CN113359470B - Designated transient time stability control method for restraining structural change of rehabilitation robot - Google Patents
Designated transient time stability control method for restraining structural change of rehabilitation robot Download PDFInfo
- Publication number
- CN113359470B CN113359470B CN202110756756.9A CN202110756756A CN113359470B CN 113359470 B CN113359470 B CN 113359470B CN 202110756756 A CN202110756756 A CN 202110756756A CN 113359470 B CN113359470 B CN 113359470B
- Authority
- CN
- China
- Prior art keywords
- robot
- hidden layer
- transient time
- rehabilitation
- rehabilitation training
- 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
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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The invention discloses a designated transient time stability control method for inhibiting structural change of a rehabilitation robot. The method is characterized in that: utilizing a dynamic model of the rehabilitation training robot to split physical quantities of structural changes related to the eccentricity and the eccentric angle, and establishing a dynamic model for describing the uncertainty of the structure of the rehabilitation training robot; providing a random configuration network (Finite Stochastic Configuration Networks, FSCN) with a limited hidden layer nodes, taking a track tracking error as network input, obtaining structure uncertainty change estimation, and constructing a dynamic model of the rehabilitation training robot with the structure uncertainty estimation; the method for stabilizing and controlling the designated transient time is provided, and the influence of structural change on the tracking performance of the man-machine system is restrained by respectively designing the transient time controller and the steady state time controller, so that the stable tracking control of the designated transient time is realized.
Description
Technical field:
the invention relates to the field of control of wheeled rehabilitation robots, in particular to a control method of a wheeled lower limb rehabilitation robot.
The background technology is as follows:
with the increase of the elderly population, the old people have walking dysfunction caused by diseases and natural aging, and if rehabilitation training is not performed in time, the leg muscle strength and balance ability are recovered, so that the daily life of the old people is seriously affected. Along with the application of the rehabilitation walking robot in places such as rehabilitation centers, nursing homes and the like, the walking rehabilitation problem of the old is solved in time. However, in practical application, the pose of the trainer is changed to cause the misalignment of the centers of the person and the robot, so that the eccentricity and the eccentric angle are generated by the man-machine system, and the structure of the man-machine system is subjected to uncertain change, which seriously affects the tracking precision of the robot on the designated motion track of the doctor, so that not only is the rehabilitation effect of the trainer not ideal, but also the collision danger of the robot is caused by the overlarge track tracking error, and the safety of the man-machine system is affected. Therefore, the method for improving the tracking precision and the safety of the rehabilitation robot has important significance for solving the structural change problem of the man-machine system.
In recent years, trace tracking control of rehabilitation walking robots has many research results, but none of the results can solve the problem of uncertain change of the structure of a man-machine system. If the robot cannot adapt to the structural change of the system in walking training, not only the tracking precision is affected, but also the robot collides with surrounding objects due to excessive track tracking errors, so that the safety of a trainer is threatened. Up to now, there is no estimation method for uncertain changes of a man-machine system structure and a stabilization control method for specified transient time, and the invention establishes an estimation model for uncertain changes of the structure based on a new view angle by using an FSCN method, and provides a stabilization tracking control method for specified transient time for inhibiting the changes of the structure, which has important significance for improving the rapid stabilization tracking of a robot.
The invention comprises the following steps:
the invention aims to:
in order to solve the problems, the invention provides a specified transient time stability control method for inhibiting structural change of a rehabilitation robot.
The technical scheme is as follows:
the invention is realized by the following technical scheme:
a specified transient time stability control method for restraining structural change of a rehabilitation robot is characterized in that:
1) Utilizing a dynamic model of the rehabilitation training robot to split physical quantities of structural changes related to the eccentricity and the eccentric angle, and establishing a dynamic model for describing the uncertainty of the structure of the rehabilitation training robot;
2) Providing a random configuration network FSCN with a limited hidden layer nodes, taking a track tracking error as network input, obtaining structure uncertainty change estimation, and constructing a dynamic model of the rehabilitation training robot with the structure uncertainty estimation;
3) The method for stabilizing and controlling the designated transient time is provided, and the influence of structural change on the tracking performance of the man-machine system is restrained by respectively designing the transient time controller and the steady state time controller, so that the stable tracking control of the designated transient time is realized.
