CN109520507B - Unmanned aerial vehicle real-time path planning method based on improved RRT - Google Patents
Unmanned aerial vehicle real-time path planning method based on improved RRT Download PDFInfo
- Publication number
- CN109520507B CN109520507B CN201811477631.7A CN201811477631A CN109520507B CN 109520507 B CN109520507 B CN 109520507B CN 201811477631 A CN201811477631 A CN 201811477631A CN 109520507 B CN109520507 B CN 109520507B
- Authority
- CN
- China
- Prior art keywords
- local
- planning
- target point
- rolling window
- window
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000005096 rolling process Methods 0.000 claims abstract description 99
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 38
- 238000005070 sampling Methods 0.000 claims abstract description 5
- 238000001514 detection method Methods 0.000 claims description 10
- 238000013507 mapping Methods 0.000 claims description 5
- 230000000737 periodic effect Effects 0.000 claims description 4
- 230000004888 barrier function Effects 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 4
- 238000004364 calculation method Methods 0.000 abstract description 2
- 230000007613 environmental effect Effects 0.000 abstract description 2
- 238000005259 measurement Methods 0.000 abstract 1
- 238000012959 renal replacement therapy Methods 0.000 description 44
- 238000005381 potential energy Methods 0.000 description 12
- 238000004088 simulation Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 3
- 238000009499 grossing Methods 0.000 description 2
- 230000002829 reductive effect Effects 0.000 description 2
- 230000002441 reversible effect Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000010391 action planning Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000009123 feedback regulation Effects 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000036961 partial effect Effects 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000013138 pruning Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000035899 viability Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention belongs to the technical field of unmanned aerial vehicle operation, and discloses an unmanned aerial vehicle real-time path planning method based on improved RRT (remote distance measurement), which comprises the steps of constructing a local rolling window; setting a local sub-target point; planning random sampling of a local RRT tree algorithm; the algorithm terminates the principle. The invention improves and fuses the algorithm which is originally only suitable for global planning, constructs a window by utilizing locally known environment information, determines a sub-target point by a certain method, and explores by utilizing the global planning algorithm in the environment. And in the process of the rolling advance of the window, the environmental information in the window is continuously updated, the planning, drawing and feedback are realized, and the target point is finally reached. Compared with global planning, the improved method based on the rolling window does not need to randomly explore the whole space, limits the planning in an infinite number of continuously updated windows, reduces the range of random exploration, reduces the calculation amount and can realize online planning.
Description
Technical Field
The invention belongs to the technical field of unmanned aerial vehicle operation, and particularly relates to an unmanned aerial vehicle real-time path planning method based on improved RRT.
Background
Currently, the current state of the art commonly used in the industry is such that: the fast-expansion random tree (RRT) path planning method has the advantages of strong adaptability, high generation speed, probability completeness and the like, but planning can be performed only by depending on a known global map. Therefore, generally speaking, RRT is a global planning method, and real-time route generation of the unmanned aerial vehicle cannot be realized. However, with the widespread application of the unmanned aerial vehicle, it is not enough to consider only global static obstacles in the indoor task environment, and corresponding obstacle avoidance strategies should be researched if uncertain and dynamic obstacles exist in the indoor environment. Local planning is to collect local information and plan feasible paths in a certain area, and when the environment is dynamic and uncertain, the local planning can follow new scanning data in real time, so that a route is re-planned to react to the environment change. The online planning system well meets the planning instantaneity, and aims to truly realize autonomous navigation and flight of the unmanned aerial vehicle. Common local planning methods include rolling window planning, dynamic planning, artificial potential field planning, and the like.
The rolling Window planning is also called Dynamic Window Approach (DWA), and its basic principle is to make planning Window according to local environment information, to make prediction analysis on action, and to feedback and regulate action planning. The rolling window method is a feasible method for generating an online path, only a path in a local environment is generated for each window planning, the path reaches the next track point, the next planning is carried out according to the detected environment information, and the global environment detection does not need to be completed at one time.
