CN108115678A - Robot and its method of controlling operation and device - Google Patents
Robot and its method of controlling operation and device Download PDFInfo
- Publication number
- CN108115678A CN108115678A CN201611069128.9A CN201611069128A CN108115678A CN 108115678 A CN108115678 A CN 108115678A CN 201611069128 A CN201611069128 A CN 201611069128A CN 108115678 A CN108115678 A CN 108115678A
- Authority
- CN
- China
- Prior art keywords
- things
- robot
- behavior
- predetermined action
- control instruction
- 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
- 238000000034 method Methods 0.000 title claims abstract description 37
- 230000009471 action Effects 0.000 claims abstract description 78
- 230000008451 emotion Effects 0.000 claims abstract description 54
- 238000011156 evaluation Methods 0.000 claims abstract description 22
- 238000012549 training Methods 0.000 claims abstract description 12
- 238000004364 calculation method Methods 0.000 claims description 7
- 239000012141 concentrate Substances 0.000 claims description 6
- 230000006399 behavior Effects 0.000 description 87
- 238000010408 sweeping Methods 0.000 description 7
- 230000007246 mechanism Effects 0.000 description 6
- 230000002787 reinforcement Effects 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 230000000638 stimulation Effects 0.000 description 5
- 238000005728 strengthening Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 206010027476 Metastases Diseases 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000009401 metastasis Effects 0.000 description 2
- 230000007996 neuronal plasticity Effects 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000003014 reinforcing effect Effects 0.000 description 1
- 230000004936 stimulating effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000004148 unit process Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
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
- B25J9/1656—Programme controls characterised by programming, planning systems for manipulators
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Manipulator (AREA)
Abstract
The invention discloses a kind of robot and its method of controlling operation and device.This method includes:Control instruction is received, wherein, control instruction is used to indicate robot and performs predetermined action;Control instruction is inputted into preset model, is exported as a result, wherein, output result includes the corresponding behavior emotion model of predetermined action performed with robot, and Affective Evaluation parameter training of the preset model at least based on user feedback obtains;Predetermined action is performed according to behavior emotion model according to output output control robot.By the present invention, user demand can not be met by solving the problems, such as that robot behavior performance is single.
Description
Technical field
The present invention relates to robot field, in particular to a kind of robot and its method of controlling operation and device.
Background technology
Robot is developed so far, and function is more and more perfect, and the requirement to robot is also higher and higher, existing intensified learning
Technology is widely used in robot, but most of concentrate on is moved in decision-making, such as control balance, control robot ambulation
Method, but the not yet depth development in decision-making mode, in addition intensified learning (Reinforcement Learning, referred to as
For RL) technology also developing in itself, ripe degree is far not achieved.In principle, intensified learning inspiration comes from behaviour psychology,
The experiments such as the dog of Pavlov, Skinner box are embodied in earliest, biologically also there is the support of neural plasticity scheduling theory.It is but strong
Chemical learning method does not touch the training pattern of emotion, does not have similar realization method in robot yet, rests on computer
The degree such as vision, natural language processing, robot can not have different behavior expressions according to the hobby of user.
Single the problem of can not meeting user demand is showed for robot behavior in correlation technique, not yet proposes have at present
The solution of effect.
The content of the invention
It is a primary object of the present invention to provide a kind of robot and its method of controlling operation and device, to solve behavior table
Existing single the problem of can not meeting user demand.
To achieve these goals, according to an aspect of the invention, there is provided a kind of robot motion control method, is somebody's turn to do
Method includes:Control instruction is received, wherein, control instruction is used to indicate robot and performs predetermined action;Control instruction is inputted
Preset model is exported as a result, wherein, output result includes the corresponding behavior emotion of predetermined action performed with robot
Pattern, Affective Evaluation parameter training of the preset model at least based on user feedback obtain;It is pressed according to output output control robot
Predetermined action is performed according to behavior emotion.
Further, the behavior emotion model is used to determine behavior emotion shape of the robot to the predetermined action
State, when the robot performs the predetermined action, the state of the robot is the behavior affective state.
Further, the control instruction is being inputted into preset model, before obtaining output result, this method further includes:
The corresponding behavior affection index of multiple things that things is concentrated is received, wherein, behavior affection index is used to represent robot to thing
The fancy grade for multiple things that object is concentrated;The corresponding behavior affection index of multiple things concentrated according to things establishes default mould
Type.
Further, multiple things that things is concentrated are the corresponding things of predetermined action that robot performs.
Further, the quantity of things assembling is n, and n is the integer more than 1;The calculation formula of preset model isWherein, wxFor default underlying parameter, wiFor the affecting parameters for i-th things that things is concentrated, xiFor
The on off state parameter that i-th things that things is concentrated currently is set, biFor the evaluation result for i-th things that things is concentrated
Numerical value.
