CN109291052A - A kind of massaging manipulator training method based on deeply study - Google Patents
A kind of massaging manipulator training method based on deeply study Download PDFInfo
- Publication number
- CN109291052A CN109291052A CN201811261282.5A CN201811261282A CN109291052A CN 109291052 A CN109291052 A CN 109291052A CN 201811261282 A CN201811261282 A CN 201811261282A CN 109291052 A CN109291052 A CN 109291052A
- Authority
- CN
- China
- Prior art keywords
- pressure
- massaging manipulator
- method based
- training method
- action
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1628—Programme controls characterised by the control loop
- B25J9/163—Programme controls characterised by the control loop learning, adaptive, model based, rule based expert control
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J11/00—Manipulators not otherwise provided for
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
- G16H40/40—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the management of medical equipment or devices, e.g. scheduling maintenance or upgrades
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
- G16H40/60—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices
- G16H40/63—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices for local operation
Abstract
The invention discloses a kind of massaging manipulator training methods based on deeply study, it solves the problems, such as that massaging manipulator movement only exists in that simulation status, massaging action are inaccurate in the prior art, has the effect of that the craftsmenship of massaging manipulator can be enhanced, provides the fatigue of the accurate massage of profession, reduction manual massage；Its technical solution are as follows: acquisition movement, pressure data handle the data, construct reference action collection, reference pressure collection, and pressure value comfort level range is arranged；The data, reference action, reference pressure input neural network are subjected to prediction and decision, execute the corresponding action value of neural network output decision, pressure value, and compare with reference action, pressure value comfort level range；Meet and trained network is connected with the control system of massaging manipulator after imposing a condition.
Description
Technical field
The present invention relates to manipulator field more particularly to a kind of massaging manipulator training sides based on deeply study
Method.
Background technique
Nowadays, not diversified for the mechanical equipment of massage, it is single function or multi-functional massager, massage mostly
Chair etc. acts less, mechanization, is difficult to hold the use of power, more comfortable, more professional service cannot be provided to user.
Manual massage's movement is fine and smooth soft, and especially professional person massages, and craftsmenship is stronger, and gimmick dynamics is consummate.But profession massage
The negligible amounts of teacher, and can not accomplish to service whenever and wherever possible, cost is larger, so not being able to satisfy the demand of ordinary people.
With the development of artificial intelligence and the continuous promotion of productivity demand, industrial machinery arm is in more and more fields
Conjunction is used.Deep layer intensified learning is applied to more and more control problems, in robotic arm path planning field and animation simulation
Field shows great advantage.Since nitrification enhancement has higher-dimension sample complexity and other physical limits, so
The dimension and complexity of data are greatly reduced by deep learning and intensified learning combined training, but at present only in emulation shape
State can not be entirely applied to actual conditions.
Mainly capturing a little for mechanical arm control field is path planning, trajectory planning problem at present, but for mechanical arm mould
The case where imitating movement, especially carrying out mechanical arm action imitation with the method that deeply learns is very rare, and realizes
Come very difficult.
Summary of the invention
For overcome the deficiencies in the prior art, the present invention provides a kind of massaging manipulator instructions based on deeply study
Practice method, there is the craftsmenship that can enhance massaging manipulator, the accurate massage of profession is provided, reduces manual massage's workload
Effect.
The present invention adopts the following technical solutions:
A kind of massaging manipulator training method based on deeply study, acquisition movement, pressure data handle the number
According to, building reference action collection, reference pressure collection, and pressure value comfort level range is set；
The data, reference action, reference pressure input neural network are subjected to prediction and decision, it is defeated to execute neural network
The corresponding action value of decision, pressure value out, and compared with reference action, pressure value comfort level range；
Meet and trained network is connected with the control system of massaging manipulator after imposing a condition.
Further, data are acquired by motion capture gloves, motion capture gloves are for capturing each finger-joint, wrist pass
Action data at section.
Further, the motion capture gloves correspond to each finger-joint, wrist joint installs pressure sensor.
Further, data procedures are handled are as follows: by each movement segment editing of acquisition to set length, and will be after editing
Movement segment be divided into several pieces.
Further, the initial state value and pressure value of extraction movement segment, using action value as reference action, by pressure
Reference pressure value is used as after value normalization.
