Method for robot friction-induced vibration suppression and industrial robot

Abstract

The application relates to a robot friction-induced vibration suppression method and an industrial robot, and belongs to the technical field of industrial robots. The robot friction-induced vibration suppression method comprises the following steps: S1. designing a high-frequency sinusoidal speed signal, generating a high-frequency sinusoidal speed signal with a variable constant a in amplitude and a variable constant b in frequency; S2. combining the high-frequency sinusoidal speed signal obtained in step S1 with a desired speed of a robot joint to obtain a corrected speed instruction; and S3. issuing and executing the speed instruction: the corrected speed instruction synthesized in step S2 is issued to a servo driver of the robot through an EtherCAT bus at a preset period through an upper computer, the speed instruction is executed by an internal speed loop of the servo driver, and the robot joint is driven to move. The application offsets friction disturbance through an adjustable high-frequency signal, guarantees the real-time performance of the instruction through EtherCAT, dynamically starts and stops to adapt to working conditions, and improves the operation stability of the robot, and has the advantages of simple debugging, low cost and strong robustness.

Description

Technical Field

[0001] This invention relates to the field of industrial robot technology, and in particular to a method for suppressing friction-induced vibration in robots and an industrial robot. Background Technology

[0002] Low-speed, high-precision motion is one of the core requirements of high-end equipment manufacturing. Applications such as semiconductor packaging, laser precision cutting, and orthopedic surgical robots generally require robots to achieve micron-level trajectory accuracy at joint speeds of 0.1° / s or even slower. However, when joints approach zero speed, the switching between static and dynamic friction causes stick-slip crawling, which becomes a major source of error in high-precision motion. This phenomenon is called friction-induced vibration, also known as stick-slip vibration. Stick-slip vibration has two motion states: one is stick-like motion when the relative velocity between two objects in frictional contact is zero, and the other is sliding motion when there is relative motion between them. At ultra-low speeds, due to the influence of friction-induced vibration, the robot constantly switches between these two motion states, resulting in significant vibration during robot operation and an inability to maintain stable low-speed operation.

[0003] Traditional solutions to this phenomenon rely on lubrication or high-cost mechanical structures. However, these mechanical structures are prone to structural resonance when flexible joints are present, further amplifying trajectory errors and increasing both cost and system reliability. Algorithmic control using a Coulomb-viscosity model is another approach. However, this model cannot describe the bristle hysteresis and Stribeck effect, has large compensation residuals near zero velocity, and requires manual readjustment of compensation parameters due to temperature, wear, or load changes, resulting in high on-site maintenance workload. Summary of the Invention

[0004] Therefore, it is necessary to provide a method for suppressing robot friction-induced vibration to address the problem of unstable operation at low speed and high precision caused by robot friction-induced vibration.

[0005] A method for suppressing friction-induced vibration in robots, characterized in that the method includes the following steps:

[0006] S1. Design a high-frequency sinusoidal velocity signal to generate a high-frequency sinusoidal velocity signal with a variable amplitude constant a and a variable frequency constant b;

[0007] S2. The high-frequency sinusoidal velocity signal obtained in step S1 is combined with the desired velocity of the robot joint to obtain the corrected velocity command;

[0008] S3. Issuing and executing speed commands: The modified speed command synthesized in step S2 is sent from the host computer to the robot's servo driver via the EtherCAT bus at a preset cycle. The speed command is executed by the speed loop inside the servo driver to drive the robot's joint movement.

[0009] In one embodiment, the expression for the high-frequency sinusoidal velocity signal is as follows:

[0010] ,in, The high-frequency disturbances are superimposed. denoted as sampling time, k as discrete sampling number, representing the kth control cycle, a as amplitude, and b as frequency.

[0011] In one embodiment, the expression for the speed command is as follows:

[0012] ,in, The corrected speed command. The desired speed.

[0013] In one embodiment, the frequency of the high-frequency sinusoidal velocity signal in step S1 is 200Hz and the amplitude is 0.003 rad / s.

