CN109512518A - A kind of operating robot man-machine coordination motion blur model reference learning control method - Google Patents
A kind of operating robot man-machine coordination motion blur model reference learning control method Download PDFInfo
- Publication number
- CN109512518A CN109512518A CN201811555139.7A CN201811555139A CN109512518A CN 109512518 A CN109512518 A CN 109512518A CN 201811555139 A CN201811555139 A CN 201811555139A CN 109512518 A CN109512518 A CN 109512518A
- Authority
- CN
- China
- Prior art keywords
- robot
- fuzzy
- learning
- model
- follows
- 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.)
- Pending
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
- A61B34/30—Surgical robots
-
- 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
Abstract
The invention discloses a kind of operating robot man-machine coordination motion blur model reference learning control methods, it is characterized in that, including learning system and robot, the learning system includes fuzzy controller, learning organization and the admittance controller being cooperatively connected, the tractive force and movement speed that robot end is subject to are as the input parameter of learning system, using the admittance value of robot as the output parameter of learning system；It is the following steps are included: the input and output parameter of learning system is blurred and establishes fuzzy set；Initial fuzzy rules table is established in fuzzy controller according to fuzzy set and establishes the reference model of learning system；Learning organization is modified to initial fuzzy rules table according to input parameter and reference model and is exported after adjusting admittance value, realizes that robot matches surgical action.The present invention can effectively make the motion process of robot matching doctor's operation, improve the compliance and safety of operating robot, obtain ideal man-machine interaction effect.
Description
Technical field
The present invention relates to medical operating robot fields, and in particular to a kind of operating robot man-machine coordination motion blur mould
Type reference learning control method.
Background technique
In recent years in operating robot field, the direct physical contact between robot and doctor is an important research
Direction.Currently, in the research of many operating robot human-computer interaction campaigns, impedance or the admittance control of robot by
It is extensive to use.But due to the strong nonlinearity of operating robot Cooperation controlling process, high order, time variation and random disturbances etc.
Factor causes control rule incomplete, and the individual factor model of doctor is difficult to build by mathematical modeling in the course of surgery
It is vertical, cause movement effects to rely primarily on the subjective intention of designer, so that ideal man-machine interaction effect cannot be obtained.
In the course of surgery, doctor needs the robot moment to keep its optimal movement state, therefore, in view of the above-mentioned problems,
It is necessary to propose a kind of fuzzy model reference learning control method of operating robot man-machine collaboration control.
Summary of the invention
The technical problem to be solved in the present invention is to provide a kind of operating robot man-machine coordination motion blur model references
Control method is practised, can effectively make the motion process of robot matching doctor's operation, improve the compliance and peace of operating robot
Quan Xing obtains ideal man-machine interaction effect.
In order to solve the above-mentioned technical problems, the present invention provides a kind of operating robot man-machine coordination motion blur model ginsengs
Learning control method, including learning system and robot are examined, the learning system includes the fuzzy controller being cooperatively connected, study
Mechanism and admittance controller, the movement speed of tractive force and robot end that robot end is subject to is as learning system
Parameter is inputted, using the admittance value of robot as the output parameter of learning system；
Itself the following steps are included:
The input and output parameter of learning system is blurred and establishes fuzzy set by step 1)；
Step 2) establishes initial fuzzy rules table in fuzzy controller according to fuzzy set and establishes learning system
Reference model；
Step 3) learning organization modifies to initial fuzzy rules table according to input parameter and reference model, is obscured
Rules modification table；
Step 4) fuzzy controller adjusts admittance value according to fuzzy rule modification table and exports, and realizes robot matching operation
Movement.
Further, the dynamic that the fuzzy controller is used to establish between the tractive force of robot and movement speed is closed,
Its formula are as follows:
mdV+BdV=Fh；
Wherein mdFor virtual inertia, BdFor admittance value, FhFor the tractive force for being applied to robot end, V is actual speed.
Further, the fuzzy set of input and output parameter is established are as follows:
Selection becomes admittance control, inputs the tractive force F being subject to for robot endh, domain is [- 7N, 7N],
The domain of its fuzzy subset is { -4, -3, -2, -1,0,1,2,3,4 }；
The speed V of robot end, domain are [- 400mm/s, 400mm/s],
The domain of its fuzzy subset is { -4, -3, -2, -1,0,1,2,3,4 }；
The admittance value B of robotd, domain is [5mm/ (sN), 50mm/ (sN)]；
The domain of its fuzzy subset is { 1,2,3,4,5 }.
