CN108917769A - A kind of adaptive grating map creating method of robot based on nine fork trees - Google Patents
A kind of adaptive grating map creating method of robot based on nine fork trees Download PDFInfo
- Publication number
- CN108917769A CN108917769A CN201810772043.XA CN201810772043A CN108917769A CN 108917769 A CN108917769 A CN 108917769A CN 201810772043 A CN201810772043 A CN 201810772043A CN 108917769 A CN108917769 A CN 108917769A
- Authority
- CN
- China
- Prior art keywords
- grid
- map
- robot
- sub
- probability
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
- Manipulator (AREA)
Abstract
The invention discloses a kind of adaptive grating map creating methods of robot based on nine fork trees, entire environmental map is divided into nine grids structure by this method, the state of judgement wherein each grid, which belongs to, to be fully taken up, partially occupies or vacant, the grid partially occupied is further subdivided into the nine grids of small level-one, judge that the state of these small grids belongs to again fully take up, partially occupy or vacant, above-mentioned cutting procedure is repeated, until entire map search finishes and meets required precision.Adaptive feature is presented independent of artificial experience in the determination of final grating map scale.This method and uniform scale Grid Method are compared, the simulation experiment result shows：The map structuring algorithm not only has convergence, but also compared with uniform scale grating map, dramatically saves memory space.
Description
Technical field
It is based on nine fork trees in computer science the invention belongs to the method for indoor mobile robot environmental map creation
Thought carries out the division and expression of environmental map, and this method shows very strong superiority in the creation of large scale environment.
Background technique
In recent years, indoor mobile robot medical treatment, food and drink, transport, in terms of play it is more and more important
Effect, especially attacking along with artificial intelligence upsurge, sweeping robot, education a batch novel product such as star NAO robot
The visual field for swarming into us, is paid close attention to more and more widely.In this field, the map building problem of robot is indoor shifting
The hot and difficult issue of mobile robot research.
There are three types of the representation methods of robot environment's map：Geometrical measurers, topological approach and Grid Method.Geometrical characteristic map
Environment is indicated with related geometrical characteristic (such as point, straight line, face).This method can search shortest path, currently, common processing
Method is to extract linear feature or geometrical characteristic from laser radar data to describe environment, and data volume to be processed is smaller, but
As barrier increases, the time cost of search is also greatly increased, and the search particularly with round barrier may fail.It opens up
The figure that falls to the ground is indoor environment to be expressed as the topology diagram with node and associated wiring.Composition is quickly succinct, but is difficult to create
Large scale map is built, for similar or symmetrical scene undiscerning.Grid Method is suitable for various complex environments, builds in realization
Mould, update and processing aspect are even more to be better than geometrical measurers and topological approach, and therefore, this method is quite by scholar and researcher in recent years
Favor.The thought of grating map is the grid that entire working environment is divided into uniform scale, and each grid is pointed out wherein
A possibility that there are barriers.Although Grid Method shows many advantages in terms of robot environment's cognition and map structuring,
When constructing large scale environmental map, there are still obvious deficiencies for it, i.e., as body of a map or chart expands and the promotion of raster resolution,
Data volume is in exponential increase, which will consume great memory space, and brings sizable computation complexity,
This makes general CPU processor be difficult to cope with.Therefore, the existing necessity of the massive storage space of grating map is solved the problems, such as,
It is again significant.
Under the premise of retaining the intrinsic superiority of grating map, reduces the consumption to memory space, that is, be directed to
Uniform scale grating map occupies massive storage space problem during large scale map building, proposes a kind of based on nine fork trees
The adaptive grating map creating algorithm of robot.
