CN110309383A - Ship trajectory clustering analysis method based on improved DBSCAN algorithm - Google Patents
Ship trajectory clustering analysis method based on improved DBSCAN algorithm Download PDFInfo
- Publication number
- CN110309383A CN110309383A CN201910521851.3A CN201910521851A CN110309383A CN 110309383 A CN110309383 A CN 110309383A CN 201910521851 A CN201910521851 A CN 201910521851A CN 110309383 A CN110309383 A CN 110309383A
- Authority
- CN
- China
- Prior art keywords
- track
- ship
- neighborhood
- eps
- length
- 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
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 32
- 238000004458 analytical method Methods 0.000 title claims abstract description 19
- 238000000034 method Methods 0.000 claims abstract description 28
- 230000004927 fusion Effects 0.000 claims abstract description 25
- 238000005259 measurement Methods 0.000 claims abstract description 12
- 238000012360 testing method Methods 0.000 claims description 5
- 241000208340 Araliaceae Species 0.000 claims 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims 1
- 235000003140 Panax quinquefolius Nutrition 0.000 claims 1
- 238000000151 deposition Methods 0.000 claims 1
- 235000008434 ginseng Nutrition 0.000 claims 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 8
- 238000010586 diagram Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 230000003542 behavioural effect Effects 0.000 description 3
- 238000011144 upstream manufacturing Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000009412 basement excavation Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000013256 coordination polymer Substances 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 231100001263 laboratory chemical safety summary Toxicity 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
- 239000003643 water by type Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/906—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Business, Economics & Management (AREA)
- Databases & Information Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of ship trajectory clustering analysis methods based on improved DBSCAN algorithm, include the following steps: S1, extract effective ship track data from AIS database；S2, track similarity measurement is carried out using fusion distance MD, obtains fused most short super track between track；S3, the length that most short super track is calculated using secondary time O (mn)；S4, fusion distance MD is obtained from the length of track；S5, the global parameter for determining improved DBSCAN algorithm；S6, entire track segment data set is scanned to obtain the set of cluster by improved DBSCAN algorithm；S7, the spatial movement mode for acquisition, if a new track meets one of vessel motion mode, then it is assumed that the track is a normality track.This method improves the accuracy of measurement, is compared to traditional DBSCAN algorithm and reduces the consumption of time, while can handle Multi-density data set well, has preferable robustness and adaptability, is also correctly classified to track.
Description
Technical field
The present invention relates to navigation safety technical field more particularly to a kind of ships based on improved DBSCAN algorithm
Trajectory clustering analysis method.
Background technique
With the development of economy, the enhancing of waterway transportation and international trade leads to the rise of maritime traffic, produces simultaneously
A large amount of ship motion profiles, this is of great significance to the transportation analysis of whole world sea-freight and supervision.Nowadays the development of technology introduces
Ship automatic identification system (AIS) tracks ship and monitors maritime affairs, it can cooperate global positioning system (GPS) and other ships
A series of Ship dynamic situations and the static information such as oceangoing ship and AIS base station exchange position, course, the speed of a ship or plane promote marine ships movement
Tracking and monitoring.By these AIS data, ship motor pattern is excavated, predicted motion state simultaneously carries out behavioural analysis, track number
According to analysis and excavation have become a new research hotspot in the field of data mining.
It generallys use DBSCAN algorithm and clustering, but the precision of existing DBSCAN algorithm measurement is carried out to ship track
Lower, the classification of obtained track is not especially accurate and long to the time consumed by ship trajectory clustering, robustness and suitable
Answering property is also poor.
Summary of the invention
To overcome the above deficiencies, the invention provides a kind of ship tracks based on improved DBSCAN algorithm
Clustering method.
The present invention overcomes the technical solution used by its technical problem to be:
A kind of ship trajectory clustering analysis method based on improved DBSCAN algorithm, includes the following steps:
S1, effective ship track data is extracted from AIS database；
S2, track similarity measurement is carried out using fusion distance MD, obtains fused most short super track between track；
S3, the length that most short super track is calculated using secondary time O (mn)；
S4, fusion distance MD is obtained from the length of track；
S5, the global parameter for determining improved DBSCAN algorithm, i.e. sweep radius Eps and minimum include points
MinPts, parameter (Eps, MinPts) are used to describe the sample distribution tightness degree of neighborhood, wherein Eps describes a certain sample
Neighborhood distance threshold, MinPts describe a certain sample distance be Eps neighborhood in number of samples threshold value；
S6, entire track segment data set is scanned by improved DBSCAN algorithm, finds out the rail of a core
Then mark section carries out traversal queries with the Eps neighborhood of some sample point, obtains the set of cluster；
S7, the spatial movement mode for acquisition, if a new track meets one of vessel motion mode,
Think that the track is a normality track, the motor behavior of target also belongs to a kind of typical operational mode.
Further, in the step S2, merge between track similarity measurement obtains track using fusion distance MD
Most short super track afterwards method particularly includes: it is assumed that two tracks a and b are respectively by the series of points sequence in two-dimensional space
(a1,...,an) and (b1,...,bm) constitute, use d (ai,bj) indicate Euclidean distance on two-dimensional surface between two o'clock, sequence
The most short super track s (a, b) for arranging a and b is the shortest track of length, this makes a and b is the subsequence of s (a, b), most short super rail
The length of mark is indicated with L (a, b).
Further, in the step S3, the specific method of the length of most short super track is calculated using secondary time O (mn)
Are as follows:
S3.1, a [1, i] and b [1, j] respectively (a is set1,a2,...,ai) and (b1,b2,...,bj) super track, respectively
Use Ai jAnd Bi jIt indicates；
S3.2, the method for Dynamic Programming is used to calculate all value A in secondary time O (mn)i jAnd Bi j, most short super rail
The length L (a, b) of mark isWithIn minimum value.
Further, in the step S4, obtain fusion distance MD's from the length of track method particularly includes:
Fusion distance MD (a, b) is obtained from length L (a), the L (b) of track a and b by following formula (1), (2), (3),
And fusion distance MD (a, b) is normalized:
Further, the value of the fusion distance MD (a, b) after the normalized is between (0,1), MD's (a, b)
Value more tends to 0, shows that a is more similar to b, and as a=b, MD (a, b)=0 shows that a is identical as b.
Further, in the step S6, the track for finding a core is being scanned to entire track segment data set
Duan Shi, if only one kernel object, directly choosing the kernel object is in core orbit segment and entire track segment data set
Other non-core object samples are all in the Eps neighborhood of this kernel object；If there is multiple cores object, it is arbitrarily chosen
In one be core orbit segment and any one core for expand and in entire track segment data set to the core orbit segment
Centainly there is an other kernel object in the Eps neighborhood of object.
Further, the specific steps for the set that the step S6 obtains cluster include the following:
S6.1, input trajectory data set D, sweep radius Eps and minimum include points MinPts；
The object p not yet checked in S6.2, Test database；
S6.3, judge whether the number of objects that p includes is not less than MinPts, if being not less than, enter in next step；If being less than,
Then marking p is noise；
S6.4, new cluster C1 is established, and Candidate Set N is added in all the points therein；
S6.5, each of Candidate Set N point q is judged, if some point be core node and with the neighborhood of object p not
It there is no core node in overlapping or overlapping region, is then marked with new class table C2, then return step S6.5 continues pair
Remaining point q judgement；Otherwise, into next step；
S6.6, judge whether some point is be overlapped with the neighborhood of object p again and whether there is core node in overlapping region,
If it exists, then all unlabelled nodes in object q neighborhood are marked with the label C1 of object p, then return step
S6.5 continues to judge remaining point q；Otherwise, into next step；
S6.7, judge whether object q is be overlapped with the neighborhood of the core node of multiple and different labels and in overlapping region is
It is no to have core node, if so, then Candidate Set N, then return step so is added in these objects by these different classes of merging
S6.5 continues to judge remaining point q；
Until point q judgement all in Candidate Set N is finished, the set of cluster is finally exported.
Further, further include step S8 after the step S7: behavior pattern point is carried out to the motion profile of ship
Analysis, the behavior pattern are included at least to upper downstream shipping behavior, berth behavior and deviation behavior.
The beneficial effects of the present invention are:
Ship trajectory clustering analysis method provided by the invention based on improved DBSCAN algorithm uses fusion distance MD
Similitude track is measured, the accuracy of measurement is improved；It is compared to traditional DBSCAN algorithm and reduces disappearing for time
Consumption, while the similarity measurement based on MD can handle Multi-density data set well, have preferable robustness and adaptability,
Also correctly track is classified；Traffic safety monitoring at sea and warning aspect, the present invention pass through the cluster of ship track
Analysis can to berth, the ships behavioural characteristic such as upper downstream shipping, deviation has identification well, be convenient for maritime control portion
Door key monitoring track abnormal area, avoids accident.
Detailed description of the invention
Fig. 1 is the schematic diagram of the ship trajectory clustering analysis method of the present invention based on improved DBSCAN algorithm.
Fig. 2 is the structural schematic diagram for asking most short super track described in the embodiment of the present invention using fusion distance MD algorithm.
Fig. 3 is the flow chart for obtaining the set of cluster described in the embodiment of the present invention by DBSCAN algorithm.
Fig. 4 is that minimum described in the embodiment of the present invention includes to count MinPts to the influence curve figure of number of clusters amount, wherein Fig. 4
(a), Fig. 4 (b), Fig. 4 (c), Fig. 4 (d), Fig. 4 (e), Fig. 4 (f), Fig. 4 (g), Fig. 4 (h) are Eps=0.0025, Eps=respectively
0.0026, Eps=0.0027, Eps=0.0028, Eps=0.0029, Eps=0.0030, Eps=0.0031, Eps=
Minimum includes influence curve figure of the points MinPts to number of clusters amount when 0.0032.
Fig. 5 is that minimum described in the embodiment of the present invention includes influence curve figure of the points MinPts to noise point quantity.
Fig. 6 is the schematic diagram that the section ship of Shanghai Lujiazui described in the embodiment of the present invention clusters track.
Fig. 7 is that histogram is compared in influence of the difference similarity algorithm to cluster result described in the embodiment of the present invention.
Fig. 8 is the speed of a ship or plane distribution histogram of upper downstream shipping described in the embodiment of the present invention.
Fig. 9 is to berth described in the embodiment of the present invention and propagate space-time trajectory figure, and wherein Fig. 9 (a) is to propagate from distally driving towards code
Head is simultaneously stopped at the harbour, and Fig. 9 (b) is that ship starts at harbour and undocks after a period of time.
Figure 10 is ship left and right deviation navigation channel track schematic diagram described in the embodiment of the present invention, and wherein Figure 10 (b) is Figure 10 (a)
The enlarged drawing of middle Blocked portion.
Specific embodiment
For a better understanding of the skilled in the art, being done in the following with reference to the drawings and specific embodiments to the present invention
It is further described, it is following to be merely exemplary that the scope of protection of the present invention is not limited.
As shown in Figure 1, a kind of ship trajectory clustering analysis side based on improved DBSCAN algorithm described in the present embodiment
Method includes the following steps:
Step S1, effective ship track data is extracted from AIS database.
Step S2, track similarity measurement is carried out using fusion distance MD, obtains fused most short super track between track.
Specifically, as shown in Fig. 2, using fusion distance MD carry out track similarity measurement obtain track between it is fused most
Short super track method particularly includes: it is assumed that two tracks a and b are respectively by the series of points sequence (a in two-dimensional space1,...,an)
(b1,...,bm) constitute, use d (ai,bj) indicating Euclidean distance on two-dimensional surface between two o'clock, sequence a and b are most
Short super track s (a, b) is the shortest track of length, this makes a and b is the subsequence of s (a, b), the length L of most short super track
(a, b) is indicated.
Step S3, the length L (a, b) that most short super track is calculated using secondary time O (mn), is specifically comprised the following steps:
S3.1, a [1, i] and b [1, j] respectively (a is set1,a2,...,ai) and (b1,b2,...,bj) super track, respectively
Use Ai jAnd Bi jIt indicates；
S3.2, the method for Dynamic Programming is used to calculate all value A in secondary time O (mn)i jAnd Bi j, most short super rail
The length L (a, b) of mark isWithIn minimum value.
Step S4, fusion distance MD is obtained from the length of track, method particularly includes:
Fusion distance MD (a, b) is obtained from length L (a), the L (b) of track a and b by following formula (1), (2), (3),
And fusion distance MD (a, b) is normalized:
Fusion distance MD (a, b) is more than or equal to any value in L (a) and L (b), passes through being averaged divided by L (a) and L (b)
Value normalizes, and after subtracting 1, the value of the fusion distance MD (a, b) after the normalized is between (0,1), really
Protected MD (a, b) and be consistently greater than 0, that is, ensure MD (a, b) be under rigid motion it is constant, the value of MD (a, b) more tends to 0,
Show that a is more similar to b, as a=b, MD (a, b)=0 shows that a is identical as b, also ensures that MD (a, b) is under rigid motion
Constant.As shown in Fig. 2 right figure, as a and b apart from each other, MD (a, b) is very big, and L (a, b) is than L (a) and L in this case
(b) big more, i.e., as track a and track b intensive sampling from same curves, fusion distance MD (a, b) should be close to
0, because during this time, L (a), L (b) and L (a, b) are generally equal.
Step S5, determine that the global parameter of improved DBSCAN algorithm, i.e. sweep radius Eps and minimum include points
MinPts, parameter (Eps, MinPts) are used to describe the sample distribution tightness degree of neighborhood, wherein Eps describes a certain sample
Neighborhood distance threshold, MinPts describe a certain sample distance be Eps neighborhood in number of samples threshold value；
Step S6, entire track segment data set is scanned by improved DBSCAN algorithm, finds out a core
Orbit segment, traversal queries are then carried out with the Eps neighborhood of some sample point, obtain the set of cluster.
In the step S6, there can be one or more kernel objects in the cluster of DBSCAN, if only one core pair
As, then other non-core object samples are all in the Eps neighborhood of this kernel object in cluster, if there is multiple cores object,
Then centainly there is an other kernel object in the Eps neighborhood of any one kernel object in cluster, otherwise the two cores pair
As can not density it is reachable, the DBSCAN clustering cluster that the collection of all samples is combined into the Eps neighborhood of these kernel objects.
In the concrete realization, the orbit segment for finding a core is scanned to entire track segment data set, if only one core pair
As then directly choosing the kernel object is core orbit segment；If there is multiple cores object, it is any choose wherein one be core
Orbit segment simultaneously expands the core orbit segment.When carrying out traversal queries with the Eps neighborhood of some sample point, in neighborhood
Point first temporarily excludes, and inquires core node in the Eps neighborhood of core node but unlabelled point, and node is avoided to repeat to look into
It askes, to reduce Region Queries number and query time.
As shown in figure 3, the process for the set that the present embodiment obtains cluster by improved DBSCAN algorithm is as follows:
S6.1, input trajectory data set D, sweep radius Eps and minimum include points MinPts；
The object p not yet checked in S6.2, Test database；
S6.3, judge whether the number of objects that p includes is not less than MinPts, if being not less than, enter in next step；If being less than,
Then marking p is noise；
S6.4, new cluster C1 is established, and Candidate Set N is added in all the points therein；
S6.5, each of Candidate Set N point q is judged, if some point be core node and with the neighborhood of object p not
It there is no core node in overlapping or overlapping region, is then marked with new class table C2, then return step S6.5 continues pair
Remaining point q judgement；Otherwise, into next step；
S6.6, judge whether some point is be overlapped with the neighborhood of object p again and whether there is core node in overlapping region,
If it exists, then all unlabelled nodes in object q neighborhood are marked with the label C1 of object p, then return step
S6.5 continues to judge remaining point q；Otherwise, into next step；
S6.7, judge whether object q is be overlapped with the neighborhood of the core node of multiple and different labels and in overlapping region is
It is no to have core node, if so, then Candidate Set N, then return step so is added in these objects by these different classes of merging
S6.5 continues to judge remaining point q；
Until point q judgement all in Candidate Set N is finished, the set of cluster is finally exported, as shown in Figure 6.
Existing DBSCAN algorithm needs artificially select the threshold value MinPts of neighborhood distance threshold Eps and number of samples
Select, and input parameter sensitivity of the improved DBSCAN algorithm to Eps and MinPts described in the present embodiment, that is to say, that Eps and
The value of MinPts directly affects cluster result, by improved DBSCAN algorithm without artificially to neighborhood distance threshold Eps and
The threshold value MinPts of number of samples is selected, but by realizing that result is selected, it is selected just by experimental result
There is selection gist, improves accuracy.
If Fig. 4 (a)-(h) shows the relationship between MinPts and number of clusters amount, trunnion axis is that MinPts from 2 changes to 20,
The longitudinal axis is the quantity of cluster, and each figure indicates different Eps, it can be seen that the quantity of cluster is reduced with the increase of MinPts, when
When MinPts is between 7 and 13, steady state is presented in most of parameter broken line, therefore, using the MinPts in this section
It can produce good Clustering Effect.
If Fig. 5 shows the relationship between MinPts value and noise spot quantity, it can be found that making an uproar when MinPts value becomes larger
Volume increases, and reduces with the increase of radius.Eps with smaller noise number is respectively 0.0030 to 0.0032, but
When Eps is 0.0031 and 0.0032, curve fluctuation is larger, more sensitive to the variation of MinPts, therefore works as Eps=0.0030
When can produce good Clustering Effect.In summary test result, the energy as input parameter MinPts=7, Eps=0.0030
Obtain ideal result.
Fig. 7 describes 5 kinds of algorithms corresponding CP (compactness) and CA (accuracy) after 40,000 AIS data clusters, from reality
Test it can be seen that the similitude based on MD judge on last Clustering Effect better than other 4 kinds of algorithms, this is because DTW with
Euclid is to carry out full matching measurement, i.e., regards track as an entirety, and matched result is needed comprising each track
The distance between corresponding point of point, and noise spot is also required to find corresponding match point, so they are quicker to noise
Sense；And for EDR and LCSS, they are local matching measure, although solving global registration method cannot identify only
There is a similar track in part, but this will lead to algorithm and only focuses on similar portion between track, without considering between similar sub-sequence
Dissimilar part, this may result in the inaccuracy of judgement.Therefore it is calculated using the improvement DBSCAN judged based on MD similarity
Method is able to carry out efficient clustering under the premise of guaranteeing validity, has efficiently accomplished the merger to similar track and has intended
It closes, so that cluster result Stability and dependability with higher.
Step S7, for the spatial movement mode of acquisition, if a new track meets one in vessel motion mode
Kind, then it is assumed that the track is a normality track, and the motor behavior of target also belongs to a kind of typical operational mode.It is usually right
The frequent region in track can be focused first in the excavation of track behavior, and excavates and finds the potential row of target under these trajectory models
For ship Activity recognition method is a kind of based on statistical method, it is intended to find by analysis space track and effectively have
Behavior pattern.
Step S8, Behavior Pattern Analysis carried out to the motion profile of ship, the behavior pattern is included at least to being above lauched
Navigation behavior, berth behavior and deviation behavior.
Specifically, the behavior pattern obtained based on trajectory analysis is as follows:
(1) downstream shipping on
As shown in figure 8, the distribution histogram of cluster 1 and the corresponding all watercraft AIS data speed of a ship or plane of cluster 2.1 speed of a ship or plane histogram of cluster
Distribution center between 10-14, and the distribution center of 2 histogram of cluster is between 6-10.Due to the ship in inland waterway navigation by
Water currents are larger, and upstream shipping is this means that sail against the stream, and by the impact of water flow, the speed of a ship or plane is generally less than the ship sailed against the current
Speed when navigating by water downstream, it is slow compared to the ship average speed sailed with the current, therefore cluster 1 is fair current, i.e. downstream shipping；And cluster
2 is sail against the current, i.e. upstream shipping.By the waters water (flow) direction where actual verification track, Lujiazui section water flow westerly to
East, and the corresponding direction along ng a path of cluster 1 is also from West to East, it was confirmed that analysis result.
(2) it berths behavior
Include the information of this dimension of time in AIS data, be added in the Visualization Model of temporal information, analyst can be with
Extract more abundant ship behavior pattern.The track in cluster 1 and cluster 2 is chosen, two-dimensional space and one-dimensional time is whole
Closing into a three-dimensional system of coordinate indicates, the space coordinate of each x for recording point, y-coordinate corresponding record point, that is, ship warp
Latitude information, the time value of third dimension coordinate then corresponding record point represent ship's speed to expression binding time information indirect in this way,
Stem is to equal related ship attribute.
As shown in attached drawing 9 (a), occur a lot of of vertical ascent in the track in the upper right corner, there is shown position is constant
And time increased trace information, this illustrates the speed of the ship here close to 0, due to being 1 table of cluster in mode in lastrow
Show that ship is downstream shipping, deducibility ship is then stopped at the harbour from harbour is distally driven towards.Similarly it is inferred to attached drawing
Occurs the region of long string time vertical ascent and the constant track in position in 9 (b), and cluster 2 indicates upstream shipping, it will be appreciated that be ship
Start at harbour, then undocks after a period of time.
(3) deviation
The position that ship navigates by water in navigation channel meets in International Rules of the Road Article 9 the rule of " keeping right " navigation,
The ship meeting that in two-way endurance, should be travelled, and must not be travelled near middle line and relatively along outer rim on the right side of navigation channel.However,
Different from the track for the vehicle being strictly limited in land transportation network, marine ships traffic is more like to move in different spaces
Dynamic, because boat trip is only possible to be limited on rough course line, and for security reasons, every ship must be protected with other ships
It holds remote, it is therefore necessary to ship track be monitored in real time, should be issued in time to ship when an abnormal situation occurs pre-
It is alert.
The track in cluster 1 is fitted using least square method, obtains typical motion track.As shown in Fig. 10, have
Shown in body such as Figure 10 (b), the solid line with dot indicates the typical ship trajectory of cluster 1 after cluster, positioned at two sides and with five-pointed star
Solid line is the confidence interval sideline of exemplary trajectory 90%, and the selection of confidence interval is often related with the unidirectional navigation width in navigation channel
System, if distance of monitoring ship during navigation away from exemplary trajectory is greater than the confidence interval of exemplary trajectory, the ship
There will be the behavioural characteristic of deviation, and the ship navigated by water in section belongs to normal/cruise.Band circle in attached drawing 10 (b)
The ship of the solid line left end black triangles solid line of point is greater than left side confidence in navigation distance of the process away from exemplary trajectory
The range in section, such left avertence will lead to from behavior to be bumped against with traveling ship in opposite directions.Solid line right end in attached drawing 10 (b) with dot
The ship of black triangles solid line is greater than the range of right side confidence interval during navigation, deviation boat right in this way
Row will cause Calculation of Ship Grounding's accident if the depth of water is less than drauht.
Above only describes basic principle of the invention and preferred embodiment, those skilled in the art can be according to foregoing description
Many changes and improvements are made, these changes and improvements should be within the scope of protection of the invention.
Claims (8)
1. a kind of ship trajectory clustering analysis method based on improved DBSCAN algorithm, which comprises the steps of:
S1, effective ship track data is extracted from AIS database；
S2, track similarity measurement is carried out using fusion distance MD, obtains fused most short super track between track；
S3, the length that most short super track is calculated using secondary time O (mn)；
S4, fusion distance MD is obtained from the length of track；
S5, the global parameter for determining improved DBSCAN algorithm, i.e. sweep radius Eps and minimum include points MinPts, ginseng
Number (Eps, MinPts) is used to describe the sample distribution tightness degree of neighborhood, wherein Eps describes the neighborhood distance of a certain sample
Threshold value, MinPts describe the distance of a certain sample as the threshold value of number of samples in the neighborhood of Eps；
S6, entire track segment data set is scanned by improved DBSCAN algorithm, finds out the orbit segment of a core,
Then traversal queries are carried out with the Eps neighborhood of some sample point, obtains the set of cluster；
S7, the spatial movement mode for acquisition, if a new track meets one of vessel motion mode, then it is assumed that
The track is a normality track, and the motor behavior of target also belongs to a kind of typical operational mode.
2. the method according to claim 1, wherein carrying out track using fusion distance MD in the step S2
Fused most short super track between similarity measurement acquisition track method particularly includes: it is assumed that two tracks a and b are respectively by two
Series of points sequence (a in dimension space1,...,an) and (b1,...,bm) constitute, use d (ai,bj) indicate between two o'clock in two dimension
The most short super track s (a, b) of Euclidean distance in plane, sequence a and b is the shortest track of length, this make a and b be s (a,
B) length of subsequence, most short super track is indicated with L (a, b).
3. according to the method described in claim 2, it is characterized in that, being calculated most in the step S3 using secondary time O (mn)
The length of short super track method particularly includes:
S3.1, a [1, i] and b [1, j] respectively (a is set1,a2,...,ai) and (b1,b2,...,bj) super track, use respectively
WithIt indicates；
S3.2, the method for Dynamic Programming is used to calculate all values in secondary time O (mn)WithMost short super track
Length L (a, b) isWithIn minimum value.
4. according to the method described in claim 3, it is characterized in that, being merged from the length of track in the step S4
Distance MD's method particularly includes:
Fusion distance MD (a, b) is obtained from length L (a), the L (b) of track a and b by following formula (1), (2), (3), and right
Fusion distance MD (a, b) is normalized:
5. according to the method described in claim 4, it is characterized in that, fusion distance MD (a, b) after the normalized
For value between (0,1), the value of MD (a, b) more tends to 0, shows that a is more similar to b, as a=b, MD (a, b)=0, show a and
B is identical.
6. the method according to claim 1, wherein in the step S6, to entire track segment data set into
When row scans for the orbit segment of a core, if only one kernel object, directly choosing the kernel object is core rail
Mark section；If there is multiple cores object, any selection is wherein core orbit segment and expands the core orbit segment for one.
7. the method according to claim 1, wherein the specific steps for the set that the step S6 obtains cluster include
It is as follows:
S6.1, input trajectory data set D, sweep radius Eps and minimum include points MinPts；
The object p not yet checked in S6.2, Test database；
S6.3, judge whether the number of objects that p includes is not less than MinPts, if being not less than, enter in next step；If being less than, mark
Note p is noise；
S6.4, new cluster C1 is established, and Candidate Set N is added in all the points therein；
S6.5, each of Candidate Set N point q is judged, if some point is core node and not be overlapped with the neighborhood of object p
Or core node is not present in overlapping region, then it is marked with new class table C2, then return step S6.5 continues to residue
Point q judgement；Otherwise, into next step；
S6.6, judge whether some point is be overlapped with the neighborhood of object p and whether there is core node in overlapping region again, if depositing
, then all unlabelled nodes in object q neighborhood are marked with the label C1 of object p, then return step S6.5 after
It is continuous that remaining point q is judged；Otherwise, into next step；
S6.7, judge whether object q is be overlapped with the neighborhood of the core node of multiple and different labels and whether has in overlapping region
Core node, if so, then Candidate Set N, then return step S6.5 so is added in these objects by these different classes of merging
Continue to judge remaining point q；
Until point q judgement all in Candidate Set N is finished, the set of cluster is finally exported.
8. method according to claim 1-7, which is characterized in that further include step S8 after the step S7:
Behavior Pattern Analysis is carried out to the motion profile of ship, the behavior pattern is included at least to upper downstream shipping behavior, berth row
For with deviation behavior.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910521851.3A CN110309383B (en) | 2019-06-17 | 2019-06-17 | Ship track clustering analysis method based on improved DBSCAN algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910521851.3A CN110309383B (en) | 2019-06-17 | 2019-06-17 | Ship track clustering analysis method based on improved DBSCAN algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110309383A true CN110309383A (en) | 2019-10-08 |
| CN110309383B CN110309383B (en) | 2021-07-13 |
Family
ID=68077311
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910521851.3A Active CN110309383B (en) | 2019-06-17 | 2019-06-17 | Ship track clustering analysis method based on improved DBSCAN algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110309383B (en) |
Cited By (30)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110941278A (en) * | 2019-12-20 | 2020-03-31 | 交控科技股份有限公司 | In-station dynamic security analysis method |
| CN111539454A (en) * | 2020-03-30 | 2020-08-14 | 武汉理工大学 | Vehicle track clustering method and system based on meta-learning |
| CN111582380A (en) * | 2020-05-09 | 2020-08-25 | 中国人民解放军92493部队试验训练总体研究所 | Ship track density clustering method and device based on space-time characteristics |
| CN111858565A (en) * | 2020-06-10 | 2020-10-30 | 北京航空航天大学 | Multi-dimensional high-precision track intelligent prediction method based on segment clustering |
| CN111897805A (en) * | 2020-06-24 | 2020-11-06 | 东南大学 | Hot spot path mining method based on longest common sub-track density clustering |
| CN111985529A (en) * | 2020-07-06 | 2020-11-24 | 东南大学 | Ship AIS track hybrid clustering method and system |
| CN112187550A (en) * | 2020-10-16 | 2021-01-05 | 温州职业技术学院 | Log analysis method based on density peak value multi-attribute clustering |
| CN112487116A (en) * | 2020-11-19 | 2021-03-12 | 江苏锐天信息科技有限公司 | Ship track analysis method based on AIS big data |
| CN112488196A (en) * | 2020-11-30 | 2021-03-12 | 青岛科技大学 | Track data parameter self-adaptive clustering method |
| CN112541591A (en) * | 2020-11-26 | 2021-03-23 | 中远海运科技股份有限公司 | Parking position identification method and platform |
| CN112556698A (en) * | 2020-12-17 | 2021-03-26 | 广州博进信息技术有限公司 | Ship behavior change and abnormality recognition method, equipment and medium |
| CN112613455A (en) * | 2020-12-29 | 2021-04-06 | 武汉理工大学 | Ship anchoring gathering area identification method and device |
| CN112733935A (en) * | 2021-01-08 | 2021-04-30 | 上海大学 | Fishing boat fishing mode prediction method based on Stacking algorithm |
| CN112749743A (en) * | 2021-01-04 | 2021-05-04 | 清华大学 | Track space-time clustering method, system and storage device |
| CN112906782A (en) * | 2021-02-07 | 2021-06-04 | 江西科技学院 | Track static inspection historical data matching method based on DTW and least square estimation |
| 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 |
| CN113032502A (en) * | 2021-02-09 | 2021-06-25 | 北京工业大学 | Ship anomaly detection method based on improved track segment DBSCAN clustering |
| CN113032378A (en) * | 2021-03-05 | 2021-06-25 | 北京工业大学 | Ship behavior pattern mining method based on clustering algorithm and pattern mining |
| CN113065604A (en) * | 2021-04-15 | 2021-07-02 | 北京理工大学 | Air target grouping method based on DTW-DBSCAN algorithm |
| CN113129644A (en) * | 2021-03-05 | 2021-07-16 | 中国人民解放军国防科技大学 | Ship three-dimensional characteristic route extraction method based on spherical scanning lines |
| CN113298195A (en) * | 2021-07-27 | 2021-08-24 | 中国电子科技集团公司第十五研究所 | Method and device for generating classical trajectory of offshore target and storage medium |
| CN113312438A (en) * | 2021-03-09 | 2021-08-27 | 中南大学 | Sea target position prediction method integrating airline extraction and trend judgment |
| CN113435639A (en) * | 2021-06-23 | 2021-09-24 | 交通运输部规划研究院 | Port water area planning method and system integrating AIS data mining |
| CN113505836A (en) * | 2021-07-14 | 2021-10-15 | 金睛兽数字科技(重庆)有限公司 | Ship route automatic generation method based on big data |
| CN113521749A (en) * | 2021-07-15 | 2021-10-22 | 珠海金山网络游戏科技有限公司 | Abnormal account detection model training method and abnormal account detection method |
| CN113689739A (en) * | 2021-08-24 | 2021-11-23 | 重庆大学 | Historical data-based judgment method for controlling river reach ship to enter or exit water |
| CN113887590A (en) * | 2021-09-22 | 2022-01-04 | 中国电子科技集团公司第二十九研究所 | Target typical track and area analysis method |
| CN114118243A (en) * | 2021-11-18 | 2022-03-01 | 中交疏浚技术装备国家工程研究中心有限公司 | Track data-based construction track identification method for trailing suction hopper dredger |
| CN115270920A (en) * | 2022-06-21 | 2022-11-01 | 中国人民解放军91977部队 | Ship target classical trajectory generation method based on density space clustering |
| CN115828120A (en) * | 2023-02-17 | 2023-03-21 | 山东科技大学 | Self-adaptive identification method and system for ship traffic behavior pattern and computer equipment |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104899263A (en) * | 2015-05-22 | 2015-09-09 | 华中师范大学 | Ship trajectory mining, analysis and monitoring method based on specific region |
| CN107679558A (en) * | 2017-09-19 | 2018-02-09 | 电子科技大学 | A kind of user trajectory method for measuring similarity based on metric learning |
| CN108304879A (en) * | 2018-02-05 | 2018-07-20 | 济南浪潮高新科技投资发展有限公司 | A kind of ship method of trajectory clustering and device |
| CN109214462A (en) * | 2018-09-25 | 2019-01-15 | 东北大学 | A kind of spatial data on-line talking method based on distributed incremental DBSCAN algorithm |
| WO2019014184A1 (en) * | 2017-07-14 | 2019-01-17 | Allstate Insurance Company | Distributed data processing systems for processing remotely captured sensor data |
| CN109447880A (en) * | 2018-10-10 | 2019-03-08 | 广州市勤思网络科技有限公司 | Road network abstracting method waterborne and system |
| CN109448442A (en) * | 2018-10-12 | 2019-03-08 | 天津大学 | The efficient monitoring method of more foul danger levels in a kind of sea area |
| CN109543715A (en) * | 2018-10-23 | 2019-03-29 | 武汉理工大学 | A kind of ship air route is extracted and the method for track deviation detection |
| CN109840358A (en) * | 2019-01-09 | 2019-06-04 | 上海海洋大学 | A kind of track segmentation method based on track time-domain difference |
-
2019
- 2019-06-17 CN CN201910521851.3A patent/CN110309383B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104899263A (en) * | 2015-05-22 | 2015-09-09 | 华中师范大学 | Ship trajectory mining, analysis and monitoring method based on specific region |
| WO2019014184A1 (en) * | 2017-07-14 | 2019-01-17 | Allstate Insurance Company | Distributed data processing systems for processing remotely captured sensor data |
| CN107679558A (en) * | 2017-09-19 | 2018-02-09 | 电子科技大学 | A kind of user trajectory method for measuring similarity based on metric learning |
| CN108304879A (en) * | 2018-02-05 | 2018-07-20 | 济南浪潮高新科技投资发展有限公司 | A kind of ship method of trajectory clustering and device |
| CN109214462A (en) * | 2018-09-25 | 2019-01-15 | 东北大学 | A kind of spatial data on-line talking method based on distributed incremental DBSCAN algorithm |
| CN109447880A (en) * | 2018-10-10 | 2019-03-08 | 广州市勤思网络科技有限公司 | Road network abstracting method waterborne and system |
| CN109448442A (en) * | 2018-10-12 | 2019-03-08 | 天津大学 | The efficient monitoring method of more foul danger levels in a kind of sea area |
| CN109543715A (en) * | 2018-10-23 | 2019-03-29 | 武汉理工大学 | A kind of ship air route is extracted and the method for track deviation detection |
| CN109840358A (en) * | 2019-01-09 | 2019-06-04 | 上海海洋大学 | A kind of track segmentation method based on track time-domain difference |
Non-Patent Citations (2)
| Title |
|---|
| 肖潇: ""基于AIS信息的船舶轨迹聚类模型研究"", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 陈加顺: ""海洋环境下聚类算法的研究"", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (43)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110941278A (en) * | 2019-12-20 | 2020-03-31 | 交控科技股份有限公司 | In-station dynamic security analysis method |
| CN111539454B (en) * | 2020-03-30 | 2023-04-18 | 武汉理工大学 | Vehicle track clustering method and system based on meta-learning |
| CN111539454A (en) * | 2020-03-30 | 2020-08-14 | 武汉理工大学 | Vehicle track clustering method and system based on meta-learning |
| CN111582380A (en) * | 2020-05-09 | 2020-08-25 | 中国人民解放军92493部队试验训练总体研究所 | Ship track density clustering method and device based on space-time characteristics |
| CN111582380B (en) * | 2020-05-09 | 2024-05-24 | 中国人民解放军92493部队试验训练总体研究所 | Ship track density clustering method and device based on space-time characteristics |
| CN111858565A (en) * | 2020-06-10 | 2020-10-30 | 北京航空航天大学 | Multi-dimensional high-precision track intelligent prediction method based on segment clustering |
| CN111858565B (en) * | 2020-06-10 | 2023-04-21 | 北京航空航天大学 | Multi-dimensional high-precision track intelligent prediction method based on line segment clustering |
| CN111897805A (en) * | 2020-06-24 | 2020-11-06 | 东南大学 | Hot spot path mining method based on longest common sub-track density clustering |
| CN111985529A (en) * | 2020-07-06 | 2020-11-24 | 东南大学 | Ship AIS track hybrid clustering method and system |
| CN112187550A (en) * | 2020-10-16 | 2021-01-05 | 温州职业技术学院 | Log analysis method based on density peak value multi-attribute clustering |
| CN112487116A (en) * | 2020-11-19 | 2021-03-12 | 江苏锐天信息科技有限公司 | Ship track analysis method based on AIS big data |
| CN112541591A (en) * | 2020-11-26 | 2021-03-23 | 中远海运科技股份有限公司 | Parking position identification method and platform |
| CN112488196A (en) * | 2020-11-30 | 2021-03-12 | 青岛科技大学 | Track data parameter self-adaptive clustering method |
| CN112556698A (en) * | 2020-12-17 | 2021-03-26 | 广州博进信息技术有限公司 | Ship behavior change and abnormality recognition method, equipment and medium |
| CN112556698B (en) * | 2020-12-17 | 2023-12-15 | 广州博进信息技术有限公司 | Ship behavior change and abnormality identification method, equipment and medium |
| CN112613455A (en) * | 2020-12-29 | 2021-04-06 | 武汉理工大学 | Ship anchoring gathering area identification method and device |
| CN112749743A (en) * | 2021-01-04 | 2021-05-04 | 清华大学 | Track space-time clustering method, system and storage device |
| CN112733935A (en) * | 2021-01-08 | 2021-04-30 | 上海大学 | Fishing boat fishing mode prediction method based on Stacking algorithm |
| CN112906782A (en) * | 2021-02-07 | 2021-06-04 | 江西科技学院 | Track static inspection historical data matching method based on DTW and least square estimation |
| CN112906782B (en) * | 2021-02-07 | 2024-01-26 | 江西科技学院 | Track static inspection historical data matching method based on DTW and least square estimation |
| CN113032502B (en) * | 2021-02-09 | 2024-02-02 | 北京工业大学 | Ship anomaly detection method based on improved track segment DBSCAN clustering |
| CN113032502A (en) * | 2021-02-09 | 2021-06-25 | 北京工业大学 | Ship anomaly detection method based on improved track segment DBSCAN clustering |
| CN113129644A (en) * | 2021-03-05 | 2021-07-16 | 中国人民解放军国防科技大学 | Ship three-dimensional characteristic route extraction method based on spherical scanning lines |
| CN113032378A (en) * | 2021-03-05 | 2021-06-25 | 北京工业大学 | Ship behavior pattern mining method based on clustering algorithm and pattern mining |
| CN113129644B (en) * | 2021-03-05 | 2022-03-08 | 中国人民解放军国防科技大学 | Ship three-dimensional characteristic route extraction method based on spherical scanning lines |
| CN113312438A (en) * | 2021-03-09 | 2021-08-27 | 中南大学 | Sea target position prediction method integrating airline extraction and trend judgment |
| CN113312438B (en) * | 2021-03-09 | 2023-09-15 | 中南大学 | Marine target position prediction method integrating route extraction and trend judgment |
| CN112906830B (en) * | 2021-04-14 | 2023-12-19 | 武汉理工大学 | Automatic generation method of ship optimal route based on AIS big data |
| 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 |
| CN113065604A (en) * | 2021-04-15 | 2021-07-02 | 北京理工大学 | Air target grouping method based on DTW-DBSCAN algorithm |
| CN113435639A (en) * | 2021-06-23 | 2021-09-24 | 交通运输部规划研究院 | Port water area planning method and system integrating AIS data mining |
| CN113505836B (en) * | 2021-07-14 | 2022-11-15 | 金睛兽数字科技(重庆)有限公司 | Ship route automatic generation method based on big data |
| CN113505836A (en) * | 2021-07-14 | 2021-10-15 | 金睛兽数字科技(重庆)有限公司 | Ship route automatic generation method based on big data |
| CN113521749A (en) * | 2021-07-15 | 2021-10-22 | 珠海金山网络游戏科技有限公司 | Abnormal account detection model training method and abnormal account detection method |
| CN113521749B (en) * | 2021-07-15 | 2024-02-13 | 珠海金山数字网络科技有限公司 | Abnormal account detection model training method and abnormal account detection method |
| CN113298195A (en) * | 2021-07-27 | 2021-08-24 | 中国电子科技集团公司第十五研究所 | Method and device for generating classical trajectory of offshore target and storage medium |
| CN113689739A (en) * | 2021-08-24 | 2021-11-23 | 重庆大学 | Historical data-based judgment method for controlling river reach ship to enter or exit water |
| CN113887590B (en) * | 2021-09-22 | 2023-06-09 | 中国电子科技集团公司第二十九研究所 | Target typical track and area analysis method |
| CN113887590A (en) * | 2021-09-22 | 2022-01-04 | 中国电子科技集团公司第二十九研究所 | Target typical track and area analysis method |
| CN114118243B (en) * | 2021-11-18 | 2023-07-07 | 中交疏浚技术装备国家工程研究中心有限公司 | Construction track identification method for trailing suction hopper dredger based on track data |
| CN114118243A (en) * | 2021-11-18 | 2022-03-01 | 中交疏浚技术装备国家工程研究中心有限公司 | Track data-based construction track identification method for trailing suction hopper dredger |
| CN115270920A (en) * | 2022-06-21 | 2022-11-01 | 中国人民解放军91977部队 | Ship target classical trajectory generation method based on density space clustering |
| CN115828120A (en) * | 2023-02-17 | 2023-03-21 | 山东科技大学 | Self-adaptive identification method and system for ship traffic behavior pattern and computer equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110309383B (en) | 2021-07-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110309383A (en) | Ship trajectory clustering analysis method based on improved DBSCAN algorithm | |
| Liu et al. | Adaptive douglas-peucker algorithm with automatic thresholding for AIS-based vessel trajectory compression | |
| CN113032502B (en) | Ship anomaly detection method based on improved track segment DBSCAN clustering | |
| CN112699315B (en) | AIS data-based ship abnormal behavior detection method | |
| CN113312438B (en) | Marine target position prediction method integrating route extraction and trend judgment | |
| CN106710313A (en) | Method and system for ship in bridge area to actively avoid collision based on laser three-dimensional imaging technique | |
| CN110210352A (en) | Ship track method for detecting abnormality based on navigation channel model | |
| Zhang et al. | An interpretable knowledge-based decision support method for ship collision avoidance using AIS data | |
| CN111157982A (en) | Intelligent ship and shore cooperative target tracking system and method based on shore-based radar | |
| CN110188093A (en) | A kind of data digging system being directed to AIS information source based on big data platform | |
| Yan et al. | Vessel movement analysis and pattern discovery using density-based clustering approach | |
| CN112966332B (en) | Conflict detection method based on multi-ship motion uncertainty, memory and processor | |
| Han et al. | Big data-driven automatic generation of ship route planning in complex maritime environments | |
| CN103531043A (en) | Point-to-point ship sailing time estimation method based on ship route matching | |
| CN114066354A (en) | Intelligent air route recommendation method and system based on global ship historical track | |
| Chang et al. | Vessel traffic analysis for maritime intelligent transportation system | |
| Liu et al. | A data mining method to extract traffic network for maritime transport management | |
| CN114139642A (en) | Ship route extraction method combining ship journey semantic object and graph theory | |
| Xu et al. | Research on ship trajectory extraction based on multi-attribute DBSCAN optimisation algorithm | |
| Sánchez-Heres | Simplification and event identification for AIS trajectories: the equivalent passage plan method | |
| Zhang et al. | A data-driven approach for ship-bridge collision candidate detection in bridge waterway | |
| Zhou et al. | Compressing AIS trajectory data based on the multi-objective peak Douglas–Peucker algorithm | |
| Stein et al. | Model-based rail detection in mobile laser scanning data | |
| Zhang et al. | Map-matching approach based on link factor and hidden Markov model | |
| Kwun et al. | Prediction of vessel arrival time using auto identification system data |
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 |