The method comprises the following steps:
step one), a dynamic model of the rehabilitation training robot is utilized to split physical quantities of structural changes related to eccentricity and eccentric angles, and a dynamic model describing that the rehabilitation training robot has an uncertain structure is established. The method is characterized in that: the kinetic model of the system is described below
Wherein the method comprises the steps of
X (t) represents the actual motion trail of the rehabilitation training robot, u (t) represents the control input force, M represents the mass of the robot, M represents the mass of the rehabilitation person, I 0 Represent moment of inertia, M 0 ，K(θ)，B (θ) is a coefficient matrix. θ represents the angle between the horizontal axis and the connection between the robot center and the first wheel center, i.e., θ=θ 1 As known from the structure of the rehabilitation training robot,θ 3 ＝θ+π，/>l i representing the distance of the center of gravity of the system from the center of each wheel, r 0 Represents the distance from center to center of gravity, phi i Representing the x' axis and the l corresponding to each wheel i Included angles (i=1, 2,3, 4).
From the dynamics model (1) of the system, the physical quantity generating uncertain change of the system structure is eccentricity and eccentric angle, and the related physical quantity is split and M is calculated 0 K (θ) is represented as M 0 K(θ)＝M 1 +ΔM 1 Wherein M is 1 Consists of rehabilitation training robot mass, rehabilitation person mass and moment of inertia, and delta M 1 Representing a physical quantity of structural change, an
At the same time, coefficient matrixThe eccentric distance and the eccentric angle of the structural change are used as physical quantities of the structural change; while coefficient matrix B (θ) contains variable λ resulting from structural changes i Decompose it into B (θ) =b 1 (θ)+ΔB 1 (θ) wherein
Wherein D is a steady physical quantity and represents the distance from the center of the man-machine system to the center of each wheel; physical quantity affecting structural change in the split dynamics model (1) is divided, and a dynamics model for describing the uncertainty of the structure of the rehabilitation training robot is established as follows
Wherein the method comprises the steps ofPhysical quantities representing uncertain changes in the structure of the rehabilitation training robot.
Step two), a random configuration network FSCN with a limited hidden layer nodes is provided, a track tracking error is taken as network input, the structure uncertainty change estimation is obtained, and a dynamics model with the structure uncertainty estimation of the rehabilitation training robot is constructed, and the method is characterized in that: based on the rehabilitation training robot dynamics model (2) with uncertain structure, the method comprises the following steps ofThe system state equation is obtained as follows:
let the appointed motion trail of rehabilitation robot be x d (t)，ε 1 (t)＝x 1 (t)-x d (t) is the track-following error of the system,is the speed tracking error of the system.
By epsilon 1 (t) is a network input, S (p, q, r) 0 ) For network output, a FSCN estimation model with uncertain structure is established. When the node number of the hidden layer of the FSCN model is L-1, the output of the model is made to beAnd f 0 =0, a gaussian function is selected as the basis function, at which time the structural change estimation error is e L-1 ＝S(p，q,r 0 )-f L-1 (ε 1 (t)) when the error e L-1 Adding a node to the hidden layer when the expected requirement is not met, and randomly generating a meeting supervision mechanism delta L Is input weight w of (2) L And bias b L 。
Wherein e L-1 ＝[e L-1，1 ，e L-1,2 ,e L-1,3 ] T Is the estimated error when the node number of the hidden layer is L-1, beta j ＝[β j，1 ，β j，2 ，β j,3 ] T Output weight, w, of j hidden layer node j ＝[w j,1 ,w j,2 ,w j,3 ] T And b j The input weights and offsets for the j-th hidden layer node,and L is the number of hidden layer nodes for outputting the j-th hidden layer node.
The FSCN supervision mechanism is designed as follows:
wherein alpha is q For the adjustable parameter (q=1, 2, 3), parameter σ<0。
Next, the FSCN under supervision mechanism δ of a limited number of hidden layer nodes is further described L The structure uncertainty estimate may be made to trend toward zero. Due to i e L || 2 Is positively determined to obtain
Since the expression (5) has a discrete characteristic with respect to L, a difference operation is performed thereon to obtain
Due to beta L,q A constant can be obtained after calculation, and then is set to be constantNumber alpha q Beta is made to be L,q ＝α q Further, from the formula (6)
From (4), it can be seen that
Let L be max The number of hidden layer nodes when the structure uncertainty estimation tends to zero is known by discrete Lyapunov criteriaFrom equation (8), it can be seen that for any hidden layer node:
respectively performing accumulation operation on two ends of the inequality of (9) to obtain
Thus, it is obtainable by (10)Due to->Is bounded so that the structure estimation error can be made to go to zero when a limited number of hidden layer nodes are added.