Summary of the invention
The present invention proposes a kind of adaptive grating map creating algorithm of robot based on nine fork trees.Based on computer science
In nine fork tree thoughts environmental map is divided into a certain size grid, vacant, part is belonged to according to the state of grid every time
Occupy or fully take up, then carry out subdivision gradually deepening, be divided into the nine grids of small level-one, until entire map only have it is vacant
With until fully taking up two states.The invention solves the uniform scale grating map of tradition during large scale map building
It is quick-fried to avoid data caused by expansion and raster resolution because of search range improve for the problem of occupying massive storage space
It is fried.Adaptive grating map creating algorithm used in the present invention and traditional uniform scale grating map creating algorithm are carried out
Comparison, the simulation experiment result shows that the algorithm is not only with convergence, and dramatically saves memory space.
The present invention is achieved by the following technical solutions, and the present invention includes the following steps：
A kind of adaptive grating map creating algorithm of robot based on nine fork trees, it is characterised in that include the following steps：
It is considered as the map of a Status unknown, robot that the first step, robot, which need the entire circumstances not known recognized,
Around a circle in entire circumstances not known, entire circumstances not known is divided into nine grids structure, that is, by map cutting be uniform
Nine sub- grids.
Second step, the state for determining every sub- grid in map, the state of which grid be vacant or be occupied, which
It is Status unknown.Need to calculate every sub- grid herein there are the probability of barrier, probability is that 1 expression occupies, and indicates empty for 0
It sets, is that the number between 0~1 indicates Status unknowns, the sub- grid for continuing to Status unknown is divided.
Third step updates map, i.e., will be occupied to use respectively with vacant sub- grid 1 and 0 indicates that remaining grid temporary is general
Rate value indicates.Whether judgement meets the division condition that terminates (half of the interpolation scale less than or equal to robot length and width smaller value at this time
When, it does not continue to divide, be occupied at this point, the grid of Status unknown is also considered as, unified in map to be indicated with 1), if satisfied,
Patterning process terminates, and final map only 1 and 0 two states, entire map will become known state from unknown state；If no
Meet, enters step four.
The sub- grid of each Status unknown is regarded as a unknown sub- map by the 4th step respectively, repeats above procedure,
The method cutting map for pressing step 1, indicates map by the method for step 2, updates map by the method for step 3.
In the first step：
It is considered as the map of a Status unknown that robot, which needs the entire circumstances not known that recognizes, and robot is not entirely
Know around a circle in environment, entire circumstances not known is divided into nine grids structure, that is, by map cutting is uniform nine sons
Grid.The calculation method of the centre coordinate (x, y) of every sub- grid is as follows：
The length and width of environmental map scale indicate that the number that grid is obtained after n times subdivision is with a, b respectively
X1X2X3...Xn(n), i chooses from 0,1,2,3,4,5,6,7,8.Work as X1,X2,X3,...,XnWhen taking 0,3,6, λ1,λ2,
λ3,....,λn=-1；Work as X1,X2,X3,...,XnWhen taking 1,4,7, λ1,λ2,λ3,....,λn=0；Work as X1,X2,X3,...,XnIt takes
2,5,8 when, λ1,λ2,λ3,....,λn=1.Work as X1,X2,X3,...,XnWhen taking 0,1,2,Work as X1,X2,
X3,...,XnWhen taking 3,4,5,Work as X1,X2,X3,...,XnWhen taking 6,7,8,This
Grid number up to 4 in embodiment.After father's grid of Status unknown is split, Position Number mode such as Fig. 1 of sub- grid
It is shown.
In second step：
Determine that the state of every sub- grid in map, the state of which grid are vacant or are occupied which is state
Unknown.Need to calculate every sub- grid herein there are the probability of barrier, probability is that 1 expression occupies, and indicates vacant for 0, is 0
Number between~1 indicates Status unknowns, and the sub- grid for continuing to Status unknown is divided.There is barrier in i-th of sub- grid
Hinder the method for calculating probability of object as follows：
p{xi(s)=OCC | XZ(s), D (s) }=
p{xi(s)=OCC | XZ(s),d(s)}×
p{xi(s)=OCC | XZ(s),φ(s)}/
[1+2×p{xi(s)=OCC | XZ(s),d(s)}×
p{xi(s)=OCC | XZ(s),φ(s)}-
p{xi(s)=OCC | XZ(s),d(s)}-
p{xi(s)=OCC | XZ(s),φ(s)}]
Wherein,
The pose X that robot motion walks to sz(s) it indicates, diFor the distance between the center of robot and grid i,
φiFor the angle between robot coordinate system's x-axis and the center of grid i.The distance range of robot sensor detection is [dmin,
dmax], angular range is [φmin,φmax].If xiInitial shape probability of state is p { xi=OCC }=p { xi=EMP }=0.5, OCC
Expression fully takes up, and EMP indicates vacant, i.e., in the initial state, the probability that grid is occupied and is not occupied is 0.5.Machine
When device people moves to s step, sensor reading D (s)=[d (s), φ (s)]T, d (s) indicate sensor apart from return value and
φ (s) indicates sensor angles return value.
