CN-122008257-A - Textile mechanical arm vision servo track tracking control method and system based on fuzzy observer

CN122008257ACN 122008257 ACN122008257 ACN 122008257ACN-122008257-A

Abstract

The invention belongs to the technical field of tracking control of textile mechanical arm tracks, and discloses a textile mechanical arm visual servo track tracking control method and system based on a fuzzy observer. Aiming at the problems that visual speed information is difficult to acquire and the uncertainty of a system model is high, the invention designs a fuzzy observer to acquire a visual speed estimated value, and approximates to unknown nonlinear dynamics by using a fuzzy logic system. In addition, the invention solves the problem of calculation complexity of the mechanical arm visual servo system by using the instruction filtering technology, eliminates adverse effects caused by filtering errors by introducing an error compensation mechanism, improves the control effect of the visual servo system, and compensates negative effects of an input dead zone on the system performance. The invention can effectively solve the problems that visual speed information is difficult to obtain, the calculation complexity exists when a controller is designed, and the like, has good control effect, and is suitable for scenes with high control precision requirements of production robots in textile industry.

Inventors

YU JINPENG
LIU JIAPENG
TIAN MINGWEI
Lv Youlong
WANG XIAOLING
DONG YUANBAO
GU TIANHAO
CHENG SHUAI
LI WEI
LIU BINGWEN
LIU SONG
TIAN XINCHENG
MA YUMEI
CHEN YUSHUO
ZHAO LIN
WANG BAOFANG
NI JINGDA
ZHANG WENXIN
WANG RONGWU
FU CHENG

Assignees

青岛大学

Dates

Publication Date: 20260512
Application Date: 20260414

Claims (10)