Further, the method for building up of the reference model are as follows:
Human body optimal motion characteristic model is gone out by minimum acceleration theory deduction,
X (τ)=X0+(Xf-X0)(6τ5-15τ4+10τ3)
Wherein, X0And XfThe whole story position of movement is respectively represented, τ is time scale constant, τ=t/tf, wherein t is movement
Time, tfThe time is terminated for movement；
Further, the learning organization includes obscuring reverse model and Knowledge-based modifier, and foundation obscures reverse model
Fuzzy reverse rule list, establish adjustment rule according to reverse rule list is obscured, when triggering adjustment rule, Knowledge-based modifier
To the admittance value in fuzzy controller, the i.e. admittance value of robot.
Further, the method for building up of rule is adjusted are as follows:
It first differentiates to human body optimal motion characteristic model, solution can obtain:
Wherein, VjerkFor the target value of reference model；
Error between reference model and controlled device output valve are as follows:
ye(kT)=Vjerk-V；
According to the input for obscuring reverse rule list and obtain obscuring reverse model are as follows:
ye(kT)=Vjerk-V
yc(kT)=(ye(kT)-ye(kT-T))/T
Output is p (kT)；
Wherein, y is inputtede(kT) fuzzy set domain are as follows: { -5, -4, -3, -2, -1,0,1,2,3,4,5 }；
Input yc(kT) fuzzy set domain are as follows: { -5, -4, -3, -2, -1,0,1,2,3,4,5 }；
Export the fuzzy set domain of p (kT) are as follows: -1, -0.8, -0.6, -0.4, -0.2,0,0.2,0.4,0.6,0.8,
1}
Therefore, Cm(kT)=Bd(kT-T)+P (kT), wherein CmIt (KT) is B after modifyingdValue；
By obscuring hidden letter ruleRealize triggering activation modification, activity are as follows:
Only activity isWhen, condition is set up,
Hidden letter rule is obscured to be activated.
Further, work as ye(kT) when ≈ 0, adjustment terminates.
Further, the learning organization is off-line learning mechanism.
Beneficial effects of the present invention:
By the self-teaching method of fuzzy model, it can be effectively solved the problems such as control rule is incomplete, obtain
Ideal man-machine interaction effect, and improve the compliance and safety of operating robot.Operating robot is assisted suitable for medical treatment.
And since self-teaching method is carried out according to the case where doctor's practical operation, it can be considered that doctor
People's operating habit, to improve the intelligence degree of robot.
Detailed description of the invention
Fig. 1 is the structural schematic diagram of fuzzy model reference learning control system of the invention.
Specific embodiment
The present invention will be further explained below with reference to the attached drawings and specific examples, so that those skilled in the art can be with
It more fully understands the present invention and can be practiced, but illustrated embodiment is not as a limitation of the invention.
Shown in referring to Fig.1, operating robot man-machine coordination motion blur model reference learning control method of the invention
One embodiment, initially sets up the overall structure of control system, including learning system and robot, and learning system includes being cooperatively connected
Fuzzy controller, reference model, learning organization and admittance controller, the tractive force that robot end is subject to and robot end
Input parameter of the movement speed at end as learning system, using the admittance value of robot as the output parameter of learning system；
In the course of surgery, doctor is it is not possible that training operating robot again and again is optimal motion mode, therefore
What the present invention selected is off-line training learning organization, improves the performance of heuristic fuzzy controller, and Optimizing Fuzzy Controller is being cured
When life needs robot assisted to move, robot can remain optimal motion state constantly.
Specifically it is, the input and output parameter of learning system is blurred and establishes fuzzy set；According to fuzzy
It is integrated into the reference model established initial fuzzy rules table in fuzzy controller and establish learning system；Learning organization is according to defeated
Enter parameter and reference model to modify to initial fuzzy rules table, obtains fuzzy rule modification table；Fuzzy controller is according to mould
Paste rules modification table adjustment admittance value simultaneously exports, and realizes that robot matches surgical action.