Further, the pressure data that the pressure value comfort level range is repeatedly collected feedback by pressure sensor obtains.
Further, the massaging manipulator include 14 articulations digitorum manus, 1 wrist joint and an elbow joint, articulations digitorum manus and
Wrist joint installation has the tentacle of pressure sensor.
Further, the tentacle is cushion.
Further, the neural network uses convolutional neural networks, and movement distribution is modeled with Gauss.
Further, it is finely adjusted by the movement of collection massaging manipulator, pressure data.
Compared with prior art, the beneficial effects of the present invention are:
(1) present invention makes manipulator reach the gimmick of professional person by deeply study, in continuous study, imitates ginseng
It examines movement simultaneously, carries out adjustment appropriate according to the actual situation, preferably adapt to different environment and massage object, mentioned to user
For more comfortable, more professional massage experience；
(2) The present invention reduces the weariness working of human treatment teacher, reduce cost, improve the professional of massage.
Detailed description of the invention
The accompanying drawings constituting a part of this application is used to provide further understanding of the present application, and the application's shows
Meaning property embodiment and its explanation are not constituted an undue limitation on the present application for explaining the application.
Fig. 1 is the flow chart of the application；
Fig. 2 is the neural metwork training figure of the application.
Specific embodiment
It is noted that following detailed description is all illustrative, it is intended to provide further instruction to the application.Unless another
It indicates, all technical and scientific terms used herein has usual with the application person of an ordinary skill in the technical field
The identical meanings of understanding.
It should be noted that term used herein above is merely to describe specific embodiment, and be not intended to restricted root
According to the illustrative embodiments of the application.As used herein, unless the context clearly indicates otherwise, otherwise singular
Also it is intended to include plural form, additionally, it should be understood that, when in the present specification using term "comprising" and/or " packet
Include " when, indicate existing characteristics, step, operation, device, component and/or their combination.
As background technique is introduced, massaging manipulator movement exists in the prior art and only exists in simulation status, massage
Inaccurate deficiency is acted, in order to solve technical problem as above, present applicant proposes a kind of based on deeply study
Massaging manipulator training method.
In a kind of typical embodiment of the application, as Figure 1-Figure 2, provide a kind of based on deeply study
Massaging manipulator training method, comprising the following steps:
Step 1, acquisition hand motion and pressure data:
Movement acquisition is that profession massage personage or movement supplier is allowed to wear existing motion capture gloves, and the movement is caught
Catching gloves can capture and record 14 joints of finger (there are two joints for thumb, and respectively there are three joints for remaining four finger) and wrist joint
The action data at place, as reference action collection；Since elbow joint with carpal angle by being adjusted, there is no need to conducts
Motion capture data.
Correspond to finger-joint on motion capture gloves, wrist position installs pressure sensor.
The massaging action and pressure value for acquiring multiple dynamics, facilitate user to select dynamics according to their needs.
Step 2, the collected data of processing, construct reference action collection, reference pressure collection, and the comfortable model of pressure value is arranged
It encloses:
Handle data procedures are as follows: by each movement segment editing of acquisition be setting length, and by the action movie after editing
Section is divided into several pieces.
Movement segment initial state value and pressure value, action value after obtaining editing return pressure value as reference action
Reference pressure value is used as after one change.
Pressure value comfort level range is obtained according to the pressure data collected and repeatedly test is fed back.
In some embodiments, each movement segment of acquisition is trimmed into 1.5 seconds, the segment vacancy less than 1.5 seconds
Time is set as 0；Each 1.5 seconds movement segments are divided into 5 parts again, every part 0.3 second.Due to hand massage actuation time
It is general shorter, it repeats similarly to act in a cycle, a movement can be completed in 1.5 seconds substantially, save the time, increased
Efficiency, and being divided equally into can guarantee to be combined into efficient situation by what 5 movement segments in 1.5 seconds linked up for 0.3 second substantially
One complete movement.
Step 3, apish swivel of hand construct massaging manipulator structure:
The massaging manipulator includes 14 articulations digitorum manus, 1 wrist joint and an elbow joint, and articulations digitorum manus and wrist joint are pacified
Dress has the tentacle of pressure sensor.
In some embodiments, the tentacle is cushion, for increasing comfort level；Pressure sensor is installed on cushion
It is interior.
Further, the cushion is made of rubber material.
Step 4 is trained using neural network:
Collected movement, reference action, reference pressure value input convolutional neural networks are subjected to prediction and decision, are executed
Network exports the corresponding movement of decision, pressure value, and compares with reference action, pressure value comfort standard.
When acting (similarity reaches 99%) similar enough to reference action, pressure value and being suitable for enough, movement is executed, it will
The control system of network connection massaging manipulator after training；
When movement, pressure value are unsatisfactory for above-mentioned condition, convolutional neural networks prediction and decision process are repeated.
Tactful network π is indicated that movement distribution is modeled with Gauss by one convolutional neural networks,
π (a | s)=N (μ (s), Σ) (1)
And the target learnt is exactly to find optimal policy π *=argπmaxJ(π)。
If every collection is started with fixed original state, expected return can be rewritten as pre- since the first step
Phase returns,
J (π)=E (R0| π)=Eτ~p (τ | π)[∑r(st, at)] (2)
It is above it is various in, J (π) is that long-term accumulation is rewarded, stFor current state, st+1For NextState, atFor current action,
s0For original state, τ is to sample tuple, and p (τ | π) it represents at tactful π a possibility that the τ of track.
The movement a that one layer of input is state s above neural network, previous step generatesi-1；One layer of input below is shape
State s and reference action agi, reference pressure, upper and lower level passes through the layer being fully connected with 512 units respectively, then can be with
The linear convergent rate layer for passing through two 128 units jointly, exports the movement of decision, as shown in Figure 2.
Network inputs state, reference action (element as target and measurement reward value), reference pressure value, previous step are raw
At movement, generated strategy by reward, value function V, each corresponding output action of strategy acts the state conduct of generation
Next state continues as input.
The particular content of network are as follows:
(1) the state s of manipulator:
By tuple θ=(θ of 47 dimensions1, θ2, θ3, θ4, θ5, θ6, θ7, θ8, θ9, θ10, θ11, θ12, θ13, θ14, θ15, θ16) composition,
First 14 respectively define from thumb to little finger, from finger tip to the joint for referring to root, and the 15th is defined as wrist joint, and the 16th
It is defined as elbow joint.
Each joint includes two angle, angular speed components again, and articulations digitorum manus junction and wrist joint include 15 pressures altogether
Force snesor.
Pressure value is determined by angle and angular speed, but is combined not unique.θ is normalized, convenient for nerve
The accuracy of the training of network.
(2) tuple ψ=(θ that the movement a of manipulator is tieed up by 321+ψ1... ..., θ16+ψ16) composition.
ψ is the angle for needing to rotate in the case where current state and angular speed.
ψ i is also normalized, if θi+ψiGreater than 1, then θi+ψi=θi+ψi-1。
ψ16Due to being elbow joint without reference to movement, so can be needed voluntarily to learn according to position.
(3) setting of reward function:
If pressure is not in comfort level range of pressure values, r=-10；If pressure within the scope of comfort level,
R=wa*ra+ww*rw+wy*ry+wt*rt+c+wp* peo,
W in formulaa=-0,55, ww=-0.05, wy=-0.3, wt=-0.1, c=1, wp=5.
raIt is the difference of angle in joint angles and reference action, rwIt is the difference of the angular speed in joint, ryIt is actual pressure
The difference of value and reference pressure value, rtIt is the difference of practical frame speed and reference action frame (0.3 second).
Peo is defaulted as 0, when user is by intensity button is lowered, peo=- | and gear after gear-tune before adjusting |, to help machine
Tool hand is switched to next gear faster.
All differences use the following form of Euclidean distance of exponential form, it may be assumed that
R=exp (∑ | | y-y ' | |2)
Y is real variable value in formula, and y ' is the value of reference variable.
Step 5, fine tuning
It is acted in true environment by collecting mechanical hand and the feedback data of pressure is finely adjusted again.
The foregoing is merely preferred embodiment of the present application, are not intended to limit this application, for the skill of this field
For art personnel, various changes and changes are possible in this application.Within the spirit and principles of this application, made any to repair
Change, equivalent replacement, improvement etc., should be included within the scope of protection of this application.
Claims (10)
1. a kind of massaging manipulator training method based on deeply study, which is characterized in that acquisition movement, pressure data,
The data are handled, construct reference action collection, reference pressure collection, and pressure value comfort level range is set；
The data, reference action, reference pressure input neural network are subjected to prediction and decision, neural network output is executed and determines
The corresponding action value of plan, pressure value, and compared with reference action, pressure value comfort level range；
Meet and trained network is connected with the control system of massaging manipulator after imposing a condition.
2. a kind of massaging manipulator training method based on deeply study according to claim 1, which is characterized in that
Data are acquired by motion capture gloves, motion capture gloves are for capturing each finger-joint, wrist joint action data.
3. a kind of massaging manipulator training method based on deeply study according to claim 2, which is characterized in that
The motion capture gloves correspond to each finger-joint, wrist joint installs pressure sensor.
4. a kind of massaging manipulator training method based on deeply study according to claim 1, which is characterized in that
Handle data procedures are as follows: by each movement segment editing of acquisition be setting length, and by the movement segment average mark after editing
At several pieces.
5. a kind of massaging manipulator training method based on deeply study according to claim 4, which is characterized in that
Extraction acts the initial state value and pressure value of segment, using action value as reference action, as ginseng after pressure value is normalized
Examine pressure value.
6. a kind of massaging manipulator training method based on deeply study according to claim 1, which is characterized in that
The pressure data that the pressure value comfort level range is repeatedly collected feedback by pressure sensor obtains.
7. a kind of massaging manipulator training method based on deeply study according to claim 1, which is characterized in that
The massaging manipulator includes 14 articulations digitorum manus, 1 wrist joint and an elbow joint, and articulations digitorum manus and wrist joint installation are with pressure
The tentacle of force snesor.
8. a kind of massaging manipulator training method based on deeply study according to claim 7, which is characterized in that
The tentacle is cushion.
9. a kind of massaging manipulator training method based on deeply study according to claim 1, which is characterized in that
The neural network uses convolutional neural networks, and movement distribution is modeled with Gauss.
10. a kind of massaging manipulator training method based on deeply study according to claim 1, feature exist
In by collecting the movement of massaging manipulator, pressure data is finely adjusted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811261282.5A CN109291052B (en) | 2018-10-26 | 2018-10-26 | Massage manipulator training method based on deep reinforcement learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811261282.5A CN109291052B (en) | 2018-10-26 | 2018-10-26 | Massage manipulator training method based on deep reinforcement learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109291052A true CN109291052A (en) | 2019-02-01 |
| CN109291052B CN109291052B (en) | 2021-11-09 |
Family
ID=65158970
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811261282.5A Active CN109291052B (en) | 2018-10-26 | 2018-10-26 | Massage manipulator training method based on deep reinforcement learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109291052B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110039537A (en) * | 2019-03-15 | 2019-07-23 | 北京精密机电控制设备研究所 | A kind of automatic measure on line multi joint motion planing method neural network based |
| CN110147891A (en) * | 2019-05-23 | 2019-08-20 | 北京地平线机器人技术研发有限公司 | Method, apparatus and electronic equipment applied to intensified learning training process |
| CN110516389A (en) * | 2019-08-29 | 2019-11-29 | 腾讯科技（深圳）有限公司 | Learning method, device, equipment and the storage medium of behaviour control strategy |
| CN110561430A (en) * | 2019-08-30 | 2019-12-13 | 哈尔滨工业大学（深圳） | robot assembly track optimization method and device for offline example learning |
