CN110788859B - Controller parameter universe self-adaptive adjustment system - Google Patents
Controller parameter universe self-adaptive adjustment system Download PDFInfo
- Publication number
- CN110788859B CN110788859B CN201911022463.7A CN201911022463A CN110788859B CN 110788859 B CN110788859 B CN 110788859B CN 201911022463 A CN201911022463 A CN 201911022463A CN 110788859 B CN110788859 B CN 110788859B
- Authority
- CN
- China
- Prior art keywords
- controller
- reference point
- module
- parameters
- sample
- 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.)
- Active
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
- B25J9/1602—Programme controls characterised by the control system, structure, architecture
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a controller parameter universe self-adaptive adjusting system, which comprises the following steps of firstly, determining a clustering center of three driving joint load inertias of a parallel mechanism in a dynamic platform reference point working space universe and the membership degree of a sample inertia about each center by means of a fuzzy clustering algorithm; then, off-line setting parameters of feedback and feedforward controllers at corresponding characteristic points of each clustering center, and storing the parameters and membership data in a control system database; and finally, calculating and updating the controller parameters of the three driving joints in real time by using the data in the database and the current coordinate information of the reference point of the movable platform of the parallel mechanism by adopting a gravity center method.
Description
Technical Field
The invention relates to a controller parameter universe self-adaptive adjusting method of a five-degree-of-freedom hybrid robot, relates to the field of robot technology and automation, and can effectively improve the motion control precision of an end effector of the five-degree-of-freedom hybrid robot.
Background
The robot control system generally adopts a composite control strategy of 'feedback + feedforward', and a feedback controller determines the system bandwidth and ensures the stability; the feedforward controller further improves the following accuracy of the system. However, for a hybrid robot system including a parallel mechanism, the load inertia converted to the parallel mechanism driving joint changes along with the robot configuration. The controller adopting the fixed gain is difficult to meet the requirements of high-speed and high-precision motion of the robot in a working space, and the application of the controller in occasions with higher motion precision requirements (such as machining) is restricted.
For such strong nonlinear time-varying systems, advanced control algorithms such as dynamics control algorithms, neural networks, genetic algorithms, and the like have been proposed, but due to the complexity of the algorithms, the amount of calculation, and lack of hardware support, it is difficult to put the algorithms into practical use. Therefore, a controller parameter adaptive adjustment method which is suitable for such a robot system and has a simple algorithm and is easy to implement is needed, so that parameters can be matched with the change of load inertia in the whole working space, and online adjustment of the robot under different configurations is realized.
Disclosure of Invention
Aiming at a five-degree-of-freedom hybrid robot with a rotating support disclosed in patent CN104985596A, the invention provides a controller parameter universe self-adaptive adjusting system which can effectively cope with the influence of the factor that the load inertia changes along with the configuration change on the control quality when a parallel mechanism drives a joint to move. The method is characterized by firstly determining the clustering centers of the load inertia of the three driving joints of the parallel mechanism in the whole working space of the reference point of the moving platform and the membership degree of the sample inertia about each center by means of a fuzzy clustering algorithm, setting the controller parameters at the corresponding characteristic points of each clustering center in an off-line manner, and then estimating the controller parameters of the three driving joints when the reference point of the moving platform is positioned at any point in the working space on line by using a gravity center method. The method has the advantages that the parameters can be matched with the load inertia change to realize global self-adaptive adjustment only by offline setting the controller parameters at a plurality of limited characteristic points, the algorithm is simple, the occupied hardware resources are less, and the method is easy to realize.
A controller parameter universe self-adaptive adjusting system comprises a coarse interpolation module, a position inverse solution module and a control algorithm module, and further comprises a controller parameter estimation module, wherein the controller parameter estimation module executes the following steps to adjust parameter matching load inertia variation, and the system comprises:
using the membership degree obtained by clustering analysis and the controller parameter obtained by off-line setting as a global variable to be written into a corresponding register for calling, and using the controller parameter set at the midpoint in a working space as an initial value;
the coarse interpolation module performs coarse interpolation on the NC codes according to the motion rule and calculates the end pose after a coarse interpolation period is finished;
the position inverse solution module calculates the corresponding movable platform reference point coordinates and driving joint instructions, and writes the movable platform reference point coordinates into a temporary stack register as global variables;
the rough interpolation module calls the controller parameter estimation module in an interruption mode according to integral multiples of a rough interpolation period, and reads the reference point coordinates, membership and n characteristic point controller parameters K of the moving platform from corresponding registers i,k (k is 1,2 …, n), and estimating the controller parameter of the i-th driving joint corresponding to the coordinate of the reference point by using the barycentric method
And as the global variable writes into the corresponding register;
calling control algorithm module to read from corresponding register
Using updated
Calculating the output command of the controller and holding it until the next estimation
And is not changed.
The controller parameter
The parameters of the controller of the driving joint when the reference point of the movable platform is positioned at any point in the working space; the joint is subjected to weighted estimation on the controller parameters at all the characteristic points, the weight is determined by the sample point closest to the current reference point, and the serial number of the sample point is defined as m (j is m); the parameter estimation algorithm of the controller is
The cluster analysis step comprises:
(1) the grid divides the working space of the reference point of the movable platform of the parallel mechanism, and defines grid nodes as sample points
Calculating load inertia of each driving joint by using sample points according to a robot mathematical model, and defining j-th sample inertia of the ith joint as I i,j ；
(2) According to the fuzzy clustering analysis algorithm, given the number of clustering classes N (N < N), the degree of membership mu of the j-th sample inertia with respect to the k-th (k is 1,2, …, N) clustering center at the l-th iteration is calculated according to a formula i,j,k,l And a new clustering center I i,k,l+1 ；
In the formula, α represents a blur index larger than 1, and is usually 2;
(3) continuously updating mu according to the above formula i,j,k,l And I i,k,l+1 Until, | | I is satisfied at the first iteration i,k,l+1 -I i,k,l And | | | is less than or equal to epsilon. Here, epsilon represents the cluster center iteration precision.
Advantageous effects
In order to ensure the stability of the system operation, the task priority of the calling controller parameter estimation module is lower than the priority of the calling servo algorithm module, and the calling of the module is realized by adopting a servo task interruption mode. The calling period of the controller parameter estimation module can be determined according to task requirements and hardware computing capability, and can be generally set as an integer (n) of the coarse interpolation period T ) And (4) doubling. Accordingly, at every n T The coarse interpolation period performs one controller parameter estimation and update and remains unchanged until the next estimation.
The five-degree-of-freedom hybrid robot controller parameter universe self-adaptive adjusting method determines the membership degrees of the feature points and any points in a working space relative to the feature points by means of a clustering analysis algorithm, and estimates the controller parameters when the reference points are located at any points on line according to the membership degrees. The method has the advantages that parameters can be matched with load inertia change to realize global self-adaptive adjustment only by offline setting of controller parameters at a plurality of limited characteristic points; the algorithm is simple, and the occupied hardware resources are less; the servo algorithm is independent from the servo algorithm, and parameters can be adjusted in real time on the premise of ensuring the stability of servo control.
Drawings
FIG. 1 is a block diagram of a hybrid robot universe adaptive control strategy
FIG. 2 schematic representation of workspace grid partitioning and feature points
FIG. 3 flow chart of controller parameter adaptive tuning calculation
Detailed Description
In order to make the technical scheme of the invention clearer, the invention is further described in detail with reference to the accompanying drawings. It should be understood that the specific examples described herein are intended to illustrate the invention, but are not intended to limit the invention to these examples. The following describes an embodiment of the present invention, taking an example of a five-degree-of-freedom hybrid robot with a rotating bracket disclosed in patent CN 104985596A.
The invention relates to a controller parameter universe adaptive adjustment system (as shown in figure 1), which comprises the following steps:
1. fuzzy clustering analysis
The grid divides the working space of the reference point (hereinafter referred to as the reference point) of the movable platform of the parallel mechanism, and defines grid nodes as sample points. And calculating the load inertia of each driving joint by using the sample points according to the robot mathematical model, and defining the load inertia as sample inertia. And analyzing the inertia of the samples by means of a fuzzy clustering algorithm, and calculating the clustering center and the membership degree of each sample inertia about the clustering center.
2. Offline tuning of controller parameters
Defining a sample point corresponding to the clustering center inertia as a characteristic point, setting off line and recording controller parameters of three driving joints of the parallel mechanism when a reference point reaches the characteristic point.
3. Defining global variables
In order to increase the update speed of the controller parameters, the controller parameters of the characteristic points, the membership of the sample inertia to the clustering center (the sample points to the characteristic points) and the current coordinates of the moving platform reference points need to be read in real time. For this purpose, the above parameters are defined as global variables to improve the speed of reading and writing them.
4. Controller parameter online estimation and adjustment
In the process of executing the motion program, the current coordinate of the reference point is obtained by utilizing the position inverse solution module through calculation and is written into the temporary stack register as a global variable, the coordinate of the current reference point and the controller parameter at the characteristic point are read by the controller parameter estimation module, the sample point closest to the current reference point is judged, the membership degree of the sample point relative to each characteristic point is read, the driving joint controller parameter corresponding to the coordinate of the current reference point is estimated by utilizing the gravity center method, and the controller parameter in the control algorithm module is updated by utilizing the driving joint controller parameter.
The method comprises the following steps: fuzzy clustering analysis
The grid divides the working space of the reference point (hereinafter referred to as the reference point) of the movable platform of the parallel mechanism, and defines grid nodes as sample points. And calculating the load inertia of each driving joint by using the sample points according to the robot mathematical model, and defining the load inertia as sample inertia. And analyzing the inertia of the samples by means of a fuzzy clustering algorithm, and calculating the clustering center and the membership degree of each sample inertia about the clustering center.
(1) As shown in FIG. 2, the grid divides the working space of the reference point (hereinafter referred to as reference point) of the movable platform of the parallel mechanism, and defines grid nodes as sample points
According to the robot mathematical model, calculating the load inertia of each driving joint by using the sample points, and defining the jth sample inertia of the ith joint as I i,j 。
(2) According to the fuzzy clustering analysis algorithm, given the clustering class number N (N < N), the j sample inertia is calculated according to a formula in the first iteration with respect to the k (N)k 1,2, …, n) cluster centers, mu i,j,k,l And a new clustering center I i,k,l+1 ：
In the formula, α represents a blur index larger than 1, and is usually 2.
(3) Continuously updating mu according to the above formula i,j,k,l And I i,k,l+1 Until, | | I is satisfied at the first iteration i,k,l+1 -I i,k,l And | | | is less than or equal to epsilon. Here, epsilon represents the cluster center iteration precision.
Step two: offline tuning of controller parameters
Defining and clustering center inertias I i,k,l The corresponding sample point is a characteristic point
Setting off line and recording controller parameters of three driving joints of parallel mechanism when reference point reaches these characteristic points
K i,k ＝{K i,k,1 K i,k,2 K i,k,3 K i,k,4 K i,k,5 }
＝{K i,k,P K i,k,I K i,k,D K i,k,vff K i,k,aff }
In the formula, K i,k The i-th (i-1, 2,3) driving joint is shown as the k-th (k-1) driving joint , 2, …, n) characteristic points. K i,k,P ,K i,k,I ,K i,k,D ,K i,k,vff ,K i,k,aff Respectively representing the gains of the proportional, integral, derivative, velocity feedforward and acceleration feedforward controllers in the controller parameter sequence.
According to the symmetry of the mechanism and the working space, the characteristic points of the No. 2 and No. 3 driving joints are symmetrical about the midplane, and the parameter of the other joint can be obtained by setting the parameter of the characteristic point of one of the two joints.
Step three: defining global variables
In order to increase the update speed of the controller parameters, the controller parameters of the characteristic points, the membership of the sample inertia to the clustering center (the sample points to the characteristic points) and the current coordinates of the moving platform reference points need to be read in real time. For this purpose, the above parameters are defined as global variables to improve the speed of reading and writing them.
Step four: controller parameter online estimation and adjustment
In the process of executing the motion program, the current coordinate of the reference point is obtained by utilizing the position inverse solution module through calculation and is written into the temporary stack register as a global variable, the coordinate of the current reference point and the controller parameter at the characteristic point are read by the controller parameter estimation module, the sample point closest to the current reference point is judged, the membership degree of each characteristic point is read, the driving joint controller parameter corresponding to the coordinate of the current reference point is estimated by utilizing the gravity center method, and the controller parameter in the control algorithm module is updated by utilizing the driving joint controller parameter.
The barycenter method assumes that the controller parameters when the reference point is located at any point in the working space are weighted estimates of all the characteristic point controller parameters, the weight (i.e., membership) is determined by the sample point closest to the current reference point, and the serial number of the sample point is defined as m (j equals m). Accordingly, the controller parameter estimation algorithm may be expressed as
The calculation flow of the global adaptive adjustment of the controller parameters is shown in fig. 3. As can be seen from the figure, in the implementation process, controller parameters obtained by membership obtained by clustering analysis and offline tuning need to be written into corresponding registers as global variables for calling, and the controller parameters tuned in the middle of the working space need to be used as initial values. During the execution of the numerical control program, firstly, the NC codes are subjected to coarse interpolation according to the motion rule, the terminal pose after a coarse interpolation period is completed is calculated, then, the position inverse solution module is used for calculating the moving platform reference point coordinate and the driving joint instruction which correspond to the position inverse solution module, and the moving platform reference point coordinate is used as a global variable and is written into the temporary stack register. According toThe integral multiple of the coarse interpolation period, the controller parameter estimation module is called in an interruption mode, the reference point coordinates of the movable platform, the membership degree and the characteristic point controller parameters are read from the corresponding register, and the controller parameters of the ith driving joint corresponding to the reference point coordinates are estimated by using a gravity center method
And written as a global variable into the corresponding register. Calling control algorithm module to read from corresponding register
Using updated
Calculating the output command of the controller and holding it until the next estimation
And is not changed.
Claims (3)
1. A controller parameter universe self-adaptive adjusting system comprises a coarse interpolation module, a position inverse solution module and a control algorithm module, and is characterized by further comprising a controller parameter estimation module, wherein the controller parameter estimation module executes the following steps to adjust parameter matching load inertia variation, and the system comprises:
taking the membership degree obtained by clustering analysis and the characteristic point controller parameter obtained by off-line setting as a global variable to be written into a corresponding register for calling, and taking the controller parameter set in a working space as an initial value;
the coarse interpolation module performs coarse interpolation on the NC codes according to the motion rule and calculates the end pose after a coarse interpolation period is finished;
the position inverse solution module calculates the corresponding movable platform reference point coordinates and driving joint instructions, and writes the movable platform reference point coordinates into a temporary stack register as global variables;
the coarse interpolation module is used for performing interpolation according to integral multiples of a coarse interpolation periodCalling a controller parameter estimation module in an interrupt mode, reading the reference point coordinates of the movable platform, the membership degree and the characteristic point controller parameters from corresponding registers, and estimating the controller parameters of the ith driving joint corresponding to the reference point coordinates by using a gravity center method
And as the global variable writes into the corresponding register;
calling control algorithm module to read from corresponding register
Using updated
Calculating the output command of the controller and holding it until the next estimation
And is not changed.
2. The system of claim 1, wherein the controller parameters are adapted to the global domain
The parameters of the controller of the driving joint when the reference point of the movable platform is positioned at any point in the working space; the joint is subjected to weighted estimation on the controller parameters at all the characteristic points, the weight is determined by the sample point closest to the current reference point, and the serial number of the sample point is defined as m (j is m); the parameter estimation algorithm of the controller is
3. The system of claim 1, wherein the step of writing the global variable into the corresponding register for calling comprises:
(1) the grid divides the working space of the reference point of the movable platform of the parallel mechanism, and defines grid nodes as sample points
Calculating load inertia of each driving joint by using sample points according to a robot mathematical model, and defining j-th sample inertia of the ith joint as I i,j ；
(2) According to the fuzzy clustering analysis algorithm, given the number of clustering classes N (N < N), the membership degree mu of the j-th sample inertia about the k-th (k is 1,2, …, N) clustering center at the l-th iteration is calculated according to a formula i,j,k,l And a new clustering center I i,k,l+1 ；
Wherein α represents a blur index greater than 1;
(3) continuously updating mu according to the above formula i,j,k,l And I i,k,l+1 Until, | | I is satisfied at the first iteration i,k,l+1 -I i,k,l Less than or equal to epsilon; here, epsilon represents the cluster center iteration precision.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911022463.7A CN110788859B (en) | 2019-10-25 | 2019-10-25 | Controller parameter universe self-adaptive adjustment system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911022463.7A CN110788859B (en) | 2019-10-25 | 2019-10-25 | Controller parameter universe self-adaptive adjustment system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110788859A CN110788859A (en) | 2020-02-14 |
| CN110788859B true CN110788859B (en) | 2022-08-05 |
Family
ID=69441177
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911022463.7A Active CN110788859B (en) | 2019-10-25 | 2019-10-25 | Controller parameter universe self-adaptive adjustment system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110788859B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113359458B (en) * | 2021-06-22 | 2023-02-28 | 天津理工大学 | Fuzzy feedforward control method of high-speed parallel robot |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101971874A (en) * | 2010-09-09 | 2011-02-16 | 华南农业大学 | Variable humidifying system for fruit and vegetable refreshing and conveying and realizing method thereof |
| CN103353987A (en) * | 2013-06-14 | 2013-10-16 | 山东大学 | Superpixel segmentation method based on fuzzy theory |
