CN106444738A - Mobile robot path planning method based on dynamic motion primitive learning model - Google Patents
Mobile robot path planning method based on dynamic motion primitive learning model Download PDFInfo
- Publication number
- CN106444738A CN106444738A CN201610348356.3A CN201610348356A CN106444738A CN 106444738 A CN106444738 A CN 106444738A CN 201610348356 A CN201610348356 A CN 201610348356A CN 106444738 A CN106444738 A CN 106444738A
- Authority
- CN
- China
- Prior art keywords
- robot
- motion
- dynamic motion
- dynamic
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 38
- 230000006870 function Effects 0.000 claims abstract description 52
- 230000004888 barrier function Effects 0.000 claims description 30
- 238000006073 displacement reaction Methods 0.000 claims description 22
- 230000001133 acceleration Effects 0.000 claims description 17
- 230000008859 change Effects 0.000 claims description 6
- 230000008878 coupling Effects 0.000 claims description 6
- 238000010168 coupling process Methods 0.000 claims description 6
- 238000005859 coupling reaction Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 5
- 230000001419 dependent effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 238000004321 preservation Methods 0.000 claims description 3
- 238000010411 cooking Methods 0.000 claims description 2
- 239000002245 particle Substances 0.000 claims description 2
- 230000007423 decrease Effects 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000003247 decreasing effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000002068 genetic effect Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
- Manipulator (AREA)
Abstract
The invention discloses a mobile robot path planning method based on a dynamic motion primitive learning model. The method includes: controlling the motion of a robot through a handle, and recording the motion track of the robot; then regarding the recorded track as a sample of the dynamic motion primitive model, performing training by employing the track sample and establishing the dynamic motion primitive model to obtain dynamic motion primitive model parameters and realize autonomous path planning of the robot; and on the basis, changing a target position of the motion of the robot to complete generalized promotion of a new target. According to the path planning method, the intelligence level of the mobile robot is improved, when the target position of the motion of the robot is changed, the robot can autonomously arrive at the new target position, the robot can complete not just a certain specified task but has the capability of generalized promotion for other tasks, and the efficiency of path planning is improved with the combination of online learning characteristics of the dynamic motion primitive model and an autonomous obstacle-avoiding function thereof.
Description
Technical field
The present invention relates to mobile robot path planning field, specifically a kind of based on dynamic motion primitive learning model
Method for planning path for mobile robot.
Background technology
Path planning is one of key technology of mobile robot, and it indicates mobile intelligent robot to a certain extent
The height of level, can rapidly find out a path convenient, collisionless and not only ensure that the mobile robot safety of self, more body
The high efficiency of Xian Liao robot and reliability.
At present, commonly used robot path planning method has the models such as Artificial Potential Field Method, fuzzy logic model, heredity.
Artificial Potential Field Method is more ripe in path planning model and more efficient planing method, is extensively made with its simple mathematical computations
With.But there is the problem such as local minimum point and goal nonreachable in traditional Artificial Potential Field Method.At present, there is multiple solution local pole
The way of point, such as heuristic search, random escape method etc., but robot is simply applied attached by these improved Artificial Potential Field Methods
The control power adding, does not tackle the problem at its root.Genetic model is a kind of many search models based on heredity and natural selection,
Have robust, flexibly, in population search be difficult to fall into the advantages such as local minimum points.But genetic model is carrying out robot path
There is the problems such as population scale is big, search volume big, be easily trapped into local minimum point, convergence rate is slow during planning.
Above traditional robot path planning's model be primarily present following two in terms of problem:
(1) task is specific, has fine performance only for a certain task, and does not have extensive Generalization Ability;
(2) study is often off-line, which results in and wants re-training to learn new scene, and real-time is very poor.
Content of the invention
The technical problem to be solved in the present invention is:For the reality in the presence of above-mentioned method for planning path for mobile robot
When property difference, and mobile robot completes the single problem of task, proposes a kind of road based on dynamic motion primitive learning model
Footpath planing method.Can searching route in real time, combine the effect that can be effectively improved path planning with its automatic obstacle avoiding function
Rate, additionally, robot is when completing new task, can keep the characteristic of original sample trace to arrive without re-training sample
Reach new target location.
For solving above-mentioned technical problem, the present invention provides following technical scheme:
A kind of method for planning path for mobile robot based on dynamic motion primitive learning model, it is characterised in that mainly wrap
Include following steps:
Step 1:The two-dimensional environment of robot motion is modeled, the two-dimensional environment interface of dummy robot's motion, machine
Device people replaces by little filled circles, and barrier is various planar graphs;
Step 2:Utilize the manipulation to robot for the handle, make robot can avoid colliding with barrier from starting point and reach mesh
Punctuate;
Step 3:During step 2 is carried out, gather robot motion trace data and learn as dynamic motion primitive
The sample point of model, described robot motion's track data includes displacement, speed and acceleration;
Step 4:Displacement during the robot motion's track obtaining according to step 3, speed and acceleration information, by these numbers
It according to as training sample, is trained obtaining corresponding to robot motion's track to sample by dynamic motion primitive algorithm
Good weighted value;
Step 5:Arranging initial parameter for particular task, described initial parameter includes starting point and the end of robot motion
Point, the best weights weight values obtaining according to step 4, the path after cooking up by the study of dynamic motion basic-element model, this path has
Have the characteristic of former sample trace, i.e. beginning and end is consistent, and its running orbit is roughly the same with sample trace;
Step 6:On the basis of step 5, add circular barrier, and in original kinetics equation, add coupling
, thus build the dynamic system with barrier avoiding function, it is achieved the automatic obstacle avoiding function of dynamic motion primitive learning model;
Step 7:On the basis of step 5, change the target location of robot motion, in the premise of not re-training sample
Under, only changing the parameter of target location, robot remains to independently reach new aiming spot, i.e. robot can complete not
For a certain appointed task, and other task is also had to the ability of extensive popularization.
In technique scheme, in step 1 being modeled the two-dimensional environment of robot motion, the requirement of modeling is:Move
The scope of activities of mobile robot is at a limited two-dimensional space;With move robot size as benchmark, by the chi of barrier
Robot is regarded as a particle by very little outward expansion;Barrier is made up of various planar graphs, Limited Number, and at machine
In people's moving process, these barriers will not change and move.
In technique scheme, step 2 detailed process is as follows:
Step 2-1:The data of read machine people's handle, when handle to up and down or left and right promote when, this interface shows in real time
Show displacement, speed and acceleration that robot moves in modeling environment;
Step 2-2:Remote-control handle, artificial cooks up the optimal path that a robot can reach home from starting point,
In view of robot typically can only before and after and side-to-side movement, before and after the path therefore planning out is also or the road of side-to-side movement
Footpath, planning track out is also referred to as sample trace;
Step 2-3:When path planning, avoiding obstacles, and by the method for data preservation by the position of sample trace
The value of shifting, speed and acceleration is recorded, and as sample data.
In technique scheme, step 4 comprises the following specific steps that:
Step 4-1:Set up the Mathematical Modeling of dynamic motion primitive:Dynamic motion primitive is generally used to form discrete fortune
Dynamic, for single free degree displacement y, introduce with constant coefficients linear differential equation referred to as dynamic system, this is
System is as the basis to motor learning:
In formula:
X and v is displacement and the speed of system respectively;x0It is initial position and target location respectively with g;τ is to stretch the time
The factor;K is the coefficient of elasticity of spring;D is the damped coefficient that system is under critical condition;F is nonlinear function, is used for generating
Arbitrarily complicated motion;
Step 4-2:Initial parameter is set, starting point x of robot motion0With impact point g, timeconstantτ, the bullet of spring
Property coefficient K, system is in the damped coefficient D under critical condition;Nonlinear function f, for forming arbitrarily complicated motion, defines f
For:
In formula:
ψiS () is Radial basis kernel function, i represents i-th Radial basis kernel function ψi(s), its span be 1 arrive N, wherein N
Represent the number of Radial basis kernel function;Radial basis kernel function is defined as:
ψi(s)=exp (-hi(s-ci)2) (4)
In formula:
ciIt is the center of Radial basis kernel function, hi>0 and determine kernel function width;Wherein hN=hN-1, i=1 ... N, α are any normal number;
Function f in formula (3) is not dependent on time parameter, and is depending on phase variant s, and the expression-form of s is:
In formula:
S is the function with regard to time t, and α is any normal number, and τ is time contraction-expansion factor;Equation (5) is understood that s is by 1
To 0 monotone decreasing, therefore equation (5) is referred to as cannoncial system;
Step 4-3:The sample data obtaining in step 3 is substituted in formula (1) and formula (2), because cannoncial system is
Integrable, i.e. s can calculate according to parameter τ, so nonlinear disturbance f ' (s) in training sample can be expressed as:
Solve best weights weight values w according to minimum error principle function Ji, the wherein expression formula of minimum error principle function
For:
J=∑s(f′(s)-f(s))2(7)
W when J takes minimumiIt is exactly optimal weighted value.
In technique scheme, step 5 comprises the following specific steps that:
Step 5-1:When robot performs specifying of task, start position and the final position of robot are set;
Step 5-2:Sample data is two-dimentional, namely includes the data on x-axis direction and the data on y-axis direction, by x
Data on direction of principal axis are trained according to step 4, obtain the best weights weight values on x-axis direction, substitute into the starting point in step 5-1
And end point values, calculate x side upwardly through the displacement after the study of dynamic motion basic-element model, speed and acceleration;
Step 5-3:Data on y-axis direction are trained according to step 4, obtain the best weights weight values on y-axis direction,
Substitute into the beginning and end value in step 5-1, calculate y side upwardly through the displacement after the study of dynamic motion basic-element model, speed
Degree and acceleration;
Step 5-4:Read in the data obtaining in step 5-2 and step 5-3, respectively obtain x-axis and the fortune of y-axis both direction
Dynamic data, export the track emulation figure of motion on two dimensional surface, i.e. complete based on dynamic motion primitive learning model to movement
The path planning of robot.
In technique scheme, the barrier that added in step 6 is to be central coordinate of circle with (0.4,0.4), and radius is
The circle of 0.1m.
The method of the present invention starts to begin one's study with simple linear dynamic system (one group of differential equation), by conversion is
Simple linear dynamic system is converted into nonlinear system by system, and forms arbitrarily complicated motion by attractor, this
Sample is with regard to can better simply study to nonlinear system.Wherein, it is that error can be certainly by the advantage that the differential equation represents
Dynamic is corrected, and the differential equation is all to be formed with fixing form, only has only to letter according to this fixing form
Single one target component of change, just adapts to new environment, i.e. can carry out extensive to fresh target;Based on dynamic motion primitive
The method of study is on-line study, need not relearn for new situation, can real-time tracking position of object.Thus,
In avoidance aspect, realize automatic obstacle avoiding by building dynamic system with barrier avoiding function, and dynamic motion basic-element model
On-line study feature and its automatic obstacle avoiding function combine and improve the efficiency of path planning.
Compared with prior art, the invention has the beneficial effects as follows:
(1) present invention proposes a kind of method for planning path for mobile robot based on dynamic motion primitive learning model, should
Learning model has extensive Generalization Ability, and robot, when completing new task, can keep former without re-training sample
The characteristic carrying out sample trace reaches new target location.
(2) the path planning model that the present invention proposes is real-time when searching route, is combined with its automatic obstacle avoiding function
Get up and can be effectively improved the efficiency of path planning.
Brief description
Fig. 1 is the mobile robot path planning process schematic based on dynamic motion primitive learning model for the present invention;
Fig. 2 is the two-dimensional environment simulating robot motion in the present invention;Wherein linear pattern track representative sample track;Various
The figure (such as rectangle, circle, ellipse) of shape represents the barrier in two-dimensional environment;
Fig. 3 is the automatic obstacle avoiding analogous diagram of dynamic motion primitive learning model in the present invention;
Fig. 4 is the analogous diagram that in the present invention, dynamic motion primitive learning model is had extensive Generalization Ability;
Fig. 5 be training sample trace with by dynamic motion basic-element model study after track comparison diagram.
Detailed description of the invention
In order to further illustrate technical scheme, the present invention will be described in detail for 1-5 below in conjunction with the accompanying drawings.
Step 1:The two-dimensional environment interface of dummy robot's motion, on its interface, robot is replaced by little filled circles, barrier
Thing is hindered to be various planar graphs;The two-dimensional environment interface arranging robot motion is square (long and width is all 1m), robot
Replace by the little filled circles of an a diameter of 5mm.
Step 2:OPENCV (Open Source Computer Vision Library) is utilized to realize handle to robot
Manipulation, make robot can from starting point avoid and barrier collision reach impact point;
Step 2-1:Write a upper computer software based on MFC (Microsoft Foundation Classes) interface,
This software can be with the data of read machine people's handle, when handle is to up and down or when left and right promotes, this interface can show in real time
Displacement, speed and the acceleration that robot moves in modeling environment;
Step 2-2:Remote-control handle, artificial cooks up the optimal path that a robot can reach home from starting point,
In view of robot typically can only before and after and side-to-side movement, before and after the path therefore planning out is also or the road of side-to-side movement
Footpath, planning track out is also referred to as sample trace;
Step 2-3:When path planning, avoiding obstacles, and by the method for data preservation by the position of sample trace
The value of shifting, speed and acceleration is recorded, and as sample data;
Step 2 uses the upper computer software write based on MFC, by can be achieved with the manipulation of handle to machine
The control of people.It is provided with the velocity magnitude when displacement of handle push rod is robot motion, wherein control robot motion's speed
Degree is in the range of-5mm/s~5mm/s.
Step 3:During step 2 is carried out, gather robot motion trace data and learn as dynamic motion primitive
The sample point of model, wherein robot motion's track data includes the size of its displacement, speed and accekeration;
Step 4:These data are made by the displacement during robot motion's track obtaining according to step 3, speed and acceleration
For the training sample of DMP learning model, by obtaining the best weights weight values corresponding to robot motion's track to the training of sample;
Step 4-1:Set up the Mathematical Modeling of dynamic motion primitive.Dynamic motion primitive is generally used to form discrete fortune
Dynamic, for single free degree displacement y, introduce with constant coefficients linear differential equation referred to as dynamic system, this is
System is as the basis to motor learning:
In formula:
X and v is displacement and the speed of system respectively;x0It is initial position and target location respectively with g;τ is to stretch the time
The factor;K is the coefficient of elasticity of spring;D is the damped coefficient that system is under critical condition;F is nonlinear function, is used for generating
Arbitrarily complicated motion;
Step 4-2:Initial parameter is set, starting point x of robot motion0With impact point g, timeconstantτ, the bullet of spring
Property coefficient K, system is in the damped coefficient D under critical condition;Nonlinear function f, for forming arbitrarily complicated motion, defines
For:
In formula:
ψiS () is Radial basis kernel function, i represents i-th Radial basis kernel function ψi(s), its span be 1 arrive N, wherein N
Represent the number of Radial basis kernel function;Radial basis kernel function is defined as:
ψi(s)=exp (-hi(s-ci)2) (4)
In formula:
ciIt is the center of Radial basis kernel function, hi>0 and determine kernel function width;Wherein hN=hN-1, i=1 ... N, α are any normal number;
Function f in formula (3) is not dependent on time parameter, and is depending on phase variant s, and the expression-form of s is:
In formula:
S is the function with regard to time t, and α is any normal number, and τ is time contraction-expansion factor;Equation (5) is understood that s is by 1
To 0 monotone decreasing, therefore equation (5) is referred to as cannoncial system;
Step 4-3:The sample data obtaining in step 3 is substituted in above-mentioned formula, because cannoncial system is integrable,
I.e. s can calculate according to parameter τ, so nonlinear disturbance f ' (s) in training sample can be expressed as:
Solve best weights weight values w according to minimum error principle function Ji, the wherein expression formula of minimum error principle function
For:
J=∑s(f′(s)-f(s))2(7)
W when J takes minimumiIt is exactly optimal weighted value;
Step 5:Initial parameter (beginning and end of robot motion) is set for particular task, obtains according to step 4
Best weights weight values, cook up by dynamic motion basic-element model study after path, this path has the spy of former sample trace
Property;
Step 5-1:When robot performs specifying of task, start position and the final position of robot are set;
Step 5-2:Sample data is two-dimentional (data on x-axis direction and the data on y-axis direction), by x-axis direction
On data be trained according to step 4, obtain the best weights weight values on x-axis direction, substitute into the beginning and end in step 5-1
Value, it is possible to calculate x side upwardly through the displacement after the study of dynamic motion basic-element model, speed and acceleration;
Step 5-3:Data on y-axis direction are trained according to step 4, obtain the best weights weight values on y-axis direction,
Substitute into the beginning and end value in step 5-1, it is possible to calculate y side upwardly through the position after the study of dynamic motion basic-element model
Shifting, speed and acceleration;
Step 5-4:The data obtaining in step 5-2 and step 5-3 are read in by MATLAB, obtains x-axis and y-axis two
The exercise data in direction, exports the track emulation figure of motion on two dimensional surface, i.e. completes to learn mould based on dynamic motion primitive
The path planning to mobile robot for the type.
Step 6:On the basis of step 5, add circular barrier, and in original kinetics equation, add coupling
, thus build the dynamic system with barrier avoiding function, it is achieved that the automatic obstacle avoiding function of dynamic motion primitive learning model;
Step 6-1:On the basis of step 5-4, adding circular barrier, wherein barrier is for circle with (0.4,0.4)
Heart coordinate, radius is the circle of 0.1m;
Step 6-2:(x v) builds band to add coupling terms P on the basis of the dynamic systems equation that step 4-1 provides
Having the dynamic system of barrier avoiding function, wherein (x, expression formula v) is coupling terms P:
In formula:
Be withFor axle,For the spin matrix of the anglec of rotation, vectorIt is the position of barrier, γ and β
Being constant, θ is the angle on the distance vector of the point on track and barrier and track between the relative velocity of that;
Step 6-3:Given coupling terms P (x, v) the constant term initial value in formula, wherein γ=8,Spin matrix R
It is expressed as:
Step 6-4:By building the dynamic system with barrier avoiding function, add the barrier described in step 6-1, machine
Device people remains to avoiding obstacles and reaches impact point, and wherein the mathematic(al) representation with the dynamic system of barrier avoiding function is:
As seen from Figure 3, dynamic motion basic-element model has the function of automatic obstacle avoiding;
Step 7:On the basis of step 5, change the target location of robot motion, in the premise of not re-training sample
Under, only changing the parameter of target location, robot remains to the new aiming spot of autonomous arrival, i.e. robot can complete
It is not for a certain appointed task, and other task is also had to the ability of extensive popularization.
Step 7-1:On the basis of step 5-4, the position changing robot target point is (0.5,0.5), substitutes into step
4, obtain on the premise of not re-training sample, the track after the study of dynamic motion basic-element model;
Step 7-2:On the basis of step 5-4, the position changing robot target point is (0.8,0.8), substitutes into step
4, obtain on the premise of not re-training sample, the track after the study of dynamic motion basic-element model;
Step 7-3:As shown in Figure 4, in step 7-1 and step 7-2, robot can reach new target location, and
And the characteristic of this track of keeping intact, hence it is demonstrated that the extensive Generalization Ability that dynamic motion primitive learning model is had;
To sum up, the present invention realizes the path planning to robot, this learning model based on dynamic motion primitive learning model
On-line study feature and its automatic obstacle avoiding function combine and improve the efficiency of path planning, and this model has extensive pushing away
Wide ability.The proposition of the present invention, improves the intelligent of mobile robot, and leads in path planning, avoidance for mobile robot
The association areas such as boat provide reference.
Claims (6)
1. the method for planning path for mobile robot based on dynamic motion primitive learning model, it is characterised in that mainly include
Following steps:
Step 1:The two-dimensional environment of robot motion is modeled, the two-dimensional environment interface of dummy robot's motion, robot
Replacing by little filled circles, barrier is various planar graphs;
Step 2:Utilize the manipulation to robot for the handle, make robot can avoid colliding with barrier from starting point and reach target
Point;
Step 3:During step 2 is carried out, gather robot motion trace data as dynamic motion primitive learning model
Sample point, described robot motion's track data includes displacement, speed and acceleration;
Step 4:These data are made by the displacement during robot motion's track obtaining according to step 3, speed and acceleration information
It for training sample, is trained obtaining the best weights corresponding to robot motion's track to sample by dynamic motion primitive algorithm
Weight values;
Step 5:Arranging initial parameter for particular task, described initial parameter includes the beginning and end of robot motion, root
The best weights weight values obtaining according to step 4, the path after cooking up by the study of dynamic motion basic-element model, this path has former state
The characteristic of this track, i.e. beginning and end are consistent, and its running orbit is roughly the same with sample trace;
Step 6:On the basis of step 5, add circular barrier, and in original kinetics equation, add coupling terms,
Thus build the dynamic system with barrier avoiding function, it is achieved the automatic obstacle avoiding function of dynamic motion primitive learning model;
Step 7:On the basis of step 5, change the target location of robot motion, on the premise of not re-training sample,
Only changing the parameter of target location, robot remains to independently reach new aiming spot, i.e. robot can complete not pin
To a certain appointed task, and other task is also had to the ability of extensive popularization.
2. the method for planning path for mobile robot based on dynamic motion primitive learning model according to claim 1, its
It is characterised by:In step 1 being modeled the two-dimensional environment of robot motion, the requirement of modeling is:The activity of mobile robot
Scope is at a limited two-dimensional space;With move robot size as benchmark, by the size outward expansion of barrier, by machine
Device people regards a particle as;Barrier is made up of various planar graphs, Limited Number, and in robot moving process these
Barrier will not change and move.
3. the method for planning path for mobile robot based on dynamic motion primitive learning model according to claim 1, its
It is characterised by:Step 2 detailed process is as follows:
Step 2-1:The data of read machine people's handle, when handle to up and down or left and right promote when, this interface shows machine in real time
Displacement, speed and the acceleration that device people moves in modeling environment;
Step 2-2:Remote-control handle, artificial cooks up the optimal path that a robot can reach home from starting point, it is considered to
To robot typically can only before and after and side-to-side movement, before and after the path therefore planning out is also or the path of side-to-side movement,
Planning track out is also referred to as sample trace;
Step 2-3:When path planning, avoiding obstacles, and by the method for data preservation by the displacement of sample trace, speed
The value of degree and acceleration is recorded, and as sample data.
4. the method for planning path for mobile robot based on dynamic motion primitive learning model according to claim 1, its
It is characterised by:Step 4 comprises the following specific steps that:
Step 4-1:Set up the Mathematical Modeling of dynamic motion primitive:Dynamic motion primitive is generally used to be formed discrete motion, right
In single free degree displacement y, introduce with constant coefficients linear differential equation referred to as dynamic system, this system conduct
Basis to motor learning:
In formula:
X and v is displacement and the speed of system respectively;x0It is initial position and target location respectively with g;τ is time contraction-expansion factor;K
It is the coefficient of elasticity of spring;D is the damped coefficient that system is under critical condition;F is nonlinear function, is used for generating arbitrarily multiple
Miscellaneous motion;
Step 4-2:Initial parameter is set, starting point x of robot motion0With impact point g, timeconstantτ, the elastic system of spring
Number K, system is in the damped coefficient D under critical condition;Nonlinear function f is for forming arbitrarily complicated motion, and definition f is:
In formula:
ψiS () is Radial basis kernel function, i represents i-th Radial basis kernel function ψiS (), its span is 1 to N, and wherein N represents
The number of Radial basis kernel function;Radial basis kernel function is defined as:
ψi(s)=exp (-hi(s-ci)2) (4)
In formula:
ciIt is the center of Radial basis kernel function, hi>0 and determine kernel function width;Wherein hN=hN-1, i=1 ... N, α are any normal number;
Function f in formula (3) is not dependent on time parameter, and is depending on phase variant s, and the expression-form of s is:
In formula:
S is the function with regard to time t, and α is any normal number, and τ is time contraction-expansion factor;Equation (5) is understood that s is single by 1 to 0
Tune successively decreases, and therefore equation (5) is referred to as cannoncial system;
Step 4-3:The sample data obtaining in step 3 is substituted in formula (1) and formula (2), because cannoncial system is to amass
Point, i.e. s can calculate according to parameter τ, so nonlinear disturbance f ' (s) in training sample can be expressed as:
Solve best weights weight values w according to minimum error principle function Ji, wherein the expression formula of minimum error principle function is:
J=∑s(f′(s)-f(s))2(7)
W when J takes minimumiIt is exactly optimal weighted value.
5. the method for planning path for mobile robot based on dynamic motion primitive learning model according to claim 1, its
It is characterised by:Step 5 comprises the following specific steps that:
Step 5-1:When robot performs specifying of task, start position and the final position of robot are set;
Step 5-2:Sample data is two-dimentional, namely includes the data on x-axis direction and the data on y-axis direction, by x-axis side
Data upwards are trained according to step 4, obtain the best weights weight values on x-axis direction, substitute into the starting point in step 5-1 and end
Point value, calculates x side upwardly through the displacement after the study of dynamic motion basic-element model, speed and acceleration;
Step 5-3:Data on y-axis direction are trained according to step 4, obtain the best weights weight values on y-axis direction, substitute into
Beginning and end value in step 5-1, calculate y side upwardly through dynamic motion basic-element model study after displacement, speed and
Acceleration;
Step 5-4:Read in the data obtaining in step 5-2 and step 5-3, respectively obtain the motion number of x-axis and y-axis both direction
According to exporting the track emulation figure of motion on two dimensional surface, i.e. complete based on dynamic motion primitive learning model to mobile machine
The path planning of people.
6. the method for planning path for mobile robot based on dynamic motion primitive learning model according to claim 1, its
It is characterised by:The barrier being added in step 6 is to be central coordinate of circle with (0.4,0.4), and radius is the circle of 0.1m.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610348356.3A CN106444738B (en) | 2016-05-24 | 2016-05-24 | Method for planning path for mobile robot based on dynamic motion primitive learning model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610348356.3A CN106444738B (en) | 2016-05-24 | 2016-05-24 | Method for planning path for mobile robot based on dynamic motion primitive learning model |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106444738A true CN106444738A (en) | 2017-02-22 |
CN106444738B CN106444738B (en) | 2019-04-09 |
Family
ID=58183456
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610348356.3A Active CN106444738B (en) | 2016-05-24 | 2016-05-24 | Method for planning path for mobile robot based on dynamic motion primitive learning model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106444738B (en) |
Cited By (29)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107368076A (en) * | 2017-07-31 | 2017-11-21 | 中南大学 | Robot motion's pathdepth learns controlling planning method under a kind of intelligent environment |
CN107436604A (en) * | 2017-07-31 | 2017-12-05 | 中南大学 | Controlling planning method is intelligently decomposed in carrying robot path under a kind of intelligent environment |
CN107450535A (en) * | 2017-07-31 | 2017-12-08 | 中南大学 | A kind of intelligent carrying robot optimal path combination chart discusses controlling planning method |
CN108255182A (en) * | 2018-01-30 | 2018-07-06 | 上海交通大学 | A kind of service robot pedestrian based on deeply study perceives barrier-avoiding method |
CN108469732A (en) * | 2018-03-13 | 2018-08-31 | 同济大学 | Robot working space's adaptability traveling control system based on DMP and method |
CN108582072A (en) * | 2018-04-28 | 2018-09-28 | 北京邮电大学 | A kind of space manipulator mission planning method based on improvement figure planning algorithm |
CN108692734A (en) * | 2017-04-07 | 2018-10-23 | 北京图森未来科技有限公司 | A kind of paths planning method and device |
CN108919794A (en) * | 2018-06-01 | 2018-11-30 | 广州视源电子科技股份有限公司 | Global path planning method and device for double-wheel differential mobile robot |
CN109108942A (en) * | 2018-09-11 | 2019-01-01 | 武汉科技大学 | The mechanical arm motion control method and system of the real-time teaching of view-based access control model and adaptive DMPS |
CN109313036A (en) * | 2017-04-27 | 2019-02-05 | 北京嘀嘀无限科技发展有限公司 | The system and method for route planning |
CN109940619A (en) * | 2019-04-09 | 2019-06-28 | 深圳前海达闼云端智能科技有限公司 | Trajectory planning method, electronic device and storage medium |
CN110315544A (en) * | 2019-06-24 | 2019-10-11 | 南京邮电大学 | A kind of robot manipulation's learning method based on video image demonstration |
CN110789530A (en) * | 2019-11-19 | 2020-02-14 | 中国科学院深圳先进技术研究院 | Four-wheel independent steering-independent driving vehicle trajectory tracking method and system |
CN111108457A (en) * | 2017-09-29 | 2020-05-05 | 罗伯特·博世有限公司 | Method, device and computer program for operating a robot control system |
US10670412B2 (en) | 2018-02-20 | 2020-06-02 | Veoneer Us, Inc. | System and method for generating a target path for a vehicle |
CN111319044A (en) * | 2020-03-04 | 2020-06-23 | 达闼科技(北京)有限公司 | Article grabbing method and device, readable storage medium and grabbing robot |
CN111633646A (en) * | 2020-05-22 | 2020-09-08 | 北京理工大学 | Robot motion planning method based on DMPs and modified obstacle avoidance algorithm |
CN111699079A (en) * | 2018-02-15 | 2020-09-22 | 罗伯特·博世有限公司 | Coordination system, operating device and method |
CN112180726A (en) * | 2020-09-29 | 2021-01-05 | 北京航空航天大学 | Spacecraft relative motion trajectory planning method based on meta-learning |
CN112207835A (en) * | 2020-09-18 | 2021-01-12 | 浙江大学 | Method for realizing double-arm cooperative work task based on teaching learning |
CN112549028A (en) * | 2020-12-02 | 2021-03-26 | 中国科学院自动化研究所 | Double-arm robot track planning method based on dynamic motion primitives and artificial potential field |
CN112631269A (en) * | 2019-10-08 | 2021-04-09 | 国立大学法人静冈大学 | Autonomous mobile robot and control program for autonomous mobile robot |
CN112959330A (en) * | 2021-02-02 | 2021-06-15 | 浙江大学 | Robot double-arm motion man-machine corresponding device and method based on master-slave dynamic motion elements |
CN112975986A (en) * | 2021-03-25 | 2021-06-18 | 珞石(北京)科技有限公司 | Mechanical arm point-to-point trajectory planning method and device based on radial basis function |
CN113043251A (en) * | 2021-04-23 | 2021-06-29 | 江苏理工学院 | Robot teaching reproduction track learning method |
CN114378791A (en) * | 2022-01-13 | 2022-04-22 | 中国科学技术大学 | Robot task execution method, device, equipment and readable storage medium |
CN114571458A (en) * | 2022-03-21 | 2022-06-03 | 苏州大学 | Micro-nano robot assembly track learning method based on dynamic motion primitives |
CN116901055A (en) * | 2023-05-19 | 2023-10-20 | 兰州大学 | Human-simulated interaction control method and device, electronic equipment and storage medium |
CN118003325A (en) * | 2024-02-28 | 2024-05-10 | 北京中科慧灵机器人技术有限公司 | Target track data information determining method, device, equipment and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010058260A (en) * | 2008-09-01 | 2010-03-18 | Korea Advanced Inst Of Sci Technol | Method for controlling motion of robot based on evolutionary algorithm and imitation learning |
US8843236B2 (en) * | 2012-03-15 | 2014-09-23 | GM Global Technology Operations LLC | Method and system for training a robot using human-assisted task demonstration |
CN105045260A (en) * | 2015-05-25 | 2015-11-11 | 湖南大学 | Mobile robot path planning method in unknown dynamic environment |
CN105137967A (en) * | 2015-07-16 | 2015-12-09 | 北京工业大学 | Mobile robot path planning method with combination of depth automatic encoder and Q-learning algorithm |
-
2016
- 2016-05-24 CN CN201610348356.3A patent/CN106444738B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010058260A (en) * | 2008-09-01 | 2010-03-18 | Korea Advanced Inst Of Sci Technol | Method for controlling motion of robot based on evolutionary algorithm and imitation learning |
US8843236B2 (en) * | 2012-03-15 | 2014-09-23 | GM Global Technology Operations LLC | Method and system for training a robot using human-assisted task demonstration |
CN105045260A (en) * | 2015-05-25 | 2015-11-11 | 湖南大学 | Mobile robot path planning method in unknown dynamic environment |
CN105137967A (en) * | 2015-07-16 | 2015-12-09 | 北京工业大学 | Mobile robot path planning method with combination of depth automatic encoder and Q-learning algorithm |
Non-Patent Citations (1)
Title |
---|
陈鹏震等: "基于动态运动基元的微小型四旋翼无人机路径规划", 《高技术通讯》 * |
Cited By (41)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108692734B (en) * | 2017-04-07 | 2020-07-07 | 北京图森未来科技有限公司 | Path planning method and device |
CN108692734A (en) * | 2017-04-07 | 2018-10-23 | 北京图森未来科技有限公司 | A kind of paths planning method and device |
CN109313036A (en) * | 2017-04-27 | 2019-02-05 | 北京嘀嘀无限科技发展有限公司 | The system and method for route planning |
CN109313036B (en) * | 2017-04-27 | 2022-04-15 | 北京嘀嘀无限科技发展有限公司 | Route planning system and method |
CN107436604A (en) * | 2017-07-31 | 2017-12-05 | 中南大学 | Controlling planning method is intelligently decomposed in carrying robot path under a kind of intelligent environment |
CN107450535A (en) * | 2017-07-31 | 2017-12-08 | 中南大学 | A kind of intelligent carrying robot optimal path combination chart discusses controlling planning method |
CN107368076B (en) * | 2017-07-31 | 2018-03-27 | 中南大学 | Robot motion's pathdepth learns controlling planning method under a kind of intelligent environment |
CN107368076A (en) * | 2017-07-31 | 2017-11-21 | 中南大学 | Robot motion's pathdepth learns controlling planning method under a kind of intelligent environment |
CN111108457A (en) * | 2017-09-29 | 2020-05-05 | 罗伯特·博世有限公司 | Method, device and computer program for operating a robot control system |
CN111108457B (en) * | 2017-09-29 | 2024-03-29 | 罗伯特·博世有限公司 | Method, device and computer program for operating a robot control system |
CN108255182A (en) * | 2018-01-30 | 2018-07-06 | 上海交通大学 | A kind of service robot pedestrian based on deeply study perceives barrier-avoiding method |
CN111699079B (en) * | 2018-02-15 | 2023-09-29 | 罗伯特·博世有限公司 | Coordination system, operation device and method |
US11992956B2 (en) | 2018-02-15 | 2024-05-28 | Robert Bosch Gmbh | Coordination system, handling device, and method |
CN111699079A (en) * | 2018-02-15 | 2020-09-22 | 罗伯特·博世有限公司 | Coordination system, operating device and method |
US10670412B2 (en) | 2018-02-20 | 2020-06-02 | Veoneer Us, Inc. | System and method for generating a target path for a vehicle |
CN108469732A (en) * | 2018-03-13 | 2018-08-31 | 同济大学 | Robot working space's adaptability traveling control system based on DMP and method |
CN108582072A (en) * | 2018-04-28 | 2018-09-28 | 北京邮电大学 | A kind of space manipulator mission planning method based on improvement figure planning algorithm |
CN108919794A (en) * | 2018-06-01 | 2018-11-30 | 广州视源电子科技股份有限公司 | Global path planning method and device for double-wheel differential mobile robot |
CN109108942A (en) * | 2018-09-11 | 2019-01-01 | 武汉科技大学 | The mechanical arm motion control method and system of the real-time teaching of view-based access control model and adaptive DMPS |
CN109940619A (en) * | 2019-04-09 | 2019-06-28 | 深圳前海达闼云端智能科技有限公司 | Trajectory planning method, electronic device and storage medium |
CN110315544A (en) * | 2019-06-24 | 2019-10-11 | 南京邮电大学 | A kind of robot manipulation's learning method based on video image demonstration |
CN112631269A (en) * | 2019-10-08 | 2021-04-09 | 国立大学法人静冈大学 | Autonomous mobile robot and control program for autonomous mobile robot |
CN112631269B (en) * | 2019-10-08 | 2024-06-11 | 国立大学法人静冈大学 | Autonomous mobile robot and control program for autonomous mobile robot |
CN110789530A (en) * | 2019-11-19 | 2020-02-14 | 中国科学院深圳先进技术研究院 | Four-wheel independent steering-independent driving vehicle trajectory tracking method and system |
CN111319044A (en) * | 2020-03-04 | 2020-06-23 | 达闼科技(北京)有限公司 | Article grabbing method and device, readable storage medium and grabbing robot |
CN111633646A (en) * | 2020-05-22 | 2020-09-08 | 北京理工大学 | Robot motion planning method based on DMPs and modified obstacle avoidance algorithm |
CN111633646B (en) * | 2020-05-22 | 2021-08-27 | 北京理工大学 | Robot motion planning method based on DMPs and modified obstacle avoidance algorithm |
CN112207835A (en) * | 2020-09-18 | 2021-01-12 | 浙江大学 | Method for realizing double-arm cooperative work task based on teaching learning |
CN112180726A (en) * | 2020-09-29 | 2021-01-05 | 北京航空航天大学 | Spacecraft relative motion trajectory planning method based on meta-learning |
CN112549028A (en) * | 2020-12-02 | 2021-03-26 | 中国科学院自动化研究所 | Double-arm robot track planning method based on dynamic motion primitives and artificial potential field |
CN112959330A (en) * | 2021-02-02 | 2021-06-15 | 浙江大学 | Robot double-arm motion man-machine corresponding device and method based on master-slave dynamic motion elements |
CN112975986A (en) * | 2021-03-25 | 2021-06-18 | 珞石(北京)科技有限公司 | Mechanical arm point-to-point trajectory planning method and device based on radial basis function |
CN113043251A (en) * | 2021-04-23 | 2021-06-29 | 江苏理工学院 | Robot teaching reproduction track learning method |
CN114378791A (en) * | 2022-01-13 | 2022-04-22 | 中国科学技术大学 | Robot task execution method, device, equipment and readable storage medium |
CN114378791B (en) * | 2022-01-13 | 2023-09-01 | 中国科学技术大学 | Robot task execution method, device, equipment and readable storage medium |
CN114571458A (en) * | 2022-03-21 | 2022-06-03 | 苏州大学 | Micro-nano robot assembly track learning method based on dynamic motion primitives |
CN114571458B (en) * | 2022-03-21 | 2022-11-15 | 苏州大学 | Micro-nano robot assembly track learning method based on dynamic motion elements |
CN116901055A (en) * | 2023-05-19 | 2023-10-20 | 兰州大学 | Human-simulated interaction control method and device, electronic equipment and storage medium |
CN116901055B (en) * | 2023-05-19 | 2024-04-19 | 兰州大学 | Human-simulated interaction control method and device, electronic equipment and storage medium |
CN118003325A (en) * | 2024-02-28 | 2024-05-10 | 北京中科慧灵机器人技术有限公司 | Target track data information determining method, device, equipment and storage medium |
CN118003325B (en) * | 2024-02-28 | 2024-07-23 | 北京中科慧灵机器人技术有限公司 | Target track data information determining method, device, equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN106444738B (en) | 2019-04-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106444738A (en) | Mobile robot path planning method based on dynamic motion primitive learning model | |
CN110989576B (en) | Target following and dynamic obstacle avoidance control method for differential slip steering vehicle | |
Wang et al. | Collision-free navigation of autonomous vehicles using convex quadratic programming-based model predictive control | |
Chen et al. | Stabilization approaches for reinforcement learning-based end-to-end autonomous driving | |
EP2360629A2 (en) | Device for the autonomous bootstrapping of useful information | |
CN107145161A (en) | Unmanned plane accesses the path planning method and device of multiple target point | |
CN106338919A (en) | USV (Unmanned Surface Vehicle) track tracking control method based on enhanced learning type intelligent algorithm | |
CN112550314B (en) | Embedded optimization type control method suitable for unmanned driving, driving control module and automatic driving control system thereof | |
CN108983605B (en) | Rigid body control method for fluid guidance based on deep reinforcement learning | |
Huang et al. | Deductive reinforcement learning for visual autonomous urban driving navigation | |
CN110181508A (en) | Underwater robot three-dimensional Route planner and system | |
CN114895707B (en) | Agricultural unmanned aerial vehicle path planning method and system based on variable frequency bat algorithm | |
Zou et al. | An end-to-end learning of driving strategies based on DDPG and imitation learning | |
CN117606490B (en) | Collaborative search path planning method for autonomous underwater vehicle | |
Yuan et al. | Hierarchical dynamic movement primitive for the smooth movement of robots based on deep reinforcement learning | |
Sun et al. | Event-triggered reconfigurable reinforcement learning motion-planning approach for mobile robot in unknown dynamic environments | |
CN113379027A (en) | Method, system, storage medium and application for generating confrontation interactive simulation learning | |
Zhang et al. | Deepreinforcementlearningforreal-world quadrupedal locomotion: a comprehensive review | |
CN114609925B (en) | Training method of underwater exploration strategy model and underwater exploration method of bionic machine fish | |
CN116430891A (en) | Deep reinforcement learning method oriented to multi-agent path planning environment | |
CN118043824A (en) | Retrieval enhanced reinforcement learning | |
Metoyer et al. | Animating athletic motion planning by example | |
Yang et al. | Mapless navigation for UAVs via reinforcement learning from demonstrations | |
Vrajitoru | Global to local for path decision using neural networks | |
Bussola et al. | Efficient Reinforcement Learning for 3D Jumping Monopods |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |