CN108297105A - A kind of mechanical arm task level time optimal trajectory planning method - Google Patents
A kind of mechanical arm task level time optimal trajectory planning method Download PDFInfo
- Publication number
- CN108297105A CN108297105A CN201810042868.6A CN201810042868A CN108297105A CN 108297105 A CN108297105 A CN 108297105A CN 201810042868 A CN201810042868 A CN 201810042868A CN 108297105 A CN108297105 A CN 108297105A
- Authority
- CN
- China
- Prior art keywords
- mechanical arm
- task level
- level time
- time optimal
- path
- 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
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J11/00—Manipulators not otherwise provided for
- B25J11/005—Manipulators for mechanical processing tasks
- B25J11/0065—Polishing or grinding
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1656—Programme controls characterised by programming, planning systems for manipulators
- B25J9/1664—Programme controls characterised by programming, planning systems for manipulators characterised by motion, path, trajectory planning
Abstract
The present invention relates to a kind of mechanical arm task level time optimal trajectory planning methods, which is characterized in that cooks up machining path track between any two points first；Then machining path track is rearranged, obtains task level time optimal machining path track；Then it is relocated by starting point, converts task level time optimal machining path track to the machining path track that grinding machine arm can execute.The optimizing algorithm rearranged to trajectory path point uses simulated annealing.This programme, which can not only provide, meets the needs of task level collision prevention track in polishing processing, and satisfaction processing total time is optimal, can be used for high-speed automated production, has significant economic benefit.
Description
Technical field
The present invention relates to a kind of mechanical arm task level time optimal trajectory planning methods, more particularly to processing is with multiple discrete
The Motion trajectory of the mechanical arm of non-interconnected polishing region.
Background technology
One critical function of intelligent manufacturing system is can to plan that the task level time optimal of grinding machine arm is kept away automatically
Track is touched, this is directly related to the processing efficiency and intelligence degree of manufacture system.By dimensional visual measurement system detectio
It, can be in each Sanding Area in conjunction with the size of milling tools to after the distribution of the striped protrusion of multiple discrete non-interconnected polishing regions
The position of a certain number of machining path points is determined on domain.As long as the milling tools on mechanical arm traverses these machining path points
It can complete the polishing of workpiece surface.
It, usually can be in conjunction with polishing when workpiece surface has multiple discrete non-interconnected polishing regions that mechanical arm is needed to polish
Technique extracts a series of machining path points from these polishing regions, and milling tools is since the starting point of a polishing region
Terminating point is polished to terminate then to enter polishing of another polishing region until completing all polishing regions.It is entire in order to realize
Polishing task spend total time the entire bruting process of a minimum of requirement planning grinding machine arm the total track of collision prevention.
Simulated annealing (SA) is derived from statistical physics, is deposited according between solid annealing process and combinatorial optimization problem
Similarity and the iteration combinatorial optimization algorithm that introduces.SA is the extension of local search algorithm, it be different from local search it
Place is with optimal state of value in certain probability selection neighborhood.Theoretically it has been proved that it is a globally optimal solution, and with
Probability 1 is close to optimal solution, therefore simulated annealing can be used for optimizing the arrangement of polishing track and obtain task level time optimal and keep away
Touch trajectory planning.
Invention content
The purpose of the present invention is when providing the initial position of grinding machine arm, for waiting for that polishing region plans grinding machine
The time optimal machining locus of task level of arm.
The present invention provides a kind of mechanical arm task level time optimal trajectory planning methods, which is characterized in that including following
Step：
Step 1, machining path track between any two points is cooked up；
Step 2, trajectory path point is rearranged, obtains task level time optimal machining path；
Step 3, starting point reorientation converts globally optimal solution to the machining path that grinding machine arm can execute.
Further, step 1 includes following sub-step：
Step 11, the point cloud data for needing to polish is obtained by Three Dimensions Inspection System；
Step 12, collision prevention path planner initializes；
Step 13, using BiTRRT collision prevention path planning algorithms planning collision prevention path；
Step 14, collision detection and mechanical arm trajectory planning are carried out to the path of planning；
Step 15, repeatedly step 13, step 14 within the set time, preserve the paths that all successes are planned；
Step 16, a preferably time optimal collision prevention track.
Further, step 14 collision detection uses FCL algorithms.
Further, step 14 mechanical arm trajectory planning uses the inverse kinematics of TRAC-IK computing redundancy degree robot
Solution.
Further, step 14 before the inverse kinetics solution of computing redundancy degree robot in the collision prevention path cooked up
It is upper to be inserted into intensive point.
Further, the optimizing algorithm that step 2 rearranges trajectory path point uses simulated annealing, packet
Include following sub-step：
Step 21, Cooling -schedule parameter is initialized, the variation range of each parameter is given, it is random within the scope of this
An initial path m0 is selected, and calculates corresponding object function E (m0)；
Step 22, to tracing point sequence do a disturbance, generate it is new put in order, calculate corresponding object function E
(m), Δ E=E (m)-E (m0) is obtained；
Step 23, it executes and receives decision process；If Δ E<0, then new model m received；If Δ E≤0, new model m is pressed
Probability P=exp (- Δ E/T) is received, and T is Current Temperatures.When model is received, m0=m is set.
Step 24, at temperature T, step 22 and 23L times are repeated；
Step 25, temperature T is slowly reduced.
Step 26, step 24 and 25 is repeated, until convergence terminates.
Further, object function E is with total time at least for object function in step 21.
Further, T is reduced in geometry rule in step 25.
Further, the condition terminated in step 26 is T<0.001.
Further, step 3 includes following sub-step：
Step 31, the search optimal alignment trajectory path middle-range point nearest from mechanical arm initial position；
Step 32, using the nearest point of mechanical arm initial position as first position, it is sequentially arranged path point again.
Description of the drawings
Fig. 1 is multiple discrete non-interconnected polishing region machining path schematic diagrames.
Fig. 2 is time optimal sub-trajectory planning flow chart.
Fig. 3 is simulated annealing flow chart.
Specific implementation mode
Embodiment
To keep technical scheme of the present invention clearer, technical scheme of the present invention is made below in conjunction with attached drawing further
Detailed description.
Shown in Fig. 1, multiple discrete non-interconnected polishing region A1, A2 ... An for being distributed on the surface of the workpiece and randomly generate
Machining path.It can be seen that waiting for the m for including inside polishing region An for n-th from the partial enlarged view positioned at the lower left corners Fig. 1
A machining path points, i.e. vn1,vn2,…,vnm.The processing tasks of grinding machine arm refer to end effector (milling tools) from
Beginning, position P0 started, and traverses the polishing starting point in all polishing regions, polishing intermediate point and polishing terminating point, is eventually returned to end
Stop bit sets the process of P0.The irregular multiple polishing regions of distribution of workpiece surface, these polishing regions are used as can see from Figure 1
Oval lines surround.
In this programme the striped for needing to polish off is obtained using the Three Dimensions Inspection System in automatically grinding system first
Shape protrusion, in conjunction with the size of striated protrusion, the parameter of milling tools and technique for grinding three, obtains in these polishing regions
The three-dimensional coordinate of point.Set the starting point P of planning tasksstartWith terminating point PendPose, start time optimal sub-trajectory planning
Device starts path planning.Particular flow sheet is as shown in Figure 2.
201 indicate the initialization of subplan device, and setting planning time is 8 seconds, using BiTRRT collision prevention path planning algorithms；
202 indicate otherwise execution route planning is planned again if it is successful, executing 203；
203 indicate collision detection and mechanical arm trajectory planning.
Intensive point is inserted on the collision prevention path cooked up, using the inverse movement of TRAC-IK computing redundancy degree robot
Solution is learned, the success rate of trajectory planning is not only increased, the calculating time of inverse solution can be also reduced to a certain extent, in same time
It can obtain more collision prevention tracks.The present embodiment runs two Arithmetic of inverse kinematics, i.e. KDL simultaneously using inverse Jacobian method
And SQP.Under default situations, when any of these algorithms obtain answer, inverse kinetics solution calculates search and is immediately finished.
Collision detection uses FCL algorithms.In order to calculate static-obstacle cost, Euclidean distance variation (EDT) has been used
With geometry collision detection.Working space is divided into three-dimensional voxel grid by the present embodiment, and each voxel of precomputation with it is quiet recently
The distance of state obstacles borders.In addition, by using the shape of the approximate robot of one group of overlapping sphere.
204 indicate to preserve current collision prevention track data and make successfully to plan that number counter adds one, check and expend the time
It whether more than 201 setting values, if it does, terminating planner operation, preserves in the Q to container of track, if when not up to default
Between, repeat 201~204.
205 indicate to select time optimal collision prevention sub-trajectory.
The track returned in container is all acceptable solution, and for the same planning tasks, we can obtain
To a plurality of different collision prevention track.In order to assess the performance for these collision prevention tracks cooked up and select time optimal keep away
Track is touched, the present embodiment has used a kind of evaluation function of collision prevention track.Collision prevention track Q is defined hereiniEvaluation function beAnd
Wherein i ∈ [1, M], M indicate the number of the collision prevention track obtained in given time period.
So, optimal collision prevention track is exactly the evaluation of estimate F of trackoptimalIt is minimized fminWhen track, i.e.,：
By above step, time optimal collision prevention sub-trajectory is obtained, tracing point in n polishing region in corresponding diagram 1.It is all
The sum of process time of polishing region is fixed, need not do the optimization of collision prevention track again, but in a different order time
Different processing total times will be obtained by going through polishing region, and the direction of black arrow is exactly that milling tools is polished from one in Fig. 1
Mobile route of the region to another polishing region.Next, most using simulated annealing (SA) optimization task level total time
Few path.
Assuming that given polishing region v1,v2,…,vr.Milling tools initial position P0 and n-1 is located at polishing region both ends
Machining path points v11,v1m,v21,v2n,…,vr1,vro, these are put and are denoted as π1,π2,…,πn.Milling tools is from P0, successively
Enter polishing region from the initial point of polishing region to complete to exit into next polishing region from terminating point after polishing, until complete
At returning to P0 after whole polishing tasks.
Therefore, a solution of the task level collision prevention track cooked up can be expressed as the cycle of the sub-trajectory between any two points
Arrangement
π=(π1,π2,…,πn)
It can also be expressed as
π1→π2→…→πn→π1
A paths,
πi(1≤i≤n) is the point of i-th of process in the path.Obviously, meet
πi≠πjWhen the solution of i ≠ j is only feasible solution.
The set of all feasible solutions constitutes solution space S
S={ arrangements of all cycles of n point }
The total duration of the collision prevention track of non-processing process is
The object function that can be obtained by task level time optimal trajectory planning problem in this way is
Thus grinding machine arm preferably one group of collision prevention track, make each point merely through primary, and spend it is total when
Between shortest problem, be converted into the minimum value for asking object function (3), which can be realized by simulated annealing.
Fig. 3 is simulated annealing flow chart, is described in detail below：
301 indicate that model initialization, the variation range of each parameter of setting models randomly choose one within the scope of this
A initial model m0, and calculate corresponding object function E (m0)；
302 indicate a disturbance is done to track dot sequency, generate it is new put in order, calculate corresponding object function E
(m), Δ E=E (m)-E (m0) is obtained；
303 indicate to receive decision process.If Δ E<0, then new model m received.If Δ E≤0, new model m presses probability P
=exp (- Δ E/T) is received, and T is Current Temperatures.When model is received, m0=m is set；
304 indicate at temperature T, repeat L disturbance and receive decision process, that is, repeat 302,303L times；
305 indicate slowly to reduce temperature T, and wherein T is reduced in geometry rule, T=T*q, q=0.92.
306 indicate to repeat 304,305, until T<0.001 convergence.
After completing above-mentioned steps, we can get milling tools and traverse all polishing point v from any point11,v12,
v13,…,vn1,vn2,…,vnmOptimal arrangement, this can ensure according to such track processing take it is most short.But processing road
The arrangement of diameter point cannot directly be polished mechanical arm use.In actual production, the default location of grinding machine arm milling tools
What P0 was usually fixed, it is therefore desirable to using P0 as the starting of machining path and final position.So needing from these points, search
Rope goes out P0 and corresponding proximal most position Mu and places it in first position of new arrangement N in former arrange, other points are then in order
It is mobile, obtain the executable final machining path of mechanical arm.
Above scheme, which can not only provide, meets the needs of task level collision prevention track in polishing processing, and meets processing rail
The total time of mark is optimal, can be used for high-speed automated production, has significant economic benefit.
Certain above-described embodiment only technical concepts and features to illustrate the invention, its object is to allow the common skill of this field
Art personnel can understand technical solution of the present invention and implement accordingly, and it is not intended to limit the scope of the present invention.Especially exist
The track cooked up rearrange in searching process, and simulated annealing should not be taken as a limitation, can realize optimizing
Other algorithms of function, such as genetic algorithm should all be regarded as protection scope of the present invention.Before not departing from the principle of the invention
It puts, several improvement and optimization can also be made, these are improved and optimization also should be regarded as protection scope of the present invention.
Claims (10)
1. a kind of mechanical arm task level time optimal trajectory planning method, which is characterized in that include the following steps：
Step 1, machining path track between any two points is cooked up；
Step 2, machining path track is rearranged, obtains task level time optimal machining path track；
Step 3, starting point relocates, and task level time optimal machining path track is converted into what grinding machine arm can execute
Machining path track.
2. mechanical arm task level time optimal trajectory planning method as described in claim 1, which is characterized in that the step 1
Including following sub-step：
Step 11, the point cloud data for needing to polish is obtained by Three Dimensions Inspection System；
Step 12, collision prevention path planner initializes；
Step 13, using BiTRRT collision prevention path planning algorithms planning collision prevention path；
Step 14, collision detection and mechanical arm trajectory planning are carried out to the path of planning；
Step 15, repeatedly step 13, step 14 within the set time, preserve the paths that all successes are planned；
Step 16, time optimal collision prevention track is selected.
3. mechanical arm task level time optimal trajectory planning method as claimed in claim 2, which is characterized in that the sub-step
14 collision detections use FCL algorithms.
4. mechanical arm task level time optimal trajectory planning method as claimed in claim 2, which is characterized in that the sub-step
14 mechanical arm trajectory plannings use the inverse kinetics solution of TRAC-IK computing redundancy degree robot.
5. mechanical arm task level time optimal trajectory planning method as claimed in claim 4, which is characterized in that the sub-step
14 are inserted into intensive point before the inverse kinetics solution of computing redundancy degree robot on the collision prevention path cooked up.
6. mechanical arm task level time optimal trajectory planning method as described in claim 1, which is characterized in that the step 2
The optimizing algorithm rearranged to trajectory path point uses simulated annealing, including following sub-step：
Step 21, Cooling -schedule parameter is initialized, the variation range of each parameter is given, is randomly choosed within the scope of this
One initial path m0, and calculate corresponding object function E (m0)；
Step 22, to tracing point sequence do a disturbance, generate it is new put in order, calculate corresponding object function E (m), obtain
To Δ E=E (m)-E (m0)；
Step 23, it executes and receives decision process；If Δ E<0, then new model m received；If Δ E≤0, new model m presses probability P
=exp (- Δ E/T) is received, and T is Current Temperatures.When model is received, m0=m is set；
Step 24, at temperature T, step 22 and 23L times are repeated；
Step 25, temperature T is slowly reduced；
Step 26, step 24 and 25 is repeated, until convergence terminates.
7. mechanical arm task level time optimal trajectory planning method as claimed in claim 6, which is characterized in that the step 21
Middle object function E is with total time at least for object function.
8. mechanical arm task level time optimal trajectory planning method as claimed in claim 6, which is characterized in that the step 25
Middle T is reduced in geometry rule.
9. mechanical arm task level time optimal trajectory planning method as claimed in claim 6, which is characterized in that the step 26
The condition of middle end is T<0.001.
10. mechanical arm task level time optimal trajectory planning method as described in claim 1, which is characterized in that the step 3
Including following sub-step：
Step 31, the search optimal alignment trajectory path middle-range point nearest from mechanical arm initial position；
Step 32, using the nearest point of mechanical arm initial position as first position, it is sequentially arranged path point again.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810042868.6A CN108297105B (en) | 2018-01-17 | 2018-01-17 | Optimal trajectory planning method for task-level time of mechanical arm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810042868.6A CN108297105B (en) | 2018-01-17 | 2018-01-17 | Optimal trajectory planning method for task-level time of mechanical arm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108297105A true CN108297105A (en) | 2018-07-20 |
| CN108297105B CN108297105B (en) | 2021-07-06 |
Family
ID=62869035
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810042868.6A Active CN108297105B (en) | 2018-01-17 | 2018-01-17 | Optimal trajectory planning method for task-level time of mechanical arm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108297105B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109676610A (en) * | 2019-01-25 | 2019-04-26 | 温州大学 | A kind of breaker puts together machines people and its method of realizing working trajectory optimization |
| CN109910015A (en) * | 2019-04-12 | 2019-06-21 | 成都天富若博特科技有限责任公司 | The terminal end path planning algorithm of construction robot is smeared in a kind of mortar spray |
| CN110196590A (en) * | 2019-04-23 | 2019-09-03 | 华南理工大学 | A kind of time optimal trajectory planning system and method for robot path tracking |
| WO2021109575A1 (en) * | 2019-12-02 | 2021-06-10 | 广东技术师范大学 | Global vision and local vision integrated robot vision guidance method and device |
| CN113062601A (en) * | 2021-03-17 | 2021-07-02 | 同济大学 | Q learning-based concrete distributing robot trajectory planning method |
| WO2021253391A1 (en) * | 2020-06-19 | 2021-12-23 | 浙江大学 | Method of using robotic arm to complete non-repeatable covering task with the minimum number of lifts |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008001275A2 (en) * | 2006-06-30 | 2008-01-03 | Philips Intellectual Property & Standards Gmbh | Method of controlling the position of at least part of an autonomously moveable device |
| CN106041946A (en) * | 2016-05-23 | 2016-10-26 | 广东工业大学 | Image-processing-based robot polishing production method and production system applying same |
| CN106563900A (en) * | 2016-11-16 | 2017-04-19 | 华南理工大学 | Method for planning optimal welding track of automobile inner trims and exterior parts |
| CN106695802A (en) * | 2017-03-19 | 2017-05-24 | 北京工业大学 | Improved RRT<*> obstacle avoidance motion planning method based on multi-degree-of-freedom mechanical arm |
| CN106891339A (en) * | 2017-02-10 | 2017-06-27 | 广东省智能制造研究所 | Polishing process and milling robot with gravity compensation |
| CN107443373A (en) * | 2017-07-20 | 2017-12-08 | 广东工业大学 | Collision prevention method for planning track and device based on articulated arm robots |
-
2018
- 2018-01-17 CN CN201810042868.6A patent/CN108297105B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008001275A2 (en) * | 2006-06-30 | 2008-01-03 | Philips Intellectual Property & Standards Gmbh | Method of controlling the position of at least part of an autonomously moveable device |
| CN106041946A (en) * | 2016-05-23 | 2016-10-26 | 广东工业大学 | Image-processing-based robot polishing production method and production system applying same |
| CN106563900A (en) * | 2016-11-16 | 2017-04-19 | 华南理工大学 | Method for planning optimal welding track of automobile inner trims and exterior parts |
| CN106891339A (en) * | 2017-02-10 | 2017-06-27 | 广东省智能制造研究所 | Polishing process and milling robot with gravity compensation |
| CN106695802A (en) * | 2017-03-19 | 2017-05-24 | 北京工业大学 | Improved RRT<*> obstacle avoidance motion planning method based on multi-degree-of-freedom mechanical arm |
| CN107443373A (en) * | 2017-07-20 | 2017-12-08 | 广东工业大学 | Collision prevention method for planning track and device based on articulated arm robots |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109676610A (en) * | 2019-01-25 | 2019-04-26 | 温州大学 | A kind of breaker puts together machines people and its method of realizing working trajectory optimization |
| CN109910015A (en) * | 2019-04-12 | 2019-06-21 | 成都天富若博特科技有限责任公司 | The terminal end path planning algorithm of construction robot is smeared in a kind of mortar spray |
| CN110196590A (en) * | 2019-04-23 | 2019-09-03 | 华南理工大学 | A kind of time optimal trajectory planning system and method for robot path tracking |
| CN110196590B (en) * | 2019-04-23 | 2021-12-14 | 华南理工大学 | Time optimal trajectory planning method for robot path tracking |
| JP2022516852A (en) * | 2019-12-02 | 2022-03-03 | 広東技術師範大学 | Robot visual guidance method and device by integrating overview vision and local vision |
| WO2021109575A1 (en) * | 2019-12-02 | 2021-06-10 | 广东技术师范大学 | Global vision and local vision integrated robot vision guidance method and device |
| JP7212236B2 (en) | 2019-12-02 | 2023-01-25 | 広東技術師範大学 | Robot Visual Guidance Method and Apparatus by Integrating Overview Vision and Local Vision |
| WO2021253391A1 (en) * | 2020-06-19 | 2021-12-23 | 浙江大学 | Method of using robotic arm to complete non-repeatable covering task with the minimum number of lifts |
| JP2022542729A (en) * | 2020-06-19 | 2022-10-07 | 浙江大学 | How to use manipulators to complete non-repeatable covering tasks with the fewest number of lifts |
| JP7254384B2 (en) | 2020-06-19 | 2023-04-10 | 浙江大学 | How to use manipulators to complete non-repeatable covering tasks with the fewest number of lifts |
| US11897143B2 (en) | 2020-06-19 | 2024-02-13 | Zhejiang University | Method for performing non-revisiting coverage task by manipulator with least number of lift-offs |
| CN113062601B (en) * | 2021-03-17 | 2022-05-13 | 同济大学 | Q learning-based concrete distributing robot trajectory planning method |
| CN113062601A (en) * | 2021-03-17 | 2021-07-02 | 同济大学 | Q learning-based concrete distributing robot trajectory planning method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108297105B (en) | 2021-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108297105A (en) | A kind of mechanical arm task level time optimal trajectory planning method | |
| EP2657800B1 (en) | Iterative packing optimization | |
| CN102814813B (en) | The method and system of deadlock is automatically prevented from multi-robot system | |
| CN102880113B (en) | Laser cutting path optimizing method | |
| Lian et al. | Optimization of process planning with various flexibilities using an imperialist competitive algorithm | |
| CN109984678B (en) | Cleaning robot and cleaning method thereof | |
| CN112232545B (en) | AGV task scheduling method based on simulated annealing algorithm | |
| CN107330214A (en) | Spatial configuration optimal method based on discretization Yu heuristic evolution algorithm | |
| Wilson et al. | How to generate a thousand master plans: A framework for computational urban design | |
| Lenin et al. | Multi-objective optimization in single-row layout design using a genetic algorithm | |
| CN109460859A (en) | A kind of plant layout's optimization method | |
| CN113854906B (en) | Control method, device and equipment for cooperative operation of multiple cleaning robots | |
| CN116523165B (en) | Collaborative optimization method for AMR path planning and production scheduling of flexible job shop | |
| Zhou et al. | Multi-objective optimization of greening scheduling problems of part feeding for mixed model assembly lines based on the robotic mobile fulfillment system | |
| Motoda et al. | Bimanual shelf picking planner based on collapse prediction | |
| Huang et al. | Planning irregular object packing via hierarchical reinforcement learning | |
| Tanaka et al. | Simultaneous planning for item picking and placing by deep reinforcement learning | |
| Power et al. | Constrained stein variational trajectory optimization | |
| CN107066654A (en) | Towards the TRAJECTORY CONTROL dot picking method and apparatus in the enveloped box face of point cloud model | |
| CN114310872B (en) | Automatic vegetable-beating method for mechanical arm based on DGG point cloud segmentation network | |
| Fan et al. | Hierarchical path planner for unknown space exploration using reinforcement learning-based intelligent frontier selection | |
| CN116224926A (en) | Dynamic scheduling optimization method and device for single-piece small-batch flexible manufacturing workshops | |
| Ma et al. | Action planning for packing long linear elastic objects into compact boxes with bimanual robotic manipulation | |
| Tejer et al. | Robust and efficient task scheduling for robotics applications with reinforcement learning | |
| Gao et al. | Co-Optimization of Environment and Policies for Decentralized Multi-Agent Navigation |
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 |