| CN103473813A (en) * | 2013-09-18 | 2013-12-25 | 南京大学 | Automatic extracting method for three-dimensional model members |
| CN104985596A (en) * | 2015-07-09 | 2015-10-21 | 天津大学 | Five-freedom hybrid robot with multi-axis rotation brackets |
| CN107490958A (en) * | 2017-07-31 | 2017-12-19 | 天津大学 | A kind of Fuzzy Adaptive Control Scheme of series parallel robot in five degrees of freedom |
| CN107511830A (en) * | 2017-09-22 | 2017-12-26 | 天津大学 | A kind of series parallel robot in five degrees of freedom controller parameter adaptively adjusts implementation method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020103761A1 (en) * | 2001-01-27 | 2002-08-01 | Glassco David H.J. | Method and apparatus for managing and administering licensing of multi-function offering applications |
-
2019
- 2019-10-25 CN CN201911022463.7A patent/CN110788859B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101971874A (en) * | 2010-09-09 | 2011-02-16 | 华南农业大学 | Variable humidifying system for fruit and vegetable refreshing and conveying and realizing method thereof |
| CN103353987A (en) * | 2013-06-14 | 2013-10-16 | 山东大学 | Superpixel segmentation method based on fuzzy theory |
| CN103473813A (en) * | 2013-09-18 | 2013-12-25 | 南京大学 | Automatic extracting method for three-dimensional model members |
| CN104985596A (en) * | 2015-07-09 | 2015-10-21 | 天津大学 | Five-freedom hybrid robot with multi-axis rotation brackets |
| CN107490958A (en) * | 2017-07-31 | 2017-12-19 | 天津大学 | A kind of Fuzzy Adaptive Control Scheme of series parallel robot in five degrees of freedom |
| CN107511830A (en) * | 2017-09-22 | 2017-12-26 | 天津大学 | A kind of series parallel robot in five degrees of freedom controller parameter adaptively adjusts implementation method |
Non-Patent Citations (3)
| Title |
|---|
| Continuous Friction Feedforward Sliding Mode Controller for a TriMule Hybrid Robot;Xu Yang 等;《IEEE/ASME Transactions on mechatronics》;20180831;第23卷(第4期);全文 * |
| 模糊聚类分析及其应用研究;李新社 等;《网络安全技术与应用》;20140131;全文 * |
| 非线性系统的模糊建模与自适应控制及其应用;王永富;《中国博士学位论文全文数据库 信息科技辑》;20061130;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110788859A (en) | 2020-02-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Pane et al. | Reinforcement learning based compensation methods for robot manipulators | |
| Hu et al. | Deep GRU neural network prediction and feedforward compensation for precision multiaxis motion control systems | |
| CN111360830B (en) | Vibration control method of flexible mechanical arm based on cooperative tracking | |
| CN110221541B (en) | Feedforward coefficient acquisition device and method of feedforward controller in servo system | |
| EP0463934A1 (en) | Time delay controlled processes | |
| CN101114166A (en) | Contour outline control method for complicated track | |
| JPH0740204B2 (en) | Controller for multi-degree-of-freedom nonlinear mechanical system | |
| CN112417755A (en) | Master-slave mode surgical robot track prediction control method | |
| Kuang et al. | Simplified newton-based CEE and discrete-time fractional-order sliding-mode CEC | |
| CN111897224B (en) | Multi-agent formation control method based on actor-critic reinforcement learning and fuzzy logic | |
| CN111106772B (en) | Induction motor state strong tracking filtering estimation method including parameter tracking | |
| CN111702767A (en) | Manipulator impedance control method based on inversion fuzzy self-adaptation | |
| Li et al. | On-line self tuning of contouring control for high accuracy robot manipulators under various operations | |
| CN109062040B (en) | PID (proportion integration differentiation) predicting method based on system nesting optimization | |
| CN114169230A (en) | Robot dynamics parameter identification method | |
| CN107511830B (en) | Adaptive adjustment realization method for parameters of five-degree-of-freedom hybrid robot controller | |
| CN112445181A (en) | Machine learning device, servo control system, and machine learning method | |
| CN110788859B (en) | Controller parameter universe self-adaptive adjustment system | |
| Liu et al. | Robust control of dual-linear-motor-driven gantry stage for coordinated contouring tasks based on feed velocity | |
| CN114545864A (en) | Friction compensation control method of electro-hydraulic servo system | |
| CN114310851A (en) | Robot dragging-free teaching method for torque sensor | |
| CN113219825A (en) | Single-leg track tracking control method and system for quadruped robot | |
| CN116501061A (en) | Model predictive control path tracking method based on hierarchical optimization | |
| WO2020133881A1 (en) | Learning control method for mechanical apparatus, and mechanical apparatus learning control system having learning function | |
| US20030173928A1 (en) | Servo control method |
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 |