In third step：
Map is updated, s step lattice-shaped probability of state, benefit are calculated based on current robot pose and sensor observation
S-1 is updated with it and walks trellis states could probability value, finally obtains s step trellis states could probability value.Method is as follows：
After indicating map rejuvenation, pose that robot motion to s walks；DsAfter indicating map rejuvenation, robot motion
The sensor reading walked to s；After indicating map rejuvenation, pose that robot motion to s-1 walks；Ds-1Indicate map more
After new, sensor reading that robot motion to s-1 walks.
In 4th step：
It regards the sub- grid of each Status unknown as a unknown sub- map respectively, repeats above procedure, i.e., by step
Rapid one method cutting map, indicates map by the method for step 2, updates map with the method for step 3.It is that search is received below
Holding back property proves：
When the occupied probability of grid is more than or equal to pmaxWhen, it is believed that the grid is occupied, and use 1 indicates；When grid is occupied
Probability be less than or equal to pminWhen, it is believed that the grid is sky, and use 0 indicates.When interpolation scale is smaller less than or equal to robot length and width
When the half of value, no longer segment down.Under this condition, if the occupied probability of grid of minimum level-one is greater than 0, then it is assumed that the grid
Lattice are occupied, and use 1 indicates；If the occupied probability of grid of minimum level-one is equal to 0, it is believed that the grid is sky, and use 0 indicates.Root
According to above-mentioned analysis it is found that the environmental map of algorithm building has the condition for clearly terminating to divide, hence it is demonstrated that adaptive
The convergence of Raster.
The principle of the present invention is：The basic thought of adaptive grating map creating algorithm derived from nine fork trees, but with it is general
Nine fork trees are different, which must comply with two basic principles：(1) all nodes are in total there are three types of state, vacant, account for completely
According to partially occupy；(2) each part occupies node and will continue to be divided into 9 child nodes down, vacant and fully take up section
Point is leaf node, can not be divided again.According to above-mentioned nine forks tree principle, a kind of adaptive grating map is created, the specific steps are：
It is considered as the map of a Status unknown that robot, which needs the entire circumstances not known recognized, and robot is in entire circumstances not known
Around a circle, entire circumstances not known is divided into nine grids structure, that is, by map cutting is uniform nine sub- grids；It determines
The state of every sub- grid, the state of which grid are vacant or are occupied which is Status unknown, need herein in map
Every sub- grid is calculated there are the probability of barrier, probability is that 1 expression occupies, and indicates vacant for 0, is the number table between 0~1
Show Status unknown, the sub- grid for continuing to Status unknown is divided；Map is updated, i.e., will be occupied and vacant sub- grid
It is indicated respectively with 1 and 0, remaining grid temporary probability value indicates.Whether judgement meets end division condition at this time, and (interpolation scale is small
When being equal to the half of robot length and width smaller value, does not continue to divide, be occupied at this point, the grid of Status unknown is also considered as,
It is unified in map to be indicated with 1), if satisfied, patterning process terminates, final map will only have 1 and 0 two states, entirely
Figure becomes known state from unknown state；If not satisfied, entering in next step；The sub- grid of each Status unknown is regarded as respectively
One unknown sub- map repeats above procedure, that is, presses the method cutting map of step 1, indicates ground by the method for step 2
Figure updates map with the method for step 3.
