CN111590581A - Positioning compensation method and device for robot - Google Patents

Positioning compensation method and device for robot Download PDF

Info

Publication number
CN111590581A
CN111590581A CN202010456895.5A CN202010456895A CN111590581A CN 111590581 A CN111590581 A CN 111590581A CN 202010456895 A CN202010456895 A CN 202010456895A CN 111590581 A CN111590581 A CN 111590581A
Authority
CN
China
Prior art keywords
robot
error compensation
moving
error
compensation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202010456895.5A
Other languages
Chinese (zh)
Other versions
CN111590581B (en
Inventor
郭东生
钟文涛
高小云
张志波
万文洁
周家裕
张睿
王佳威
衷镇宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Gree Electric Appliances Inc of Zhuhai
Zhuhai Gree Intelligent Equipment Co Ltd
Original Assignee
Gree Electric Appliances Inc of Zhuhai
Zhuhai Gree Intelligent Equipment Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Gree Electric Appliances Inc of Zhuhai, Zhuhai Gree Intelligent Equipment Co Ltd filed Critical Gree Electric Appliances Inc of Zhuhai
Priority to CN202010456895.5A priority Critical patent/CN111590581B/en
Publication of CN111590581A publication Critical patent/CN111590581A/en
Priority to PCT/CN2020/139939 priority patent/WO2021238191A1/en
Application granted granted Critical
Publication of CN111590581B publication Critical patent/CN111590581B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Programme-controlled manipulators
    • B25J9/16—Programme controls
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Programme-controlled manipulators
    • B25J9/16—Programme controls
    • B25J9/1656—Programme controls characterised by programming, planning systems for manipulators
    • B25J9/1664—Programme controls characterised by programming, planning systems for manipulators characterised by motion, path, trajectory planning
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Programme-controlled manipulators
    • B25J9/16—Programme controls
    • B25J9/1679—Programme controls characterised by the tasks executed
    • B25J9/1692—Calibration of manipulator

Landscapes

  • Engineering & Computer Science (AREA)
  • Robotics (AREA)
  • Mechanical Engineering (AREA)
  • Manipulator (AREA)
  • Numerical Control (AREA)

Abstract

The invention discloses a positioning compensation method and device for a robot. Wherein, the method comprises the following steps: collecting a moving point set of the robot when the robot moves in a preset space grid, wherein the space grid refers to a grid which is divided according to a cube with a preset step length in a working area of the robot; training by utilizing the moving point set to obtain an error compensation model, wherein the error compensation model is a model established according to error compensation parameters between a theoretical position and an actual position of a flange central point of the robot; acquiring actual coordinates of a target position to which the robot is supposed to arrive; calculating a compensation parameter of the robot moving to the actual coordinate by using the error compensation model, wherein the compensation parameter is used for compensating the positioning precision of the robot moving to the actual coordinate; and controlling the robot to reach the target position based on the compensation parameters. The invention solves the technical problems that the robot positioning error compensation speed is low, the adaptability is low and the positioning precision supplement requirement of the robot cannot be met in the related technology.

