KR102607486B1 - Method and Apparatus for Reducing Vibration Noise in a Robot Arm - Google Patents

Abstract

This relates to a vibration noise removal method and device for a robot arm that removes vibration noise by reflecting the effects of vibration noise generated on each axis. It receives information on the structure and kinematic characteristics of the robot arm and formulates the kinematic characteristics of the robot arm. A modeling step, performing a kinematic calculation for at least one axis of the robot arm using the modeling result from the modeling step, determining vibration noise for at least one axis of the robot arm, performed Based on the kinematic calculation for at least one axis and the identified vibration noise, a step of correcting the position control data for each axis by reflecting the vibration noise generated on the axis and the vibration noise generated on the adjacent axis for each axis. By removing the vibration noise generated on the axis of the robot arm by the vibration noise removal method of the robot arm, it can provide accurate and safe operation of the robot arm by reflecting the effect of vibration noise generated on other nearby axes. The effect is derived.

Description

Method and Apparatus for Reducing Vibration Noise in a Robot Arm}

This relates to a method and device for removing vibration noise from a robot arm, and relates to a method and device for removing vibration noise from a robot arm that removes vibration noise by reflecting the effects of vibration noise generated for each axis.

Generally, a robot arm is equipped with a motor for each axis, and the movement of the robot arm can be manipulated by controlling these motors. And the movement of each axis can be detected using a sensor such as an encoder.

Additionally, robotic arms can be equipped with a variety of sensors, such as position, velocity, and acceleration sensors, and using these sensor data, the robotic arm can perform accurate and repeatable tasks. The movement of each axis of the robot arm can be identified, and various tasks can be performed based on this information. Each axis of the robot arm is controlled independently of each other, and the movement of each axis can affect each other.

Meanwhile, it is desirable to eliminate the vibration noise of the robot arm as it can interfere with work performance. If the vibration noise is large, the robot arm may not move to the target position accurately or the work speed may slow down. Removing vibration noise plays an important role in improving the work performance of the robot arm.

KR 10-2022-0058674 A KR 10-2001-0081236 A

The present invention was derived from this technical background, and provides accurate and safe operation of the robot arm by reflecting the effect of vibration noise generated on other nearby axes in eliminating vibration noise generated on the axis of the robot arm. There is a purpose.

The present invention for achieving the above problems includes the following configuration.

That is, the method of removing vibration noise of a robot arm according to an embodiment of the present invention is a method performed on a computing device having one processor and a memory that stores one or more programs executed by the one or more processors, Step of receiving structural and kinematic characteristic information of the robot arm and formulating and modeling the kinematic characteristics of the robot arm; performing kinematic calculations on at least one axis of the robot arm using the modeling results from the modeling step A step of determining vibration noise for at least one axis of the robot arm, based on the performed kinematic calculation for at least one axis and the identified vibration noise, for each axis adjacent to the vibration noise generated in the axis. It includes a step of correcting the position control data for each axis by reflecting the vibration noise generated from the axis.

Meanwhile, a computer device having one or more processors and a memory that stores one or more programs executed by the one or more processors receives structural and kinematic characteristic information of the robot arm and formulates the kinematic characteristics of the robot arm. a modeling module that performs modeling, a computation performance module that performs a kinematic calculation on at least one axis of the robot arm using the modeling results from the modeling module, and a grasping module that determines vibration noise on at least one axis of the robot arm. Based on the module, the performed kinematic calculation for at least one axis, and the identified vibration noise, position control data for each axis is generated by reflecting the vibration noise generated on the corresponding axis and the vibration noise generated on the adjacent axis. Includes a correction module for correction.

According to the present invention, in removing the vibration noise generated on the axis of the robot arm, the effect of providing accurate and safe operation of the robot arm is derived by reflecting the effect of vibration noise generated on other nearby axes.

Figure 1 is a block diagram showing the configuration of a vibration noise removal device for a robot arm according to an embodiment of the present invention.
Figure 2 is a flowchart of a method for removing vibration noise of a robot arm according to an embodiment of the present invention.

