CN109799820A - Unmanned ship local paths planning method based on the random road sign figure method of comparison expression - Google Patents
Unmanned ship local paths planning method based on the random road sign figure method of comparison expression Download PDFInfo
- Publication number
- CN109799820A CN109799820A CN201910059813.0A CN201910059813A CN109799820A CN 109799820 A CN109799820 A CN 109799820A CN 201910059813 A CN201910059813 A CN 201910059813A CN 109799820 A CN109799820 A CN 109799820A
- Authority
- CN
- China
- Prior art keywords
- paths
- path
- pop
- ship
- point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Navigation (AREA)
- Traffic Control Systems (AREA)
Abstract
The present invention provides a kind of unmanned ship local paths planning method based on the random road sign figure method of comparison expression, comprising: S1 obtains environmental information；S2 obtains starting point, target point, determines road sign point number K and maximum number of iterations G；S3, initial ship's navigation route scheme set pop is obtained using multiple target collision checking method0, S4, foundation Revised genetic algorithum are to pop0It is iterated processing, the set of paths pop after obtaining iterationm；S5, the set of paths pop after iteration is obtainedmThe fitness value of middle each route scheme；S6, according to fitness value, choose the most short ship's navigation route scheme for output.The above method solves the problems such as slow prior art storing process difficulty data, ship reaction speed, consuming more multiple-energy-source.
Description
Technical field
The present invention relates to Path Planning Technique, especially a kind of unmanned ship part based on the random road sign figure method of comparison expression
Paths planning method.
Background technique
In mobile robot field, random walk figure method can effectively solve the problem that the path in higher dimensional space and Complex Constraints is advised
The problem of drawing, but the random walk figure method that uses of mobile robot carries out feasible path search frequently with A* algorithm, and A* algorithm
Search feasible path can generate a large amount of process data.For being applied to for Intelligent unattended ship, need to comprehensively consider ocean
Complexity, Ship Maneuverability and the energy consumption of environment, the random walk figure method that mobile robot uses lack practical marine use
Feasibility.
The random walk figure method that mobile robot uses is described as follows below.
Random walk figure (PRM) method finds the path in given map between two o'clock, and PRM carries out the substantially former of path planning
Reason:
Assuming that robot running environment is that the two-dimensional closed plane space containing several barriers is denoted as C.Robot
It can be known as free space with the free-moving region i.e. region at removing barrier and its edge and be denoted as Cfree, by mobile robot
It regards particle in C as, and does not consider that kinematics constrain.
The basic thought of random walk figure method is the freely sky that robot system operation is indicated with a random network
Between Cfree, this random network is the probability map established by robot system.Probability map is a Undirected Circulant Graphs,
It is indicated with G=(Ga, Ge).Wherein Ga is node collection, and element therein is in free planning space CfreeIn for robot it is random
The position shape of selection, the element in the collection Ge of side then indicate the line between node, represent the local feasible path between the shape of position.Then
A feasible path is searched for robot system using searching algorithm in this random network.The algorithm can pass through 2
Stage completes: off-line learning stage and online query stage.
1) the off-line learning stage.The task that the stage mainly completes is in CfreeIn establish a probability map for robot
G=(Ga, Ge).
First in a random way in CfreeThe middle a large amount of robot location's point of stochastical sampling constitutes the node in map G
Collect Ga；Secondly, sector planning device is used to find their neighbor node for each node in Ga, and connection is established for them,
To constitute the side collection Ge in map.The road sign figure that the study stage is set up is planned for online query and is used.
2) the online query stage.The task that the stage mainly completes is the item in given initial position s and target point g
Under part, according to the road sign figure that off-line phase constructs, the feasible path of a connection s and g is searched out.Firstly, utilizing sector planning
S and g are connected to 2 nodes nearest with its distance in road sign figure by device respectivelyWithThen, using heuristic search algorithm
A connection is searched out from road sign figureWithFeasible path.
Heuristic search algorithm is frequently with A* algorithm.A* algorithm is that fan-shaped Dissatisfied rope is started to spread out from starting point, passes through calculating
The cost estimated value of present node and its 8 direction neighbors, it is to answer that the path for finding out minimum cost value, which is optimal path,
With wide efficient method.
The problem of above method exists in practical applications is as follows:
A1) side one that connecting node can not be completely secured in the sector planning device that above-mentioned random walk figure method uses is positioned at
Free space；
A2) above-mentioned random walk figure is often used A* algorithm to search for feasible path, can generate a large amount of process data, data
Amount of storage is big；
A3) for being applied to for Intelligent unattended ship, need to comprehensively consider complexity, the ship's manoeuverability of marine environment
Energy and energy consumption use the random walk figure method of A* algorithm search feasible path to lack the feasibility of practical marine use.
For this purpose, need it is a kind of can be applied to Intelligent unattended ship domain and using the random road sign figure method of comparison expression nothing
People's ship local paths planning method.
Summary of the invention
For the problems of the prior art, the present invention provides a kind of unmanned ship office based on the random road sign figure method of comparison expression
Portion's paths planning method.
In a first aspect, the present invention provides a kind of unmanned ship local paths planning side based on the random road sign figure method of comparison expression
Method, comprising:
S1, the environment navigated by water according to predetermined unmanned ship, obtain environmental information；
S2, starting point when being navigated by water according to the environmental information and unmanned ship, target point, determine road sign point number K
Maximum number of iterations G required for optimizing with local path avoidance；
The position of S3, the random each road sign point for generating road sign point number K, starting point, target point and road sign point are connected two-by-two
It connects, initial ship's navigation route scheme set pop is obtained using multiple target collision checking method0, the initial ship's navigation road
Diameter scheme set pop0It include: the path that a plurality of unmanned ship can navigate by water；Each path is in the map of environmental information
It is internal and belong to non-barrier region；
S4, foundation Revised genetic algorithum and the maximum number of iterations G are to the initial ship's navigation route scheme collection
Close pop0It is iterated processing, the set of paths pop after obtaining iterationm；
S5, the set of paths pop after iteration is obtainedmThe fitness value of middle each route scheme；
S6, according to the set of paths pop after iterationmThe fitness value of middle each route scheme is chosen for output
Most short ship's navigation route scheme.
Optionally, step S3 includes:
S31, according to the environmental information, generate K two-dimensional coordinate at random, each two-dimensional coordinate is as a road sign point
Position；
S32, to judge whether each road sign point position is located at using multiple target collision detection method non-in the environmental information
Inside barrier region and map, if so, retaining, otherwise give up the road sign point position；
S33, combined starting point, target point and an any random road sign point based on multiple target collision detection method to be formed it is a plurality of
Path；And
Whether the line in each paths after judging combination between any two points is located at non-barrier region, and whether
Belong to inside map；
If being, using the path of judgement as initial ship's navigation route scheme set pop0In element.
Optionally, whether the line in each paths judged after combination in step S33 between any two points is located at
Non- barrier region, and whether belong to inside map, comprising:
The angle for obtaining line and positive direction of the x-axis, determines an exploration range according to the angle,
Judge whether determining exploration range includes whether barrier region or the exploration range are in outside map
Region；
If not including barrier region and not also in map perimeter, the paths are as initial ship's navigation road
Diameter scheme set pop0In element.
Optionally, step S5 includes:
According to formula (1) to acquisite approachs set popmIn spacing and D in each paths between inflection point；
Wherein, n is each paths inflection point number, and the path inflection point includes: starting point, target point and road sign point, xj
And yjFor the coordinate of path inflection point j, xj+1And yj+1 is the coordinate with path inflection point j adjacent comers j+1；
Using the inverse of spacing and D as the fitness value fit of the paths；
Optionally, step S4 includes:
S41, it is directed to initial ship's navigation route scheme set pop0In pending crossover operation two paths, generate one
A random chance；The same length of two paths of the pending crossover operation；
Judge whether the random chance generated is greater than preset crossover probability threshold value；
If more than two paths of pending crossover operation are then carried out crossover operation, obtain two new paths；
If being less than, crossover operation is not carried out to two paths of pending crossover operation；
S42, the initial ship's navigation route scheme set pop is traversed according to the crossover operation mode of step S410Middle institute
After some paths, new path will be obtained and without the original route combination of crossover operation, obtain first path set pop1；
S43, it is directed to first path set pop1In pending mutation operation each paths, it is random to generate a variation
Probability；
Judge whether the variation random chance generated is greater than preset mutation probability threshold value；
If more than the path of pending mutation operation is then carried out mutation operation, obtains new path；
If being less than, mutation operation is not carried out to the path of pending mutation operation；
S44, the first path set pop is traversed according to the mutation operation mode of step S431In all path it
Afterwards, new path will be obtained and without the original route combination of mutation operation, obtain the second set of paths pop2；
S45, it is directed to the second set of paths pop2, obtain the second set of paths pop2In each paths fitness value；
Ratio selection operation is carried out according to the fitness value in all paths using Propertional model, by ratio selection operation
Path afterwards is as third set of paths pop3；
S46, judge whether current iteration number i is equal to maximum number of iterations G；
If S47, being less than, from third set of paths pop3Start to repeat the above steps the process of S41 to step S46；Directly
It is equal to maximum number of iterations G to the number of iterations i；
The set of paths pop that will be obtained when being equal to maximum number of iterations GmAs the set of paths pop after iterationm。
Optionally, two paths by pending crossover operation in step S41 carry out crossover operation, obtain new two
Path, comprising:
Two paths essentially identical to path length randomly choose cut-out point on the two route segment and carry out breakpoint section
It is disconnected, if cut-out point position is essentially identical, two paths truncation parts are exchanged with each other；
If cut-out point position is different, crossover operation is not carried out to two paths of pending crossover operation.
Optionally, the path by pending mutation operation in S43 carries out mutation operation, obtains new path, comprising:
Deflecting operation is carried out to the route segment direction being linked to be between two road sign points of arbitrary neighborhood in each paths.
Optionally, the step S1 includes:
Environmental information is read from the BMP picture file of checkout area sea chart polar plot.
The invention has the benefit that
Method of the invention can generate safe and stable, Ship's Optimum Route in unmanned ship navigation, and which solve stored
Journey difficulty data, ship reaction speed are slow, expend the problems such as more multiple-energy-source.Method of the invention, which can reach, not to be needed to store
Number of passes according to, reduce optimization and calculate time equivalence fruit, increase the maneuverability of Intelligent unattended ship, random road sign figure method answered
For Intelligent unattended ship domain, foundation is provided to realizing that Intelligent unattended ship intelligently navigates by water.
Detailed description of the invention
Fig. 1 is the unmanned ship local paths planning based on the random road sign figure method of comparison expression that one embodiment of the invention provides
The schematic diagram of method；
Fig. 2 is the schematic diagram for the simulation unmanned ship navigation that one embodiment of the invention provides；
Fig. 3 be another embodiment of the present invention provides based on the random road sign figure method of comparison expression unmanned ship local path rule
The schematic diagram for the method for drawing.
Specific embodiment
In order to preferably explain the present invention, in order to understand, with reference to the accompanying drawing, by specific embodiment, to this hair
It is bright to be described in detail.
Random walk figure method used in mobile robot is applied to the path planning of unmanned ship by insider, is deposited
It is as follows in problem:
B1) node that the sector planning device that traditional random walk figure method uses is cooked up can guarantee its in free space,
However, not can avoid its line one when connecting node is positioned at free space；
It B2 is to select cost estimated value most every time when) A* algorithm carries out the search for the node that will be walked in next step
Small node, therefore the optimal path found.But A* algorithm will extend the whole of present node Just because of this every time after
After node, their cost estimated value is calculated with heuristic function, then selects the smallest node of cost estimated value as next
The node that step is walked.In this process, OPEN table needs to save a large amount of nodal information；
B3) when being applied to Intelligent unattended ship, used sea chart is mostly more huge, and path is more complicated, and path is searched
Rope process may then calculate thousands of node, and calculation amount is very huge.Therefore, traditional random walk figure method searches for one
Path needs the time longer, this means that Intelligent unattended ship reaction speed can decline.
Defect based on the prior art, method problem to be solved of the invention are as follows:
First, this method is quasi- to be applied to checkout area shiphoard measurement, has practical application.
Second, being filtered out a plurality of by multiple target collision detection method includes starting point and target point and navigation path whole position
In area of feasible solutions navigation path and generate initial navigation path scheme set, be different from traditional random walk figure method and only saved
Point collision detection.
The multiple target collision detection method that the present invention uses is that collision detection is all carried out to node and node line, solves tradition
The sector planning device that random walk figure method uses not can avoid the line one between the node that it is cooked up and be positioned at free space
The problem of.
Third optimizes each path in the navigation path scheme set of generation using genetic algorithm, final relatively more each
The path length in path, selects shortest path, this, which is different from, carries out the such heuristic search of feasible path search using A* algorithm
Traditional random walk figure method of formula, the random road sign figure method of comparison expression can be to avoid the generation computed repeatedly with excessive process data；
4th, since input parameter and output result are all real number, algorithm coding mode is real coding, avoids making
The complex process of code conversion is needed when with binary coding；
5th, what path planning solved is that path is optimal or the most short problem in path, and the present invention comprehensively considers that path is most short, rule
The factors such as obstacle avoidance object, ship domain and feasibility have formulated the fitness of distance sum between reflection Intelligent unattended ship track inflection point
Function compares according to the size of the fitness function value, obtain suboptimum navigation path scheme set for compare select it is most short
Path is to solve path planning problem.
Embodiment one
In conjunction with shown in Fig. 1 to Fig. 3, the unmanned ship local path rule based on the random road sign figure method of comparison expression of the present embodiment
The method of drawing comprises the following steps that
S1, the environment navigated by water according to predetermined unmanned ship, obtain environmental information.
For example, environmental information can be read from the BMP picture file of checkout area sea chart polar plot.
S2, starting point when being navigated by water according to the environmental information and unmanned ship, target point, determine road sign point number K
Maximum number of iterations G required for optimizing with local path avoidance.
The road sign point number K of the present embodiment is artificial determining according to environmental information, starting point and target point.Greatest iteration
Number G is also artificial empirically determining.
The position of S3, the random each road sign point for generating road sign point number K, starting point, target point and road sign point are connected two-by-two
It connects, initial ship's navigation route scheme set pop is obtained using multiple target collision checking method0, the initial ship's navigation road
Diameter scheme set pop0It include: the path that a plurality of unmanned ship can navigate by water；Each path is in the map of environmental information
It is internal and belong to non-barrier region.
To better understand step S3, step S3 is illustrated below:
S31, according to the environmental information, generate K two-dimensional coordinate at random, each two-dimensional coordinate is as a road sign point
Position；
S32, to judge whether each road sign point position is located at using multiple target collision detection method non-in the environmental information
Inside barrier region and map, if so, retaining, otherwise give up the road sign point position；
S33, combined starting point, target point and an any random road sign point based on multiple target collision detection method to be formed it is a plurality of
Path；And
Whether the line in each paths after judging combination between any two points is located at non-barrier region, and whether
Belong to inside map；
If being, using the path of judgement as initial ship's navigation route scheme set pop0In element.
For example, the angle of the line and positive direction of the x-axis in each paths after combination between any two points is obtained, according to
The angle determines an exploration range, judges whether determining exploration range includes barrier region or the exploration model
It whether encloses in map perimeter；
If not including barrier region and not also in map perimeter, the paths are as initial ship's navigation road
Diameter scheme set pop0In element.
The positive direction of the x-axis of the present embodiment is defined according to environmental information in advance.
S4, foundation Revised genetic algorithum and the maximum number of iterations G are to the initial ship's navigation route scheme collection
Close pop0It is iterated processing, the set of paths pop after obtaining iterationm；
S5, the set of paths pop after iteration is obtainedmThe fitness value of middle each route scheme.
For example, according to formula (1) to acquisite approachs set popmIn spacing and D in each paths between inflection point；
Wherein, n is each paths inflection point number, and the road sign inflection point includes: starting point, target point and road sign point, xj
And yjFor the coordinate of path inflection point j, xj+1And yj+1 is the coordinate with path inflection point j adjacent comers j+1；
Using the inverse of spacing and D as the fitness value fit of the paths；
S6, according to the set of paths pop after iterationmThe fitness value of middle each route scheme is chosen for output
Most short ship's navigation route scheme.
In addition, being illustrated using following sub-steps to step S4 to better understand the process of above-mentioned steps S4
Include: as follows
S41, it is directed to initial ship's navigation route scheme set pop0In pending crossover operation two paths, generate one
A random chance；The same length of two paths of the pending crossover operation；
Judge whether the random chance generated is greater than preset crossover probability threshold value；
If more than two paths of pending crossover operation are then carried out crossover operation, obtain two new paths；
If being less than, crossover operation is not carried out to two paths of pending crossover operation；
S42, the initial ship's navigation route scheme set pop is traversed according to the crossover operation mode of step S410Middle institute
After some paths, new path will be obtained and without the original route combination of crossover operation, obtain first path set pop1；
S43, it is directed to first path set pop1In pending mutation operation each paths, it is random to generate a variation
Probability；
Judge whether the variation random chance generated is greater than preset mutation probability threshold value；
If more than the path of pending mutation operation is then carried out mutation operation, obtains new path；
If being less than, mutation operation is not carried out to the path of pending mutation operation；
S44, the first path set pop is traversed according to the mutation operation mode of step S431In all path it
Afterwards, new path will be obtained and without the original route combination of mutation operation, obtain the second set of paths pop2；
S45, it is directed to the second set of paths pop2, obtain the second set of paths pop2In each paths fitness value；
Ratio selection operation is carried out according to the fitness value in all paths using Propertional model, by ratio selection operation
Path afterwards is as third set of paths pop3；
S46, judge whether current iteration number i is equal to maximum number of iterations G；
If S47, being less than, from third set of paths pop3Start to repeat the above steps the process of S41 to step S46；Directly
It is equal to maximum number of iterations G to the number of iterations i；
The set of paths pop that will be obtained when being equal to maximum number of iterations GmAs the set of paths pop after iterationm。
Further, two paths by pending crossover operation in step S41 carry out crossover operation, obtain new two
Paths, comprising:
Two paths essentially identical to path length randomly choose cut-out point on the two route segment and carry out breakpoint section
It is disconnected, if cut-out point position is essentially identical, two paths truncation parts are exchanged with each other；
If cut-out point position is different, crossover operation is not carried out to two paths of pending crossover operation.
Further, the path by pending mutation operation in S43 carries out mutation operation, obtains new path, comprising:
Deflecting operation is carried out to the route segment direction being linked to be between two road sign points of arbitrary neighborhood in each paths.
The method of the present embodiment can generate safe and stable, Ship's Optimum Route in unmanned ship navigation, and which solve storages
Process data is difficult, ship reaction speed is slow, expends the problems such as more multiple-energy-source.Method of the invention, which can reach, not to be needed to store
Process data reduces optimization and calculates time equivalence fruit, increases the maneuverability of Intelligent unattended ship, so that random road sign figure method can be with
Applied to Intelligent unattended ship domain, foundation is provided to realizing that Intelligent unattended ship intelligently navigates by water.
Embodiment two
In conjunction with shown in Fig. 2 and Fig. 3, the method for the present embodiment mainly includes the steps that following:
1) environmental information is read from the BMP picture file of checkout area sea chart polar plot；
2) starting point [X1, Y1] is set, target point [X2, Y2], calculates required road sign point number K and local path
Avoidance optimizes maximum number of iterations G；
3) K road sign point position is generated at random, starting point, target point and road sign point are connected two-by-two, records line non-
Inside barrier region and map and include starting point and target point set of paths, obtain what a plurality of ship's navigation path was formed
Navigation path network, and then obtain the initial ship's navigation route scheme set pop that will be used for navigation path selected section0。
In the present embodiment, in each of initial ship's navigation route scheme set route scheme corresponding network figure
One paths, it may be said that the image of initial ship's navigation route scheme set shows it is navigation path network.
The concrete operations of step 3) pass through following sub-step 3a) it is described as follows to 3d):
K two-dimensional coordinate 3a) is generated at random in entire simulated environment map, as random road sign point position；
3b) judge whether random road sign point position is located inside non-barrier region and map, if so, it is random to retain this
Road sign point；If it is not, then giving up the random road sign point；
3c) judge whether the line after starting point, target point and all random road sign points connect two-by-two is located at non-barrier
Inside region and map, if so, the line that record can directly or indirectly be connected with starting point, i.e., by starting point
All feasible paths；If it is not, then giving up the line；
Wherein, judging the method used when line is whether inside non-barrier region and map is: according to line and x-axis
The angle calcu-lation of positive direction goes out an exploration range, if this, which is explored, does not include barrier region or map perimeter in range
Then judge that this line is located inside non-barrier region and map, otherwise judges that this line is not located at non-barrier region and map
It is internal.
The mulitpath comprising target point 3d) is filtered out from all feasible paths by starting point, these paths
Then form the initial ship's navigation route scheme set pop for navigation path selected section0。
4) initial ship's navigation route scheme set pop is evaluated0In in each ship's navigation route scheme between inflection point away from
From with obtain the fitness value of each path；
The distance between inflection point and D in ship's navigation route scheme are evaluated according to formula (1)
Wherein, n is inflection point number in ship's navigation route scheme, xjAnd yjFor the coordinate of path inflection point j, xj+1And yj+1For
With the coordinate of path inflection point j adjacent comers j+1.
According to formula (2) using the inverse of the distance between inflection point in ship's navigation route scheme and D as fitness value
fit；
5) it is directed to initial path set pop0(initial ship's navigation route scheme set i.e. above-mentioned) implements the following,
Until generating updated set of paths popm；
For the concrete operations of step 5), using following step 5a) it is described as follows to step 5d):
5a) it is directed to the initial path set pop0In path select two-by-two carry out crossover operation, generate first path
Set pop1, it specifically includes:
Judge whether the corresponding random chance of two paths of pending crossover operation is greater than crossover probability threshold value to judge.
The quantity of random chance is initial path set pop0In half, for initial path set pop0In two paths generate
One random chance；The value of random chance is between 0~1.The crossover probability threshold value of the present embodiment is given value, value model
Enclose 0.6~0.9.
If random chance is greater than crossover probability, the crossover operation of two paths is carried out: path identical to path length
Scheme combination of two randomly chooses cut-out point on the two route segment and carries out breakpoint truncation, if cut-out point position is identical, by two
Kind scheme truncation part is exchanged with each other；
If although random chance is less than crossover probability or random chance and is greater than crossover probability but cut-out point position not phase
Together, path length is not identical, then retains original ship's navigation set of paths and then carry out step 5b)；
Traverse initial path set pop0In all paths carry out crossover operation after, generate first path set pop1。
5b) it is directed to first path set pop1Each path, selection carry out mutation operation, generate the second set of paths
pop2, it specifically includes:
Judge whether the corresponding variation random chance of each paths of pending mutation operation is greater than mutation probability threshold value.
The quantity of the variation random chance of the present embodiment is first path set pop1In number of paths, for first path set
pop1In one variation random chance of each coordinates measurement；The value for the random chance that makes a variation is between 0~1.This implementation
The mutation probability threshold value of example is given value, value range 0.001~0.1.
If random chance is greater than mutation probability, the path for treating mutation operation carries out mutation operation: in route scheme
Path direction carries out deflecting；Otherwise, retain original ship's navigation set of paths and then carry out step 5c)；
Traverse first path set pop1In all paths carry out mutation operation after, generate the second set of paths pop2。
5c) it is directed to the second set of paths pop2, the second set of paths pop is obtained according to formula (1)2In each paths turn
Spacing and D between point carry out ratio selection operation according to the fitness value fit that formula (2) calculate each path, generate third
Set of paths pop3Then step 5d is carried out).
The ratio selection operation (or being Propertional model/wheel disc bet method) of the present embodiment can be regarded as: individual quilt
The probability chosen is directly proportional to its fitness value size, wherein individual is ship's navigation route scheme.
5d) judge whether current iteration number i reaches preset local path avoidance optimization maximum number of iterations G, if i≤
G, then current iteration number i increases preset value 1, and third set of paths pop3For the updated set of paths of this iteration
popm, go to step 5a), repeat step 5a) to the treatment process of step 5d)；Step 6) is gone to if reaching.
6) it is directed to set of paths popm, according to formula (1) to acquisite approachs set popmIn between each paths inflection point between
Away from D or fitness value, and according to the size of acquired spacing and D or fitness value from multiple groups ship's navigation route scheme
It selects most short ship's navigation route scheme and exports.
The method of the present embodiment has the advantage that
(1) it is suitable for Intelligent unattended ship kinetic characteristic, to realize that it is new that the local obstacle-avoiding route planning of ship provides
Solution；
(2) it is to rely on coastal checkout area, has practical application；
(3) feasible navigation path is screened by multiple target collision detection method, it is ensured that the safety of Intelligent unattended ship's navigation；
(4) it is different from and traditional random walk figure method that feasible path searches for such heuristic search formula is carried out using A* algorithm,
The random road sign figure method of comparison expression can improve to avoid the generation computed repeatedly with excessive process data, effectively quickening calculating speed
Intelligent unattended ship avoidance efficiency, may be implemented Intelligent unattended mooring stability and efficiently navigates by water.
Above-mentioned each embodiment can be cross-referenced, and the present embodiment is not defined each embodiment.
Finally, it should be noted that above-described embodiments are merely to illustrate the technical scheme, rather than to it
Limitation；Although the present invention is described in detail referring to the foregoing embodiments, those skilled in the art should understand that:
It can still modify to technical solution documented by previous embodiment, or to part of or all technical features into
Row equivalent replacement；And these modifications or substitutions, it does not separate the essence of the corresponding technical solution various embodiments of the present invention technical side
The range of case.
Claims (8)
1. a kind of unmanned ship local paths planning method based on the random road sign figure method of comparison expression characterized by comprising
S1, the environment navigated by water according to predetermined unmanned ship, obtain environmental information；
S2, starting point when being navigated by water according to the environmental information and unmanned ship, target point, determine road sign point number K drawn game
Maximum number of iterations G required for portion path avoidance optimizes；
The position of S3, the random each road sign point for generating road sign point number K, starting point, target point and road sign point are connected two-by-two, adopted
Initial ship's navigation route scheme set pop is obtained with multiple target collision checking method0, the initial ship's navigation route scheme
Set pop0It include: the path that a plurality of unmanned ship can navigate by water；Each path inside the map of environmental information and
Belong to non-barrier region；
S4, foundation Revised genetic algorithum and the maximum number of iterations G are to the initial ship's navigation route scheme set pop0
It is iterated processing, the set of paths pop after obtaining iterationm；
S5, the set of paths pop after iteration is obtainedmThe fitness value of middle each route scheme；
S6, according to the set of paths pop after iterationmThe fitness value of middle each route scheme chooses the most short ship for output
Oceangoing ship navigation path scheme.
2. the method according to claim 1, wherein step S3 includes:
S31, according to the environmental information, generate K two-dimensional coordinate at random, each two-dimensional coordinate is as a road sign point
It sets；
S32, judge whether each road sign point position is located at non-obstacle in the environmental information using multiple target collision detection method
Inside object area and map, if so, retaining, otherwise give up the road sign point position；
S33, it combines starting point, target point and any random road sign point to form a plurality of road based on multiple target collision detection method
Diameter；And
Whether whether the line in each paths after judging combination between any two points is located at non-barrier region, and belong to
Inside map；
If being, using the path of judgement as initial ship's navigation route scheme set pop0In element.
3. according to the method described in claim 2, it is characterized in that, in step S33 judgement combination after each paths in
Whether whether the line between any two points is located at non-barrier region, and belong to inside map, comprising:
The angle for obtaining line and positive direction of the x-axis, determines an exploration range according to the angle,
Judge whether determining exploration range includes whether barrier region or the exploration range are in map outside area
Domain；
If not including barrier region and not also in map perimeter, the paths are as initial ship's navigation path side
Case set pop0In element.
4. method according to claim 1 or 3, which is characterized in that step S5 includes:
According to formula (1) to acquisite approachs set popmIn spacing and D in each paths between inflection point；
Wherein, n is each paths inflection point number, and the path inflection point includes: starting point, target point and road sign point, xjAnd yjFor
The coordinate of path inflection point j, xj+1And yj+1 is the coordinate with path inflection point j adjacent comers j+1；
Using the inverse of spacing and D as the fitness value fit of the paths；
5. according to the method described in claim 4, it is characterized in that, step S4 includes:
S41, it is directed to initial ship's navigation route scheme set pop0In pending crossover operation two paths, generate one with
Machine probability；The same length of two paths of the pending crossover operation；
Judge whether the random chance generated is greater than preset crossover probability threshold value；
If more than two paths of pending crossover operation are then carried out crossover operation, obtain two new paths；
If being less than, crossover operation is not carried out to two paths of pending crossover operation；
S42, the initial ship's navigation route scheme set pop is traversed according to the crossover operation mode of step S410In it is all
After path, new path will be obtained and without the original route combination of crossover operation, obtain first path set pop1；
S43, it is directed to first path set pop1In pending mutation operation each paths, generate a variation random chance；
Judge whether the variation random chance generated is greater than preset mutation probability threshold value；
If more than the path of pending mutation operation is then carried out mutation operation, obtains new path；
If being less than, mutation operation is not carried out to the path of pending mutation operation；
S44, the first path set pop is traversed according to the mutation operation mode of step S431In after all path, will
Original route to new path and without mutation operation combines, and obtains the second set of paths pop2；
S45, it is directed to the second set of paths pop2, obtain the second set of paths pop2In each paths fitness value；
Ratio selection operation is carried out according to the fitness value in all paths using Propertional model, after ratio selection operation
Path is as third set of paths pop3；
S46, judge whether current iteration number i is equal to maximum number of iterations G；
If S47, being less than, from third set of paths pop3Start to repeat the above steps the process of S41 to step S46；Until iteration
Number i is equal to maximum number of iterations G；
The set of paths pop that will be obtained when being equal to maximum number of iterations GmAs the set of paths pop after iterationm。
6. according to the method described in claim 5, it is characterized in that, the Liang Tiaolu by pending crossover operation in step S41
Diameter carries out crossover operation, obtains two new paths, comprising:
Two paths essentially identical to path length randomly choose cut-out point on the two route segment and carry out breakpoint truncation, if
Cut-out point position is essentially identical, then is exchanged with each other two paths truncation parts；
If cut-out point position is different, crossover operation is not carried out to two paths of pending crossover operation.
7. according to the method described in claim 5, it is characterized in that, the path by pending mutation operation in S43 becomes
ETTHER-OR operation obtains new path, comprising:
Deflecting operation is carried out to the route segment direction being linked to be between two road sign points of arbitrary neighborhood in each paths.
8. the method according to the description of claim 7 is characterized in that the step S1 includes:
Environmental information is read from the BMP picture file of checkout area sea chart polar plot.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910059813.0A CN109799820B (en) | 2019-01-22 | 2019-01-22 | Unmanned ship local path planning method based on comparative random road map method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910059813.0A CN109799820B (en) | 2019-01-22 | 2019-01-22 | Unmanned ship local path planning method based on comparative random road map method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109799820A true CN109799820A (en) | 2019-05-24 |
| CN109799820B CN109799820B (en) | 2020-12-22 |
Family
ID=66560101
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910059813.0A Active CN109799820B (en) | 2019-01-22 | 2019-01-22 | Unmanned ship local path planning method based on comparative random road map method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109799820B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110955239A (en) * | 2019-11-12 | 2020-04-03 | 中国地质大学（武汉） | Unmanned ship multi-target trajectory planning method and system based on inverse reinforcement learning |
| CN113341992A (en) * | 2021-06-18 | 2021-09-03 | 广东工业大学 | Unmanned ship multitask path planning method |
| CN114355918A (en) * | 2021-12-27 | 2022-04-15 | 北京航天数据股份有限公司 | Deicing vehicle path planning method and device and storage medium |
| CN114536326A (en) * | 2022-01-19 | 2022-05-27 | 深圳市灵星雨科技开发有限公司 | Road sign data processing method and device and storage medium |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001078951A1 (en) * | 2000-04-13 | 2001-10-25 | Zhimin Lin | Semi-optimal path finding in a wholly unknown environment |
| US20090125225A1 (en) * | 2004-03-24 | 2009-05-14 | Bbn Technologies Corp | Vehicle routing and path planning |
| CN101916312A (en) * | 2010-08-09 | 2010-12-15 | 哈尔滨工程大学 | Ship route intelligent generation method based on combination of artificial immunity and ant colony algorithm |
| CN102169347A (en) * | 2011-03-08 | 2011-08-31 | 浙江工业大学 | Multi-robot path planning system based on cooperative co-evolution and multi-population genetic algorithm |
| CN102929286A (en) * | 2012-11-26 | 2013-02-13 | 北京理工大学 | Rapid planning method for surface global path of planet |
| CN104864863A (en) * | 2014-02-21 | 2015-08-26 | 联想(北京)有限公司 | Route selection method and electronic equipment |
| CN106444755A (en) * | 2016-09-22 | 2017-02-22 | 江苏理工学院 | Mobile robot path planning method and system based on improved genetic algorithm |
| US9618934B2 (en) * | 2014-09-12 | 2017-04-11 | 4D Tech Solutions, Inc. | Unmanned aerial vehicle 3D mapping system |
| CN106598054A (en) * | 2017-01-16 | 2017-04-26 | 深圳优地科技有限公司 | Method and device for adjusting robot path |
| CN106843211A (en) * | 2017-02-07 | 2017-06-13 | 东华大学 | A kind of method for planning path for mobile robot based on improved adaptive GA-IAGA |
| CN106845716A (en) * | 2017-01-25 | 2017-06-13 | 东南大学 | A kind of unmanned surface vehicle local delamination paths planning method based on navigation error constraint |
| CN108871351A (en) * | 2018-08-02 | 2018-11-23 | 哈尔滨工程大学 | A kind of matched dynamic path planning method of AUV sea-floor relief |
-
2019
- 2019-01-22 CN CN201910059813.0A patent/CN109799820B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001078951A1 (en) * | 2000-04-13 | 2001-10-25 | Zhimin Lin | Semi-optimal path finding in a wholly unknown environment |
| US20090125225A1 (en) * | 2004-03-24 | 2009-05-14 | Bbn Technologies Corp | Vehicle routing and path planning |
| CN101916312A (en) * | 2010-08-09 | 2010-12-15 | 哈尔滨工程大学 | Ship route intelligent generation method based on combination of artificial immunity and ant colony algorithm |
| CN102169347A (en) * | 2011-03-08 | 2011-08-31 | 浙江工业大学 | Multi-robot path planning system based on cooperative co-evolution and multi-population genetic algorithm |
| CN102929286A (en) * | 2012-11-26 | 2013-02-13 | 北京理工大学 | Rapid planning method for surface global path of planet |
| CN104864863A (en) * | 2014-02-21 | 2015-08-26 | 联想(北京)有限公司 | Route selection method and electronic equipment |
| US9618934B2 (en) * | 2014-09-12 | 2017-04-11 | 4D Tech Solutions, Inc. | Unmanned aerial vehicle 3D mapping system |
| CN106444755A (en) * | 2016-09-22 | 2017-02-22 | 江苏理工学院 | Mobile robot path planning method and system based on improved genetic algorithm |
| CN106598054A (en) * | 2017-01-16 | 2017-04-26 | 深圳优地科技有限公司 | Method and device for adjusting robot path |
| CN106845716A (en) * | 2017-01-25 | 2017-06-13 | 东南大学 | A kind of unmanned surface vehicle local delamination paths planning method based on navigation error constraint |
| CN106843211A (en) * | 2017-02-07 | 2017-06-13 | 东华大学 | A kind of method for planning path for mobile robot based on improved adaptive GA-IAGA |
| CN108871351A (en) * | 2018-08-02 | 2018-11-23 | 哈尔滨工程大学 | A kind of matched dynamic path planning method of AUV sea-floor relief |
Non-Patent Citations (3)
| Title |
|---|
| OZER OZDIKIS: "Genetic Algorithms with Random Coordinates for Route Planning on a 3D Terrain", 《2011 FIFTH INTERNATIONAL CONFERENCE ON GENETIC AND EVOLUTIONARY COMPUTING》 * |
| 刘二辉,等: "基于改进遗传算法的自动导引小车动态路径规划及其实现", 《计算机集成制造系统》 * |
| 徐纪锋，等: "基于自适应蚁群算法的最短路径搜索方法研究", 《中国科技信息》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110955239A (en) * | 2019-11-12 | 2020-04-03 | 中国地质大学（武汉） | Unmanned ship multi-target trajectory planning method and system based on inverse reinforcement learning |
| CN110955239B (en) * | 2019-11-12 | 2021-03-02 | 中国地质大学（武汉） | Unmanned ship multi-target trajectory planning method and system based on inverse reinforcement learning |
| CN113341992A (en) * | 2021-06-18 | 2021-09-03 | 广东工业大学 | Unmanned ship multitask path planning method |
| CN113341992B (en) * | 2021-06-18 | 2023-10-27 | 广东工业大学 | Unmanned ship multitasking path planning method |
| CN114355918A (en) * | 2021-12-27 | 2022-04-15 | 北京航天数据股份有限公司 | Deicing vehicle path planning method and device and storage medium |
| CN114536326A (en) * | 2022-01-19 | 2022-05-27 | 深圳市灵星雨科技开发有限公司 | Road sign data processing method and device and storage medium |
| CN114536326B (en) * | 2022-01-19 | 2024-03-22 | 深圳市灵星雨科技开发有限公司 | Road sign data processing method, device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109799820B (en) | 2020-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109799820A (en) | Unmanned ship local paths planning method based on the random road sign figure method of comparison expression | |
| Shah et al. | Long-distance path planning for unmanned surface vehicles in complex marine environment | |
| Cao et al. | Target search control of AUV in underwater environment with deep reinforcement learning | |
| Faigl | Approximate solution of the multiple watchman routes problem with restricted visibility range | |
| CN109839110A (en) | A kind of multiple target point path planning method based on quick random search tree | |
| Kala et al. | Robotic path planning in static environment using hierarchical multi-neuron heuristic search and probability based fitness | |
| Ghorbani et al. | Using genetic algorithm for a mobile robot path planning | |
| CN108489491A (en) | A kind of Three-dimensional Track Intelligent planning method of autonomous underwater vehicle | |
| Wang et al. | Scene mover: Automatic move planning for scene arrangement by deep reinforcement learning | |
| CN108984741A (en) | A kind of ground drawing generating method and device, robot and computer readable storage medium | |
| CN108413963A (en) | Bar-type machine people's paths planning method based on self study ant group algorithm | |
| Shivashankar et al. | Real-time planning for covering an initially-unknown spatial environment | |
| CN112612290B (en) | Underwater vehicle three-dimensional multi-task path planning method considering ocean currents | |
| Du et al. | An optimized path planning method for coastal ships based on improved DDPG and DP | |
| CN115129064A (en) | Path planning method based on fusion of improved firefly algorithm and dynamic window method | |
| Gu et al. | Path planning for mobile robot in a 2.5‐dimensional grid‐based map | |
| Funk et al. | Graph-based Reinforcement Learning meets Mixed Integer Programs: An application to 3D robot assembly discovery | |
| Shen et al. | CT-CPP: Coverage path planning for 3D terrain reconstruction using dynamic coverage trees | |
| Lin et al. | Application of improved RRT algorithm in unmanned surface vehicle path planning | |
| CN115951682A (en) | Constraint four-target AUV global path planning method under condition of considering ocean current existence | |
| CN115759199A (en) | Multi-robot environment exploration method and system based on hierarchical graph neural network | |
| Wu et al. | Complete Coverage Path Planning Based on Improved Genetic Algorithm for Unmanned Surface Vehicle | |
| Opoku et al. | The Ar-Star (Ar) Pathfinder | |
| Wang et al. | Welding robot path optimization based on hybrid discrete PSO | |
| Yan | Research on the A Star Algorithm for Finding Shortest Path |
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 |