Further, output result further includes the desired value of predetermined action, according to output output control robot according to
After behavior emotion model performs predetermined action, this method further includes:Receive the feedback result that predetermined action is performed to robot;
According to feedback result and desired value update preset model.
Further, control instruction includes at least one of:Image control instructs;Phonetic control command;Bio signal
Control instruction.
To achieve these goals, according to another aspect of the present invention, a kind of robot motion control device is additionally provided,
The device includes:First receiving unit, for receiving control instruction, wherein, control instruction, which is used to indicate robot and performs, to be made a reservation for
Action;Input unit for control instruction to be inputted preset model, is exported as a result, wherein, output result includes and machine
The corresponding behavior emotion model of predetermined action that people performs, Affective Evaluation parameter instruction of the preset model at least based on user feedback
It gets;Control unit, for performing predetermined action according to behavior emotion model according to output output control robot.
Further, behavior emotion model is for determining behavior affective state of the robot to predetermined action, in robot
When performing predetermined action, the state of robot is behavior affective state.
Further, which further includes:Second receiving unit for control instruction to be inputted preset model, obtains
Before exporting result, the corresponding behavior affection index of multiple things that things is concentrated is received, wherein, behavior affection index is used for table
Show the fancy grade for multiple things that robot concentrates things;Unit is established, for the multiple things pair concentrated according to things
The behavior affection index answered establishes preset model.
Further, multiple things that things is concentrated are the corresponding things of predetermined action that robot performs.
Further, the quantity of things assembling is n, and n is the integer more than 1;The calculation formula of preset model isWherein, wxFor default underlying parameter, wiFor the affecting parameters for i-th things that things is concentrated, xiFor
The on off state parameter that i-th things that things is concentrated currently is set, biFor the evaluation result for i-th things that things is concentrated
Numerical value.
Further, output result further includes the desired value of predetermined action, which further includes:3rd receiving unit is used
In after predetermined action is performed according to behavior emotion model according to output output control robot, receive and robot is performed in advance
Surely the feedback result acted;Updating block, for updating preset model according to feedback result and desired value.
Further, control instruction includes at least one of:Image control instructs;Phonetic control command;Bio signal
Control instruction.
To achieve these goals, according to another aspect of the present invention, a kind of robot is additionally provided, the robot bag
It includes:The robot motion control device of the embodiment of the present invention.
The present invention by receiving control instruction, wherein, control instruction be used to indicate robot perform predetermined action;It will control
Instruction input preset model is exported as a result, wherein, output result include with the predetermined action that robot performs compared with
Behavior emotion model, Affective Evaluation parameter training of the preset model at least based on user feedback obtain;According to output output control
Robot performs predetermined action according to behavior emotion model, and solving robot behavior performance single can not meet user demand
Problem, and then needed to show the effect of different behaviors according to user.
Description of the drawings
The attached drawing for forming the part of the application is used for providing a further understanding of the present invention, schematic reality of the invention
Example and its explanation are applied for explaining the present invention, is not constituted improper limitations of the present invention.In the accompanying drawings:
Fig. 1 is the flow chart of robot motion control method according to a first embodiment of the present invention;
Fig. 2 is the flow chart of robot motion control method according to a second embodiment of the present invention;And
Fig. 3 is the schematic diagram of robot motion control device according to embodiments of the present invention.
Specific embodiment
It should be noted that in the case where there is no conflict, the feature in embodiment and embodiment in the application can phase
Mutually combination.The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
In order to which those skilled in the art is made to more fully understand application scheme, below in conjunction in the embodiment of the present application
The technical solution in the embodiment of the present application is clearly and completely described in attached drawing, it is clear that described embodiment is only
The embodiment of the application part, instead of all the embodiments.Based on the embodiment in the application, ordinary skill people
Member's all other embodiments obtained without making creative work should all belong to the model of the application protection
It encloses.
It should be noted that term " first " in the description and claims of this application and above-mentioned attached drawing, "
Two " etc. be the object for distinguishing similar, without being used to describe specific order or precedence.It should be appreciated that it so uses
Data can exchange in the appropriate case, so as to embodiments herein described herein.In addition, term " comprising " and " tool
Have " and their any deformation, it is intended that cover it is non-exclusive include, for example, containing series of steps or unit
Process, method, system, product or equipment are not necessarily limited to those steps or unit clearly listed, but may include without clear
It is listing to Chu or for the intrinsic other steps of these processes, method, product or equipment or unit.
An embodiment of the present invention provides a kind of robot motion control methods.
Fig. 1 is the flow chart of robot motion control method according to a first embodiment of the present invention, as shown in Figure 1, the party
Method comprises the following steps:
Step S102:Control instruction is received, wherein, control instruction is used to indicate robot and performs predetermined action;
Step S104:Control instruction is inputted into preset model, is exported as a result, wherein, output result includes and machine
The corresponding behavior emotion model of predetermined action that people performs, Affective Evaluation parameter instruction of the preset model at least based on user feedback
It gets.
Step S106:Predetermined action is performed according to behavior emotion model according to output output control robot.
The embodiment using receive control instruction, control instruction is then inputted into preset model, exported as a result, according to
It exports output control robot and performs predetermined action according to behavior emotion model, since preset model is the feelings based on user feedback
Sense evaluating is trained, and behavior emotion model when robot performs predetermined action is can determine, therefore, according to model
Output output control robot according to behavior emotion model perform predetermined action can solve robot behavior performance it is single can not
The problem of meeting user demand, and then needed to show the effect of different behaviors according to user.
The robot of the embodiment of the present invention can be intelligent Service humanoid robot, accompany humanoid robot etc., optionally, control
Instruction includes at least one in image control instruction, phonetic control command, bio signal control instruction, and control instruction can also
It is other kinds of instruction, receiving control instruction, can be robot detect action that owner sends by camera, receives
The image control arrived instructs or receives phonetic control command, for example, receiving " dancing " that owner sends to robot
Phonetic control command, robot can also receive wearable bio signal control instruction, for example, some biologies of owner's wearing
The equipment of signal detection detects that owner's physical condition exception occurs and sends control instruction, and robot receives bio signal control
Instruction.Control instruction control robot performs predetermined action, and predetermined action can be that robot such as dances, sweeps the floor, turn-taking at the actions,
After control instruction is received, control instruction is inputted into preset model, carries out model calculating, exported as a result, according to
The output output control robot arrived performs predetermined action according to behavior emotion.Pass through the technical solution of the embodiment of the present invention, machine
Device people can have a variety of emotion models and user interaction.
Optionally, control instruction is inputted into preset model, before obtaining output result, receives multiple things that things is concentrated
The corresponding behavior affection index of object, wherein, behavior affection index is used to represent the happiness for multiple things that robot concentrates things
Good degree;Preset model is established according to the corresponding behavior affection index of multiple things that things is concentrated.
The foundation of preset model needs the corresponding index of things that things is concentrated, wherein, multiple things that things is concentrated can
To be the things for the behavior emotion that can reflect robot, for example, singing and dancing, the things such as sweep the floor, the things pair that things is concentrated
The index answered can be fancy grade of the user to the things, for example, index is higher when user relatively takes notice of some things,
When user less takes notice of some things, index is than relatively low.After the corresponding index of multiple things of things concentration is received, root
Preset model is established according to corresponding behavior affection index.Behavior emotion model is used to determine behavior feelings of the robot to predetermined action
Sense state, when robot performs predetermined action, the state of robot is behavior affective state, and behavior affective state can be with
Behavior emotion model is corresponding, under a kind of behavior emotion model, can have a kind of behavior affective state.
Optionally, output result includes the desired value of predetermined action, according to exporting output control robot according to behavior
After emotion performs predetermined action, the feedback result of the action to robot is received;It is pre- according to feedback result and desired value update
If model.
The output result that control instruction input preset model obtains is included to the desired value of predetermined action, the phase of predetermined action
Prestige value can be used to indicate that the desired value of behavior emotion when robot performs predetermined action, for example, robot perform it is predetermined
Whether it is happy execution during action, after robot performs predetermined action according to behavior emotion, receives and robot is moved
The feedback result of work, feedback result are sent by user, for example, owner gives a mark to the behavior of robot, are received to machine
Then the feedback result of the action of device people updates preset model according to the desired value of feedback result and predetermined action, for example, can be with
Preset model is updated according to the difference of feedback result and the desired value of predetermined action.
Optionally, the quantity of things assembling is n, wherein, n is the integer more than 1, the calculation formula of preset model
ForWherein, wxFor default underlying parameter, wiFor the affecting parameters for i-th things that things is concentrated, xi
For the on off state parameter that i-th things that things is concentrated currently is set, biFor the evaluation knot for i-th things that things is concentrated
Fruit numerical value.
The calculation formula of preset model isWherein, wxIt can be one for default underlying parameter
Definite value is set, w by user setting or before dispatching from the factoryiFor the affecting parameters for i-th things that things is concentrated, the shadow of i-th things
Sound parameter can be i-th things to the significance level of owner, xiThe switch being currently set for i-th things that things is concentrated
State parameter, xiWith two values, when taking the first numerical value, for example, xi=1, represent influence of i-th things for owner
State is opens, when taking second value, for example, xi=0, influence state of i-th things for owner is represented to close.