| CN113211441A (en) * | 2020-11-30 | 2021-08-06 | 湖南太观科技有限公司 | Neural network training and robot control method and device |
| CN114053112A (en) * | 2021-10-19 | 2022-02-18 | 奥佳华智能健康科技集团股份有限公司 | Massage method, device, terminal equipment and medium |
| CN114609918A (en) * | 2022-05-12 | 2022-06-10 | 齐鲁工业大学 | Four-footed robot motion control method, system, storage medium and equipment |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003052771A (en) * | 2001-08-20 | 2003-02-25 | Sintokogio Ltd | Method and system for controlling massage machine |
| CN102727362A (en) * | 2012-07-20 | 2012-10-17 | 上海海事大学 | NUI (Natural User Interface)-based peripheral arm motion tracking rehabilitation training system and training method |
| US20150190246A1 (en) * | 2012-08-02 | 2015-07-09 | Korea University Of Technology And Education Industry-University Cooperation Foundation | Motion control device based on winding string |
| CN105311792A (en) * | 2014-07-02 | 2016-02-10 | 北京蝶禾谊安信息技术有限公司 | Rehabilitation training device and data collecting method thereof |
| CN107280910A (en) * | 2017-05-15 | 2017-10-24 | 武汉理工大学 | A kind of autonomous intelligence massager and operating method based on data acquisition |
| CN107825393A (en) * | 2017-12-14 | 2018-03-23 | 北京工业大学 | A kind of total joint measurement type data glove |
| CN108171329A (en) * | 2017-12-13 | 2018-06-15 | 华南师范大学 | Deep learning neural network training method, number of plies adjusting apparatus and robot system |
| US20180177666A1 (en) * | 2016-12-22 | 2018-06-28 | Rehab-Robotics Company Ltd. | Power Assistive Device For Hand Rehabilitation And A Method of Using The Same |
| CN108543216A (en) * | 2018-01-26 | 2018-09-18 | 南京航空航天大学 | A kind of hand function reconstructing device and its implementation based on master ＆ slave control |
| CN108621159A (en) * | 2018-04-28 | 2018-10-09 | 首都师范大学 | A kind of Dynamic Modeling in Robotics method based on deep learning |
-
2018
- 2018-10-26 CN CN201811261282.5A patent/CN109291052B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003052771A (en) * | 2001-08-20 | 2003-02-25 | Sintokogio Ltd | Method and system for controlling massage machine |
| CN102727362A (en) * | 2012-07-20 | 2012-10-17 | 上海海事大学 | NUI (Natural User Interface)-based peripheral arm motion tracking rehabilitation training system and training method |
| US20150190246A1 (en) * | 2012-08-02 | 2015-07-09 | Korea University Of Technology And Education Industry-University Cooperation Foundation | Motion control device based on winding string |
| CN105311792A (en) * | 2014-07-02 | 2016-02-10 | 北京蝶禾谊安信息技术有限公司 | Rehabilitation training device and data collecting method thereof |
| US20180177666A1 (en) * | 2016-12-22 | 2018-06-28 | Rehab-Robotics Company Ltd. | Power Assistive Device For Hand Rehabilitation And A Method of Using The Same |
| CN107280910A (en) * | 2017-05-15 | 2017-10-24 | 武汉理工大学 | A kind of autonomous intelligence massager and operating method based on data acquisition |
| CN108171329A (en) * | 2017-12-13 | 2018-06-15 | 华南师范大学 | Deep learning neural network training method, number of plies adjusting apparatus and robot system |
| CN107825393A (en) * | 2017-12-14 | 2018-03-23 | 北京工业大学 | A kind of total joint measurement type data glove |
| CN108543216A (en) * | 2018-01-26 | 2018-09-18 | 南京航空航天大学 | A kind of hand function reconstructing device and its implementation based on master ＆ slave control |
| CN108621159A (en) * | 2018-04-28 | 2018-10-09 | 首都师范大学 | A kind of Dynamic Modeling in Robotics method based on deep learning |
Non-Patent Citations (3)
| Title |
|---|
| HONGLI CHEN; XINYU WU; WEI FENG; TIANTIAN XU; YONG HE: "Design and path planning of Massagebot: One massaging robot climbing along the acupuncture points", 《INFORMATION AND AUTOMATION (ICIA)》 * |
| 杨庆华，张立彬，阮健，胥芳，鲍官军，沈建冰: "人类手指抓取过程关节的运动规律研究", 《中国机械工程》 * |
| 王洪伟: "足部按摩机器人的技能学习与融合方法研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110039537A (en) * | 2019-03-15 | 2019-07-23 | 北京精密机电控制设备研究所 | A kind of automatic measure on line multi joint motion planing method neural network based |
| CN110147891A (en) * | 2019-05-23 | 2019-08-20 | 北京地平线机器人技术研发有限公司 | Method, apparatus and electronic equipment applied to intensified learning training process |
