CN108896049A - A kind of motion positions method in robot chamber - Google Patents
A kind of motion positions method in robot chamber Download PDFInfo
- Publication number
- CN108896049A CN108896049A CN201810558892.5A CN201810558892A CN108896049A CN 108896049 A CN108896049 A CN 108896049A CN 201810558892 A CN201810558892 A CN 201810558892A CN 108896049 A CN108896049 A CN 108896049A
- Authority
- CN
- China
- Prior art keywords
- robot
- filtering
- moment
- observation
- motion positions
- 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.)
- Withdrawn
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
- G01C21/206—Instruments for performing navigational calculations specially adapted for indoor navigation
Abstract
The invention discloses a kind of motion positions methods in robot chamber, include the following steps：1) encoder data synchronous acquisition step, for acquiring the corner cumulative data of original two driving wheel of left and right；2) substitution of left and right sidesing driving wheel data is calculated cumulative movement distance and current real-time attitude angle according to the kinematics model of the differential trolley of two-wheel by encoder data scaling step；3) with the cumulative movement distance and real-time attitude angle obtained in a sampling period, the displacement increment in current sample period is calculated；4) it according to the displacement increment in step 3), carries out adaptive α β filtering algorithm (a kind of stable state under Kalman filtering algorithm) and calculates, obtain the estimated coordinates of the robot displacement of subsequent time.The present invention provides a kind of robot ambulation localization method that can complete adaptive-filtering, is filtered estimation using adaptive α β filter to reduce model error, inhibits filtering divergence, obtains the current robot high quality coordinates of motion.
Description
Technical field
The present invention relates to automatic control fields, more particularly to a kind of motion positions method in robot chamber.
Background technique
Robot during the motion, needs to monitor the coordinate of its movement constantly, and robot generallys use differential two-wheel and drives
It is dynamic to move, rotary encoder is installed on motor shaft, the real-time angle data of two motors can be exported in real time, it is defeated to acquire the data
Enter into the kinematics model of the differential trolley of two-wheel, obtains the path data and attitude angle information of dandy horse, can calculate
The current world coordinates of trolley, but this group of coordinate can carry certain system noise, measurement noise, random noise.The prior art
In, it there is no a kind of motion positions method in robot chamber, various noises can be eliminated, make during the motion, to can be realized room
The precision of interior positioning.
Therefore those skilled in the art are dedicated to developing a kind of motion positions method in robot chamber, and it is fixed accurately to realize
Position.
Summary of the invention
In view of the above drawbacks of the prior art, technical problem to be solved by the invention is to provide in a kind of robot chamber
Motion positions method can accurately realize indoor positioning.
To achieve the above object, the present invention provides a kind of motion positions method in robot chamber, include the following steps：
1) encoder data synchronous acquisition step, for acquiring the corner cumulative data of original two driving wheel of left and right；
2) encoder data scaling step, according to the kinematics model of the differential trolley of two-wheel, left and right sidesing driving wheel number
Cumulative movement distance and current real-time attitude angle are calculated according to substitution；
3) it with the cumulative movement distance and real-time attitude angle obtained in a sampling period, calculates in current sample period
Displacement increment；
4) it according to the displacement increment in step 3), carries out Kalman filtering calculating or adaptive α, β filtering calculates, obtain
The estimated coordinates of the robot displacement of subsequent time.
Preferably, this method is further comprising the steps of：
5) the code-disc data conversion in step 1) is calculated, obtains distance increment Δ s and yaw angle Δ θ, passes through Descartes
Coordinate conversion, obtains observation coordinate；
6) it calculates and obtains observation；
7) observation coordinate in step 5) and the estimated coordinates in the step 4) are subjected to the statistical parameters meters such as covariance
Point counting analysis, obtains the filtering parameter and gain matrix at current time；
8) by the result bonding state equation of transfer of step 7), coordinate after new filtering is calculated.
Preferably, in the step 5), the observation coordinate of calculating robot according to the following formula：
Δxk=Δ sk·sin(θk(k))
Δyk=Δ sk·cos(θk(k))
Wherein, (Δ xk, Δ yk) be robot observation coordinate.
Preferably, in the step 4), the predictive displacement of the robot of subsequent time is obtained according to the following formula：
Wherein,It is displaced for the robot predicting of subsequent time.
In the step 6), by following state transition equation, the k+1 moment can be released according to the filter result of last moment
Observation：
X (k+1)=Φ (l (k), r (k))+G (k) W (k)
Wherein, Φ is state-transition matrix；
L (k), r (k) are the code-disc signal of left and right two-wheeled respectively；
G (k) is constant matrices G (k)=[T, T2/2]T, T is measurement period；
W (k) is noise signal；
X (k+1) is the observation at k+1 moment.
Preferably, in the step 7), filtering parameter and gain matrix are calculated using α β filter.
Preferably, in the step 7), gain matrix is calculated according to the following formula：
Wherein, K (k+1) is the prediction gain matrix at k+1 moment.
Preferably, it in the step 7), is carried out according to the following steps：
71) judge whether to add up to set periodicity；
72) if not up to set periodicity, calculates parameter, β according to the following equation：
73) filtering parameter is calculated according to following equation：
Wherein,
σvIt is the standard deviation for filtering estimated value and measurement error；
W is zero-mean, i.e. the Gauss of 1 variance divides white noise；
74) gain matrix is calculated according to following equation：
Preferably, it in the step 8), calculates according to the following equation：
Wherein, Y (k+1) is the observation at (k+1) moment；
Transfer matrix H (k)=[1 0]；
It is the estimated value at (k+1) moment；
X (k+1/k+1) is the filter result at (k+1) moment.
The beneficial effects of the invention are as follows：The present invention, which provides one kind, can complete adaptive robot ambulation localization method；
Firstly, the technical program carries out location prediction using adaptive α β filtering algorithm, for common filtering algorithm,
Adaptive α β filtering algorithm can show better tracking and positioning effect when robot target occurs motor-driven.
Also, the present invention, using a kind of adaptive filter method, is utilizing measurement number in exercise data filtering link
While according to Recursive Filtering is carried out, continuous estimation and inaccurate parameter and noise variance matrix in correction model, to reduce mould
Type error inhibits filtering divergence, improves the estimated accuracy of current robot coordinate.
Detailed description of the invention
Fig. 1 is the flow chart of the embodiment of the invention.
Fig. 2 is statistical error calculation flow chart in Fig. 1.
Fig. 3 is adaptive α β filtering signal process flow diagram in Fig. 1.
Specific embodiment
Present invention will be further explained below with reference to the attached drawings and examples：
As shown in Figure 1, a kind of motion positions method in robot chamber, includes the following steps：
1) encoder data synchronous acquisition step, for acquiring the corner cumulative data of original two driving wheel of left and right；
2) encoder data scaling step, according to the kinematics model of the differential trolley of two-wheel, left and right sidesing driving wheel number
Cumulative movement distance and current real-time attitude angle are calculated according to substitution；
3) it with the cumulative movement distance and real-time attitude angle obtained in a sampling period, calculates in current sample period
Displacement increment；
4) it according to the displacement increment in step 4), carries out adaptive α β filtering or Kalman filtering calculates, obtain next
The estimated coordinates of the robot displacement at moment.
The differential robot of two-wheel has unique movement characteristic, when two driving wheels operate in the same direction at the same speed, machine
People will do close to linear uniform motion, and it is more when two driving wheels be that will not meet synchronized rotation condition in the same direction.
When robot target makees linear uniform motion, using basic filtering and prediction technique can position well with
Track target.But in practice, robot target tends to occur motor-driven, and the algorithm based on Kalman filtering can be realized more
Accurate tracking and locating effect.
Further, further comprising the steps of：
5) the code-disc data conversion in step 1) is calculated, obtains distance increment Δ s and yaw angle Δ θ, passes through Descartes
Coordinate conversion, obtains observation coordinate；
5) the code-disc data in step 1) are calculated into observation coordinate at that time according to observational equation；
Observation is input to adaptive α β filter by the observation for obtaining robot motion's information.It carves at the beginning
After system starting, filter is started to work, output the result is that displacement increment after being handled by filtering algorithm.
6) it calculates and obtains observation；
With the cumulative movement distance and real-time attitude angle obtained in a sampling period, calculate in current sample period
Displacement increment, the displacement increment are exactly the observation input value of adaptive α β filter.
