CN114167722A - Parallel robot tracking control method based on super-exponential convergence neural network - Google Patents

Parallel robot tracking control method based on super-exponential convergence neural network Download PDF

Info

Publication number
CN114167722A
CN114167722A CN202111421980.9A CN202111421980A CN114167722A CN 114167722 A CN114167722 A CN 114167722A CN 202111421980 A CN202111421980 A CN 202111421980A CN 114167722 A CN114167722 A CN 114167722A
Authority
CN
China
Prior art keywords
time
matlab
real
parallel robot
leg
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
CN202111421980.9A
Other languages
Chinese (zh)
Other versions
CN114167722B (en
Inventor
卓琳
陈德潮
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.)
Hangzhou Dianzi University
Original Assignee
Hangzhou Dianzi University
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 Hangzhou Dianzi University filed Critical Hangzhou Dianzi University
Priority to CN202111421980.9A priority Critical patent/CN114167722B/en
Publication of CN114167722A publication Critical patent/CN114167722A/en
Application granted granted Critical
Publication of CN114167722B publication Critical patent/CN114167722B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
    • G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Numerical Control (AREA)
  • Feedback Control In General (AREA)

Abstract

The invention discloses a tracking control method of a parallel robot based on a super-exponential convergence neural network, which comprises the following specific steps: a user initiates a communication request at an MATLAB end, establishes connection and inputs an expected path and an expected speed; MATLAB reads the real-time position, speed and acceleration of the parallel robot end effector in VREP, and reads the real-time acceleration of each supporting leg; calculating by utilizing real-time data at an MATLAB end to obtain a real-time control signal and a coefficient matrix updating signal; calculating the MATLAB end control signal and the updating signal to obtain an updated coefficient matrix and control information of each supporting leg; and inputting the obtained leg control information into the VREP and executing a dynamic simulation process. The invention utilizes real-time state and data of the parallel robot with uncertainty, realizes accurate and rapid tracking and control of the parallel robot through an online algorithm, and dynamically simulates the experimental process through a VREP simulation platform.