Further, the L2 regularization method is utilized to solve the output weight value beta= [ beta ] 1 ，...β L ] T Is the optimum value beta of (2) * The following are provided:
wherein the method comprises the steps ofThe vector is output for the hidden layer. Solving (11) by least squares method to obtain
β＝(H T H+I L ) -1 H T S(p,q,r 0 ) (12)
Wherein I is L And is an identity matrix of L x L. Since formula (12) contains a structurally indeterminate amount S (p, q, r 0 ) Cannot be directly solved, and then the estimated value of beta is designedIs adapted to make ∈>
The dynamic model of the rehabilitation training robot with the structure uncertainty change estimation is obtained by the method that:
and thirdly), providing a designated transient time stability control method, and respectively designing transient time controllers and steady state time controllers to inhibit the influence of structural change on the tracking performance of the man-machine system and realize the stable tracking control of the designated transient time. The method is characterized in that: the error state equation is established as follows:
the transient motion process of the system is recorded as t is less than or equal to t f Wherein t is f For the specified transient time, a specified transient time stability controller is designed according to an error state equation (14). Setting variableη 1 Is the parameter and eta 1 Not less than 1, and designing intermediate variable z 2 (t)＝ε 2 (t)-ε 2d (t)＝ε 2 (t)+ψ 1 (t) for z 2 And (t) deriving to obtain:
wherein the method comprises the steps ofIs psi 1 (t) vs. ε 1 (t) deviation determination, ->Is psi 1 (t) biasing t. Thus the system (14) is
Let the weight estimation errorThe following weight self-adaptive rate is designed
Wherein the method comprises the steps ofη 3 Is the parameter and eta 3 And is more than or equal to 1. The lyapunov function was designed as:
deriving equation (18) along equation (16) to obtain
For transient motion processes, i.e. t.ltoreq.t f Designing a designated transient time tracking controller as
Wherein the method comprises the steps oft f To specify transient time, eta 2 Is the parameter and eta 2 ≥1，Is B 1 (θ) pseudo-inverse matrix. Substituting formula (20) into formula (19)
As can be seen from the lyapunov function (18),thus, +.>Further substituting into (21) to obtain
Inequality transformation of formula (22) is performed to obtain
Wherein ζ=η 1 +η 2 +η 3 >1. And further has
By solving the differential inequality (24), we obtain
V 1 (t)≤ln(C(t f -t) ξ +1) (25)
Wherein the method comprises the steps oft 0 Is the initial time.
As can be seen from equation (25), when t=t f V at the time of 1 (t f ) =0, i.eFrom this, the system can be known to be at the designated transient time t f And stably tracking the training track at the moment.
For steady state motion processes, i.e. t>t f Designing a steady-state process controller according to the error state equation (14) such that the rehabilitation robot is at t>t f And continuing to stably track the appointed training track. Design of Lyapunov function
Deriving (26) to obtain
Designing a steady state process controller to be
Wherein R is>And 0 is a controller adjusting parameter. Substituting the controller (28) into the formula (27),the system can realize tracking control of steady-state process.
Step four), based on STM32G4 series singlechip will export PWM signal and provide motor drive module, make rehabilitation training robot can help the patient to follow doctor appointed walking track, its characterized in that: STM32G4 series single-chip microcomputer is used as a main controller, and the input of the main controller is connected with a motor speed measuring module, and the output of the main controller is connected with a motor driving module; the motor driving module is connected with the direct current motor; the power supply system supplies power to the respective electrical devices. The control method of the main controller is to read the feedback signal of the motor encoder and the control command signal X given by the main controller d (t) andan error signal is calculated. According to the error signal, the main controller calculates the control quantity of the motor according to a preset control algorithm and sends the control quantity to the motor driving module, and the motor rotates to drive the wheels to maintain self balance and move in a specified mode.
The advantages and effects:
the invention relates to a specified transient time stability control method for inhibiting structural change of a rehabilitation robot, which has the following advantages:
according to the invention, a man-machine system dynamics model with uncertain structural change estimated by an FSCN method is established by splitting physical quantities influencing structural change in a rehabilitation training robot dynamics model; the rehabilitation robot specified transient time stable controller for inhibiting the structural change is designed, stable tracking control of the system with specified transient time is realized, tracking precision of a man-machine system is improved, and safety of a trainer is guaranteed.
Description of the drawings:
FIG. 1 is a block diagram of the operation of a controller according to the present invention;
FIG. 2 is a graph of a system of the present invention;
FIG. 3 shows a STM32G4 single-chip microcomputer minimum system according to the present invention;
FIG. 4 is a power circuit of the present invention;
FIG. 5 is a circuit of the motor drive and speed measurement module of the present invention;
fig. 6 is a circuit of the general principles of the hardware of the present invention.
The specific embodiment is as follows:
the present invention will be further described with reference to the accompanying drawings, but the scope of the present invention is not limited by the examples.
The dynamic model of the rehabilitation training robot is utilized to split physical quantities of structural changes related to the eccentricity and the eccentric angle, and the dynamic model for describing the uncertainty of the structure of the rehabilitation training robot is established; providing a random configuration network (Finite Stochastic Configuration Networks, FSCN) with a limited hidden layer nodes, taking a track tracking error as network input, obtaining structure uncertainty change estimation, and constructing a dynamic model of the rehabilitation training robot with the structure uncertainty estimation; providing a designated transient time stability control method, and respectively designing transient time and steady state time controllers to inhibit the influence of structural change on the tracking performance of a man-machine system and realize the stable tracking control of the designated transient time;
a specified transient time stability control method for restraining structural change of a rehabilitation robot is characterized in that:
1) Utilizing a dynamic model of the rehabilitation training robot to split physical quantities of structural changes related to the eccentricity and the eccentric angle, and establishing a dynamic model for describing the uncertainty of the structure of the rehabilitation training robot;
2) Providing a random configuration network FSCN with a limited hidden layer nodes, taking a track tracking error as network input, obtaining structure uncertainty change estimation, and constructing a dynamic model of the rehabilitation training robot with the structure uncertainty estimation;
3) The method for stabilizing and controlling the designated transient time is provided, and the influence of structural change on the tracking performance of the man-machine system is restrained by respectively designing the transient time controller and the steady state time controller, so that the stable tracking control of the designated transient time is realized.
The method comprises the following steps:
step one), a dynamic model of the rehabilitation training robot is utilized to split physical quantities of structural changes related to eccentricity and eccentric angles, and a dynamic model describing that the rehabilitation training robot has an uncertain structure is established. The method is characterized in that: the kinetic model of the system is described below
Wherein the method comprises the steps of
X (t) represents the actual motion trail of the rehabilitation training robot, u (t) represents the control input force, M represents the mass of the robot, M represents the mass of the rehabilitation person, I 0 Represent moment of inertia, M 0 ,K(θ),B (θ) is a coefficient matrix. θ represents the angle between the horizontal axis and the connection between the robot center and the first wheel center, i.e., θ=θ 1 As known from the structure of the rehabilitation training robot,θ 3 ＝θ+π，/>l i representing the distance of the center of gravity of the system from the center of each wheel, r 0 Represents the distance from center to center of gravity, phi i Representing the x' axis and the l corresponding to each wheel i Included angles (i=1, 2,3, 4).
From the dynamics model (1) of the system, the physical quantity generating uncertain change of the system structure is eccentricity and eccentric angle, and the related physical quantity is split and M is calculated 0 K (θ) is represented as M 0 K(θ)＝M 1 +ΔM 1 Wherein M is 1 Consists of rehabilitation training robot mass, rehabilitation person mass and moment of inertia, and delta M 1 Representing a physical quantity of structural change, an
At the same time, coefficient matrixThe eccentric distance and the eccentric angle of the structural change are used as physical quantities of the structural change; while coefficient matrix B (θ) contains variable λ resulting from structural changes i Decompose it into B (θ) =b 1 (θ)+ΔB 1 (θ) wherein
Wherein D is a steady physical quantity and represents the distance from the center of the man-machine system to the center of each wheel; physical quantity affecting structural change in the split dynamics model (1) is divided, and a dynamics model for describing the uncertainty of the structure of the rehabilitation training robot is established as follows
Wherein the method comprises the steps ofPhysical quantities representing uncertain changes in the structure of the rehabilitation training robot.
Step two), a random configuration network FSCN with a limited hidden layer nodes is provided, a track tracking error is taken as network input, the structure uncertainty change estimation is obtained, and a dynamics model with the structure uncertainty estimation of the rehabilitation training robot is constructed, and the method is characterized in that: based on the rehabilitation training robot dynamics model (2) with uncertain structure, the method comprises the following steps ofThe system state equation is obtained as follows:
let the appointed motion trail of rehabilitation robot be x d (t)，ε 1 (t)＝x 1 (t)-x d (t) is the track-following error of the system,is the speed tracking error of the system.
By epsilon 1 (t) is a network input, S (p, q, r) 0 ) For network output, a FSCN estimation model with uncertain structure is established. When the node number of the hidden layer of the FSCN model is L-1, the output of the model is made to beAnd f 0 =0, a gaussian function is selected as the basis function, at which time the structural change estimation error is e L-1 ＝S(p，q，r 0 )-f L-1 (ε 1 (t)) when the error e L-1 Adding a node to the hidden layer when the expected requirement is not met, and randomly generating a meeting supervision mechanism delta L Is input weight w of (2) L And bias b L 。
Wherein e L-1 ＝[e L-1,1 ,e L-1,2 ,e L-1,3 ] T Is the estimated error when the node number of the hidden layer is L-1, beta j ＝[β j,1 ,β j,2 ,β j,3 ] T Output weight, w, of j hidden layer node j ＝[w j,1 ,w j,2 ,w j,3 ] T And b j The input weights and offsets for the j-th hidden layer node,and L is the number of hidden layer nodes for outputting the j-th hidden layer node.
The FSCN supervision mechanism is designed as follows:
wherein alpha is q For the adjustable parameter (q=1, 2, 3), parameter σ<0。
Next, the FSCN under supervision mechanism δ of a limited number of hidden layer nodes is further described L The structure uncertainty estimate may be made to trend toward zero. Due to i e L || 2 Is positively determined to obtain
Since the expression (5) has a discrete characteristic with respect to L, a difference operation is performed thereon to obtain
Due to beta L,q A constant can be obtained after calculation, and then the constant alpha is set q Beta is made to be L,q ＝α q Further, from the formula (6)
From (4), it can be seen that
Let L be max The number of hidden layer nodes when the structure uncertainty estimation tends to zero is known by discrete Lyapunov criteriaFrom equation (8), it can be seen that for any hidden layer node:
respectively performing accumulation operation on two ends of the inequality of (9) to obtain
Thus, it is obtainable by (10)Due to->Is bounded so that the structure estimation error can be made to go to zero when a limited number of hidden layer nodes are added.
Further, the L2 regularization method is utilized to solve the output weight value beta= [ beta ] 1 ，...β L ] T Is the optimum value beta of (2) * The following are provided:
wherein the method comprises the steps ofThe vector is output for the hidden layer. Solving (11) by least squares method to obtain
β＝(H T H+I L ) -1 H T S(p,q，r 0 ) (12)
Wherein I is L And is an identity matrix of L x L. Due to (12)) Containing structurally undefined amounts S (p, q, r) 0 ) Cannot be directly solved, and then the estimated value of beta is designedIs adapted to make ∈>
The dynamic model of the rehabilitation training robot with the structure uncertainty change estimation is obtained by the method that:
and thirdly), providing a designated transient time stability control method, and respectively designing transient time controllers and steady state time controllers to inhibit the influence of structural change on the tracking performance of the man-machine system and realize the stable tracking control of the designated transient time. The method is characterized in that: the error state equation is established as follows:
the transient motion process of the system is recorded as t is less than or equal to t f Wherein t is f For the specified transient time, a specified transient time stability controller is designed according to an error state equation (14). Setting variableη 1 Is the parameter and eta 1 Not less than 1, and designing intermediate variable z 2 (t)＝ε 2 (t)-ε 2d (t)＝ε 2 (t)+ψ 1 (t) for z 2 And (t) deriving to obtain:
wherein the method comprises the steps ofIs psi 1 (t) vs. ε 1 (t) deviation determination, ->Is psi 1 (t) biasing t. Thus the system (14) is
Let the weight estimation errorThe following weight self-adaptive rate is designed
Wherein the method comprises the steps ofη 3 Is the parameter and eta 3 And is more than or equal to 1. The lyapunov function was designed as:
deriving equation (18) along equation (16) to obtain
For transient motion processes, i.e. t.ltoreq.t f Designing a designated transient time tracking controller as
Wherein the method comprises the steps oft f To specify transient time, eta 2 Is the parameter and eta 2 ≥1，Is B 1 (θ) pseudo-inverse matrix. Substituting formula (20) into formula (19)
As can be seen from the lyapunov function (18),thus, +.>Further substituting into (21) to obtain
Inequality transformation of formula (22) is performed to obtain
Wherein ζ=η 1 +η 2 +η 3 >1. And further has
By solving the differential inequality (24), we obtain
V 1 (t)≤ln(C(t f -t) ξ +1) (25)
Wherein the method comprises the steps oft 0 Is the initial time.
As can be seen from equation (25), when t=t f V at the time of 1 (t f ) =0, i.eFrom this, the system can be known to be at the designated transient time t f And stably tracking the training track at the moment.
For steady state motion processes, i.e. t>t f Designing a steady-state process controller according to the error state equation (14) such that the rehabilitation robot is at t>t f And continuing to stably track the appointed training track. Design of Lyapunov function
Deriving (26) to obtain
Designing a steady state process controller to be
Wherein R is>And 0 is a controller adjusting parameter. Substituting the controller (28) into the formula (27),the system can realize tracking control of steady-state process.
Step four), based on STM32G4 series singlechip will export PWM signal and provide motor drive module, make rehabilitation training robot can help the patient to follow doctor appointed walking track, its characterized in that: STM32G4 series single-chip microcomputer is used as a main controller, and the input of the main controller is connected with a motor speed measuring module, and the output of the main controller is connected with a motor driving module; the motor driving module is connected with the direct current motor; the power supply system supplies power to the respective electrical devices. The control method of the main controller is to read the feedback of the motor encoderSignal and control command signal X given by main controller d (t) andan error signal is calculated. According to the error signal, the main controller calculates the control quantity of the motor according to a preset control algorithm and sends the control quantity to the motor driving module, and the motor rotates to drive the wheels to maintain self balance and move in a specified mode.
The invention solves the problem of stable control of the specified transient time of the rehabilitation training robot for inhibiting the structural change. Establishing a rehabilitation walking robot dynamics model with structure change estimation; the method for stabilizing and controlling the specified transient time of the rehabilitation robot for inhibiting the uncertain change of the structure realizes the stable tracking control of the human-machine system with the specified transient time by respectively designing the transient process and the steady process controller, effectively improves the tracking precision of the human-machine system and ensures the safety of a trainer.
The invention provides the output PWM signal for the motor driving module based on the STM32G4 series singlechip, so that the robot adapts to structural change and helps a rehabilitation person to track a training track appointed by a doctor. The control method improves the tracking precision of the man-machine system from a new view angle for inhibiting the structural change, and simultaneously, the rehabilitation training robot is enabled to be quickly stabilized by designing the designated transient time stabilization controller, so that the risk of collision of the robot caused by overlarge track errors is avoided, and the safety of a trainer is ensured.
Claims (1)
1. The specified transient time stability control method for inhibiting structural change of rehabilitation robot is characterized by comprising the following steps:
1) Utilizing a dynamic model of the rehabilitation training robot to split physical quantities of structural changes related to the eccentricity and the eccentric angle, and establishing a dynamic model for describing the uncertainty of the structure of the rehabilitation training robot;
2) Providing a random configuration network FSCN with a limited hidden layer nodes, taking a track tracking error as network input, obtaining structure uncertainty change estimation, and constructing a dynamic model of the rehabilitation training robot with the structure uncertainty estimation;
3) Providing a designated transient time stability control method, and respectively designing transient time and steady state time controllers to inhibit the influence of structural change on the tracking performance of a man-machine system and realize the stable tracking control of the designated transient time;
the kinetic model of the system is described below
Wherein the method comprises the steps of
X (t) represents the actual motion trail of the rehabilitation training robot, u (t) represents the control input force, M represents the mass of the robot, M represents the mass of the rehabilitation person, I 0 Represent moment of inertia, M 0 ,K(θ),B (theta) is a coefficient matrix; θ represents the angle between the horizontal axis and the connection between the robot center and the first wheel center, i.e., θ=θ 1 As known from the structure of the rehabilitation training robot,θ 3 ＝θ+π，/>l i representing the distance of the center of gravity of the system from the center of each wheel, r 0 Represents the distance from center to center of gravity, phi i Representing the x' axis and the l corresponding to each wheel i The included angles between the two are i=1, 2,3 and 4;
from the dynamics model (1) of the system, the physical quantity generating uncertain change of the system structure is eccentricity and eccentric angle, and the related physical quantity is split and M is calculated 0 K (θ) is represented as M 0 K(θ)＝M 1 +ΔM 1 Wherein M is 1 Consists of rehabilitation training robot mass, rehabilitation person mass and moment of inertia, and delta M 1 Representing a physical quantity of structural change, an
At the same time, coefficient matrixThe eccentric distance and the eccentric angle of the structural change are used as physical quantities of the structural change; while coefficient matrix B (θ) contains variable λ resulting from structural changes i Decompose it into B (θ) =b 1 (θ)+ΔB 1 (θ) wherein
Wherein D is a steady physical quantity and represents the distance from the center of the man-machine system to the center of each wheel; physical quantity affecting structural change in the split dynamics model (1) is divided, and a dynamics model for describing the uncertainty of the structure of the rehabilitation training robot is established as follows
Wherein the method comprises the steps ofPhysical quantity representing uncertain change of the structure of the rehabilitation training robot;
based on the rehabilitation training robot dynamics model (2) with uncertain structure, the method comprises the following steps ofThe system state equation is obtained as follows:
let the appointed motion trail of rehabilitation robot be x d (t)，ε 1 (t)＝x 1 (t)-x d (t) is the track-following error of the system,tracking errors for the speed of the system;
by epsilon 1 (t) is a network input, S (p, q, r) 0 ) Establishing an FSCN estimation model with uncertain structure for network output; when the node number of the hidden layer of the FSCN model is L-1, the output of the model is made to beAnd f 0 =0, a gaussian function is selected as the basis function, at which time the structural change estimation error is e L-1 ＝S(p,q,r 0 )-f L-1 (ε 1 (t)) when the error e L-1 Adding a node to the hidden layer when the expected requirement is not met, and randomly generating a meeting supervision mechanism delta L Is input weight w of (2) L And bias b L ；