Description

Positioning compensation method and device for robot
Technical Field
The invention relates to the technical field of robot control, in particular to a positioning compensation method and device for a robot.
Background
Because the industrial robot is adopted for processing and assembling, the absolute positioning precision is mainly depended on, if the absolute positioning precision is too low, the product quality is seriously influenced, and the scheme for improving the positioning precision of the robot at present comprises the following steps: off-line error compensation and on-line error compensation. The method comprises the following steps of performing offline error compensation, namely compensating the positioning accuracy of the robot by a certain method before the robot is used for working, such as a mathematical approximation method, online error compensation and the like, wherein the methods for compensating the positioning accuracy of the robot have some defects, and the mathematical approximation method has a poor effect of solving a complex nonlinear model and is difficult to meet the high-accuracy requirements of the application of drilling and riveting of the airplane and the like; on-line error compensation usually requires the addition of a real-time feedback device at the end of the robot to enable the robot to continuously adjust the end to the desired position during operation. On-line error compensation can generally obtain higher positioning accuracy, but the robot is not easy to operate in some complex scenes due to the fact that a feedback device is added on the robot end effector, so that the positioning accuracy compensation speed of the robot is low easily, and the adaptability is low.
In view of the above problems, no effective solution has been proposed.
Disclosure of Invention
The embodiment of the invention provides a positioning compensation method and a positioning compensation device for a robot, which are used for at least solving the technical problems that the robot positioning error compensation speed is low, the adaptability is low and the requirement for supplementing the positioning precision of the robot cannot be met in the related technology.
According to an aspect of an embodiment of the present invention, there is provided a positioning compensation method of a robot, including: collecting a moving point set of a robot when the robot moves in a preset space grid, wherein the space grid is a grid divided according to a cube with a preset step length in a working area of the robot; training by utilizing the moving point set to obtain an error compensation model, wherein the error compensation model is a model established according to error compensation parameters between a theoretical position and an actual position of a flange central point of the robot; acquiring actual coordinates of a target position to which the robot is supposed to arrive; calculating a compensation parameter of the robot moving to the actual coordinate by using the error compensation model, wherein the compensation parameter is used for compensating the positioning precision of the robot moving to the actual coordinate; and controlling the robot to reach the target position based on the compensation parameter.
Optionally, the step of collecting a moving point set of the robot when moving in the preset spatial grid includes: dividing the working area of the robot into a plurality of space cubes by taking the basic coordinate of the robot as a reference and the preset step length as a dividing rule; collecting three-dimensional coordinates of sampling points at the tail end of the robot in a plurality of space cubes according to a preset collection rule from a first base point of a working area of the robot; based on the three-dimensional coordinates, a set of movement points of the robot while moving is determined.
Optionally, after acquiring the moving point set of the robot moving on the preset spatial grid, the positioning compensation method further includes: dividing the plurality of spatial cubes into a plurality of sampling layers; collecting the change parameters and the change characteristics of the robot in the directions of the coordinate systems in each sampling layer; and establishing the error compensation model based on the change parameters and the change characteristics.
Optionally, the training of the moving point set to obtain an error compensation model includes: representing the corresponding actual coordinate of each sampling point in the moving point set by a first vector, and representing the corresponding theoretical coordinate of each sampling point in the moving point set by a second vector; determining a feedback neural network that constructs the error compensation model, the feedback neural network comprising: an input layer, a hidden layer and an output layer; inputting a vector set consisting of the first vector and the second vector to an input layer of the feedback neural network; determining a first connection weight between the neuron of the input layer and the neuron of the hidden layer, and determining a second connection weight between the neuron of the hidden layer and the neuron of the output layer; determining a theoretical coordinate set and a test coordinate set of a sampling point based on the first connection weight and the second connection weight; and determining the error compensation model based on the theoretical coordinate set, and testing the error compensation model by using the test coordinate set.
Optionally, determining a first connection weight between the neuron of the input layer and the neuron of the hidden layer, and determining a second connection weight between the neuron of the hidden layer and the neuron of the output layer comprises: determining the number of neurons of the hidden layer; generating a first connection weight between the neuron of the input layer and the neuron of the hidden layer based on the number of the neurons; determining an activation function and a hidden layer output matrix for constructing an error compensation model; determining a second connection weight between the neuron of the hidden layer and the neuron of the output layer based on the activation function, the hidden layer output matrix, the first connection weight, and the neuron number threshold of the hidden layer.
Optionally, after training with the moving point set to obtain an error compensation model, the positioning compensation method further includes: determining an error vector and an absolute error value corresponding to each sampling point based on the actual coordinate and the theoretical coordinate of each sampling point; determining a prediction bias value corresponding to each sample point based on the error vector and the absolute error value; calculating a prediction deviation average value based on the prediction deviation value of each sampling point; and screening a target value range based on the prediction deviation average value, wherein the target value range is used for optimizing the calculation accuracy of the error compensation model.
Optionally, the error compensation model is constructed using an extreme learning algorithm ELM.
According to another aspect of the embodiments of the present invention, there is also provided a positioning compensation apparatus for a robot, including: the system comprises a collecting unit, a calculating unit and a processing unit, wherein the collecting unit is used for collecting a moving point set of the robot when a preset space grid moves, and the space grid is a grid divided according to a cube with a preset step length in a working area of the robot; the training unit is used for training by utilizing the moving point set to obtain an error compensation model, wherein the error compensation model is a model established according to error compensation parameters between a theoretical position and an actual position of a flange central point of the robot; an acquisition unit for acquiring actual coordinates of a target position to which the robot is supposed to arrive; a moving unit, configured to calculate a compensation parameter for the robot to move to the actual coordinate by using the error compensation model, where the compensation parameter is used to compensate for a positioning accuracy of the robot to move to the actual coordinate; and the control unit is used for controlling the robot to reach the target position based on the compensation parameter.
Optionally, the acquisition unit comprises: the first dividing module is used for dividing the working area of the robot into a plurality of space cubes by taking the basic coordinate of the robot as a reference and the preset step length as a dividing rule; the first acquisition module is used for acquiring three-dimensional coordinates of sampling points at the tail end of the robot in a plurality of space cubes from a first base point of a working area of the robot according to a preset acquisition rule; and the first determination module is used for determining a moving point set of the robot when moving based on the three-dimensional coordinates.
Optionally, the positioning compensation apparatus further comprises: the second division module is used for dividing the plurality of space cubes into a plurality of sampling layers after a moving point set of the robot moving in a preset space grid is collected; the second acquisition module is used for acquiring the change parameters and the change characteristics of the robot in the directions of the coordinate systems in each sampling layer; and the first establishing module is used for establishing the error compensation model based on the change parameters and the change characteristics.
Optionally, the training unit comprises: the second determining module is used for representing the corresponding actual coordinate of each sampling point in the moving point set by using a first vector and representing the corresponding theoretical coordinate of each sampling point in the moving point set by using a second vector; a third determining module, configured to determine a feedback neural network for constructing the error compensation model, where the feedback neural network includes: an input layer, a hidden layer and an output layer; the first input module is used for inputting a vector set consisting of the first vector and the second vector to an input layer of the feedback neural network; a fourth determining module, configured to determine a first connection weight between the neuron of the input layer and the neuron of the hidden layer, and determine a second connection weight between the neuron of the hidden layer and the neuron of the output layer; a fifth determining module, configured to determine a theoretical coordinate set and a test coordinate set of the sampling point based on the first connection weight and the second connection weight; and the sixth determining module is used for determining the error compensation model based on the theoretical coordinate set and testing the error compensation model by using the test coordinate set.
Optionally, the fourth determining module includes: the first determining submodule is used for determining the number of the neurons of the hidden layer; a first generation module, configured to generate a first connection weight between the neuron of the input layer and the neuron of the hidden layer based on the number of the neurons; the second determining submodule is used for determining an activation function and a hidden layer output matrix for constructing an error compensation model; a third determining submodule, configured to determine a second connection weight between the neuron of the hidden layer and the neuron of the output layer based on the activation function, the hidden layer output matrix, the first connection weight, and the neuron number threshold of the hidden layer.
Optionally, the positioning compensation apparatus further comprises: the seventh determining module is used for determining an error vector and an absolute error value corresponding to each sampling point based on the actual coordinate and the theoretical coordinate of each sampling point after an error compensation model is obtained by utilizing the moving point set training; an eighth determining module, configured to determine a prediction bias value corresponding to each sampling point based on the error vector and the absolute error value; the calculation module is used for calculating a prediction deviation average value based on the prediction deviation value of each sampling point; and the screening module is used for screening a target value interval based on the prediction deviation average value, wherein the target value interval is used for optimizing the calculation accuracy of the error compensation model.
Optionally, the error compensation model is constructed using an extreme learning algorithm ELM.
According to another aspect of an embodiment of the present invention, there is also provided an industrial robot including: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the method of positioning compensation of a robot of any of the above via execution of the executable instructions.
According to another aspect of the embodiments of the present invention, there is also provided a computer storage medium, where the computer storage medium includes a stored program, and when the program runs, the apparatus where the computer storage medium is located is controlled to execute any one of the above-mentioned methods for positioning compensation of a robot.
In the embodiment of the invention, when the positioning accuracy of the robot is compensated, a moving point set of the robot moving in a preset space grid is collected, an error compensation model is obtained by training the moving point set, then the actual coordinates of the target position to which the robot is supposed to arrive can be obtained, the compensation parameters of the robot moving to the actual coordinates are calculated by using the error compensation model, and the robot is controlled to arrive at the target position based on the compensation parameters. In this embodiment, can adopt the moving point set of different sampling intervals as training set to carry out error compensation model training and compensation test, and establish the error compensation model between robot flange central point theoretical position and the actual position, can compensate the absolute positioning accuracy of arbitrary point in the work area of robot through the error compensation model, the compensation speed is fast, and absolute positioning accuracy is high, the generalization performance is good, can satisfy the positioning accuracy demand of robot, thereby it is slower to solve robot positioning error compensation speed among the correlation technique, and adaptability is low, can not satisfy the technical problem of the positioning accuracy replenishment demand of robot.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the invention without limiting the invention. In the drawings:
FIG. 1 is a flow chart of an alternative method of robot position compensation according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of an alternative spatial grid sampling point plan in accordance with an embodiment of the present invention;
FIG. 3 is a schematic diagram of an alternative feedback neural network according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of an alternative method of constructing an error compensation model, according to an embodiment of the invention;
fig. 5 is a schematic diagram of an alternative industrial robot coordinate system transformation relationship according to an embodiment of the present invention;
FIG. 6 is a flow chart of an alternative method of converting coordinate systems of a robot in accordance with embodiments of the present invention;
fig. 7 is a schematic diagram of an alternative positioning compensation device for a robot according to an embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
To facilitate understanding of the present invention by those skilled in the art, some terms or nouns referred to in the embodiments of the present invention are explained below:
the ELM, Extreme learning Machine for short is a Machine learning system or method constructed based on a feedforward neural network. The method applies the ELM to the positioning accuracy compensation of the robot, establishes an error compensation model based on the ELM algorithm, and can compensate the absolute positioning accuracy of any point in a working area.
The error compensation model provided by the embodiment of the invention can optimize the neuron value of the hidden layer and improve the positioning precision.
In accordance with an embodiment of the present invention, there is provided an embodiment of a method for position compensation of a robot, it is noted that the steps illustrated in the flowchart of the drawings may be performed in a computer system such as a set of computer executable instructions and that, although a logical order is illustrated in the flowchart, in some cases the steps illustrated or described may be performed in an order different than that presented herein.
Fig. 1 is a flowchart of an alternative method for compensating the positioning of a robot according to an embodiment of the present invention, as shown in fig. 1, the method includes the following steps:
step S102, collecting a moving point set of the robot when moving in a preset space grid, wherein the space grid refers to a grid which is divided according to a cube with a preset step length in a working area of the robot;
step S104, training by utilizing the moving point set to obtain an error compensation model, wherein the error compensation model is a model established according to error compensation parameters between a theoretical position and an actual position of a flange central point of the robot;
step S106, acquiring the actual coordinates of the target position to which the robot is supposed to arrive;
step S108, calculating a compensation parameter of the robot moving to the actual coordinate by using the error compensation model, wherein the compensation parameter is used for compensating the positioning precision of the robot moving to the actual coordinate;
and step S110, controlling the robot to reach the target position based on the compensation parameters.
Through the steps, when the positioning accuracy of the robot is compensated, a moving point set of the robot when the robot moves in a preset space grid is collected, an error compensation model is obtained through the moving point set training, actual coordinates of a target position to which the robot is supposed to arrive can be obtained, compensation parameters of the robot moving to the actual coordinates are calculated through the error compensation model, and the robot is controlled to arrive at the target position based on the compensation parameters. In this embodiment, can adopt the moving point set of different sampling intervals as training set to carry out error compensation model training and compensation test, and establish the error compensation model between robot flange central point theoretical position and the actual position, can compensate the absolute positioning accuracy of arbitrary point in the work area of robot through the error compensation model, the compensation speed is fast, and absolute positioning accuracy is high, the generalization performance is good, can satisfy the positioning accuracy demand of robot, thereby it is slower to solve robot positioning error compensation speed among the correlation technique, and adaptability is low, can not satisfy the technical problem of the positioning accuracy replenishment demand of robot.
The robot related to the embodiment of the invention includes but is not limited to: industrial robots (such as six-axis robots) and educational robots can be applied to scenes such as welding, drilling and riveting, polishing and paint spraying by using the robots, and the absolute positioning precision of the robots during automatic operation of parts is improved.
The present invention will be described below with reference to the above-described embodiments.
Step S102, a moving point set of the robot when moving in a preset space grid is collected, wherein the space grid refers to a grid which is divided according to a cube with a preset step length in a working area of the robot.
By sampling the spatial grid of the robot, a set of moving points can be obtained.
Optionally, the step of collecting a moving point set of the robot when the robot moves in a preset space grid includes: dividing a working area of the robot into a plurality of space cubes by taking a base coordinate of the robot as a reference and taking a preset step length as a division rule; collecting three-dimensional coordinates of sampling points at the tail end of the robot in a plurality of space cubes according to a preset collection rule from a first base point of a working area of the robot; based on the three-dimensional coordinates, a set of movement points of the robot while moving is determined.
Fig. 2 is a schematic diagram of an alternative spatial grid sampling point planning according to an embodiment of the present invention, as shown in fig. 2, with a robot base coordinate as a reference, an actual working area, i.e. a cube P1P2P3P4P5P6P7P8, is first planned in a robot working space, and a distance d in the working area is a predetermined step (fixed sampling step), where a sampling path may be: zigzag sampling is carried out on the plane P1P2P3P4, namely, the 2 nd line is collected from the starting point P1 to the position P2 and then returns to the position P1P3 which is a preset step distance away from P1 until the sampling reaches P4, and the 2 nd layer is collected in the same way until the sampling end point P8 is completely collected. And taking the zero point position of the robot as an initial position for each sampling, reaching a sampling point in the same posture, and then measuring the three-dimensional coordinate of the actual position of the tail end of the robot by using measuring equipment to obtain a moving point set.
Optionally, after acquiring the moving point set of the robot when moving on the preset spatial grid, the positioning compensation method further includes: dividing a plurality of space cubes into a plurality of sampling layers; collecting the change parameters and the change characteristics of the robot in the directions of the coordinate systems in each sampling layer; and establishing an error compensation model based on the change parameters and the change characteristics.
The coordinate system directions of the sampling layers include: x-direction, Y-direction, Z-direction.
Alternatively, after the error compensation model is obtained by training using the moving point set, the positioning compensation method further includes: determining an error vector and an absolute error value corresponding to each sampling point based on the actual coordinate and the theoretical coordinate of each sampling point; determining a prediction bias value corresponding to each sample point based on the error vector and the absolute error value; calculating a prediction deviation average value based on the prediction deviation value of each sampling point; and screening a target value range based on the prediction deviation average value, wherein the target value range is used for optimizing the calculation accuracy of the error compensation model.
Based on the sampling space shown in FIG. 2, let the theoretical coordinate of the ith sampling point (i.e. black dot in the figure) be Ti(xi，yi，zi) (i ═ 1: n), the actual coordinate is Pi (x'i，y′i，z′i) (i ═ 1: n) representing an error vector of an ith sample point by a first formula, wherein the first formula is:
ei＝Pi-Ti＝(x′i-xi，y′i-yi，z′i-zi) Formula (1);
let Δ xi＝x′i-xi,Δyi＝y′i-yi,Δzi＝z′i-ziThen the first formula may be represented by the following second formula:
ei＝(Δxi，Δyi，Δzi) Formula (2);
then, an absolute error value, which is an euclidean distance, is expressed by a third formula:
Figure BDA0002509696110000071
by combining the acquisition paths of all sampling points, the absolute error values of all sampling layers (parallel to the surface P1P2P3P 4) are compared to reflect the change condition of the positioning error along the Z direction of the robot base coordinate system, the absolute error values of all sampling rows (parallel to the line segment P1P 2) on the single sampling layer are compared to reflect the change condition of the positioning error along the X direction of the robot base coordinate system, and the absolute error values of all sampling points on the single sampling row are compared to reflect the change condition of the positioning error along the Y direction of the robot base coordinate system.
In the embodiment of the present invention, by comparing the directions of the positioning error vectors with the points on the line segment parallel to P1P5, P1P3, or P1P2, the variation characteristics of the positioning error direction along each direction of the robot base coordinate system are obtained, and the magnitude and direction of the positioning error of the robot have determined continuous variation characteristics along different coordinate axis directions of the base coordinate system, and the variation characteristics may include:
(1) the absolute error value is basically unchanged along with the increase of the Z coordinate of the sampling point;
(2) the absolute error value increases along with the decrease of the X coordinate of the sampling point;
(3) the absolute error value decreases with the increase of the Y coordinate of the sampling point, but increases abruptly at Y ═ 0.
After the change characteristics are obtained, an error model can be further established based on the change parameters and the change characteristics, and then the robot positioning accuracy compensation is carried out.
And step S104, training by using the moving point set to obtain an error compensation model, wherein the error compensation model is a model established according to error compensation parameters between the theoretical position and the actual position of the flange central point of the robot.
In the embodiment of the invention, an error compensation model can be constructed by using an extreme learning algorithm ELM, and the structure of the ELM can be understood as a feedback neural network with a single hidden layer.
Optionally, the obtaining of the error compensation model by training using the moving point set includes: representing the corresponding actual coordinate of each sampling point in the moving point set by using a first vector, and representing the corresponding theoretical coordinate of each sampling point in the moving point set by using a second vector; determining a feedback neural network for constructing an error compensation model, wherein the feedback neural network comprises: an input layer, a hidden layer and an output layer; inputting a vector set consisting of the first vector and the second vector to an input layer of a feedback neural network; determining a first connection weight between the neuron of the input layer and the neuron of the hidden layer, and determining a second connection weight between the neuron of the hidden layer and the neuron of the output layer; determining a theoretical coordinate set and a test coordinate set of the sampling point based on the first connection weight and the second connection weight; and determining an error compensation model based on the theoretical coordinate set, and testing the error compensation model by using the test coordinate set.
Fig. 3 is a schematic structural diagram of an alternative feedback neural network according to an embodiment of the present invention, as shown in fig. 3, which includes an input layer, an implicit layer, and an output layer.
The input layer and the hidden layer, and the hidden layer and the output layer are connected with each other. Through I1，I2，I3……INRepresenting input layer neurons, by O1，O2，O3……OMRepresenting output layer neurons by q1，q2，q3……qLRepresenting hidden layer neurons, wijDenotes a first connection weight between the i (i ═ 1: N) th neuron of the input layer and the j (j ═ 1: L) th neuron of the hidden layer, βjkA second connection weight between the j (j 1: L) th neuron of the hidden layer and the k (k 1: M) th neuron of the output layer is representedi、ti(i ═ 1: n) denotes the actual and theoretical coordinates of the sample point, respectively.
The actual coordinates are determined by a fourth formula, the theoretical coordinates are determined by a fifth formula,
the fourth formula: pi＝[x′iy′iz′i]TI ═ 1: n, formula (4);
the fifth formula: t is ti＝[xiyizi]T，i＝1: n, formula (5);
with pi、tiFor a set of corresponding inputs and outputs of the ELM algorithm, the present invention may first define M-N-3, x'iy′iz′iAnd xiyiziRespectively correspond to I1，I2，I3And O1，O2，O3Then, all the inputs P (represented by the sixth formula) and the outputs T (represented by the seventh formula) of the ELM algorithm are:
the sixth formula:
Figure BDA0002509696110000091
a seventh formula:
Figure BDA0002509696110000092
alternatively, the determining a first connection weight between the neuron of the input layer and the neuron of the hidden layer, and the determining a second connection weight between the neuron of the hidden layer and the neuron of the output layer includes: determining the number of neurons of the hidden layer; generating a first connection weight between the neuron of the input layer and the neuron of the hidden layer based on the number of the neurons; determining an activation function and a hidden layer output matrix for constructing an error compensation model; determining a second connection weight between the neuron of the hidden layer and the neuron of the output layer based on the activation function, the hidden layer output matrix, the first connection weight and the neuron number threshold of the hidden layer.
And setting the number L of neurons in the hidden layer, and randomly generating a first connection weight w between the input layer and the hidden layer and a neuron threshold b of the hidden layer. The first connection weight w is represented by an eighth formula, and the neuron threshold is represented by a ninth formula, where the eighth formula is:
Figure BDA0002509696110000093
the ninth formula is:
Figure BDA0002509696110000101
setting the activation function of the ELM algorithm to be f (x) (f (x)) infinitely differentiable in any interval, only the second connection weight beta between the hidden layer and the output layer is not determined at the moment, and the beta is an L multiplied by 3 dimensional matrix. According to the linear network of fig. 3, the fifth formula can be re-expressed by a tenth formula, which is:
Figure BDA0002509696110000102
wherein: w is aj＝[wj1wj2w3]，j＝1：L。
Formula (10) may be further represented by formula (11): h β ═ T.
H is a hidden layer output matrix in the feedback neural network and is represented by four known quantities, namely input P, an activation function f (x), a first connection weight w and a threshold b.
To prevent the over-fitting problem of the ELM algorithm, the value of the number L of hidden layer neurons is smaller than the number n of training samples, so the solution of the second connection weight β between the hidden layer and the output layer is formula (12):
Figure BDA0002509696110000103
wherein H+Is the generalized inverse of the output matrix.
Because the value of the number L of the hidden layer neurons as the ELM algorithm parameters has a large influence on the generalization performance of the error compensation model, too few hidden layer neurons can cause large errors of the fitting model, too many hidden layer neurons can cause the error model to be over-fitted, and the training efficiency of the model is reduced, the L value is further adjusted to optimize the compensation model in the embodiment of the invention.
Fig. 4 is a schematic diagram of an optional error compensation model construction according to an embodiment of the present invention, as shown in fig. 4, when an ELM algorithm is used for training, training is performed by combining a selected activation function and a set number of neurons in a hidden layer through a theoretical coordinate of a sampling point and an actual coordinate of the sampling point, so as to obtain an error compensation model, and in an optimization stage of the error compensation model, a predicted coordinate of the test point is obtained by combining the actual coordinate of the test point, and based on the theoretical coordinate of the test point, whether positioning accuracy meets a requirement is determined, if yes, the error compensation model can be further optimized, and if not, the number of neurons in the hidden layer needs to be readjusted, so as to further optimize the compensation model.
In order to optimize the value of the number L of the neurons, another group of collection points in the envelope range of the sampling points are taken as the test points of the optimization target in the error compensation algorithm, and the theoretical coordinate set and the actual measurement coordinate set are respectively recorded as XtAnd Yp. The theoretical coordinate set is expressed by equation (13), and the actual measurement coordinate set is expressed by equation (14).
Figure BDA0002509696110000111
Figure BDA0002509696110000112
Where m represents the number of test point set samples.
With the actual set of coordinates YpAs input of ELM algorithm, the theoretical coordinate predicted by the error model can be calculated by the trained formula (10) and is marked as Xf. The theoretical coordinates are represented by the formula (15),
Figure BDA0002509696110000113
the prediction deviation E of the error compensation model is the theoretical coordinate X of the test point settAnd theoretical coordinate X predicted by error modelfThe difference of (a) is represented by the formula (16):
Figure BDA0002509696110000114
from equations (1) and (3), the prediction error of the jth test point (j ═ 1: m) is expressed by equation (17):
Figure BDA0002509696110000115
wherein the content of the first and second substances,
Figure BDA0002509696110000116
predicted deviation mean EARepresented by formula (18):
Figure BDA0002509696110000117
predicted deviation EAThe smaller the value is, the better the generalization performance of the error model established under the corresponding value of L is, and the programming automatically adjusts the number L of the neurons in the hidden layer. Specifically, the value is taken from 1% of the number n of samples in the training set until the number n of samples is the same as the number n of samples in the training set according to EAAnd (3) selecting an L value interval with good generalization performance, wherein the L value determines the number of equations in an ELM algorithm linear equation set, so that weights w and β and a threshold b with smaller L values are taken in the L value interval with good generalization performance to replace an equation (10) to obtain an optimized error compensation model in order to improve the training solution and prediction efficiency.
Step S106, acquiring the actual coordinates of the target position to which the robot is supposed to arrive.
After the actual coordinates of the target position to be reached by the robot are determined, the theoretical coordinates of the robot during movement are calculated through an error compensation model, and then the robot is controlled to accurately reach the target position through compensation parameters.
And S108, calculating a compensation parameter of the robot moving to the actual coordinate by using the error compensation model, wherein the compensation parameter is used for compensating the positioning precision of the robot moving to the actual coordinate.
And step S110, controlling the robot to reach the target position based on the compensation parameters.
The invention is illustrated below by means of a further alternative embodiment.
The embodiment of the invention is explained by taking GR20A type 6-degree-of-freedom industrial robot as an example, so as to analyze the positioning error rule of the robot and complete an error compensation experiment based on an ELM algorithm, and a 3-dimensional three-optical measuring instrument is used as measuring equipment.
After acquiring data of a plurality of sampling points (for example, acquiring data of 2 groups of sampling points 2197 of 13 points × 13 rows × 13 layers), randomly acquiring a first group of data points as test points for an error model optimization test, and randomly acquiring a second group of data points as verification points for an actual compensation effect of an error compensation model in a planned sampling cube, where fig. 5 is a schematic diagram of a transformation relation of coordinate systems of an optional industrial robot according to an embodiment of the present invention, and as shown in fig. 5, the transformation relation includes establishing a base coordinate system with a center point of a base of the robot, a flange center coordinate system, a world coordinate system, and a measurement system coordinate system.