[0014] In one embodiment, the specific steps for verifying the high-frequency sinusoidal velocity signal in step S1 are as follows:

[0015] S11. Design a point-to-point linear motion simulation;

[0016] S12. Build a LuGre friction model, observe the source of friction in the system and calculate the total friction in the system;

[0017] S13. By comparing the changes in torque and displacement over time before and after the injection of a high-frequency sinusoidal velocity signal, the effectiveness of the high-frequency positive linear velocity signal in reducing the friction of the system is verified.

[0018] In one embodiment, step S11, specifically designing a point-to-point straight-line running simulation step, is as follows:

[0019] S111. Given the target position of the object;

[0020] S112. Calculate the position error between the current position and the given position;

[0021] S113. Input the position error into the PID controller, and the PID controller outputs the control torque to make the object accelerate or decelerate.

[0022] S114. Determine if the object has reached the target position. If not, return to step S112 and continue the loop; if it has reached the target position, the motion is complete.

[0023] In one embodiment, the specific steps for building the LuGre friction model in step S12 are as follows:

[0024] S121. Determine the Stribeck function:

[0025] ,in, For Coulomb friction, For static friction, The corrected speed command. Stribeck speed;

[0026] S122. Calculate the mane deformation:

[0027] ,in, For the first The mane-shaped variable of each control cycle, Sampling time, This is the stiffness coefficient;

[0028] S123. Calculate the total frictional force:

[0029] ,in, The damping coefficient is... The coefficient of viscous friction is... Represents the elastic friction component. Indicates the damping friction component. It represents the viscous friction component.

[0030] In one embodiment, the specific steps of step S3 are as follows:

[0031] S31. The host computer monitors the absolute speed of the robot's motor shafts in real time during each control cycle;

[0032] S32. When the absolute speed is detected to be lower than a preset speed threshold, the generation and injection of a high-frequency sinusoidal speed signal are initiated;

[0033] S33. When the absolute speed is detected to be higher than the preset speed threshold, immediately cut off the generation and injection of the high-frequency sinusoidal speed signal, and keep the total friction force at the current moment at the last effective value;

[0034] S34. When the absolute speed is detected to be in the preset speed threshold transition zone, a 50ms hysteresis timer is used to prevent threshold jitter.

[0035] In one embodiment, the host computer is equipped with a parameter adjustment module, which allows for one-click adjustment of the preset speed threshold, the amplitude and frequency of the high-frequency sinusoidal speed signal.

[0036] An industrial robot includes: a body, joints, a servo driver, a host computer, and a robot friction-induced vibration suppression method algorithm program stored on the host computer. When the host computer executes the algorithm program, it implements the steps of the methods described in the above embodiments.

[0037] The beneficial effects of this invention are as follows: The high-frequency sinusoidal speed signal is essentially a micro-perturbation compensation signal. Its high-frequency characteristics can overcome the friction dead zone. The corrected speed command superimposed with the high-frequency sinusoidal speed signal can actively offset the speed fluctuations caused by frictional disturbances, preventing joints from experiencing stick-slip effects due to nonlinear friction, thus avoiding vibrations during movement. Simultaneously, the variable amplitude and frequency of the high-frequency sinusoidal speed signal can adapt to the frictional characteristics of different joints. It can be specifically adjusted according to the robot joint's load characteristics, motion conditions, and friction parameters, without requiring refactoring the control algorithm, and is suitable for multiple robot models and scenarios. The EtherCAT bus possesses microsecond-level synchronization and low-latency characteristics, ensuring that the corrected speed command is transmitted to the servo driver in real time and accurately, avoiding compensation lag caused by command transmission delays, and further reducing trajectory tracking and positioning errors. From the function expression of the total frictional force, it can be seen that the total frictional force of the system consists of three parts: elastic friction component, damped friction component, and viscous friction component. When step S3 is executed, Much larger , Approaching 0, that is This causes friction. Significantly reduced. With The decrease of the mane shape variable can be seen from the expression of the mane shape variable function. The elastic friction component is significantly reduced, and consequently, the elastic friction component is also significantly reduced. This is due to the injection of a high-frequency sinusoidal velocity signal. Dominated by high-frequency components, the integral over a complete cycle is 0, meaning the damped friction component is 0. Therefore, the total friction force... Total friction within the macroeconomic cycle Averaged This drives the joint to continuously cross the static-dynamic friction boundary at the microscopic scale, transforming the originally intense nonlinear friction into a predictable, feedforward-adaptive average disturbance. Simultaneously, during low-speed motion, due to... The value is very small. The friction force becomes very small, thus greatly reducing the performance requirements of the controller. This allows for easy overcoming of friction and prevention of vibration, enabling the robot to maintain stable operation. However, the risk of joint friction-induced vibration is not uniformly distributed. During startup, positioning, and low-speed crawling, the nonlinear switching between static and dynamic friction is more frequent, resulting in more significant stick-slip vibrations—critical scenarios for friction suppression. At high speeds, friction is primarily stable dynamic friction, with low vibration risk; injecting high-frequency signals in these situations may become unnecessary disturbances. By monitoring the absolute speed threshold through a host computer, dynamic start-stop of the high-frequency sinusoidal speed signal is achieved, ensuring friction averaging while eliminating energy waste and noise pollution at high speeds. Furthermore, during robot movement, the speed may oscillate slightly around the threshold due to load fluctuations and sensor noise. Directly triggering the signal start-stop would cause frequent switching of the high-frequency signal, leading to new vibration interference. Step S34 addresses this issue using a 50ms hysteresis timer. When the speed enters the threshold transition zone, the timer starts. Signal interruption or reactivation is only performed if the speed remains consistently above or below the threshold for more than 50ms. This enhances the anti-interference capability of the control logic and ensures the stability of the robot's motion system near the threshold. The entire process of suppressing friction-induced vibration requires no complex model calculations or algorithm optimizations, no on-site parameter calibration, and no need for a special controller. This reduces the operational difficulty for engineers and the system debugging cycle, while also reducing the computational power required by the control algorithm for the robot's main control unit. This reserves resources for other precision control functions, such as force control and vision guidance, offering advantages of low cost, easy deployment, and high robustness. Attached Figure Description

[0038] Figure 1 Flowchart of a method for suppressing friction-induced vibration in robots;

[0039] Figure 2 A flowchart for verifying high-frequency sinusoidal velocity signals;

[0040] Figure 3 Flowchart for linear motion simulation design;

[0041] Figure 4 The graph shows the velocity-time and friction-time relationships in the LuGre friction model.

[0042] Figure 5 A comparison of torque changes over time before and after injecting a high-frequency sinusoidal velocity signal;

[0043] Figure 6 A comparison of displacement versus time before and after the injection of a high-frequency sinusoidal velocity signal;

[0044] Figure 7 This is the control logic diagram for dynamic start-stop of a high-frequency sinusoidal speed signal. Detailed Implementation

[0045] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0046] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0047] The following describes a method for suppressing robot friction-induced vibration according to some embodiments of the present invention, with reference to the accompanying drawings.

[0048] like Figures 1 to 7 As shown, a method for suppressing friction-induced vibration in a robot is provided, comprising:

[0049] S1. Design a high-frequency sinusoidal velocity signal to generate a high-frequency sinusoidal velocity signal with a variable amplitude constant a and a variable frequency constant b;

[0050] S2. The high-frequency sinusoidal velocity signal obtained in step S1 is combined with the desired velocity of the robot joint to obtain the corrected velocity command;

[0051] S3. Issuing and executing speed commands: The modified speed command synthesized in step S2 is sent from the host computer to the robot's servo driver via the EtherCAT bus at a preset cycle. The speed command is executed by the speed loop inside the servo driver to drive the robot's joint movement.

[0052] In addition to the features of the above embodiments, this embodiment further specifies that the frequency of the high-frequency sinusoidal velocity signal in step S1 is 200Hz and the amplitude is 0.003 rad / s.

[0053] In addition to the features of the above embodiments, this embodiment further defines the expression for the high-frequency sinusoidal velocity signal as follows:

[0054]

[0055] Specifically, in step S1, parameters are set with one click via the host computer parameter adjustment module. This embodiment takes a six-degree-of-freedom serial robot with rigid links and flexible joints as an example, taking amplitude a = 0.003 rad / s and frequency b = 200 Hz. The expression for the obtained high-frequency sinusoidal velocity signal is:

[0056]

[0057] In addition to the features of the above embodiments, this embodiment further defines the expression of the speed command as follows:

[0058]

[0059] like Figure 2 , Figure 5 and Figure 6 As shown, in addition to the features of the above embodiments, this embodiment further defines the specific steps for verifying the high-frequency sinusoidal velocity signal in step S1 as follows:

[0060] S11. Design a point-to-point linear motion simulation;

[0061] S12. Build a LuGre friction model, observe the source of friction in the system and calculate the total friction in the system;

[0062] S13. By comparing the changes in torque and displacement over time before and after the injection of a high-frequency sinusoidal velocity signal, the effectiveness of the high-frequency positive linear velocity signal in reducing the friction of the system is verified.

[0063] like Figure 3 As shown, in addition to the features of the above embodiments, this embodiment further specifies that in step S11, the step of designing a point-to-point straight-line running simulation is as follows:

[0064] S111. Given the target position of the object;

[0065] S112. Calculate the position error between the current position and the given position;

[0066] S113. Input the position error into the PID controller, and the PID controller outputs the control torque to make the object accelerate or decelerate.

[0067] S114. Determine if the object has reached the target position. If not, return to step S112 and continue the loop; if it has reached the target position, the motion is complete.

[0068] Specifically, a linear motion simulation is designed to describe the motion of the object. Before the high-frequency positive linear velocity signal is injected, due to the presence of friction, the system will form a limit cycle near the target position, making it impossible for the object to stop accurately at the target position. The PID controller continuously outputs control torque, and the object swings back and forth around 1m.

[0069] like Figures 4 to 6 As shown, in addition to the features of the above embodiments, this embodiment further defines the specific steps of building the LuGre friction model in step S12 as follows:

[0070] S121. Determine the Stribeck function:

[0071] ,in, For Coulomb friction, For static friction, The corrected speed command. Stribeck speed;

[0072] S122. Calculate the mane deformation:

[0073] ,in, For the first The mane-shaped variable of each control cycle, Sampling time, This is the stiffness coefficient;

[0074] S123. Calculate the total frictional force:

[0075] ,in, The damping coefficient is... The coefficient of viscous friction is... Represents the elastic friction component. Indicates the damping friction component. It represents the viscous friction component.

[0076] Specifically, a LuGre friction model was built to observe the source of friction in the system and calculate the total friction. The velocity-time curves and friction-time curves obtained from the LuGre friction model matched the actual feedback phenomena, successfully simulating friction in real-world motion scenarios. Based on this, a high-frequency sinusoidal velocity signal was injected. By comparing the changes in torque-time and object displacement-time before and after the injection of the high-frequency sinusoidal velocity signal, the effectiveness and feasibility of the high-frequency sinusoidal velocity signal were verified.