Detailed description of the invention
Fig. 1, adaptive grating map creating process.
Fig. 2, STATIC SIMULATION environment, grey parts are barrier.
Fig. 3, uniform scale Grid Method composition.
Fig. 4, adaptive Grid Method composition.
Specific embodiment
Elaborate below to the embodiment of the present invention, the present embodiment premised on technical solution of the present invention under carry out it is real
It applies, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to following embodiments.
Embodiment：
The first step, it is considered as the map of a Status unknown, robot that robot, which needs the entire circumstances not known recognized,
Around a circle in entire circumstances not known, entire circumstances not known is divided into nine grids structure, that is, by map cutting be uniform
Nine sub- grids.The calculation method of the centre coordinate (x, y) of every sub- grid is as follows：
The length and width of environmental map scale indicate that the number that grid is obtained after n times subdivision is with a, b respectively
X1X2X3...Xn(n), i chooses from 0,1,2,3,4,5,6,7,8.Work as X1,X2,X3,...,XnWhen taking 0,3,6, λ1,λ2,
λ3,....,λn=-1；Work as X1,X2,X3,...,XnWhen taking 1,4,7, λ1,λ2,λ3,....,λn=0；Work as X1,X2,X3,...,XnIt takes
2,5,8 when, λ1,λ2,λ3,....,λn=1.Work as X1,X2,X3,...,XnWhen taking 0,1,2,Work as X1,X2,
X3,...,XnWhen taking 3,4,5,Work as X1,X2,X3,...,XnWhen taking 6,7,8,This
Grid number up to 4 in embodiment.After father's grid of Status unknown is split, Position Number mode such as Fig. 1 of sub- grid
It is shown.
Second step determines that the state of every sub- grid in map, the state of which grid are vacant or are occupied, which
It is Status unknown.Need to calculate every sub- grid herein there are the probability of barrier, probability is that 1 expression occupies, and indicates empty for 0
It sets, is that the number between 0~1 indicates Status unknowns, the sub- grid for continuing to Status unknown is divided.In i-th of sub- grid
There are the method for calculating probability of barrier is as follows：
p{xi(s)=OCC | XZ(s), D (s) }=
p{xi(s)=OCC | XZ(s),d(s)}×
p{xi(s)=OCC | XZ(s),φ(s)}/
[1+2×p{xi(s)=OCC | XZ(s),d(s)}×
p{xi(s)=OCC | XZ(s),φ(s)}-
p{xi(s)=OCC | XZ(s),d(s)}-
p{xi(s)=OCC | XZ(s),φ(s)}]
Wherein,
The pose X that robot motion walks to sz(s) it indicates, diFor the distance between the center of robot and grid i,
φiFor the angle between robot coordinate system's x-axis and the center of grid i.The distance range of robot sensor detection is [dmin,
dmax], angular range is [φmin,φmax].If xiInitial shape probability of state is p { xi=OCC }=p { xi=EMP }=0.5, OCC
Expression fully takes up, and EMP indicates vacant, i.e., in the initial state, the probability that grid is occupied and is not occupied is 0.5.Machine
When device people moves to s step, sensor reading D (s)=[d (s), φ (s)]T, d (s) indicate sensor apart from return value and
φ (s) indicates sensor angles return value.
Third step updates map, i.e., will be occupied to use respectively with vacant sub- grid 1 and 0 indicates that remaining grid temporary is general
Rate value indicates.Whether judgement meets the division condition that terminates (half of the interpolation scale less than or equal to robot length and width smaller value at this time
When, it does not continue to divide, be occupied at this point, the grid of Status unknown is also considered as, unified in map to be indicated with 1), if satisfied,
Patterning process terminates, and final map only 1 and 0 two states, entire map will become known state from unknown state；If no
Meet, enters step four.S step lattice-shaped probability of state, benefit are calculated based on current robot pose and sensor observation
S-1 is updated with it and walks trellis states could probability value, finally obtains s step trellis states could probability value.Method is as follows：
After indicating map rejuvenation, pose that robot motion to s walks；DsAfter indicating map rejuvenation, robot motion
The sensor reading walked to s；After indicating map rejuvenation, pose that robot motion to s-1 walks；Ds-1Indicate map more
After new, sensor reading that robot motion to s-1 walks.
4th step regards the sub- grid of each Status unknown as a unknown sub- map respectively, repeats above procedure,
The method cutting map for pressing step 1, indicates map by the method for step 2, updates map with the method for step 3.It is below
Search convergence proves：
When the occupied probability of grid is more than or equal to pmaxWhen, it is believed that the grid is occupied, and use 1 indicates；When grid is occupied
Probability be less than or equal to pminWhen, it is believed that the grid is sky, and use 0 indicates.When interpolation scale is smaller less than or equal to robot length and width
When the half of value, no longer segment down.Under this condition, if the occupied probability of grid of minimum level-one is greater than 0, then it is assumed that the grid
Lattice are occupied, and use 1 indicates；If the occupied probability of grid of minimum level-one is equal to 0, it is believed that the grid is sky, and use 0 indicates.Root
According to above-mentioned analysis it is found that the environmental map of algorithm building has the condition for clearly terminating to divide, hence it is demonstrated that adaptive
The convergence of Raster.
Implementation result
Assuming that the static indoor environment that simulated environment is 50 × 50, as shown in Figure 2.Emulation experiment is in Turtlebot2 machine
It is carried out in people's analogue system, robot diameter is 42cm.The data collection task of motion model is completed by odometer and gyroscope,
And the data collection task of measurement model is completed by RPLIDAR-A2360 ° of laser radar scanner.The height of laser radar is
40.5mm, width 72.5mm, ranging range with 0.15m~8m be it is best, 0 °~360 ° of scanning angle, resolution of ranging<
0.5mm, 0.9 ° of angular resolution, single ranging time 0.25ms, measurement frequency>=4000Hz, scan frequency 10Hz, scanning one
The case where frequency representative value in week is single pass lucky 400 sampled points.
Assuming that may be implemented quickly and accurately to position under this simulated environment, covariance Q (s)=diag of motion artifacts
{0.12,0.12,0.022,(1.0×π/360)2,...,(1.0×π/360)2, covariance R (s)=diag of observation noise
{0.022,(0.2×π/360)2}。
Simulated environment of the adaptive Raster to Fig. 2 for being utilized respectively the uniform scale Grid Method of tradition and pitching tree based on nine
It is patterned, simulation result comparison is as shown in Figure 3,4.
The uniform scale Grid Method of tradition and the adaptive Raster Simulation result data set based on nine forks are counted such as 1 institute of table
Show.
| Occupy memory space | Runing time | |
| Uniform Grid Method | 617M | 0.88s |
| Adaptive Grid Method | 4.76M | 1.92s |
Claims (5)
1. a kind of adaptive grating map creating method of robot based on nine fork trees, which is characterized in that include the following steps：
It is considered as the map of a Status unknown that the first step, robot, which need the entire circumstances not known recognized, and robot is whole
Around a circle in a circumstances not known, entire circumstances not known is divided into nine grids structure, that is, by map cutting is uniform nine
A sub- grid；
Second step, the state for determining every sub- grid in map, the state of which grid are vacant or are occupied which is shape
State is unknown；Needing to calculate every sub- grid herein there are the probability of barrier, probability is that 1 expression occupies, indicate vacant for 0,
Status unknowns are indicated for the number between 0~1, and the sub- grid for continuing to Status unknown is divided；
Third step updates map, i.e., will be occupied to use respectively with vacant sub- grid 1 and 0 indicates, remaining grid temporary probability value
It indicates；When whether judgement meets the half for terminating division condition i.e. interpolation scale less than or equal to robot length and width smaller value at this time,
It does not continue to divide, be occupied at this point, the grid of Status unknown is also considered as, it is unified in map to be indicated with 1, if satisfied, composition
Process terminates, and final map only 1 and 0 two states, entire map will become known state from unknown state；If discontented
Foot, into the 4th step；
The sub- grid of each Status unknown is regarded as a unknown sub- map by the 4th step respectively, is repeated above procedure, that is, is pressed
The method cutting map of the first step, indicates map by the method for second step, updates map by the method for third step.
2. a kind of adaptive grating map creating method of robot based on nine fork trees according to claim 1, feature
It is, in the first step：
It is considered as the map of a Status unknown that robot, which needs the entire circumstances not known recognized, and robot is in entire unknown ring
Around a circle in border, entire circumstances not known is divided into nine grids structure, that is, by map cutting is uniform nine sub- grids；
The calculation method of the centre coordinate (x, y) of every sub- grid is as follows：
The length and width of environmental map scale indicate that the number that grid is obtained after n times subdivision is X with a, b respectively1X2X3...Xn, n
Position, i choose from 0,1,2,3,4,5,6,7,8；Work as X1,X2,X3,...,XnWhen taking 0,3,6, λ1,λ2,λ3,....,λn=-1；When
X1,X2,X3,...,XnWhen taking 1,4,7, λ1,λ2,λ3,....,λn=0；Work as X1,X2,X3,...,XnWhen taking 2,5,8, λ1,λ2,
λ3,....,λn=1；Work as X1,X2,X3,...,XnWhen taking 0,1,2,Work as X1,X2,X3,...,XnTake 3,4,
When 5,Work as X1,X2,X3,...,XnWhen taking 6,7,8,Grid number up to 4
Position.
3. a kind of adaptive grating map creating method of robot based on nine fork trees according to claim 1, feature
It is, in second step：
Determine that the state of every sub- grid in map, the state of which grid are vacant or are occupied which is Status unknown
's；Need to calculate every sub- grid herein there are the probability of barrier, probability is that 1 expression occupies, and indicates vacant for 0, is 0~1
Between number indicate Status unknowns, the sub- grid for continuing to Status unknown is divided；There are barriers in i-th of sub- grid
Method for calculating probability it is as follows：
p{xi(s)=OCC | XZ(s), D (s) }=
p{xi(s)=OCC | XZ(s),d(s)}×
p{xi(s)=OCC | XZ(s),φ(s)}/
[1+2×p{xi(s)=OCC | XZ(s),d(s)}×
p{xi(s)=OCC | XZ(s),φ(s)}-
p{xi(s)=OCC | XZ(s),d(s)}-
p{xi(s)=OCC | XZ(s),φ(s)}]
Wherein,
The pose X that robot motion walks to sz(s) it indicates, diFor the distance between the center of robot and grid i, φiFor
Angle between robot coordinate system's x-axis and the center of grid i；The distance range of robot sensor detection is [dmin,dmax],
Angular range is [φmin,φmax]；If xiInitial shape probability of state is p { xi=OCC }=p { xi=EMP }=0.5, OCC expression
It fully takes up, EMP indicates vacant, i.e., in the initial state, the probability that grid is occupied and is not occupied is 0.5；Robot
When moving to s step, sensor reading D (s)=[d (s), φ (s)]T, d (s) indicate sensor apart from return value and φ (s)
Indicate sensor angles return value.
4. a kind of adaptive grating map creating method of robot based on nine fork trees according to claim 1, feature
It is, in third step：
Map is updated, s step lattice-shaped probability of state is calculated based on current robot pose and sensor observation, utilizes it
S-1 step trellis states could probability value is updated, s step trellis states could probability value is finally obtained；Method is as follows：
After indicating map rejuvenation, pose that robot motion to s walks；DsAfter indicating map rejuvenation, robot motion to s
The sensor reading of step；After indicating map rejuvenation, pose that robot motion to s-1 walks；Ds-1After indicating map rejuvenation,
The sensor reading that robot motion walks to s-1.
5. a kind of adaptive grating map creating method of robot based on nine fork trees according to claim 1, feature
It is, in the 4th step：
It regards the sub- grid of each Status unknown as a unknown sub- map respectively, repeats above procedure, that is, press step 1
Method cutting map, indicate map by the method for second step, update map with the method for third step；It is search convergence below
It proves：
When the occupied probability of grid is more than or equal to pmaxWhen, it is believed that the grid is occupied, and use 1 indicates；When grid is occupied general
Rate is less than or equal to pminWhen, it is believed that the grid is sky, and use 0 indicates；When interpolation scale is less than or equal to robot length and width smaller value
When half, no longer segment down；Under this condition, if the occupied probability of grid of minimum level-one is greater than 0, then it is assumed that the grid quilt
It occupies, use 1 indicates；If the occupied probability of grid of minimum level-one is equal to 0, it is believed that the grid is sky, and use 0 indicates；According to upper
Analysis is stated it is found that the environmental map of building has the condition for clearly terminating to divide, hence it is demonstrated that adaptive Raster
Convergence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810772043.XA CN108917769A (en) | 2018-07-13 | 2018-07-13 | A kind of adaptive grating map creating method of robot based on nine fork trees |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810772043.XA CN108917769A (en) | 2018-07-13 | 2018-07-13 | A kind of adaptive grating map creating method of robot based on nine fork trees |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108917769A true CN108917769A (en) | 2018-11-30 |
Family
ID=64411989
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810772043.XA Pending CN108917769A (en) | 2018-07-13 | 2018-07-13 | A kind of adaptive grating map creating method of robot based on nine fork trees |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108917769A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109528095A (en) * | 2018-12-28 | 2019-03-29 | 深圳市愚公科技有限公司 | It sweeps the floor calibration method, sweeping robot and the mobile terminal of record figure |
| CN109631919A (en) * | 2018-12-28 | 2019-04-16 | 芜湖哈特机器人产业技术研究院有限公司 | A kind of hybrid navigation map constructing method for merging reflector and occupying grid |
| CN110673947A (en) * | 2019-08-12 | 2020-01-10 | 江苏博人文化科技有限公司 | Method for reducing memory required by laser slam graph building |
| CN111329398A (en) * | 2020-03-27 | 2020-06-26 | 上海高仙自动化科技发展有限公司 | Robot control method, robot, electronic device, and readable storage medium |
| WO2021003958A1 (en) * | 2019-07-09 | 2021-01-14 | 苏州科瓴精密机械科技有限公司 | Method for creating and system for creating raster map |
| CN113377097A (en) * | 2021-01-25 | 2021-09-10 | 杭州易享优智能科技有限公司 | Path planning and obstacle avoidance method for blind person guide |
| CN113885531A (en) * | 2021-11-05 | 2022-01-04 | 上海肇观电子科技有限公司 | Method for moving robot, circuit, medium, and program |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102506881A (en) * | 2011-10-28 | 2012-06-20 | 雷雅雯 | Method for rapidly and accurately locating by utilizing relative coordinate |
| CN104634352A (en) * | 2015-03-02 | 2015-05-20 | 吉林大学 | Road matching method based on fusion of probe vehicle movement track and electronic map |
-
2018
- 2018-07-13 CN CN201810772043.XA patent/CN108917769A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102506881A (en) * | 2011-10-28 | 2012-06-20 | 雷雅雯 | Method for rapidly and accurately locating by utilizing relative coordinate |
| CN104634352A (en) * | 2015-03-02 | 2015-05-20 | 吉林大学 | Road matching method based on fusion of probe vehicle movement track and electronic map |
Non-Patent Citations (3)
| Title |
|---|
| JING ZHOU: ""Adaptive GridMap Building Algorithm for Mobile Robot Based on Multiway Tree"", 《INTERNATIONAL CONFERENCE ON ELECTRICAL ENGINEERING, CONTROL AND ROBOTICS (EECR 2018)》 * |
| 胡引翠等: "基于"自适应九叉树"空间信息多级格网的城市管理模式 ", 《地理与地理信息科学》 * |
| 郭利进等: "基于四叉树的自适应栅格地图创建算法", 《控制与决策》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109528095A (en) * | 2018-12-28 | 2019-03-29 | 深圳市愚公科技有限公司 | It sweeps the floor calibration method, sweeping robot and the mobile terminal of record figure |
| CN109631919A (en) * | 2018-12-28 | 2019-04-16 | 芜湖哈特机器人产业技术研究院有限公司 | A kind of hybrid navigation map constructing method for merging reflector and occupying grid |
| CN109631919B (en) * | 2018-12-28 | 2022-09-30 | 芜湖哈特机器人产业技术研究院有限公司 | Hybrid navigation map construction method integrating reflector and occupied grid |
| WO2021003958A1 (en) * | 2019-07-09 | 2021-01-14 | 苏州科瓴精密机械科技有限公司 | Method for creating and system for creating raster map |
| WO2021003959A1 (en) * | 2019-07-09 | 2021-01-14 | 苏州科瓴精密机械科技有限公司 | Method for updating and system for updating raster map |
| CN110673947A (en) * | 2019-08-12 | 2020-01-10 | 江苏博人文化科技有限公司 | Method for reducing memory required by laser slam graph building |
| CN110673947B (en) * | 2019-08-12 | 2022-04-05 | 江苏博人文化科技有限公司 | Method for reducing memory required by laser slam graph building |
| CN111329398A (en) * | 2020-03-27 | 2020-06-26 | 上海高仙自动化科技发展有限公司 | Robot control method, robot, electronic device, and readable storage medium |
| CN113377097A (en) * | 2021-01-25 | 2021-09-10 | 杭州易享优智能科技有限公司 | Path planning and obstacle avoidance method for blind person guide |
| CN113885531A (en) * | 2021-11-05 | 2022-01-04 | 上海肇观电子科技有限公司 | Method for moving robot, circuit, medium, and program |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108917769A (en) | A kind of adaptive grating map creating method of robot based on nine fork trees | |
| WO2020134082A1 (en) | Path planning method and apparatus, and mobile device | |
| Wu et al. | Plant phenotyping by deep-learning-based planner for multi-robots | |
| KR101912233B1 (en) | Target positioning method and system | |
| CN104298734B (en) | Mobile updating method for change pattern spots in land use status change survey | |
| CN101510225B (en) | STL model boolean operation method of products | |
| CN106774314A (en) | A kind of home-services robot paths planning method based on run trace | |
| CN105868582B (en) | Using drosophila optimization method identification of protein compound | |
| CN105716604A (en) | Mobile robot indoor positioning method and system based on geomagnetic sequences | |
| CN109190704A (en) | The method and robot of detection of obstacles | |
| CN106548484A (en) | Product model dispersion point cloud Boundary characteristic extraction method based on two-dimentional convex closure | |
| CN104853434A (en) | Indoor positioning method based on SVM and K mean value clustering algorithm | |
| CN110986956B (en) | Autonomous learning global positioning method based on improved Monte Carlo algorithm | |
| Rusu et al. | Perception for mobile manipulation and grasping using active stereo | |
| CN109886155A (en) | Man power single stem rice detection localization method, system, equipment and medium based on deep learning | |
| CN108256060A (en) | A kind of closed loop detection method, device, terminal and storage medium | |
| CN103236043B (en) | A kind of plant organ point cloud restoration method | |
| CN107918953A (en) | The extracting method and device of laser scanning power line point cloud based on three dimensions | |
| CN113505646B (en) | Target searching method based on semantic map | |
| Kumar et al. | A unified grid-based wandering pattern detection algorithm | |
| CN101783016A (en) | Crown appearance extract method based on shape analysis | |
| Rodrigues et al. | Clustering of architectural floor plans: A comparison of shape representations | |
| CN112356031A (en) | On-line planning method based on Kernel sampling strategy under uncertain environment | |
| CN107123138B (en) | Based on vanilla-R point to the point cloud registration method for rejecting strategy | |
| Han et al. | A double branch next-best-view network and novel robot system for active object reconstruction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181130 |