Fig. 6 is a flowchart of an alternative method for converting coordinate systems of a robot according to an embodiment of the present invention, and as shown in fig. 6, a relationship among a flange coordinate system, a world coordinate system, and a measurement system coordinate system may be established by the measurement system, so as to rotate axes a1 and a6 of the robot, respectively, measure coordinates of a marker point at each rotational position in the world coordinate system, measure initial position coordinates of a marker point in the world coordinate system, and measure coordinates of a marker point at each theoretical position in the world coordinate system. After the coordinates of the mark points at each rotation position in the world coordinate system are measured, circle fitting can be respectively carried out on the mark points obtained by measurement when the A1 and A6 axes rotate, the x axis and the z axis of the robot in the world coordinate system are further obtained, the conversion relation between the world coordinate system and the robot base coordinate system can be determined by combining the coordinates of the flange plate center at the initial position in the world coordinate system, and the translation relation between the world coordinate system and the robot base coordinate system is determined. After the coordinates of the marking points at each theoretical position in the world coordinate system are measured, the coordinates of the flange center of the robot at each theoretical position in the world coordinate system are determined by combining the deviation parameters of the marking points and the flange center, and further the actual coordinates of the flange center of the robot at each theoretical position in the base coordinate system are determined.
And (3) after coordinate conversion is carried out on the 3-dimensional coordinates of the first group of verification points acquired in the experiment, absolute error calculation is carried out according to the formulas (1) to (3), and an error range and an error value before compensation are obtained. According to the proposed error compensation model, the absolute error condition after compensation is obtained, the absolute error condition is compared with the absolute error value before error compensation, and the result shows that the absolute positioning accuracy of the robot can be effectively improved by using an ELM algorithm to train the error compensation model and perform compensation test by using point sets with different sampling intervals as a training set.
The invention is illustrated below by means of a further alternative embodiment.
Fig. 7 is a schematic diagram of an alternative positioning compensation device for a robot according to an embodiment of the present invention, and as shown in fig. 7, the positioning compensation device may include: an acquisition unit 71, a training unit 73, an acquisition unit 75, a movement unit 77, a control unit 79, wherein,
the acquisition unit 71 is configured to acquire a moving point set of the robot when moving in a preset spatial grid, where the spatial grid is a grid divided according to a cube with a predetermined step length in a working area of the robot;
the training unit 73 is used for obtaining an error compensation model by utilizing the moving point set training, wherein the error compensation model is a model established according to error compensation parameters between the theoretical position and the actual position of the flange central point of the robot;
an acquisition unit 75 for acquiring actual coordinates of a target position to which the robot is supposed to arrive;
a moving unit 77 for calculating a compensation parameter of the robot moving to the actual coordinate by using the error compensation model, wherein the compensation parameter is used for compensating the positioning accuracy of the robot moving to the actual coordinate;
and a control unit 79 for controlling the robot to reach the target position based on the compensation parameter.
When the positioning accuracy of the robot is compensated, the positioning compensation device of the robot firstly acquires a moving point set of the robot moving in a preset space grid through the acquisition unit 71, then obtains an error compensation model through the training unit 73 by using the moving point set training, then obtains an actual coordinate of a target position to which the robot is supposed to arrive through the acquisition unit 75, calculates a compensation parameter of the robot moving to the actual coordinate through the moving unit 77 by using the error compensation model, and controls the robot to arrive at the target position through the control unit 79 based on the compensation parameter. In this embodiment, can adopt the moving point set of different sampling intervals as training set to carry out error compensation model training and compensation test, and establish the error compensation model between robot flange central point theoretical position and the actual position, can compensate the absolute positioning accuracy of arbitrary point in the work area of robot through the error compensation model, the compensation speed is fast, and absolute positioning accuracy is high, the generalization performance is good, can satisfy the positioning accuracy demand of robot, thereby it is slower to solve robot positioning error compensation speed among the correlation technique, and adaptability is low, can not satisfy the technical problem of the positioning accuracy replenishment demand of robot.
Optionally, the collecting unit includes: the first division module is used for dividing the working area of the robot into a plurality of space cubes by taking the basic coordinate of the robot as a reference and taking a preset step length as a division rule; the first acquisition module is used for acquiring three-dimensional coordinates of sampling points at the tail end of the robot in a plurality of space cubes from a first base point of a working area of the robot according to a preset acquisition rule; the first determination module is used for determining a moving point set of the robot when the robot moves on the basis of the three-dimensional coordinates.
In an embodiment of the present invention, the positioning compensation apparatus further includes: the second division module is used for dividing the plurality of space cubes into a plurality of sampling layers after the moving point set of the robot moving in the preset space grid is collected; the second acquisition module is used for acquiring the change parameters and the change characteristics of the robot in the directions of the coordinate systems in each sampling layer; and the first establishing module is used for establishing an error compensation model based on the change parameters and the change characteristics.
Optionally, the training unit includes: the second determining module is used for representing the corresponding actual coordinate of each sampling point in the moving point set by using the first vector and representing the corresponding theoretical coordinate of each sampling point in the moving point set by using the second vector; a third determining module, configured to determine a feedback neural network for constructing the error compensation model, where the feedback neural network includes: an input layer, a hidden layer and an output layer; the first input module is used for inputting a vector set consisting of the first vector and the second vector to an input layer of the feedback neural network; the fourth determination module is used for determining a first connection weight between the neuron of the input layer and the neuron of the hidden layer and determining a second connection weight between the neuron of the hidden layer and the neuron of the output layer; the fifth determining module is used for determining a theoretical coordinate set and a test coordinate set of the sampling point based on the first connection weight and the second connection weight; and the sixth determining module is used for determining the error compensation model based on the theoretical coordinate set and testing the error compensation model by utilizing the test coordinate set.
Alternatively, the fourth determining module includes: the first determining submodule is used for determining the number of the neurons of the hidden layer; the first generation module is used for generating a first connection weight between the neuron of the input layer and the neuron of the hidden layer based on the number of the neurons; the second determining submodule is used for determining an activation function and a hidden layer output matrix for constructing an error compensation model; and the third determining submodule is used for determining a second connection weight value between the neuron of the hidden layer and the neuron of the output layer based on the activation function, the hidden layer output matrix, the first connection weight value and the neuron number threshold of the hidden layer.
Optionally, the positioning compensation apparatus further includes: the seventh determining module is used for determining an error vector and an absolute error value corresponding to each sampling point based on the actual coordinate and the theoretical coordinate of each sampling point after an error compensation model is obtained by utilizing the training of the moving point set; an eighth determining module, configured to determine a prediction bias value corresponding to each sampling point based on the error vector and the absolute error value; the calculation module is used for calculating a prediction deviation average value based on the prediction deviation value of each sampling point; and the screening module is used for screening a target value range based on the prediction deviation average value, wherein the target value range is used for optimizing the calculation accuracy of the error compensation model.
Optionally, an error compensation model is constructed by using an extreme learning algorithm ELM.
The above-mentioned positioning compensation device for the robot may further include a processor and a memory, where the above-mentioned acquisition unit 71, the training unit 73, the acquisition unit 75, the moving unit 77, the control unit 79, and the like are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to implement corresponding functions.
The processor comprises a kernel, and the kernel calls a corresponding program unit from the memory. The kernel can be set to be one or more, and the precision of the robot in positioning is compensated by adjusting the kernel parameters.
The memory may include volatile memory in a computer readable medium, Random Access Memory (RAM) and/or nonvolatile memory such as Read Only Memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip.
According to another aspect of an embodiment of the present invention, there is also provided an industrial robot including: a processor; and a memory for storing executable instructions for the processor; wherein the processor is configured to perform the positioning compensation method of the robot of any one of the above via execution of the executable instructions.
According to another aspect of the embodiments of the present invention, there is also provided a computer storage medium including a stored program, wherein when the program runs, an apparatus in which the computer storage medium is located is controlled to execute any one of the above-mentioned methods for robot positioning compensation.
The present application further provides a computer program product adapted to perform a program for initializing the following method steps when executed on a data processing device: collecting a moving point set of the robot when the robot moves in a preset space grid, wherein the space grid refers to a grid which is divided according to a cube with a preset step length in a working area of the robot; training by utilizing the moving point set to obtain an error compensation model, wherein the error compensation model is a model established according to error compensation parameters between a theoretical position and an actual position of a flange central point of the robot; acquiring actual coordinates of a target position to which the robot is supposed to arrive; calculating a compensation parameter of the robot moving to the actual coordinate by using the error compensation model, wherein the compensation parameter is used for compensating the positioning precision of the robot moving to the actual coordinate; and controlling the robot to reach the target position based on the compensation parameters.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
In the above embodiments of the present invention, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
In the embodiments provided in the present application, it should be understood that the disclosed technology can be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units may be a logical division, and in actual implementation, there may be another division, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, units or modules, and may be in an electrical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic or optical disk, and other various media capable of storing program codes.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.