[0077] According to the LuGre friction model, the total frictional force of the system is... It consists of three parts: elastic friction, damped friction, and viscous friction. When the host computer injects a high-frequency sinusoidal velocity and combines it with the desired velocity, a speed command is obtained. , Much larger , Approaching 0, that is This causes friction. Significantly reduced. With The decrease of the mane shape variable can be seen from the expression of the mane shape variable function. The elastic friction component is significantly reduced, and consequently, the elastic friction component is also significantly reduced. This is due to the injection of a high-frequency sinusoidal velocity signal. Dominated by high-frequency components, the integral over a complete cycle is 0, meaning the damped friction component is 0. Therefore, the total friction force... Total friction within the macroeconomic cycle Averaged This drives the joint to continuously cross the static-dynamic friction boundary at the microscopic scale, transforming the originally intense nonlinear friction into a predictable, feedforward-adaptive average disturbance. Simultaneously, during low-speed motion, due to... The value is very small. The friction force is thus significantly reduced, greatly decreasing the performance requirements of the controller and easily overcoming it. Therefore, in the linear motion simulation system, after the injection of the high-frequency sinusoidal velocity signal, the object can eventually reach and stop at the target position. By comparing the torque-time and object displacement-time graphs, it can be seen that after signal injection, the PID controller's torque output is 0, while the object displacement remains unchanged. This successfully verifies the effectiveness and feasibility of the high-frequency velocity signal in reducing friction. Using the verified high-frequency sinusoidal velocity signal, a synthesized speed command is sent to the servo driver at a certain period via the EtherCAT bus. The servo driver's internal speed loop executes this speed command, driving the robot's joint movement.

[0078] In such Figure 7 As shown, in addition to the features of the above embodiments, this embodiment further defines the specific steps of step S3 as follows:

[0079] S31. The host computer monitors the absolute speed of the robot's motor shafts in real time during each control cycle;

[0080] S32. When the absolute speed is detected to be lower than a preset speed threshold, the generation and injection of a high-frequency sinusoidal speed signal are initiated;

[0081] S33. When the absolute speed is detected to be higher than the preset speed threshold, immediately cut off the generation and injection of the high-frequency sinusoidal speed signal, and keep the total friction force at the current moment at the last effective value;

[0082] S34. When the absolute speed is detected to be in the preset speed threshold transition zone, a 50ms hysteresis timer is used to prevent threshold jitter.

[0083] Specifically, a speed threshold is set via a host computer parameter adjustment module. The host computer monitors the motor shaft speed in real time at a 1kHz cycle. When the speed falls into the preset low-speed range, the host computer immediately injects a fixed sinusoidal speed component of 200 Hz and 0.003 rad / s and synthesizes a speed command, which is sent to the robot's servo driver via the EtherCAT bus. This drives the joint to continuously cross the static-dynamic friction boundary at a microscale, transforming the originally intense nonlinear friction into a predictable and feedforward average disturbance. When the speed exceeds the preset speed threshold range, the high-frequency signal immediately becomes silent, maintaining the last effective value of the total friction force of the system. This ensures both friction averaging and eliminates energy waste and noise pollution at high speeds. When the speed enters the threshold transition zone, a timer starts. The signal is only cut off or turned on after the speed has been continuously higher or lower than the threshold for more than 50ms, improving the anti-interference capability of the control logic and ensuring the stability of the robot's motion system near the threshold.

[0084] In addition to the features of the above embodiments, this embodiment further specifies that the host computer is equipped with a parameter adjustment module, through which the preset speed threshold, the amplitude and frequency of the high-frequency sinusoidal speed signal can be adjusted with one click.

[0085] Specifically, all algorithm programs run on the host computer, and the parameters do not need to be calibrated on-site or compared with a special controller. The parameters can be flexibly optimized for different robot joints, different loads, and different wear conditions without the need to reconstruct the control algorithm, making it suitable for multiple robot models and complex industrial scenarios.

[0086] An industrial robot includes: a body, joints, a servo driver, a host computer, and a robot friction-induced vibration suppression method algorithm program stored on the host computer. When the host computer executes the algorithm program, it implements the steps of the methods described in the above embodiments.

[0087] 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 specification.

[0088] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for suppressing friction-induced vibration in robots, characterized in that, The method includes the following steps: S1. Design a high-frequency sinusoidal velocity signal to generate a high-frequency sinusoidal velocity signal with a variable amplitude constant a and a variable frequency constant b; S2. The high-frequency sinusoidal velocity signal obtained in step S1 is combined with the desired velocity of the robot joint to obtain the corrected velocity command; S3. Issuing and executing speed commands: The modified speed command synthesized in step S2 is sent from the host computer to the robot's servo driver via the EtherCAT bus at a preset cycle. The speed command is executed by the speed loop inside the servo driver to drive the robot's joint movement. The specific steps of step S3 are as follows: S31. The host computer monitors the absolute speed of the robot's motor shafts in real time during each control cycle; S32. When the absolute speed is detected to be lower than a preset speed threshold, the generation and injection of a high-frequency sinusoidal speed signal are initiated; S33. When the absolute speed is detected to be higher than the preset speed threshold, immediately cut off the generation and injection of the high-frequency sinusoidal speed signal, and keep the total friction force at the current moment at the last effective value; S34. When the absolute speed is detected to be in the preset speed threshold transition zone, a 50ms hysteresis timer is used to prevent threshold jitter. The host computer is equipped with a parameter adjustment module, which allows for one-click adjustment of the preset speed threshold, the amplitude and frequency of the high-frequency sinusoidal speed signal. The host computer sets the speed threshold through the parameter adjustment module. The host computer monitors the motor shaft speed in real time at a 1kHz cycle. When the speed falls into the preset low-speed range, the host computer immediately injects a fixed sinusoidal speed component of 200 Hz and 0.003 rad / s and synthesizes a speed command. This command is then sent to the robot's servo driver via the EtherCAT bus, driving the joint to continuously traverse the static range at a microscopic scale. The dynamic friction boundary transforms the originally intense nonlinear friction into a predictable, feedforward-capable average disturbance.

2. The method for suppressing robot friction-induced vibration according to claim 1, characterized in that, The expression for the high-frequency sinusoidal velocity signal is as follows: ； in, The high-frequency disturbances are superimposed. denoted as sampling time, k as discrete sampling number, representing the kth control cycle, a as amplitude, and b as frequency.

3. The method for suppressing robot friction-induced vibration according to claim 2, characterized in that, The expression for the speed command is as follows: ； in, The corrected speed command. The desired speed.

4. The method for suppressing robot friction-induced vibration according to claim 1, characterized in that, The frequency of the high-frequency sinusoidal velocity signal in step S1 is 200Hz, and the amplitude is 0.003 rad / s.

5. The method for suppressing robot friction-induced vibration according to claim 1, characterized in that, In step S1, the specific steps for verifying the high-frequency sinusoidal velocity signal are as follows: S11. Design a point-to-point linear motion simulation; S12. Build a LuGre friction model, observe the source of friction in the system and calculate the total friction in the system; S13. By comparing the changes in torque and displacement over time before and after the injection of a high-frequency sinusoidal velocity signal, the effectiveness of the high-frequency sinusoidal velocity signal in reducing the friction of the system is verified.

6. The method for suppressing robot friction-induced vibration according to claim 5, characterized in that, In step S11, the specific steps for designing a point-to-point straight-line running simulation are as follows: S111. Given the target position of the object; S112. Calculate the position error between the current position and the given position; S113. Input the position error into the PID controller, and the PID controller outputs the control torque to make the object accelerate or decelerate. S114. Determine if the object has reached the target position. If not, return to step S112 and continue the loop; if it has reached the target position, the motion is complete.

7. The method for suppressing robot friction-induced vibration according to claim 5, characterized in that, In step S12, the specific steps for building the LuGre friction model are as follows: S121. Determine the Stribeck function: ； in, For Coulomb friction, For static friction, The corrected speed command. Stribeck speed; S122. Calculate the mane deformation: ； in, For the first The mane-shaped variable of each control cycle, Sampling time, This is the stiffness coefficient; S123. Calculate the total frictional force: ； in, The damping coefficient is... The coefficient of viscous friction is... Represents the elastic friction component. Indicates the damping friction component. It represents the viscous friction component.

8. An industrial robot, comprising: The robot body, joints, servo drivers, host computer, and robot friction-induced vibration suppression algorithm program stored on the host computer are characterized in that the host computer implements the steps of the method described in any one of claims 1 to 7 when executing the algorithm program.

Images