The artificial potential field method is an idea based on potential energy in analog physics, and is used for analogizing a path planning problem into a problem of searching an optimal solution of potential energy in a space. The target point is regarded as a potential energy minimum point (which can be regarded as a zero potential energy point approximately) in the space, the potential energy of the barrier in the space is infinite, the attraction force of the target point and the repulsion force of the barrier on the unmanned aerial vehicle are simulated, so that an artificially defined potential energy field is obtained in the space, the unmanned aerial vehicle tends to move towards the direction of low potential energy to form a track with gradually reduced potential energy, and if the potential energy field is continuous in the space and only has a minimum value, the potential energy minimum point, namely the target point, can be reached. However, when the potential energy field in the space has multiple minimum values, the local minimum point of the non-target point may cause that the unmanned aerial vehicle cannot find a point with lower potential energy after reaching the local minimum, and thus the unmanned aerial vehicle cannot reach the target point forever, resulting in failure of path planning.
In summary, the problems of the prior art are as follows:
(1) the algorithm in the prior art is only suitable for global planning, so that the limitation is caused, the operation is inconvenient, and the use efficiency is reduced.
(2) In the prior art, random exploration needs to be carried out on the whole space, the range of the random exploration is large, the calculated amount is increased, and online planning cannot be effectively realized.
The difficulty and significance for solving the technical problems are as follows:
the existing mature technology is path planning under the condition that a global map is known. In fact, in many cases, drones perform tasks in unknown environments. Therefore, a global map cannot be obtained in advance, and thus global path planning cannot be performed. This makes the global planning method limited in its application.
Thus, an improved RRT algorithm based on a rolling window is proposed. Since only the path near the drone is planned, the global map need not be known. This makes unmanned aerial vehicle's viability at unknown environment promote greatly.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides an unmanned aerial vehicle real-time path planning method based on improved RRT.
The invention is realized in this way, an unmanned aerial vehicle real-time path planning method based on improved RRT, comprising the following steps:
the method comprises the following steps: constructing a local rolling window, and acquiring local environment information through a sensor carried by an unmanned aerial vehicle to construct the local rolling window;
step two: setting local child target points, wherein the child target points are positioned on the boundary of a rolling window, namely selecting feasible points which meet certain global mapping conditions on the boundary as local child target points;
step three: the local RRT tree algorithm randomly samples and plans, the planning in the window is finished when the sub-target is not reached, the condition that whether the planning of the feasible path reaches the window boundary is finished is taken as the condition that whether the planning of the feasible path reaches the window boundary, when the RRT extends to the window boundary, the planning to the child target point is stopped, and a new rolling plan is established by taking a new node as the center;
step four: the algorithm terminates the principle.
Further, in the first step, a local rolling window is constructed, specifically:
local environment information is obtained through a sensor carried by an unmanned aerial vehicle, a planned window is driven in a periodic mode, and a window area planned each time is Ein (q)R(tt)＝{p|p∈C，d(q，qR(t))≤R}，qR(t) denotes the center of the rolling window, and R is the radius of the rolling window, typically the detection radius of the sensor.
Further, in the second step, the setting of the local child target point specifically includes:
will be located on the border of the rolling window and at the current point and the target point qgoalPoints on the connecting line as child target points qtemp goal；
Will (x)c，yc) As the center of the current rolling window, the dotted circle with radius R is the rolling window range, (x)g，yg) Is a global target point, (x)t0，yto) Is a windowThe intersection point of the line connecting the center and the target point and the dotted line circle is defined as a local sub-target point qtemp goal。
Because in RRT branch expansion, the growth direction of the branch depends on qrandIf the target point is not reached, only the generated q is selected randomlyrandWithin the feasible region and close to the child target point qtemp goalThe branches will grow towards the feasible and target point. If the child target point is on the obstacle, the standard RRT plan cannot be reached, and at this time, a feasible child target point is randomly generated on the circle, so as to quickly generate a feasible path reaching the window boundary in the area.
theta＝rand()·2π (2)
At this time, the random node selection of the RRT within the window should consider the relationship with the child target point and how to stop the search.
Further, in step three, the random sampling planning of the local RRT tree algorithm specifically includes:
defining the angle between a random point on the circle and the horizontal line as theta ═ beta + alpha, beta is qrandThe angle of the ion target point is deflected, and the selection of theta is constructed into a normal function which is symmetrically distributed about alpha; namely, it is
β＝randn·π
Thus, randomly distributed angles beta within (-180 DEG, 180 DEG) are constructed, and the ion target point selection q is biased at the angle betarand；
And when the RRT extends to the window boundary, stopping planning to the child target point and establishing a new rolling plan by taking the new node as the center.
