CN113485356A - Robot rapid movement planning method - Google Patents
Robot rapid movement planning method Download PDFInfo
- Publication number
- CN113485356A CN113485356A CN202110851774.5A CN202110851774A CN113485356A CN 113485356 A CN113485356 A CN 113485356A CN 202110851774 A CN202110851774 A CN 202110851774A CN 113485356 A CN113485356 A CN 113485356A
- Authority
- CN
- China
- Prior art keywords
- robot
- new
- path
- collision
- collision detection
- 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 33
- 238000001514 detection method Methods 0.000 claims abstract description 47
- 238000005070 sampling Methods 0.000 claims abstract description 24
- 230000008569 process Effects 0.000 claims abstract description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 63
- 238000012544 monitoring process Methods 0.000 claims description 22
- 238000010586 diagram Methods 0.000 claims description 16
- 238000005457 optimization Methods 0.000 claims description 13
- 239000012636 effector Substances 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000009191 jumping Effects 0.000 claims description 3
- 238000009826 distribution Methods 0.000 abstract description 7
- 230000004888 barrier function Effects 0.000 abstract description 3
- 238000004088 simulation Methods 0.000 description 15
- 238000011156 evaluation Methods 0.000 description 9
- 230000008859 change Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000001133 acceleration Effects 0.000 description 2
- 230000003247 decreasing effect Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0217—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with energy consumption, time reduction or distance reduction criteria
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0221—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving a learning process
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)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention provides a robot rapid movement planning method, which takes a collision detection result in the process of sampling and generating a new path point as a reference, dynamically adjusts target offset probability, guides sampling point generation according to the dynamically adjusted target offset probability, and carries out robot movement path planning. The method can improve the rationality of the target offset probability under the condition of different barrier distributions, and can optimize the track while improving the planning speed.
Description
Technical Field
The invention relates to the field of robot motion planning, in particular to a robot rapid motion planning method.
Background
Algorithms proposed and used for the robot motion planning problem can be mainly divided into three types, namely a space search method based on geometric construction, a motion planning algorithm based on track optimization, a motion planning algorithm based on sampling and the like. When the high-dimensional robot motion planning problem is processed, the motion planning based on sampling is more efficient. The basic idea is to determine a limited obstacle avoidance configuration set which can fully represent free space connectivity and is used for solving a path diagram containing a motion planning problem in an obstacle environment. Two algorithms most commonly used in Random sampling planning are Probabilistic Roadmap (PRM) that is learned first and then queried, and Rapidly-expanding Random Tree (RRT) that is learned while querying. Compared with the PRM and the RRT algorithm, the PRM and RRT algorithm fully considers the constraints (such as incomplete constraint, kinematic dynamics constraint and the like) objectively existing in the robot, and the obtained track is more reasonable. Step M.LaValle introduces The idea of target bias in The Randomized resistive Planning [ J ] (The International Journal of Robotics Research,2001,20(5):378-400), proposes RRT algorithm (GOAL _ BIASED _ RRT, GB-RRT) based on target bias, and improves The search efficiency of The basic RRT algorithm by guiding to The target. However, the fixed target offset probability makes the path the same as the target offset probability when the path is expanded in the configuration space with sparse obstacles and the configuration space with dense obstacles, and when the obstacles are not uniformly distributed in the space, the fixed target offset probability cannot deal with the rapidity well. At the same time, a fixed value of the target bias probability also does not optimize the trajectory or path. A SPARSE-STABLE RRT algorithm (STABLE-SPARSE-SST, RRT) was proposed in Lavalle S M, Kuffner J.in Rapid-expanding Random Trees: Progress and projects [ C ] (4th International works on the Algorithmic bases of Robotics,2001:293-308), which is an asymptotically near-optimal planning method. The algorithm does not consider the action of target bias, still faces to all configuration space sampling, is not consistent with the actual direct target path from the initial path as far as possible, and simultaneously consumes a large amount of time to iterate to obtain the most effective track.
Disclosure of Invention
The invention aims to provide a robot rapid movement planning method aiming at the problems in the prior art, so that the reasonability of target offset probability under different obstacle distribution conditions is improved, and the trajectory optimization is carried out while the planning speed is improved.
The invention is realized by the following technical scheme:
a robot rapid movement planning method takes a collision detection result in the process of sampling and generating a new path point as a reference, dynamically adjusts target offset probability, guides sampling point generation according to the dynamically adjusted target offset probability, and carries out robot movement path planning.
Preferably, the method comprises the following steps:
step 2, defining a collision detection array Collision result for recording the collision detection result n times before the current state; the array has n elements, each element being 0 or 1; 0 represents that the collision detection result is no collision, and 1 represents that the collision detection result is collision; initializing a Collision result matrix into an all-zero matrix;
step 3, calculating the target offset probability p according to the collision detection array Collision resultgoal；
Step 4, generating probability p through random number, if p<pgoalThen the target configuration x is takengoalIs a sampling point ssampleOtherwise, sampling randomly;
step 5, calculating the position x of each joint of the robot and the position x of the end effector in the Cartesian space according to the sampling resultnearest(ii) a According to position xnearestObtaining an obstacle collision detection state;
step 6, according to the position xnearestPerforming node expansion through Monte Carlo, and combining positive kinematics to obtain the positions of each joint of the robot and the position x of the end effectornewThen according to position xnewPerforming collision detection with an obstacle collision detection state;
step 7, if collision occurs, updating a collision detection array and skipping to step 8;
if no collision occurs, updating a collision detection array and the motion state of the robot, updating a path diagram by combining a near optimal algorithm and a sparse algorithm, and jumping to the step 8;
step 8, judging whether the target is reached or the iteration times are exceeded, and if not, returning to the step 3; if yes, outputting a path diagram, and finishing the motion planning.
Further, in step 3, a target bias probability p is calculatedgoalThe formula is as follows:
further, in step 2, n is 10.
Preferably, in step 5, the method is based on a neighborhood optimization algorithm and positive kinematicsCalculating the position of each joint of the robot and the position x of the end effector in Cartesian spacenearest。
Preferably, in step 7, if a collision occurs, the collision detection array is updated as follows:
CollsionResult＝[CollsionResult(2,n),1]。
preferably, in step 7, if no collision occurs, the collision detection array is updated as follows:
CollsionResult＝[CollsionResult(2,n),0]。
preferably, in step 7, the path map is updated by combining the near optimal algorithm and the sparse algorithm, and the process jumps to step 8, specifically: at position xnewCentered on the sparse radius δsFor radius, the nearest monitoring state S is sought in the monitoring set Snew，xpeerIs a representative monitoring state S in the monitoring set SnewCalculating x at the path point in the state spacenewAnd xpeerCost function of (1), cost (x) respectivelynew) And cost (x)peer) (ii) a If cost (x)new)≥cost(xpeer) Then go to step 8; otherwise will xnearestTo xnewAdd the path of (c) to the existing path graph and update the active set of path points VactiveAnd a set of inactive path points VinactiveThen, the path is thinned, and the step 8 is performed.
Further, in step 7, if there is no position x in the monitoring set SnewNearby monitoring state snewThen x isnewAdded to the monitoring set S, and looked at xnewIs s isnew。
Compared with the prior art, the invention has the following beneficial technical effects:
the DP-SST planning algorithm of the dynamic target bias probability based on the collision feedback enables the path to be more inclined to the target point expansion in the environment with sparse obstacles and more inclined to the global expansion in the environment with dense obstacles by dynamically adjusting the target bias probability, and realizes the purpose of improving the planning speed and optimizing the track by combining the adjacent optimal algorithm. The invention can rapidly realize motion planning under the condition of uneven distribution of obstacles and simultaneously carry out path optimization. In addition, a reasonable target bias probability can be searched for different obstacle distribution environments, and reference is provided for other motion planning methods for fixing the target bias probability.
Drawings
FIG. 1 is an overall flow chart of the movement planning method (DP-SST) of the present invention;
FIG. 2 is a diagram of a DP-SST motion planning in a general obstacle environment;
FIG. 3 is a diagram of a change of a cost function in a DP-SST motion planning process in a general obstacle environment;
FIG. 4 shows a target offset probability p in the DP-SST motion planning process under a general obstacle environmentgoalA change in (c);
FIG. 5 is a schematic diagram of motion planning for other motion planning algorithms in a general obstacle environment; (a) a GB-RRT movement planning schematic diagram; (b) SST motion planning schematic;
FIG. 6 is a diagram illustrating a DP-SST movement plan in a special obstacle environment;
FIG. 7 is a diagram of a change in a cost function during a DP-SST motion planning process in a special obstacle environment;
FIG. 8 shows a target offset probability p in the DP-SST motion planning process under the environment of a special obstaclegoalA change in (c);
FIG. 9 is a schematic diagram of motion planning by other motion planning algorithms in a special obstacle environment; (a) a GB-RRT movement planning schematic diagram; (b) SST motion planning schematic;
Detailed Description
The present invention will now be described in further detail with reference to specific examples, which are intended to be illustrative, but not limiting, of the invention.
The invention provides an SST algorithm (DP-SST) of DYNAMIC target bias PROBABILITY based on collision detection feedback. The algorithm dynamically adjusts the target bias probability by taking a collision detection result in the process of generating a new path point by sampling as a reference. The DP-SST defines a Collision detection array Collision result which represents the dynamic change of the collision detection result, and records the collision detection result n times before the current state. Taking n-10 as an example, the array has 10 elements, each element being 0 or 1. 0 represents that the collision detection is no collision, and 1 represents that the collision detection is a collision. For example
CollisionResult＝[0,0,0,1,0,0,1,0,0,0]
The array indicates that the 4th and 7 th collision detection results in this state are that a collision has occurred, and the rest are not. Since the impact detection result too far ahead does not have a great influence on the expansion of the current state path diagram, only the impact detection result 10 times before the current state is considered. The target bias probability calculation rule is as follows
In subsequent expansion, the collision detection array sequentially records the collision detection results of the previous ten times, and data are updated. If all the elements of the collision detection array Collision result are 0, the condition is shown that no obstacle exists in the range contacted by the path diagram in the current state, and the target offset probability p is at the momentgoalLet the target point become the sampling point, the acceleration path map expands towards the target point. If all the elements of the collision detection array Collision result are 1, the range obstacles contacted by the path diagram are dense under the current state, and the target offset probability p is at the momentgoalAnd (5) driving the sampling points to sample in the global state, and expanding the acceleration path graph to the global state. The greater the number of times a collision is detected, the lower the target bias probability, and conversely the higher the target bias probability. By setting the target bias probability, the path is more biased to target point expansion in an environment with sparse obstacles and more biased to global expansion in an environment with dense obstacles, and the trajectory optimization can be realized while the planning speed is improved by combining a nearby optimal algorithm.
The overall flow chart of the movement planning method is shown in fig. 1, and specifically comprises the following steps:
And 2, initializing the Collision result matrix into an all-zero matrix.
Step 3, calculating the target bias probability pgoalThe formula is as follows:
and 4, generating the probability p through random numbers. If p is<pgoalThen the target configuration x is takengoalIs a sampling point ssampleOtherwise, random sampling is carried out.
And 5, calculating the positions of all joints of the robot and the position x of the end effector in the Cartesian space according to the sampling result and the adjacent optimal algorithm and positive kinematicsnearest. And according to position xnearestAn obstacle collision detection state is obtained. I.e. at the sampling point ssampleAs a center, solving the path point x with the minimum cost and the best quality through a near optimal algorithm (Bestnear) within the query radius range given in the step 1nearest。
Step 6, according to the position xnearestPerforming node expansion through Monte Carlo (Monte Carlo-Prop), and combining positive kinematics to obtain the positions of each joint of the robot and the position x of the end effectornewThen according to position xnewAnd performing collision detection with the obstacle collision detection state.
Step 7, if collision occurs, the collision detection array is updated as follows
CollsionResult＝[CollsionResult(2,n),1]
After the update is completed, it is determined whether the target is reached or the number of iterations is exceeded, i.e., step 8 is performed.
If no collision detection occurs, then
(1) The collision detection array is updated as follows
CollsionResult＝[CollsionResult(2,n),0]
(2) And updating the motion state of the robot, including the motion information of the base, the joint angle and the end effector of the robot, and updating the monitoring set S.
(3) At position xnewCentered on the sparse radius δsFor radius, the nearest monitoring state S is sought in the monitoring set Snew. If there is no location x in the monitoring set SnewNearby monitoring state snewThen x isnewAdded to the monitoring set S, and looked at xnewIs s isnew。xpeerIs a representative monitoring state S in the monitoring set SnewWaypoints in the state space. Calculating xnewAnd xpeerCost function of (1), cost (x) respectivelynew) And cost (x)peer) And comparing the path quality of the two nodes.
If cost (x)new)≥cost(xpeer) Then step 8 is performed, i.e. it is determined whether the target is reached or the number of iterations is exceeded, otherwise x is usednearestTo xnewAdd the path of (c) to the existing path graph and update the active set of path points VactiveAnd a set of inactive path points VinactiveThen, step 8 is performed.
Step 8, judging whether the target is reached or the iteration times are exceeded, and if not, returning to the step 3; if yes, outputting the path and finishing the motion planning.
Examples
Numerical simulation is carried out by taking the movement plan of plane particles in the environment of obstacles as an example, and the numerical simulation is combined with general pgoalAnd comparing the GB-RRT algorithm with the SST algorithm of 0.5, and performing simulation verification under two simulation environments of general obstacle distribution and special obstacle distribution with target traps, wherein the two simulation environments are considered that a planning algorithm containing target offsets can be trapped in the target traps wrapped by local obstacles and is difficult to or incapable of automatically jumping out of the target traps.
Setting evaluation indexes of path points under two environments
cost(xi)＝klocalgcost1(xi)+kgoalgcost2(xi)
Wherein the content of the first and second substances,
the length, x, that the waypoint has traveledjFor nodes in existing paths, xj_parentAs its parent node, cost2(xi)＝||xi-xgoalAnd | is the length of the path point from the target point. G | | | is the Euclidean distance between two nodes, i.e.
If only cost is set1(xi) For the evaluation index, the path always finds the minimum point, and the phenomenon of going back and going back occurs. If only cost is set2(xi) For evaluating the index, a point which is close to the target point but has a more tortuous path may be selected, and the quality of the path may not be measured. Thus passing through the scaling factor klocal∈[0,1]And kgoal∈[0,1]The ratio of the two is adjusted, the influence of the path which the path point has already traveled and the target point on the path point is weighed, and cost (x) is usedi) To characterize the path quality corresponding to the current point.
In order to compare the advantages and disadvantages of the three motion planning algorithms, the calculation method for evaluating the planning efficiency and the track optimization effect of the algorithms is designed as follows
Wherein n is1、n2、n3The number of iterations of the three motion planning algorithms reflects the planning efficiency of the motion planning, Ji1Smaller values indicate a relatively higher algorithm planning efficiency. L is1、L2、L3The optimization index result of the three motion planning algorithms, namely the path length obtained by planning, can reflectEffect of trajectory optimization, Ji2Smaller values indicate a relatively better optimization of the algorithm trajectory. In order to measure the planning efficiency and the track optimization effect of the algorithm as a whole, the calculation method of the overall evaluation index of the algorithm is as follows:
Jithe smaller the algorithm is, the better the effect is, the more the algorithm can optimize the track while improving the planning efficiency, and the value is not more than 1.
1. Motion planning in general obstacle environments
In the MATLAB simulation environment, the map size is set to [800,680 ]]Initial point is [50,50 ]]The target point is [700,600 ]]And a straight distance 851 is provided therebetween with an obstacle. Calculating the evaluation index according to the actual length, and taking klocalAnd kgoalAll are 1, and the simulation results are shown in fig. 2, 3 and 4.
Simulation results show that the planned path completely avoids the obstacles and no collision occurs. Meanwhile, in order to ensure the shortest path, the robot walks along a straight line as much as possible, and the length of the robot is 1281. And as the movement plan progresses, the evaluation function cost2(xi) The overall trend is decreasing, i.e. the path is always expanding towards the target point. cost (x)i) The total path is increased all the time, which accords with the actual situation that the total path is increased. Also from fig. 4, in this obstacle environment, if a fixed target bias probability is adopted, p is recommendedgoal0.3 to 0.4.
In the same obstacle environment, GB-RRT and SST simulation results are shown in FIG. 5. Table 1 shows simulation results of three planning algorithms in a general obstacle environment, which can be obtained from table 1, and compared with the GB-RRT algorithm, although the DP-SST algorithm has a large number of iterations and a low speed, the path obtained by the planning is shorter and better. Compared with SST, the DP-SST algorithm has a longer path length, but the iteration number is far smaller than that of the SST algorithm. The iteration times and the path length are comprehensively considered, and the DP-SST algorithm is better and consistent with the display result of the overall evaluation index. The DP-SST algorithm is also described from the side, so that not only is the planning efficiency improved, but also the trajectory optimization is carried out, and the problem that the planning efficiency conflicts with the trajectory optimization in the motion planning algorithm based on sampling is solved.
TABLE 1 comparison of the results of DP-SST algorithm with GB-RRT, SST algorithm in general obstacle environment
2. Movement planning in the context of special obstacles
The special obstacle environment is a "target trap" type obstacle environment. In the MATLAB simulation environment, the map size is set to [800,680 ]]Initiation point [50,50 ]]The target point is [550,600 ]]At a linear distance 743 between them, and between them a "target trap" type obstacle is arranged. Calculating the evaluation index according to the actual length, and taking klocalAnd kgoalAll are 1, and the simulation results are shown in fig. 6, 7 and 8.
Simulation results show that one path in the graph enters a target trap and directly advances towards a target, but the global expansion starts rapidly after the path meets an obstacle, and finally the path reaches a target point. The final path length is 1167. And as the movement plan progresses, the evaluation function cost2(xi) The overall trend is decreasing, i.e. the path is always expanding towards the target point. cost (x)i) Is increasing all the time, and accords with the actual situation. From fig. 8, in the obstacle environment, if the fixed target offset probability is adopted, p is recommendedgoal0.3 to 0.4.
The GB-RRT and SST simulation results are shown in FIG. 9 under the same obstacle environment. Table 2 shows the simulation results of three planning algorithms under a special barrier environment, and the DP-SST algorithm is superior to the GB-RRT algorithm in terms of iteration times and path length compared with the GB-RRT algorithm under the environment of target trap type barrier distribution. Although the path length obtained by the SST algorithm is shorter than that obtained by the DP-SST algorithm, the iteration number of the SST algorithm is far greater than that of the DP-SST algorithm. The iteration times and the path length are comprehensively considered, and the DP-SST algorithm is better and consistent with the display result of the overall evaluation index.
TABLE 2 comparison of the results of DP-SST algorithm with GB-RRT, SST algorithm in the context of special obstacles
Claims (9)
1. A robot rapid motion planning method is characterized in that a collision detection result in the process of sampling and generating a new path point is used as a reference, target offset probability is dynamically adjusted, and a sampling point is guided to be generated according to the dynamically adjusted target offset probability so as to plan a robot motion path.
2. The robot fast motion planning method of claim 1, comprising:
step 1, initializing a robot motion state and an obstacle motion state;
step 2, defining a collision detection array Collision result for recording the collision detection result n times before the current state; the array has n elements, each element being 0 or 1; 0 represents that the collision detection result is no collision, and 1 represents that the collision detection result is collision; initializing a Collision result matrix into an all-zero matrix;
step 3, calculating the target offset probability p according to the collision detection array Collision resultgoal；
Step 4, generating probability p through random number, if p<pgoalThen the target configuration x is takengoalIs a sampling point ssampleOtherwise, sampling randomly;
step 5, calculating the position x of each joint of the robot and the position x of the end effector in the Cartesian space according to the sampling resultnearest(ii) a According to position xnearestObtaining an obstacle collision detection state;
step 6, according to the position xnearestPerforming node expansion through Monte Carlo, and combining positive kinematics to obtain the positions of each joint of the robot and the position x of the end effectornewThen according to position xnewPerforming collision detection with an obstacle collision detection state;
step 7, if collision occurs, updating a collision detection array and skipping to step 8;
if no collision occurs, updating a collision detection array and the motion state of the robot, updating a path diagram by combining a near optimal algorithm and a sparse algorithm, and jumping to the step 8;
step 8, judging whether the target is reached or the iteration times are exceeded, and if not, returning to the step 3; if yes, outputting a path diagram, and finishing the motion planning.
3. The robot fast motion planning method of claim 2, wherein in step 3, a target bias probability p is calculatedgoalThe formula is as follows:
4. the robot fast motion planning method of claim 2, wherein in step 2, n is 10.
5. The method for rapid robot motion planning according to claim 1, wherein in step 5, the positions of the joints of the robot and the position x of the end effector in cartesian space are calculated according to a vicinity optimization algorithm and positive kinematicsnearest。
6. The method for planning rapid movement of a robot according to claim 1, wherein in step 7, if a collision occurs, the collision detection array is updated as follows:
CollsionResult＝[CollsionResult(2,n),1]。
7. the method for planning rapid movement of a robot according to claim 1, wherein in step 7, if no collision occurs, the collision detection array is updated as follows:
CollsionResult＝[CollsionResult(2,n),0]。
8. the robot rapid motion planning method according to claim 1, wherein in step 7, the path map is updated by combining a near optimal algorithm and a sparse algorithm, and the method jumps to step 8, specifically: at position xnewCentered on the sparse radius δsFor radius, the nearest monitoring state S is sought in the monitoring set Snew，xpeerIs a representative monitoring state S in the monitoring set SnewCalculating x at the path point in the state spacenewAnd xpeerCost function of (1), cost (x) respectivelynew) And cost (x)peer) (ii) a If cost (x)new)≥cost(xpeer) Then go to step 8; otherwise will xnearestTo xnewAdd the path of (c) to the existing path graph and update the active set of path points VactiveAnd a set of inactive path points VinactiveThen, the path is thinned, and the step 8 is performed.
9. The method for planning rapid movement of robot of claim 8, wherein in step 7, if there is no position x in the monitoring set SnewNearby monitoring state snewThen x isnewAdded to the monitoring set S, and looked at xnewIs s isnew。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110851774.5A CN113485356B (en) | 2021-07-27 | 2021-07-27 | Robot rapid movement planning method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110851774.5A CN113485356B (en) | 2021-07-27 | 2021-07-27 | Robot rapid movement planning method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113485356A true CN113485356A (en) | 2021-10-08 |
| CN113485356B CN113485356B (en) | 2022-06-21 |
Family
ID=77944168
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110851774.5A Active CN113485356B (en) | 2021-07-27 | 2021-07-27 | Robot rapid movement planning method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113485356B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107438754A (en) * | 2015-02-10 | 2017-12-05 | 御眼视觉技术有限公司 | Sparse map for autonomous vehicle navigation |
| CN107943053A (en) * | 2017-12-15 | 2018-04-20 | 陕西理工大学 | A kind of paths planning method of mobile robot |
| CN108776492A (en) * | 2018-06-27 | 2018-11-09 | 电子科技大学 | A kind of four-axle aircraft automatic obstacle avoiding and air navigation aid based on binocular camera |
| CN110497403A (en) * | 2019-08-05 | 2019-11-26 | 上海大学 | A kind of manipulator motion planning method improving two-way RRT algorithm |
| CN110703768A (en) * | 2019-11-08 | 2020-01-17 | 福州大学 | Improved dynamic RRT mobile robot motion planning method |
| CN110962130A (en) * | 2019-12-24 | 2020-04-07 | 中国人民解放军海军工程大学 | Heuristic RRT mechanical arm motion planning method based on target deviation optimization |
| CN111207767A (en) * | 2020-02-20 | 2020-05-29 | 大连理工大学 | Vehicle planning algorithm improved based on RRT algorithm |
| CN111752306A (en) * | 2020-08-14 | 2020-10-09 | 西北工业大学 | Unmanned aerial vehicle route planning method based on fast-expanding random tree |
| CN111897328A (en) * | 2020-07-17 | 2020-11-06 | 武汉理工大学 | Path planning method, device and equipment based on improved artificial potential field method |
| CN112356033A (en) * | 2020-11-09 | 2021-02-12 | 中国矿业大学 | Mechanical arm path planning method integrating low-difference sequence and RRT algorithm |
| CN112650306A (en) * | 2020-12-25 | 2021-04-13 | 北京理工大学 | Unmanned aerial vehicle motion planning method based on dynamics RRT |
| CN112809665A (en) * | 2020-12-16 | 2021-05-18 | 安徽工业大学 | Mechanical arm motion planning method based on improved RRT algorithm |
-
2021
- 2021-07-27 CN CN202110851774.5A patent/CN113485356B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107438754A (en) * | 2015-02-10 | 2017-12-05 | 御眼视觉技术有限公司 | Sparse map for autonomous vehicle navigation |
| CN107943053A (en) * | 2017-12-15 | 2018-04-20 | 陕西理工大学 | A kind of paths planning method of mobile robot |
| CN108776492A (en) * | 2018-06-27 | 2018-11-09 | 电子科技大学 | A kind of four-axle aircraft automatic obstacle avoiding and air navigation aid based on binocular camera |
| CN110497403A (en) * | 2019-08-05 | 2019-11-26 | 上海大学 | A kind of manipulator motion planning method improving two-way RRT algorithm |
| CN110703768A (en) * | 2019-11-08 | 2020-01-17 | 福州大学 | Improved dynamic RRT mobile robot motion planning method |
| CN110962130A (en) * | 2019-12-24 | 2020-04-07 | 中国人民解放军海军工程大学 | Heuristic RRT mechanical arm motion planning method based on target deviation optimization |
| CN111207767A (en) * | 2020-02-20 | 2020-05-29 | 大连理工大学 | Vehicle planning algorithm improved based on RRT algorithm |
| CN111897328A (en) * | 2020-07-17 | 2020-11-06 | 武汉理工大学 | Path planning method, device and equipment based on improved artificial potential field method |
| CN111752306A (en) * | 2020-08-14 | 2020-10-09 | 西北工业大学 | Unmanned aerial vehicle route planning method based on fast-expanding random tree |
| CN112356033A (en) * | 2020-11-09 | 2021-02-12 | 中国矿业大学 | Mechanical arm path planning method integrating low-difference sequence and RRT algorithm |
| CN112809665A (en) * | 2020-12-16 | 2021-05-18 | 安徽工业大学 | Mechanical arm motion planning method based on improved RRT algorithm |
| CN112650306A (en) * | 2020-12-25 | 2021-04-13 | 北京理工大学 | Unmanned aerial vehicle motion planning method based on dynamics RRT |
Non-Patent Citations (3)
| Title |
|---|
| 李猛: "基于智能优化与RRT算法的无人机任务规划方法研究", 《中国优秀博硕士学位论文全文数据库(博士)工程科技Ⅱ辑》 * |
| 杨大庆等: "基于快速扩展随机树的机械臂路径优化算法研究", 《组合机床与自动化加工技术》 * |
| 蔡文涛等: "基于改进RRT算法的机械臂路径规划", 《传感器与微系统》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113485356B (en) | 2022-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110887484B (en) | Mobile robot path planning method based on improved genetic algorithm and storage medium | |
| CN109764886B (en) | Path planning method | |
| Ostafew et al. | Learning‐based nonlinear model predictive control to improve vision‐based mobile robot path tracking | |
| Tuncali et al. | Utilizing S-TaLiRo as an automatic test generation framework for autonomous vehicles | |
| Kurniawati et al. | Motion planning under uncertainty for robotic tasks with long time horizons | |
| CN107063280A (en) | A kind of intelligent vehicle path planning system and method based on control sampling | |
| CN109597425B (en) | Unmanned aerial vehicle navigation and obstacle avoidance method based on reinforcement learning | |
| Huang et al. | Differentiable integrated motion prediction and planning with learnable cost function for autonomous driving | |
| CN109542117B (en) | Underwater vehicle rolling planning algorithm based on improved RRT | |
| Kim et al. | Bi-directional value learning for risk-aware planning under uncertainty | |
| Janoš et al. | Multi-goal path planning using multiple random trees | |
| CN113296520B (en) | Routing inspection robot path planning method integrating A and improved gray wolf algorithm | |
| CN116661469B (en) | Robot track error correction method and system | |
| CN114879660B (en) | Robot environment sensing method based on target drive | |
| CN114608585A (en) | Method and device for synchronous positioning and mapping of mobile robot | |
| CN113485356B (en) | Robot rapid movement planning method | |
| Heide et al. | Performance optimization of autonomous platforms in unstructured outdoor environments using a novel constrained planning approach | |
| Lu et al. | An information potential approach for tracking and surveilling multiple moving targets using mobile sensor agents | |
| CN116734877A (en) | Robot dynamic obstacle avoidance method based on improved A-algorithm and dynamic window method | |
| CN110849385A (en) | Trajectory planning method and system for searching conjugate gradient descent based on double-layer heuristic search | |
| CN115903816A (en) | Low-energy-consumption mobile robot path planning method | |
| Knepper et al. | Empirical sampling of path sets for local area motion planning | |
| Zhang et al. | An adaptive artificial potential function approach for geometric sensing | |
| Guo et al. | HPO-RRT*: a sampling-based algorithm for UAV real-time path planning in a dynamic environment | |
| Wurm et al. | Improved Simultaneous Localization and Mapping using a Dual Representation of the Environment. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |