically, filtering the spring-damped gain aims to smooth out any spring-damped gain values ​​that may arise during adaptive adjustment, eliminating noise or abrupt changes and ensuring the continuity and stability of the gain. For example, a mean filter can be used, with a suitable cutoff frequency set to remove high-frequency noise. The filtered spring-damped gain is the result of the above filtering process, characterized by being more stable and continuous than the original gain. This filtered spring-damped gain will serve as a key input parameter for subsequent compensation processing, ensuring the stability of the compensation calculation. Based on the filtered spring-damped gain, combined with the robot's current arm speed and the corresponding preset feedforward coefficient, compensation processing is performed. This aims to accurately calculate the desired output force at the end of the main arm using the smoothed and stabilized spring-damped gain, combined with the robot's current arm speed and the preset feedforward coefficient. This desired output force at the end of the main arm will be directly used in subsequent main arm drive stages, and is crucial for achieving stable and precise teleoperation.

[0056] Before calculating the desired output force at the end effector of the main arm, this application first filters the adaptively adjusted spring-damping gain. This filtering effectively smooths the gain curve, removes high-frequency noise and instantaneous fluctuations, resulting in a more stable and continuous filtered spring-damping gain. Subsequently, based on this filtered spring-damping gain, compensation is performed using the robot's current main arm speed and the corresponding preset feedforward coefficient to calculate the desired output force at the end effector. This method ensures that the gain parameters used for compensation are more reliable, avoiding output force instability caused by gain fluctuations and guaranteeing the accuracy and stability of the calculated desired output force. This processing mechanism, combined with basic teleoperation control methods, significantly improves the smoothness of the output force and the reliability of control while maintaining the adaptive capability of the entire force feedback system, thus providing a solid foundation for the precise actuation of the main arm.

[0057] By employing the aforementioned technical solution, the adaptively adjusted spring-damping gain is filtered, effectively smoothing the gain curve and eliminating potential noise and abrupt changes. This makes the spring-damping gain used in subsequent compensation processing more stable and continuous, significantly improving the accuracy and stability of the calculated desired output force at the boom end. During teleoperation, the operator receives a more stable and reliable force feedback signal, avoiding system oscillations or operational uncertainties caused by unstable force output, thereby improving the system's response accuracy and operational safety. Especially when the boom comes into contact with the environment or collides, stable force feedback helps the operator more accurately perceive the remote contact status, improving precision operation capabilities and operational immersion.

[0058] In some implementations, compensation is performed based on the filtered spring-damping gain, combined with the robot's current arm speed and the robot's corresponding preset feedforward coefficient, to obtain the desired output force at the end of the arm, including: Based on the filtered spring-damping gain, combined with the pose error and the preset spring coefficient, spring compensation processing is performed to obtain spring force information; Based on the filtered spring-damping gain, combined with the speed error and the preset damping coefficient, damping compensation is performed to obtain the damping force information; By combining the current speed of the robot's main arm with the robot's corresponding preset feedforward coefficient, feedforward compensation is performed to obtain feedforward force information; The desired output force at the end of the main boom is obtained by superimposing the spring force information, damping force information, and feedforward force information.

[0059] Spring compensation processing involves calculating a force component proportional to the pose error. This force component is determined by the filtered spring-damping gain, the pose error, and a preset spring constant. Its function is to provide a restoring force, causing the master arm to move towards the target pose, thus simulating the characteristics of a physical spring. For example, spring compensation processing can directly calculate spring force information by multiplying the filtered spring-damping gain, the pose error, and the preset spring constant. Spring compensation processing is a core component of the force feedback mechanism, ensuring that the master arm accurately reflects the positional deviation between the slave arm and the target pose, providing the operator with an intuitive sense of position.

[0060] Damping compensation refers to calculating a force component proportional to the velocity error, determined by the filtered spring-damping gain, the velocity error, and a preset damping coefficient. Its function is to dissipate system energy and suppress oscillations, thereby providing stability and a smoother force feedback experience. Damping compensation can be calculated by multiplying the filtered spring-damping gain, velocity error, and preset damping coefficient. Damping compensation is crucial in teleoperation control, effectively suppressing instability during system motion or contact, preventing boom oscillations, and thus improving operational smoothness and safety.