7) observation coordinate in step 5) and the estimated coordinates in the step 4) are subjected to the systems such as covariance
It counts parameter and calculates analysis, obtain the filtering parameter and gain matrix at current time；
8) by the result bonding state equation of transfer of step 7), coordinate after new filtering is calculated.
Further, such as Fig. 2, according to process shown in Fig. 2, the data l of two code-discs of synchronization gain robot or so
(k),r(k).Then distance increment Δ s and yaw angle Δ θ (attitude angle) is obtained by the kinematics model on chassis, then
Being converted into cartesian coordinate representation, (Δ x, Δ y), here it is observation coordinate, the motion mathematical model in the technical program can
To be realized using conventional kinematics model.
Further, in the step 5), the observation coordinate of calculating robot according to the following formula：
Δxk=Δ sk·sin(θk(k))
Δyk=Δ sk·cos(θk(k))
Wherein, (Δ xk, Δ yk) be robot observation coordinate and displacement increment.
(Δ x, Δ y) carry system noise, random noise and measurement noise composition to displacement increment, and observational equation is single order
Linear transformation can only occur for linear equation, above-mentioned noise.The two-dimentional ground of traveling is not the level ground X-Y of stricti jurise, institute
Mass center with robot is to have vibration displacement on Z axis in a strict sense, this will generate comparable random noise, while machine
People's drive system can have certain skidding, this process can also generate random noise.Tooth in other such as drive systems
Wheel friction, two-wheeled installation decentraction etc. belong to system noise.Miscellaneous measurement noise can be wiped when acquiring encoder wherein.This
Three kinds of noises can position noise cumulative errors to the walking of entire robot system.
Further, the filter result input filter at k moment is estimated equation, to obtain the estimated value at k+1 momentEstimation equation in the present invention uses the accelerated motion model of newtonian motion mechanics.In the step 4), press
The estimated coordinates of the robot of subsequent time are obtained according to following equation：
Wherein,For the robot estimated coordinates of subsequent time.
During data are observed in processing, a predictive displacement increment can be generated, for this position prediction displacement increment and
Observation displacement increment will do it an assessment.Evaluation process is completed by the algorithm of filter design, and filter can be according to history
Filter result and current observation calculate the displacement increment of this period, the current coordinate data of final updating robot.
The overview flow chart of walking positioning method in the robot chamber of implementing regulations of the present invention, as shown in Figure 1, having obtained
The exercise data of the robot of last moment, the world coordinates (x including last momentt,yt) filtering estimation seat valueFiltering
The history filtering data in 20 periods in estimated coordinates and setting sliding window length.When robot moves it in current period
Afterwards, sensor can capture new code-disc signal increment.The signal increment is calculated by the kinematics model of dandy horse
To the path increment Δ s and current yaw angle θ (t) of trolley.
Various adaptive-filterings and prediction technique based on Kalman filtering will show better tracking and positioning effect.This
Exercise data filtering link in invention is carrying out Recursive Filtering using metric data using a kind of adaptive α β filter
While, constantly estimation and inaccurate parameter and noise variance matrix in correction model inhibit filtering to reduce model error
Diverging improves the estimated accuracy of current robot coordinate.
(Δ x, Δ y) can enter α β filter process to the displacement increment of robot obtained above, which is to take
With system noise and measure noise.α β filter is the Kalman filter under a kind of stable state, the gain K matrix that it is calculated
Calculation amount is much smaller than the former, and is directed to the moving target of two-dimensional surface, its convergence and robustness is better than Kalman's filter
Wave device.
Filtering estimated value is obtained, then can calculate error between filtering estimated value and observed data value, unite to error
Meter analysis.Statistical data length is determined by the sliding window cycle length defined.Filter sliding window in the present invention is designed as 20 and adopts
The sample period.In conjunction with sample frequency 100Hz, 0.2 second before history data are exactly saved.
In the step 6), by following state transition equation, the k+1 moment can be released according to the filter result of last moment
Observation：
X (k+1)=Φ (l (k), r (k))+G (k) W (k)
Wherein, Φ is state-transition matrix；
L (k), r (k) are the code-disc signal of left and right two-wheeled respectively；
G (k) is constant matrices G (k)=[T, T2/2]T, T is measurement period；
W (k) is noise signal；
X (k+1) is the observation at k+1 moment.
In the step 7), filtering parameter and gain matrix are calculated using α β filter.
Further, it in the step 7), is carried out according to the following steps：
71) judge whether to add up to set periodicity；In the present embodiment, going through for 20 periods in sliding window length is set
History filtering data is setting periodicity.
If 72) calculate parameter, β according to the following equation when not up to set periodicity i.e. 20 period：
73) filtering parameter is calculated according to following equation：
Wherein,
σvIt is the standard deviation for filtering estimated value and measurement error；
W is zero-mean, i.e. the Gauss of 1 variance divides white noise；It can be seen that the adaptive α β filtering of implementing regulations according to the present invention
Device calculates the statistical relationship between estimated value and measured value in filtering, then can dynamic adjust gain matrix K,
Obtain preferable dynamic characteristic.
74) link 4 can calculate filtering gain matrix in Fig. 2, calculate gain matrix according to following equation：
The parameter alpha β of common α β filtering is the definite value used, so gain matrix K is a constant coefficient matrix, is had in target
When maneuverable, it is difficult to reach ideal convergence effect.Adaptive α β filter can be according between estimated value and measured value
Error relationship adjustment parameter α β achieve the purpose that dynamic convergence.
For the prediction gain matrix at k+1 moment
Shown in Fig. 3 is a kind of adaptive α β filter signal process flow diagram used in the present invention.It is a total of in Fig. 2
Six steps.Link 1 is to link 3 according to aforementioned execution.
Further, link 4 obtains gain matrix K (k+1) after the completion of executing in Fig. 3, and link 5 will be according to following formula
It calculates in the step 8), calculates according to the following equation：
Wherein, Y (k+1) is the observation at (k+1) moment；
Transfer matrix H (k)=[1 0]；
It is the estimated value at (k+1) moment；
X (k+1/k+1) is the filter result at (k+1) moment.
Fig. 2 link 5 is finished, and entire adaptive-filtering has just completely executed an iteration, obtains filtered seat
Cursor position (xk+1,yk+1), the signal processing flow of filter also will enter (k+2) moment from (k+1) moment.
Motion positions method in a kind of robot chamber provided by the invention is equipped with rotation on differential Two-wheeled motor shaft
Encoder can export the real-time angle data of two motors in real time, acquire the kinematics that the data are input to the differential trolley of two-wheel
In model, obtain the path data and attitude angle information of dandy horse, can family calculate the current world coordinates of trolley, but this
Group coordinate can carry certain system noise, measurement noise, random noise.Present invention employs a kind of adaptive α β filters to go
Above-mentioned noise is eliminated, adaptive α β filter is a kind of extension linear kalman filter, for highly maneuvering target
Tracking has convergence well.
The preferred embodiment of the present invention has been described in detail above.It should be appreciated that those skilled in the art without
It needs creative work according to the present invention can conceive and makes many modifications and variations.Therefore, all technologies in the art
Personnel are available by logical analysis, reasoning, or a limited experiment on the basis of existing technology under this invention's idea
Technical solution, all should be within the scope of protection determined by the claims.
Claims (8)
1. a kind of motion positions method in robot chamber, it is characterized in that：Include the following steps：
1) encoder data synchronous acquisition step, for acquiring the corner cumulative data of original two driving wheel of left and right；
2) encoder data scaling step, according to the kinematics model of the differential trolley of two-wheel, in left and right sidesing driving wheel data generation
Enter to calculate cumulative movement distance and current real-time attitude angle；
3) with the cumulative movement distance and real-time attitude angle obtained in a sampling period, the position in current sample period is calculated
Move increment；
4) it according to the displacement increment in step 3), carries out Kalman filtering calculating or adaptive α, β filtering calculates, obtain next
The estimated coordinates of the robot displacement at moment.
2. motion positions method in robot chamber as described in claim 1, it is characterized in that：It is further comprising the steps of：
5) the code-disc data conversion in step 1) is calculated, obtains distance increment Δ s and yaw angle Δ θ, passes through cartesian coordinate
Conversion, obtains observation coordinate；
6) it calculates and obtains observation；
7) observation coordinate in step 5) and the estimated coordinates in the step 4) are subjected to the statistical parameters calculating point such as covariance
Analysis, obtains the filtering parameter and gain matrix at current time；
8) by the result bonding state equation of transfer of step 7), coordinate after new filtering is calculated.
3. motion positions method in robot chamber as claimed in claim 2, it is characterized in that：In the step 5), according to following public affairs
The observation coordinate of formula calculating robot：
Δxk=Δ sk·sin(θk(k))
Δyk=Δ sk·cos(θk(k))
Wherein, (Δ xk, Δ yk) be robot observation coordinate.
4. motion positions method in robot chamber as described in claim 1, it is characterized in that：In the step 4), according to following
Formula obtains the predictive displacement of the robot of subsequent time：
Wherein,It is displaced for the robot predicting of subsequent time.
In the step 6), by following state transition equation, the sight at k+1 moment can be released according to the filter result of last moment
Measured value：
X (k+1)=Φ (l (k), r (k))+G (k) W (k)
Wherein, Φ is state-transition matrix；
L (k), r (k) are the code-disc signal of left and right two-wheeled respectively；
G (k) is constant matrices G (k)=[T, T2/2]T, T is measurement period；
W (k) is noise signal；
X (k+1) is the observation at k+1 moment.
5. motion positions method in robot chamber as claimed in claim 2, it is characterized in that：In the step 7), filtered using α β
Wave device calculates filtering parameter and gain matrix.
6. motion positions method in robot chamber as claimed in claim 2, it is characterized in that：In the step 7), according to following
Formula calculates gain matrix：
Wherein, K (k+1) is the prediction gain matrix at k+1 moment.
7. motion positions method in robot chamber as claimed in claim 2, it is characterized in that：In the step 7), according to following
Step carries out：
71) judge whether to add up to set periodicity；
72) if not up to set periodicity, calculates parameter, β according to the following equation：
73) filtering parameter is calculated according to following equation：
Wherein,
σvIt is the standard deviation for filtering estimated value and measurement error；
W is zero-mean, i.e. the Gauss of 1 variance divides white noise；
74) gain matrix is calculated according to following equation：
8. motion positions method in robot chamber as described in claim 1, it is characterized in that：In the step 8), according to following
Formula calculates：
Wherein, Y (k+1) is the observation at (k+1) moment；
Transfer matrix H (k)=[1 0]；
It is the estimated value at (k+1) moment；
X (k+1/k+1) is the filter result at (k+1) moment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810558892.5A CN108896049A (en) | 2018-06-01 | 2018-06-01 | A kind of motion positions method in robot chamber |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810558892.5A CN108896049A (en) | 2018-06-01 | 2018-06-01 | A kind of motion positions method in robot chamber |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108896049A true CN108896049A (en) | 2018-11-27 |
Family
ID=64343817
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810558892.5A Withdrawn CN108896049A (en) | 2018-06-01 | 2018-06-01 | A kind of motion positions method in robot chamber |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108896049A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112025706A (en) * | 2020-08-26 | 2020-12-04 | 北京市商汤科技开发有限公司 | Method and device for determining state of robot, robot and storage medium |
| CN112923966A (en) * | 2021-01-29 | 2021-06-08 | 东方红卫星移动通信有限公司 | Kalman filtering-based angle estimation method for double-reading-head photoelectric encoder |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1759797A (en) * | 2004-10-12 | 2006-04-19 | 三星光州电子株式会社 | Robot cleaner coordinates compensation method and a robot cleaner system using the same |
| CN104316059A (en) * | 2014-11-05 | 2015-01-28 | 中国科学院嘉兴微电子与系统工程中心 | Dead-reckoning navigation positioning method and system for acquiring longitude and latitude of automobile by using speedometer |
| CN104931045A (en) * | 2015-05-18 | 2015-09-23 | 哈尔滨工程大学 | Positioning coded disc-based positioning method of omni-directional mobile robot |
| CN106393104A (en) * | 2016-08-25 | 2017-02-15 | 北京创想智控科技有限公司 | Stroke calibration method of mobile robot |
| CN106643694A (en) * | 2016-11-04 | 2017-05-10 | 航天科工智能机器人有限责任公司 | Method for indoor positioning of robot |
| CN107219542A (en) * | 2017-04-14 | 2017-09-29 | 北京克路德人工智能科技有限公司 | GNSS/ODO-based robot double-wheel differential positioning method |
-
2018
- 2018-06-01 CN CN201810558892.5A patent/CN108896049A/en not_active Withdrawn
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1759797A (en) * | 2004-10-12 | 2006-04-19 | 三星光州电子株式会社 | Robot cleaner coordinates compensation method and a robot cleaner system using the same |
| CN104316059A (en) * | 2014-11-05 | 2015-01-28 | 中国科学院嘉兴微电子与系统工程中心 | Dead-reckoning navigation positioning method and system for acquiring longitude and latitude of automobile by using speedometer |
| CN104931045A (en) * | 2015-05-18 | 2015-09-23 | 哈尔滨工程大学 | Positioning coded disc-based positioning method of omni-directional mobile robot |
| CN106393104A (en) * | 2016-08-25 | 2017-02-15 | 北京创想智控科技有限公司 | Stroke calibration method of mobile robot |
| CN106643694A (en) * | 2016-11-04 | 2017-05-10 | 航天科工智能机器人有限责任公司 | Method for indoor positioning of robot |
| CN107219542A (en) * | 2017-04-14 | 2017-09-29 | 北京克路德人工智能科技有限公司 | GNSS/ODO-based robot double-wheel differential positioning method |
Non-Patent Citations (3)
| Title |
|---|
| 卢海进 等: "自适应α-β滤波算法的研究和应用", 《大众科技》 * |
| 李旭: ""基于移动设备的室内定位系统的研究与实现"", 《中国优秀硕士学位论文全文数据库》 * |
| 潘桂彬: ""轮式机器人控制系统的研究与设计"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112025706A (en) * | 2020-08-26 | 2020-12-04 | 北京市商汤科技开发有限公司 | Method and device for determining state of robot, robot and storage medium |
| CN112025706B (en) * | 2020-08-26 | 2022-01-04 | 北京市商汤科技开发有限公司 | Method and device for determining state of robot, robot and storage medium |
| CN112923966A (en) * | 2021-01-29 | 2021-06-08 | 东方红卫星移动通信有限公司 | Kalman filtering-based angle estimation method for double-reading-head photoelectric encoder |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106123890A (en) | A kind of robot localization method of Fusion | |
| CN111398984B (en) | Self-adaptive laser radar point cloud correction and positioning method based on sweeping robot | |
| CN110297462B (en) | Gear grinding precision prediction modeling method considering influence of machine tool geometric error | |
| CN109916431A (en) | A kind of wheel encoder calibration algorithm for four wheel mobile robots | |
| CN111238471B (en) | Sideslip angle estimation method and estimator suitable for agricultural machine linear navigation | |
| CN106871904A (en) | A kind of mobile robot code-disc positioning correction method based on machine vision | |
| CN108508471A (en) | A kind of automatic driving vehicle localization method and device | |
| CN110345936B (en) | Track data processing method and processing system of motion device | |
| CN108896049A (en) | A kind of motion positions method in robot chamber | |
| CN111238535A (en) | IMU error online calibration method based on factor graph | |
| AU2020104234A4 (en) | An Estimation Method and Estimator for Sideslip Angle of Straight-line Navigation of Agricultural Machinery | |
| CN110941183A (en) | Industrial robot dynamics identification method based on neural network | |
| CN110231620A (en) | A kind of noise correlation system tracking filter method | |
| CN112326990A (en) | Method and system for measuring speed of working vehicle | |
| US11314220B2 (en) | Non-contact method and system for controlling an industrial automation machine | |
| CN109856622B (en) | Single-radar linear track target state estimation method under constraint condition | |
| CN112744293A (en) | Steering wheel zero offset angle measuring method and vehicle | |
| CN106408589A (en) | Vehicle-mounted overlooking camera based vehicle movement measurement method | |
| CN104075710B (en) | A kind of motor-driven Extended target based on Trajectory Prediction axial attitude real-time estimation method | |
| CN111546337B (en) | Industrial robot full-coverage path generation method and system based on free-form surface | |
| CN111222225B (en) | Method and device for determining pose of sensor in robot | |
| CN114219852A (en) | Multi-sensor calibration method and device for automatic driving vehicle | |
| CN113203419B (en) | Indoor inspection robot correction positioning method based on neural network | |
| CN108107882B (en) | Automatic calibration and detection system of service robot based on optical motion tracking | |
| CN104849697A (en) | Alpha-beta filter method based on depolarization coordinate transformation |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20181127 |