CN113190026B - Energy-saving underwater path planning method based on re-excitation mechanism particle swarm algorithm - Google Patents

Energy-saving underwater path planning method based on re-excitation mechanism particle swarm algorithm Download PDF

Info

Publication number
CN113190026B
CN113190026B CN202110504234.XA CN202110504234A CN113190026B CN 113190026 B CN113190026 B CN 113190026B CN 202110504234 A CN202110504234 A CN 202110504234A CN 113190026 B CN113190026 B CN 113190026B
Authority
CN
China
Prior art keywords
path
particle swarm
auv
swarm algorithm
algorithm
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
Application number
CN202110504234.XA
Other languages
Chinese (zh)
Other versions
CN113190026A (en
Inventor
黄浩乾
金超
吴昊
李光辉
魏嘉颖
杨晨
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hohai University HHU
Original Assignee
Hohai University HHU
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hohai University HHU filed Critical Hohai University HHU
Priority to CN202110504234.XA priority Critical patent/CN113190026B/en
Publication of CN113190026A publication Critical patent/CN113190026A/en
Application granted granted Critical
Publication of CN113190026B publication Critical patent/CN113190026B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/04—Control of altitude or depth
    • G05D1/06—Rate of change of altitude or depth
    • G05D1/0692—Rate of change of altitude or depth specially adapted for under-water vehicles

Landscapes

  • Engineering & Computer Science (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Feedback Control In General (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

The invention discloses an energy-saving underwater path planning method based on a re-excitation mechanism particle swarm algorithm, which comprises the steps of firstly generating a convex polygonal obstacle by utilizing a Graham algorithm, and expanding the obstacle outwards to generate a danger zone; then, by using an MAKLINK graph theory method, connecting the vertex lines of the convex polygonal obstacles to generate a undirected network graph; carrying out primary suboptimal path planning by utilizing a Dijkstra algorithm to obtain a shortest link line; the particle swarm algorithm is improved by using a re-excitation mechanism and is integrated into the influence of ocean current, so that the autonomous underwater robot (AUV) is guaranteed to maximally utilize ocean current, and energy is saved; and then, path optimization is carried out by using a final improved particle swarm algorithm, and the Bessel algorithm is used for smoothing the optimal path, so that the AUV is ensured to find the optimal path under the condition of saving energy, and safe obstacle avoidance is realized.

Description

Energy-saving underwater path planning method based on re-excitation mechanism particle swarm algorithm
Technical Field
The invention relates to a path planning method of an underwater robot, in particular to an energy-saving underwater path planning method based on a re-excitation mechanism particle swarm algorithm.
Background
The AUV is used as a main tool for detecting the sea by the human, safely avoids obstacles when interaction is carried out between underwater and an external environment, and a reasonable path is planned, so that the safety of the underwater robot is guaranteed. The underwater environment is complex and complicated, various obstacles exist, and the AUV is influenced most by ocean currents, so that the consideration of ocean currents during underwater path planning has important significance for saving robot energy.
At present, path planning of underwater robots is researched in a two-dimensional environment, such as a typical a-algorithm, an Ant colony Algorithm (ACO), an Artificial Potential Field Method (APF), and the like, but all have respective defects, for example, the a-algorithm is slow in operation speed when processing a large-scale map, the Ant colony algorithm is slow in early convergence speed, and the Artificial Potential Field Method is easy to fall into local optimization. The current general environment modeling method is a grid method, but the calculation is complex, and the environment information is not easy to express.
Although the MAKLINK map can only be applied to a two-dimensional environment, the MAKLINK map can better express environment information, and can set corresponding obstacle areas according to actual environment information, so that the MAKLINK map is flexible and changeable. For complex marine environments, ocean currents are considered first, so that the influence of the ocean currents must be added when environment modeling and path planning are carried out.
The particle swarm algorithm has the advantages of simple parameters and high search speed, but the self-adaptive capacity of the particle swarm algorithm needs to be improved, and how to combine the particle swarm algorithm with other methods to plan a path is a technical problem which needs to be solved urgently.
Disclosure of Invention
The invention aims to: aiming at the technical problems that the existing AUV is easily influenced by ocean currents when working in a complex marine environment and the power consumption of an underwater robot needs to be considered, the invention provides an energy-saving underwater path planning method based on a re-excitation mechanism particle swarm algorithm, namely, the negative influence of the ocean currents is counteracted by using a speed synthesis method, the ocean currents are fully utilized, and therefore, the energy of the AUV is saved; meanwhile, in order to improve the self-adaptive capacity of the particle swarm optimization, the particle swarm optimization is improved underwater based on a re-excitation mechanism, and an optimal path is planned.
The technical scheme is as follows: the invention relates to an energy-saving underwater path planning method based on a re-excitation mechanism particle swarm algorithm, which comprises the following steps of:
step (1), a convex polygonal barrier is established according to a Graham algorithm, and a set distance is expanded outwards to establish a danger zone;
step (2), constructing a MAKLINK two-dimensional map based on the convex polygon barrier, and constructing a undirected network map on the basis of the MAKLINK two-dimensional map;
Step (3), performing suboptimal path optimization on the constructed two-dimensional map by using a Dijkstra algorithm to obtain a suboptimal path based on an MAKLINK map;
step (4), then, optimizing a suboptimal path by utilizing an improved particle swarm algorithm to finally obtain a shortest path;
and (5) sampling a plurality of equidistant points on the path at the broken line generated by the improved particle swarm optimization, and reducing the turning point of the shortest path by using a Bezier curve so as to smooth the path.
Wherein, step (1) includes the following content:
step (1.1), constructing a convex polygonal barrier by utilizing a Graham algorithm, and searching points on a convex hull to generate the convex polygonal barrier according to polar coordinate sequencing through a random number of points on a two-dimensional coordinate system;
step (1.2), on the basis of generating the convex-edge-shaped barrier, manually expanding a set distance outwards, wherein the distance value is dmin。
When the AUV carries out underwater path planning, the distance d between the AUV and an obstacle meets the following constraint:
d≥dmin (1)
in the step (2), the specific steps of constructing the MAKLINK two-dimensional map are as follows:
step (2.1), connecting a certain vertex of the convex polygonal barrier constructed in the step (1) with other vertexes or space boundary points to form a link line, wherein the link line does not penetrate through the barrier, setting a node on each link line, determining a feasible line segment of which the link line between the nodes does not penetrate through the barrier according to the actual situation of the MAKLINK diagram, and forming a feasible line segment set;
And (2.2) selecting the midpoint of the link line as a node, and connecting the nodes with each other to form a undirected network graph.
The suboptimal path optimization in the step (3) comprises the following steps:
on the basis of generating the undirected network map in the step (2), searching the shortest link line from the target to the starting point on the MAKLINK two-dimensional map according to a Dijkstra algorithm to form a suboptimal path, and obtaining the shortest path point, thereby obtaining the link line where each path point is located.
In the step (4), a new shortest path point on the link lines is determined by improving the particle swarm optimization, and then the determined new path points on each link line are connected to obtain the optimal planning path. Specifically, the particle swarm optimization is adopted to simulate the activities of birds, the optimal position is randomly searched, each particle is regarded as a search individual of a D-dimensional search space, and the speed V of each particle in a two-dimensional mapidAnd position XidRepresenting the velocity V and the position X, V of the ith particle in a two-dimensional map in D-dimensional spaceidAnd position XidFormula (2) and formula (3) after the particle swarm optimization is adopted for updating:
Vid＝ωVid+C1r1(Pid-Xid)+C2r2(Pgd-Xid) (2)
Xid＝Xid+Vid (3)
wherein, omega is an inertia factor, the value of omega is nonnegative, the global optimization capability is strong when the value is larger, and the local optimization capability is strong when the value is smaller; c 1And C2Is a learning factor and is a constant; r is1And r2Is [0,1 ]]Random number, PidFor an individual optimal solution, PgdIs a global optimal solution.
The process of adjusting ω according to the environment information is:
firstly, dividing each link line of a shortest path graph in an MAKLINK two-dimensional map into ten equally-divided points as sampling points, calculating the distance L between the sampling points, and searching the shortest distance L between all the sampling pointsminAnd LmaxThe cost calculation formulas from node i to node j are (4) and (5):
Figure BDA0003057688860000031
J＝∑cost (5)
j is the cost sum of the selected path cost value and the node to be selected; updating the inertial weight ω of the particle using J as the re-excitation signal, the update formula is (6):
ω＝|exp(KJ)-1| (6)
wherein K is a non-negative constant.
Then, the particle state is evaluated by adopting a velocity synthesis method, and when the ocean current velocity is known, the path planning is carried out by adjusting the velocity and the direction of the AUV, wherein the AUV has the expected velocity VEAUV desired Angle α of equation (7)3Is represented by formula (8):
Figure BDA0003057688860000032
α3＝arcsin(VC*sin(α2-α1)/VA)+α2 (8)
wherein, VcIs the ocean current velocity vector, VAIs the velocity vector of AUV, α1Angle of ocean current, α2Angle of line connecting AUV to target, alpha3Is the angle of the AUV.
Judging the current position of the particle according to an evaluation function, and updating X in the formula (2) by the particle according to the value of the evaluation function idAnd VidConsidering the influence of ocean currents on the AUV, the evaluation function is of formula (9):
Figure BDA0003057688860000033
Figure BDA0003057688860000034
Figure BDA0003057688860000035
wherein E, A, B, C are a non-negative constant; p is0,P1,P2…,Pi,Pi+1,PDThe shortest path point is obtained through a Dijkstra algorithm, and D represents the number of routes in the shortest path; viIs the velocity of the ith particle, αiIs the angle of the ith particle, VexpectedAnd alphaexpectedThe desired speed and angle are obtained by equations (7) and (8).
The working principle is as follows: firstly, generating a convex polygonal barrier by using a Graham algorithm, expanding a barrier model outwards by a set distance to generate a danger zone, and preventing an underwater robot from being close to the barrier and colliding; then, connecting the vertex lines of the final convex polygonal barriers by using an MAKLINK graph theory method to generate a undirected network graph; performing primary suboptimal path planning by utilizing a Dijkstra algorithm to obtain the shortest link line; the particle swarm algorithm is improved by using a re-excitation mechanism, and is integrated into the influence of ocean current, so that the Autonomous Underwater robot (AUV) is guaranteed to maximally utilize ocean current, and energy consumption is saved; and then, path optimization is carried out by using a final improved particle swarm optimization, and because the method can generate a group of unsmooth and oscillatory paths when planning the path of the robot, the complexity of robot control is increased, the Bessel algorithm is applied to the optimal path to smooth the path, so that the number of corners of the robot is reduced, and the AUV is ensured to find the optimal path under the condition of saving energy.
Has the beneficial effects that: compared with the prior art, the invention has the following advantages:
(1) in the environment modeling process, the convex polygonal barrier is expanded out of a dangerous area through the Graham algorithm, so that the danger that the underwater robot is easy to collide because the underwater robot is too close to the barrier is avoided.
(2) The method utilizes the MAKLINK to construct the two-dimensional environment map, can accurately express the barrier information, and ignores the height information in the marine environment under the actual condition; compared with a common grid map, the map calculation is quicker and quicker, and can be well applied to maps in a larger range.
(3) The path planning is performed for the first time by utilizing the Dijkstra algorithm, so that an optimized node is provided for the subsequent particle swarm algorithm, and the time for searching the path by the particle swarm algorithm is reduced.
(4) Because the MAKLINK graph can calculate the relevant information of the path according to the nodes, the particle swarm algorithm continuously judges whether the surrounding environment is good or bad according to the information of the nodes by utilizing a re-excitation mechanism in the searching process, so that the parameters of the particles are adjusted in a self-adaptive manner.
(5) Under the actual condition, the influence of ocean current exists in the ocean, and the path planning of the AUV is influenced most by the ocean current, so that the speed synthesis method is adopted to adjust the speed and the direction of the AUV to counteract and utilize the influence of the ocean current to generate a shortest path without collision and capable of saving energy, namely, the safe obstacle avoidance is realized, and the energy consumption of the underwater robot is saved.
Drawings
FIG. 1 is a flow chart of an energy-saving underwater path planning method based on a re-excitation mechanism particle swarm algorithm;
FIG. 2 is a schematic diagram of a convex hull in the underwater path planning method of the present invention;
FIG. 3 is a diagram illustrating a danger zone expansion in the underwater path planning method according to the present invention;
FIG. 4 is a multidirectional network diagram based on MAKLINK in the underwater path planning method of the present invention;
FIG. 5 is a schematic diagram of a re-excitation mechanism in the underwater path planning method of the present invention;
FIG. 6 is a schematic diagram of a velocity synthesis method in the underwater path planning method of the present invention;
FIG. 7 is a schematic diagram of a link line in the underwater path planning method of the present invention;
fig. 8 is a path optimization diagram in the underwater path planning method of the present invention.
Detailed Description
Example (b):
as shown in figure 1, the energy-saving underwater path planning method based on the re-excitation mechanism particle swarm optimization carries out environment modeling by utilizing a Graham algorithm and an MAKLINK graph theory, carries out path sub-optimization by utilizing a Dijkstra algorithm, carries out final optimization on a path by adopting an improved particle swarm algorithm, and meanwhile, carries out smoothing treatment on the path by using a Bezier curve in order to reduce the number of corners of a robot.
The invention relates to an underwater path planning method based on a re-excitation mechanism particle swarm algorithm, which comprises the following steps of:
Step (1), a convex polygonal obstacle model is established according to the Graham algorithm, a set distance is expanded outwards to establish a danger zone, and the specific process is as follows:
step (1.1), as shown in FIG. 2, a random number of points P are designed on a two-dimensional coordinate system0、P1、P2、P3And P4Specifically, the smallest one on the ordinate is found firstPoint P0Then translating the coordinate system to P0And (3) sequencing polar angles of all points to obtain a coordinate origin, searching the points on the convex hull anticlockwise to perform modeling of the convex polygon barrier, and generating a convex polygon barrier until all the points are searched, wherein the barrier is shown in fig. 3. The modeling of the obstacles is also constructed from a sea map to be more realistic.
Step (1.2), then, manually expanding a certain distance outwards on the basis of the convex polygonal barrier, wherein the distance value is dminMeter, fig. 3, to prevent collision when the AUV is operated under water too close to an obstacle. The dashed portions in fig. 3 are convex polygonal obstacles and the solid lines represent extended danger zones. When the AUV carries out underwater path planning, the distance d between the AUV and the obstacle satisfies the constraint of the formula (1):
d≥dmin (1)
step (2), constructing a MAKLINK two-dimensional map based on the convex polygon barrier, and constructing a undirected network map on the basis of the two-dimensional map; the method specifically comprises the following steps:
And (2.1) connecting a certain vertex of the convex polygonal obstacle constructed in the step (1) with other vertexes or space boundary points to form a link line, wherein the link line is required to be incapable of passing through the obstacle, a node is arranged on each link line, a connection line between the node and the node is determined according to the actual situation of the MAKLINK two-dimensional map, feasible line segments which do not pass through the obstacle are formed, a feasible line segment set is formed, and the formed dotted line is the MAKLINK two-dimensional map as shown in fig. 4.
And (2.2) selecting the midpoint of the link line as a node, and connecting the nodes with each other to form a undirected network graph. As shown in fig. 4, the solid line is a multidirectional network map based on MAKLINK two-dimensional map, in which the black convex polygon is an obstacle and the gray portion is a danger zone.
Step (3), finding a suboptimal path for the constructed two-dimensional map of the MAKLINK by utilizing a Dijkstra algorithm to obtain a suboptimal path based on the two-dimensional map of the MAKLINK, namely an initial optimized path, such as a gray solid line shown in FIG. 8, and the specific process comprises the following steps:
generating a undirected network graph in step (2)On the basis, a Dijkstra algorithm is utilized to search a child node closest to the starting point from the starting point, and the child node closest to the current node is selected when the next node is selected during each updating of the Dijkstra algorithm, so that the shortest path is ensured, a suboptimal path is finally formed, as shown in FIG. 8, and the shortest path point P is obtained 1,P2,P3,P4,P5,P12,P13,P11So as to obtain the link line P where each path point is locatedi1Pi2As shown in fig. 7.
Step (4), optimizing a suboptimal path by utilizing an improved particle swarm algorithm to obtain a shortest path; the specific analysis is as follows:
because the motion space of the AUV is composed of convex polygonal obstacles, the path slides on the link line and cannot intersect with the obstacles, so that a new shortest path point on the link line is determined by improving the particle swarm algorithm, and then the determined new shortest path points on each link line are connected to finally obtain the optimal planned path, and the process is as follows:
(4.1) simulating the activities of birds by adopting a particle swarm algorithm, randomly searching for an optimal position, and regarding each particle as a searching individual of a D-dimensional searching space, wherein the individual has 2 attributes: velocity V of each particle in a two-dimensional mapidAnd position XidAnd (3) representing the speed V and the position X of the ith particle in the two-dimensional map in the D-dimensional space, wherein the formulas after the particle swarm optimization is adopted for updating are (2) and (3).
Vid＝ωVid+C1r1(Pid-Xid)+C2r2(Pgd-Xid) (2)
Xid＝Xid+Vid (3)
Wherein, omega is an inertia factor, the value of omega is nonnegative, and the global optimization capability is strong when the value is larger; when the size is small, the local optimizing capacity is strong. C1And C2Is a learning factor and is a constant. r is1And r2Is [0,1 ]]Random number, P idFor the individual best solution, PgdIs a globally optimal solution.
The reasonable design of omega can improve the searching efficiency of the particle swarm algorithm and avoid falling into local optimum. Therefore, the present invention introduces a re-excitation mechanism of reinforcement learning, as shown in fig. 5, to adaptively adjust ω so that it automatically adjusts ω according to the environmental information.
As shown in fig. 5, the underwater robot observes the current environment and makes corresponding action, then the environment gives feedback to the underwater robot, and the underwater robot improves the next action, and the operation is repeated until making the optimal action. The process of automatically adjusting ω is as follows:
dividing each link line of a shortest path graph in a suboptimal path of the MAKLINK two-dimensional map into ten equal division points, taking the ten equal division points as sampling points, calculating the distance L between the sampling points, and searching the shortest distance L between all the sampling pointsminAnd LmaxThe motion cost value cost from node i to node j is calculated as equation (4):
Figure BDA0003057688860000061
J＝∑cost (5)
j in the equation (5) is the sum of the cost value of the selected path and the cost of the node to be selected. When J is larger, the particle position is worse and ω needs to be enhanced to improve the global optimization capability. When J is smaller, omega needs to be reduced, and the local optimizing capacity of the particle needs to be improved. Updating the inertial weight ω of the particle using J as the re-excitation signal, the update formula is (6):
ω＝|exp(KJ)-1| (6)
Wherein K is a non-negative number.
And (4.2) when the path is planned by adopting the improved particle swarm algorithm, considering the complex marine environment, the AUV is greatly influenced by ocean current during path planning, and the state of the particles is evaluated by combining a speed synthesis method. As shown in FIG. 6, VcIs the ocean current velocity vector, VAIs the velocity vector of AUV, VEIs the desired velocity vector. Alpha is alpha1Angle of ocean current, α2Angle of line connecting AUV to target, alpha3Is AUVThe angle of (c). VEIs pointed by the AUV towards the target point. VcnFor ocean currents at VEVertical component of (A) VanFor robot at VEThe vertical component of (c). When V iscnAnd VanWhen the side effects of ocean currents are counteracted, the side effects of the ocean currents are counteracted; under the condition that the speed of the ocean current is known, the speed and the direction of the AUV are adjusted, so that the ocean current can be fully utilized for path planning. At this time AUV desired speed VEFor equation (7), the desired angle is equation (8):
Figure BDA0003057688860000071
α3＝arcsin(VC*sin(α2-α1)/VA)+α2 (8)
the evaluation function f (xi) comprehensively judges whether the position of the current particle is good or bad, and the particle updates X in the formula (2) according to the value of the evaluation functionidAnd Vid. Considering the effect of ocean currents on AUV, an evaluation function of the binding velocity synthesis method is presented herein.
Figure BDA0003057688860000072
Figure BDA0003057688860000073
Figure BDA0003057688860000074
Wherein E, A, B, C are a non-negative constant, P i,Pi+1Is the shortest path point as in fig. 7. In equation (9), i represents the number of shortest path points obtained by Dijkstra, and D represents a D-segment route. I.C. AiIs (10), HiIs (11), ViIs the velocity of the ith particle, αiIs the angle of the ith particle, VexpectedAnd alphaexpectedThe desired speed and angle obtained by (7) and (8). Among the shortest paths obtained by Dijkstra, the number of routes in the shortest path is regarded as D.
The AUV has a motion space formed by convex polygonal obstacles and a path Pi1Pi2The upper slide does not intersect with the obstacle, so P is determined by improving the particle swarm algorithmi1Pi2New shortest path point P ofiAnd then connecting the determined new path points on each link line to finally obtain an optimal planning path, wherein a black curve shown in fig. 8 is a final particle swarm optimization path.
And (5) sampling a plurality of equidistant points at the broken line on the path generated by the improved particle swarm optimization in order to reduce the broken points, and smoothing the path by using a method based on a Bezier curve, because the finally generated shortest path has a plurality of broken points.
The complexity of the robot control is greatly increased because a set of unsmooth and oscillatory paths will be created when planning the paths of a swarm of robots. Therefore, a Bezier curve optimization method is integrated in the algorithm to smooth the path so as to reduce the number of corners of the robot. In the embodiment of the invention, a 3-order Bezier curve optimization path is adopted, and the formula is (12):
P(t)＝P0(1-t)3+3P1(1-t)2t+3P2(1-t)t2+P3t3 (12)
Wherein t is a random number between 0 and 1, P (t) is a coordinate value under t, 4 points which are equidistant at a broken line on a path generated by the improved particle swarm optimization are sampled, and the 4 points are P0,P1,P2,P3From equation (12), a series of points P (t) is obtained, and connecting these points yields a smooth curve. The black curve shown in fig. 8 is the particle swarm optimization path after smoothing.

Claims (8)

1. An energy-saving underwater path planning method based on a re-excitation mechanism particle swarm algorithm is characterized in that: the method comprises the following steps:
step (1), a convex polygonal barrier is established according to a Graham algorithm, and a set distance is expanded outwards to establish a danger zone;
step (2), constructing a MAKLINK two-dimensional map based on convex polygon obstacles, and constructing a undirected network map on the basis of the MAKLINK two-dimensional map;
step (3), performing suboptimal path optimization on the constructed two-dimensional map by using a Dijkstra algorithm to obtain a suboptimal path based on the MAKLINK map;
step (4), then, optimizing the suboptimal path by utilizing an improved particle swarm algorithm to obtain a shortest path; the specific process is as follows: simulating the activities of birds by adopting a particle swarm algorithm, randomly searching for an optimal position, regarding each particle as a search individual of a D-dimensional search space, and regarding the speed V of each particle in a two-dimensional map idAnd position XidRepresenting the velocity V and the position X, V of the ith particle in a two-dimensional map in D-dimensional spaceidAnd position XidAfter the particle swarm algorithm is adopted for updating, the formula (2) and the formula (3) are shown:
Vid＝ωVid+C1r1(Rid-Xid)+C2r2(Pgd-Xid) (2)
Xid＝Xid+Vid (3)
wherein, omega is an inertia factor, the value of omega is nonnegative, the global optimizing ability is strong when the value is larger, and the local optimizing ability is strong when the value is smaller; c1And C2Is a learning factor, is a constant; r is a radical of hydrogen1And r2Is [0,1 ]]Random number, PidFor an individual optimal solution, PgdIs a global optimal solution;
adjusting ω according to the environment information:
firstly, dividing each link line of a shortest path graph in an MAKLINK two-dimensional map into ten equally-divided points as sampling points, calculating the distance L between the sampling points, and searching the shortest distance L between all the sampling pointsminAnd LmaxThe cost calculation formulas from node i to node j are (4) and (5):
Figure FDA0003588881210000011
J＝∑cost (5)
j is the cost sum of the selected path cost value and the node to be selected; updating the inertial weight ω of the particle using J as the re-excitation signal, the update formula is (6):
ω＝|exp(KJ)-1 (6)
wherein K is a non-negative constant;
and (5) sampling a plurality of equidistant points on the path at the broken line generated by the improved particle swarm optimization, and reducing the turning point of the shortest path by using a Bezier curve so as to smooth the path.
2. The energy-saving underwater path planning method based on the re-excitation mechanism particle swarm algorithm according to claim 1, wherein: the step (1) comprises the following steps:
step (1.1), constructing a convex polygonal barrier by utilizing a Graham algorithm, and searching points on a convex hull to generate the convex polygonal barrier according to polar coordinate sequencing through a random number of points on a two-dimensional coordinate system;
step (1.2), on the basis of generating the convex-edge-shaped barrier, manually expanding the set distance outwards, wherein the distance value is dmin。
3. The energy-saving underwater path planning method based on the re-excitation mechanism particle swarm algorithm according to claim 2, wherein: in the step (1.2), when the AUV carries out underwater path planning, the distance d between the AUV and the obstacle meets the following constraint:
d≥dmin (1) 。
4. the energy-saving underwater path planning method based on the re-excitation mechanism particle swarm algorithm according to claim 1, wherein: the specific steps of constructing the MAKLINK two-dimensional map in the step (2) are as follows:
step (2.1), connecting a certain vertex of the convex polygonal barrier constructed in the step (1) with other vertexes or space boundary points to form a link line, setting a node on each link line, determining that the link line between the nodes does not pass through the barrier as a feasible line segment according to the actual situation of the MAKLINK graph, and forming a feasible line segment set;
And (2.2) selecting the midpoint of the link line as a node, and connecting the nodes with each other to form a undirected network graph.
5. The energy-saving underwater path planning method based on the re-excitation mechanism particle swarm algorithm according to claim 1, wherein: the suboptimal path optimization of the step (3) comprises the following steps:
on the basis of generating the undirected network map in the step (2), searching the shortest link line from the target to the starting point on the MAKLINK two-dimensional map according to the Dijkstra algorithm to form a suboptimal path, and obtaining the shortest path point, thereby obtaining the link line where each path point is located.
6. The energy-saving underwater path planning method based on the re-excitation mechanism particle swarm algorithm according to claim 5, wherein: and determining new shortest path points on the link lines by improving a particle swarm algorithm, and then connecting the determined new path points on each link line to obtain an optimal planning path.
7. The energy-saving underwater path planning method based on the re-excitation mechanism particle swarm algorithm according to claim 1, wherein: the particle state is evaluated by adopting a speed synthesis method, and when the ocean current speed is known, the path planning is carried out by adjusting the speed and the direction of an AUV (autonomous Underwater vehicle), wherein the AUV has a desired speed V EAUV desired Angle α of equation (7)3Is represented by formula (8):
Figure FDA0003588881210000021
α3＝arcsin(VC*sin(α2-α1)/VA)+α2 (8)
wherein, VcIs an ocean current velocity vector, VAIs the velocity vector of AUV, α1Angle of ocean current, α2Angle of line connecting AUV to target, alpha3Is the angle of the AUV.
8. The energy-saving underwater path planning method based on the re-excitation mechanism particle swarm algorithm according to claim 7, wherein: judging the current position of the particle according to an evaluation function, and updating X in the formula (2) by the particle according to the value of the evaluation functionidAnd VidConsidering the influence of ocean currents on the AUV, the evaluation function is formula (9):
Figure FDA0003588881210000031
Figure FDA0003588881210000032
Figure FDA0003588881210000033
wherein E, A, B, C are a non-negative constant; p0,P1,P2…,Pi,Pi+1,PDThe shortest path point is obtained by Dijkstra algorithm, and D represents the number of routes in the shortest path; viIs the velocity of the ith particle, αiIs the angle of the ith particle, VexpectedAnd alphaexpectedThe desired speed and angle are obtained by equations (7) and (8).
CN202110504234.XA 2021-05-10 2021-05-10 Energy-saving underwater path planning method based on re-excitation mechanism particle swarm algorithm Active CN113190026B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110504234.XA CN113190026B (en) 2021-05-10 2021-05-10 Energy-saving underwater path planning method based on re-excitation mechanism particle swarm algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110504234.XA CN113190026B (en) 2021-05-10 2021-05-10 Energy-saving underwater path planning method based on re-excitation mechanism particle swarm algorithm

Publications (2)

Publication Number Publication Date
CN113190026A CN113190026A (en) 2021-07-30
CN113190026B true CN113190026B (en) 2022-06-28

Family

ID=76988577

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110504234.XA Active CN113190026B (en) 2021-05-10 2021-05-10 Energy-saving underwater path planning method based on re-excitation mechanism particle swarm algorithm

Country Status (1)

Country Link
CN (1) CN113190026B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113485371B (en) * 2021-08-11 2023-04-25 河海大学 Underwater multi-AUV path planning method based on improved sparrow search algorithm
CN113934218B (en) * 2021-11-16 2022-03-25 杭州云象商用机器有限公司 Cleaning robot path planning method, device, equipment and storage medium

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109685274A (en) * 2018-12-26 2019-04-26 绥化学院 The method of high-speed rail path planning design based on maklink figure Multi-node link
CN112161627A (en) * 2020-09-23 2021-01-01 同济大学 Intelligent path planning method for fire-fighting robot
CN112306097A (en) * 2020-10-29 2021-02-02 杭州电子科技大学 Novel unmanned aerial vehicle path planning method
CN112650246B (en) * 2020-12-23 2022-12-09 武汉理工大学 Ship autonomous navigation method and device

Also Published As

Publication number Publication date
CN113190026A (en) 2021-07-30

Similar Documents

Publication Publication Date Title
CN110083165B (en) Path planning method of robot in complex narrow environment
CN113190026B (en) Energy-saving underwater path planning method based on re-excitation mechanism particle swarm algorithm
Rao et al. Large-scale path planning for underwater gliders in ocean currents
CN108564202B (en) Unmanned ship route optimization method based on environment forecast information
Cao et al. Target search control of AUV in underwater environment with deep reinforcement learning
CN111399506A (en) Global-local hybrid unmanned ship path planning method based on dynamic constraints
CN110703752B (en) Unmanned ship double-layer path planning method based on immune heredity-artificial potential field method
CN109144102A (en) A kind of Path Planning for UAV based on improvement bat algorithm
CN105629974A (en) Robot path planning method and system based on improved artificial potential field method
Bai et al. UAV Path Planning Based on Improved A ∗ and DWA Algorithms
CN109542117A (en) Based on the submarine navigation device Rolling Planning algorithm for improving RRT
CN108388250B (en) Water surface unmanned ship path planning method based on self-adaptive cuckoo search algorithm
Guo et al. Research progress of path planning methods for autonomous underwater vehicle
CN111338350A (en) Unmanned ship path planning method and system based on greedy mechanism particle swarm algorithm
CN110196598A (en) A kind of dynamic collision prevention method of unmanned boat
CN113189984B (en) Unmanned ship path planning method based on improved artificial potential field method
Lan et al. Improved RRT algorithms to solve path planning of multi-glider in time-varying ocean currents
Zeng et al. Optimal path planning based on annular space decomposition for AUVs operating in a variable environment
CN114077256A (en) Overwater unmanned ship path planning method
Yan et al. A novel 3-D bio-inspired neural network model for the path planning of an AUV in underwater environments
CN108445894A (en) A kind of secondary paths planning method considering unmanned boat movenent performance
Sànchez et al. A darwinian swarm robotics strategy applied to underwater exploration
CN110320907B (en) Double-layer collision avoidance method for unmanned surface vessel based on improved ant colony algorithm and elliptic collision cone deduction model
Zhang et al. Global path planning for mobile robot based on A∗ algorithm and genetic algorithm
Zhang et al. Research on Path Planning Method of Unmanned Boat Based on Improved Artificial Potential Field Method

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