When influence state of i-th things for owner is closes, robot will not generate shadow to the behavior emotion of the things to owner
It rings.biCan be row when owner performs the things for robot for the evaluation result numerical value for i-th things that things is concentrated
For the marking of emotion.
In an optional application scenarios, robot is after the phonetic order of owner " sweeping the floor " is received, according to pre-
If behavior emotion of model when obtaining sweeping the floor, action of sweeping the floor then is performed, if the behavior emotion model that preset model exports
For happy pattern, then robot show it is happy sweep the floor, for example, action is brisk, while plays music etc., owner is to robot
Behavior emotion feel quite pleased, marking it is higher, then robot according to owner marking may determine that, owner likes showing when sweeping the floor
Therefore the happy behavior affective state gone out, when robot sweeps the floor next time, still shows happy behavior emotion, such as
Fruit owner has been weary of the happy behavior emotion that robot is shown when sweeping the floor, then makes low score to robot, robot according to
Marking judges that owner does not like the happy behavior affective state shown when sweeping the floor, then when sweeping the floor next time, shows
Unhappy behavior affective state.
Fig. 2 is the flow chart of robot motion control method according to a second embodiment of the present invention, which can make
For the preferred embodiment of above-mentioned first embodiment, as shown in Fig. 2, the robot motion control method comprises the following steps:
Step S201:Receive control instruction.
Control instruction can be visual input signal (Vision input), language in-put (Language input) and can
Dress one or more of bio signal input (Wearable biosignal input).
Step S202:Online strengthening is trained.
Online strengthening training (Soul Model Reinforcement Learning Core) can be by control instruction
It inputs default model to be calculated, be exported as a result, the behavior for determining robot according to output result exports.It is also, pre-
If model can be updated according to the feedback of rewards and punishments mechanism, to realize that online strengthening is trained.
Step S203:Behavior exports.
After output result is obtained according to preset model, exported according to the behavior of output output control robot
(Behavior output)。
Step S204:User's evaluation is fed back.
User's evaluation feedback (Human evaluation feedback) is being received, for example, row of the user to robot
For marking.
Step S205:Feedback, rewards and punishments mechanism.
By introducing rewards and punishments mechanism feedback (Feedback Reward/punishment block), preset model is carried out
It corrects, realizes online strengthening training.
Step S206:Desired value.
Desired value (Expectation) is obtained according to preset model, it would be desirable to which value and user feedback evaluation are as feedback prize
The foundation of mechanism is punished, can be using the difference of user feedback evaluation and desired value as feedback parameter.
The technical solution of the embodiment of the present invention employs the principle of behaviour psychology in principle, is applied earliest in Ba Fuluo
In the experiments such as the dog of husband, Skinner box, biologically also there is the support of neural plasticity scheduling theory.But the extensive chemical of the prior art
The training pattern for not touching emotion is practised, does not also have similar realization method in robot, rests on computer vision, nature
The degree such as Language Processing, the embodiment of the present invention are put forward for the first time structure and the realization of emotion model.
The present invention relates to machine life how to be allowed preferably to carry out emotion communication with the mankind, online intensified learning is used
It adds in the positive negative reinforcement of psychology concept and positive and negative punishment is trained for a long time, it is therefore intended that user interacted with robot
Cheng Zhongneng experiences the emotion feedback close to the mankind.
The reinforcing punishment theory of behaviour psychology covers positive negative reinforcement and positive and negative punishment.It is simply to introduce below:
1. positive reinforcement:Give a kind of good stimulation.In order to establish a kind of behavior pattern of adaptability, by way of reward,
Repeat this behavior pattern, and remain behind.Such as enterprise is to actively proposing that the worker of conductive suggestion issues bonus.
2. negative reinforcement:Remove a bad stimulation.Appearance to trigger desirable behavior is set up.Such as enterprise does not allow
Personal call is beaten between at work, as soon as an employee has this custom, being scolded occurs in this behavior one, once but he stop
This behavior should just stop the censure to him immediately.
3. positive punishment:Apply a bad stimulation.This is a kind of method that punishment is given when unsuitable behavior occurs.
4. negative punishment:Remove a good stimulation.It is this punishment than it is positive punishment it is more commonly used.When unsuitable behavior occurs
When, no longer give original reward.
Basic thought based on intensified learning, backfeed loop are reward driving (reward-driven, referred to as RL),
It is expected that wx+b can be expressed as, b is expressed as reward parameter, judges assignment by evaluation function.But psychology thinks the row of the mankind
It is that emotion derives from two aspects, is to get well to stimulate with badly stimulating respectively, the behavior of people can be understood as plural good stimulation and plural number
The bad multiply-add effect stimulated, but existing RL does not add in this kind of classification mechanism, and the technical solution of the embodiment of the present invention adds
Enter this mechanism, and applied to human-computer interaction robot.Therefore the expectational model of RL is wx+(w1x1+b1)+(w2x2+b2)+……。
Dimension is extended on the basis of original intensified learning.On the dimension newly increased, a part is derived from presetting.According to public affairs
Formula, w1x1+b1The expectation Contribution Model being expressed as under the things, w1For definite value, designer is defaulted in, is conceptually equal to people
Class is to the fancy grade of certain things, and b1It is represented as the rewards and punishments parameter of the things.If it is a kind of things to mop floor, for child
Detest things is probably for son, then w1For negative, x1The on off state for representing the things for 0 or 1, b1It represents current
The rewards and punishments degree of things.Things collection can increase, but need to preset the fancy grade of the things.The expectation tribute of comprehensive all things
Model is offered, then can obtain the rewards and punishments degree of the individual current point in time, is fed back in the individual input.The change of model is direct
The training method of the model is affected, though human-computer interaction robot can use advance trained metastasis model, people
It is always that This is what people generally disapprove of in terms of propertyization, therefore trained metastasis model makes into the technical solution of the embodiment of the present invention in advance by this
On-line study mode allows user to train the robot in person, and the behavior pattern of the robot is made gradually to be close to the users, final to service
In user.
It should be noted that step shown in the flowchart of the accompanying drawings can be in such as a group of computer-executable instructions
It is performed in computer system, although also, show logical order in flow charts, it in some cases, can be with not
The order being same as herein performs shown or described step.
An embodiment of the present invention provides a kind of robot motion control device, which can be used for performing implementation of the present invention
The robot motion control method of example.
Fig. 3 is the schematic diagram of robot motion control device according to embodiments of the present invention, as shown in figure 3, the device bag
It includes:
First receiving unit 10, for receiving control instruction, wherein, the control instruction is used to indicate robot and performs in advance
Fixed action;
Input unit 20 for the control instruction to be inputted preset model, is exported as a result, wherein, exporting result
The corresponding behavior emotion model of predetermined action including being performed with robot, the preset model is at least based on user feedback
Affective Evaluation parameter training obtains;
Control unit 30 performs for the robot according to the output output control according to the behavior emotion model
The predetermined action.
The embodiment uses the first receiving unit, for receiving control instruction, wherein, control instruction is used to indicate robot
Perform predetermined action;Input unit for control instruction to be inputted preset model, is exported as a result, wherein, preset model is used
Behavior emotion model when determining that robot performs predetermined action, Affective Evaluation ginseng of the preset model at least based on user feedback
Number training obtains;Control unit, for performing predetermined action according to behavior emotion model according to output output control robot, from
And user demand can not be met by solving the problems, such as that robot behavior performance is single, and then according to user need to show
The effect of different behaviors.
Optionally, behavior emotion model is held for determining behavior affective state of the robot to predetermined action in robot
During row predetermined action, the state of robot is behavior affective state.
Optionally, which further includes:Second receiving unit for the control instruction to be inputted preset model, obtains
To before output result, the corresponding behavior affection index of multiple things that things is concentrated is received, wherein, the behavior affection index
For representing the fancy grade of multiple things that the robot concentrates the things;Unit is established, for according to the thing
The corresponding behavior affection index of multiple things that object is concentrated establishes the preset model.
Optionally, multiple things that the things is concentrated are the corresponding thing of the predetermined action that the robot performs
Object.
Optionally, the quantity of things assembling is n, and n is that the calculation formula of the integer preset model more than 1 isWherein, wxFor default underlying parameter, wiFor the affecting parameters for i-th things that things is concentrated, xiFor
The on off state parameter that i-th things that things is concentrated currently is set, biFor the evaluation result for i-th things that things is concentrated
Numerical value.
Optionally, output result includes the desired value of predetermined action, which further includes:3rd receiving unit, for
After performing predetermined action according to behavior emotion model according to output output control robot, the anti-of action to robot is received
Present result;Updating block, for updating preset model according to feedback result and desired value.
Optionally, control instruction includes at least one of:Image control instructs;Phonetic control command;Bio signal control
System instruction.
The embodiment of the present invention additionally provides a kind of robot, which includes the robot motion control of the embodiment of the present invention
Device processed.
In the above embodiment of the present invention, all emphasize particularly on different fields to the description of each embodiment, do not have in some embodiment
The part of detailed description may refer to the associated description of other embodiment.
Obviously, those skilled in the art should be understood that each module of the above-mentioned present invention or each step can be with general
Computing device realize that they can concentrate on single computing device or be distributed in multiple computing devices and be formed
Network on, optionally, they can be realized with the program code that computing device can perform, it is thus possible to which they are stored
In the storage device by computing device come perform either they are fabricated to respectively each integrated circuit modules or by they
In multiple modules or step be fabricated to single integrated circuit module to realize.In this way, the present invention is not limited to any specific
Hardware and software combines.
