CN113625780A - Distributed unmanned cluster cooperative motion path planning method capable of avoiding deadlock - Google Patents
Distributed unmanned cluster cooperative motion path planning method capable of avoiding deadlock Download PDFInfo
- Publication number
- CN113625780A CN113625780A CN202111184687.5A CN202111184687A CN113625780A CN 113625780 A CN113625780 A CN 113625780A CN 202111184687 A CN202111184687 A CN 202111184687A CN 113625780 A CN113625780 A CN 113625780A
- Authority
- CN
- China
- Prior art keywords
- constraint
- planning
- distributed unmanned
- distributed
- path planning
- 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
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
- G05D1/10—Simultaneous control of position or course in three dimensions
- G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
- G05D1/104—Simultaneous control of position or course in three dimensions specially adapted for aircraft involving a plurality of aircrafts, e.g. formation flying
Abstract
The invention discloses a deadlock-free distributed unmanned cluster cooperative motion path planning method, wherein N objects in a distributed unmanned cluster move in the same dimension space; the object is at the initial momentFrom the initial positionMove to respective target positionsAnd do not collide with each other during the movement; the method is characterized in that aiming at the distributed unmanned cluster movement path planning, the optimization problem and the model function of the distributed unmanned cluster cooperative movement path planning are constructed by defining constraint, target and variable, an infinite view is defined by adding terminal constraint, and then an early warning band is usedThe buffer Vono unit constructs obstacle avoidance constraints, and by introducing a penalty function related to an early warning zone into a target function, the problem of planning the cooperative motion path of the cluster system with complex dynamics is solved, the deadlock problem is avoided, and the feasibility is ensured.
Description
Technical Field
The invention relates to an unmanned intelligent cluster cooperative control method, in particular to a distributed unmanned cluster cooperative motion path planning method which is based on distributed model prediction control, avoids deadlock and ensures feasibility.
Background
Distributed unmanned cluster cooperative control is a technical problem of a complex system, individuals in a cluster can be in various forms such as unmanned aerial vehicles, unmanned vehicles and unmanned boats, and key technologies to be solved mainly include an information interaction mechanism, a multi-machine cooperative environment perception technology, a cooperative decision, a cluster cooperative motion path planning technology and the like. And the cluster coordinated movement path planning is a core part of the cluster coordinated movement path planning.
Aiming at a cluster cooperative motion path planning algorithm, the current relatively mature cluster distributed cooperative motion path planning algorithm mainly comprises the following steps: dispute-resolution (contentions-resolution), artificial potential field (posititialflies), distributed model predictive control (distributed model predictive control), and geometry (geometry). The dispute resolution method and the artificial potential field method are two commonly used distributed collaborative motion path planning algorithms at present, only play a limited mutual avoidance role under more conditions, cannot ensure the obstacle avoidance between clusters, cannot consider the motion model of a controlled object by a geometric rule, and are only suitable for a few application scenes such as a simple speed motion control system. In the distributed model predictive control method, through constructing a convex optimization problem containing position constraint and dynamic constraint, mutual obstacle avoidance can be ensured, and a complex kinematic model of a controlled object is considered. But the existing methods can not process the deadlock problem caused by the lack of a central coordinator in the distributed unmanned cluster. Deadlock problems, which are more likely to occur in dense clustering situations, refer to multiple objects blocking each other so that no parties can reach the destination. The prior art lacks a distributed unmanned cluster collaborative motion path planning method which can avoid deadlock and ensure feasibility.
Disclosure of Invention
The invention aims to provide a distributed unmanned cluster cooperative motion path planning method capable of ensuring feasibility and avoiding deadlock, which is used for solving the problem of cluster system cooperative motion path planning with complex dynamics, avoiding the deadlock problem and ensuring the feasibility.
Based on distributed model predictive control, aiming at distributed unmanned cluster motion path planning, an optimization problem and a model function of the distributed unmanned cluster collaborative motion path planning are constructed by defining constraints, targets and variables, specifically, firstly, a concept of infinite view is introduced by adding terminal constraints, so that feasibility of an optimization problem solution in the model predictive control is realized, and therefore, feasibility of a motion path planning algorithm is ensured, secondly, a buffer Weino unit (warnebufferdVoronoi cell) with an early warning band is used for constructing an obstacle avoidance constraint, and a penalty function related to the early warning band is introduced into a target function, so that feasibility of the motion path planning algorithm is ensured, and a deadlock problem is thoroughly avoided.
The technical scheme provided by the invention is as follows:
a method for planning cooperative motion paths of a distributed unmanned cluster to avoid deadlock is disclosed, wherein individual objects in the distributed unmanned cluster comprise unmanned planes, unmanned vehicles and unmanned boats in various forms; the distributed unmanned cluster has N objects moving in the same dimension space;Nthe objects are identicalMoving in space at an initial momentFrom the initial positionMove to respective target positionsAnd do not collide with each other during the movement; aiming at the distributed unmanned cluster movement path planning, an optimization problem and a model function of the distributed unmanned cluster cooperative movement path planning are constructed by defining constraint, target and variable, specifically, firstly, an infinite view is introduced by adding terminal constraint to realize the feasibility of an optimization problem solution in model prediction control, thereby ensuring the feasibility of a movement path planning algorithm, and secondly, the method uses a solution containing a target and a variableConstructing obstacle avoidance constraints by buffer Vono units of the early warning zone, and avoiding deadlock by introducing penalty functions related to the early warning zone into the objective function; the distributed unmanned cluster cooperative motion path planning method comprises the following steps:
1) initializing a preset path in a distributed unmanned cluster to an objectiFor example, the predetermined path is recorded as,In order to be the initial position of the device,ibeing an object in a distributed unmanned cluster;is the initial time; horizon of the eyeWhereinKIs the length of the visual field;
2) any one object in distributed unmanned clusteriCommunicating respective preset path information with other objects;
3) acquiring state information including position information and speed information through a real-time positioning system;
4) obtaining an avoidance coefficient;As an objectiAbout an objectjThe avoidance angle parameter of (1);is a constant;
5) according to the obtained state information and the avoidance coefficient, an infinite view field is defined by adding terminal constraint, the avoidance constraint is constructed by using a buffer Voronoi unit containing the early warning zone, a penalty function related to the early warning zone is introduced into a target function, and distribution is constructedOptimizing problems and model functions of the formula unmanned cluster collaborative motion path planning; and solving to obtain input informationAnd informationNamely, the planning of the distributed unmanned cluster cooperative motion path for avoiding deadlock is realized; the specific process is as follows:
51) constructing a distributed unmanned cluster cooperative motion path planning optimization problem (model function), and realizing motion path planning by adopting a distributed model prediction control method; the method comprises the following steps:
any one object in distributed unmanned clusteriFrom an initial positionMove to the target movement positionIt has the dynamics of a second order system;
each object in the distributed unmanned cluster at each elapsed time intervalhLength of field of viewConstructing and solving the convex-down optimization problem for re-planning; the constructed distributed unmanned cluster cooperative motion optimization path planning model function is expressed as follows:
Wherein the content of the first and second substances,is the field of visionkA control input of;view obtained for planningkThe state quantity of (c); ;indicating the horizonkA sequence of positions;indicating the horizonkA sequence of velocities;Cplanning an objective function of an optimization model for the distributed unmanned cluster cooperative motion path;
52) defining optimization variables, including control inputsState quantity ofAnd early warning belt variables(ii) a Self-defining early warning zone variables;
initial position specification in planning variablesAnd an initial speed is defined asWhereinAndis an object oftActual position of time and inThe actual speed at the moment is obtained by measurement;
53) defining constraints comprising: the method comprises the following steps of dynamic constraint, obstacle avoidance constraint, early warning zone constraint, input constraint, speed constraint and terminal constraint; self-defining obstacle avoidance constraint, early warning zone constraint and terminal constraint;
the kinetic constraint, expressed as:
wherein:
obstacle avoidance and restraint: the following linear constraints are obtained by dividing the space through a buffer voronoi unit with an early warning zone:
is the last momentWhen in the visual fieldkThe planning result is also called as a preset path; first planning time takingWhereinIs an initial position; extend minimum distance(ii) a Wherein the content of the first and second substances,is the minimum distance between two objects;
the early warning band constraint is expressed as:
wherein the content of the first and second substances,the width of the buffer zone is indicated and is set manually.
The input constraints are:
the speed constraint is:
the terminal constraints are:
target function of constructed distributed unmanned cluster cooperative motion path planning optimization modelRepresented by formula (2):
Wherein the content of the first and second substances,Cis an objective function;is the visual fieldkThe weight constant of (1);is a target position of the object; is an objectiAbout an objectjOf whereinIs a constant number of times that the number of the first,is an objectiAbout an objectjThe avoidance angle parameter;
54) solving the cluster motion path planning problem, wherein the obtained result comprises input informationAnd location information;
55) The solution result is obtainedAstInput of time of dayThe motion is input into a lower-layer motion control system, so that the motion control is realized, and the purpose of motion obstacle avoidance is achieved;
6) location information results obtained from optimization problemsObtaining a predetermined path(ii) a Wherein h is a set time interval;
namely: generating a position information resultSet as a preset path(ii) a Wherein the content of the first and second substances,kfirst fingerkAt each FOV;his a set time interval;tis the time;ifirst fingeriAn object.
Through the steps, the planning of the distributed unmanned cluster cooperative motion path for avoiding deadlock is realized.
The invention is implemented by adopting a crazyfiles unmanned aerial vehicle system to realize the process of the method. In specific implementation, the optimization model is solved by the convex optimization function CVXPY and a solver MOSEK matched with the CVXPY in python language, so that an accurate global optimal solution can be obtained in a short time, and the reliability is high.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides an unmanned cluster cooperative motion path planning method which ensures feasibility and avoids deadlock. Based on the distributed model predictive control, the concept of infinite view is introduced by adding terminal constraint, so that the feasibility of optimizing the problem solution is realized, and the feasibility of planning the motion path is ensured. Secondly, a buffer Voronoi cell (a warning buffered Voronoi cell) with an early warning band is introduced to divide the motion space of different objects, and a penalty function related to the early warning band is introduced into a target function to avoid the deadlock problem. The method can ensure the feasibility of the unmanned cluster in motion path planning and avoid deadlock, and is particularly suitable for motion path planning in a dense cluster environment in a dynamic environment, complex dynamics requirements and high maneuverability. The method can be applied to the situations such as multi-missile cooperative strike, multi-robot cooperative transportation, multi-unmanned aerial vehicle cooperative reconnaissance and the like.
Drawings
Fig. 1 is a block diagram of a platform module according to an embodiment of the present invention, which includes a sensing layer, a planning layer, and an execution layer.
FIG. 2 is a block flow diagram of the method of the present invention.
FIG. 3 is a block diagram of a process for implementing the method of the present invention based on optitrack and crazyflies according to an embodiment of the present invention.
Wherein the content of the first and second substances,xOya reference plane that is artificially defined,as an objectiIn the visual fieldKIs located at a predetermined position in the vertical direction,as an objectjIn the visual fieldKIs located at a predetermined position in the vertical direction,is an objectiAbout an objectjThe avoidance angle of (2).
Detailed Description
The invention will be further described by way of examples, without in any way limiting the scope of the invention, with reference to the accompanying drawings.
The invention provides a distributed unmanned cluster cooperative motion path planning method capable of guaranteeing feasibility and avoiding deadlock, which can solve the problem of cluster system cooperative motion path planning with complex dynamics, avoid the deadlock problem and guarantee the feasibility.
Fig. 1 shows a structure of an experimental platform for implementing the method of the present invention, and the overall flow of the method of the present invention is shown in fig. 2. The following is a specific process of each object in the distributed unmanned cluster collaborative motion path planning to avoid deadlock and guarantee feasibility:
1) initializing a preset path in a distributed unmanned cluster to an objectiFor example, the initialized default path is recorded as ,Is the initial position of the preset path,ibeing an object in a distributed unmanned cluster;is the initial time;kis the view serial number; then entering a main cycle;
2) any one object in distributed unmanned clusterCommunicating respective preset path information with other objects;
3) acquiring state information including position information and speed information through a real-time positioning system (such as Vicon, Optitrack or GPS);
4) obtaining an avoidance coefficient;As an objectiAbout an objectjThe avoidance angle parameter of (1);is a constant;
5) after state information and avoidance coefficients are obtained, a CVXPY is utilized to construct an optimization problem, a MOSEK solver is used for solving in the CVXPY, and the obtained result comprises input informationAnd location information(ii) a Input information in the obtained resultSetting as a control input;
6) and then using the location information in the results of the above optimization problemObtaining a new preset path;
7) Finally inputting the control quantityTo the underlying motion control systems (crazyfiles). The motion process of the objects in the distributed unmanned cluster is realized through control input.
In the field of distributed unmanned cluster cooperative control technology, the most common problem of cluster motion path planning is described as follows:
assuming cluster existenceNAn object. Need thisNThe objects are identicalThe motion in the dimensional space is, at an initial moment,from the initial positionMove to respective target positionsAnd do not collide with each other during movement.
Is numbered for any one of themiThe object of (2), which needs to move from an initial position to a target motion position, has the dynamics of a second order system, approximating as much as possible the actual situation:
the status of the system is indicated,respectively, representing the position, velocity and control input of the object. And the following speed and input constraints need to be satisfied:
wherein the content of the first and second substances,the acceleration is the maximum acceleration, and the acceleration is the maximum acceleration,is the maximum speed;representing the modulus of the vector. Any two objectsiAndjthe following conditions are required to be met for avoiding the obstacle:
Aiming at the problem of cluster motion path planning, a distributed model predictive control method is adopted to realize motion path planning. Distributed model predictive control entails having each object at each elapsed time intervalhLength of field of viewAnd constructing and solving the following optimization problem for re-planning.
Wherein the content of the first and second substances,the time of this rescheduled time is indicated. The solution result is obtainedCan be used as a system intInput of time of dayAnd the motion is input into a lower-layer motion control system, so that the motion control is realized, and the purpose of avoiding obstacles by motion is achieved.
The motion path planning problem with the obstacle avoidance requirement is a typical non-convex optimization problem or a nonlinear programming problem after being converted into an optimization problem, and has the characteristics of low solving speed and no global optimal solution. Therefore, how to convert the convex optimization problem into a convex optimization problem is a core problem in model predictive control, and therefore, the construction of the convex optimization problem needs to be described in detail.
Optimizing variables including control inputsState quantity ofAnd an early warning band variable, wherein the invention defines the early warning band variable.
Is the field of visionkThe control input of (a) is received,to the view field to be plannedkA state quantity of (b), whereinIndicating the horizonkThe sequence of positions of (a) and (b),indicating the horizonkA velocity sequence of (a). Furthermore, the initial position in the planning variables is specified asAnd an initial speed is defined asWhereinAndis an object oftActual position of time and inThe actual speed at the moment is obtained by measurement;is a warning tape variable.
The constraint includes: the method comprises the following steps of dynamic constraint, obstacle avoidance constraint, early warning zone constraint, input constraint, speed constraint and terminal constraint; the invention defines obstacle avoidance constraint, early warning zone constraint and terminal constraint.
Wherein:
obstacle avoidance and restraint:
the following linear constraints can be obtained by dividing the space by the buffer voronoi unit with the early warning zone:
is the last momentWhen in the visual fieldkThe planning result is also called as a preset path; first planning time takingWhereinIs an initial position; extend minimum distance(ii) a Wherein the content of the first and second substances,is the minimum distance between two objects;
the early warning band constraint is expressed as:
wherein the content of the first and second substances,the width of the buffer zone is indicated and is set manually.
Inputting constraints:
speed constraint:
and (4) terminal constraint:
the existence of the terminal constraint ensures the existence of the optimization problem solution.
Target function of constructed distributed unmanned cluster cooperative motion path planning optimization modelCRepresented by formula (3):
Wherein the content of the first and second substances,Cis an objective function;is the visual fieldkThe weight constant of (1);is a target position of the object; is an objectiAbout an objectjOf whereinIs a constant number of times that the number of the first,is an objectiAbout an objectjThe avoidance angle parameter of (1), which is obtained as shown in fig. 4. In the context of figure 4, it is shown,xOya reference plane that is artificially defined,as an objectiIn the visual fieldKIs located at a predetermined position in the vertical direction,as an objectjIn the visual fieldKAt a preset position.Is composed ofAndthe included angle therebetween.
Fig. 3 shows a flow of implementing the method of the present invention using a crazyfiles unmanned aerial vehicle system in the embodiment of the present invention.
When the method is implemented specifically, the optimization model is solved by the convex optimization function CVXPY and the solver MOSEK matched with the CVXPY in the python language, so that an accurate global optimal solution can be obtained in a short time, and the method has high reliability.
According to the method, terminal constraints are added into the optimization problem model, the view length in model prediction control can be equivalent to infinity, and further, the concept of infinite view is introduced. Then, the space is divided by using the buffer Voronoi unit containing the early warning zone based on the preset path, so that the preset path used in the optimization process can be proved to meet all the limits in the convex optimization, the feasible region of the convex optimization is guaranteed to be non-empty, and the recursion feasibility of a plurality of continuous convex optimization problems (namely sequence convex optimization) at different moments is realized. The invention uses and expands the minimum distanceAlternative minimum distancesForming obstacle avoidance constraints (equations (11) - (12)) compared with the original minimum distanceThe obstacle avoidance condition at the sampling point can only be considered, and the extended distance is minimumThe obstacle avoidance situation in the time period between sampling points can be considered, so that the obstacle avoidance feasibility in the whole process is further ensured. Therefore, the method ensures the obstacle avoidance feasibility.
The invention also adds the variable of the early warning zone into the obstacle avoidance constraintSo that a plurality of objects can generate interaction force in the process of mutual blocking, and then the avoidance coefficient is setThe magnitude of the interaction force is controlled to generate a right-handed force, so that the deadlock state caused by balanced force among multiple objects is broken. This dextrorotatory acting force can be optimally solved by the convex optimization problemKKT (Karush-Kuhn-Tucker) conditions of (1) were demonstrated and further elucidatedFor the purpose of breaking the force balance, by settingAnd deadlock states among multiple objects due to balanced acting force are broken.
To sum up, the invention realizes the feasibility of optimizing the problem solution by adding the concept of introducing infinite view by terminal constraint based on the distributed model predictive control, thereby ensuring the feasibility of motion path planning. And dividing motion spaces of different objects by introducing a buffering Voronoi cell (warning buffered Voronoi cell) containing an early warning band, and avoiding a deadlock problem by introducing a penalty function related to the early warning band into a target function. The specific implementation shows that the method can ensure the feasibility of the unmanned cluster in motion path planning and avoid deadlock, and is particularly suitable for motion path planning in a dense cluster environment in a dynamic environment, complex dynamics requirements and high maneuverability. The method can be applied to the situations such as multi-missile cooperative strike, multi-robot cooperative transportation, multi-unmanned aerial vehicle cooperative reconnaissance and the like.
It is noted that the disclosed embodiments are intended to aid in further understanding of the invention, but those skilled in the art will appreciate that: various alternatives and modifications are possible without departing from the invention and scope of the appended claims. Therefore, the invention should not be limited to the embodiments disclosed, but the scope of the invention is defined by the appended claims.
Claims (6)
1. A method for planning cooperative motion paths of a distributed unmanned cluster to avoid deadlock is disclosed, wherein N objects in the distributed unmanned cluster move in the same dimensional space; the object is at the initial momentFrom the initial positionMove to respective target positionsAnd do not collide with each other during the movement; the method is characterized in that aiming at the distributed unmanned cluster movement path planning, an optimization problem and a model function of the distributed unmanned cluster cooperative movement path planning are constructed by defining constraints, targets and variables, an infinite view is defined by adding terminal constraints, then an obstacle avoidance constraint is constructed by using a buffer Weinuo unit containing an early warning zone, and a penalty function related to the early warning zone is introduced into a target function;
the method for planning the distributed unmanned cluster cooperative motion path for avoiding deadlock comprises the following steps:
1) initializing a preset path in the distributed unmanned cluster;
distributing any one object in unmanned clusteriIs recorded as,Is an initial position;is the initial time; horizon of the eyeWhereinKIs the length of the visual field;
2) icommunicating respective preset path information with other objects;
3) acquiring state information including position information and speed information through a real-time positioning system;
4) obtainTaking avoidance coefficient;As an objectiAbout an objectjThe avoidance angle parameter of (1);is a constant;
5) according to the obtained state information and avoidance coefficients, an infinite view field is defined by adding terminal constraints, an obstacle avoidance constraint is constructed by using a buffer Vono unit containing an early warning zone, penalty functions related to the early warning zone are introduced into a target function, and an optimization problem and a model function of the distributed unmanned cluster cooperative motion path planning are constructed; and solving to obtain input informationAnd location informationNamely, the planning of the distributed unmanned cluster cooperative motion path for avoiding deadlock is realized; the specific process is as follows:
51) constructing a model function of a distributed unmanned cluster cooperative motion path planning optimization problem, and realizing motion path planning by adopting a distributed model prediction control method; the method comprises the following steps:
any one object in distributed unmanned clusteriFrom an initial positionMove to the target movement positionEach object at each elapsed time intervalhLength of field of viewConstructing and solving the convex-down optimization problem for re-planning;
the model function of the constructed distributed unmanned cluster cooperative motion optimization path planning optimization problem is represented as follows:
Wherein the content of the first and second substances,is the field of visionkA control input of;view obtained for planningkThe state quantity of (c); ;indicating the horizonkA sequence of positions;indicating the horizonkA sequence of velocities;Cas a distributed unmanned collectionPlanning an objective function of an optimization model by using the group collaborative motion path;
52) defining optimization variables, including control inputsState quantity ofAnd early warning belt variables(ii) a Self-defining early warning zone variables;
initial position specification in planning variablesAnd an initial speed is defined asWhereinAndis an object oftActual position of time and inThe actual speed of the moment is obtained by measurement;
53) defining constraints comprising: the method comprises the following steps of dynamic constraint, obstacle avoidance constraint, early warning zone constraint, input constraint, speed constraint and terminal constraint; self-defining obstacle avoidance constraint, early warning zone constraint and terminal constraint;
the kinetic constraint, expressed as:
wherein:
the obstacle avoidance constraint is a linear constraint obtained by dividing the space through a buffer voronoi unit with an early warning zone, and is expressed as follows:
is the last momentWhen in the visual fieldkThe planning result is also called as a preset path; first planning time takingWhereinIs an initial position; extend a minimum distance of(ii) a Wherein the content of the first and second substances,is the minimum distance between two objects;
the early warning band constraint is expressed as:
the input constraints are:
the speed constraint is:
the terminal constraints are:
target function of constructed distributed unmanned cluster cooperative motion path planning optimization modelCRepresented by formula (2):
Wherein the content of the first and second substances,Cis an objective function;is the visual fieldkThe weight constant of (1);is a target position of the object; is an objectiAbout an objectjOf whereinIs a constant number of times that the number of the first,is an objectiAbout an objectjThe avoidance angle parameter of (1);
54) solving the cluster motion path planning problem, wherein the obtained result comprises input informationAnd location information;
55) The solution result is obtainedAstInput of time of dayThe motion is input into a lower-layer motion control system, so that the motion control is realized, and the purpose of motion obstacle avoidance is achieved;
2. The deadlock avoidance distributed unmanned cluster cooperative motion path planning method of claim 1, wherein the individual objects in the distributed unmanned cluster include unmanned vehicles, unmanned vehicles and unmanned boats.
4. The deadlock avoidance distributed unmanned cluster cooperative motion path planning method of claim 1, wherein any one of the distributed unmanned clusters is numbered asiThe process of moving from the initial position to the target moving position is represented as:
the status of the system is indicated,the position, velocity and control inputs of the object are represented separately, and the following velocity and input constraints need to be satisfied:
wherein the content of the first and second substances,the acceleration is the maximum acceleration, and the acceleration is the maximum acceleration,is the maximum speed;modulo representing a vector, any two objectsiAndjthe following conditions are required to be met for avoiding the obstacle:
5. The deadlock avoidance distributed unmanned cluster cooperative motion path planning method of claim 1, wherein the real time positioning system specifically employs Vicon, Optitrack or GPS.
6. The deadlock avoidance distributed unmanned cluster cooperative motion path planning method of claim 1, wherein the lower layer motion control system specifically employs crazyfiles.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111184687.5A CN113625780B (en) | 2021-10-12 | 2021-10-12 | Distributed unmanned cluster cooperative motion path planning method capable of avoiding deadlock |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111184687.5A CN113625780B (en) | 2021-10-12 | 2021-10-12 | Distributed unmanned cluster cooperative motion path planning method capable of avoiding deadlock |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113625780A true CN113625780A (en) | 2021-11-09 |
CN113625780B CN113625780B (en) | 2022-01-28 |
Family
ID=78391012
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111184687.5A Active CN113625780B (en) | 2021-10-12 | 2021-10-12 | Distributed unmanned cluster cooperative motion path planning method capable of avoiding deadlock |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113625780B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115237157A (en) * | 2022-08-08 | 2022-10-25 | 南京理工大学 | Road network constraint-based air-to-ground unmanned cluster multi-task point path planning method |
CN115328161A (en) * | 2022-09-15 | 2022-11-11 | 安徽工程大学 | Welding robot path planning method based on K-view ant colony algorithm |
CN115774455A (en) * | 2023-02-13 | 2023-03-10 | 北京大学 | Distributed unmanned cluster trajectory planning method for avoiding deadlock in complex obstacle environment |
CN115328161B (en) * | 2022-09-15 | 2024-04-26 | 安徽工程大学 | Welding robot path planning method based on K vision ant colony algorithm |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109582027A (en) * | 2019-01-14 | 2019-04-05 | 哈尔滨工程大学 | A kind of USV cluster collision-avoidance planning method based on Modified particle swarm optimization algorithm |
CN111291984A (en) * | 2020-01-21 | 2020-06-16 | 北京大学 | Multi-unmanned aerial vehicle distributed task selection and trajectory design method and device |
CN111811511A (en) * | 2020-06-23 | 2020-10-23 | 北京理工大学 | Unmanned aerial vehicle cluster real-time track generation method based on dimension reduction decoupling mechanism |
US20200348696A1 (en) * | 2018-07-18 | 2020-11-05 | The Trustees Of The University Of Pennsylvania | Control of multi-drone fleets with temporal logic objectives |
CN112000131A (en) * | 2020-09-09 | 2020-11-27 | 中国人民解放军国防科技大学 | Unmanned aerial vehicle cluster path planning method and system based on artificial potential field method |
CN113253764A (en) * | 2021-07-14 | 2021-08-13 | 北京大学 | Unmanned cluster affine formation control method based on dimensionality reduction observer |
-
2021
- 2021-10-12 CN CN202111184687.5A patent/CN113625780B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20200348696A1 (en) * | 2018-07-18 | 2020-11-05 | The Trustees Of The University Of Pennsylvania | Control of multi-drone fleets with temporal logic objectives |
CN109582027A (en) * | 2019-01-14 | 2019-04-05 | 哈尔滨工程大学 | A kind of USV cluster collision-avoidance planning method based on Modified particle swarm optimization algorithm |
CN111291984A (en) * | 2020-01-21 | 2020-06-16 | 北京大学 | Multi-unmanned aerial vehicle distributed task selection and trajectory design method and device |
CN111811511A (en) * | 2020-06-23 | 2020-10-23 | 北京理工大学 | Unmanned aerial vehicle cluster real-time track generation method based on dimension reduction decoupling mechanism |
CN112000131A (en) * | 2020-09-09 | 2020-11-27 | 中国人民解放军国防科技大学 | Unmanned aerial vehicle cluster path planning method and system based on artificial potential field method |
CN113253764A (en) * | 2021-07-14 | 2021-08-13 | 北京大学 | Unmanned cluster affine formation control method based on dimensionality reduction observer |
Non-Patent Citations (2)
Title |
---|
XIANGKE WANG 等: "Distributed sliding mode control for leader-follower formation flight of fixed-wing unmanned aerial vehicles subject to velocity constraints", 《INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL》 * |
刘淼 等: "不确定通讯下的异构多智能体网络鲁棒编队控制", 《空间控制技术与应用》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115237157A (en) * | 2022-08-08 | 2022-10-25 | 南京理工大学 | Road network constraint-based air-to-ground unmanned cluster multi-task point path planning method |
CN115237157B (en) * | 2022-08-08 | 2024-01-23 | 南京理工大学 | Air-ground unmanned cluster multi-task point path planning method under road network constraint |
CN115328161A (en) * | 2022-09-15 | 2022-11-11 | 安徽工程大学 | Welding robot path planning method based on K-view ant colony algorithm |
CN115328161B (en) * | 2022-09-15 | 2024-04-26 | 安徽工程大学 | Welding robot path planning method based on K vision ant colony algorithm |
CN115774455A (en) * | 2023-02-13 | 2023-03-10 | 北京大学 | Distributed unmanned cluster trajectory planning method for avoiding deadlock in complex obstacle environment |
Also Published As
Publication number | Publication date |
---|---|
CN113625780B (en) | 2022-01-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Yang et al. | Collision free 4D path planning for multiple UAVs based on spatial refined voting mechanism and PSO approach | |
Singla et al. | Memory-based deep reinforcement learning for obstacle avoidance in UAV with limited environment knowledge | |
Chen et al. | Cyber-physical system enabled nearby traffic flow modelling for autonomous vehicles | |
Hong et al. | Rules of the road: Predicting driving behavior with a convolutional model of semantic interactions | |
CN109032168B (en) | DQN-based multi-unmanned aerial vehicle collaborative area monitoring airway planning method | |
Grigorescu et al. | Neurotrajectory: A neuroevolutionary approach to local state trajectory learning for autonomous vehicles | |
Zheng et al. | Evolutionary route planner for unmanned air vehicles | |
CN113625780B (en) | Distributed unmanned cluster cooperative motion path planning method capable of avoiding deadlock | |
Oh et al. | Rendezvous and standoff target tracking guidance using differential geometry | |
Zhang et al. | Pigeon-inspired optimization approach to multiple UAVs formation reconfiguration controller design | |
Hou et al. | Time-coordinated control for unmanned aerial vehicle swarm cooperative attack on ground-moving target | |
US20210325891A1 (en) | Graph construction and execution ml techniques | |
Brittain et al. | Scalable autonomous separation assurance with heterogeneous multi-agent reinforcement learning | |
Li et al. | A deep reinforcement learning based approach for autonomous overtaking | |
Huang et al. | Recoat: A deep learning-based framework for multi-modal motion prediction in autonomous driving application | |
Al-Sharman et al. | Self-learned autonomous driving at unsignalized intersections: A hierarchical reinforced learning approach for feasible decision-making | |
Ali et al. | Feature selection-based decision model for UAV path planning on rough terrains | |
Sanders et al. | Optimal offline path planning of a fixed wing unmanned aerial vehicle (UAV) using an evolutionary algorithm | |
CN112034880A (en) | Novel multi-unmanned aerial vehicle collaborative route planning method | |
Regier et al. | Improving navigation with the social force model by learning a neural network controller in pedestrian crowds | |
CN112258896A (en) | Unmanned aerial vehicle fusion airspace operation method based on flight path | |
CN116339321A (en) | Global information driven distributed multi-robot reinforcement learning formation surrounding method based on 5G communication | |
Chronis et al. | Dynamic Navigation in Unconstrained Environments Using Reinforcement Learning Algorithms | |
Skyrda | Decentralized autonomous unmanned aerial vehicle swarm formation and flight control | |
CN112161626B (en) | High-flyability route planning method based on route tracking mapping network |
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 |