CN106774327B - A kind of robot path planning method and device - Google Patents
A kind of robot path planning method and device Download PDFInfo
- Publication number
- CN106774327B CN106774327B CN201611209442.2A CN201611209442A CN106774327B CN 106774327 B CN106774327 B CN 106774327B CN 201611209442 A CN201611209442 A CN 201611209442A CN 106774327 B CN106774327 B CN 106774327B
- Authority
- CN
- China
- Prior art keywords
- moving target
- robot
- leg
- moving
- map
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0214—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with safety or protection criteria, e.g. avoiding hazardous areas
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0221—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving a learning process
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0231—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0257—Control of position or course in two dimensions specially adapted to land vehicles using a radar
Abstract
The invention discloses a kind of robot path planning method and devices, human bioequivalence is carried out by the tracking to moving target, obtain position and movement velocity of the moving target relative to robot, position and movement velocity to the moving target according to acquisition relative to robot, the position of predicted motion target next step;And this map is moved into relative to what the position of robot and movement velocity, the position of next step and the first Gaussian function obtained moving target according to moving target;It is merged the waiting cost map of the moving target of acquisition, the cost map for moving into this map and acquired barrier to obtain main cost map;According to the main cost map and default rule, the path of robot is planned.Therefore, so not only can be with the position of predicted motion target next step, and the path of robot can be planned according to the motion profile of moving target, so that robot can be not influencing the movable under the premise of carry out activity of people.
Description
Technical field
The present invention relates to robot field more particularly to a kind of robot path planning methods and device.
Background technique
With the development of robot technology, robot participates in human lives in a manner of directly or indirectly can not keep away
Exempt from, when people and robot take action in space simultaneously, the movement track that robot only knows about people can just avoid people, so that
Robot can not influence people it is movable under the premise of activity.
But the running track of people is indefinite, is difficult to predict the movement track of people, the method provided in the prior art,
Robot can not be made to plan the motion path of itself according to the running track of people, so that robot can not influence people
Carry out activity under the premise of movable.
Summary of the invention
In view of this, the invention discloses a kind of robot path planning method and device, solve in the prior art without
The problem of motion path of the Motion trajectory robot of method foundation people, so that robot can not influence people's activity
Under the premise of carry out activity.
A kind of robot path planning method disclosed by the embodiments of the present invention, comprising:
The body and leg of moving target are tracked, the body data and leg data for respectively obtaining tracking input
Into preset machine learning model, the first position of body and the second position of leg are obtained;The machine learning model is
The data stood and walked based on human body are trained;
The second position of the first position of obtained body and leg is merged to obtain the position of moving target;It is described
The position for target of doing exercises is position of the moving target relative to robot;
Fortune is predicted according to the position of the moving target and by the movement velocity of the moving target obtained to body tracking
The position of moving-target next step;
According to the position of the moving target, movement velocity, the position of next step and the first Gaussian function, movement is generated
Target moves into this map;First Gaussian function is the function that moving direction upward peak is incremented by;
Data and the second Gaussian function when the moving target obtained when according to tracking moving object is static generate fortune
The waiting cost map of moving-target;Second Gaussian function is circular Gaussian function;
By the waiting cost map of the moving target, move into the cost map of this map and acquired barrier into
Row fusion obtains main cost map;
According to the main cost map and default rule, the path of robot is planned.
Optionally, the second position of the first position of obtained body and leg is merged to obtain the position of moving target
It sets, comprising:
According to the weight of body and leg, the second position of the first position of obtained body and leg merge
To the position of moving target.
Optionally, according to the main cost map and default rule, path planning is carried out to robot, comprising:
Obtain the road conditions in the main cost map;
When the road is clear, order robot constant speed passes through;
When cost is relatively low for road, the deceleration of order robot passes through;
When road cost is medium, order robot stops action;
When road higher cost, the opposite direction of moving target described in order Robot Selection passes through.
Optionally, the body of moving target and leg are tracked, comprising:
The body of moving target is tracked by the camera being mounted in robot;
The leg of moving target is tracked by the laser radar sensor being mounted in robot.
Optionally, the machine learning model is convolutional neural networks learning model.