[0061] Feedforward compensation processing refers to calculating a force component based on the current speed of the boom and a predefined feedforward coefficient. Unlike feedback control, feedforward control pre-calculates the required force based on the desired motion, aiming to reduce system latency and improve responsiveness. Feedforward compensation processing obtains feedforward force information by directly multiplying the current boom speed by the preset feedforward coefficient. Feedforward compensation processing introduces active actuation capability into the system, pre-compensating for inherent system latency and inertia, resulting in a faster and more precise boom response, especially in fast-moving or dynamically changing environments.

[0062] Superposition processing refers to combining the individual force components (spring force information, damping force information, and feedforward force information) to generate the final desired output force at the boom end. For example, it can be done according to the formula for calculating the desired output force at the boom end: Based on spring force information Damping force information and feedforward force information By superimposing the values, the desired output force in each direction is obtained. Therefore, it is possible to determine the expected output force in each direction. By performing force synthesis, the final desired output force at the end of the main arm is obtained. ;in, for Expected output force in the direction, for Spring-damping gain after directional filtering for The direction corresponds to the preset spring constant. for The direction corresponds to the preset damping coefficient. for The direction corresponds to the preset feedforward coefficient. for The pose error corresponding to the direction for The velocity error corresponding to the direction, for The current velocity of the main boom corresponding to the direction. This superposition ensures that all aspects of the control strategy contribute to the overall force feedback. Superposition processing is usually achieved through vector addition, which linearly superimposes spring force information, damping force information, and feedforward force information in their respective dimensions. It integrates the force components from different compensation mechanisms, ensuring that the expected output force at the end of the main boom can comprehensively and accurately reflect the interaction state between the slave boom and the environment, as well as the operator's intention.

[0063] This application's solution refines the compensation process into three independent yet coordinated stages: spring compensation, damping compensation, and feedforward compensation, which are then superimposed to create a more precise and accurate force feedback signal generation process. This decomposed processing method allows the system to provide targeted compensation for different types of errors (pose error, velocity error) and dynamic characteristics (main arm speed), effectively solving the problem of inaccurate force feedback signals or unstable system responses caused by insufficiently detailed compensation processing in traditional methods. Through precise spring force, stable damping force, and rapidly responding feedforward force, this application's solution significantly improves the realism of force feedback and system stability, enabling operators to more accurately perceive the remote environment and improving the accuracy and safety of teleoperation.

[0064] Through the above technical solution, this application effectively solves the problem of inaccurate force feedback signals or unstable system response caused by insufficient compensation processing in traditional teleoperation control methods. Specifically, the compensation processing is decomposed into independent spring compensation, damping compensation, and feedforward compensation, enabling the system to calculate and apply force feedback more accurately. Spring compensation ensures that position errors are accurately converted into force feedback, enhancing the operator's perception of the slave arm's position. Damping compensation effectively suppresses system oscillations, improving operational smoothness and safety, while the introduction of filter gain avoids the impact of excessive damping on response sensitivity. Feedforward compensation introduces an active drive mechanism, pre-compensating for system delays based on the current speed of the master arm, significantly improving the master arm's response speed and operational immediacy, providing a smoother experience for operators performing rapid or dynamic operations. Finally, by superimposing these independent force components, this application ensures that the expected output force at the master arm's end effector comprehensively and realistically reflects the interaction between the slave arm and the environment, as well as the operator's operational intentions, thereby significantly improving the immersion, accuracy, and safety of teleoperation. This sophisticated compensation mechanism maintains system stability while greatly optimizing human-computer interaction performance, enabling operators to complete delicate tasks more accurately and efficiently.

[0065] In some implementations, determining the joint torque vector based on the expected output force at the end of the main arm and driving the main arm based on the joint torque vector includes: obtaining the current gravity compensation information of the main arm; performing gravity compensation processing on the main arm based on the expected output force at the end of the main arm and in combination with the gravity compensation information to obtain the main arm joint torque vector; and driving the main arm based on the main arm joint torque vector.