Further, in step four, the algorithm termination principle is specifically
When the rolling window rolls to the global target point one time, when the global target point qgoalWithin the range of the rolling window, i.e. d (q)goal，qR(t)) < R, then q is directly addedgoalPlanning as target point to reach qgoalAnd the algorithm ends.
Another object of the present invention is to provide a rolling RRT algorithm with the following general flow:
(1) initializing variables of the RRT tree in a rolling window;
(2) judging qgoalWhether the position is in the rolling window or not, if so, turning to the step 3, and not turning to the step 4;
(3) q is to begoalAs target points, nodes q are randomly generated on the boundaryrand；
(4) Obtaining local sub-target point q on rolling window boundarytempgoalGenerating random nodes q on the boundaryrand；
(5) Selecting the nearest tree node q in the current tree speciesnearAnd generating a node q according to a certain step lengthnew；
(6) At qnear、qnewCarrying out collision detection on the connecting line;
(7) judging qnewWhether it is in a flyable region and qnear、qnewDistributing obstacles on the connecting line, if the obstacles are not distributed, returning to the step 2, and if the obstacles are distributed, turning to the step 6;
(8) q is to benewAdding into the tree list, judging to reach qgoalIf not, go to step 9; if yes, ending the algorithm;
(9) judging whether the boundary of the rolling window is reached, if not, returning to the step 2, and if so, turning to the step 10;
(10) from qnewSearch in reverse directionTo a local path and with qnewEstablishing a new rolling window for the center, and returning to the step 1;
in summary, the advantages and positive effects of the invention are:
the invention can improve and fuse the algorithm which is originally only suitable for global planning, construct a window by utilizing locally known environment information, determine a sub-target point by a certain method, and then explore by utilizing the global planning algorithm in the environment. And in the process of the rolling advance of the window, the environmental information in the window is continuously updated, the planning, drawing and feedback are realized, and the target point is finally reached. Compared with global planning, the improved method based on the rolling window does not need to randomly explore the whole space, limits the planning in an infinite number of continuously updated windows, reduces the range of random exploration, reduces the calculation amount and can realize online planning.
Drawings
Fig. 1 is a flowchart of a method for planning a real-time path of an unmanned aerial vehicle based on an improved RRT according to an embodiment of the present invention.
Fig. 2 is a selection diagram of the rolling RRT sub-target provided by the embodiment of the present invention.
FIG. 3 is a simulation result of a rolling RRT in a circular obstacle provided by an embodiment of the present invention;
in the figure: a is q is 50; b is q ═ 30; c is q ═ 20; and d is q-10.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The following detailed description of the principles of the invention is provided in connection with the accompanying drawings.
As shown in fig. 1, the method for planning the real-time path of the unmanned aerial vehicle based on the improved RRT provided by the embodiment of the present invention includes the following steps:
s101: constructing a local rolling window, and acquiring local environment information through a sensor carried by an unmanned aerial vehicle to construct the local rolling window;
s102: setting local child target points, wherein the child target points are positioned on the boundary of a rolling window, namely selecting feasible points which meet certain global mapping conditions on the boundary as local child target points;
s103: the local RRT tree algorithm randomly samples and plans, the planning in the window is finished when the sub-target is not reached, the condition that whether the planning of the feasible path reaches the window boundary is finished is taken as the condition that whether the planning of the feasible path reaches the window boundary, when the RRT extends to the window boundary, the planning to the child target point is stopped, and a new rolling plan is established by taking a new node as the center;
s104: the algorithm terminates the principle.
In step S101, the constructing a local rolling window provided in the embodiment of the present invention specifically includes:
local environment information is obtained through a sensor carried by an unmanned aerial vehicle, a planned window is driven in a periodic mode, and a window area planned each time is Win (q)R(t))＝{p|p∈C，d(q，qR(t))≤R}，qR(t) denotes the center of the rolling window, and R is the radius of the rolling window, typically the detection radius of the sensor.
In step S102, the setting of the local child target point provided by the embodiment of the present invention specifically includes:
will be located on the border of the rolling window and at the current point and the target point qgoalPoints on the connecting line as child target points qtemp goal；
Will (x)c，yc) As the center of the current rolling window, the dotted circle with radius R is the rolling window range, (x)g，yg) Is a global target point, (x)t0，yto) Is the intersection point of the connecting line of the window center and the target point and the dotted line circle, and is defined as a local sub-target point qtrmp goal。
Because in RRT branch expansion, the growth direction of the branch depends on qrandIs followed byMachine selection, when not reaching the target point, as long as q is generatedrandWithin the feasible region and close to the child target point qtemp goalThe branches will grow towards the feasible and target point. If the child target point is on the obstacle, the standard RRT plan cannot be reached, and at this time, a feasible child target point is randomly generated on the circle, so as to quickly generate a feasible path reaching the window boundary in the area.
theta＝rand()·2π (2)
At this time, the random node selection of the RRT within the window should consider the relationship with the child target point and how to stop the search.
In step S103, the random sampling planning of the local RRT tree algorithm provided in the embodiment of the present invention specifically includes:
defining the angle between a random point on the circle and the horizontal line as theta ═ beta + alpha, beta is qrandThe angle of the ion target point is deflected, and the selection of theta is constructed into a normal function which is symmetrically distributed about alpha; namely, it is
β＝randn·π
Thus, randomly distributed angles beta within (-180 DEG, 180 DEG) are constructed, and the ion target point selection q is biased at the angle betarand；
And when the RRT extends to the window boundary, stopping planning to the child target point and establishing a new rolling plan by taking the new node as the center.
In step S105, the algorithm termination principle provided in the embodiment of the present invention is specifically
Scrolling to the global target point one time at a scrolling windowWhen the global target point qgoalWithin the range of the rolling window, i.e. d (q)goal，qR(t)) < R, then q is directly addedgoalPlanning as target point to reach qgoalAnd the algorithm ends.
The rolling RRT algorithm provided by the embodiment of the invention has the following general flow:
(1) initializing variables of the RRT tree in a rolling window;
(2) judging qgoalWhether the position is in the rolling window or not, if so, turning to the step 3, and not turning to the step 4;
(3) q is to begoalAs target points, nodes q are randomly generated on the boundaryrand；
(4) Obtaining local sub-target point q on rolling window boundarytemp goalGenerating random nodes q on the boundaryrand；
(5) Selecting the nearest tree node q in the current tree speciesnearAnd generating a node q according to a certain step lengthnew；
(6) At qnear、qnewCarrying out collision detection on the connecting line;
(7) judging qnewWhether it is in a flyable region and qnear、qnewDistributing obstacles on the connecting line, if the obstacles are not distributed, returning to the step 2, and if the obstacles are distributed, turning to the step 6;
(8) q is to benewAdding into the tree list, judging to reach qgoalIf not, go to step 9; if yes, ending the algorithm;
(9) judging whether the boundary of the rolling window is reached, if not, returning to the step 2, and if so, turning to the step 10;
(10) from qnewReverse search results in a local path and qnewEstablishing a new rolling window for the center, and returning to the step 1;
the working principle part provided by the embodiment of the invention is as follows:
and taking a rolling window planning algorithm as a basic frame, taking an RRT algorithm as a path planning strategy in each window, forming an RRT expansion tree of a local area by pruning and smoothing of a path, connecting a current point and a local sub target point, realizing motion control and feedback regulation, updating window information after reaching a new path point, and planning the path by using the RRT method again.
The application principle of the present invention will be further described in detail with reference to the following specific embodiments;
example 1;
1. building a partially scrolling window
The local environment information is obtained by a sensor carried by an unmanned aerial vehicle, a planned window is driven in a periodic mode, the window area of each planning is defined as Win (q _ R (t)) { p | p ∈ C, d (q, q _ R (t)) ≦ R }, q _ R (t) represents the center of a rolling window, and R is the radius of the rolling window and is generally the detection radius of the sensor. The above window is also referred to as the field of view of the drone.
The unmanned aerial vehicle can preferably keep a certain speed to continuously fly in the real-time route planning and autonomous navigation processes, and does not need to stop waiting for the next time after the window planning generates a path. Therefore, for continuous flight, the next window planning must be performed in advance before the local sub-target point is reached, and the currently planned target point is taken as the center of the next planning window. After the next rolling planning route is obtained, the unmanned aerial vehicle finishes the stepping and path smoothing processing of the planning remaining points and the next path. Here, it is necessary to know the partial environment information of the next window in advance, so that R is defined to be slightly smaller than the detection radius of the sensor, so as to construct more than one rolling window under the local environment obtained by the detector.