The embodiment of the invention also discloses a kind of robot path planning's device, described device includes:
Tracing unit, for moving target body and leg be tracked, respectively will the obtained body data of tracking
It is input in preset machine learning model with leg data, obtains the first position of body and the second position of leg;It is described
Machine learning model is that the data stood and walked based on human body are trained;
First integrated unit is transported for being merged the second position of the first position of obtained body and leg
The position of moving-target;The position of the moving target is position of the moving target relative to robot;
Predicting unit, the speed for the position according to the moving target and the moving target by being obtained to body tracking
Spend the position of predicted motion target next step;
First generation unit, for high according to the position of the moving target, movement velocity, next step position and first
This function generates the cost map of moving target;First Gaussian function is the function that moving direction upward peak is incremented by;
Second generation unit, the data and second when the moving target obtained when for according to tracking moving object is static
The waiting cost map of Gaussian function generation moving target;Second Gaussian function is circular Gaussian function;
Second integrated unit, for by the waiting cost map of the moving target, move into this map and acquired
The cost map of barrier is merged to obtain main cost map;
Path planning unit, for being advised to the path of robot according to the main cost map and default rule
It draws.
Optionally, first integrated unit includes:
First fusion subelement, for the weight according to body and leg, by the first position and leg of obtained body
The second position merged to obtain the position of moving target.
Optionally, path planning unit, comprising:
Subelement is obtained, for obtaining the road conditions in the main cost map;
First order subelement, for when the road is clear, order robot constant speed to pass through;
Second order subelement, for when cost is relatively low for road, the deceleration of order robot to pass through;
Third order subelement, for when road cost is medium, order robot to stop action;
4th order subelement is used for the negative side of moving target described in order Robot Selection when road higher cost
To passing through.
Optionally, tracing unit, comprising:
First tracking subelement, chases after for body of the camera by being mounted in robot to moving target
Track;
Second tracking subelement, for the laser radar sensor by being mounted in robot to the leg of moving target
It is tracked.
Optionally, the machine learning model is convolutional neural networks learning model.
A kind of robot path planning method and device disclosed by the embodiments of the present invention, by the tracking to moving target into
Row human bioequivalence obtains position and movement velocity of the moving target relative to robot, thus according to the moving target phase of acquisition
Position and movement velocity for robot, the position of predicted motion target next step;And according to moving target relative to machine
What the position of people and movement velocity, the position of next step and the first Gaussian function obtained moving target moves into this map;It will
The waiting cost map of the moving target of acquisition moves into this map and the cost map of acquired barrier merge
To main cost map;According to the main cost map and default rule, the path of robot is planned.Therefore, in this way
Not only can be with the position of predicted motion target next step, and it can be according to the motion profile of moving target to the path of robot
It is planned, so that robot can be not influencing the movable under the premise of carry out activity of people.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
The embodiment of invention for those of ordinary skill in the art without creative efforts, can also basis
The attached drawing of offer obtains other attached drawings.
Fig. 1 shows a kind of flow diagram of paths planning method provided in an embodiment of the present invention;
Fig. 2 shows a kind of structural schematic diagrams of path planning apparatus provided in an embodiment of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
With reference to Fig. 1, a kind of flow diagram of robot path planning method of the embodiment of the present invention is shown.In this implementation
In example, the method may include:
S101: being tracked the body and leg of moving target, the body data and leg number for respectively obtaining tracking
According to being input in preset machine learning model, the first position of body and the second position of leg are obtained;The machine learning
Model is that the data stood and walked based on human body are trained.
In the present embodiment, since the camera shooting being mounted in robot can be used during to tracking moving object
Head is tracked the body of moving target, but since the leg of moving target sometimes is possible to be blocked by barrier,
It therefore can also be using the leg of other method tracing movement targets, the present embodiment while tracing movement intended body
In, the leg of moving target is tracked by the laser radar sensor being mounted in robot, acquires moving target
Two-dimensional distance data of the leg relative to robot.
In the present embodiment, the machine learning model is that the data stood and walked based on human body are trained, and is wrapped
An input interface and output interface are included, the input that the body data tracked inputs the machine learning model can be connect
Mouthful, from the first position of the available moving target body of output interface;The data of the leg tracked are input to the machine
The input interface of device learning model obtains the second position of moving target leg from output interface.Wherein for the machine of use
Learning model can use convolutional neural networks learning model.Using the machine learning model, the accurate of output result is improved
Degree.
In the present embodiment, it should be noted that when being tracked to moving target, due to the zone of action of robot tracking
Moving target may be more than to exist, may be with the presence of multiple moving targets, when tracking, can be same to multiple moving targets
When track, obtain the body of multiple moving targets and the data of leg.It is then possible to according to the multiple moving targets tracked
The data of body and leg are separately input in preset machine learning model, obtain the first of the body of each moving target
The second position of position and leg, and the operation of following S102-S107 is carried out respectively.
S102: the second position of the first position of obtained body and leg is merged to obtain the position of moving target
It sets;The position of the moving target is position of the moving target relative to robot.
In the present embodiment, the second position of the first position of the body obtained in the machine learning model and leg is carried out
Fusion, the two can be merged according to different weights, specifically, S102 can specifically include: according to body and leg
Weight is merged the second position of the first position of obtained body and leg to obtain the position of moving target.
For example: the fusion of first position and the second position can based on the average output of two results, i.e., body and
The leg weight shared in fusion is respectively 50%;However, since the data reliability obtained to body tracking is more preferable, it can
A biggish value is arranged in the weight of body, the weight of leg is arranged the value for being less than body weight, when fusion according to
According to the weighted value set, first position and the second position are merged to the position for obtaining moving target relative to robot.
S103: pre- according to the position of the moving target and the movement velocity of the moving target by being obtained to body tracking
Survey the position of moving target next step.
In the present embodiment, the movement velocity of moving target can be obtained when the body to moving target is tracked,
In the present embodiment, the movement velocity of acquisition is a vector, i.e., the movement velocity can both indicate the speed of moving target movement
The value for spending size, can also indicate direction of the moving target relative to robot.
Further, when the position of the next step to moving target is predicted, can predict moving target may
Multiple positions.
For example: when the next step position to moving target is predicted, 10 can be predicted in the range of 1s
The next step position of moving target.
S104: it is generated according to the position of the moving target, movement velocity, the position of next step and the first Gaussian function
Moving target moves into this map;First Gaussian function is the function that moving direction upward peak is incremented by.
In the present embodiment, the first Gaussian function used is peak value for the position where moving target, is successively decreased to both sides, when
When human motion, the mobile direction peak value of moving target is incremented by, and the direction peak value moved before moving target successively decreases.
In the present embodiment, the position of the position of the moving target, movement velocity and moving target next step is expressed as
The parameter of first Gaussian function moves into this map using the first Gaussian function generation moving target.
Wherein, the opposite position for moving into the size in Gaussian function region and moving target and robot in this map of generation
It is set to inverse ratio;I.e. relative position is remoter, then the region influenced is smaller (it can be appreciated that region area is smaller);It gets over relative position
Closely, then the region influenced is bigger (it can be appreciated that the area in region is smaller).In addition to this, Gauss in the cost map of generation
The size in function influences region and the movement velocity of moving target are directly proportional, and movement velocity is bigger, with equidistant influence area
It is smaller, but the influence of movement velocity is limited to moving target and robot is in when moving in the same direction.
S105: data and the second Gaussian function when the moving target obtained when according to tracking moving object is static,
Generate the waiting cost map of moving target;Second Gaussian function is circular Gaussian function.
In the present embodiment, number of moving target when static can be obtained when the body to moving target is tracked
According to waiting cost map can be generated according to data of moving target when static and the second Gaussian function.
S106: by the cost of the waiting cost map of the moving target, the barrier for moving into this map and having obtained
Map is merged to obtain main cost map.
In the present embodiment, the cost map of the barrier can be and be pre-stored in robot, is also possible to
It before S101, is tracked by the environment to surrounding, and data and third Gaussian function according to tracking, acquired disturbance object
Cost map, wherein third Gaussian function can be circular Gaussian function.
In the present embodiment, due to being mentioned above, the moving target of tracking can be waiting costs that are multiple, therefore obtaining
Map and move into this map be also possible to it is multiple, therefore can by the waiting cost map of multiple moving targets and movement cost
The cost map of map and the barrier obtained is merged, and main cost map is obtained.It is main at local in the present embodiment
Figure, when moving target is static, Gaussian Profile figure be it is round, when moving target movement, moving target moving direction at
This consumption value will increase, i.e., moving direction peak value increases, moreover, when distance of the robot apart from moving target is closer or machine
When people is bigger relative to the movement velocity of moving target, consumption value cost is higher.
S107: according to the main cost map and default rule, the path of robot is planned.
In the present embodiment, main cost map can provide robot can with the motion profile of moving target, so as to for
The case where the case where robot offer road, road mentioned here, it can be understood as the phase of moving target with robot in fact
The case where to position and relative velocity.
In the present embodiment, S107 be can specifically include: obtain the road conditions in the main cost map;When the road is clear
When, order robot constant speed passes through;When cost is relatively low for road, the deceleration of order robot passes through;When road higher cost, life
The opposite direction of moving target described in Robot Selection is enabled to pass through.
For example: the road is clear can be understood as moving target and remains static, and gauss of distribution function is round at this time
Shape;When cost is relatively low for road, it can be understood as moving target is kept in motion with the relative position of robot farther out, at this time
Can order robot deceleration pass through;When road cost be it is medium when, it can be understood as moving target be kept in motion and with
The relative position of robot is closer, at this time can order robot stop action;When road higher cost, it can be understood as fortune
Moving-target is kept in motion and very close with the relative position of robot, if robot may if continuing movement at this time
Encounter moving target, thus can the direction of moving target described in order Robot Selection pass through.
In the present embodiment, human bioequivalence is carried out by the tracking to moving target, obtains moving target relative to robot
Position and movement velocity, thus position and movement velocity according to the moving target of acquisition relative to robot, predicted motion
The position of target next step;And according to moving target relative to the position of robot and movement velocity, prediction moving target under
The position of one step and the first Gaussian function, obtain moving target moves into this map;In the process to tracking moving object
In, can also obtain moving target it is static when data, the static state of moving target is generated according to the data and the second Gaussian function
Cost map;By the waiting cost map of the moving target, move into the cost map of this map and acquired barrier
It is merged to obtain main cost map;According to the main cost map and default rule, the path of robot is planned.
Therefore, so not only can be with the position of predicted motion target next step, and it can be according to the motion profile of moving target to machine
The path of device people plans, so that robot can be not influencing the movable under the premise of carry out activity of people.
With reference to Fig. 2, a kind of structural schematic diagram of path planning apparatus of the embodiment of the present invention is shown.In the present embodiment,
Described device for example may include:
Tracing unit 201, for moving target body and leg be tracked, respectively will the obtained body number of tracking
It is input in preset machine learning model according to leg data, obtains the first position of body and the second position of leg;Institute
Stating machine learning model is that the data stood and walked based on human body are trained;
First integrated unit 202, for merge the second position of the first position of obtained body and leg
To the position of moving target;The position of the moving target is position of the moving target relative to robot;
Predicting unit 203, for the position according to the moving target and the moving target by being obtained to body tracking
Prediction of speed moving target next step position;
First generation unit 204, for according to the position of the moving target, movement velocity, next step position and
One Gaussian function generates the cost map of moving target;First Gaussian function is the function that moving direction upward peak is incremented by;
Second generation unit 205, the data when moving target obtained when for according to tracking moving object is static and
The waiting cost map of second Gaussian function generation moving target;Second Gaussian function is circular Gaussian function;
Second integrated unit 206, for by the waiting cost map of the moving target, move into this map and obtained
The cost map of barrier merged to obtain main cost map;
Path planning unit 207, for being carried out to the path of robot according to the main cost map and default rule
Planning.
Optionally, first integrated unit includes:
First fusion subelement, for the weight according to body and leg, by the first position and leg of obtained body
The second position merged to obtain the position of moving target.
Optionally, path planning unit, comprising:
Subelement is obtained, for obtaining the road conditions in the main cost map;
First order subelement, for when the road is clear, order robot constant speed to pass through;
Second order subelement, for when cost is relatively low for road, the deceleration of order robot to pass through;
Third order subelement, for when road cost is medium, order robot to stop action;
4th order subelement is used for the negative side of moving target described in order Robot Selection when road higher cost
To passing through.
Optionally, tracing unit, comprising:
First tracking subelement, chases after for body of the camera by being mounted in robot to moving target
Track;
Second tracking subelement, for the laser radar sensor by being mounted in robot to the leg of moving target
It is tracked.
Optionally, the machine learning model is convolutional neural networks learning model.
Disclosed device through this embodiment carries out human bioequivalence by the tracking to moving target, obtains moving target
Position and movement velocity relative to robot, thus according to the moving target of acquisition relative to the position of robot and movement speed
Degree, the position of predicted motion target next step;And according to moving target relative to the position of robot and movement velocity, next step
Position and the first Gaussian function obtain moving target move into this map;By the waiting cost of the moving target of acquisition
Scheme, move into this map and the cost map of acquired barrier is merged to obtain main cost map;According to it is described it is main at
This map and default rule plan the path of robot.It therefore, so not only can be with predicted motion target in next step
Position, and the path of robot can be planned according to the motion profile of moving target, so that robot can
With not influencing the movable under the premise of carry out activity of people.
It should be noted that all the embodiments in this specification are described in a progressive manner, each embodiment weight
Point explanation is the difference from other embodiments, and the same or similar parts between the embodiments can be referred to each other.
The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention.
Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, of the invention
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest scope of cause.
Claims (10)
1. a kind of robot path planning method, which is characterized in that the described method includes:
The body and leg of moving target are tracked, are respectively input to body data and leg data that tracking obtains pre-
If machine learning model in, obtain the first position of body and the second position of leg;The machine learning model is base
It is trained in the data that human body stands and walks;
The second position of the first position of obtained body and leg is merged to obtain the position of moving target;The movement
The position of moving target is position of the moving target relative to robot;
According to the position of the moving target and the movement velocity predicted motion mesh of the moving target by being obtained to body tracking
Mark the position of next step;
According to the position of the moving target, movement velocity, the position of next step and the first Gaussian function, moving target is generated
Move into this map;First Gaussian function is the function that moving direction upward peak is incremented by;
Data and the second Gaussian function when the moving target obtained when according to tracking moving object is static generate movement mesh
Target waiting cost map;Second Gaussian function is circular Gaussian function;
The waiting cost map of the moving target, the cost map for moving into this map and acquired barrier are melted
Conjunction obtains main cost map;
According to the main cost map and default rule, the path of robot is planned.
2. the method according to claim 1, wherein by the second of the first position of obtained body and leg
Set the position for being merged to obtain moving target, comprising:
According to the weight of body and leg, the second position of the first position of obtained body and leg is merged and is transported
The position of moving-target.
3. the method according to claim 1, wherein according to the main cost map and default rule, to machine
Device people carries out path planning, comprising:
Obtain the road conditions in the main cost map;
When the road is clear, order robot constant speed passes through;
When cost is relatively low for road, the deceleration of order robot passes through;
When road cost is medium, order robot stops action;
When road higher cost, the opposite direction of moving target described in order Robot Selection passes through.
4. the method according to claim 1, wherein the body and leg to moving target are tracked, comprising:
The body of moving target is tracked by the camera being mounted in robot;
The leg of moving target is tracked by the laser radar sensor being mounted in robot.
5. the method according to claim 1, wherein the machine learning model is that convolutional neural networks learn mould
Type.
6. a kind of robot path planning's device, which is characterized in that described device includes:
Tracing unit, for moving target body and leg be tracked, respectively will the obtained body data of tracking and leg
Portion's data are input in preset machine learning model, obtain the first position of body and the second position of leg;The machine
Learning model is that the data stood and walked based on human body are trained;
First integrated unit obtains movement mesh for being merged the second position of the first position of obtained body and leg
Target position;The position of the moving target is position of the moving target relative to robot;
Predicting unit, the speed for the position according to the moving target and the moving target by obtaining to body tracking are pre-
Survey the position of moving target next step;
First generation unit, for according to the position of the moving target, movement velocity, next step position and the first Gaussian function
Number, generates the cost map of moving target;First Gaussian function is the function that moving direction upward peak is incremented by;
Second generation unit, data and the second Gauss when the moving target obtained when for according to tracking moving object is static
The waiting cost map of function generation moving target;Second Gaussian function is circular Gaussian function;
Second integrated unit, for by the waiting cost map of the moving target, move into this map and acquired obstacle
The cost map of object is merged to obtain main cost map;
Path planning unit, for planning the path of robot according to the main cost map and default rule.
7. device according to claim 6, which is characterized in that first integrated unit includes:
First fusion subelement, for the weight according to body and leg, by the of the first position of obtained body and leg
It is merged to obtain the position of moving target in two positions.