[0066] Gravity compensation information refers to the torque or force caused by gravity acting on the joints of the robot's main arm. Obtaining this information aims to accurately quantify the gravitational influence on the main arm in its current posture, providing accurate input for subsequent compensation processing. For example, torque sensors can be installed at each joint of the main arm to directly measure and acquire the current gravitational torque. The purpose of gravity compensation processing is to effectively separate the force the operator expects to apply (i.e., the expected output force at the end of the main arm) from the torque generated by the main arm's own gravity, ensuring that the force feedback felt by the operator is purely operational force, rather than a gravitational burden. This process involves mapping the expected output force at the end of the main arm to joint space through the transpose of the Jacobian matrix, obtaining the torque generated at the joint by the operational force, and then superimposing this torque with the pre-acquired gravity compensation information (i.e., the torque generated by gravity at the joint) to obtain the final torque vector to be applied to the main arm joints. For example, a preset joint torque vector calculation formula can be used. The desired output force at the end of the main arm is obtained by transposing the Jacobian matrix. Mapping to joint space yields the torque generated at the joint by the operating force. Subsequently, the torque generated at the joint can be based on the operating force. and gravity compensation information The moment vector of the main arm joint is obtained by superposition processing. ;in, Represents the moment vector of the main arm joint. Represents gravity compensation information. Represents the transpose of the Jacobian matrix of the main arm. This represents the desired output force at the end of the main arm in all directions (x, y, z, rx, ry, rz). Driving the main arm refers to converting the calculated joint torque vectors into actual physical motion, enabling the main arm to respond according to the operator's intentions. This is typically achieved through actuators (such as servo motors) within the main arm. Specifically, the calculated joint torque vectors are sent as commands to the controllers of each joint of the main arm. The controllers then adjust the motor current or voltage to generate corresponding torque, thereby driving the movement of each joint of the main arm and achieving precise control and force feedback.

[0067] This solution incorporates a gravity compensation element into the joint torque calculation, enabling the main arm's driving torque to accurately reflect the force the operator intends to apply, without requiring the operator to bear the additional torque generated by the main arm's own weight. In the entire teleoperation control method, the aforementioned steps have already generated the desired output force at the main arm's end effector through mechanisms such as adaptive adjustment of spring-damping gain, effectively suppressing system oscillations and improving the clarity of contact. Building upon this, this solution further ensures that this desired output force is transmitted to the operator realistically and naturally, avoiding interference from gravity on the force feedback signal. Specifically, after the system adaptively adjusts the spring-damping gain based on the slave arm's pose and velocity errors, and calculates the desired output force at the main arm's end effector by combining the current speed of the main arm and the feedforward coefficient, this solution calculates the final joint torque vector by combining this desired output force with the gravity compensation information of the main arm. This combination allows the main arm to automatically counteract its own gravity when responding to the operator's intentions, resulting in a purer and more realistic force feedback perceived by the operator, greatly enhancing the immersion and safety of the operation.

[0068] Through the above technical solution, this application effectively solves the problems of inaccurate joint torque calculation and the need for operators to overcome additional gravity burden caused by ignoring the influence of gravity in traditional teleoperation. By acquiring the current gravity compensation information of the main arm and incorporating it into the joint torque calculation, the driving torque of the main arm can accurately offset its own gravity, thereby ensuring that the force feedback felt by the operator is a pure operating force, rather than a superposition of gravity and operating force. This significantly reduces operator fatigue and improves the comfort and naturalness of operation. At the same time, because the realism of the force feedback signal is enhanced, the operator can more accurately perceive the contact status of the remote environment, thereby improving the safety, precision operation capability, and immersion of teleoperation. In the entire teleoperation control method, this solution, combined with the aforementioned spring-damping gain adaptive adjustment and compensation processing, jointly constructs a more complete and efficient force feedback mechanism, enabling the system to provide a more sensitive and realistic interactive experience while maintaining stability.

[0069] In some implementations, determining the target pose of the robot's slave arm based on the current pose of the robot's master arm end effector includes: acquiring the current pose of the robot's master arm end effector; and performing low-pass filtering on the current pose of the master arm end effector to obtain the target pose of the slave arm.