It should be noted that the technical terms used in the present invention are only used to describe specific embodiments and are not intended to limit the present invention. In addition, the technical terms used in the present invention, unless specifically defined in a different sense in the present invention, should be interpreted as meanings generally understood by those skilled in the art in the technical field to which the present invention pertains, and are not overly comprehensive. It should not be interpreted in a literal or excessively reduced sense.

Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the attached drawings.

The apparatus for removing vibration noise of a robot arm according to embodiments of the present invention may be implemented by at least one computer device, and the method for removing vibration noise of a robot arm according to embodiments of the present invention may remove vibration noise of a robot arm. It may be performed through at least one computer device included in the device. At this time, a computer program according to an embodiment of the present invention may be installed and driven in the computer device, and the computer device may perform a method of removing vibration noise of a robot arm according to embodiments of the present invention under the control of the driven computer program. It can be done. The above-described computer program can be combined with a computer device and stored in a computer-readable recording medium to execute the method of removing vibration noise of the robot arm on the computer.

Figure 1 is a block diagram showing the configuration of a vibration noise removal device for a robot arm according to an embodiment of the present invention.

The manager terminal 30 is interpreted to encompass all terminal devices through which the manager performing the overall operation of the robot arm according to one embodiment receives information. In one embodiment, it is possible to control the movement of the robot arm performed by the vibration noise removal device 10 and provide various operation status information through the user interface of the administrator terminal 30.

The administrator terminal 30 is a smart phone, a portable terminal, a mobile terminal, a foldable terminal, a personal digital assistant (PDA), and a portable multimedia terminal (PMP). Player terminal, telematics terminal, navigation terminal, personal computer, laptop computer, Slate PC, Tablet PC, ultrabook, wearable device Device (e.g., smartwatch, smart glass, HMD (Head Mounted Display), etc.), Wibro terminal, IPTV (Internet Protocol Television) terminal, smart TV, digital broadcasting It can be applied to various terminals such as terminals, AVN (Audio Video Navigation) terminals, A/V (Audio/Video) systems, flexible terminals, and digital signage devices.

The network 20 is a network such as a personal area network (PAN), a local area network (LAN), a campus area network (CAN), a metropolitan area network (MAN), a wide area network (WAN), a broadband network (BBN), and the Internet. It may include one or more arbitrary networks. Additionally, the network 20 may include any one or more of network topologies including a bus network, star network, ring network, mesh network, star-bus network, tree or hierarchical network, etc. Not limited.

The apparatus 10 for removing vibration noise of a robot arm according to an embodiment includes a communication interface 110, a memory 120, an input/output interface 130, and a processor 140.

For example, if a robot arm has six axes, let's assume that there are axes 1, 2, and 3 to 6. Here, axis 2 will be affected by the vibration of axis 1 and will also be affected by the vibration of axis 3.

When analyzing the vibration of axis 2, the vibration noise removal device 10 of a robot arm according to one embodiment processes the application of noise generated from axis 1 and axis 3 to axis 2 based on data analysis. Here, the basic movement and vibration for each axis can be identified and further analyzed to select meaningful vibration. As a result, it is possible to effectively remove vibration noise, thereby maintaining the work accuracy and stability of the robot arm designed to perform precise work.

The communication interface 110 may provide a function for the vibration noise removal device 10 of the robot arm to communicate with other devices (eg, the storage devices described above) through the network 20. For example, requests, commands, data, files, etc. generated by the processor 140 of the vibration noise removal device 10 of the robot arm according to the program code stored in a recording device such as the memory 120 are transmitted through the communication interface 110. Depending on the control, it may be transmitted to other devices through the network 20. Conversely, signals, commands, data, files, etc. from other devices pass through the network 20 to the vibration noise removal device 10 of the robot arm through the communication interface 110 of the vibration noise removal device 10 of the robot arm. It can be received as . Signals, commands, data, etc. received through the communication interface 110 may be transmitted to the processor 140 or memory 120, and files, etc. may be further included in the vibration noise removal device 10 of the robot arm. Can be stored on a storage medium (persistent storage described above)

The memory 120 is a computer-readable recording medium and may include a non-permanent mass storage device such as random access memory (RAM), read only memory (ROM), and a disk drive. Here, non-destructive large-capacity recording devices such as ROM and disk drives may be included in the vibration noise removal device 10 of the robot arm as a separate permanent storage device separate from the memory 120.