Claims (10)

1. A method for compensating for a position of a robot, comprising:
collecting a moving point set of a robot when the robot moves in a preset space grid, wherein the space grid is a grid divided according to a cube with a preset step length in a working area of the robot;
training by utilizing the moving point set to obtain an error compensation model, wherein the error compensation model is a model established according to error compensation parameters between a theoretical position and an actual position of a flange central point of the robot;
acquiring actual coordinates of a target position to which the robot is supposed to arrive;
calculating a compensation parameter of the robot moving to the actual coordinate by using the error compensation model, wherein the compensation parameter is used for compensating the positioning precision of the robot moving to the actual coordinate;
and controlling the robot to reach the target position based on the compensation parameter.
2. The positioning compensation method according to claim 1, wherein the step of collecting the moving point set of the robot when the robot moves in the preset spatial grid comprises:
dividing the working area of the robot into a plurality of space cubes by taking the basic coordinate of the robot as a reference and the preset step length as a dividing rule;
collecting three-dimensional coordinates of sampling points at the tail end of the robot in a plurality of space cubes according to a preset collection rule from a first base point of a working area of the robot;
based on the three-dimensional coordinates, a set of movement points of the robot while moving is determined.
3. The positioning compensation method according to claim 2, wherein after acquiring the set of moving points of the robot while moving on the preset spatial grid, the positioning compensation method further comprises:
dividing the plurality of spatial cubes into a plurality of sampling layers;
collecting the change parameters and the change characteristics of the robot in the directions of the coordinate systems in each sampling layer;
and establishing the error compensation model based on the change parameters and the change characteristics.
4. The method of claim 2, wherein the training of the set of moving points to obtain an error compensation model comprises:
representing the corresponding actual coordinate of each sampling point in the moving point set by a first vector, and representing the corresponding theoretical coordinate of each sampling point in the moving point set by a second vector;
determining a feedback neural network that constructs the error compensation model, the feedback neural network comprising: an input layer, a hidden layer and an output layer;
inputting a vector set consisting of the first vector and the second vector to an input layer of the feedback neural network;
determining a first connection weight between the neuron of the input layer and the neuron of the hidden layer, and determining a second connection weight between the neuron of the hidden layer and the neuron of the output layer;
determining a theoretical coordinate set and a test coordinate set of a sampling point based on the first connection weight and the second connection weight;
and determining the error compensation model based on the theoretical coordinate set, and testing the error compensation model by using the test coordinate set.
5. The method of claim 4, wherein determining first connection weights between the neurons of the input layer and the neurons of the hidden layer, and determining second connection weights between the neurons of the hidden layer and the neurons of the output layer comprises:
determining the number of neurons of the hidden layer;
generating a first connection weight between the neuron of the input layer and the neuron of the hidden layer based on the number of the neurons;
determining an activation function and a hidden layer output matrix for constructing an error compensation model;
determining a second connection weight between the neuron of the hidden layer and the neuron of the output layer based on the activation function, the hidden layer output matrix, the first connection weight, and the neuron number threshold of the hidden layer.
6. The location compensation method of claim 4, wherein after training with the moving point set to obtain an error compensation model, the location compensation method further comprises:
determining an error vector and an absolute error value corresponding to each sampling point based on the actual coordinate and the theoretical coordinate of each sampling point;
determining a prediction bias value corresponding to each sample point based on the error vector and the absolute error value;
calculating a prediction deviation average value based on the prediction deviation value of each sampling point;
and screening a target value range based on the prediction deviation average value, wherein the target value range is used for optimizing the calculation accuracy of the error compensation model.
7. The position compensation method of claim 1, wherein the error compensation model is constructed using an extreme learning algorithm (ELM).
8. A positioning compensation device of a robot, comprising:
the system comprises a collecting unit, a calculating unit and a processing unit, wherein the collecting unit is used for collecting a moving point set of the robot when a preset space grid moves, and the space grid is a grid divided according to a cube with a preset step length in a working area of the robot;
the training unit is used for training by utilizing the moving point set to obtain an error compensation model, wherein the error compensation model is a model established according to error compensation parameters between a theoretical position and an actual position of a flange central point of the robot;
an acquisition unit for acquiring actual coordinates of a target position to which the robot is supposed to arrive;
a moving unit, configured to calculate a compensation parameter for the robot to move to the actual coordinate by using the error compensation model, where the compensation parameter is used to compensate for a positioning accuracy of the robot to move to the actual coordinate;
and the control unit is used for controlling the robot to reach the target position based on the compensation parameter.
9. An industrial robot, characterized by comprising:
a processor; and
a memory for storing executable instructions of the processor;
wherein the processor is configured to perform the method of positioning compensation of a robot of any of claims 1 to 7 via execution of the executable instructions.
10. A computer storage medium, characterized in that the computer storage medium comprises a stored program, wherein when the program runs, the computer storage medium is controlled by a device to execute the positioning compensation method of the robot according to any one of claims 1 to 7.
CN202010456895.5A 2020-05-26 2020-05-26 Positioning compensation method and device for robot Active CN111590581B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202010456895.5A CN111590581B (en) 2020-05-26 2020-05-26 Positioning compensation method and device for robot
PCT/CN2020/139939 WO2021238191A1 (en) 2020-05-26 2020-12-28 Positioning compensation method and apparatus for robot

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010456895.5A CN111590581B (en) 2020-05-26 2020-05-26 Positioning compensation method and device for robot

Publications (2)

Publication Number Publication Date
CN111590581A true CN111590581A (en) 2020-08-28
CN111590581B CN111590581B (en) 2021-10-22

Family

ID=72181550

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010456895.5A Active CN111590581B (en) 2020-05-26 2020-05-26 Positioning compensation method and device for robot

Country Status (2)

Country Link
CN (1) CN111590581B (en)
WO (1) WO2021238191A1 (en)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112551044A (en) * 2020-12-10 2021-03-26 惠州市恒泰科技股份有限公司 Method and device for feeding to-be-formed battery cell
CN112643669A (en) * 2020-12-04 2021-04-13 广州机械科学研究院有限公司 Robot position deviation compensation method, system, device and storage medium
CN113514050A (en) * 2021-04-01 2021-10-19 中车青岛四方机车车辆股份有限公司 Positioning method, system, medium, equipment, mobile platform and overhaul production line
CN113635312A (en) * 2021-10-18 2021-11-12 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） Industrial robot motion precision compensation method and system and computer equipment
WO2021238191A1 (en) * 2020-05-26 2021-12-02 珠海格力智能装备有限公司 Positioning compensation method and apparatus for robot
CN114001655A (en) * 2021-11-02 2022-02-01 深圳市腾盛精密装备股份有限公司 Method and device for determining precision of machine equipment
CN115416018A (en) * 2022-08-17 2022-12-02 雅客智慧（北京）科技有限公司 End effector deformation compensation method and device, electronic equipment and storage medium
CN115648228A (en) * 2022-12-28 2023-01-31 广东隆崎机器人有限公司 Industrial robot multi-source error compensation method, device, equipment and storage medium
CN116079696A (en) * 2023-04-07 2023-05-09 广东毕要科技有限公司 SCARA robot positioning updating method and device based on vibration signals
CN113514050B (en) * 2021-04-01 2024-05-28 佛山中车四方轨道车辆有限公司 Positioning method, system, medium, equipment, mobile platform and overhaul production line

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114347018B (en) * 2021-12-20 2024-04-16 上海大学 Mechanical arm disturbance compensation method based on wavelet neural network
CN114378816A (en) * 2021-12-27 2022-04-22 歌尔股份有限公司 Equipment motion track deviation rectifying method, control device and storage medium
CN114683279B (en) * 2022-02-17 2023-05-23 广东博智林机器人有限公司 Navigation precision determining method and device and electronic equipment
CN114523475A (en) * 2022-03-01 2022-05-24 南京理工大学 Automatic error calibration and compensation device and method for robot assembly system
CN114474004B (en) * 2022-03-18 2023-10-03 河北工业大学 Error compensation planning control strategy for multi-factor coupling vehicle-mounted building robot
CN114734440B (en) * 2022-04-15 2023-09-05 同济大学 Precise calibration method for kinematic parameters of hybrid double-arm transfer robot
CN115183716B (en) * 2022-09-10 2023-01-13 武汉光昱明晟智能科技有限公司 Earth measurement method and system based on intelligent navigation robot
CN115543096B (en) * 2022-12-02 2023-04-07 世优（北京）科技有限公司 Dynamic capture data processing method, device and system based on neural network
CN116061176B (en) * 2022-12-27 2024-02-23 哈尔滨思哲睿智能医疗设备股份有限公司 Motion compensation method, motion compensation device, electronic equipment and storage medium
CN116113229B (en) * 2023-04-14 2023-06-09 合肥安迅精密技术有限公司 Method and device for guaranteeing consistency of feeding points of multiple femtocells and storage medium

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0589394A1 (en) * 1992-09-21 1994-03-30 Toyoda Koki Kabushiki Kaisha Method and apparatus for controlling a robot using a neural network
CN105353301A (en) * 2015-10-23 2016-02-24 华北电力大学（保定） Circuit breaker state assessment parameter obtaining method
CN106407659A (en) * 2016-08-31 2017-02-15 中国地质大学（武汉） Air quality index (AQI) predicting method and device
CN106737855A (en) * 2016-08-22 2017-05-31 南京理工大学 A kind of robot precision compensation method of comprehensive position and attitude error model and rigidity compensation
CN108818540A (en) * 2018-08-28 2018-11-16 珠海格力智能装备有限公司 Method and device, the pocessor and storage media of compensating parameter
CN110039537A (en) * 2019-03-15 2019-07-23 北京精密机电控制设备研究所 A kind of automatic measure on line multi joint motion planing method neural network based
CN110385720A (en) * 2019-07-26 2019-10-29 南京航空航天大学 A kind of robot localization error compensating method based on deep neural network

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH07261821A (en) * 1994-03-16 1995-10-13 Fanuc Ltd Robot track planning method taking deflection due to load into consideration
CN111055293A (en) * 2019-11-21 2020-04-24 东南大学 High-precision constant-force grinding method for industrial robot based on curved surface self-adaption
CN111590581B (en) * 2020-05-26 2021-10-22 珠海格力智能装备有限公司 Positioning compensation method and device for robot

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0589394A1 (en) * 1992-09-21 1994-03-30 Toyoda Koki Kabushiki Kaisha Method and apparatus for controlling a robot using a neural network
CN105353301A (en) * 2015-10-23 2016-02-24 华北电力大学（保定） Circuit breaker state assessment parameter obtaining method
CN106737855A (en) * 2016-08-22 2017-05-31 南京理工大学 A kind of robot precision compensation method of comprehensive position and attitude error model and rigidity compensation
CN106407659A (en) * 2016-08-31 2017-02-15 中国地质大学（武汉） Air quality index (AQI) predicting method and device
CN108818540A (en) * 2018-08-28 2018-11-16 珠海格力智能装备有限公司 Method and device, the pocessor and storage media of compensating parameter
CN110039537A (en) * 2019-03-15 2019-07-23 北京精密机电控制设备研究所 A kind of automatic measure on line multi joint motion planing method neural network based
CN110385720A (en) * 2019-07-26 2019-10-29 南京航空航天大学 A kind of robot localization error compensating method based on deep neural network

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021238191A1 (en) * 2020-05-26 2021-12-02 珠海格力智能装备有限公司 Positioning compensation method and apparatus for robot
CN112643669A (en) * 2020-12-04 2021-04-13 广州机械科学研究院有限公司 Robot position deviation compensation method, system, device and storage medium
CN112551044B (en) * 2020-12-10 2022-09-27 惠州市恒泰科技股份有限公司 Method and device for feeding to-be-formed battery cell
CN112551044A (en) * 2020-12-10 2021-03-26 惠州市恒泰科技股份有限公司 Method and device for feeding to-be-formed battery cell
CN113514050A (en) * 2021-04-01 2021-10-19 中车青岛四方机车车辆股份有限公司 Positioning method, system, medium, equipment, mobile platform and overhaul production line
CN113514050B (en) * 2021-04-01 2024-05-28 佛山中车四方轨道车辆有限公司 Positioning method, system, medium, equipment, mobile platform and overhaul production line
CN113635312A (en) * 2021-10-18 2021-11-12 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） Industrial robot motion precision compensation method and system and computer equipment
US11660745B2 (en) 2021-10-18 2023-05-30 China Electronics Reliability And Environmental Testing Institute ((The Fifth Institute Of Electronics, Ministry Of Industry And Information Technology) (China Saibao Laboratory)) Industrial robot motion accuracy compensation method and system, and computer device
CN114001655A (en) * 2021-11-02 2022-02-01 深圳市腾盛精密装备股份有限公司 Method and device for determining precision of machine equipment
CN115416018A (en) * 2022-08-17 2022-12-02 雅客智慧（北京）科技有限公司 End effector deformation compensation method and device, electronic equipment and storage medium
CN115416018B (en) * 2022-08-17 2024-03-15 雅客智慧（北京）科技有限公司 End effector deformation compensation method, device, electronic equipment and storage medium
CN115648228A (en) * 2022-12-28 2023-01-31 广东隆崎机器人有限公司 Industrial robot multi-source error compensation method, device, equipment and storage medium
CN116079696A (en) * 2023-04-07 2023-05-09 广东毕要科技有限公司 SCARA robot positioning updating method and device based on vibration signals
CN116079696B (en) * 2023-04-07 2023-06-13 广东毕要科技有限公司 SCARA robot positioning updating method and device based on vibration signals

Also Published As

Publication number Publication date
CN111590581B (en) 2021-10-22
WO2021238191A1 (en) 2021-12-02

Similar Documents

Publication Publication Date Title
CN111590581B (en) Positioning compensation method and device for robot
CN110385720B (en) Robot positioning error compensation method based on deep neural network
CN108908327B (en) Robot positioning error grading compensation method
CN111964575B (en) Digital twin modeling method for milling of mobile robot
CN110900598B (en) Robot three-dimensional motion space action simulation learning method and system
CN109782601B (en) Design method of self-adaptive neural network synchronous robust controller of coordinated mechanical arm
CN105259786B (en) The inertial parameter discrimination method and device of target to be identified
CN102183205A (en) Method for matching optimal assembly poses of large-sized parts
CN111260649B (en) Close-range mechanical arm sensing and calibrating method
CN111975771A (en) Mechanical arm motion planning method based on deviation redefinition neural network
CN110715769A (en) Method for calibrating stress point position of weighing sensor of multi-point method centroid measuring equipment
CN111752151A (en) Adaptive tracking and compensating method and system for grinding and polishing industrial blade
CN111673733B (en) Intelligent self-adaptive compliance control method of robot in unknown environment
CN110728088A (en) Method and device for optimizing transfer station parameters of tracker for three-dimensional thermal expansion deformation of workpiece
CN113878581B (en) Error prediction and real-time compensation method for five-degree-of-freedom hybrid robot
CN113910218B (en) Robot calibration method and device based on kinematic and deep neural network fusion
CN114398049A (en) Self-adaptive dynamic updating method for digital twin model of discrete manufacturing workshop
US20220258336A1 (en) Model generation apparatus, model generation method, control apparatus, and control method
CN110065068A (en) A kind of robotic asssembly operation programming by demonstration method and device based on reverse-engineering
Landgraf et al. A hybrid neural network approach for increasing the absolute accuracy of industrial robots
CN115229796A (en) Robot end pose tracking compensation method for aviation large member milling
CN114131595B (en) Robot 6D pose estimation system and method
CN108717262B (en) Special-shaped curved surface tracking method and system based on moment feature learning neural network
CN113799137A (en) Mobile processing robot rapid visual positioning method based on neural network
CN113370208A (en) Control method and system for reconfigurable mechanical arm subsystem

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CB03 Change of inventor or designer information

Inventor after: Wen Hui

Inventor after: Zhong Zhenyu

Inventor after: Zhong Wentao

Inventor after: Zhang Tianyi

Inventor after: Guo Dongsheng

Inventor after: Gao Xiaoyun

Inventor after: Zhang Zhibo

Inventor after: Wan Wenjie

Inventor after: Zhou Jiayu

Inventor after: Zhang Rui

Inventor after: Wang Jiawei

Inventor before: Guo Dongsheng

Inventor before: Zhong Wentao

Inventor before: Gao Xiaoyun

Inventor before: Zhang Zhibo

Inventor before: Wan Wenjie

Inventor before: Zhou Jiayu

Inventor before: Zhang Rui

Inventor before: Wang Jiawei

Inventor before: Zhong Zhenyu

CB03 Change of inventor or designer information