Description

Parallel robot tracking control method based on super-exponential convergence neural network
Technical Field
The invention relates to the field of parallel mechanical arm motion control algorithms, in particular to a method for realizing accurate tracking control of a parallel robot with uncertainty by utilizing an ZNN model and a super-exponential convergence factor.
Background
The parallel robot is a closed-loop mechanism which is driven in a parallel mode by connecting a movable platform and a fixed platform through more than two independent kinematic chains. The parallel robot has the excellent characteristics of no accumulated error, high precision, quick response, high bearing capacity and the like, and is widely applied to the industries of aviation, aerospace, seabed operation, underground mining and the like. In the case of the present invention, the stewart platform is a typical robot with six degrees of freedom.
The real-time tracking control of the parallel robot mainly controls the movement of an end effector installed on a mobile platform along a desired path in a robot working space through online calculation and guidance. The real-time control parallel robot has wide application in the work with intensive labor and high precision requirement, such as assembly, welding, paint spraying and the like.
Recurrent Neural Networks (RNNs) are often applied to solve real-time problems, while return-to-Zero Neural Networks (ZNN) can handle multiple state-dimensional problems and thus can be considered as systematic approaches to solving various time-varying problems, including time-varying tracking control problems. For the parallel robot with uncertainty, the real-time tracking control problem is solved by combining the super-exponential convergence factor with the return-to-zero neural network, and the SEC-ZNN has the advantages of both the super-exponential convergence factor and the return-to-zero neural network, so that the effect of accurate control is achieved while the robot is quickly converged.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a tracking control method of a parallel robot based on a super-exponential convergence neural network, which has the characteristic of super-exponential convergence.
In order to solve the technical problems, the invention is realized by the following technical scheme:
a tracking control method of a parallel robot based on a super-exponential convergence neural network comprises the following steps:
1) a user initiates a communication request at an MATLAB end to establish connection;
2) the user inputs the expected path r at the MATLAB endd(t) and desired speed
Figure BDA0003377827890000011
3) MATLAB reads real-time position r of parallel robot end effector in VREPa(t), real time velocity
Figure BDA0003377827890000012
And acceleration
Figure BDA0003377827890000013
Reading real-time acceleration of each leg
Figure BDA0003377827890000014
4) Calculating to obtain a real-time control signal at an MATLAB end by using the real-time data in the step 3) according to a kinetic equation
Figure BDA0003377827890000015
Sum coefficient matrix update signal
Figure BDA0003377827890000016
5) Calculating to obtain an updated coefficient matrix at the MATLAB terminal by using the control signal and the updating signal in the step 4)
Figure BDA0003377827890000017
And each leg control information l (t + Δ t).
6) Inputting the leg control information l (t + delta t) obtained by the MATLAB terminal into VREP and executing a dynamic simulation process.
The dynamic model of the step 4) comprises the following specific methods:
Figure BDA0003377827890000021
wherein the formula (1) is a landing leg control signal iterative formula,
Figure BDA0003377827890000022
the leg control signal is the control signal for the leg at time t,
Figure BDA0003377827890000023
for the matrix of coefficients at the time t,
Figure BDA0003377827890000024
for the desired velocity of the end effector at time t,
Figure BDA0003377827890000025
is to set a parameter for controlling convergence in the tracking process, Ψ (-) is an activation function matrix and each element therein is an odd function that monotonically increases, rd(t) and ra(t) represents the desired path and the actual path of the end effector at time t, respectively.
Equation (2) is an iterative equation of the coefficient matrix, in which
Figure BDA0003377827890000026
The coefficient matrix update signal is updated for time t,
Figure BDA0003377827890000027
for the acceleration of each leg at time t,
Figure BDA0003377827890000028
for the matrix of coefficients at the time t,
Figure BDA0003377827890000029
and
Figure BDA00033778278900000210
respectively representing the actual velocity and the actual acceleration of the end effector at time t,
Figure BDA00033778278900000211
is to set parameters for controlling convergence in the adaptation process, superscript
Figure BDA00033778278900000212
Representing a pseudo-inverse result of a vector or matrix.
The planning problem is solved by using an SEC-ZNN algorithm and real-time dynamic simulation is carried out at a VREP end, in the step 4), in an iteration process, an index factor exp (t) is used, so that a path is rapidly converged and is more accurately controlled, in the step 6), a VREP platform is used for real-time dynamic simulation, and the motion process of the robot is really shown.
The invention fully utilizes the real-time state and data of the parallel robot with uncertainty, realizes the accurate and rapid tracking and control of the parallel robot through the online algorithm, and dynamically simulates the experimental process through the VREP simulation platform, thereby providing more accurate reference for transferring the algorithm to the real object platform in the future.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a flow chart of the SEC-ZNN algorithm;
FIG. 3 is the result of an operation on the simulation platform.
Detailed Description
The invention is further described with reference to the following detailed description and accompanying drawings:
FIG. 1 shows a tracking control method of a parallel robot based on a super-exponential convergence neural network, which comprises the following steps:
1) performing tracking control on the parallel robot with uncertainty by using an SEC-ZNN algorithm;
2) the MATLAB is used as a remote control end, the VREP platform is used as an actual simulation end, and the simulation parallel robot carries out tracking control in the actual motion process with uncertainty;
therefore, the process of the invention is as follows:
1) and initiating a connection establishment request at the MATLAB end, and initiating a VREP end to establish connection corresponding to the request.
2) And sending a control instruction by the MATLAB terminal to read and set related information such as the initial robot posture, the control joint handle, the communication time interval and the like of the VREP terminal, and generating a simplified robot model at the MATLAB terminal.
3) And when the iteration is not finished and the connection is not disconnected, calling a synchronous trigger and reading the dynamic parameters.
4) And calculating by using an SEC-ZNN algorithm to obtain the leg control information. The SEC-ZNN algorithm comprises the following specific contents:
according to the kinetic model:
Figure BDA0003377827890000031
and calculating the iteration quantity of the leg length and the coefficient matrix corresponding to the time t. For the obtained iteration quantity, according to an iteration formula:
Figure BDA0003377827890000032
and iteratively updating the leg length control quantity and the coefficient matrix.
5) The MATLAB end sends out a control signal, and the VREP end receives the signal and then utilizes the PID controller to dynamically simulate the leg movement.
Fig. 3 shows the tracking control result of a simulation experiment. The track in the figure is the motion track of the end effector, and the end effector is arranged on the upper surface of the Stewart parallel robot.
The above-described embodiments of the present invention do not limit the scope of the present invention. Any modification, equivalent replacement, and improvement made within the spirit and scope of the present invention shall be included in the protection scope of the claims of the present invention.

Claims (2)

1. A tracking control method of a parallel robot based on a super-exponential convergence neural network is characterized by comprising the following steps:
1) a user initiates a communication request at an MATLAB end to establish connection;
2) the user inputs the expected path r at the MATLAB endd(t) and desired speed
Figure FDA0003377827880000011
3) MATLAB reads real-time position r of parallel robot end effector in VREPa(t), real time velocity
Figure FDA0003377827880000012
And acceleration
Figure FDA0003377827880000013
Reading real-time acceleration of each leg
Figure FDA0003377827880000014
4) Calculating to obtain a real-time control signal at an MATLAB end by using the real-time data in the step 3) according to a kinetic equation
Figure FDA0003377827880000015
Sum coefficient matrix update signal
Figure FDA0003377827880000016
The specific method for calculating according to the kinetic equation comprises the following steps:
Figure FDA0003377827880000017
wherein the formula (1) is a landing leg control signal iterative formula,
Figure FDA0003377827880000018
the leg control signal is the control signal for the leg at time t,
Figure FDA0003377827880000019
for the matrix of coefficients at the time t,
Figure FDA00033778278800000110
for the desired velocity of the end effector at time t,
Figure FDA00033778278800000111
is to set a parameter for controlling convergence in the tracking process, Ψ (-) is an activation function matrix and each element therein is an odd function that monotonically increases, rd(t) and ra(t) respectively representing the expected path and the actual path of the end effector at the time t, exp (t) being an exponential factor;
equation (2) is an iterative equation of the coefficient matrix, in which
Figure FDA00033778278800000112
The coefficient matrix update signal is updated for time t,
Figure FDA00033778278800000113
for the acceleration of each leg at time t,
Figure FDA00033778278800000114
for the matrix of coefficients at the time t,
Figure FDA00033778278800000115
and
Figure FDA00033778278800000116
respectively representing the actual velocity and the actual acceleration of the end effector at time t,
Figure FDA00033778278800000117
is to set parameters for controlling convergence in the adaptation process, superscript
Figure FDA00033778278800000120
A pseudo-inverse matrix result representing a vector or matrix;
5) calculating to obtain an updated coefficient matrix at the MATLAB terminal by using the control signal and the updating signal in the step 4)
Figure FDA00033778278800000118
And each leg control information iota (t + delta t);
6) inputting the leg control information l (t + delta t) obtained by the MATLAB terminal into VREP and executing a dynamic simulation process.
2. The method for tracking and controlling the parallel robot based on the super-exponential convergent neural network according to claim 1, wherein the method comprises the following steps: the step 5) is specifically as follows:
Figure FDA00033778278800000119
CN202111421980.9A 2021-11-26 2021-11-26 Parallel robot tracking control method based on super-exponential convergence neural network Active CN114167722B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111421980.9A CN114167722B (en) 2021-11-26 2021-11-26 Parallel robot tracking control method based on super-exponential convergence neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111421980.9A CN114167722B (en) 2021-11-26 2021-11-26 Parallel robot tracking control method based on super-exponential convergence neural network

Publications (2)

Publication Number Publication Date
CN114167722A true CN114167722A (en) 2022-03-11
CN114167722B CN114167722B (en) 2024-03-29

Family

ID=80481126

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111421980.9A Active CN114167722B (en) 2021-11-26 2021-11-26 Parallel robot tracking control method based on super-exponential convergence neural network

Country Status (1)

Country Link
CN (1) CN114167722B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018176854A1 (en) * 2017-03-27 2018-10-04 华南理工大学 Method for programming repeating motion of redundant robotic arm
CN110134013A (en) * 2019-05-08 2019-08-16 杭州电子科技大学 The parallel mechanical arm finite time convergence control motion control method of external disturbance can be fought
CN111443712A (en) * 2020-03-30 2020-07-24 杭州电子科技大学 Three-dimensional path planning method based on longicorn group search algorithm
CN111872933A (en) * 2019-11-25 2020-11-03 浙江大学宁波理工学院 SCARA robot trajectory tracking control method based on improved quadratic iterative learning control
CN111975768A (en) * 2020-07-08 2020-11-24 华南理工大学 Mechanical arm motion planning method based on fixed parameter neural network
CN112894819A (en) * 2021-01-29 2021-06-04 佛山树客智能机器人科技有限公司 Robot dynamic motion control method and device based on double neural networks

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018176854A1 (en) * 2017-03-27 2018-10-04 华南理工大学 Method for programming repeating motion of redundant robotic arm
CN110134013A (en) * 2019-05-08 2019-08-16 杭州电子科技大学 The parallel mechanical arm finite time convergence control motion control method of external disturbance can be fought
CN111872933A (en) * 2019-11-25 2020-11-03 浙江大学宁波理工学院 SCARA robot trajectory tracking control method based on improved quadratic iterative learning control
CN111443712A (en) * 2020-03-30 2020-07-24 杭州电子科技大学 Three-dimensional path planning method based on longicorn group search algorithm
CN111975768A (en) * 2020-07-08 2020-11-24 华南理工大学 Mechanical arm motion planning method based on fixed parameter neural network
CN112894819A (en) * 2021-01-29 2021-06-04 佛山树客智能机器人科技有限公司 Robot dynamic motion control method and device based on double neural networks

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
彭志文;高宏力;梁超;文刚;: "RBF神经网络补偿的并联机器人控制研究", 机械设计与制造, no. 03, 8 March 2018 (2018-03-08) *

Also Published As

Publication number Publication date
CN114167722B (en) 2024-03-29

Similar Documents

Publication Publication Date Title
CN107490965B (en) Multi-constraint trajectory planning method for space free floating mechanical arm
CN106945020B (en) A kind of space double mechanical arms system motion control method for coordinating
Jung et al. Force tracking impedance control of robot manipulators under unknown environment
CN112959325B (en) High-precision control method for collaborative machining of double-moving mechanical arm in large scene
Wen et al. Elman fuzzy adaptive control for obstacle avoidance of mobile robots using hybrid force/position incorporation
CN107962566B (en) Repetitive motion planning method for mobile mechanical arm
US11498213B2 (en) Robot joint space graph path planning and move execution
Tan et al. Minimum time trajectory planner for the discrete dynamic robot model with dynamic constraints
CN111168684B (en) On-orbit assembly sequence planning method for large-scale spatial structure
CN105598968B (en) A kind of motion planning and control method of parallel mechanical arm
CN110076783A (en) A kind of plane activation lacking mechanical arm position control method based on fuzzy control and liapunov function
CN108393886B (en) Distributed multi-mobile manipulator cooperative carrying method capable of optimizing energy and operation degree
CN112091829B (en) Sand blasting and rust removing parallel robot friction force mutation compensating fuzzy self-adaptive sliding mode control method
Dehio et al. Modeling and control of multi-arm and multi-leg robots: Compensating for object dynamics during grasping
CN110598285A (en) Method and device for solving inverse kinematics of manipulator track and storage medium
CN108908347A (en) One kind is towards redundancy mobile mechanical arm error-tolerance type repetitive motion planning method
CN101143443A (en) Six-freedom parallel robot decoupling method
CN109940615B (en) Terminal state network optimization method for synchronous repeated motion planning of double-arm manipulator
CN109623812B (en) Mechanical arm trajectory planning method considering spacecraft body attitude motion
CN114942593A (en) Mechanical arm self-adaptive sliding mode control method based on disturbance observer compensation
CN114571444A (en) Q-learning-based impedance control method for slagging-off robot
Kang et al. Kinematic path‐tracking of mobile robot using iterative learning control
CN114167722A (en) Parallel robot tracking control method based on super-exponential convergence neural network
JP2020049554A (en) Track formation method, track formation device, and robot system
CN113103262A (en) Robot control device and method for controlling robot

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