Additionally, an operating system and at least one program code may be stored in the memory 120. These software components may be loaded into the memory 120 from a computer-readable recording medium separate from the memory 120. Such separate computer-readable recording media may include computer-readable recording media such as floppy drives, disks, tapes, DVD/CD-ROM drives, and memory cards. In another embodiment, software components may be loaded into the memory 120 through the communication interface 110 rather than a computer-readable recording medium. For example, software components may be loaded into the memory 120 of the vibration noise removal device 10 of the robotic arm based on a computer program installed by files received through the network 20.

The input/output interface 130 may be a means for interfacing with an input/output device. For example, input devices may include devices such as a microphone, keyboard, or mouse, and output devices may include devices such as displays and speakers. As another example, the input/output interface 130 may be a means for interfacing with a device that integrates input and output functions into one, such as a touch screen. The input/output device may be composed of a single device with the vibration noise removal device 10 of the robot arm.

The processor 140 may be configured to process instructions of a computer program by performing basic arithmetic, logic, and input/output operations. Commands may be provided to the processor 140 by the memory 120 or the communication interface 110. For example, the processor 140 may be configured to execute received instructions according to program codes stored in a recording device such as memory 120.

In more detail, the processor 140 includes a modeling module 1410, an operation performance module 1420, a grasping module 1430, and a correction module 1440.

The modeling module 1410 receives information on the structure and kinematic characteristics of the robot arm and formulates and models the kinematic characteristics of the robot arm.

The calculation performance module 1420 performs a kinematic calculation on at least one axis of the robot arm using the modeling result from the modeling module 1410.

The detection module 1430 detects vibration noise about at least one axis of the robot arm.

The correction module 1440 calculates the kinematic calculation for at least one axis performed by the calculation performance module 1420 and determines the axis adjacent to the vibration noise generated in the corresponding axis for each axis based on the vibration noise identified by the identification module 1430. The position control data for each axis is corrected by reflecting the vibration noise generated.

Hereinafter, a specific embodiment of a method for removing vibration noise of a robot arm will be described.

Figure 2 is a flowchart of a method for removing vibration noise of a robot arm according to an embodiment of the present invention.

The method for removing vibration noise of a robot arm according to an embodiment first receives structural and kinematic characteristic information of the robot arm and models the kinematic characteristics by formulating them (S200).

The movement of the robot arm can be controlled by modeling the robot arm using a mathematical model of the robot arm.

Modeling analysis of robotic arms plays an important role in understanding and predicting the movements of robotic arms. At this time, modeling analysis of the robot arm includes kinematic modeling and dynamic modeling.

Kinematic modeling is the process of mathematically modeling the movement of a robot arm. A robotic arm has various components connected, and each component performs different movements. By mathematically modeling the interactions between these components, the movement of the robot arm can be explained. An example is kinematic modeling using Denavit-Hartenberg (D-H) parameters.

For modeling of a robot arm, information about each axis of the robot arm, joint angles, link length, mass and center of gravity, and moment of inertia are collected.

Using modeling, kinematic calculations for each axis of the robot arm can be performed and various kinematic characteristics of the robot arm, such as position, speed, acceleration, torque, and force, can be estimated.

Meanwhile, in order to estimate the position, speed, acceleration, torque, force, etc. of each axis of the robot arm, the equation of motion of the robot arm must be formulated. To this end, the mass, length, and moment of inertia of each part of the robot arm can be measured, and the equation of motion of the robot arm can be derived.

It is a mathematical equation that explains the movement of the robot arm according to the force and acceleration applied to the robot arm during the modeling process, and the equation of motion of the robot arm based on Newton's laws of mechanics can be derived.

Each joint of a robot arm generally moves around a rotation axis. Therefore, the torque acting on each joint is proportional to the moment of inertia and angular acceleration of each joint. In robot arms, torque loss and friction must be taken into consideration in consideration of the efficiency with which torque is consumed.

In one embodiment, the equation of motion of the robot arm is expressed as Equation 1.

