CN106444738B - Method for planning path for mobile robot based on dynamic motion primitive learning model - Google Patents
Method for planning path for mobile robot based on dynamic motion primitive learning model Download PDFInfo
- Publication number
- CN106444738B CN106444738B CN201610348356.3A CN201610348356A CN106444738B CN 106444738 B CN106444738 B CN 106444738B CN 201610348356 A CN201610348356 A CN 201610348356A CN 106444738 B CN106444738 B CN 106444738B
- Authority
- CN
- China
- Prior art keywords
- robot
- dynamic motion
- motion
- data
- dynamic
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 230000006870 function Effects 0.000 claims abstract description 52
- 230000008859 change Effects 0.000 claims abstract description 6
- 230000004888 barrier function Effects 0.000 claims description 29
- 238000006073 displacement reaction Methods 0.000 claims description 19
- 230000001133 acceleration Effects 0.000 claims description 18
- 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
- 238000010411 cooking Methods 0.000 claims description 3
- 230000003247 decreasing effect Effects 0.000 claims description 3
- 230000001419 dependent effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 238000013178 mathematical model Methods 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 3
- 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
- 238000005516 engineering process 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 kind of method for planning path for mobile robot based on dynamic motion primitive learning model.Robot motion, the motion profile of recorder people are controlled with handle first.Then it using the track of record as the sample of dynamic motion basic-element model, by establishing dynamic motion basic-element model, is trained using track sample and obtains dynamic motion basic-element model parameter, to realize robot autonomous path planning.On this basis, change the target position of robot motion, complete the extensive popularization to fresh target.Paths planning method of the invention improves the intelligent level of mobile robot, when the target position of robot motion changes, robot independently can reach new target position, i.e. robot can complete not for a certain appointed task, and also have the ability of extensive popularization for other tasks;And the on-line study feature of dynamic motion basic-element model and its automatic obstacle avoiding function combine the efficiency for improving path planning.
Description
Technical field
It is specifically a kind of based on dynamic motion primitive learning model the present invention relates to mobile robot path planning field
Method for planning path for mobile robot.
Background technique
Path planning is one of key technology of mobile robot, it indicates mobile robot intelligence to a certain extent
Horizontal height, can rapidly find out a convenient, collisionless path not only ensure that the safety of mobile robot itself, more body
The high efficiency and reliability of Xian Liao robot.
Currently, 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 mature and more efficient planing method in path planning model, is made extensively with its simple mathematical computations
With.But traditional Artificial Potential Field Method the problems such as there are local minimum point and goal nonreachables.Currently, there are many solve local pole
The method of dot, such as heuristic search, random escape method etc., but these improved Artificial Potential Field Methods only apply robot attached
The control force added, does not tackle the problem at its root.Genetic model is a kind of more search models based on heredity and natural selection,
Have many advantages, such as that robust, flexible, search is not easy to fall into local minimum points in population.But genetic model is carrying out robot path
That there are population scales when planning is big, search space is big, is easily trapped into local minimum point, the problems such as convergence rate is slow.
Above traditional robot path planning's model is primarily present the problem of following two aspect:
(1) task is specific, has fine performance only for a certain task, without extensive Generalization Ability;
(2) study is often offline, and which results in wanting re -training to learn new scene, real-time is very poor.
Summary of the invention
The technical problem to be solved by the present invention is for the reality in the presence of above-mentioned method for planning path for mobile robot
The problem of when property is poor and mobile robot completes job order one, proposes a kind of road based on dynamic motion primitive learning model
Diameter planing method.Can real-time searching route, the effect that can effectively improve path planning is combined with its automatic obstacle avoiding function
Rate, in addition, robot when completing new task, can not have to re -training sample and the characteristic of original sample trace is kept to arrive
Up to new target position.