Wherein e L-1 ＝[e L-1,1 ,e L-1,2 ,e L-1,3 ] T Is the estimated error when the node number of the hidden layer is L-1, beta j ＝[β j,1 ,β j,2 ,β j,3 ] T Output weight, w, of j hidden layer node j ＝[w j,1 ,w j,2 ,w j,3 ] T And b j The input weights and offsets for the j-th hidden layer node,the output of the j hidden layer node is that L is the number of hidden layer nodes;
the FSCN supervision mechanism is designed as follows:
wherein alpha is q For an adjustable parameter (q=1, 2, 3), the parameter σ < 0;
next, the FSCN under supervision mechanism δ of a limited number of hidden layer nodes is further described L Under, the structure uncertainty estimate may be made to trend toward zero; due to i e L || 2 Is positively determined to obtain
Since the expression (5) has a discrete characteristic with respect to L, a difference operation is performed thereon to obtain
Due to beta L,q A constant can be obtained after calculation, and then the adjustable parameter alpha is set q Beta is made to be L,q ＝α q Further, from the formula (6)
From (4), it can be seen that
Let L be max The number of hidden layer nodes when the structure uncertainty estimation tends to zero is known by discrete Lyapunov criteriaFrom equation (8), it can be seen that for any hidden layer node:
respectively performing accumulation operation on two ends of the inequality of (9) to obtain
Thus, it is obtainable by (10)Due to->Bounded, thus when a limited number of hidden layer nodes are added, the structure estimation error can be made to trend to zero;
solving output weight value beta= [ beta ] by using L2 regularization method 1 ,...β L ] T Is the optimum value beta of (2) * The following are provided:
wherein the method comprises the steps ofOutputting a vector for an implicit layer; solving (11) by least squares method to obtain
β＝(H T H+I L ) -1 H T S(p,q,r 0 ) (12)
Wherein I is L An identity matrix of L x L; since formula (12) contains a structurally indeterminate amount S (p, q, r 0 ) Cannot be directly solved, and then the estimated value of beta is designedIs adapted to make ∈>
The dynamic model of the rehabilitation training robot with the structure uncertainty change estimation is obtained by the method that:
the error state equation is established as follows:
the transient motion process of the system is recorded as t is less than or equal to t f Wherein t is f Designing a designated transient time stability controller according to an error state equation (14) for the designated transient time; setting variableη 1 Is the parameter and eta 1 Not less than 1, and designing intermediate variable z 2 (t)＝ε 2 (t)-ε 2d (t)＝ε 2 (t)+ψ 1 (t) for z 2 And (t) deriving to obtain:
wherein the method comprises the steps ofIs psi 1 (t) vs. ε 1 (t) deviation determination, ->Is psi 1 (t) biasing t; thus the system (14) is
Let the weight estimation errorThe following weight self-adaptive rate is designed
Wherein the method comprises the steps ofη 3 Is the parameter and eta 3 1 or more; the lyapunov function was designed as:
deriving equation (18) along equation (16) to obtain
For transient motion processes, i.e. t.ltoreq.t f Designing a designated transient time tracking controller as
Wherein the method comprises the steps oft f To specify transient time, eta 2 Is the parameter and eta 2 ≥1，Is B 1 A pseudo-inverse of (θ); substituting formula (20) into formula (19)
As can be seen from the lyapunov function (18),thus, +.>Further substituting into (21) to obtain
Inequality transformation of formula (22) is performed to obtain
Wherein ζ=η 1 +η 2 +η 3 > 1; and further has
By solving the differential inequality (24), we obtain
V 1 (t)≤ln(C(t f -t) ξ +1) (25)
Wherein the method comprises the steps oft 0 Is the initial time;
as can be seen from equation (25), when t=t f V at the time of 1 (t f ) =0, i.eFrom this, the system can be known to be at the designated transient time t f The stable tracking training track is achieved at the moment;
for steady state motion processes, i.e. t > t f Designing a steady-state process controller according to the error state equation (14) so that the rehabilitation robot is at t > t f Continuing to stably track the appointed training track; design of Lyapunov function
Deriving (26) to obtain
Designing a steady state process controller to be
Wherein R > 0 is a controller adjustment parameter; substituting the controller (28) into the formula (27),the system can realize tracking control of steady-state process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110756756.9A CN113359470B (en) | 2021-07-05 | 2021-07-05 | Designated transient time stability control method for restraining structural change of rehabilitation robot |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110756756.9A CN113359470B (en) | 2021-07-05 | 2021-07-05 | Designated transient time stability control method for restraining structural change of rehabilitation robot |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113359470A CN113359470A (en) | 2021-09-07 |
| CN113359470B true CN113359470B (en) | 2023-08-11 |
Family
ID=77538165