8. device according to claim 6, which is characterized in that path planning unit, comprising:
Subelement is obtained, for obtaining the road conditions in the main cost map;
First order subelement, for when the road is clear, order robot constant speed to pass through;
Second order subelement, for when cost is relatively low for road, the deceleration of order robot to pass through;
Third order subelement, for when road cost is medium, order robot to stop action;
4th order subelement, for when road higher cost, the opposite direction of moving target described in order Robot Selection to be logical
It crosses.
9. device according to claim 6, which is characterized in that tracing unit, comprising:
First tracking subelement, is tracked for body of the camera by being mounted in robot to moving target;
Second tracking subelement is carried out for leg of the laser radar sensor by being mounted in robot to moving target
Tracking.
10. device according to claim 6, which is characterized in that the machine learning model is convolutional neural networks study
Model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611209442.2A CN106774327B (en) | 2016-12-23 | 2016-12-23 | A kind of robot path planning method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611209442.2A CN106774327B (en) | 2016-12-23 | 2016-12-23 | A kind of robot path planning method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106774327A CN106774327A (en) | 2017-05-31 |
CN106774327B true CN106774327B (en) | 2019-09-27 |
Family
ID=58920299
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611209442.2A Active CN106774327B (en) | 2016-12-23 | 2016-12-23 | A kind of robot path planning method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106774327B (en) |
Families Citing this family (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
AU2017418043B2 (en) | 2017-07-13 | 2020-05-21 | Beijing Voyager Technology Co., Ltd. | Systems and methods for trajectory determination |
US10395144B2 (en) * | 2017-07-24 | 2019-08-27 | GM Global Technology Operations LLC | Deeply integrated fusion architecture for automated driving systems |
CN107861508B (en) * | 2017-10-20 | 2021-04-20 | 纳恩博(北京)科技有限公司 | Local motion planning method and device for mobile robot |
WO2019095108A1 (en) * | 2017-11-14 | 2019-05-23 | 深圳先进技术研究院 | Robot imitation learning method and apparatus, robot and storage medium |
CN109405843B (en) * | 2018-09-21 | 2020-01-03 | 北京三快在线科技有限公司 | Path planning method and device and mobile device |
CN109144069A (en) * | 2018-09-28 | 2019-01-04 | 徐常柏 | A kind of sweeper charge control method and control system |
CN109167410A (en) * | 2018-09-28 | 2019-01-08 | 徐常柏 | A kind of sweeper charge control method and control system |
US20200160210A1 (en) * | 2018-11-20 | 2020-05-21 | Siemens Industry Software Ltd. | Method and system for predicting a motion trajectory of a robot moving between a given pair of robotic locations |
CN111413957B (en) | 2018-12-18 | 2021-11-02 | 北京航迹科技有限公司 | System and method for determining driving actions in autonomous driving |
CN111796586A (en) * | 2019-03-19 | 2020-10-20 | 北京奇虎科技有限公司 | Local path planning method and device for intelligent mobile equipment |
CN112286049A (en) * | 2019-07-27 | 2021-01-29 | 华为技术有限公司 | Motion trajectory prediction method and device |
CN111089590B (en) * | 2019-12-09 | 2021-10-15 | 泉州装备制造研究所 | Method for tracking human leg by mobile robot through fusion of vision and laser |
CN112748733A (en) * | 2020-12-16 | 2021-05-04 | 广东电网有限责任公司 | Cable drum carriage path planning method, device, equipment and storage medium |
CN112907672B (en) * | 2021-05-07 | 2021-10-08 | 上海擎朗智能科技有限公司 | Robot avoidance method and device, electronic equipment and storage medium |
CN113253256B (en) * | 2021-06-04 | 2021-11-19 | 智道网联科技(北京)有限公司 | Monitoring method and device based on sensor fusion equipment and sensor fusion equipment |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101303726A (en) * | 2008-06-06 | 2008-11-12 | 重庆大学 | System for tracking infrared human body target based on corpuscle dynamic sampling model |
CN101587592A (en) * | 2008-05-21 | 2009-11-25 | 上海新联纬讯科技发展有限公司 | Target motion identification system |
CN103558856A (en) * | 2013-11-21 | 2014-02-05 | 东南大学 | Service mobile robot navigation method in dynamic environment |
JP2014164665A (en) * | 2013-02-27 | 2014-09-08 | Brother Ind Ltd | Autonomous mobile device |
KR101497096B1 (en) * | 2014-02-25 | 2015-03-03 | 경남대학교 산학협력단 | Shopping Assistant Robot for the Visually Impaired |
-
2016
- 2016-12-23 CN CN201611209442.2A patent/CN106774327B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101587592A (en) * | 2008-05-21 | 2009-11-25 | 上海新联纬讯科技发展有限公司 | Target motion identification system |
CN101303726A (en) * | 2008-06-06 | 2008-11-12 | 重庆大学 | System for tracking infrared human body target based on corpuscle dynamic sampling model |
JP2014164665A (en) * | 2013-02-27 | 2014-09-08 | Brother Ind Ltd | Autonomous mobile device |
CN103558856A (en) * | 2013-11-21 | 2014-02-05 | 东南大学 | Service mobile robot navigation method in dynamic environment |
KR101497096B1 (en) * | 2014-02-25 | 2015-03-03 | 경남대학교 산학협력단 | Shopping Assistant Robot for the Visually Impaired |
Non-Patent Citations (2)
Title |
---|
基于随机森林与卡尔曼滤波的人体跟踪方法研究;扈彬;《中国优秀硕士学位论文全文数据库 信息科技辑》;20121031(第10期);第30、41-44页 * |
机器人视觉路径跟踪识别定位优化仿真;姚永一等;《计算机仿真》;20160531;第33卷(第5期);第401-404页 * |
Also Published As
Publication number | Publication date |
---|---|
CN106774327A (en) | 2017-05-31 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106774327B (en) | A kind of robot path planning method and device | |
CN104181926B (en) | The navigation control method of robot | |
Sergiyenko et al. | Data transferring model determination in robotic group | |
CN106873585B (en) | A kind of navigation method for searching, robot and system | |
US20200004266A1 (en) | Method of performing cloud slam in real time, and robot and cloud server for implementing the same | |
Marjovi et al. | Distributed graph-based convoy control for networked intelligent vehicles | |
CN106873599A (en) | Unmanned bicycle paths planning method based on ant group algorithm and polar coordinate transform | |
KR101048098B1 (en) | Robot route planning device and method | |
CN105955262A (en) | Mobile robot real-time layered path planning method based on grid map | |
CN105867365A (en) | Path programming and navigation system based on improved artificial potential field method and method thereof | |
CN109425352A (en) | Self-movement robot paths planning method | |
JP2011161624A (en) | Trajectory planning method, trajectory planning system and robot | |
CN108549403A (en) | A kind of collaborative obstacle avoidance method and device | |
KR20130133778A (en) | Method for animating characters, with collision avoidance based on tracing information | |
CN107662205A (en) | Robot and its joint motions control method and device | |
Nawaz et al. | A study of mobility models for UAV communication networks | |
Sasaki et al. | A3C based motion learning for an autonomous mobile robot in crowds | |
Feng et al. | Decision-making and path planning for highway autonomous driving based on spatio-temporal lane-change gaps | |
Valecce et al. | Robotic‐aided IoT: automated deployment of a 6TiSCH network using an UGV | |
Van Toll et al. | Connecting global and local agent navigation via topology | |
CN109559337A (en) | Positioning track acquisition methods under three-dimensional scenic | |
CN111673729B (en) | Path determination method | |
Ruan et al. | Dynamic cellular learning automata for evacuation simulation | |
Amado et al. | End-to-end deep learning applied in autonomous navigation using multi-cameras system with RGB and depth images | |
Johora et al. | Dynamic path planning and movement control in pedestrian simulation |
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 | ||
CB02 | Change of applicant information |
Address after: 518000 Shekou Torch Pioneering Building, Nanshan District, Shenzhen City, Guangdong Province, 2nd Floor Applicant after: New wisdom Technology Co., Ltd. Address before: 518000 Shekou Torch Pioneering Building, Nanshan District, Shenzhen City, Guangdong Province, 2nd Floor Applicant before: Sino wisdom Machinery Co., Ltd. |
|
CB02 | Change of applicant information | ||
GR01 | Patent grant | ||
GR01 | Patent grant |