The foregoing is only a preferred embodiment of the present invention, is not intended to limit the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.Within the spirit and principles of the invention, that is made any repaiies
Change, equivalent substitution, improvement etc., should all be included in the protection scope of the present invention.
Claims (15)
1. a kind of robot motion control method, which is characterized in that including:
Control instruction is received, wherein, the control instruction is used to indicate robot and performs predetermined action;
The control instruction is inputted into preset model, is exported as a result, wherein, the output result includes and the robot
The corresponding behavior emotion model of the predetermined action performed, the preset model at least Affective Evaluation based on user feedback
Parameter training obtains;
The predetermined action is performed according to the behavior emotion model according to robot described in the output output control.
2. according to the method described in claim 1, it is characterized in that, the behavior emotion model is used to determine the robot pair
The behavior affective state of the predetermined action, when the robot performs the predetermined action, the state of the robot is
The behavior affective state.
3. according to the method described in claim 1, it is characterized in that, by the control instruction input preset model, obtain defeated
Go out before result, the method further includes:
The corresponding behavior affection index of multiple things that things is concentrated is received, wherein, the behavior affection index is used to represent institute
State the fancy grade for multiple things that robot concentrates the things;
The preset model is established according to the corresponding behavior affection index of multiple things that the things is concentrated.
4. according to the method described in claim 3, it is characterized in that, multiple things that the things is concentrated are held for the robot
The capable corresponding things of the predetermined action.
5. according to the method described in claim 3, it is characterized in that, the quantity of the things assembling is n, n is more than 1
Integer;The calculation formula of the preset model is
Wherein, wxFor default underlying parameter, wiFor the affecting parameters for i-th things that the things is concentrated, xiFor the things
The on off state parameter that i-th things concentrated currently is set, biFor the evaluation result for i-th things that the things is concentrated
Numerical value.
6. according to the method described in claim 1, it is characterized in that, the output result further includes the expectation of the predetermined action
Value, after the robot according to the output output control performs the predetermined action according to the behavior emotion model,
The method further includes:
Receive the feedback result that the predetermined action is performed to the robot;
The preset model is updated according to the feedback result and the desired value.
7. according to the method described in claim 1, it is characterized in that, the control instruction includes at least one of:
Image control instructs;
Phonetic control command;
Bio signal control instruction.
8. a kind of robot motion control device, which is characterized in that including:
First receiving unit, for receiving control instruction, wherein, the control instruction is used to indicate robot and performs predetermined move
Make;
Input unit for the control instruction to be inputted preset model, is exported as a result, wherein, the output result bag
The corresponding behavior emotion model of the predetermined action performed with the robot is included, the preset model is at least based on user
The Affective Evaluation parameter training of feedback obtains;
Control unit, it is described pre- according to behavior emotion model execution for the robot according to the output output control
Fixed action.
9. device according to claim 8, which is characterized in that the behavior emotion model is used to determine the robot pair
The behavior affective state of the predetermined action, when the robot performs the predetermined action, the state of the robot is
The behavior affective state.
10. device according to claim 8, which is characterized in that described device further includes:
Second receiving unit for the control instruction to be inputted preset model, before obtaining output result, receives things collection
In the corresponding behavior affection index of multiple things, wherein, the behavior affection index is for representing the robot to described
The fancy grade for multiple things that things is concentrated;
Unit is established, the corresponding behavior affection index of multiple things for being concentrated according to the things is established described default
Model.
11. device according to claim 10, which is characterized in that multiple things that the things is concentrated are the robot
The corresponding things of the predetermined action performed.
12. device according to claim 10, which is characterized in that the quantity of the things assembling is n, and n is big
In 1 integer;The calculation formula of the preset model is
Wherein, wxFor default underlying parameter, wiFor the affecting parameters for i-th things that the things is concentrated, xiFor the things
The on off state parameter that i-th things concentrated currently is set, biFor the evaluation result for i-th things that the things is concentrated
Numerical value.
13. device according to claim 8, which is characterized in that the output result further includes the phase of the predetermined action
Prestige value, described device further include:
3rd receiving unit, for being performed in the robot according to the output output control according to the behavior emotion model
After the predetermined action, the feedback result that the predetermined action is performed to the robot is received;
Updating block, for updating the preset model according to the feedback result and the desired value.
14. device according to claim 8, which is characterized in that the control instruction includes at least one of:
Image control instructs;
Phonetic control command;
Bio signal control instruction.
15. a kind of robot, which is characterized in that control dress including the robot motion any one of claim 8 to 14
It puts.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611069128.9A CN108115678B (en) | 2016-11-28 | 2016-11-28 | Robot and motion control method and device thereof |
PCT/CN2017/092038 WO2018095041A1 (en) | 2016-11-28 | 2017-07-06 | Robot, and action control method and device therefor |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611069128.9A CN108115678B (en) | 2016-11-28 | 2016-11-28 | Robot and motion control method and device thereof |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108115678A true CN108115678A (en) | 2018-06-05 |
CN108115678B CN108115678B (en) | 2020-10-23 |
Family
ID=62194768
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611069128.9A Active CN108115678B (en) | 2016-11-28 | 2016-11-28 | Robot and motion control method and device thereof |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN108115678B (en) |
WO (1) | WO2018095041A1 (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109074502A (en) * | 2018-07-26 | 2018-12-21 | 深圳前海达闼云端智能科技有限公司 | Method, apparatus, storage medium and the robot of training artificial intelligence model |
CN110576433A (en) * | 2018-06-08 | 2019-12-17 | 香港商女娲创造股份有限公司 | robot motion generation method |
CN113001539A (en) * | 2019-12-19 | 2021-06-22 | 本田技研工业株式会社 | Robot control device, robot control method, and recording medium |
CN114578720A (en) * | 2020-12-01 | 2022-06-03 | 合肥欣奕华智能机器股份有限公司 | Control method and control system |
CN116935497A (en) * | 2023-09-19 | 2023-10-24 | 广州中鸣数码科技有限公司 | Game control method and device |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112861804A (en) * | 2021-03-17 | 2021-05-28 | 上海庞勃特科技有限公司 | Human body action evaluation method, evaluation device and evaluation system |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020052672A1 (en) * | 1999-05-10 | 2002-05-02 | Sony Corporation | Robot and control method |
US20020103576A1 (en) * | 1999-05-10 | 2002-08-01 | Sony Corporation | Robot and its control method |
CN1463215A (en) * | 2001-04-03 | 2003-12-24 | 索尼公司 | Leg type moving robot, its motion teaching method and storage medium |
JP2004098252A (en) * | 2002-09-11 | 2004-04-02 | Ntt Docomo Inc | Communication terminal, control method of lip robot, and control device of lip robot |
CN103456314A (en) * | 2013-09-03 | 2013-12-18 | 广州创维平面显示科技有限公司 | Emotion recognition method and device |
CN104350541A (en) * | 2012-04-04 | 2015-02-11 | 奥尔德巴伦机器人公司 | Robot capable of incorporating natural dialogues with a user into the behaviour of same, and methods of programming and using said robot |
CN105930374A (en) * | 2016-04-12 | 2016-09-07 | 华南师范大学 | Emotion robot conversation method and system based on recent feedback, and robot |
CN105945949A (en) * | 2016-06-01 | 2016-09-21 | 北京光年无限科技有限公司 | Information processing method and system for intelligent robot |
CN105988591A (en) * | 2016-04-26 | 2016-10-05 | 北京光年无限科技有限公司 | Intelligent robot-oriented motion control method and intelligent robot-oriented motion control device |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101014321B1 (en) * | 2009-02-24 | 2011-02-14 | 한국전자통신연구원 | Method for emotion recognition based on Minimum Classification Error |
CN101618280B (en) * | 2009-06-30 | 2011-03-23 | 哈尔滨工业大学 | Humanoid-head robot device with human-computer interaction function and behavior control method thereof |
KR20110002757A (en) * | 2009-07-02 | 2011-01-10 | 삼성전자주식회사 | Emotion model device, apparatus and method for adaptive learning personality of emotion model |
US9117168B2 (en) * | 2012-09-28 | 2015-08-25 | Korea Institute Of Industrial Technology | Apparatus and method for calculating internal state for artificial emotion |
CN105260745A (en) * | 2015-09-30 | 2016-01-20 | 西安沧海网络科技有限公司 | Information push service system capable of carrying out emotion recognition and prediction based on big data |
CN105807933B (en) * | 2016-03-18 | 2019-02-12 | 北京光年无限科技有限公司 | A kind of man-machine interaction method and device for intelligent robot |
-
2016
- 2016-11-28 CN CN201611069128.9A patent/CN108115678B/en active Active
-
2017
- 2017-07-06 WO PCT/CN2017/092038 patent/WO2018095041A1/en active Application Filing
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020052672A1 (en) * | 1999-05-10 | 2002-05-02 | Sony Corporation | Robot and control method |
US20020103576A1 (en) * | 1999-05-10 | 2002-08-01 | Sony Corporation | Robot and its control method |
CN1463215A (en) * | 2001-04-03 | 2003-12-24 | 索尼公司 | Leg type moving robot, its motion teaching method and storage medium |
JP2004098252A (en) * | 2002-09-11 | 2004-04-02 | Ntt Docomo Inc | Communication terminal, control method of lip robot, and control device of lip robot |