Here, τ represents the torque acting on the robot arm, q represents the joint angle of the robot arm, q' represents the angular velocity of each joint, and q'' represents the angular acceleration of each joint.

Additionally, M(q) is the inertia matrix of the robot arm and includes the moment of inertia of each joint. C(q, q') is the Coriolis and gravity term of the robot arm, which affects the movement of the robot arm, and G(q) represents the gravitational force acting on the robot arm.

Using these equations of motion, the motion of the robot arm can be mathematically modeled. Equations of motion can also be used in a variety of application fields, such as robot arm control and path planning.

Additionally, the Denavit-Hartenberg notation, which describes geometric transformations that describe the joint angles and lengths connecting each part of the robot arm, may be applied.

This is a robotics model that explains the position and direction of each joint of the robot arm. The lines connecting each joint of the robot arm are expressed as a fixed coordinate system called an axis, and through this, the position and direction of each joint are expressed. indicates.

Additionally, a transform matrix that describes the relative position and direction between each joint of the robot arm can be used. Each transpose matrix is defined by the DH parameter and is calculated using each joint angle.

You can also use a robot arm dynamics model that describes the forces and moments acting on the robot arm, and a robot arm control model that uses control algorithms such as a PID controller to calculate joint angles that allow the robot arm to move in the desired position and direction. .

In one aspect, the modeling step models interactions between adjacent axes.

In the modeling step, multi-axis robot dynamics modeling of the robot arm is performed to analyze the motion and force generated in each axis of the robot arm in order to calculate the vibration noise effect by modeling the interaction between adjacent axes. This allows us to predict the impact of interactions between adjacent axes and find ways to optimize the motion and forces of each axis.

At this time, the modeling step can be performed by cross-coupling modeling.

Cross-coupling modeling is one of the multi-axis robot dynamics modeling techniques that considers the interaction between adjacent axes.

Cross-coupling modeling is a method that considers the interaction between adjacent axes in a multi-axis dynamic model of a robot arm. To achieve this, the multi-axis robot dynamics model is modeled using general Newton equations and Lagrange equations, and cross-coupling interactions are additionally considered. Cross-coupling interaction can refer to a situation where the movement of one axis affects the movement of an adjacent axis.

By considering the interactions between adjacent axes using cross-coupling modeling, the motions and forces of multi-axis robots can be more accurately modeled.

In addition, the modeling step collects experimental data to measure the motion of the robot arm, and through this, the influence of interaction between adjacent axes can be analyzed to analyze the impact of vibration noise on adjacent axes. For this purpose, the movement of the robot arm is measured using sensors such as accelerometers or gyroscopes, and the frequency and amplitude of vibration noise are analyzed through this. Based on the analyzed values, the impact of interactions between adjacent axes can be estimated and implemented to reflect this in modeling.

The movement of the robot arm can be measured using sensors such as accelerometers or gyroscopes, and the frequency and amplitude of vibration noise can be analyzed through this. Based on this, in the correction step, which will be described later, the influence of interaction between adjacent axes can be estimated and reflected in modeling.

Then, a kinematic calculation is performed on at least one axis of the robot arm using the modeling results from the modeling step (S210).

The step of performing the calculation receives the modeling results of the robot arm from the modeling stage, as well as the target position and direction vector of the movement that the robot arm wants to perform. This can be input through the user interface of the administrator terminal 30. And based on the input values, the joint angles of the robot arm and the movement of the connected links are calculated.

Specifically, forward kinematics calculations for each axis of the robot arm can be performed. For this purpose, the position, speed, and acceleration of each joint of the robot arm are input along with the modeling results of the robot arm, and based on this, the position, speed, and acceleration of the end of the robot arm can be calculated.

At this time, motion planning is performed using information obtained through kinematic calculations of the robot arm. Through this, you can control the movement of the robot arm and allow the robot arm to move in the desired position and direction.

Additionally, the robot arm can be controlled using the motion planning results of the robot arm. For this purpose, the motor or actuator of the robot arm is controlled so that the robot arm can move in the correct position and direction.

In the step of performing kinematic calculations, forward kinematics formulas that calculate the position and direction of the robot arm can be derived through kinematic modeling of the robot arm using each joint angle of the robot arm.