[0070] The method first acquires the current pose of the robot's end effector. This step aims to obtain the position and orientation information of the end effector in the workspace in real time as the operator manipulates it. This information is the fundamental input for teleoperation control and directly reflects the operator's intent. For example, the pose data of the end effector in the world coordinate system can be calculated and acquired in real time using sensors such as encoders and inertial measurement units (IMUs) integrated inside the end effector, combined with the robot's kinematic model. Alternatively, the pose information of the end effector or specific markers on it can be directly measured using non-contact sensors such as external optical tracking systems and magnetic tracking systems.

[0071] Subsequently, the current pose of the end effector is low-pass filtered. This step is used to eliminate high-frequency noise and unintentional operator jitter that may exist in the current pose data of the end effector, in order to obtain a smooth and stable pose signal. The function of the low-pass filter is to allow signals below a preset cutoff frequency to pass through, while attenuating signals above the preset cutoff frequency, thereby effectively removing high-frequency interference.

[0072] Finally, the target pose of the slave arm is obtained. The end effector pose of the master arm, after low-pass filtering, is the target pose of the slave arm. This target pose has been smoothed to more accurately and stably reflect the operator's true intention, providing a reliable reference for the subsequent following control of the slave arm.

[0073] This application's solution addresses the issue of target pose instability caused by noise or jitter in the master arm's end effector pose by introducing a low-pass filtering mechanism into the teleoperation control method, thereby improving system stability, operational accuracy, and the reliability of force feedback. Specifically, at the beginning of the teleoperation control phase, the current pose of the robot's master arm's end effector is first acquired. This step ensures that the system can perceive the operator's input in real time. Given that the current pose of the master arm's end effector may be affected by factors such as operator hand tremors or sensor noise, directly using it as the target pose of the slave arm may lead to unstable slave arm movement or even system oscillation. Therefore, this application further performs low-pass filtering on the acquired current pose of the master arm's end effector. This filtering effectively removes high-frequency interference components from the pose data, preserving the smooth, low-frequency motion trajectory of the operator's intention, thus obtaining a stable and accurate target pose of the slave arm. This filtered target pose of the slave arm, used as the input for subsequent slave arm following control, can significantly reduce slave arm movement jitter and improve the smoothness of the slave arm's response to the operator's intention. Based on this, the slave arm follows the target pose of the slave arm and returns the current slave arm pose in real time. Subsequently, the system performs error processing based on the current slave arm pose and the target slave arm pose to obtain the control error. Because the stability of the target slave arm pose is improved, the calculated control error is more accurate and stable, avoiding false error signals caused by target pose fluctuations. This provides a reliable input for subsequent adaptive adjustment of the spring-damping gain, enabling the gain adjustment to more accurately reflect the actual contact state and operational requirements, thereby optimizing the calculation of the desired output force at the end of the master arm. Ultimately, the master arm is driven through the joint torque vector, providing the operator with more realistic and stable force feedback. Overall, this scheme, by introducing low-pass filtering in the pose determination stage, ensures the stability of the teleoperation system from the source, providing a solid foundation for subsequent force feedback control, making the entire teleoperation process safer, more precise, and more natural.

[0074] like Figure 3 As shown, this application provides a remote operation control system for a robot, comprising: The determination module 21 is used to determine the target pose of the robot's slave arm based on the current pose of the robot's main arm end effector. Control module 22 is used to perform slave arm following control based on the target pose of the slave arm to obtain the current pose of the slave arm; Processing module 23 is used to perform error processing based on the current slave arm pose and the slave arm target pose to obtain the control error; The adjustment module 24 is used to adaptively adjust the spring-damping gain based on the control error and a preset error threshold to obtain the spring-damping gain of the robot. The compensation module 25 is used to perform compensation processing based on the spring-damping gain, combined with the current speed of the robot's main arm and the corresponding preset feedforward coefficient of the robot, to obtain the desired output force at the end of the main arm; The drive module 26 is used to determine the joint torque vector based on the expected output force at the end of the main arm, and drive the main arm based on the joint torque vector.