CN104350541A (en) * | 2012-04-04 | 2015-02-11 | 奥尔德巴伦机器人公司 | Robot capable of incorporating natural dialogues with a user into the behaviour of same, and methods of programming and using said robot |
CN103456314A (en) * | 2013-09-03 | 2013-12-18 | 广州创维平面显示科技有限公司 | Emotion recognition method and device |
CN105930374A (en) * | 2016-04-12 | 2016-09-07 | 华南师范大学 | Emotion robot conversation method and system based on recent feedback, and robot |
CN105988591A (en) * | 2016-04-26 | 2016-10-05 | 北京光年无限科技有限公司 | Intelligent robot-oriented motion control method and intelligent robot-oriented motion control device |
CN105945949A (en) * | 2016-06-01 | 2016-09-21 | 北京光年无限科技有限公司 | Information processing method and system for intelligent robot |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110576433A (en) * | 2018-06-08 | 2019-12-17 | 香港商女娲创造股份有限公司 | robot motion generation method |
CN110576433B (en) * | 2018-06-08 | 2021-05-18 | 香港商女娲创造股份有限公司 | Robot motion generation method |
CN109074502A (en) * | 2018-07-26 | 2018-12-21 | 深圳前海达闼云端智能科技有限公司 | Method, apparatus, storage medium and the robot of training artificial intelligence model |
WO2020019252A1 (en) * | 2018-07-26 | 2020-01-30 | 深圳前海达闼云端智能科技有限公司 | Artificial intelligence model training method and device, storage medium and robot |
CN113001539A (en) * | 2019-12-19 | 2021-06-22 | 本田技研工业株式会社 | Robot control device, robot control method, and recording medium |
CN114578720A (en) * | 2020-12-01 | 2022-06-03 | 合肥欣奕华智能机器股份有限公司 | Control method and control system |
CN114578720B (en) * | 2020-12-01 | 2023-11-07 | 合肥欣奕华智能机器股份有限公司 | Control method and control system |
CN116935497A (en) * | 2023-09-19 | 2023-10-24 | 广州中鸣数码科技有限公司 | Game control method and device |
CN116935497B (en) * | 2023-09-19 | 2024-01-05 | 广州中鸣数码科技有限公司 | Game control method and device |
Also Published As
Publication number | Publication date |
---|---|
WO2018095041A1 (en) | 2018-05-31 |
CN108115678B (en) | 2020-10-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108115678A (en) | Robot and its method of controlling operation and device | |
KR102296906B1 (en) | Virtual character generation from image or video data | |
Komatsu et al. | Adaptation gap hypothesis: How differences between users’ expected and perceived agent functions affect their subjective impression | |
CN106503654A (en) | A kind of face emotion identification method based on the sparse autoencoder network of depth | |
JPWO2005093650A1 (en) | Will expression model device, psychological effect program, will expression simulation method | |
CN108335345A (en) | The control method and device of FA Facial Animation model, computing device | |
JP2002018146A (en) | Interactive toy, reaction behavior generator and reaction behavior pattern generation method | |
CN106445147A (en) | Behavior management method and device of conversational system based on artificial intelligence | |
CN108364550B (en) | Experience type interactive process design method for extra-high voltage electric virtual training | |
EP3276542A1 (en) | Processing system and program | |
CN109858574B (en) | Autonomous learning method and system for intelligent agent for man-machine cooperative work | |
CN101140636B (en) | System and method for generating robot genome | |
CN106022294A (en) | Intelligent robot-oriented man-machine interaction method and intelligent robot-oriented man-machine interaction device | |
CN111582457A (en) | Method for realizing general machine intelligence by simulating human memory | |
Li et al. | Facial feedback for reinforcement learning: a case study and offline analysis using the TAMER framework | |
CN107526831A (en) | A kind of natural language processing method and apparatus | |
Khashman | Blood cell identification using emotional neural networks. | |
CN111563575B (en) | Machine intelligent implementation method simulating human intelligence | |
CN110125955A (en) | Control device and machine learning device | |
Hafez et al. | Improving robot dual-system motor learning with intrinsically motivated meta-control and latent-space experience imagination | |
CN109389417A (en) | Simulate reaction bonusing method, device and the server of participant | |
CN114003643A (en) | Robot emotion expression system and method | |
CN106489114A (en) | A kind of generation method of robot interactive content, system and robot | |
CN110826510A (en) | Three-dimensional teaching classroom implementation method based on expression emotion calculation | |
CN110008321A (en) | Information interacting method and device, storage medium and electronic device |
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 |