| CN110147891B (en) * | 2019-05-23 | 2021-06-01 | 北京地平线机器人技术研发有限公司 | Method and device applied to reinforcement learning training process and electronic equipment |
| CN110516389A (en) * | 2019-08-29 | 2019-11-29 | 腾讯科技（深圳）有限公司 | Learning method, device, equipment and the storage medium of behaviour control strategy |
| CN110516389B (en) * | 2019-08-29 | 2021-04-13 | 腾讯科技（深圳）有限公司 | Behavior control strategy learning method, device, equipment and storage medium |
| CN110561430A (en) * | 2019-08-30 | 2019-12-13 | 哈尔滨工业大学（深圳） | robot assembly track optimization method and device for offline example learning |
| CN110561430B (en) * | 2019-08-30 | 2021-08-10 | 哈尔滨工业大学（深圳） | Robot assembly track optimization method and device for offline example learning |
| CN113211441A (en) * | 2020-11-30 | 2021-08-06 | 湖南太观科技有限公司 | Neural network training and robot control method and device |
| CN113211441B (en) * | 2020-11-30 | 2022-09-09 | 湖南太观科技有限公司 | Neural network training and robot control method and device |
| CN114053112A (en) * | 2021-10-19 | 2022-02-18 | 奥佳华智能健康科技集团股份有限公司 | Massage method, device, terminal equipment and medium |
| CN114609918A (en) * | 2022-05-12 | 2022-06-10 | 齐鲁工业大学 | Four-footed robot motion control method, system, storage medium and equipment |
| CN114609918B (en) * | 2022-05-12 | 2022-08-02 | 齐鲁工业大学 | Four-footed robot motion control method, system, storage medium and equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109291052B (en) | 2021-11-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109291052A (en) | A kind of massaging manipulator training method based on deeply study | |
| US20220249937A1 (en) | Method and system of optimizing and personalizing resistance force in an exercise | |
| Noohi et al. | A model for human–human collaborative object manipulation and its application to human–robot interaction | |
| US20190009133A1 (en) | Systems and methods for data-driven movement skill training | |
| WO2020098197A1 (en) | Palm-supported finger rehabilitation training apparatus and use method | |
| Ngeo et al. | Control of an optimal finger exoskeleton based on continuous joint angle estimation from EMG signals | |
| Avraham et al. | Toward perceiving robots as humans: Three handshake models face the turing-like handshake test | |
| Strazzulla et al. | Online bimanual manipulation using surface electromyography and incremental learning | |
| CN108983973A (en) | A kind of humanoid dexterous myoelectric prosthetic hand control method based on gesture identification | |
| CN104777775A (en) | Two-wheeled self-balancing robot control method based on Kinect device | |
| Esfahlani et al. | An adaptive self-organizing fuzzy logic controller in a serious game for motor impairment rehabilitation | |
| CN109718059A (en) | Hand healing robot self-adaptation control method and device | |
| Wu et al. | A Visual-Based Gesture Prediction Framework Applied in Social Robots. | |
| Zheng et al. | A virtual reality rehabilitation training system based on upper limb exoskeleton robot | |
| Wei et al. | A novel upper limb rehabilitation system with hand exoskeleton mechanism | |
| CN107618018A (en) | A kind of manipulator behavior speed proportional control method based on myoelectricity | |
| Miao et al. | Performance-based iterative learning control for task-oriented rehabilitation: a pilot study in robot-assisted bilateral training | |
| González et al. | Planning, execution and monitoring of physical rehabilitation therapies with a robotic architecture. | |
| CN105999652A (en) | Brain injury rehabilitation training device based on pinching of double finger pulps | |
| Jamwal et al. | Utilizing the intelligence edge framework for robotic upper limb rehabilitation in home | |
| CN113616193A (en) | Rhythm coordination characteristic characterization method based on self-adaptive CPG parameter identification | |
| CN108656119A (en) | A kind of control method of humanoid robot | |
| CN113357773A (en) | Method and device for air conditioner control and air conditioner | |
| Ueyama | System identification of neural mechanisms from trial-by-trial motor behaviour: modelling of learning, impairment and recovery | |
| Tzorakoleftherakis et al. | Effects of optimal tactile feedback in balancing tasks: A pilot study |
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 |