[0075] This embodiment achieves a balance between the stability and interactive performance of the teleoperation system by adaptively adjusting the spring-damping gain by combining the control error with a preset error threshold. This dynamically enhances the clarity of force feedback when the slave arm is in contact with the environment, while maintaining system stability during non-contact or violent movements. Specifically, through the above technical solution, this application achieves dynamic optimization of the spring-damping gain during teleoperation. Since the adjustment module 24 adaptively adjusts the spring-damping gain based on the comparison between the control error and the preset error threshold, when the slave arm is in contact with the environment, the pose error increases while the velocity error is small, and the spring-damping gain is increased to enhance the clarity of force feedback, enabling the operator to accurately perceive the contact state. When the slave arm is not in contact with the environment or undergoes rapid movement, the velocity error is large or the pose error is small, and the spring-damping gain is maintained at the default value to avoid system oscillation. In addition, the compensation module 25 introduces a feedforward coefficient for compensation processing, further ensuring the accuracy of the force feedback signal.

[0076] In summary, this application effectively resolves the contradiction between stability and interactive performance through an adaptive mechanism of spring-damping gain, thereby improving operational safety and precision operation capabilities while ensuring system stability.

[0077] The teleoperation control system for a robot provided in this application embodiment is used to execute the steps in the teleoperation control method for a robot provided in the first aspect above. The principle of the teleoperation control system for a robot provided in this embodiment is the same as that of the teleoperation control method for a robot provided in the first aspect above, and will not be discussed in detail here.

[0078] like Figure 4 As shown, this application provides a robot device, including a processor 111, a communication interface 112, a memory 113 and a communication bus 114, wherein the processor 111, the communication interface 112 and the memory 113 communicate with each other through the communication bus 114, and the memory 113 is used to store computer programs. In one embodiment of this application, the processor 111, when executing a program stored in a memory, implements the steps of the teleoperation control method for the robot as described in any of the first aspects.

[0079] This application provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it performs the method in any optional implementation of the above embodiments to achieve the following functions: determining the target pose of the robot's slave arm based on the current pose of the robot's master arm end effector; performing slave arm following control based on the slave arm target pose to obtain the current slave arm pose; performing error processing based on the current slave arm pose and the slave arm target pose to obtain a control error; performing adaptive adjustment of the spring-damping gain based on the control error and a preset error threshold to obtain the robot's spring-damping gain; performing compensation processing based on the spring-damping gain, combined with the robot's master arm current speed and the robot's corresponding preset feedforward coefficient to obtain the desired output force of the master arm end effector; determining the joint torque vector based on the desired output force of the master arm end effector, and driving the master arm based on the joint torque vector. The computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0080] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and there may be other division methods in actual implementation. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interface; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0081] Furthermore, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0082] Furthermore, the functional modules in the various embodiments of this 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.

[0083] In this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations.

[0084] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for remote control of a robot, characterized in that, include: Based on the current pose of the robot's main arm end effector, determine the target pose of the robot's slave arm; Based on the target pose of the slave arm, slave arm following control is performed to obtain the current pose of the slave arm; Error processing is performed based on the current slave arm pose and the target slave arm pose to obtain the control error; Based on the control error, the spring-damping gain is adaptively adjusted in combination with a preset error threshold to obtain the spring-damping gain of the robot. Based on the spring-damping gain, and combined with the current speed of the robot's main arm and the corresponding preset feedforward coefficient of the robot, compensation processing is performed to obtain the desired output force at the end of the main arm; The joint torque vector is determined based on the expected output force at the end of the main arm, and the main arm is driven based on the joint torque vector. The control error includes pose error and velocity error, and the preset error threshold includes a velocity error threshold and a pose error threshold. The adaptive adjustment of the spring-damping gain of the robot based on the control error and the preset error threshold to obtain the spring-damping gain includes: If the speed error is not greater than the speed error threshold and the pose error is not less than the pose error threshold, obtain the preset spring-damping gain base of the robot. Based on the spring-damping gain base, and combined with the difference between the pose error and the pose error threshold, the spring-damping gain is adaptively adjusted to obtain the spring-damping gain of the robot.