Wherein, learning organization includes obscuring reverse model and Knowledge-based modifier composition, for learning and modifying output ginseng
Number obscures the fuzzy reverse rule list of reverse model by establishing, and establishes adjustment rule according to reverse rule list is obscured, works as triggering
When adjustment rule, Knowledge-based modifier is to the admittance value in fuzzy controller, the i.e. admittance value of robot.In returning for fuzzy model
Feedback information is transmitted to reference to the feedback information from robot and reference model is combined by Lu Zhong, fuzzy model study
Practise mechanism, the parameter that learning organization will be modified in fuzzy controller using obtained feedback information.Fuzzy controller is according to machine
The real-time speed of device people and the tractive force of operator control the variation of the admittance value of robot, keep the movement of robot close
Ideal motion model.
In its structure, the dynamic that fuzzy controller is used to establish between the tractive force of robot and movement speed is closed, public
Formula are as follows:
mdV+BdV=Fh；
Wherein mdFor virtual inertia, BdFor admittance value, FhFor the tractive force for being applied to robot end, V is actual speed.
The movement speed of tractive force and robot end that robot end is subject to as the input parameter of learning system,
Using the admittance value of robot as the output parameter of learning system, and fuzzy set is established after being blurred:
Selection becomes admittance control, inputs the tractive force F being subject to for robot endh, domain is [- 7N, 7N],
The domain of its fuzzy subset is { -4, -3, -2, -1,0,1,2,3,4 }；
The speed V of robot end, domain are [- 400mm/s, 400mm/s],
The domain of its fuzzy subset is { -4, -3, -2, -1,0,1,2,3,4 }；
The admittance value B of robotd, domain is [5mm/ (sN), 50mm/ (sN)]；
The domain of its fuzzy subset is { 1,2,3,4,5 }.
According to the own situation of doctors experience (operating habit) and robot, the initial fuzzy rules of fuzzy controller are established
Table:
Establish the reference model of control system:
It is this for human-computer interaction to establish minimum acceleration model for robotic training to optimal movement state
Model is most submissive motion mode, can derive following human body optimal motion characteristic model by minimum acceleration theory:
X (τ)=X0+(Xf-X0)(6τ5-15τ4+10τ3)
Wherein, X0And XfThe whole story position of movement is respectively represented, τ is time scale constant, τ=t/tf, wherein t is movement
Time, tfThe time is terminated for movement；
By differentiating to above-mentioned formula, solution can be obtained:
Wherein, VjerkFor the target value of reference model；
Error between reference model and controlled device output valve are as follows:
ye(kT)=Vjerk-V；
Error between reference model and controlled device output valve are as follows:
ye(kT)=Vjerk-V；
Off-line learning mechanism is designed, the fuzzy reverse rule list for obscuring reverse model is established:
According to the input for obscuring reverse rule list and obtain obscuring reverse model are as follows:
ye(kT)=Vjerk-V
yc(kT)=(ye(kT)-ye(kT-T))/T
Output is p (kT)；
Wherein, y is inputtede(kT) fuzzy set domain are as follows: { -5, -4, -3, -2, -1,0,1,2,3,4,5 }；
Input yc(kT) fuzzy set domain are as follows: { -5, -4, -3, -2, -1,0,1,2,3,4,5 }；
Export the fuzzy set domain of p (kT) are as follows: -1, -0.8, -0.6, -0.4, -0.2,0,0.2,0.4,0.6,0.8,
1}；
Therefore, Cm(kT)=Bd(kT-T)+P (kT), wherein CmIt (KT) is B after modifyingdValue；
By obscuring hidden letter ruleRealize triggering activation modification, activity are as follows:
Only activity isWhen, condition is set up,
It obscures hidden letter rule to be activated, admittance value is adjusted, remaining does not change.Work as ye(kT) when ≈ 0, be regarded as by
The ideal behavior of control object has arrived at, and the mechanism of learning system will no longer adjust the parameter of fuzzy controller or not do
Big adjustment out, it can fine tuning, some big movements are defaulted as no operation.
By off-line learning institutional adjustment initial fuzzy rules table, adjust robot according to modified fuzzy reasoning table
Admittance value, the motion process that matching surgical operates in the process.A kind of operating robot man-machine collaboration control provided by the invention
The fuzzy model reference learning control method of system can control the problems such as rule is incomplete with effective solution, obtain ideal people
Machine interaction effect, and improve the compliance and safety of operating robot.Operating robot is assisted suitable for medical treatment.
Embodiment described above is only to absolutely prove preferred embodiment that is of the invention and being lifted, protection model of the invention
It encloses without being limited thereto.Those skilled in the art's made equivalent substitute or transformation on the basis of the present invention, in the present invention
Protection scope within.Protection scope of the present invention is subject to claims.
Claims (8)
1. a kind of operating robot man-machine coordination motion blur model reference learning control method, which is characterized in that including study
System and robot, the learning system includes fuzzy controller, learning organization and the admittance controller being cooperatively connected, by machine
Input parameter of the movement speed of tractive force and robot end that people end is subject to as learning system, by the admittance of robot
It is worth the output parameter as learning system；
Itself the following steps are included:
The input and output parameter of learning system is blurred and establishes fuzzy set by step 1)；
Step 2) establishes initial fuzzy rules table according to fuzzy set in fuzzy controller and establishes the reference of learning system
Model；
Step 3) learning organization modifies to initial fuzzy rules table according to input parameter and reference model, obtains fuzzy rule
Modify table；
Step 4) fuzzy controller adjusts admittance value according to fuzzy rule modification table and exports, and realizes that robot matching operation is dynamic
Make.
2. operating robot man-machine coordination motion blur model reference learning control method as described in claim 1, feature
It is, the dynamic that the fuzzy controller is used to establish between the tractive force of robot and movement speed is closed, formula are as follows:
mdV+BdV=Fh；
Wherein mdFor virtual inertia, BdFor admittance value, FhFor the tractive force for being applied to robot end, V is actual speed.
3. operating robot man-machine coordination motion blur model reference learning control method as claimed in claim 2, feature
It is, the fuzzy set of input and output parameter is established are as follows:
Selection becomes admittance control, inputs the tractive force F being subject to for robot endh, domain is [- 7N, 7N],
The domain of its fuzzy subset is { -4, -3, -2, -1,0,1,2,3,4 }；
The speed V of robot end, domain are [- 400mm/s, 400mm/s],
The domain of its fuzzy subset is { -4, -3, -2, -1,0,1,2,3,4 }；
The admittance value B of robotd, domain is [5mm/ (sN), 50mm/ (sN)]；
The domain of its fuzzy subset is { 1,2,3,4,5 }.
4. operating robot man-machine coordination motion blur model reference learning control method as claimed in claim 2, feature
It is, the method for building up of the reference model are as follows:
Human body optimal motion characteristic model is gone out by minimum acceleration theory deduction,
X (τ)=Xo+(Xf-X0)(6τ5-15τ4+10τ3)
Wherein, X0And XfThe whole story position of movement is respectively represented, τ is time scale constant, τ=t/tf, wherein when t is movement
Between, tfThe time is terminated for movement.
5. operating robot man-machine coordination motion blur model reference learning control method as claimed in claim 4, feature
It is, the learning organization includes obscuring reverse model and Knowledge-based modifier, establishes the fuzzy reverse rule for obscuring reverse model
Then table establishes adjustment rule according to reverse rule list is obscured, and when triggering adjustment rule, Knowledge-based modifier is to fuzzy controller
In admittance value, i.e. the admittance value of robot.
6. operating robot man-machine coordination motion blur model reference learning control method as claimed in claim 5, feature
It is, adjusts the method for building up of rule are as follows:
It first differentiates to human body optimal motion characteristic model, solution can obtain:
Wherein, VjerkFor the target value of reference model；
Error between reference model and controlled device output valve are as follows:
ye(kT)=Vjerk-V；
According to the input for obscuring reverse rule list and obtain obscuring reverse model are as follows:
ye(kT)=Vjerk-V
yc(kT)=(ye(kT)-ye(kT-T))/T
Output is p (kT)；
Wherein, y is inputtede(kT) fuzzy set domain are as follows: { -5, -4, -3, -2, -1,0,1,2,3,4,5 }；
Input yc(kT) fuzzy set domain are as follows: { -5, -4, -3, -2, -1,0,1,2,3,4,5 }；
Export the fuzzy set domain of p (kT) are as follows: { -1, -0.8, -0.6, -0.4, -0.2,0,0.2,0.4,0.6,0.8,1 }
Therefore, Cm(kT)=Bd(kT-T)+P (kT), wherein CmIt (KT) is B after modifyingdValue；
By obscuring hidden letter ruleRealize triggering activation modification, activity are as follows:
Only activity isWhen, condition is set up, and is obscured
Hidden letter rule is activated.
7. operating robot man-machine coordination motion blur model reference learning control method as claimed in claim 6, feature
It is, works as ye(kT) when ≈ 0, adjustment terminates.
8. operating robot man-machine coordination motion blur model reference learning control method as described in claim 1, feature
It is, the learning organization is off-line learning mechanism.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811555139.7A CN109512518A (en) | 2018-12-19 | 2018-12-19 | A kind of operating robot man-machine coordination motion blur model reference learning control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811555139.7A CN109512518A (en) | 2018-12-19 | 2018-12-19 | A kind of operating robot man-machine coordination motion blur model reference learning control method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109512518A true CN109512518A (en) | 2019-03-26 |
Family
ID=65796420
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811555139.7A Pending CN109512518A (en) | 2018-12-19 | 2018-12-19 | A kind of operating robot man-machine coordination motion blur model reference learning control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109512518A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070035384A (en) * | 2005-09-27 | 2007-03-30 | 이현식 | Design of a Fuzzy State Observer |
| CN107053179A (en) * | 2017-04-21 | 2017-08-18 | 哈尔滨思哲睿智能医疗设备有限公司 | A kind of mechanical arm Compliant Force Control method based on Fuzzy Reinforcement Learning |
| CN108785997A (en) * | 2018-05-30 | 2018-11-13 | 燕山大学 | A kind of lower limb rehabilitation robot Shared control method based on change admittance |
-
2018
- 2018-12-19 CN CN201811555139.7A patent/CN109512518A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070035384A (en) * | 2005-09-27 | 2007-03-30 | 이현식 | Design of a Fuzzy State Observer |
| CN107053179A (en) * | 2017-04-21 | 2017-08-18 | 哈尔滨思哲睿智能医疗设备有限公司 | A kind of mechanical arm Compliant Force Control method based on Fuzzy Reinforcement Learning |
| CN108785997A (en) * | 2018-05-30 | 2018-11-13 | 燕山大学 | A kind of lower limb rehabilitation robot Shared control method based on change admittance |
Non-Patent Citations (1)
| Title |
|---|
| SHAOLONG KUANG等: "Intelligent Control for Human-Robot Cooperation in Orthopedics Surgery", 《ADVANCES IN EXPERIMENTAL MEDICINE AND BIOLOGY》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Deng et al. | A learning-based hierarchical control scheme for an exoskeleton robot in human–robot cooperative manipulation | |
| Dimeas et al. | Fuzzy learning variable admittance control for human-robot cooperation | |
| Jordan | Computational aspects of motor control and motor learning | |
| CN107053179B (en) | A kind of mechanical arm Compliant Force Control method based on Fuzzy Reinforcement Learning | |
| Kober et al. | Policy search for motor primitives in robotics | |
| Bhattacharyya et al. | A synergetic brain-machine interfacing paradigm for multi-DOF robot control | |
| Asl et al. | An assist-as-needed control scheme for robot-assisted rehabilitation | |
| Karniel et al. | Human motor control: learning to control a time-varying, nonlinear, many-to-one system | |
| Wu et al. | Fuzzy sliding mode control of an upper limb exoskeleton for robot-assisted rehabilitation | |
| CN111243706B (en) | Upper limb rehabilitation robot training task generation method based on variable impedance demonstration learning | |
| CN106730629A (en) | Lower limb robot and the control method of active movement is carried out using the robot | |
| CN113633521A (en) | Control system and control method for upper limb exoskeleton rehabilitation robot | |
| Wu et al. | Learning from demonstration and interactive control of variable-impedance to cut soft tissues | |
| Morasso | A vexing question in motor control: the degrees of freedom problem | |
| Zhang et al. | Performance-based assistance control for robot-mediated upper-limbs rehabilitation | |
| Xiuxiang et al. | An adaptive fuzzy sliding mode control for angle tracking of human musculoskeletal arm model | |
| CN109512518A (en) | A kind of operating robot man-machine coordination motion blur model reference learning control method | |
| Kambara et al. | Learning and generation of goal-directed arm reaching from scratch | |
| Zhang et al. | Recasting brain-machine interface design from a physical control system perspective | |
| Malhotra et al. | Reduced dimensionality control for the act hand | |
| Hu et al. | Disturbance rejection speed control based on linear extended state observer for isokinetic muscle strength training system | |
| Perdikis et al. | Evidence accumulation in asynchronous BCI | |
| Mitrovic et al. | Adaptive optimal control for redundantly actuated arms | |
| Danel | Reinforcement learning for humanoid robot control | |
| CN113995629B (en) | Mirror image force field-based upper limb double-arm rehabilitation robot admittance control method and system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190326 |
|
| RJ01 | Rejection of invention patent application after publication |