CN109947136B - Collaborative active sensing method for unmanned aerial vehicle group rapid target search - Google Patents
Collaborative active sensing method for unmanned aerial vehicle group rapid target search Download PDFInfo
- Publication number
- CN109947136B CN109947136B CN201910151483.8A CN201910151483A CN109947136B CN 109947136 B CN109947136 B CN 109947136B CN 201910151483 A CN201910151483 A CN 201910151483A CN 109947136 B CN109947136 B CN 109947136B
- Authority
- CN
- China
- Prior art keywords
- unmanned aerial
- aerial vehicle
- vehicle group
- closed
- loop
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000013507 mapping Methods 0.000 claims abstract description 25
- 230000001133 acceleration Effects 0.000 claims abstract description 22
- 238000012544 monitoring process Methods 0.000 claims abstract description 15
- 238000005457 optimization Methods 0.000 claims abstract description 9
- 238000013528 artificial neural network Methods 0.000 claims description 10
- 238000005070 sampling Methods 0.000 claims description 10
- 230000008447 perception Effects 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 6
- 230000000007 visual effect Effects 0.000 claims description 5
- 210000002569 neuron Anatomy 0.000 claims description 2
- 230000008569 process Effects 0.000 abstract description 4
- 238000004364 calculation method Methods 0.000 abstract description 3
- 238000011160 research Methods 0.000 description 4
- 239000002245 particle Substances 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000001939 inductive effect Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000004888 barrier function Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
Images
Landscapes
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention discloses a collaborative active sensing method for rapid target search of unmanned aerial vehicle group, belonging to the technical field of multi-unmanned aerial vehicle regional monitoring. The method comprises the following steps: setting a target position according to a target monitoring area of an unmanned aerial vehicle group; decoupling the multi-unmanned aerial vehicle three-dimensional trajectory planning into planning problems in the horizontal and vertical directions, determining the planning sequence of the multi-unmanned aerial vehicle collaborative search in the horizontal direction based on a tree dynamic growth self-organizing mapping algorithm, adjusting the flight heights of an unmanned aerial vehicle group, and accessing all target position neighborhoods; the smooth closed-loop three-dimensional trajectory planning of the unmanned aerial vehicle group is realized by utilizing the piecewise smooth curve; by combining unmanned aerial vehicle path speed and acceleration constraints, the problem of shortest time optimization of active sensing of a multi-unmanned aerial vehicle collaborative area is solved by utilizing speed profile calculation. The processes are carried out simultaneously, and the planned route enables the multi-unmanned aerial vehicle to visit the position of the target object as much as possible in the shortest time, so that the cooperative and rapid target search is realized, and the active sensing task is completed.
Description
Technical Field
The invention belongs to the technical field of multi-unmanned aerial vehicle regional monitoring, and particularly relates to a collaborative active perception method for unmanned aerial vehicle group rapid target search.
Background
Unmanned aerial vehicle has power drive and can carry the mission equipment, possesses beyond visual range flight and unmanned vehicles of autonomous flight ability. By virtue of the characteristics of low manufacturing cost, small volume, strong maneuverability and the like, the device plays an increasingly important role in the fields of military use, civil use and the like. The task executed by a plurality of unmanned aerial vehicles together and related research become the trend of application gradually, and the regional monitoring is an important task of the existing unmanned aerial vehicle group control system. The cooperative target search and active sensing in the large-scale complex environment are one of the main problems to be solved in the field of unmanned aerial vehicle group control systems.
At present, most of the route planning algorithm researches for unmanned aerial vehicle target search aim at avoiding barriers from reaching a flight target point, suboptimal or optimal route tracks are calculated and selected through related route planning algorithms, most of the researches stay in the field of two-dimensional track planning, and the defects of large calculated amount, poor robustness and the like exist. The research on documents in the prior art shows that the improved A-algorithm and the particle swarm algorithm are mostly used for unmanned aerial vehicle track planning, but the A-algorithm is slow in searching speed and large in calculation amount, the unmanned aerial vehicle optimal track is difficult to find under the multi-constraint condition, the particle swarm algorithm planning space division is rough, the flight constraint condition is difficult to meet, the space search outside the optional path set cannot be achieved, the particle swarm algorithm is not high in accuracy, and the optimal track is difficult to find under the complex environment and the multi-constraint condition.
The tree-shaped dynamic growth self-organizing mapping algorithm is a competitive neural network algorithm of unsupervised learning, is an improvement on a basic self-organizing mapping algorithm, adopts a flexible tree-shaped structure, and has higher algorithm execution efficiency. At present, the algorithm is applied to the field of data clustering, and a small number of papers apply the self-organizing mapping algorithm to the online trajectory planning of a single unmanned aerial vehicle and mostly concentrate on a two-dimensional space. In the field of unmanned plane group target search, no patent exists yet for realizing unmanned plane group three-dimensional track planning tasks by using the algorithm.