2. The remote operation control method for a robot according to claim 1, characterized in that, The step of adaptively adjusting the spring-damping gain of the robot based on the control error and a preset error threshold to obtain the spring-damping gain further includes: When the pose error is less than the pose error threshold, or when the velocity error is greater than the velocity error threshold, the preset default gain is determined as the spring-damping gain of the robot.

3. The teleoperation control method for a robot according to claim 1, characterized in that, The step of performing error processing based on the current slave arm pose and the target slave arm pose to obtain the control error includes: The pose error is obtained by calculating the current slave arm pose and the target slave arm pose. Based on the pose error, differential processing is performed to obtain the velocity error; The pose error and the velocity error are used as the control error.

4. The teleoperation control method for a robot according to claim 3, characterized in that, The compensation process, based on the spring-damping gain and combined with the current speed of the robot's main arm and the corresponding preset feedforward coefficient, yields the desired output force at the end of the main arm, including: The spring-damping gain is filtered to obtain the filtered spring-damping gain; Based on the filtered spring-damping gain, and combined with the current speed of the robot's main arm and the corresponding preset feedforward coefficient of the robot, compensation processing is performed to obtain the desired output force at the end of the main arm.

5. The teleoperation control method for a robot according to claim 4, characterized in that, The process involves using the filtered spring-damping gain, combined with the current speed of the robot's main arm and the corresponding preset feedforward coefficient, to perform compensation processing and obtain the desired output force at the end of the main arm, including: Based on the filtered spring-damping gain, combined with the pose error and the preset spring coefficient, spring compensation processing is performed to obtain spring force information; Based on the filtered spring-damping gain, and combined with the speed error and the preset damping coefficient, damping compensation processing is performed to obtain damping force information; By combining the current speed of the robot's main arm with the robot's corresponding preset feedforward coefficient, feedforward compensation processing is performed to obtain feedforward force information; The desired output force at the end of the main boom is obtained by superimposing the spring force information, the damping force information, and the feedforward force information.

6. The teleoperation control method for a robot according to claim 1, characterized in that, The step of determining the joint torque vector based on the desired output force at the end of the main arm, and driving the main arm based on the joint torque vector, includes: Obtain the current gravity compensation information of the main arm; Based on the expected output force at the end of the main arm, and combined with the gravity compensation information, gravity compensation processing is performed on the main arm to obtain the main arm joint torque vector; The main arm is driven according to the main arm joint torque vector.

7. The teleoperation control method for a robot according to claim 1, characterized in that, Determining the target pose of the robot's slave arm based on the current pose of the robot's master arm end effector includes: Obtain the current pose of the robot's end effector; The current pose of the main arm end is low-pass filtered to obtain the target pose of the slave arm.

8. A remote operation control system for a robot, characterized in that, include: The determination module is used to determine the target pose of the robot's slave arm based on the current pose of the robot's main arm end effector. The control module is used to perform slave arm following control based on the target pose of the slave arm to obtain the current pose of the slave arm; The processing module is used to perform error processing based on the current slave arm pose and the slave arm target pose to obtain the control error; The adjustment module is used to adaptively adjust the spring-damping gain based on the control error and a preset error threshold to obtain the spring-damping gain of the robot. The compensation module is used to perform compensation processing based on the spring-damping gain, combined with the current speed of the robot's main arm and the corresponding preset feedforward coefficient of the robot, to obtain the desired output force at the end of the main arm; A drive module is used to determine the joint torque vector based on the desired output force at the end of the main arm, and to drive the main arm based on the joint torque vector; The control error includes pose error and velocity error, and the preset error threshold includes a velocity error threshold and a pose error threshold. The adaptive adjustment of the spring-damping gain of the robot based on the control error and the preset error threshold to obtain the spring-damping gain includes: If the speed error is not greater than the speed error threshold and the pose error is not less than the pose error threshold, obtain the preset spring-damping gain base of the robot. Based on the spring-damping gain base, and combined with the difference between the pose error and the pose error threshold, the spring-damping gain is adaptively adjusted to obtain the spring-damping gain of the robot.

9. A robotic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in a memory, implements the teleoperation control method for the robot according to any one of claims 1-7.

Images