Additionally, an inverse kinematics formula that calculates each joint angle can be derived using the position and direction of the robot arm.

Additionally, as a process of analyzing the movement in the environment in which the robot arm operates, the force and torque acting on various components of the robot arm are calculated through dynamic modeling in which the robot arm performs tasks such as picking up or moving objects. Thus, the robot arm may further perform mechanical calculations to determine the force and torque required to perform a certain task.

Mechanical modeling for mechanical calculation of a robot arm includes Newton's laws of mechanics, which describe the movement of the robot arm according to the force and acceleration applied to the robot arm, and a formula used to build a dynamic model that describes the movement of the robot arm. , the Euler-Lagrange equation, which generalizes the laws of Newtonian mechanics, can be used.

However, it is not limited to this, and mathematical concepts and formulas such as PID control and differential equations may be used in modeling the robot arm.

Afterwards, vibration noise for at least one axis of the robot arm is identified. (S220).

In one embodiment, identifying vibration noise may mean obtaining a vibration noise measurement value measured from a measuring device capable of measuring vibration noise on multiple axes.

For example, an accelerometer can be used to measure vibration occurring on the axis of a robot arm, and the data can be analyzed to determine the vibration frequency and amplitude.

Alternatively, the vibration signal measured with an accelerometer may be converted to the frequency domain using FFT, and the frequency spectrum may be analyzed to determine the size of the vibration noise for each frequency band.

Another example could be measuring thermal changes around the axis of a robot arm using an infrared camera. Through this, changes in frictional heat caused by vibration can be identified and vibration noise can be measured.

In addition, various tests and verification can be performed to check the vibration noise generated from the axis of the robot arm.

In another embodiment, the step of identifying vibration noise includes building a model for the movement of each axis of the robot arm modeled in the modeling step, calculating the vibration noise value expected to occur in each axis, and providing the calculated result. Using this, you can identify vibration noise generated from adjacent axes.

To this end, calculations for motion modeling and vibration noise prediction of each axis can be performed, and modeling considering interactions between adjacent axes can be performed.

Motion modeling of each axis includes inertia modeling, which sets the equation of motion by considering the moment of inertia and weight of each axis, friction modeling, which sets the equation of motion by considering the friction force of each axis, and setting the control method by considering the motor, sensor, etc. of each axis. This may include control modeling, etc.

Based on the modeled movement model of each axis, the vibration noise value expected to occur during movement can be calculated.

By using the vibration noise values predicted to occur in each axis, the vibration noise occurring in adjacent axes can be identified. To achieve this, calculations are made taking into account the interaction model between adjacent axes.

And based on the performed kinematic calculation for at least one axis and the identified vibration noise, the position control data for each axis is corrected by reflecting the vibration noise generated on the axis and the vibration noise generated on the adjacent axis. Do it (S230).

Analysis of vibration data includes the FFT (Fast Fourier Transform) algorithm, which converts and analyzes vibration data in the time domain to the frequency domain, the Wavelet transform algorithm, which is one of various analysis techniques used for signal analysis in the time-frequency domain, and the multidimensional data. PCA (Principal Component Analysis) algorithm that extracts key information from data by considering variance, clustering algorithm technique that divides data into several groups with similar properties, regression analysis algorithm that models the relationship between input and output variables, and complex patterns. An artificial neural network machine learning algorithm is useful for finding, and one of the deep learning algorithms can be used as an extension of the artificial neural network to learn more complex patterns. However, it is not limited to this.

In one aspect, the correction step may use a machine learning model to identify vibration noise generated in adjacent axes, predict vibration noise interaction between adjacent axes, and perform noise correction using the corresponding predicted value.

The correction step uses a deep learning algorithm to identify vibration noise generated in adjacent axes by a model learned through a multi-layered neural network.

Vibration noise data generated from the robot arm is collected from various measuring devices. Then, to train the deep learning model, the collected data is preprocessed, and the model is designed and trained. The learned model can identify vibration noise generated from adjacent axes, and based on this, the movement of the robot arm can be controlled more accurately.

Using a machine learning model, the movement pattern of the robot arm can be identified through model learning, and a more efficient control algorithm can be developed to identify vibration noise generated from adjacent axes in the robot arm.

In another aspect, the correction step may predict noise generated in each axis using an interaction model, and use the corresponding predicted value to correct the influence of noise generated in adjacent axes.

An interaction model is a method of analyzing and modeling interaction relationships between various variables, and can be used in fields such as statistics, machine learning, and artificial intelligence.

In other words, when collecting noise data generated from each axis, noise data from adjacent axes is also collected by considering the interaction between adjacent axes.

We then model the noise generated from each axis and the interaction between adjacent axes. To this end, the model can be trained and predicted using noise data on each axis and noise data on adjacent axes. The value predicted by the model is the predicted value of noise generated on the corresponding axis, and by using this as the noise value generated on each axis, the impact of noise generated on adjacent axes can be determined.

And based on the predicted noise value, a more accurate control algorithm can be developed.

In one embodiment, vibration data measured in all axes of the robot arm is collected to correct position control data for each axis. Then, the frequency components of each axis are analyzed through FFT analysis on the collected vibration data.

And based on the analysis results, the vibration noise of each axis can be separated by considering the influence of vibration noise of the next axis.

By using the vibration noise data separated from each axis, the modeling of the corresponding axis is updated, and the steps of analyzing the vibration noise on all axes are repeated through the updated modeling, the vibration data measured on all axes can be analyzed simultaneously. . As a result, vibration noise can be removed by reflecting the effect of vibration noise on adjacent axes.

As an example, the correction step can be performed by independently analyzing vibration data measured on each axis.

To compensate for the effects of vibration noise generated in other axes, modeling of each axis can be used to remove the vibration predicted in other axes. For example, when analyzing vibration data measured on the To this end, the vibration noise in each axis can be modeled and analyzed, and an algorithm can be performed to calculate and remove the predicted vibration in that axis.

In one embodiment, the correction step reflects the effect on overall vibration noise by multiplying the results of performing FFT analysis on the vibration data measured on each axis, and the multiplied result is calculated through IFFT (Inverse Fast Fourier Transform). Convert it back to the time domain to derive the final vibration data.

Specifically, performing FFT analysis on the vibration data measured on each axis, multiplying the results of the FFT analysis, and converting the multiplied result back to the time domain through IFFT (Inverse Fast Fourier Transform) to derive the final vibration data. You can.

Additionally, filtering techniques can be used to remove vibration noise generated from other axes. For example, high-frequency vibration noise generated on another axis can be filtered out to remove the vibration expected from data measured on that axis.

The correction step can minimize vibration noise occurring in the robot arm axis by improving the vibration noise in a specific axis and the control algorithm of the robot arm. For this purpose, various kinematic modeling and control algorithms can be applied. In other words, the correction step uses robotic arm modeling and control algorithms to understand how the movements of each axis affect each other.

The control algorithm of the robot arm is an algorithm that determines the control method of the robot arm and controls the robot arm using the modeling results of the robot arm and sensor data. In the compensation step, this algorithm can compensate for the impact of the movement of each axis on other axes.

The above-described method may be implemented as an application or in the form of program instructions that can be executed through various computer components and recorded on a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, etc., singly or in combination.

The program instructions recorded on the computer-readable recording medium may be those specifically designed and configured for the present invention, or may be known and usable by those skilled in the computer software field.

Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floptical disks. media), and hardware devices specifically configured to store and perform program instructions, such as ROM, RAM, flash memory, etc.

Examples of program instructions include not only machine language code such as that created by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware device may be configured to operate as one or more software modules to perform processing according to the invention and vice versa.

Although the above has been described with reference to embodiments, those skilled in the art will understand that various modifications and changes can be made to the present invention without departing from the spirit and scope of the present invention as set forth in the following patent claims. You will be able to.

10: Vibration noise removal device 20: Network
30: Administrator terminal 110: Communication interface
120: memory 130: input/output interface
140: Processor 1410: Modeling module
1420: calculation performance module 1430: grasp module
1440: Calibration module

Claims (8)

one processor, and
A method performed on a computing device having a memory that stores one or more programs executed by the one or more processors, comprising:
Step of receiving structural and kinematic characteristic information of the robot arm and formulating and modeling the kinematic characteristics of the robot arm;
performing a kinematic calculation on at least one axis of the robot arm using the modeling result from the modeling step;
Determining vibration noise about at least one axis of the robot arm;
Based on the performed kinematic calculation for at least one axis and the identified vibration noise, the position control data for each axis is corrected by reflecting the vibration noise generated on the axis and the vibration noise generated on the adjacent axis for each axis. Including steps;
The modeling step is,
Model interactions between adjacent axes,
The correction step is,
Predict vibration noise generated from adjacent axes using the model modeled in the modeling step,
The correction step is,
Using a machine learning model, we identify vibration noise generated from adjacent axes, predict vibration noise interaction between adjacent axes, and use the predicted values to perform noise compensation.
An interaction model is used to predict noise on each axis, and the predicted values are used to compensate for the effects of noise on adjacent axes.
The step of identifying the vibration noise is,
In the modeling step, a model for the movement of each axis of the modeled robot arm is built to calculate the vibration noise value expected to occur in each axis, and the calculation result is used to identify vibration noise generated in adjacent axes, A model is trained and predicted using the noise data of each axis and the noise data of adjacent axes to model the interaction between noise occurring in each axis and adjacent axes, and to determine the impact of noise occurring in adjacent axes.
The correction step is,
The results of FFT analysis on the vibration data measured on each axis are multiplied to reflect the overall effect on vibration noise, and the multiplied results are converted back to the time domain through IFFT (Inverse Fast Fourier Transform) to produce the final vibration data. Derive and remove vibration noise generated on other axes by filtering out high-frequency vibration noise generated on other axes and remove vibration predicted from data measured on that axis.
A method of removing vibration noise of a robot arm that performs motion planning using information obtained through the kinematic calculation of the robot arm in the step of performing the kinematic calculation. delete According to claim 1,
The modeling step is a method of removing vibration noise of a robot arm, where cross-coupling modeling is performed. delete delete According to claim 1,
The step of identifying the vibration noise is,
A method of removing vibration noise from a robot arm that obtains vibration noise measurements from measurement equipment that can measure vibration noise on multiple axes.
delete one or more processors, and
A computer device having a memory storing one or more programs to be executed by the one or more processors,
A modeling module that receives structural and kinematic characteristic information of the robot arm and formulates and models the kinematic characteristics of the robot arm;
a computation performance module that performs a kinematic computation on at least one axis of the robot arm using modeling results from the modeling module;
A detection module that detects vibration noise about at least one axis of the robot arm;
Based on the performed kinematic calculation for at least one axis and the identified vibration noise, the position control data for each axis is corrected by reflecting the vibration noise generated on the axis and the vibration noise generated on the adjacent axis for each axis. Comprising a correction module;
The modeling module is,
Model interactions between adjacent axes,
The correction module is,
Predict vibration noise generated from adjacent axes using the model modeled in the modeling module,
The correction module is,
Using a machine learning model, we identify vibration noise generated from adjacent axes, predict vibration noise interaction between adjacent axes, and use the predicted values to perform noise compensation.
An interaction model is used to predict noise on each axis, and the predicted values are used to compensate for the effects of noise on adjacent axes.
The grasping module is,
Build a model for the movement of each axis of the robot arm modeled in the modeling module to calculate the vibration noise value expected to occur in each axis, use the calculation result to determine the vibration noise generated in adjacent axes, and calculate the vibration noise value expected to occur in each axis. A model is trained and predicted using noise data on an axis and noise data on adjacent axes to model the interaction between noise occurring on each axis and adjacent axes, and identify the impact of noise occurring on adjacent axes.
The correction module is,
The results of FFT analysis on the vibration data measured on each axis are multiplied to reflect the overall effect on vibration noise, and the multiplied results are converted back to the time domain through IFFT (Inverse Fast Fourier Transform) to produce the final vibration data. Derive and remove vibration noise generated on other axes by filtering out high-frequency vibration noise generated on other axes and remove vibration predicted from data measured on that axis.
A computer device that performs motion planning using information obtained through the kinematic calculation of a robot arm in the kinematic calculation performing module.