1. The textile mechanical arm vision servo track tracking control method based on the fuzzy observer is characterized by comprising the following steps of: step 1, establishing a camera model and a dynamic model of a rigid multi-joint textile mechanical arm with uncertainty and an input dead zone; Obtaining the relation between the projection coordinates of the characteristic points on the image plane and the coordinates of the characteristic points under the mechanical arm base coordinate system according to the visual projection model, obtaining the relation between the characteristic point speeds and the mechanical arm joint angular speeds through derivation, and obtaining a rigid multi-joint textile mechanical arm dynamics model under the image space through coordinate transformation; And 2, performing control design on the mechanical arm visual servo system to obtain a mechanical arm fuzzy self-adaptive output feedback controller under the condition of unknown visual speed information, wherein the specific design process of the controller is as follows: The method comprises the steps of designing a fuzzy observer to obtain an estimated value of visual speed, processing unknown nonlinear dynamics in a mechanical arm visual servo system by using a fuzzy logic system, solving the problem of computational complexity by adopting an instruction filtering technology, introducing an error compensation mechanism to eliminate the influence caused by a filtering error, considering the influence of an input dead zone on the system performance, and finally designing a real control law; and 3, realizing track tracking control on the rigid multi-joint spinning mechanical arm system by using a mechanical arm fuzzy self-adaptive output feedback controller.
2. The control method for tracking visual servo tracks of a textile mechanical arm based on a fuzzy observer according to claim 1, wherein the step 1 is specifically: Establishing a mechanical arm end effector coordinate system, a mechanical arm base coordinate system and a camera coordinate system, wherein the mechanical arm end effector coordinate system, the mechanical arm base coordinate system and the camera coordinate system are used for representing the movement of the mechanical arm end and the relation between the mechanical arm base coordinate system and the camera coordinate system; Homogeneous coordinates of projection of feature points on image plane The method comprises the following steps: (1) Wherein, the Representing feature points on an image projection plane The value of the axis is set to, Representing feature points on an image projection plane The value of the axis is set to, The depth of the feature points is indicated, The joint angle of the mechanical arm is represented, Is the homogeneous coordinate of the feature point under the coordinate system of the mechanical arm end effector, Is a homogeneous transformation matrix of external parameters of the camera, namely from a mechanical arm end effector coordinate system to a camera coordinate system, Is a homogeneous transformation matrix from a mechanical arm base coordinate system to a mechanical arm end effector coordinate system, Is the homogeneous coordinate of the feature point under the mechanical arm base coordinate system, Representing a matrix determined by camera internal parameters; Using Representing a perspective projection matrix, the expression of which is as follows: (2) the expression of (2) is as follows: (3) Wherein, the Representing perspective projection matrices Is a third row vector of (a); Deriving the formula (1) to obtain a speed relation shown in the formula (4): (4) Wherein, the ; Wherein the method comprises the steps of Representing perspective projection matrices Is arranged in the first row of vectors of (a), Representing perspective projection matrices Is a second row vector of (a); Indicating the joint angular velocity of the mechanical arm; Is a composite jacobian matrix; Pseudo-inverse of (2) The method comprises the following steps: ; One with input dead zone, degree of freedom of The dynamic model of the rigid multi-joint textile mechanical arm is as follows: (5) Wherein, the Representing the inertial matrix of the mechanical arm, Representing the coriolis force and centripetal force matrices of the robotic arm, Representing the gravity vector of the mechanical arm, I.e. ; The input value representing the arm joint torque with the input dead zone characteristic is the output value of the arm controller The dead zone characteristic has an output value of And the input and output of the dead zone satisfy the following relationship: (6) Wherein the method comprises the steps of And A left slope and a right slope representing the dead zone characteristic, And A left critical point and a right critical point representing the dead zone characteristic, Output vector representing a robotic arm controller Is the first of (2) An element; Known parameters 、 Are all normal values with limits, and the left and right slopes of the dead zone are the same, i.e ; And Unknown and bounded, then the dead zone characteristic, equation (6), translates into: (7) Wherein the method comprises the steps of And The definition is as follows: (8) (9) Due to parameters 、、 And Are all bounded to obtain Is also bounded, i.e , Represents an arbitrary positive constant; (10) Wherein, the , , , Representation of Is the first of (2) The number of elements to be added to the composition, Representation of Is the first of (2) The number of elements to be added to the composition, A diagonal matrix is represented and, Representation of Is the first of (2) An element; Through the following coordinate transformation: (11) (12) substituting the formula (11) and the formula (12) into the formula (5) to obtain a rigid multi-joint spinning mechanical arm dynamics model which is as follows: (13) Wherein, the , , ; ; , Is that ; 、、 Respectively defining the position, speed and acceleration of the characteristic point And Wherein for convenience of description, the following variable substitutions are made, defined Is that , Is that , Is that , Is that ; Then equation (13) is rewritten as: (14) 。
3. the control method for tracking the visual servo track of the textile mechanical arm based on the fuzzy observer according to claim 2, wherein the step 2 is specifically: the construction instruction filter is as follows: ; Wherein, the , 、 As an output signal of the instruction filter, 、 Is a constant; Is an input signal to the instruction filter; 、 Respectively representing the internal state variables and the derivative values of the filters; initial state of instruction filter , ; Wherein the method comprises the steps of Is that Is used for the initial value of (a), Is that Is used for the initial value of (a), Is that Is set to an initial value of (1); When (when) When the input signal satisfies 、 Wherein 、 Is constant, then to arbitrary Exists in the presence of And So that And (2) and 、、 Are all bounded; Wherein the method comprises the steps of , ; The construction of the fuzzy observer is as follows: ; Wherein, the Is a state estimation value; to estimate the error, it is defined as: ; Is a function of the positive design parameters, ; Is a vector estimation value; Representation of Is a function of the estimated value of (2); Representation of Is a function of the estimated value of (2); The pair is defined in a tight set Continuous function in (a) There is always a fuzzy logic system Such that: Wherein, the method comprises the steps of, The weight matrix is represented by a matrix of weights, , Representing a fuzzy basis function; Is an approximation error, and , Representing a