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110756756.9A Active CN113359470B (en) | 2021-07-05 | 2021-07-05 | Designated transient time stability control method for restraining structural change of rehabilitation robot |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113359470B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112132096A (en) * | 2020-09-30 | 2020-12-25 | 中国矿业大学 | Behavior modal identification method of random configuration network for dynamically updating output weight |
| CN112229624A (en) * | 2020-09-30 | 2021-01-15 | 中国矿业大学 | Pneumatic regulating valve fault diagnosis method based on low-deviation random configuration network |
| CN112433495A (en) * | 2020-11-27 | 2021-03-02 | 沈阳工业大学 | Rapid finite time control of rehabilitation robot based on SCN (substation configuration network) man-machine uncertain model |
| CN112506054A (en) * | 2020-11-27 | 2021-03-16 | 沈阳工业大学 | Rehabilitation robot random finite time stable control based on SCN observation active thrust |
| CN112571424A (en) * | 2020-11-27 | 2021-03-30 | 沈阳工业大学 | Direct constraint control of each axis speed of rehabilitation robot based on SCN walking force estimation |
-
2021
- 2021-07-05 CN CN202110756756.9A patent/CN113359470B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112132096A (en) * | 2020-09-30 | 2020-12-25 | 中国矿业大学 | Behavior modal identification method of random configuration network for dynamically updating output weight |
| CN112229624A (en) * | 2020-09-30 | 2021-01-15 | 中国矿业大学 | Pneumatic regulating valve fault diagnosis method based on low-deviation random configuration network |
| CN112433495A (en) * | 2020-11-27 | 2021-03-02 | 沈阳工业大学 | Rapid finite time control of rehabilitation robot based on SCN (substation configuration network) man-machine uncertain model |
| CN112506054A (en) * | 2020-11-27 | 2021-03-16 | 沈阳工业大学 | Rehabilitation robot random finite time stable control based on SCN observation active thrust |
| CN112571424A (en) * | 2020-11-27 | 2021-03-30 | 沈阳工业大学 | Direct constraint control of each axis speed of rehabilitation robot based on SCN walking force estimation |
Non-Patent Citations (1)
| Title |
|---|
| 党建猛.基于级位切换策略的高速列车制动过程建模与控制方法研究.中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑.2018,第4章. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113359470A (en) | 2021-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Neural-network-based self-tuning PI controller for precise motion control of PMAC motors | |
| CN112433495B (en) | Rehabilitation robot rapid finite time control based on SCN man-machine uncertain model | |
| CN112506054B (en) | Rehabilitation robot random finite time stable control based on SCN observation active thrust | |
| CN113359470B (en) | Designated transient time stability control method for restraining structural change of rehabilitation robot | |
| Su et al. | Trajectory tracking control of human support robots via adaptive sliding-mode approach | |
| CN113031442B (en) | Modularized mechanical arm dispersed robust fault-tolerant control method and system | |
| CN113359767B (en) | Method for controlling safe driving of limited track tracking error of robot structure with slow change | |
| Krafes et al. | Linear, nonlinear and intelligent controllers for the inverted pendulum problem | |
| Danh et al. | Comparison of estimator-based compensation schemes for hydrostatic transmissions with uncertainties | |
| CN112034842A (en) | Service robot speed constraint tracking control method suitable for different users | |
| CN116755481A (en) | PID vibration active control and reliability analysis method based on neural network | |
| CN113325720B (en) | Self-adaptive tracking control method for rehabilitation training robot with movement speed decision | |
| CN116135485A (en) | Design method of preset performance track tracking controller of two-degree-of-freedom mechanical arm | |
| CN112433474B (en) | Safety triggering control method of cushion robot based on SCN internal interference force estimation | |
| CN113419423B (en) | Tracking control method for service robot to adapt to structural change in limited time | |
| Prakash et al. | Neuro-PI controller based model reference adaptive control for nonlinear systems | |
| Aguilar-Ibáñez et al. | A linear differential flatness approach to controlling the Furuta pendulum | |
| CN115128951A (en) | Double-loop high-performance control method based on expected track limited optimization | |
| CN112571424B (en) | Rehabilitation robot shaft speed direct constraint control based on SCN walking force estimation | |
| Hernández et al. | Adaptive neural sliding mode control of an inverted pendulum mounted on a ball system | |
| Zhao et al. | Accelerated adaptive backstepping control of the chaotic PMSM via the type-2 sequential fuzzy neural network | |
| Barambones et al. | Robust speed estimation and control of an induction motor drive based on artificial neural networks | |
| CN112571424A (en) | Direct constraint control of each axis speed of rehabilitation robot based on SCN walking force estimation | |
| Li et al. | Design of Fuzzy Cross Coupling Controller for Cartesian Robot | |
| CN117032293A (en) | Random tracking control method for service robot motion environment data driving observation |
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 |