In order to solve the above technical problems, the invention provides the 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 main packet
Include following steps:
Step 1: the two-dimensional environment of robot motion being modeled, the two-dimensional environment interface of dummy robot's movement, machine
Device people is replaced with small filled circles, and barrier is various planar graphs;
Step 2: the manipulation using handle to robot, enable robot from starting point avoid with barrier collide and reach mesh
Punctuate;
Step 3: during step 2 carries out, acquisition robot motion trace data learns as dynamic motion primitive
The sample point of model, robot motion's track data include displacement, velocity and acceleration;
Step 4: displacement, velocity and acceleration data when the robot motion track obtained according to step 3 count these
According to as training sample, sample is trained to obtain corresponding to robot motion track most by dynamic motion primitive algorithm
Good weighted value;
Step 5: initial parameter being set for particular task, the initial parameter includes starting point and the end of robot motion
Point, the best weights weight values obtained according to step 4 cook up the path after learning by dynamic motion basic-element model, path tool
There is the characteristic of original sample track, i.e. beginning and end is consistent, and its running track is roughly the same with sample trace;
Step 6: on the basis of step 5, round barrier is added, and coupling is added in original kinetics equation
, so that building has the dynamic system of barrier avoiding function, realize the automatic obstacle avoiding function of dynamic motion primitive learning model;
Step 7: on the basis of step 5, changing the target position of robot motion, in the premise of not re -training sample
Under, only change the parameter of target position, robot remains to independently reach new aiming spot, i.e. robot can complete not
For a certain appointed task, and also there is the ability of extensive popularization for other tasks.
In above-mentioned technical proposal, the two-dimensional environment of robot motion is modeled in step 1, the requirement of modeling are as follows: move
The scope of activities of mobile robot is in a limited two-dimensional space;On the basis of the size of mobile robot, by the ruler of barrier
It is very little to extend to the outside, regard robot as a particle;Barrier is made of various planar graphs, Limited Number, and in machine
These barriers will not be changed and be moved in people's moving process.
In above-mentioned technical proposal, detailed process is as follows for step 2:
Step 2-1: the data of read machine manpower handle, when handle is pushed to upper and lower or left and right, which shows in real time
Show the displacement, velocity and acceleration that robot moves in modeling environment;
Step 2-2: remote-control handle, the artificial optimal path cooking up a robot and capable of reaching home from starting point,
In view of robot generally can only front and back and side-to-side movement, therefore cook up come path be also front and back or side-to-side movement road
Diameter cooks up the track come and is also referred to as sample trace;
Step 2-3: in planning path, avoiding obstacles, and the method saved with data is by the position of sample trace
Move, the value of velocity and acceleration is recorded, and as sample data.
In above-mentioned technical proposal, step 4 is comprised the following specific steps that:
Step 4-1: establish the mathematical model of dynamic motion primitive: dynamic motion primitive is generally used to form discrete fortune
It is dynamic, y is displaced for single freedom degree, introduces and has constant coefficients linear differential equation and referred to as dynamic system, this is
System is as the basis to motor learning:
In formula:
X and v is the displacement and speed of system respectively;x0It is initial position and target position respectively with g;τ is that the time is flexible
The factor;K is the coefficient of elasticity of spring;D is the damped coefficient that system is under critical state;F is nonlinear function, for generating
Arbitrarily complicated movement;
Step 4-2: setting initial parameter, the starting point x of robot motion0With target point g, timeconstantτ, the bullet of spring
Property coefficient K, system are in the damped coefficient D under critical state;Nonlinear function f is used to form arbitrarily complicated movement, defines f
Are as follows:
In formula:
ψiIt (s) is Radial basis kernel function, i indicates i-th of Radial basis kernel function ψi(s), value range is 1 to N, wherein N
Indicate 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, but depends on phase variant s, the expression-form of s are as follows:
In formula:
S is the function about time t, and α is any normal number, and τ is time contraction-expansion factor;The s known to equation (5) is by 1
To 0 monotone decreasing, therefore equation (5) is known as cannoncial system;
Step 4-3: sample data obtained in step 3 is substituted into formula (1) and formula (2), because cannoncial system is
Integrable, i.e. s can be calculated according to parameter τ, so the nonlinear disturbance f ' (s) in training sample can be expressed as:
Best weights weight values w is solved according to minimum error principle function Ji, the wherein expression formula of minimum error principle function
Are as follows:
J=∑s(f′(s)-f(s))2 (7)
W when J takes minimumiIt is exactly optimal weighted value.
In above-mentioned technical proposal, step 5 is comprised the following specific steps that:
Step 5-1: when robot executes specified task, start position and the final position of robot are set;
Step 5-2: sample data is two-dimensional, namely including the data in the data and y-axis direction in x-axis direction, by x
Data in axis direction are trained according to step 4, obtain the best weights weight values in x-axis direction, substitute into the starting point in step 5-1
And end point values, the side x is calculated upwardly through the displacement after the study of dynamic motion basic-element model, velocity and acceleration;
Step 5-3: the data on y-axis direction are trained according to step 4, obtain the best weights weight values on y-axis direction,
The beginning and end value in step 5-1 is substituted into, calculates the side y upwardly through the displacement after the study of dynamic motion basic-element model, speed
Degree and acceleration;
Step 5-4: data obtained in step 5-2 and step 5-3 are read in, the fortune of x-axis and y-axis both direction is respectively obtained
Dynamic data, export the track emulation figure of movement on two-dimensional surface, that is, complete based on dynamic motion primitive learning model to movement
The path planning of robot.
In above-mentioned technical proposal, the barrier being added in step 6 is with (0.4,0.4) for central coordinate of circle, and radius is
The circle of 0.1m.
Method of the invention starts to begin one's study with simple linear dynamic system (one group of differential equation), is by conversion
Simple linear dynamic system is converted into nonlinear system by system, and arbitrarily complicated movement is formed by attractor, this
Sample better simply can study nonlinear system.Wherein, it is that error can be certainly the advantages of expression with the differential equation
Dynamic is corrected, and the differential equation is formed with fixed format, only needs letter according to this format fixed
As soon as single change target component, adapts to new environment, it can carried out to fresh target extensive;Based on dynamic motion primitive
The method of study is on-line study, and new situation is not had to relearn, can real-time tracking position of object.Thus,
On in terms of avoidance, automatic obstacle avoiding is realized by constructing the dynamic system with barrier avoiding function, and dynamic motion basic-element model
On-line study feature and its automatic obstacle avoiding function combine the efficiency for improving path planning.
Compared with prior art, the beneficial effects of the present invention are:
(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 can not have to re -training sample and keep former when completing new task
The characteristic for carrying out sample trace reaches new target position.
(2) path planning model proposed by the present invention is in real time, in conjunction with its automatic obstacle avoiding function in searching route
The efficiency of path planning can be effectively improved by getting up.
Detailed description of the invention
Fig. 1 is that the present invention is based on the mobile robot path planning process schematics of dynamic motion primitive learning model;
Fig. 2 is the two-dimensional environment that dummy robot moves in the present invention;Wherein linear type track representative sample track;It is various
The figure (such as rectangle, circle, ellipse) of shape indicates 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 dynamic motion primitive learning model has extensive Generalization Ability in the present invention;
Fig. 5 is trained sample trace and passes through the track comparison diagram after dynamic motion basic-element model learns.
Specific embodiment
Technical solution in order to further illustrate the present invention, the present invention will be described in detail by 1-5 with reference to the accompanying drawing.
Step 1: the two-dimensional environment interface of dummy robot's movement, robot is replaced with small filled circles on interface, is hindered
Hindering object is various planar graphs;The two-dimensional environment interface that robot motion is arranged is square (long and width is all 1m), robot
It is replaced with a diameter by the small filled circles of 5mm.
Step 2: realizing handle to robot using OPENCV (Open Source Computer Vision Library)
Manipulation, enable robot from starting point avoid and barrier collision reach target point;
Step 2-1: writing a upper computer software based on the interface MFC (Microsoft Foundation Classes),
The software can be with the data of read machine manpower handle, and when handle is pushed to upper and lower or left and right, which can show in real time
Displacement that robot moves in modeling environment, velocity and acceleration;
Step 2-2: remote-control handle, the artificial optimal path cooking up a robot and capable of reaching home from starting point,
In view of robot generally can only front and back and side-to-side movement, therefore cook up come path be also front and back or side-to-side movement road
Diameter cooks up the track come and is also referred to as sample trace;
Step 2-3: in planning path, avoiding obstacles, and the method saved with data is by the position of sample trace
Move, the value of velocity and acceleration is recorded, and as sample data;
It is the upper computer software write based on MFC used in step 2, can be achieved with by the manipulation to handle to machine
The control of people.The velocity magnitude when displacement for being provided with handle push rod is robot motion, wherein control robot motion's speed
The range of degree is -5mm/s~5mm/s.
Step 3: during step 2 carries out, acquisition robot motion trace data learns as dynamic motion primitive
The sample point of model, wherein robot motion's track data includes the size of its displacement, velocity and acceleration value;
Step 4: these data are made in displacement, velocity and acceleration when the robot motion track obtained according to step 3
For the training sample of DMP learning model, best weights weight values corresponding to robot motion track are obtained by the training to sample;
Step 4-1: the mathematical model of dynamic motion primitive is established.Dynamic motion primitive is generally used to form discrete fortune
It is dynamic, y is displaced for single freedom degree, introduces and has constant coefficients linear differential equation and referred to as dynamic system, this is
System is as the basis to motor learning:
In formula:
X and v is the displacement and speed of system respectively;x0It is initial position and target position respectively with g;τ is that the time is flexible
The factor;K is the coefficient of elasticity of spring;D is the damped coefficient that system is under critical state;F is nonlinear function, for generating
Arbitrarily complicated movement;
Step 4-2: setting initial parameter, the starting point x of robot motion0With target point g, timeconstantτ, the bullet of spring
Property coefficient K, system are in the damped coefficient D under critical state;Nonlinear function f is used to form arbitrarily complicated movement, definition
Are as follows:
In formula:
ψiIt (s) is Radial basis kernel function, i indicates i-th of Radial basis kernel function ψi(s), value range is 1 to N, wherein N
Indicate 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, but depends on phase variant s, the expression-form of s are as follows:
In formula:
S is the function about time t, and α is any normal number, and τ is time contraction-expansion factor;The s known to equation (5) is by 1
To 0 monotone decreasing, therefore equation (5) is known as cannoncial system;
Step 4-3: sample data obtained in step 3 is substituted into above-mentioned formula, because cannoncial system is integrable,
That is s can be calculated according to parameter τ, so the nonlinear disturbance f ' (s) in training sample can be expressed as:
Best weights weight values w is solved according to minimum error principle function Ji, the wherein expression formula of minimum error principle function
Are as follows:
J=∑s(f′(s)-f(s))2 (7)
W when J takes minimumiIt is exactly optimal weighted value;
Step 5: for particular task setting initial parameter (beginning and end of robot motion), being obtained according to step 4
Best weights weight values, cook up by dynamic motion basic-element model learn after path, the path have original sample track spy
Property;
Step 5-1: when robot executes specified task, start position and the final position of robot are set;
Step 5-2: sample data is two-dimensional (data in data and y-axis direction in x-axis direction), by x-axis direction
On data be trained according to step 4, obtain the best weights weight values in x-axis direction, substitute into the beginning and end in step 5-1
Value, so that it may calculate the side x upwardly through the displacement after the study of dynamic motion basic-element model, velocity and acceleration;
Step 5-3: the 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, so that it may calculate the side y upwardly through the position after the study of dynamic motion basic-element model
It moves, velocity and acceleration;
Step 5-4: data obtained in step 5-2 and step 5-3 are read in by MATLAB, obtain x-axis and y-axis two
The exercise data in direction exports the track emulation figure of movement on two-dimensional surface, that is, completes to learn mould based on dynamic motion primitive
Path planning of the type to mobile robot.
Step 6: on the basis of step 5, round barrier is added, and coupling is added in original kinetics equation
, so that building has the dynamic system of barrier avoiding function, realize the automatic obstacle avoiding function of dynamic motion primitive learning model;
Step 6-1: on the basis of step 5-4, round barrier is added, wherein barrier is with (0.4,0.4) for circle
Heart coordinate, radius are the circle of 0.1m;
Step 6-2: coupling terms P (x, v) is added on the basis of the dynamic systems equation that step 4-1 is provided to construct band
There is the dynamic system of barrier avoiding function, wherein the expression formula of coupling terms P (x, v) are as follows:
In formula:
Be withFor axis,For the spin matrix of rotation angle, vectorIt is the position of barrier, γ and β
It is constant, θ is the angle on the distance vector and track of the point on track and barrier between the relative velocity of that point;
Step 6-3: giving the constant term initial value in coupling terms P (x, v) formula, wherein γ=8,Spin matrix R
It is expressed as:
Step 6-4: having the dynamic system of barrier avoiding function by constructing, and barrier described in step 6-1, machine is added
Device people remains to avoiding obstacles and reaches target point, wherein the mathematic(al) representation of the dynamic system with barrier avoiding function are as follows:
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, changing the target position of robot motion, in the premise of not re -training sample
Under, only change the parameter of target position, robot remains to the new aiming spot of autonomous arrival, i.e. robot can complete
It is not directed to a certain appointed task, and also there is the ability of extensive popularization for other tasks.
Step 7-1: on the basis of step 5-4, the position for changing robot target point is (0.5,0.5), substitutes into step
4, it obtains under 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 for changing robot target point is (0.8,0.8), substitutes into step
4, it obtains under 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 position, and
And the characteristic for this track of keeping intact, hence it is demonstrated that extensive Generalization Ability possessed by dynamic motion primitive learning model;
To sum up, the present invention is based on dynamic motion primitive learning models to realize the path planning to robot, the learning model
On-line study feature and its automatic obstacle avoiding function combine the efficiency for improving path planning, and the model has extensive push away
Wide ability.It is proposed of the invention improves the intelligence of mobile robot, in path planning, avoidance and leads for mobile robot
The related fieldss such as boat provide reference.
Claims (5)
1. a kind of 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 being modeled, the two-dimensional environment interface of dummy robot's movement, robot
It is replaced with small filled circles, barrier is various planar graphs;
Step 2: the manipulation using handle to robot, enable robot from starting point avoid with barrier collide and reach target
Point;
Step 3: during step 2 carries out, acquiring robot motion trace data as dynamic motion primitive learning model
Sample point, robot motion's track data include displacement, velocity and acceleration;
Step 4: these data are made in displacement, velocity and acceleration data when the robot motion track obtained according to step 3
For training sample, sample is trained by dynamic motion primitive algorithm to obtain best weights corresponding to robot motion track
Weight values;
Step 5: initial parameter being set for particular task, the initial parameter includes the beginning and end of robot motion, root
The best weights weight values obtained according to step 4, cook up the path after learning by dynamic motion basic-element model, which has as former state
The characteristic of this track, i.e. beginning and end are consistent, and its running track is roughly the same with sample trace;
Step 6: on the basis of step 5, round barrier is added, and coupling terms are added in original kinetics equation,
To which building has the dynamic system of barrier avoiding function, the automatic obstacle avoiding function of dynamic motion primitive learning model is realized;
Step 7: on the basis of step 5, change the target position of robot motion, under the premise of not re -training sample,
Only change the parameter of target position, robot remains to independently reach new aiming spot, i.e. robot can complete not needle
To a certain appointed task, and also there is the ability of extensive popularization for other tasks;
Step 4 dynamic motion primitive algorithm comprises the following specific steps that:
Step 4-1: establish the mathematical model of dynamic motion primitive: dynamic motion primitive is generally used to form discrete movement, right
It is displaced y in single freedom degree, introduces and has constant coefficients linear differential equation and referred to as dynamic system, this system conduct
To the basis of motor learning:
In formula:
X and v is the displacement and speed of system respectively;x0It is initial position and target position 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 state;F is nonlinear function, any multiple for generating
Miscellaneous movement;
Step 4-2: setting initial parameter, the starting point x of robot motion0With target point g, timeconstantτ, the elasticity system of spring
Number K, system are in the damped coefficient D under critical state;Nonlinear function f is used to form arbitrarily complicated movement, defines f are as follows:
In formula:
ψiIt (s) is Radial basis kernel function, i indicates i-th of Radial basis kernel function ψi(s), value range is 1 to N, and wherein N is indicated
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 the width for determining kernel function;Wherein hN=hN-1, i=1 ... N, α are any normal number;
Function f in formula (3) is not dependent on time parameter, but depends on phase variant s, the expression-form of s are as follows:
In formula:
S is the function about time t, and α is any normal number, and τ is time contraction-expansion factor;The s known to equation (5) is single by 1 to 0
What tune successively decreased, therefore equation (5) is known as cannoncial system;
Step 4-3: sample data obtained in step 3 is substituted into formula (1) and formula (2), because cannoncial system is can to accumulate
Point, i.e. s can be calculated according to parameter τ, so the nonlinear disturbance f ' (s) in training sample can be expressed as:
Best weights weight values w is solved according to minimum error principle function Ji, the wherein expression formula of minimum error principle function are as follows:
J=∑s(f′(s)-f(s))2 (7)
W when J takes minimumiIt is exactly optimal weighted value.
2. the method for planning path for mobile robot according to claim 1 based on dynamic motion primitive learning model,
It is characterized in that: the two-dimensional environment of robot motion being modeled in step 1, the requirement of modeling are as follows: the activity of mobile robot
Range is in a limited two-dimensional space;On the basis of the size of mobile robot, the size of barrier is extended to the outside, by machine
Device people regards a particle as;Barrier is made of various planar graphs, Limited Number, and in robot moving process these
Barrier will not be changed and be moved.
3. the method for planning path for mobile robot according to claim 1 based on dynamic motion primitive learning model,
Be characterized in that: detailed process is as follows for step 2:
Step 2-1: the data of read machine manpower handle, when handle is pushed to upper and lower or left and right, which shows machine in real time
Displacement that device people moves in modeling environment, velocity and acceleration;
Step 2-2: remote-control handle, the artificial optimal path cooking up a robot and capable of reaching home from starting point consider
To robot generally can only front and back and side-to-side movement, therefore cook up come path be also front and back or side-to-side movement path,
It cooks up the track come and is also referred to as sample trace;
Step 2-3: in planning path, avoiding obstacles, and the method saved with data is 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 according to claim 1 based on dynamic motion primitive learning model,
Be characterized in that: step 5 comprises the following specific steps that:
Step 5-1: when robot executes specified task, start position and the final position of robot are set;
Step 5-2: sample data is two-dimensional, namely including the data in the data and y-axis direction in x-axis direction, by x-axis side
Upward data are trained according to step 4, obtain the best weights weight values in x-axis direction, substitute into the starting point in step 5-1 and end
Point value calculates the side x upwardly through the displacement after the study of dynamic motion basic-element model, velocity and acceleration;
Step 5-3: the data on y-axis direction are trained according to step 4, obtain the best weights weight values on y-axis direction, are substituted into
Beginning and end value in step 5-1, calculate the side y upwardly through dynamic motion basic-element model study after displacement, speed and
Acceleration;
Step 5-4: data obtained in step 5-2 and step 5-3 are read in, the movement number of x-axis and y-axis both direction is respectively obtained
According to exporting the track emulation figure of movement on two-dimensional surface, that is, complete based on dynamic motion primitive learning model to mobile machine
The path planning of people.
5. the method for planning path for mobile robot according to claim 1 based on dynamic motion primitive learning model,
Be characterized in that: the barrier being added in step 6 is with (0.4,0.4) for central coordinate of circle, 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 CN106444738A (en) | 2017-02-22 |
CN106444738B true 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) |
Families Citing this family (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10710592B2 (en) * | 2017-04-07 | 2020-07-14 | Tusimple, Inc. | System and method for path planning of autonomous vehicles based on gradient |
EP3452787B1 (en) * | 2017-04-27 | 2020-12-23 | Beijing Didi Infinity Technology And Development Co., Ltd. | Systems and methods for route planning |
CN107450535B (en) * | 2017-07-31 | 2018-06-29 | 中南大学 | A kind of intelligence carrying robot optimal path combination chart discusses controlling planning method |
CN107436604B (en) * | 2017-07-31 | 2018-03-16 | 中南大学 | Controlling planning method is intelligently decomposed in carrying robot path under a kind of intelligent environment |
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 |
DE102017217412A1 (en) * | 2017-09-29 | 2019-04-04 | Robert Bosch Gmbh | Method, apparatus and computer program for operating a robot control system |
CN108255182B (en) * | 2018-01-30 | 2021-05-11 | 上海交通大学 | Service robot pedestrian perception obstacle avoidance method based on deep reinforcement learning |
DE102018202321A1 (en) | 2018-02-15 | 2019-08-22 | Robert Bosch Gmbh | Coordination system, handling equipment and procedures |
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 |
CN108582072B (en) * | 2018-04-28 | 2020-09-15 | 北京邮电大学 | Improved graph planning algorithm-based space manipulator task planning method |
CN108919794B (en) * | 2018-06-01 | 2021-08-17 | 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 |
CN109108942B (en) * | 2018-09-11 | 2021-03-02 | 武汉科技大学 | Mechanical arm motion control method and system based on visual real-time teaching 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 |
CN110315544B (en) * | 2019-06-24 | 2022-10-14 | 南京邮电大学 | Robot operation learning method based on video image demonstration |
JP7221839B2 (en) * | 2019-10-08 | 2023-02-14 | 国立大学法人静岡大学 | Autonomous Mobile Robot and Control Program for Autonomous Mobile Robot |
CN110789530B (en) * | 2019-11-19 | 2021-04-09 | 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 |
CN111319044B (en) * | 2020-03-04 | 2022-08-09 | 达闼科技(北京)有限公司 | Article grabbing method and device, readable storage medium and grabbing robot |
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 |
CN112207835B (en) * | 2020-09-18 | 2021-11-16 | 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 |
CN112959330B (en) * | 2021-02-02 | 2022-05-17 | 浙江大学 | Robot double-arm motion man-machine corresponding device and method based on master-slave dynamic motion elements |
CN112975986B (en) * | 2021-03-25 | 2022-06-24 | 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 |
CN113043251B (en) * | 2021-04-23 | 2023-07-07 | 江苏理工学院 | Robot teaching reproduction track learning method |
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 |
CN114571458B (en) * | 2022-03-21 | 2022-11-15 | 苏州大学 | Micro-nano robot assembly track learning method based on dynamic motion elements |
CN116901055B (en) * | 2023-05-19 | 2024-04-19 | 兰州大学 | Human-simulated interaction control method and device, electronic 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 |
---|
基于动态运动基元的微小型四旋翼无人机路径规划;陈鹏震等;《高技术通讯》;20160228;第26卷(第2期);第186-194页 |
Also Published As
Publication number | Publication date |
---|---|
CN106444738A (en) | 2017-02-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106444738B (en) | Method for planning path for mobile robot based on dynamic motion primitive learning model | |
CN111618847B (en) | Mechanical arm autonomous grabbing method based on deep reinforcement learning and dynamic motion elements | |
CN110083165B (en) | Path planning method of robot in complex narrow environment | |
CN108115681B (en) | Simulation learning method and device for robot, robot and storage medium | |
CN111872934B (en) | Mechanical arm control method and system based on hidden semi-Markov model | |
CN106338919A (en) | USV (Unmanned Surface Vehicle) track tracking control method based on enhanced learning type intelligent algorithm | |
Chen et al. | Stabilization approaches for reinforcement learning-based end-to-end autonomous driving | |
CN109108942A (en) | The mechanical arm motion control method and system of the real-time teaching of view-based access control model and adaptive DMPS | |
CN110450156B (en) | Optimal design method of self-adaptive fuzzy controller of multi-degree-of-freedom mechanical arm system | |
CN109241552A (en) | A kind of underwater robot motion planning method based on multiple constraint target | |
CN108827312A (en) | A kind of coordinating game model paths planning method based on neural network and Artificial Potential Field | |
CN109240091A (en) | A kind of underwater robot control method based on intensified learning and its control method tracked | |
CN111506063B (en) | Mobile robot map-free navigation method based on layered reinforcement learning framework | |
CN107168309A (en) | A kind of underwater multi-robot paths planning method of Behavior-based control | |
CN114089776B (en) | Unmanned aerial vehicle obstacle avoidance method based on deep reinforcement learning | |
CN115373415A (en) | Unmanned aerial vehicle intelligent navigation method based on deep reinforcement learning | |
CN114973650A (en) | Vehicle ramp entrance confluence control method, vehicle, electronic device, and storage medium | |
Serra et al. | An optimal trajectory planner for a robotic batting task: the table tennis example | |
Wang et al. | Learning of long-horizon sparse-reward robotic manipulator tasks with base controllers | |
CN114564008A (en) | Mobile robot path planning method based on improved A-Star algorithm | |
Sun et al. | Integrating reinforcement learning and learning from demonstrations to learn nonprehensile manipulation | |
CN107065563A (en) | A kind of monomer machine fish Behavior- Based control strategy based on three-dimensional fuzzy PTP algorithms | |
CN114609925B (en) | Training method of underwater exploration strategy model and underwater exploration method of bionic machine fish | |
US20220051106A1 (en) | Method for training virtual animal to move based on control parameters | |
Metoyer et al. | Animating athletic motion planning by example |
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 |