positive number; For functions Any positive number The following inequality holds: ; Wherein, the Is a vector two-norm; this gives: (15) Wherein the method comprises the steps of , , , And then obtaining: ; Representing an approximation error; having a symmetrical matrix There will always be a symmetric matrix Satisfies the following conditions ; Defining an error variable: (16) Wherein, the 、 Representing the variable of the error and, Which represents the preset desired signal(s), Representing the filter output signal; the compensation tracking error variable is defined as: (17) Wherein, the 、 Representing the compensation of the tracking error variable, 、 Compensating the signal for the filtered error; Selecting Function of The method comprises the following steps: ; For a pair of And (3) deriving to obtain: (18) according to the young's inequality: (19) (20) Substituting formulas (19), (20) into formula (18) to obtain: (21) Selecting Function of The method comprises the following steps: ; For a pair of And (3) deriving to obtain: (22) according to the young's inequality: (23) Designing virtual control laws And error compensation signal The method comprises the following steps: (24) Wherein, the Gain is controlled for the system, and Substituting the formulas (23) and (24) into the formula (22) to obtain: (25) Selecting Function of The method comprises the following steps: ; Wherein, the Is a positive definite matrix of the matrix and the matrix, Representing the weight matrix The column vector estimation error is used to estimate the column vector, The dimension number of the vector is the fuzzy basis function; For a pair of And (3) deriving to obtain: (26) Wherein, the Representing the compensation of the tracking error signal, Representing an adaptation law; according to the young's inequality: (27) design of true control law Compensating signal Self-adaptive law The method comprises the following steps: (28) Wherein, the Gain is controlled for the system, and ; Is a positive constant; substituting formulas (27), (28) into formula (26) yields: (29) Wherein the method comprises the steps of Representing the weight matrix A column vector; The fuzzy self-adaptive output feedback controller of the mechanical arm is designed as follows: 。
4. The method for controlling the visual servo track tracking of the textile mechanical arm based on the fuzzy observer according to claim 1, wherein in the step 2, after the design of the fuzzy self-adaptive output feedback controller of the mechanical arm is completed, stability analysis is performed on the rigid multi-joint textile mechanical arm system controlled by the fuzzy self-adaptive output feedback controller of the mechanical arm.
5. The control method for tracking the visual servo track of the textile mechanical arm based on the fuzzy observer according to claim 4, wherein in the step 2, the specific process of performing the stability analysis is as follows: Selecting Function of : (30) Deriving equation (30), and substituting equations (24), (28), and (29) to obtain: (31) Wherein: ; Wherein, the 、 Respectively representing the minimum value and the maximum value of the matrix trace; Obtained from equation (31): (32) Wherein, the Representation of At the position of Time of day The function value of the function value, Representation of At the position of Time of day The function value of the function value, ; Equation (32) shows And All belong to a tight set , Representing the weight matrix estimation error, so that all signals of the closed loop system are bounded, an error compensation signal of the instruction filter Satisfy the following requirements Can be obtained Is bounded due to And (2) and Is bounded, then tracking error Is also bounded.
6. The method for controlling the visual servo track tracking of the textile mechanical arm based on the fuzzy observer according to claim 1, wherein the step 3 is specifically: According to the relation between the characteristic point speed and the mechanical arm joint angular speed, mapping the mechanical arm dynamic model to an image space through coordinate transformation to obtain the mechanical arm dynamic model in the image space, aiming at the problems that visual speed information is difficult to obtain and the mechanical arm system contains unknown nonlinear dynamics, designing a fuzzy observer to obtain an estimated value of the visual speed information, approaching a mathematical model of the rigid multi-joint textile mechanical arm system by using a fuzzy logic system, and adopting an instruction filtering technology and introducing an error compensation mechanism to solve the problem of computational complexity and eliminate filtering errors, thereby realizing track tracking control of the rigid multi-joint textile mechanical arm system.
7. The fuzzy observer-based textile mechanical arm vision servo track tracking control method according to claim 1, wherein in the step 1, the built rigid multi-joint textile mechanical arm dynamics model is as follows A mathematical model of a dynamic system of the joint rigid mechanical arm, wherein, The number of joints, i.e., the degrees of freedom, of the robotic arm system.
8. The textile mechanical arm vision servo track tracking control system based on the fuzzy observer is characterized by comprising the following modules: the model building module is used for building a camera model and a dynamic model of the rigid multi-joint textile mechanical arm with uncertainty and input dead zone; Obtaining the relation between the projection coordinates of the characteristic points on the image plane and the coordinates of the characteristic points under the mechanical arm base coordinate system according to the visual projection model, obtaining the relation between the characteristic point speeds and the mechanical arm joint angular speeds through derivation, and obtaining a rigid multi-joint textile mechanical arm dynamics model under the image space through coordinate transformation; The controller design module is used for controlling and designing the mechanical arm visual servo system to obtain the mechanical arm fuzzy self-adaptive output feedback controller under the condition of unknown visual speed information, wherein the specific design process of the controller is as follows: The method comprises the steps of designing a fuzzy observer to obtain an estimated value of visual speed, processing unknown nonlinear dynamics in a mechanical arm visual servo system by using a fuzzy logic system, solving the problem of computational complexity by adopting an instruction filtering technology, introducing an error compensation mechanism to eliminate the influence caused by a filtering error, considering the influence of an input dead zone on the system performance, and finally designing a real control law; And the track tracking control module is used for realizing track tracking control on the rigid multi-joint textile mechanical arm system by utilizing the mechanical arm fuzzy self-adaptive output feedback controller.
9. Computer device comprising a memory and one or more processors, characterized in that executable code is stored in the memory, which when executed by the processors is adapted to carry out the steps of the fuzzy observer based textile mechanical arm vision servo track tracking control method according to any of the previous claims 1 to 7.
10. A computer readable storage medium having a program stored thereon, wherein the program, when executed by a processor, is configured to implement the steps of the fuzzy observer based textile mechanical arm vision servo track tracking control method according to any one of claims 1 to 7.

Description

Textile mechanical arm vision servo track tracking control method and system based on fuzzy observer Technical Field The invention belongs to the technical field of tracking control of textile mechanical arm tracks, and particularly relates to a method and a system for tracking control of a textile mechanical arm visual servo track based on a fuzzy observer. Background China is a textile export and production country, however, for large production workshops, the traditional manual operation production mode can not meet the increasing production demands gradually. Therefore, it is necessary to realize intelligent development of textile equipment to improve productivity. The multi-degree-of-freedom textile mechanical arm can realize free grabbing of textile production raw materials, so that manpower resources are greatly saved, the safety of a production process is improved, and the multi-degree-of-freedom textile mechanical arm becomes important production equipment of a textile workshop gradually. However, when in an unknown working environment, the robotic arm is often not able to accommodate production requirements due to lack of sensing capabilities for the surrounding environment. To address this difficulty, a mechanical arm vision servo control scheme incorporating vision perception and control techniques has been developed. The visual servo control can sense the target in real time by using the image information as a real-time feedback signal, so that the accurate positioning and flexible grabbing of the moving object in the complex environment are realized. Visual servo control is an important control means in the current industrial field, and has great development potential. At present, the control method of the mechanical arm in the textile field mainly adopts a control mode of combining proportional-integral-derivative control and a gravity term. Such control methods often require accurate mechanical arm dynamics models. However, when the robotic arm is in an unknown operating environment, it is often difficult to build an accurate mechanical arm dynamics model. Research finds that fuzzy logic systems can effectively handle uncertainty and nonlinear terms in the system, and are widely applied to nonlinear system control at present. On the other hand, the back-stepping method is widely used for solving the control problem of the rigid multi-joint mechanical arm because of the characteristic of easy combination with fuzzy control and self-adaptive control, but when designing the controller by the back-stepping method, the virtual control function needs to be continuously derived, so that the calculated amount is greatly increased. In the current research, most vision servo controllers are required to acquire a vision speed measurement value, however, in actual situations, the vision speed is difficult to acquire, and most of the existing measurement schemes acquire the vision speed value by taking a derivative of image position information, and the measurement schemes often contain large measurement noise and reduce the performance of the system. In addition, the dead zone is a typical nonlinear model, and dead zone effects can have serious impact on system performance and even possibly destroy system stability. Disclosure of Invention The invention aims to provide a fuzzy observer-based textile mechanical arm visual servo track tracking control method so as to realize accurate track tracking control on a rigid multi-joint textile mechanical arm system. In order to achieve the above purpose, the present invention adopts the following technical scheme: a textile mechanical arm vision servo track tracking control method based on a fuzzy observer comprises the following steps: step 1, establishing a camera model and a dynamic model of a rigid multi-joint textile mechanical arm with uncertainty and an input dead zone; Obtaining the relation between the projection coordinates of the characteristic points on the image plane and the coordinates of the characteristic points under the mechanical arm base coordinate system according to the visual projection model, obtaining the relation between the characteristic point speeds and the mechanical arm joint angular speeds through derivation, and obtaining a rigid multi-joint textile mechanical arm dynamics model under the image space through coordinate transformation; And 2, performing control design on the mechanical arm visual servo system to obtain a mechanical arm fuzzy self-adaptive output feedback controller under the condition of unknown visual speed information, wherein the specific design process of the controller is as follows: The method comprises the steps of designing a fuzzy observer to obtain an estimated value of visual speed, processing unknown nonlinear dynamics in a mechanical arm visual servo system by using a fuzzy logic system, solving the problem of computational complexity by adopting an instruction filtering technology, introducing an er