2. Selection of local child target points
The selection of local child target points in the rolling RRT should correspond to a map that is a global target point and can avoid obstacle areas, a feasible waypoint. In general, in order to make full use of the local environment information in the rolling window, sub-target points are defined to be located on the boundary of the rolling window, that is, points which are feasible and meet some global mapping condition are selected as local sub-target points on the boundary. Here, the
Will be located at the rolling windowOn the boundary of the mouth and at the current point and the target point qgoalPoints on the connecting line as child target points qtemp goal。
As shown in fig. 2, will (x)c，yc) As the center of the current rolling window, the dotted circle with radius R is the rolling window range, (x)g，yg) Is a global target point, (x)t0，yto) Is the intersection point of the connecting line of the window center and the target point and the dotted line circle, and is defined as a local sub-target point qtemp goal。
Because in RRT branch expansion, the growth direction of the branch depends on qrandIf the target point is not reached, only the generated q is selected randomlyrandWithin the feasible region and close to the child target point qtemp goalThe branches will grow towards the feasible and target point. If the child target point is on the obstacle, the standard RRT plan cannot be reached, and at this time, a feasible child target point is randomly generated on the circle, so as to quickly generate a feasible path reaching the window boundary in the area.
theta＝rand()·2π (2)
At this time, the random node selection of the RRT within the window should consider the relationship with the child target point and how to stop the search.
3. Random sampling strategy for local RRT algorithm
From the expansion process of each tree of RRT, a random node qrandIs chosen to have randomness throughout the space, where q israndIs restricted to the border of the rolling window, is developed, qrandWill be selected with probability p as the child target point (x) on the circumference in FIG. 2t0，yto) The probability of (1-p) is chosen as any point on the circumference, such as a random point in the figure(xt，yt). So q israndWill tend to select points on the circumference that are closer to the sub-goal points and the probability will be greater the closer they are, if q israndAnd reselecting on the obstacle.
As shown in fig. 2, an angle θ between a random point on the circle and the horizontal line is defined as β + α, and β is qrandThe angle of the ion target point is biased such that the selection of θ is constructed as a normal function of the symmetric distribution about α. Namely, it is
β＝randn·π
Thus, randomly distributed angles beta within (-180 DEG, 180 DEG) are constructed, and the ion target point selection q is biased at the angle betarand。
And when the RRT extends to the window boundary, stopping planning to the child target point and establishing a new rolling plan by taking the new node as the center.
4. Principle of algorithm termination
When the rolling window rolls to the global target point one time, when the global target point qgoalWithin the range of the rolling window, i.e. d (q)goal，qR(t)) < R, then q is directly addedgoalPlanning as target point to reach qgoalAnd the algorithm ends.
Example 2;
the rolling RRT algorithm provided by the embodiment of the invention has the following general flow:
(1) initializing variables of the RRT tree in a rolling window;
(2) judging whether the window is positioned in the rolling window, if so, turning to the step 3, and if not, turning to the step 4;
(3) will be the target point, randomly generate nodes on the boundary
(4) Obtaining a local child target point on the boundary of the rolling window, and generating a random node on the boundary;
(5) selecting a nearest tree node in the current tree species, and generating a node according to a certain step length;
(6) performing collision detection on the connecting line;
(7) judging whether the aircraft is in a flyable area and the connecting line is distributed without obstacles, if not, returning to the step 2, and if so, turning to the step 6;
(8) adding the tree list, judging that the tree list is reached, and if not, turning to the step 9; if yes, ending the algorithm;
(9) judging whether the boundary of the rolling window is reached, if not, returning to the step 2, and if so, turning to the step 10;
(10) obtaining a local path from the reverse search, establishing a new rolling window for the center, and returning to the step 1;
the application principle of the present invention is further explained in detail with reference to the specific simulation test;
simulation test 1;
the simulation environment of 500X500 is set, 4 circular obstacles are distributed, the radius of a rolling window is 100, the search step is 50, the coordinates of a starting point are (50, 50), the coordinates of an end point are (400 ), the conditions of smaller steps q of 50, 30, 20 and 10 are simulated, other parameters K of 10000, p of 0.3 and the radius R of the rolling window of 100 are considered, and the obstacle avoidance condition of the obstacles is mainly considered. And compares its performance with the standard RRT at the same step size and probability. The computer hardware equipment is an Intel (R) single-core processor 2.1GHz, and the memory is 4 GB. And programming a simulation program in an MATLAB environment.
As shown in fig. 3, the present invention provides simulation results of rolling RRTs in a circular obstacle;
wherein, in fig. 3-a, q is 50; q-30 in fig. 3-b; in fig. 3-c, q is 20; q-10 in fig. 3-d;
as shown in FIG. 3, simulation verification shows that the algorithm can finish one-time window planning in real time by taking 0.02s on average, the first-time window planning completion time is 0.01-0.08s, and the planning time is short and meets the requirement of real-time online path planning of the unmanned aerial vehicle under the condition that the time taken by the unmanned aerial vehicle to move is not considered.
The selection strategy of the random node is to tend to the growth of sub-goal points, which are the mappings of the global goal, so the generation of the local tree is to some extent to the growth of the goal. When the child target point is always directly mapped to the global target point, the planning route can be very flat, the planning efficiency is high, the step length is small, and although the number of nodes is large in the local planning, different characteristics are shown when the node is used for the global planning. The small-step generation route is shorter than the large-step, the number of generated windows is small, the route is smooth, the unmanned aerial vehicle kinematics constraint can be realized more favorably, and the advantages of the small-step generation route in the local planning are fully exerted
The algorithm has certain engineering application value for applying the RRT algorithm to online real-time planning of the unmanned aerial vehicle.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (5)
1. The unmanned aerial vehicle real-time path planning method based on the improved RRT is characterized by comprising the following steps of:
the method comprises the following steps: constructing a local rolling window, and acquiring local environment information through a sensor carried by an unmanned aerial vehicle to construct the local rolling window;
step two: setting a local sub-target point, wherein the local sub-target point is positioned on the boundary of a local rolling window, namely selecting a point which is feasible and meets the global mapping condition on the boundary as a local sub-target point;
step three: the local RRT algorithm randomly samples and plans, the planning in the local rolling window is not finished by reaching the local sub-target point, whether the planning of the feasible path reaches the boundary of the local rolling window is taken as an ending condition, when the RRT extends to the boundary of the local rolling window, the planning to the local sub-target point is stopped, and a new rolling plan is established by taking a new node as the center;
step four: an algorithm termination principle;
in the fourth step, the algorithm termination principle specifically includes:
when the local rolling window rolls to the global target point one time, when the global target point qgoalWithin a partially scrolling window, i.e. d (q)goal，qR(t)) < R, then q is directly addedgoalPlanning as the final target point to reach qgoalAnd ending the algorithm;
in the first step, a local rolling window is constructed, specifically:
local environment information is obtained through a sensor carried by an unmanned aerial vehicle, a planned local rolling window is driven in a periodic mode, and the area of the planned local rolling window every time is Win (q)R(t))＝{p|p∈C，d(q，qR(t))≤R}，qR(t) denotes the center of the local rolling window, and R is the radius of the local rolling window, which is the detection radius of the sensor.
2. The method for planning the real-time path of the unmanned aerial vehicle based on the improved RRT of claim 1, wherein in the second step, the setting of the local sub-target point specifically includes:
will be located on the boundary of the local rolling window and at the current point and the global target point qgoalPoints on the connecting line are used as local sub-target points qtempgoal；
Current point (x)c，yc) As the center of the current local rolling window, a dotted circle with a radius R is the local rolling window range, R is the radius of the local rolling window, q isgoal(xg，yg) Is a global target point, (x)T0，yto) Is the intersection point of the connecting line of the center of the local rolling window and the global target point and the dotted line circle, and is defined as a local sub-target point qtempgoal：
If the local subtarget point is on the barrier and cannot be reached by standard RRT planning, randomly generating a feasible local subtarget point on the circle, and aiming at quickly generating a feasible path reaching the boundary of the local rolling window in the area;
theta＝rand()·2π；
the random node selection of the RRT within the local rolling window takes into account the relationship to the local child target point and how to stop the search.
3. The method for unmanned aerial vehicle real-time path planning based on the improved RRT of claim 1, wherein in step three, the random sampling planning of the local RRT algorithm specifically comprises: one point q on the circle at randomrandAt an angle theta + alpha to the horizontal, alpha, beta being qrandThe angle of departure from the local sub-target point, the choice of θ is constructed as a normal function distributed symmetrically about α:
β＝randn·π；
constructing randomly distributed angles beta within (-180 DEG, 180 DEG), and deviating local sub-target point selection q by the angle betarand(ii) a And when the RRT extends to the boundary of the local rolling window, stopping planning to the local sub-target point, and establishing a new rolling plan by taking the new node as the center.
4. The method of claim 1, wherein the method of real-time path planning for the unmanned aerial vehicle based on the improved RRT comprises the steps of:
(1) initializing variables of the RRT tree in a local rolling window;
(2) judging qgoalWhether the current position is located in the local rolling window or not, if so, turning to the step (3), and not turning to the step (4);
(3) the global target point qgoalAs the final target point, a node q is randomly generated on the boundaryrand；
(4) Obtaining local sub-target point q on local rolling window boundarytempgoalGenerating random nodes q on the boundaryrand；
(5) Selecting a nearest tree node q in a current treenearAnd generating a node q according to a certain step lengthnew；
(6) At qnear、qnewCarrying out collision detection on the connecting line;
(7) judging qnewWhether it is in a flyable region and qnear、qnewDistributing no obstacles on the connecting line, returning to the step (2) if the connecting line is not satisfied, and turning to the step (6) if the connecting line is satisfied;
(8) q is to benewAdding into the tree list, judging to reach qgoalIf not, go to step (9); if yes, ending the algorithm;
(9) judging whether the boundary of the local rolling window is reached, if not, returning to the step (2), and if so, returning to the step (10);
(10) from qnewReverse search results in a local path and qnewAnd (4) establishing a new local rolling window for the center, and returning to the step (1).
5. An unmanned aerial vehicle, which applies the method for planning the real-time path of the unmanned aerial vehicle based on the improved RRT as claimed in any one of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811477631.7A CN109520507B (en) | 2018-12-05 | 2018-12-05 | Unmanned aerial vehicle real-time path planning method based on improved RRT |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811477631.7A CN109520507B (en) | 2018-12-05 | 2018-12-05 | Unmanned aerial vehicle real-time path planning method based on improved RRT |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109520507A CN109520507A (en) | 2019-03-26 |
| CN109520507B true CN109520507B (en) | 2021-05-07 |
Family
ID=65794385
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811477631.7A Expired - Fee Related CN109520507B (en) | 2018-12-05 | 2018-12-05 | Unmanned aerial vehicle real-time path planning method based on improved RRT |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109520507B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110262473B (en) * | 2019-04-29 | 2021-09-14 | 上海交通大学 | Unmanned ship automatic collision avoidance method based on improved Bi-RRT algorithm |
| CN110162104A (en) * | 2019-06-14 | 2019-08-23 | 哈尔滨工业大学 | The RRT unmanned plane paths planning method of imitative Driver Vision perception |
| CN110471441A (en) * | 2019-08-12 | 2019-11-19 | 南京理工大学 | A kind of quadrotor drone barrier-avoiding method based on laser acquisition |
| CN110531782A (en) * | 2019-08-23 | 2019-12-03 | 西南交通大学 | Unmanned aerial vehicle flight path paths planning method for community distribution |
| CN110531770B (en) * | 2019-08-30 | 2023-06-02 | 的卢技术有限公司 | RRT path planning method and system based on improvement |
| CN110906947B (en) * | 2019-12-04 | 2022-05-17 | 山东省科学院自动化研究所 | Slime RRT navigation method and system based on two-dimensional static path planning |
| CN110900611A (en) * | 2019-12-13 | 2020-03-24 | 合肥工业大学 | Novel mechanical arm target positioning and path planning method |
| CN111238518B (en) * | 2020-01-13 | 2023-09-29 | 山东交通学院 | Intelligent agricultural vehicle path planning method based on improved RRT algorithm |
| CN113625701A (en) * | 2020-05-09 | 2021-11-09 | 苏州宝时得电动工具有限公司 | Mowing robot path planning method and mowing robot |
| CN111707264A (en) * | 2020-05-30 | 2020-09-25 | 同济大学 | Improved and extended RRT path planning method, system and device |
| CN112462785B (en) * | 2020-12-04 | 2022-06-03 | 厦门大学 | Mobile robot path planning method and device and storage medium |
| CN113625764A (en) * | 2021-08-30 | 2021-11-09 | 云南民族大学 | Unmanned aerial vehicle autonomous exploration method based on boundary driving, navigation system and unmanned aerial vehicle |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104765371A (en) * | 2015-04-22 | 2015-07-08 | 福州大学 | Route planning method based on rolling window deep searching and fuzzy control fusion |
| 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 |
| CN107608372A (en) * | 2017-08-14 | 2018-01-19 | 广西师范大学 | It is a kind of that path planning method is cooperateed with improving the multiple no-manned plane that PH curves are combined based on improvement RRT algorithms |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105807760B (en) * | 2014-12-30 | 2020-01-03 | Tcl集团股份有限公司 | Intelligent robot and method and device for self-establishing path thereof |
| JP6606442B2 (en) * | 2016-02-24 | 2019-11-13 | 本田技研工業株式会社 | Mobile route plan generation device |
| CN106406320B (en) * | 2016-11-29 | 2019-08-20 | 重庆重智机器人研究院有限公司 | The robot of robot path planning method and programme path |
| CN107883962A (en) * | 2017-11-08 | 2018-04-06 | 南京航空航天大学 | A kind of dynamic Route planner of multi-rotor unmanned aerial vehicle under three-dimensional environment |
| CN108375987A (en) * | 2018-05-28 | 2018-08-07 | 济南大学 | A kind of real-time trajectory planning of low-latitude flying unmanned plane |
-
2018
- 2018-12-05 CN CN201811477631.7A patent/CN109520507B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104765371A (en) * | 2015-04-22 | 2015-07-08 | 福州大学 | Route planning method based on rolling window deep searching and fuzzy control fusion |
| 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 |
| CN107608372A (en) * | 2017-08-14 | 2018-01-19 | 广西师范大学 | It is a kind of that path planning method is cooperateed with improving the multiple no-manned plane that PH curves are combined based on improvement RRT algorithms |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109520507A (en) | 2019-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109520507B (en) | Unmanned aerial vehicle real-time path planning method based on improved RRT | |
| Wang et al. | Deep-reinforcement-learning-based autonomous UAV navigation with sparse rewards | |
| CN107145161B (en) | Flight path planning method and device for unmanned aerial vehicle to access multiple target points | |
| Ganeshmurthy et al. | Path planning algorithm for autonomous mobile robot in dynamic environment | |
| EP1733287B1 (en) | System and method for adaptive path planning | |
| Yu et al. | Probabilistic path planning for cooperative target tracking using aerial and ground vehicles | |
| Xiang et al. | Continuous control with deep reinforcement learning for mobile robot navigation | |
| CN110926477A (en) | Unmanned aerial vehicle route planning and obstacle avoidance method | |
| Gao et al. | Multi-mobile robot autonomous navigation system for intelligent logistics | |
| Liu et al. | Towards search-based motion planning for micro aerial vehicles | |
| Das et al. | An improved Q-learning algorithm for path-planning of a mobile robot | |
| Ji et al. | Mapless-planner: A robust and fast planning framework for aggressive autonomous flight without map fusion | |
| CN116804879A (en) | Robot path planning framework method for improving dung beetle algorithm and fusing DWA algorithm | |
| Habib et al. | Mobile robot motion planning by point to point based on modified ant colony optimization and Voronoi diagram | |
| Tanzmeister et al. | Path planning on grid maps with unknown goal poses | |
| Mokhtari et al. | Safe deep q-network for autonomous vehicles at unsignalized intersection | |
| Nguyen et al. | Nonmyopic planning for long-term information gathering with an aerial glider | |
| Belavadi et al. | Frontier exploration technique for 3d autonomous slam using k-means based divisive clustering | |
| Shi et al. | Research on Path Planning Strategy of Rescue Robot Based on Reinforcement Learning | |
| Zhang et al. | Integrated Design of Cooperative Area Coverage and Target Tracking with Multi-UAV System | |
| CN115826586B (en) | Path planning method and system integrating global algorithm and local algorithm | |
| Zhang et al. | A state-decomposition DDPG algorithm for UAV autonomous navigation in 3D complex environments | |
| Cook et al. | Intelligent cooperative control for urban tracking with unmanned air vehicles | |
| Zhang et al. | A multi-goal global dynamic path planning method for indoor mobile robot | |
| Chen et al. | Multi-target active object tracking with monte carlo tree search and target motion modeling |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210507 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |