CN113505836B - Ship route automatic generation method based on big data - Google Patents
Ship route automatic generation method based on big data Download PDFInfo
- Publication number
- CN113505836B CN113505836B CN202110794625.XA CN202110794625A CN113505836B CN 113505836 B CN113505836 B CN 113505836B CN 202110794625 A CN202110794625 A CN 202110794625A CN 113505836 B CN113505836 B CN 113505836B
- Authority
- CN
- China
- Prior art keywords
- ship
- navigation
- points
- point
- route
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Human Resources & Organizations (AREA)
- General Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Evolutionary Biology (AREA)
- Marketing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Development Economics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Game Theory and Decision Science (AREA)
- Artificial Intelligence (AREA)
- Entrepreneurship & Innovation (AREA)
- Evolutionary Computation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Remote Sensing (AREA)
- Computational Linguistics (AREA)
- Navigation (AREA)
Abstract
The invention discloses a ship route automatic generation method based on big data, which comprises the steps of obtaining original data of a ship navigation track and establishing an original set of the ship navigation track; extracting a starting point and a terminating point of a navigation track in the original set of the ship navigation track, and classifying the navigation track to obtain a ship navigation track clustering set; giving a navigation plan and ship information, searching similar navigation tracks, and determining a navigable path with the maximum transition probability as a planned route of the voyage; step S3 is repeatedly executed, and a ship route mapping library is established; automatically generating a planned route of the current ship by giving constraint conditions of the current ship navigation plan; the invention comprehensively considers a plurality of factors such as the navigation season of the ship, the ship information and the like, establishes the ship route mapping library, can directly carry out dynamic inquiry in the ship route mapping library according to the constraint condition in the subsequent route design process, automatically generates the current planned route, and greatly shortens the route design time.
Description
Technical Field
The invention relates to the technical field of automatic generation of ship routes, in particular to a ship route automatic generation method based on big data.
Background
Shipping undertakes most of trade transportation tasks in the world, and ship navigation accidents happen occasionally, which causes casualties and major property loss, so how to ensure safe navigation of ships has important significance for improving production efficiency and ensuring life and property safety. The ship route planning can enable a ship to run in a safe sea area and be far away from a dangerous area; meanwhile, the optimized air route can shorten the navigation time of the ship, improve the navigation efficiency and play a role in energy conservation and emission reduction. At present, ship route planning is mainly drawn manually by a driver, the quality of a route depends excessively on the navigation experience and professional background of the driver, the planned route cannot provide a scientific route for the driver for reference, the productivity is low, and the ship navigation safety cannot be well guaranteed.
Existing automatic route generation techniques include dynamic route planning and static route design. The dynamic route planning mainly aims at that a moving target changes along with time in the process of sailing of a ship, and therefore the route planning is carried out by combining an avoidance algorithm. The existing dynamic route planning technology is mainly designed based on algorithms such as A-x dynamic search, heuristic genetic search, planning grid search and the like, and requires real-time acquisition of the motion state of a moving target. Different from dynamic route planning, the static route design is a ship navigation route which needs to be designed before a ship is launched, the ship navigates according to the route in the navigation process, and actively returns to the navigation route after passing through a maneuvering obstacle avoidance in the process. The prior art of static route design is mainly realized by optimizing steering points and setting an objective function such as the shortest path, most of the static route design adopts a graph theory method to search the shortest path, does not consider the seaworthiness of a ship in a sea area, and cannot adapt to the inherent navigation rule of the ship in the sea.
Disclosure of Invention
In view of the above, the present invention aims to provide a method for automatically generating a ship route based on big data, so as to solve the problems in the prior art that route planning excessively depends on navigation experience of a driver and professional background, so that the referential is not high, the productivity is low, the navigation safety cannot be well guaranteed, and the inherent navigation law of a ship cannot be adapted.
In order to achieve the purpose, the invention provides the following technical scheme:
a ship route automatic generation method based on big data comprises the following steps
S1: acquiring original data of a ship navigation track, and sorting and cleaning the original data to establish an original set of the ship navigation track;
s2: extracting starting points and ending points of all navigation tracks in the original set of the ship navigation tracks, and classifying the ship navigation tracks through a clustering algorithm to obtain a ship navigation track clustering set;
s3: giving a navigation plan and ship information of a ship, searching similar navigation tracks in a ship navigation track cluster set, obtaining a navigable path set based on the similar navigation tracks, calculating a transition probability corresponding to each navigable path in the navigable path set, and determining the navigable path with the highest transition probability as a planned route of the voyage;
s4: step S3 is repeatedly executed for multiple times, a plurality of planned routes corresponding to the navigation plan of the given ship are calculated and stored, constraint conditions are set based on the planned routes and the corresponding ship information, and a ship route mapping base is established;
s5: and giving the constraint condition of the current ship navigation plan, carrying out route mapping in the ship route mapping library, and automatically generating the planned route of the current ship.
Further, the specific method of step S1 is:
acquiring navigation tracks of all voyages of each ship in a preset time and a preset sea area to form original data, sorting and cleaning the original data, eliminating the navigation track with abnormal data in the original data, obtaining the normal navigation track of each voyage of each ship, and establishing an original set of the ship navigation tracks.
Further, the step S2 includes the steps of:
s201: extracting the starting point and the end point of each navigation track in the original set of the navigation tracks of the ship to form a starting point set and an end point set, and marking the navigation tracks according to corresponding ship information;
s202: clustering the initial points and the end points extracted in the step S201 respectively by adopting a clustering algorithm to obtain corresponding clustering results;
s203: traversing all navigation tracks in the original set of the ship navigation tracks, and classifying the navigation tracks according to the clustering results of the starting points and the end points in the step S202 to form a ship navigation track clustering set.
Further, the step S202 adopts a density-based clustering method and a DBSCAN algorithm to cluster the start point and the end point, respectively, and the specific steps are as follows:
s2021: setting a neighborhood radius of the DBSCAN algorithm and a preset number of starting points/end points in the neighborhood radius;
s2022: randomly selecting unprocessed first starting points/first ending points in the starting point set/the ending point set, and checking the quantity of the starting points/the ending points contained in the adjacent domain radius of the first starting points/the first ending points;
s2023: judging whether the number of the start points/the end points included in the first start point/first end point neighborhood radius in the step S2022 is greater than or equal to a preset number, if so, continuing to execute the step S2024, otherwise, repeating the step S2022;
s2024: establishing a corresponding first clustering set by using a first initial point/a first end point, establishing a candidate set, and classifying initial points/end points which are not classified into any clustering set or marked as noise and are contained in the neighborhood radius of the first initial point/the first end point into the candidate set;
s2025: selecting unprocessed second starting points/second end points in the candidate set, and checking the number of the starting points/end points contained in the adjacent radius of the second starting points/second end points;
s2026: judging whether the number of the start points/end points included in the neighborhood radius of the second start point/second end point in the step S2025 is greater than or equal to the preset number, if so, adding the second start point/second end point and the start point/end point which is not classified into any cluster set or not marked as noise in the neighborhood radius to the corresponding first cluster set, otherwise, only classifying the second start point/second end point into the corresponding first cluster set;
s2027: repeating steps S2025-S2026 until the candidate set is empty, and continuing to execute step S2028;
s2028: and (5) repeating the steps (S2022) - (S2027) until all the starting points/the end points in the starting point set/the end point set belong to a certain cluster set or are marked as noise, and obtaining a cluster set of the starting points/the end points of the ship navigation track.
Further, when determining the starting point/ending point included in the neighborhood radius of the starting point/ending point, calculating the distance between the two starting points/ending points, and if the distance between the two starting points/ending points is smaller than the neighborhood radius, the starting point/ending point is included in the neighborhood radius of the corresponding starting point/ending point.
Further, the distance between the two start points/end points is calculated by using an euclidean distance formula, which is as follows:
wherein: (x) 1 ,x 2 ) And (y) 1 ,y 2 ) Two-dimensional coordinates corresponding to the two start points/end points respectively.
Further, the step S3 includes the steps of:
s301: giving a ship navigation plan, classifying the navigation plan into a category represented by a corresponding ship navigation track cluster set according to an initial point and an end point of the navigation plan, dynamically searching similar ship navigation tracks in the similar ship navigation track cluster set according to ship information of ships, and establishing an approximate track set;
s302: establishing a transition probability directed graph aiming at the approximate track set in the step S301;
s303: searching the navigable paths of the ship in the transition probability digraph, establishing a navigable path set, calculating the transition probability of each navigable path in the navigable path set, and taking the navigable path with the maximum transition probability as a planned route from a starting point to a terminating point.
Further, the specific method for establishing the transition probability directed graph in step S302 is as follows:
marking a plurality of nodes on each navigation track in the approximate track set at equal distance, calculating the transition probability from each node position to the adjacent node position, and establishing a transition probability directed graph.
Further, in step S303, the calculation formula of the transition probability of the navigable path is as follows:
wherein: p (Tr) j ) The transition probability of the jth navigable path in the navigable path set is obtained;
for the ith node in the jth navigable path, i =1,2, \ 8230, m and m are the number of nodes in any navigable path;
j =1,2, \ 8230for the (i + 1) th node in the jth navigable path, and k are the number of navigable paths;
the transition probability of the ith node in the jth navigable path is obtained;
to exist node
Time node
The transition probability of (2);
the planned route from the starting point to the ending point based on the navigable path with the highest transition probability may be represented as:
path＝argmaxP(Tr j )；
wherein: and the path is a planned route generated by the voyage based on the navigable path with the maximum transition probability.
Further, the ship information includes ship type, ship size, loading state, draft and sailing season.
According to the scheme, the historical navigation tracks of the ship are classified by adopting a density-based clustering method, so that the universality and the accuracy of ship navigation track clustering are improved, and a data base is laid for real-time route generation; in addition, when a ship route mapping library is established, a plurality of constraint conditions such as the current navigation season, ship information and the like are combined, and big data search based on the ship navigation track can improve the accuracy of real-time route generation of a ship and ensure the scientificity of route generation; finally, dynamic inquiry of the ship route is carried out in the ship route mapping library, so that the time required by route generation can be greatly shortened, the calculation cost is greatly saved, and the real-time performance of route generation is improved.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention may be realized and attained by the means of the instrumentalities and combinations particularly pointed out hereinafter.
Drawings
FIG. 1 is a flow chart of a ship route automatic generation method based on big data.
Fig. 2 is a flowchart of step S2 in fig. 1.
Fig. 3 is a flowchart of step 202 in fig. 1.
Fig. 4 is a flowchart of step S3 in fig. 1.
Detailed Description
The following is further detailed by way of specific embodiments:
examples
Fig. 1 is a flow chart of a method for automatically generating a ship route based on big data according to the present invention. The ship route automatic generation method based on big data of the embodiment specifically comprises the following steps:
s1: and establishing an original set of ship sailing tracks.
Specifically, acquiring navigation tracks of all voyages of each ship in a preset time and a preset sea area, and taking all the navigation tracks as original data; and sorting and cleaning the original data, and eliminating the navigation track with abnormal data (such as abnormal navigation speed, abnormal navigation track and the like) in the original data to obtain the normal navigation track of each ship in each voyage, so as to establish an original set of ship navigation tracks.
S2: and classifying the ship navigation tracks and establishing a ship navigation track clustering set.
Extracting the starting points and the end points of all navigation tracks in the original set of the navigation tracks of the ship, clustering all the starting points and the end points through a clustering algorithm, and classifying the navigation tracks of the ship according to the clustering results of the starting points and the end points to obtain a ship navigation track clustering set.
As shown in fig. 2, the step S2 includes the following steps:
s201: and extracting a starting point and an end point and marking ship information.
Firstly, extracting a starting point and an end point of each navigation track in an original set of navigation tracks of the ship to form a starting point set and an end point set, wherein the starting point set D is S And set of end points D E Respectively expressed as:
wherein:
as a starting point, the point of the start,
n is the number of start points or the number of end points.
Because the ship has an inherent navigation rule when sailing on the sea, in order to increase the seaworthiness of the ship in the sea and integrate the navigation rules of historical ships, each navigation track needs to be marked according to corresponding ship information. In the present embodiment, the ship information includes, but is not limited to, ship type, ship size, ship state, draft, and sailing season.
S202: and clustering the starting points and the end points.
In order to improve the universality and accuracy of ship track clustering, a clustering algorithm is adopted to cluster the initial points and the end points extracted in the step S201 respectively to obtain corresponding clustering results. In this embodiment, the density-based clustering method uses a DBSCAN algorithm to cluster the start points and the end points, respectively.
Since the same method is used when clustering is performed on the start point and the end point, in this embodiment, the start point is taken as an example for description.
As shown in fig. 3, the step S202 includes the following steps:
s2021: initializing a DBSCAN algorithm, and setting a preset number of other starting points contained in a neighborhood radius e of the DBSCAN algorithm and a neighborhood radius e corresponding to a certain starting point, namely, a minimum number minPts of the starting points or the ending points contained in the neighborhood radius e of a processing object (starting point).
S2022: randomly selecting a set of starting points D S The first starting point of the cluster that is not processed (i.e., not classified as a cluster set or marked as noise)
Is denoted as a first object p; checking the number Pts of starting points contained by the first object p within the set neighborhood radius e 1 。
When other starting points included in the neighborhood radius e of the first object p are determined, the distance rho from the first object p to other starting points in the neighborhood radius e is calculated by adopting an Euclidean distance formula:
wherein: (x) 1 ,x 2 ) And (y) 1 ,y 2 ) Two-dimensional coordinates corresponding to the two starting points respectively.
Judging whether the Euclidean distance rho between the two initial points obtained by calculation in the formula (3) is smaller than the neighborhood radius e, if so, indicating that the initial points are contained in the neighborhood radius e of the first object p, and thus obtaining the quantity Pts of the initial points contained in the neighborhood radius e of the first object p 1 。
S2023: it is determined whether the number Pts of the start points included in the neighborhood radius e of the first object p in step S2022 is greater than or equal to the preset number minPts.
If the number of start points Pts contained in the neighborhood radius e of the first object p 1 If the number is greater than or equal to the preset number minPts, the process continues to step S2024.
If the number of start points Pts contained in the neighborhood radius e of the first object p 1 If the number is less than the preset number minPts, step S2022 is repeatedly performed to continue to select other unprocessed first starting points for processing.
S2024: with the first object p (i.e. the first starting point selected in step S2021)
) Establishing a corresponding first cluster set C for core points 1 And establishing a candidate set N, and then classifying the starting points which are not classified into any cluster set or marked as noise and are contained in the neighborhood radius e of the first object p into the candidate set N, namely:
wherein:
is a starting point comprised within a radius e in the neighborhood of the first object p.
S2025: selecting a second starting point within the candidate set where N is unprocessed
Marking as a second object q, determining other starting points contained in the radius e in the neighborhood of the second object q by adopting the method in the step S2022, and obtaining the number Pts of the starting points contained in the radius e in the neighborhood of the second object q 2 。
S2026: it is determined whether the number of the starting points included in the neighborhood radius e of the second object q in step S2025 is greater than or equal to the preset number minPts.
If the number of starting points Pts contained in the neighborhood radius e of the second object q 2 If the number is greater than or equal to the predetermined number minPts, the second object q (i.e. the second starting point) is set
) Adding the initial point which is not classified into any cluster set or marked as noise in the neighborhood radius e to the corresponding first cluster set C 1 In (1). The starting point which is not classified into any cluster set or marked as noise in the neighborhood radius e of the second object q can be represented as:
wherein: r is the number of starting points within the neighborhood radius e of the second object q that have been classified into any cluster set, which is marked as noise.
If the number of starting points Pts contained in the neighborhood radius e of the second object q 2 Less than the predetermined number minPts, then only the second object q (i.e., the second starting point) is selected
) Adding to the corresponding first cluster set C 1 And (4) neutralizing.
S2027: steps S2025-S2026 are repeated until the candidate set N is emptyTime (i.e. all the second starting points in the candidate set N)
All classified as a cluster set or marked as noise), continue to execute step S2028;
s2028: repeating the steps S2022-S2027 until the starting point set D S All starting points in
And all the ship navigation tracks are classified into a certain cluster set or marked as noise, and the cluster set of the ship navigation track starting point is obtained. The set of starting points D S Can be expressed as:
D S ＝C 1 +C 2 +…+C M +ε (6)
wherein: c 1 ,…,C M Respectively clustering sets for starting points; m is the number of the starting point cluster sets, namely the starting point set D S All starting points in
The number of categories to be divided; ε is the set of starting points marked as noise.
S203: and classifying the navigation tracks according to the clustering result.
Traversing all navigation tracks in the original set of the ship navigation tracks, classifying the navigation tracks according to the clustering results of the starting points and the end points in the step S202, and classifying all the ship navigation tracks into a plurality of categories to form a ship navigation track clustering set.
S3: and calculating the transition probability and selecting a planning route.
Specifically, a navigation plan and corresponding ship information of a ship are given, a navigation track similar to the given ship navigation plan and the ship information is searched in a ship navigation track clustering set, a current navigable path set of the given ship is obtained based on the similar navigation track, a transition probability corresponding to each navigable path in the navigable path set is calculated, and the navigable path with the maximum transition probability is determined as a planned route of the current voyage of the ship.
As shown in fig. 4, the step S3 includes the following steps:
s301: dynamically searching the ship sailing track and establishing an approximate track set.
The method comprises the steps of giving a navigation plan of a ship, classifying the navigation plan into a category represented by a corresponding ship navigation track cluster set according to an initial point and an end point of the navigation plan, dynamically searching similar ship navigation tracks in the ship navigation track cluster sets of the same category according to ship information of the ship, comparing the ship information (the type, the size, the loading state, the draft, the navigation season and the like of the ship) of the given ship with the type, the size, the loading state, the draft and the navigation season of the ship marked by the navigation tracks in the ship navigation track cluster sets of the same category, searching the similar ship navigation tracks, and establishing an approximate track set of the given ship at the current navigation time.
S302: and establishing a transition probability directed graph.
For the approximate track set in step S301, a plurality of nodes are marked at equal intervals on each navigation track, the transition probability from each node position to the adjacent node position (the adjacent node position may be the node position on the same navigation track, or the node position on other navigation tracks in the approximate track set) is calculated, and a transition probability directed graph is established.
In this embodiment, the established transition probability directed graph is defined as G (V, E, W), where V represents a set of all nodes on each navigation track in the approximate track set; e represents a set of directed edges formed by connecting lines of the current node position and the next adjacent node position; w represents the transition probability of each directed edge, i.e., the probability of transitioning from the current node position to the next node position.
S303: and calculating the transition probability to obtain a planned route.
Searching the navigable path of the given ship in the current voyage according to the transition probability directed graph established in the step S302, and establishing a navigable path set { Tr 1 ,Tr 2 ,…,Tr j } (j =1,2, \ 8230;, k, k being navigablePath set { Tr 1 ,Tr 2 ,…,Tr j The number of navigable paths in (i.e. Tr) j Representing a set of navigable paths { Tr 1 ,Tr 2 ,…,Tr j J-th navigable path in (j), and then separately compute a set of navigable paths { Tr } 1 ,Tr 2 ,…,Tr j And (4) taking the navigable path with the maximum transition probability as a planning route from a starting point to an ending point.
For navigable path set { Tr 1 ,Tr 2 ,…,Tr j Each navigable path Tr in the } can be regarded as a chain with m nodes, denoted as:
for i =1,2, \ 8230;, m-1, the transition probability of the current node can be based on
Calculating the transition probability of the next node
The transition probability for each navigable path Tr can thus be calculated:
wherein: p (Tr) j ) For a navigable set of paths { Tr 1 ,Tr 2 ,…,Tr j The transition probability of the jth navigable path in the } is determined;
is the ith node in the jth navigable path.
Thus, the planned route path based on the navigable path with the highest transition probability as the starting point to the ending point can be represented as:
path＝argmaxP(Tr j ) (8)
s4: and establishing a ship route mapping library.
Giving a large number of ship navigation plans and corresponding ship information, repeatedly executing the step S3, calculating and storing a planned route corresponding to the navigation plan of each given ship, setting constraint conditions based on each planned route (including a starting point, an end point and the like) and corresponding ship information (ship type, ship size, loading state, draft, navigation season and the like), and establishing a ship route mapping library.
S5: and automatically planning a ship route.
Before the ship sails, a sailing plan and constraint conditions of the ship are given, route mapping is carried out in the ship route mapping library, and a planned route of the current ship is automatically generated.
According to the scheme, multiple factors such as the navigation season of the ship, ship information and the like are comprehensively considered, dynamic search is carried out based on large data of the ship track, a strong ship route mapping library is established, and dynamic query can be directly carried out in the ship route mapping library according to constraint conditions in the subsequent route design process, so that the current planned route is automatically generated, and the time for route design is greatly shortened.
The foregoing is merely an example of the present invention and common general knowledge of known specific structures and features of the embodiments is not described herein in any greater detail. It should be noted that, for those skilled in the art, without departing from the structure of the present invention, several variations and modifications can be made, which should also be regarded as the scope of the present invention, and these do not affect the effect of the implementation of the present invention and the practicability of the present invention.
Claims (10)
1. A ship route automatic generation method based on big data is characterized by comprising the following steps:
s1: acquiring original data of a ship navigation track, and sorting and cleaning the original data to establish an original set of the ship navigation track;
s2: extracting starting points and ending points of all navigation tracks in the original set of the ship navigation tracks, and classifying the ship navigation tracks through a clustering algorithm to obtain a ship navigation track clustering set;
s3: giving a navigation plan and ship information of a ship, searching similar navigation tracks in a ship navigation track cluster set, obtaining a navigable path set based on the similar navigation tracks, calculating a transition probability corresponding to each navigable path in the navigable path set, and determining the navigable path with the maximum transition probability as a planned route of the voyage;
s4: step S3 is repeatedly executed for multiple times, a plurality of planned routes corresponding to the navigation plan of the given ship are calculated and stored, constraint conditions are set based on the planned routes and the corresponding ship information, and a ship route mapping base is established;
s5: and giving the constraint condition of the current ship navigation plan, carrying out route mapping in the ship route mapping library, and automatically generating the planned route of the current ship.
2. The automatic ship route generation method based on big data according to claim 1, characterized in that the specific method of step S1 is:
acquiring navigation tracks of all voyages of each ship in a preset time and a preset sea area to form original data, sorting and cleaning the original data, eliminating abnormal navigation tracks in the original data to obtain normal navigation tracks of each voyage of each ship, and establishing an original set of ship navigation tracks.
3. The big data based ship route automatic generation method according to claim 1, wherein the step S2 comprises the steps of:
s201: extracting the starting point and the end point of each navigation track in the original set of the navigation tracks of the ship to form a starting point set and an end point set, and marking the navigation tracks according to corresponding ship information;
s202: clustering the initial points and the end points extracted in the step S201 respectively by adopting a clustering algorithm to obtain corresponding clustering results;
s203: traversing all navigation tracks in the original set of the ship navigation tracks, and classifying the navigation tracks according to the clustering results of the starting points and the end points in the step S202 to form a ship navigation track clustering set.
4. The automatic ship route generation method based on big data as claimed in claim 3, wherein said step S202 adopts a density-based clustering method and uses DBSCAN algorithm to cluster the starting point and the ending point respectively, and its concrete steps are:
s2021: setting a neighborhood radius of the DBSCAN algorithm and a preset number of starting points/end points in the neighborhood radius;
s2022: randomly selecting unprocessed first starting points/first ending points in the starting point set/the ending point set, and checking the number of the starting points/the ending points contained in the adjacent radius of the first starting points/the first ending points;
s2023: judging whether the number of the start points/the end points included in the neighborhood radius of the first start point/the first end point in the step S2022 is greater than or equal to a preset number, if so, continuing to execute the step S2024, otherwise, repeating the step S2022;
s2024: establishing a corresponding first clustering set by using a first initial point/a first end point, establishing a candidate set, and classifying initial points/end points which are not classified into any clustering set or marked as noise and are contained in the neighborhood radius of the first initial point/the first end point into the candidate set;
s2025: selecting unprocessed second starting points/second end points in the candidate set, and checking the number of the starting points/end points contained in the adjacent radius of the second starting points/second end points;
s2026: judging whether the number of the start points/end points included in the neighborhood radius of the second start point/second end point in the step S2025 is greater than or equal to the preset number, if so, adding the second start point/second end point and the start point/end point which is not classified into any cluster set or not marked as noise in the neighborhood radius to the corresponding first cluster set, otherwise, only classifying the second start point/second end point into the corresponding first cluster set;
s2027: repeating steps S2025-S2026 until the candidate set is empty, and continuing to execute step S2028;
s2028: and (5) repeating the steps (S2022) - (S2027) until all the starting points/the end points in the starting point set/the end point set belong to a certain cluster set or are marked as noise, and obtaining a cluster set of the starting points/the end points of the ship navigation track.
5. The automatic big-data-based ship route generation method according to claim 4, wherein when determining the start point/end point included in the neighborhood radius of the start point/end point, the distance between the two start points/end points is calculated, and if the distance between the two start points/end points is smaller than the neighborhood radius, the start point/end point is included in the neighborhood radius of the corresponding start point/end point.
6. The automatic ship route generation method based on big data of claim 5, wherein the distance between two start points/end points is calculated by Euclidean distance formula, and the calculation formula is as follows:
wherein: (x) 1 ,x 2 ) And (y) 1 ,y 2 ) Two-dimensional coordinates corresponding to the two start points/end points respectively.
7. The big data based ship route automatic generation method according to claim 4, wherein the step S3 comprises the steps of:
s301: giving a ship navigation plan, classifying the navigation plan into a category represented by a corresponding ship navigation track cluster set according to an initial point and an end point of the navigation plan, dynamically searching similar ship navigation tracks in the similar ship navigation track cluster set according to ship information of ships, and establishing an approximate track set;
s302: establishing a transition probability directed graph aiming at the approximate track set in the step S301;
s303: searching the navigable paths of the ship in the transition probability directed graph, establishing a navigable path set, calculating the transition probability of each navigable path in the navigable path set, and taking the navigable path with the maximum transition probability as a planned route from a starting point to an ending point.
8. The automatic ship route generation method based on big data as claimed in claim 7, wherein the specific method for establishing the transition probability directed graph in step S302 is as follows:
marking a plurality of nodes on each navigation track in the approximate track set at equal distance, calculating the transition probability from each node position to the adjacent node position, and establishing a transition probability directed graph.
9. The method as claimed in claim 7, wherein in step S303, the calculation formula of the transition probability of the navigable path is as follows:
wherein: p (Tr) j ) The transition probability of the jth navigable path in the navigable path set is obtained;
for the ith node in the jth navigable path, i =1,2, \8230, and m are the number of nodes in any navigable path;
j =1,2, \ 8230for the (i + 1) th node in the jth navigable path, and k are the number of navigable paths;
the transition probability of the ith node in the jth navigable path is obtained;
to exist a node
Time node
The transition probability of (2);
the planned route based on the navigable path with the highest transition probability as the starting point to the ending point may be represented as:
path＝arg max P(Tr j )；
wherein: and the path is a planned route generated by the voyage route based on the navigable path with the maximum transition probability.
10. The automatic big data-based ship route generation method according to claim 1, wherein the ship information includes a ship type, a ship type size, a ship state, a draft, and a voyage season.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110794625.XA CN113505836B (en) | 2021-07-14 | 2021-07-14 | Ship route automatic generation method based on big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110794625.XA CN113505836B (en) | 2021-07-14 | 2021-07-14 | Ship route automatic generation method based on big data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113505836A CN113505836A (en) | 2021-10-15 |
| CN113505836B true CN113505836B (en) | 2022-11-15 |
Family
ID=78013187
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110794625.XA Active CN113505836B (en) | 2021-07-14 | 2021-07-14 | Ship route automatic generation method based on big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113505836B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110309383A (en) * | 2019-06-17 | 2019-10-08 | 武汉科技大学 | Ship trajectory clustering analysis method based on improved DBSCAN algorithm |
| CN111401668A (en) * | 2020-06-05 | 2020-07-10 | 江苏海事职业技术学院 | Unmanned ship route planning method based on big data |
| CN111578941A (en) * | 2020-04-26 | 2020-08-25 | 亿海蓝（北京）数据技术股份公司 | Ship route recommendation method, system and storage medium |
| CN112562415A (en) * | 2020-11-26 | 2021-03-26 | 中远海运科技股份有限公司 | Ship route recommendation method |
| CN112906830A (en) * | 2021-04-14 | 2021-06-04 | 武汉理工大学 | Automatic generation method of optimal ship route based on AIS big data |
-
2021
- 2021-07-14 CN CN202110794625.XA patent/CN113505836B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110309383A (en) * | 2019-06-17 | 2019-10-08 | 武汉科技大学 | Ship trajectory clustering analysis method based on improved DBSCAN algorithm |
| CN111578941A (en) * | 2020-04-26 | 2020-08-25 | 亿海蓝（北京）数据技术股份公司 | Ship route recommendation method, system and storage medium |
| CN111401668A (en) * | 2020-06-05 | 2020-07-10 | 江苏海事职业技术学院 | Unmanned ship route planning method based on big data |
| CN112562415A (en) * | 2020-11-26 | 2021-03-26 | 中远海运科技股份有限公司 | Ship route recommendation method |
| CN112906830A (en) * | 2021-04-14 | 2021-06-04 | 武汉理工大学 | Automatic generation method of optimal ship route based on AIS big data |
Non-Patent Citations (2)
| Title |
|---|
| 一种基于ＡＩＳ数据的船舶航线自动规划方法;石浩 等;《上海船舶运输科学研究所学报》;20210330;25-30 * |
| 基于AIS数据库自动航线设计研究;吕建;《中国优秀硕士学位论文全文数据库（电子期刊）工程科技Ⅱ辑》;20180715;C036-198 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113505836A (en) | 2021-10-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108459503B (en) | Unmanned surface vehicle track planning method based on quantum ant colony algorithm | |
| CN113961004A (en) | Pirate area ship route planning method and system, electronic equipment and storage medium | |
| CN112906830B (en) | Automatic generation method of ship optimal route based on AIS big data | |
| CN110398250B (en) | Unmanned ship global path planning method | |
| CN110196059B (en) | Unmanned ship global path planning method | |
| CN110319837A (en) | Indoor complex condition path planning method for service robot | |
| Delling et al. | Customizable point-of-interest queries in road networks | |
| Song et al. | T $^\star $: Time-Optimal Risk-Aware Motion Planning for Curvature-Constrained Vehicles | |
| CN111562786B (en) | Multi-stage optimized unmanned ship path planning method and device | |
| CN111721296B (en) | Data driving path planning method for underwater unmanned vehicle | |
| CN114427866A (en) | Path planning method, electronic device and storage medium | |
| CN109974708B (en) | Unmanned ship track planning method, terminal equipment and storage medium | |
| Erdoğan et al. | Exact and heuristic algorithms for the carrier–vehicle traveling salesman problem | |
| CN116952251B (en) | Path planning method, device, terminal equipment and readable storage medium | |
| CN113505836B (en) | Ship route automatic generation method based on big data | |
| CN113311843B (en) | Unmanned ship path planning method based on safe distance constraint and LOS sight judgment | |
| Hu et al. | Research and realization of optimum route planning in vehicle navigation systems based on a hybrid genetic algorithm | |
| CN116224999A (en) | Ship global path planning method based on data driving | |
| CN113776535A (en) | Unmanned ship route planning method based on rasterized electronic chart | |
| CN113625709A (en) | Obstacle avoidance method for unmanned surface vehicle | |
| CN113341957A (en) | Multi-robot path planning method based on trace map A _ star algorithm | |
| Zhao | Step-by-step multi-ship collision avoidance navigation path planning based on particle swarm optimization | |
| Li et al. | DF-FS based path planning algorithm with sparse waypoints in unstructured terrain | |
| CN115454106B (en) | AUV (autonomous Underwater vehicle) docking path planning method based on bidirectional search RRT (remote radio transmitter) | |
| CN115326058B (en) | Robot map generation method, device, storage medium, and mobile robot |
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 |