Disclosure of Invention
Aiming at large-scale and incompletely known complex environments, the invention solves the problems of rapid track planning and active perception of unmanned aerial vehicle group collaborative area target search under multiple constraints by utilizing a tree-shaped dynamic growth self-organizing mapping algorithm and Bezier smooth planning. The invention aims at the designated area continuously monitored by a plurality of unmanned aerial vehicles, carries out cooperative target search on targets in a complex environment, quickly finishes target observation in the search area, plans a most effective three-dimensional space track for each unmanned aerial vehicle, searches and observes the targets at the minimum cost and the highest speed, collects regional information in real time to the maximum extent, realizes the active perception of unmanned aerial vehicle groups and achieves the purpose of effectively monitoring large-scale areas.
In order to achieve the purpose, the invention is realized by adopting the following technical scheme.
The invention provides a collaborative active perception method for fast target search of unmanned aerial vehicle group, which is characterized by comprising the following steps:
1) setting n target positions of the unmanned aerial vehicle group according to a target monitoring area of the unmanned aerial vehicle group; the unmanned aerial vehicle group comprises N unmanned aerial vehicles which move independently, each unmanned aerial vehicle is provided with a visual sensor, the monitoring radius of each unmanned aerial vehicle is set according to the sensing distance of each visual sensor, and the monitoring radius is used as the neighborhood of each target position;
2) determining an access node of an unmanned aerial vehicle group track based on a tree dynamic growth self-organizing mapping algorithm, and specifically comprising the following steps:
2.1) decoupling the three-dimensional trajectory planning of the unmanned aerial vehicle group into planning problems in the horizontal and vertical directions;
2.2) generating and updating access nodes of the closed-loop tracks of the unmanned aerial vehicle group based on a tree dynamic growth self-organizing mapping algorithm, determining the sequence of the closed-loop tracks of the unmanned aerial vehicle group in the horizontal direction in a collaborative search mode, and adjusting the flight height of the unmanned aerial vehicle group; traversing the neighborhoods of all target positions;
3) planning the smooth closed-loop three-dimensional track of the unmanned aerial vehicle group by using a piecewise smooth curve and a local iterative optimization method for the access node of the three-dimensional track of the unmanned aerial vehicle group finally determined in the step 2);
4) setting a path speed constraint condition and an acceleration constraint condition in the unmanned aerial vehicle group motion control according to the smooth closed-loop three-dimensional track of the unmanned aerial vehicle group obtained in the step 3), calculating a speed profile of the unmanned aerial vehicle group according to the set path speed constraint condition and the set acceleration constraint condition, solving a path corresponding to the shortest time actively sensed by the multi-unmanned aerial vehicle collaborative area according to the speed profile, and controlling the unmanned aerial vehicle group to realize target search through the path.
The invention provides a collaborative active perception method for unmanned aerial vehicle group rapid target search, which aims at large-scale and incompletely known complex environments and is based on a tree dynamic growth self-organizing mapping algorithm and a Bezier smooth planning method so as to solve the problem of rapid flight path planning of unmanned aerial vehicle group collaborative area target search under multiple constraints. Compared with the prior art, the method has the following beneficial effects:
1) compared with a single unmanned aerial vehicle track planning method, the unmanned aerial vehicle track planning method can solve the problem of rapid three-dimensional track planning of an unmanned aerial vehicle group, complete a target search task and realize active perception of the unmanned aerial vehicle group. The unmanned aerial vehicle can adjust the flying height and speed according to the actual environment in the navigation stage, and is suitable for different terrains; the target is searched and observed with minimum cost and fastest speed, and regional information is collected in real time to the maximum extent.
2) The method solves the problem of multi-unmanned aerial vehicle multi-collaborative track planning, applies the tree-shaped dynamic growth self-organizing mapping algorithm and Bezier smooth planning to the problems of fast track planning and active sensing in order to improve the algorithm execution efficiency, and has the advantages of high convergence speed and high convergence precision.
3) Different from the traditional Dubins curve constraint, the unmanned plane group can accelerate linearly and turn at a small radius and a small speed, is not constrained by the minimum forward speed and the minimum turning radius of the unmanned plane but is constrained by the maximum speed and the maximum acceleration of the unmanned plane, so that the method adopts a three-order Bezier curve to carry out smooth planning, and improves the practicability and the high efficiency of the method in an actual scene.
Drawings
FIG. 1 is a flow chart of a collaborative active sensing method for unmanned aerial vehicle group rapid target search;
FIG. 2 is a diagram of relative position relationships among a population of unmanned aerial vehicles, a target location, and a field of target locations;
FIG. 3 is a flow chart of node update and adjustment.
Detailed Description
The invention will be further explained with reference to the drawings.
Referring to fig. 1, a flow chart of a collaborative active sensing method for fast target search of an unmanned aerial vehicle group is provided, and the method is particularly suitable for fast target search and track planning in real time for a large-scale and incompletely known complex target area under a multi-constraint condition by an unmanned aerial vehicle group, and specifically includes the following steps:
1) and setting the number N of the unmanned aerial vehicles, starting points of the unmanned aerial vehicles and the maximum iteration number of subsequent calculation. Setting n target positions of the unmanned aerial vehicle group according to a target monitoring area of the unmanned aerial vehicle group; the unmanned aerial vehicle group includes N unmanned aerial vehicles that independently move, carries on a vision sensor on every unmanned aerial vehicle respectively, sets for every unmanned aerial vehicle's supervision radius R according to the inductive distance of this vision sensor (in this embodiment, each vision sensor is the same, so every unmanned aerial vehicle's supervision radius R is equal, and sets for the inductive distance of vision sensor). In order to save target searching time and improve the efficiency and adaptability of active sensing, the set monitoring radius is used as the neighborhood of each target position, and the unmanned aerial vehicle group only needs to observe the target in the monitoring radius, so that the observation effect can be considered to be achieved. Fig. 2 is a relative position relationship diagram of an unmanned aerial vehicle group (in this embodiment, an unmanned aerial vehicle group is formed by 3 unmanned aerial vehicles), a target position and a target position neighborhood, where points a, B, and C are starting points of the unmanned aerial vehicle group, a black point is a target position of a monitoring area, and a gray circular area is a neighborhood of the target position.
2) Determining an access node of an unmanned aerial vehicle group track based on a tree dynamic growth self-organizing mapping algorithm, and specifically comprising the following steps:
2.1) expand the three-dimensional trajectory planning of unmanned aerial vehicle crowd on two-dimensional trajectory planning basis, be about to the three-dimensional trajectory planning decoupling of unmanned aerial vehicle crowd for the planning problem of horizontal and vertical direction, each unmanned aerial vehicle in the unmanned aerial vehicle crowd is in the trajectory planning of three-dimensional space, and the position is expressed as:
wherein theta is a turning angle of the unmanned aerial vehicle, psi is a pitch angle of the unmanned aerial vehicle; establishing a coordinate system by taking the circle center of the robot chassis as an origin and taking any two directions which are perpendicular to each other on a plane where the robot chassis is located as x and y axes, wherein the z axis and the x and y axes meet the right-hand spiral rule;
2.2) generating and updating access nodes of closed-loop tracks of unmanned aerial vehicle groups (each unmanned aerial vehicle corresponds to one closed-loop track) based on a tree dynamic growth self-organizing mapping algorithm, determining the sequence of the closed-loop tracks of the unmanned aerial vehicle groups in the horizontal direction in a collaborative search, and adjusting the flight heights of the unmanned aerial vehicle groups; the neighborhood of all target locations is traversed. Referring to fig. 3, the specific implementation process is as follows (in the execution process, each drone synchronously completes the following steps):
2.2.1) constructing a tree-shaped dynamic growth self-organizing mapping double-layer neural network, randomly selecting root nodes, setting each neuron node in the double-layer neural network to correspond to one access node on a closed-loop track of an unmanned aerial vehicle group, taking the target position of the unmanned aerial vehicle group as an input layer of the tree-shaped dynamic growth self-organizing mapping double-layer neural network, taking N access nodes in total of N closed-loop tracks of the unmanned aerial vehicle group as an output layer of the tree-shaped dynamic growth self-organizing mapping double-layer neural network, and respectively positioning each access node output by the tree-shaped dynamic growth self-organizing mapping double-layer neural network in the neighborhood of the corresponding target position;
2.2.2) connecting adjacent access nodes on the closed-loop track of each unmanned aerial vehicle by a three-order Bezier curve respectively, wherein each unmanned aerial vehicle forms a smooth closed-loop track by m (m is a positive integer) segmentation curves respectively;
2.2.3) for the current target position of the unmanned aerial vehicle group, respectively traversing the piecewise curve between each access node on each unmanned aerial vehicle closed-loop track, finding out the point on the closed-loop track closest to the current target position waypoint, and taking the point as a newly added access node, wherein the target position waypoint is any point which is located in the neighborhood range of the target position and is on the connecting line of each access node and the corresponding target position; if the newly added access node is in the neighborhood of the corresponding target position, the position of the newly added access node is not adjusted, and the step 2.2.4 is executed); if the newly added access node is outside the neighborhood of the corresponding target position, adjusting the position relation between other access nodes corresponding to each target position and the corresponding target position waypoint, wherein the adjusting method comprises the following steps:
v′＝v+μf(σ，d)(v*.sp-v)
v is the position of the access node to be adjusted; v' is the adjusted position of the access node; mu is the learning rate of the tree-shaped dynamic growth self-organizing mapping algorithm, and mu is more than 0 and less than 1; f (sigma, d) is a neighborhood function of the access nodes, sigma is the gain rate of the tree-shaped dynamic growth self-organizing mapping algorithm, 0 & ltsigma & lt 1, d represents the distance from the access nodes to be adjusted to the newly added access nodes, M is the number of the access nodes before adjustment in the closed-loop track of each unmanned aerial vehicle, and v is the number of the access nodes before adjustment in the closed-loop track of each unmanned aerial vehicle*.spThe position of a target position waypoint corresponding to the access node to be adjusted; the physical meaning of f (σ, d) is: if the distance d from the original access node to the newly-added access node is within 0.2M, adjusting the corresponding target position waypoint according to a set gain rate, otherwise, not adjusting, namely f (sigma, d) is 0;
2.2.4) updating and adjusting other target positions on the closed-loop track of each unmanned aerial vehicle according to the step 2.2.2) and the step 2.2.3);
2.2.5) removing the access nodes which are not in the neighborhood of the target position after updating and adjustment or redundant access nodes in the neighborhood of the same target position from the corresponding closed-loop track to obtain winning access nodes, so as to determine the sequence of the closed-loop track of the unmanned aerial vehicle group collaborative search in the horizontal direction; solving first and second derivatives of the Z axis according to a third-order segmented Bezier curve expression so as to adjust the flying height of the unmanned aerial vehicle group;
2.2.6) repeatedly executing the steps 2.2.2) -2.2.5) until the iteration times set in the step 1) are reached or all winning access nodes in each closed-loop track are in the corresponding target position neighborhood, and ending the iteration.
3) Planning the smooth closed-loop three-dimensional track of the unmanned aerial vehicle group by using a piecewise smooth curve and a local iterative optimization method for the access node of the three-dimensional track of the unmanned aerial vehicle group finally determined in the step 2). In the embodiment, a three-order segmented Bezier curve is utilized, all curves on the track are ensured to meet smooth and continuous conditions, and the turning angle theta, the pitch angle psi and the starting point tangent length l of all segmented curves in the three-dimensional tracks of the unmanned aerial vehicle group are subjected to local iterative optimization algorithmaAnd end tangent length lbAnd optimizing the four parameters to obtain the optimized smooth closed-loop three-dimensional track of the unmanned aerial vehicle group. The specific process of the step is as follows:
selecting four control points B for each section of segmented curve of each unmanned aerial vehicle closed-loop track0，B1，B2，B3The expression of the adopted third-order segmented Bezier curve is as follows:
X(τ)＝B0(1-τ)3+3B1τ(1-τ)2+3B2τ2(1-τ)+B3τ3
wherein the parameter tau ∈ [0,1]，B0，B3For each segment, the position of the start and end points of the segmented Bezier curve, B1，B2Determining tangents at the starting point and the ending point of each segment of the segmented Bezier curve; the closed-loop track curve of each unmanned aerial vehicle consists of m segmented Bezier curves, and in order to ensure track continuity, two adjacent segmented Bezier curves X in the closed-loop track of each unmanned aerial vehiclei-1，XiThe control points of (c) need to satisfy the following relation:
is calculated according to the following formulaTangent line of starting point of i-1 th segmented Bezier curve in closed-loop track of each unmanned aerial vehicle
And tangent to the i-th segmented Bezier curve end point
Tangent line
Length of (2)
And tangent line
Length of (2)
Respectively as follows:
calculating the i-1 segmented Bezier curve X in the closed-loop track of each unmanned aerial vehicle according to the following formulai-1The tangent at the starting point is:
calculating the ith segmented Bezier curve X in the closed-loop track of each unmanned aerial vehicle according to the following formulaiThe tangent at the end point is:
the unmanned plane group track smoothness meets the expression:
four parameters theta are optimized by local iterationi，ψi，
And optimizing, wherein the two angle optimization step lengths are 0.01 pi, the two length parameter optimization step lengths are 0.5%, and when the iteration times reach 50 times, the optimization is finished, so that the smooth closed-loop three-dimensional trajectory planning of the unmanned aerial vehicle group is realized.
4) Setting path speed and acceleration constraint conditions in the unmanned aerial vehicle group motion control according to the smooth closed-loop three-dimensional track of the unmanned aerial vehicle group obtained in the step 3), and calculating a speed profile of the unmanned aerial vehicle group according to the set path speed and acceleration constraint conditions; and solving a path corresponding to the shortest time actively sensed by the multi-unmanned aerial vehicle cooperative area according to the speed profile, and controlling an unmanned aerial vehicle group to realize target search through the path.
In this step, the motion of unmanned aerial vehicle horizontal and vertical direction is restricted by respective maximum speed and maximum acceleration, and wherein, horizontal direction acceleration resolvable is tangential acceleration and radial acceleration. Tangential acceleration influences the speed of the unmanned aerial vehicle, and radial acceleration only influences the curvature of the track of the unmanned aerial vehicle. The time of flight is determined by the velocity profile of the trajectory. The problem of the shortest time optimization of active sensing of the multi-unmanned aerial vehicle collaborative area is solved by calculating a speed profile. The method specifically comprises the following steps:
4.1) uniformly sampling all the segmented Bezier curves in the smooth closed-loop three-dimensional track of the unmanned aerial vehicle group, and calculating the curvature of each sampling point of the segmented Bezier curve expression which meets the requirement determined in the step 3);
4.2) setting the initial speed and the ending speed of the unmanned aerial vehicle group as 0 to finish initialization;
4.3) calculating first and second derivatives of a smooth closed-loop three-dimensional trajectory curve equation of the unmanned aerial vehicle group along the vertical direction, and constraining and planning the trajectory according to the set maximum vertical speed and maximum vertical acceleration of the unmanned aerial vehicle group; calculating the maximum horizontal velocity at each sampling point according to the set maximum horizontal velocity and maximum horizontal acceleration and the curvature of each sampling point calculated in the step 4.1);
4.4) according to the set maximum tangential acceleration, adopting forward iteration to limit the speed (including the speed in the horizontal direction and the speed in the vertical direction) of the unmanned aerial vehicle group; and according to the set maximum tangential acceleration, backward iteration is adopted to limit the speed (including the horizontal direction and the vertical direction) of the sampling point to be calculated.
4.5) setting the total T of the navigation time of the unmanned aerial vehicle group as an objective function, wherein the expression is as follows:
if the minimum target function of the unmanned aerial vehicle group is met, returning a track curve of each unmanned aerial vehicle, namely returning each unmanned aerial vehicle path corresponding to the shortest total time of unmanned aerial vehicle group navigation, and searching the target by each unmanned aerial vehicle according to the obtained path; wherein, χ is the final trajectory of the UAV, XiAnd segmenting the Bezier curve for the ith unmanned plane.
And the unmanned aerial vehicles in the unmanned aerial vehicle cluster synchronously execute the steps, and the planned route enables the unmanned aerial vehicles to visit the position of the target object as much as possible in the shortest time, so that the cooperative and rapid target search is realized, and the active sensing task is completed.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (4)
1. A collaborative active perception method for unmanned aerial vehicle group rapid target search is characterized by comprising the following steps:
1) setting n target positions of the unmanned aerial vehicle group according to a target monitoring area of the unmanned aerial vehicle group; the unmanned aerial vehicle group comprises N unmanned aerial vehicles which move independently, each unmanned aerial vehicle is provided with a visual sensor, the monitoring radius of each unmanned aerial vehicle is set according to the sensing distance of each visual sensor, and the monitoring radius is used as the neighborhood of each target position;
2) determining an access node of an unmanned aerial vehicle group track based on a tree dynamic growth self-organizing mapping algorithm, and specifically comprising the following steps:
2.1) decoupling the three-dimensional trajectory planning of the unmanned aerial vehicle group into planning problems in the horizontal and vertical directions;
2.2) generating and updating access nodes of the closed-loop tracks of the unmanned aerial vehicle group based on a tree dynamic growth self-organizing mapping algorithm, determining the sequence of the closed-loop tracks of the unmanned aerial vehicle group in the horizontal direction in a collaborative search mode, and adjusting the flight height of the unmanned aerial vehicle group; traversing the neighborhoods of all target positions;
3) planning the smooth closed-loop three-dimensional track of the unmanned aerial vehicle group by using a piecewise smooth curve and a local iterative optimization method for the access node of the three-dimensional track of the unmanned aerial vehicle group finally determined in the step 2);
4) setting a path speed constraint condition and an acceleration constraint condition in the unmanned aerial vehicle group motion control according to the smooth closed-loop three-dimensional track of the unmanned aerial vehicle group obtained in the step 3), calculating a speed profile of the unmanned aerial vehicle group according to the set path speed constraint condition and the set acceleration constraint condition, solving a path corresponding to the shortest time actively sensed by the multi-unmanned aerial vehicle collaborative area according to the speed profile, and controlling the unmanned aerial vehicle group to realize target search through the path.
2. The collaborative active perception method for unmanned aerial vehicle group-oriented fast target search according to claim 1, wherein: the step 2.2) is realized by the following steps:
2.2.1) constructing a tree-shaped dynamic growth self-organizing mapping double-layer neural network, randomly selecting root nodes, setting each neuron node in the tree-shaped dynamic growth self-organizing mapping double-layer neural network to correspond to one access node on a closed-loop track of an unmanned aerial vehicle group, taking the target position of the unmanned aerial vehicle group as an input layer of the tree-shaped dynamic growth self-organizing mapping double-layer neural network, taking N access nodes in total of N closed-loop tracks of the unmanned aerial vehicle group as an output layer of the tree-shaped dynamic growth self-organizing mapping double-layer neural network, and respectively positioning each output access node in the neighborhood of one corresponding target position;
2.2.2) connecting adjacent access nodes on the closed-loop track of each unmanned aerial vehicle by a three-order segmented Bezier curve respectively, wherein each unmanned aerial vehicle forms a smooth closed-loop track by m segmented curves respectively;
2.2.3) for the current target position of the unmanned aerial vehicle group, respectively traversing the piecewise curve between each access node on each unmanned aerial vehicle closed-loop track, finding out the point on the closed-loop track closest to the current target position waypoint, and taking the point as a newly added access node; the target position waypoint is any point which is on a connecting line of the access node and the corresponding target position and is positioned in the neighborhood range of the target position; if the newly added access node is in the target position neighborhood, the position of the newly added access node is not adjusted, and the step 2.2.4 is executed); if the newly added access node is outside the neighborhood of the corresponding target position, adjusting other access nodes corresponding to the target positions to approach to the waypoint of the target position, wherein the adjusting method comprises the following steps:
v′＝v+μf(σ,d)(v*.sp-v)
v is the position of the access node to be adjusted; v' is the adjusted position of the access node; mu is the learning rate of the tree-shaped dynamic growth self-organizing mapping algorithm, 0<μ<1; f (sigma, d) is a neighborhood function of the access node, sigma is the gain rate of the tree dynamic growth self-organizing mapping algorithm, and 0<σ<1, d represents the distance from the access node to be adjusted to the newly added access node, M is the number of the access nodes before adjustment in the closed-loop track of each unmanned aerial vehicle, v*.spThe position of a target position waypoint corresponding to the access node to be adjusted;
2.2.4) updating and adjusting other target positions on the closed-loop track of each unmanned aerial vehicle according to the step 2.2.2) and the step 2.2.3);
2.2.5) removing the access nodes which are not in the neighborhood of the target position after updating and adjustment or redundant access nodes in the neighborhood of the same target position from the corresponding closed-loop track to obtain winning access nodes, so as to determine the sequence of the closed-loop track of the unmanned aerial vehicle group collaborative search in the horizontal direction; solving first and second derivatives of the Z axis according to the third-order segmented Bezier curve so as to adjust the flying height of the unmanned aerial vehicle group;
2.2.6) repeatedly executing the steps 2.2.2) -2.2.5) until the iteration times set in the step 1) are reached or all winning access nodes in each closed-loop track are in the corresponding target position neighborhood, and ending the iteration.
3. The cooperative active sensing method for unmanned aerial vehicle group-oriented fast target search according to claim 1 or 2, wherein a third-order segmented Bezier curve is adopted as the segmented smooth curve in step 3), and specifically comprises the following steps:
selecting four control points B for each section of segmented curve of each unmanned aerial vehicle closed-loop track0,B1,B2,B3The adopted three-order segmented Bezier curve expression is as follows:
X(τ)＝B0(1-τ)3+3B1τ(1-τ)2+3B2τ2(1-τ)+B3τ3
wherein the parameter tau ∈ [0,1]，B0,B3For the starting and end positions of each segmented Bezier curve, B1,B2Defining tangents at the starting point and the ending point of each segmented Bezier curve; the closed loop curve of each unmanned aerial vehicle consists of m segmented Bezier curves, and in order to ensure the continuous track, two adjacent segmented Bezier curves X in the closed loop track of each unmanned aerial vehiclei-1,XiThe requirements are as follows:
is the (i-1) thThe position of the end point of the segmented Bezier curve,
the starting point position of the ith segmented Bezier curve;
calculating the tangent line of the starting point of the i-1 th segmented Bezier curve in the closed-loop track of each unmanned aerial vehicle according to the following formula
And tangent to the i-th segmented Bezier curve end point
Tangent line
Length of (2)
And tangent line
Length of (2)
Respectively as follows:
wherein,
is a tangent at the starting point of the i-1 th segmented Bezier curve,
is the ith-The position of the start of the 1 segmented Bezier curve,
for the position of the end point of the i-th segmented Bezier curve,
is a tangent line at the end point of the ith segmented Bezier curve;
calculating the i-1 segmented Bezier curve X in the closed-loop track of each unmanned aerial vehicle according to the following formulai-1The tangent at the starting point is:
in the formula, theta is the turning angle of the unmanned aerial vehicle, psi is the pitch angle of the unmanned aerial vehicle;
calculating the ith segmented Bezier curve X in the closed-loop track of each unmanned aerial vehicle according to the following formulaiThe tangent at the end point is:
the unmanned plane group track smoothness meets the expression:
four parameters theta are optimized by local iterationi,ψi,
And optimizing to realize smooth closed-loop three-dimensional trajectory planning of the unmanned aerial vehicle group.
4. The collaborative active perception method for unmanned aerial vehicle group-oriented fast target search according to claim 3, wherein the step 4) specifically includes the following steps:
4.1) uniformly sampling all the segmented Bezier curves in the smooth closed-loop three-dimensional track of the unmanned aerial vehicle group, and calculating the curvature of each sampling point of the segmented Bezier curve expression which meets the requirement determined in the step 3);
4.2) setting the initial speed and the ending speed of the unmanned aerial vehicle group as 0 to finish initialization;
4.3) calculating first and second derivatives of a smooth closed-loop three-dimensional trajectory curve equation of the unmanned aerial vehicle group along the vertical direction, and constraining and planning the trajectory according to the set maximum vertical speed and maximum vertical acceleration of the unmanned aerial vehicle group; calculating the maximum horizontal velocity at each sampling point according to the set maximum horizontal velocity and maximum horizontal acceleration and the curvature of each sampling point calculated in the step 4.1);
4.4) limiting the speed of the unmanned aerial vehicle group by adopting forward iteration according to the set maximum tangential acceleration; according to the set maximum tangential acceleration, backward iteration is adopted to limit the speed of the sampling point to be calculated;
4.5) setting the total T of the navigation time of the unmanned aerial vehicle group as an objective function, wherein the expression is as follows:
if the minimum objective function of the unmanned aerial vehicle group is met, returning a track curve of each unmanned aerial vehicle, namely returning each unmanned aerial vehicle path corresponding to the shortest unmanned aerial vehicle group navigation time, and searching the target by each unmanned aerial vehicle according to the obtained path; wherein, χ is the final trajectory of the UAV, XiAnd segmenting the Bezier curve for the ith unmanned plane.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910151483.8A CN109947136B (en) | 2019-02-28 | 2019-02-28 | Collaborative active sensing method for unmanned aerial vehicle group rapid target search |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910151483.8A CN109947136B (en) | 2019-02-28 | 2019-02-28 | Collaborative active sensing method for unmanned aerial vehicle group rapid target search |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109947136A CN109947136A (en) | 2019-06-28 |
| CN109947136B true CN109947136B (en) | 2020-07-10 |
Family
ID=67007112
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910151483.8A Active CN109947136B (en) | 2019-02-28 | 2019-02-28 | Collaborative active sensing method for unmanned aerial vehicle group rapid target search |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109947136B (en) |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110427046B (en) * | 2019-07-26 | 2022-09-30 | 沈阳航空航天大学 | Three-dimensional smooth random-walking unmanned aerial vehicle cluster moving model |
| CN110362115B (en) * | 2019-07-31 | 2022-02-18 | 中国人民解放军总参谋部第六十研究所 | Time constraint simultaneous arrival multi-unmanned aerial vehicle path planning algorithm |
| CN110488835B (en) * | 2019-08-28 | 2020-12-08 | 北京航空航天大学 | Unmanned system intelligent local path planning method based on double back propagation neural networks |
| CN110413007B (en) * | 2019-08-30 | 2022-04-19 | 中国科学技术大学 | Control method and system for flight path of unmanned aerial vehicle, electronic device and medium |
| CN110646857B (en) * | 2019-09-12 | 2021-06-08 | 北京自动化控制设备研究所 | Underwater magnetic target distributed searching method |
| EP4053667A1 (en) * | 2019-11-01 | 2022-09-07 | Guangzhou Xaircraft Technology Co., Ltd. | Route altitude adjustment method, unmanned aerial vehicle operation method, and related apparatus |
| CN111207752B (en) * | 2020-01-17 | 2022-10-04 | 西北工业大学 | Unmanned aerial vehicle track planning method based on dynamic tangent point adjustment |
| CN113742607B (en) * | 2020-05-28 | 2023-12-08 | 浙江财经大学 | Stay position recommending method based on geographical track of principal |
| CN112947558B (en) * | 2021-02-07 | 2023-04-07 | 北京理工大学 | Space-time synchronization collaborative trajectory planning method |
| CN112947594B (en) * | 2021-04-07 | 2023-08-04 | 东北大学 | Unmanned aerial vehicle-oriented track planning method |
| CN114485669B (en) * | 2022-01-20 | 2024-04-19 | 南京航空航天大学 | Unmanned aerial vehicle track planning method oriented to shooting utility maximization |
| CN115356910A (en) * | 2022-07-29 | 2022-11-18 | 杭州电子科技大学 | Cluster control method based on perception control and perception tolerance |
| CN115268504B (en) * | 2022-09-29 | 2022-12-27 | 四川腾盾科技有限公司 | Ground-imitating flight path planning method for large unmanned aerial vehicle |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103197684A (en) * | 2013-04-25 | 2013-07-10 | 清华大学 | Method and system for cooperatively tracking target by unmanned aerial vehicle cluster |
| CN105425820A (en) * | 2016-01-05 | 2016-03-23 | 合肥工业大学 | Unmanned aerial vehicle cooperative search method for moving object with perception capability |
| CN105700555A (en) * | 2016-03-14 | 2016-06-22 | 北京航空航天大学 | Potential game-based multi-unmanned aerial vehicle cooperative search method |
| CN106406354A (en) * | 2016-11-29 | 2017-02-15 | 中山大学 | Distributed aircraft formation method based on three-dimensional dynamic obstacle avoidance |
| CN107037829A (en) * | 2017-05-09 | 2017-08-11 | 广东容祺智能科技有限公司 | A kind of cluster unmanned plane route planning method |
| CN107063255A (en) * | 2017-01-09 | 2017-08-18 | 北京工业大学 | A kind of three-dimensional Route planner based on improvement drosophila optimized algorithm |
| CN107168380A (en) * | 2017-06-28 | 2017-09-15 | 西安电子科技大学 | A kind of multi-step optimization method of the unmanned aerial vehicle group region overlay based on ant group algorithm |
| CN108398960A (en) * | 2018-03-02 | 2018-08-14 | 南京航空航天大学 | A kind of multiple no-manned plane collaboration target tracking method for improving APF and being combined with segmentation Bezier |
| CN108563243A (en) * | 2018-06-28 | 2018-09-21 | 西北工业大学 | A kind of unmanned aerial vehicle flight path planing method based on improvement RRT algorithms |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10586464B2 (en) * | 2015-07-29 | 2020-03-10 | Warren F. LeBlanc | Unmanned aerial vehicles |
-
2019
- 2019-02-28 CN CN201910151483.8A patent/CN109947136B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103197684A (en) * | 2013-04-25 | 2013-07-10 | 清华大学 | Method and system for cooperatively tracking target by unmanned aerial vehicle cluster |
| CN105425820A (en) * | 2016-01-05 | 2016-03-23 | 合肥工业大学 | Unmanned aerial vehicle cooperative search method for moving object with perception capability |
| CN105700555A (en) * | 2016-03-14 | 2016-06-22 | 北京航空航天大学 | Potential game-based multi-unmanned aerial vehicle cooperative search method |
| CN106406354A (en) * | 2016-11-29 | 2017-02-15 | 中山大学 | Distributed aircraft formation method based on three-dimensional dynamic obstacle avoidance |
| CN107063255A (en) * | 2017-01-09 | 2017-08-18 | 北京工业大学 | A kind of three-dimensional Route planner based on improvement drosophila optimized algorithm |
| CN107037829A (en) * | 2017-05-09 | 2017-08-11 | 广东容祺智能科技有限公司 | A kind of cluster unmanned plane route planning method |
| CN107168380A (en) * | 2017-06-28 | 2017-09-15 | 西安电子科技大学 | A kind of multi-step optimization method of the unmanned aerial vehicle group region overlay based on ant group algorithm |
| CN108398960A (en) * | 2018-03-02 | 2018-08-14 | 南京航空航天大学 | A kind of multiple no-manned plane collaboration target tracking method for improving APF and being combined with segmentation Bezier |
| CN108563243A (en) * | 2018-06-28 | 2018-09-21 | 西北工业大学 | A kind of unmanned aerial vehicle flight path planing method based on improvement RRT algorithms |
Non-Patent Citations (3)
| Title |
|---|
| 《TGSOM:一种用于数据聚类的动态自组织映射神经网络》;王莉等;《电子与信息学报》;20030331;第2 5 卷(第3 期);第313-319页 * |
| 《复杂环境下多无人机协作式地面移动目标跟踪》;王林等;《控制理论与应用》;20110331;第28 卷(第3 期);第300-308页 * |
| 《高动态自组织无人机网络架构设计与模式研究》;陈思静等;《计算机科学》;20150228;第４２卷(第２期);第50-54页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109947136A (en) | 2019-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109947136B (en) | Collaborative active sensing method for unmanned aerial vehicle group rapid target search | |
| CN109871032B (en) | Multi-unmanned aerial vehicle formation cooperative control method based on model predictive control | |
| CN110470301B (en) | Unmanned aerial vehicle path planning method under multi-dynamic task target point | |
| CN111176334B (en) | Multi-unmanned aerial vehicle cooperative target searching method | |
| CN109357678B (en) | Multi-unmanned aerial vehicle path planning method based on heterogeneous pigeon swarm optimization algorithm | |
| CN112684807A (en) | Unmanned aerial vehicle cluster three-dimensional formation method | |
| CN109917817B (en) | Multi-underwater robot collaborative path planning method | |
| CN111024085B (en) | Unmanned aerial vehicle track planning method with end point direction and time constraints | |
| CN110320919B (en) | Method for optimizing path of mobile robot in unknown geographic environment | |
| CN109254592B (en) | Unmanned aerial vehicle formation change control method and system | |
| CN108549210A (en) | Multiple no-manned plane based on BP neural network PID control cooperates with flying method | |
| CN113759935B (en) | Intelligent group formation mobile control method based on fuzzy logic | |
| CN114967744A (en) | Planning method for multi-unmanned aerial vehicle cooperative obstacle avoidance | |
| CN115202349A (en) | Multi-mobile-robot cooperative formation control method, device, equipment and storage medium based on communication interference | |
| CN110412984B (en) | Cluster safety consistency controller and control method thereof | |
| CN115202357A (en) | Autonomous mapping method based on impulse neural network | |
| CN114545971A (en) | Multi-agent distributed flyable path planning method, system, computer equipment and medium under communication constraint | |
| CN114138022A (en) | Distributed formation control method for unmanned aerial vehicle cluster based on elite pigeon swarm intelligence | |
| CN117970928A (en) | Route planning method based on cooperation of multiple mobile robots of ROS system | |
| CN115454061B (en) | Robot path obstacle avoidance method and system based on 3D technology | |
| CN114815899A (en) | Unmanned aerial vehicle three-dimensional space path planning method based on 3D laser radar sensor | |
| Zheng et al. | A method for UAV tracking target in obstacle environment | |
| Zhang et al. | Path Prediction Method for Automotive Applications Based on Cubic Spline Interpolation | |
| ZHANG et al. | AUV cluster path planning based on improved RRT* algorithm | |
| Yi et al. | Lidar Odometry and Mapping Optimized by the Theory of Functional Systems in the Parking Lot |
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 |