DE102019216449A1 - Method and device for creating a ship route map - Google Patents

Abstract

The invention relates to a method for creating a ship route map (20), tracking data (10) from ships being received or recorded by means of an input device (2), the tracking data (10) for each of the ships in at least one ordered set of geographical positions ( 41) with respectively associated travel directions (42), the tracking data (10) of the ships being combined to form clusters (40) by means of a data processing device (3) depending on a distance determined on the basis of a distance measure and at least one predetermined limit value Distance measure takes into account both a position difference between the individual geographic positions (41) and a travel direction difference between the respective associated travel directions (42), and a geographic position (41) and a travel direction (42) are or will be assigned to each cluster (40) , and wherein the clusters (40) ü Directed edges (43, 44) are linked to one another or are linked, and the set of clusters (40) linked via the directed edges (43, 43) generated in this way by means of an output device (4) as a ship route map data record (41 ) provided. The invention also relates to a device (1) for creating a ship route map (20).

Description

The invention relates to a method and a device for creating a ship route map.

Worldwide shipping traffic has been increasing for years. In the commercial sector in particular, it is common to use forecasting methods to predict an estimated time of arrival (ETA). In order to be able to carry out such a prediction, ship route maps are necessary in which the ship routes traveled by ships are stored.

It is known to create ship route maps from tracking data of the ships. Such tracking data is obtained, for example, via the worldwide Automatic Identification System (AIS), which is installed in the ships themselves. The ship route maps are created manually in this case, however, by combining the tracking data into ship routes at locations where the density of the data points is particularly high.

Out T. Zhang et al., BIRCH: an efficient data clustering method for very large databases, ACM Sigmod Record, 1996 , a method is known to combine data into clusters.

The invention is based on the object of improving a method and a device for creating a ship route map.

The object is achieved according to the invention by a method with the features of claim 1 and a device with the features of claim 10. Advantageous refinements of the invention emerge from the subclaims.

In particular, a method for creating a ship route map is provided, wherein tracking data from ships are received or recorded by means of an input device, the tracking data for each of the ships each comprising at least one ordered set of geographical positions with respectively associated directions of travel, the tracking data of the ships are combined into clusters by means of a data processing device depending on a distance determined on the basis of a distance measure and at least one predetermined limit value, the distance measure taking into account both a position difference between the individual geographical positions and a travel direction difference between the respective associated travel directions, and each cluster taking into account one geographical position and a direction of travel is assigned or is assigned, and wherein the clusters are linked to one another via directed edges or are linked, and the set of clusters linked via the directed edges generated in this way is provided as a ship route map data record by means of an output device.

Furthermore, in particular a device for creating a ship route map is created, comprising an input device, a data processing device, and an output device, the input device being configured to receive or record tracking data from ships, the tracking data for each of the ships in at least one ordered set at geographical positions with respectively associated travel directions, and wherein the data processing device is set up to combine the tracking data of the ships into clusters as a function of a distance determined on the basis of a distance measure and at least one predetermined limit value, the distance measure being both a position difference between the individual geographical Positions as well as a travel direction difference between the respective associated travel directions are taken into account, and each cluster has a geographical position and a journey Direction is assigned or is assigned, and to link the clusters to one another via directed edges, and wherein the output device is set up to provide the quantity of clusters linked via the directed edges generated in this way as a ship route map data set.

The method and the device make it possible to create a ship route map on the basis of tracking data provided, for example in the form of historical AIS data. For this purpose, the tracking data from several ships are combined into clusters, this being done on the basis of both the geographical positions contained in the tracking data and the respective associated directions of travel. In order to take into account both the geographical positions and the associated driving directions, a corresponding distance measure is selected. Starting from a geographical position and the associated direction of travel in the tracking data of a ship and the geographical position and direction of travel assigned to a cluster, it is determined whether the distance determined by means of the distance measure is below a predetermined limit value or not. If this is the case, the data point becomes the Tracking data, that is, the geographical position and the associated direction of travel, assigned to the cluster, otherwise not. A geographical position and a direction of travel are assigned to each cluster. The assigned geographical position is in particular a geographical position averaged over all geographical positions assigned to the cluster via the tracking data. In the same way, in particular, the direction of travel is also averaged over all directions of travel assigned to the cluster. If there is not yet a cluster for a geographical position in the tracking data to which the observed data point can be assigned, a new cluster is generated based on the observed geographical position of the tracking data and the associated direction of travel. The individual clusters are linked to one another via directed edges. The link takes place here in particular via a chronological sequence of the individual geographical positions in the respective tracking data of the ships. If, for example, the tracking data of two ships that have traveled parallel to one another are combined into clusters, the clusters are linked via the directed edges, in particular in the direction in which the individual geographical positions are linked to one another (temporally) or those of one (temporal) ) Corresponds to the sequence of geographical positions. If the tracking data of all ships are combined in this way to form linked clusters, the generated quantity of the clusters linked via the directed edges is provided as a ship route map data record. In particular, the ship route data record is output, for example in the form of a digital data packet at an interface provided for this purpose.

One advantage of the method and the device is that a ship route map can be created and provided in an automated manner. By means of the disclosed method it is possible in particular to identify and define ship routes as “one-way streets”, that is to say ship routes on which the ships predominantly or exclusively have one direction of travel. This represents in particular an advantage over ship route maps which were created only on the basis of a density of geographical positions without taking a direction of travel into account.

The tracking data include, in particular, for each of the ships, in each case at least one, in particular chronologically, ordered amount of data points. Each data point comprises a tuple formed from a geographical position and an associated direction of travel. In this context, ordered is intended to mean, in particular, that individual geographical positions are linked to one another, that is to say that the ship under consideration has passed through in a predetermined sequence. Such an ordered set can also be referred to as the ship's trajectory. A direction of travel is assigned to each geographical position, which indicates the direction in which the ship under consideration has passed through the geographical position. The geographical position is indicated, for example, in the form of geographical coordinates (longitude and latitude). The direction of travel is indicated, for example, by a value between 0 and 360 °. The tracking data are in particular AIS data.

The specified limit value is specified in particular in the form of a distance (e.g. 1 km, 2 km, 10 km, etc.). Depending on whether the distance determined by means of the distance measure is smaller than the predefined limit value, that is to say smaller than the predefined distance, or not, an assignment to a cluster is made or not.

In particular, it is provided that the BIRCH method is used to combine the tracking data into clusters, the distance measure used here being adapted according to the method in such a way that both a position difference between the individual geographical positions and a travel direction difference between the respective associated travel directions are taken into account.

Parts of the device can be designed individually or collectively as a combination of hardware and software, for example as program code that is executed on a microcontroller or microprocessor.

In one embodiment it is provided that a weighting value is or is assigned to the directed edges between the clusters, the weighting value between two interconnected clusters being increased for each unidirectional link between these clusters due to tracking data, and the weighting value between two with one another linked clusters is reduced for each additional link between these clusters due to tracking data that runs in the opposite direction. The weighting value here corresponds to a valency of the link or connection between two clusters and is specified, for example, as a natural number or whole number. As a result, a number of ships that have traveled the same ship route can be determined or assigned. The more ships have traveled the same route, the more geographical positions were assigned to the clusters in the same way, in particular in the same direction of travel, assigned. The weighting value of the directed edge is increased, for example by the value 1, for each ship that has traveled along a part of the ship route mapped via two clusters linked to one another via a directed edge. If, on the other hand, a ship is traveling in the opposite direction, the weighting value for the directed edge is reduced, for example by the value 1. This can in particular prevent the creation of back and forth links between adjacent clusters in which geographical positions are assigned multiple times to adjacent clusters (connections in the form of A → B, B → A, A → B). On the basis of the weighting values of the directed edges between the clusters, it can be determined in the created ship route map how often the ship route is or was driven in comparison to other ship routes.

In one embodiment it is provided that at least one static geographical position is received by means of the input device, two clusters being generated for the at least one static geographical position by means of the data processing device, and the two clusters generated with clusters that are located in a predetermined radius around the at least one static geographical position are linked, the associated directed edges of these links having different directions for the two clusters. In this way, ports can be linked to the ship routes shown in the ship route map via the linked clusters. The static geographic coordinates for a port are fixed here. In order to enable both a direction of travel in the direction of the port and a direction of travel away from the port, since a cluster can only have one direction of travel, two clusters are generated. The generated clusters are then linked to clusters in the vicinity. It can be provided that a search radius for clusters in the vicinity of the geographical position of a port is gradually increased (e.g. in steps of 2 km, 4 km, 8 km) if no cluster is initially found.

wobei d_geog eine Positionsdifferenz bzw. ein Abstand zwischen den geographischen Koordinaten (meist eines Datenpunktes der Trackingdaten und eines Clusters), F ein Faktor, um einen Einfluss einer Fahrtrichtungsdifferenz zu berücksichtigen, und d_Fahrtrichtung eine Fahrtrichtungsdifferenz ist. Haben zwei geographische Positionen (z.B. von einem Datenpunkt in den Trackingdaten und einem bereits vorhandenen Cluster) beispielsweise einen Abstand von 1000 m mit einer zugehörigen Fahrtrichtungsdifferenz von 5°, so ergibt sich mit dieser Funktion: $$\text{D}=1000\text{ m}+\text{F}*5°$$

In one embodiment it is provided that a weighted sum between a respective position difference and the associated travel direction difference is formed as a distance measure. For example, it can be provided that the following function is used as the distance measure: $$\text{D.}=\text{d}\_\text{geog}+\text{F.}*\text{d}\_\text{Direction of travel},$$

where d_geog is a position difference or a distance between the geographic coordinates (usually a data point of the tracking data and a cluster), F a factor to take into account an influence of a travel direction difference, and d_Drive direction is a travel direction difference. If two geographical positions (e.g. from a data point in the tracking data and an already existing cluster) have a distance of 1000 m with an associated travel direction difference of 5 °, this function results in: $$\text{D.}=1000\text{m}+\text{F.}*5°$$

ergeben. Dies ist ein einfaches Distanzmaß, welches jedoch bei Unterschieden in der Fahrtrichtung, die im Bereich eines Wertes von 180° liegen, das heißt bei denen die Fahrtrichtungen in entgegengesetzte Richtungen verlaufen, problematisch werden kann, insbesondere, wenn der mindestens eine Grenzwert zum Herabsetzen einer Clusterauflösung relativ groß gewählt wird.In order to determine a weighting of the difference in the directions of travel, it can be provided, for example, that F = (1000 m / 10 °). A travel direction difference of 10 ° then corresponds to a distance between the geographical positions of 1000 m $$\text{D.}=1500\text{m}$$

surrender. This is a simple distance measure, which, however, can be problematic in the case of differences in the direction of travel that are in the range of a value of 180 °, i.e. where the directions of travel run in opposite directions, especially if the at least one limit value is used to reduce a cluster resolution is chosen to be relatively large.

wobei x_cluster und y_cluster jeweils die geographischen Koordinaten des Clusters, x_data und y_data jeweils die geographischen Koordinaten eines Datenpunktes der Trackingdaten und dir_cluster und dir_data die Fahrtrichtungen des Clusters bzw. des Datenpunkts bezeichnen. Die Funktion d_equirectangular() wird dann wie folgt gewählt: $$\text{d}\_\text{equirectangular}=\text{r}\_\text{earth}*\text{sqrt}\left(\left(\text{lat}1-\text{lat}2\right)2+\left(\text{d}\left(\text{Ion}\mathrm{1,}\text{Ion}2\right)*\text{cos}\left(\left(\text{lat}1+\text{lat}2\right)/2\right)\right)2\right),$$

wobei jeweils die Abstände in den jeweiligen geographischen Koordinaten mit Hilfe von zugehörigen Breitengraden („lat“) und Längengraden („Ion“) berechnet werden und r_earth der Radius der Erde ist. Ferner: $$\text{d}\_\text{angle}=\text{tan}\left(\text{diff}/2\right)\mathrm{2,}$$

wobei diff die Fahrtrichtungsdifferenz bezeichnet und einen Wert zwischen 0° und 180° annehmen kann. Die tan^2-Funktion sorgt dafür, dass die resultierenden Werte zwischen 0 (bei diff = 0°) und unendlich (bei diff = 180°) liegen, das heißt dass eine Gewichtung der Fahrtrichtungsdifferenz in dem Distanzmaß überproportional in Abhängigkeit der Fahrtrichtungsdifferenz erhöht wird.In a further embodiment it is therefore provided that the distance measure is selected in such a way that a travel direction difference in the determined distance is taken into account disproportionately as a function of the determined travel direction difference. In this way, a large difference in the direction of travel can be taken into greater account. For example, the following function can be used as a distance measure: $$\begin{array}{l}\text{D.}=\text{d}\_\text{equirectangular}\left(\text{x}\_\text{cluster},\text{y}\_\text{cluster},\text{x}\_\text{data},\text{y}\_\text{data}\right)\\ +\text{F.}*\text{d}\_\text{angle}\left(\text{to you}\_\text{cluster},\text{to you}\_\text{data}\right),\end{array}$$

where x_cluster and y_cluster each denote the geographical coordinates of the cluster, x_data and y_data each denote the geographical coordinates of a data point of the tracking data and dir_cluster and dir_data denote the directions of travel of the cluster or of the data point. The function d_equirectangular () is then chosen as follows: $$\text{d}\_\text{equirectangular}=\text{r}\_\text{earth}*\text{sqrt}\left(\left(\text{lat}1-\text{lat}2\right)2+\left(\text{d}\left(\text{ion}\mathrm{1,}\text{ion}2\right)*\text{cos}\left(\left(\text{lat}1+\text{lat}2\right)/2\right)\right)2\right),$$

whereby the distances in the respective geographic coordinates are calculated with the help of the associated latitudes ("lat") and longitudes ("ion") and r_earth is the radius of the earth. Further: $$\text{d}\_\text{angle}=\text{tan}\left(\text{diff}/2\right)\mathrm{2,}$$

where diff denotes the difference in the direction of travel and can assume a value between 0 ° and 180 °. The tan ^ 2 function ensures that the resulting values are between 0 (at diff = 0 °) and infinite (at diff = 180 °), i.e. a weighting of the travel direction difference in the distance measure is increased disproportionately depending on the travel direction difference .

wobei als dist_threshold und angle_threshold jeweils Grenzwerte für eine Positionsdifferenz zwischen den geographischen Positionen bzw. eine Fahrtrichtungsdifferenz zwischen den Fahrtrichtungen gewählt werden, die einander entsprechen sollen. Soll beispielsweise eine Fahrtrichtungsdifferenz von 10° in dem Distanzmaß einem Abstand von 1000 m entsprechen, werden die Grenzwerte wie folgt gewählt: $$\begin{array}{l}\text{dist}\_\text{threshold}=1000\text{ m},\\ \text{angle}\_\text{threshold}=10°.\end{array}$$

In einer Ausführungsform ist vorgesehen, dass der mindestens eine vorgegebene Grenzwert in Abhängigkeit der geographischen Position gewählt ist oder gewählt wird. Dies ermöglicht es, eine Clusterauflösung, das heißt einen Abstand zwischen Clustern flexibel vorzugeben. Einfach ausgedrückt bedeutet dies, dass je größer der Grenzwert gewählt wird, desto geringer auch eine Anzahl von Clustern pro Fläche ist, das heißt desto geringer ist die Clusterauflösung. Beispielsweise kann vorgesehen sein, den mindestens einen Grenzwert auf dem offenen Meer größer zu wählen, sodass die Cluster dort einen größeren Abstand zueinander haben. In Küstennähe, insbesondere in stark befahrenen Bereichen, kann der Grenzwert hingegen deutlich kleiner gewählt werden, sodass in diesen Bereichen ein Abstand zwischen den Clustern geringer ist und eine Clusterauflösung sich hierdurch erhöht. Auf diese Weise ist es möglich, eine Datenmenge der erstellten Schiffsroutenkarte zu reduzieren, trotzdem aber das Erzeugen und Bereitstellen auch von sehr großen Schiffsroutenkarten, insbesondere von einer globalen Schiffsroutenkarte, zu ermöglichen.The factor F is selected as follows, for example: $$\text{F.}=\text{dist}\_\text{threshold}/\left(\text{tan}\left(\text{angle}\_\text{threshold}/2\right)\right)\mathrm{2,}$$

limit values for a position difference between the geographical positions or a travel direction difference between the travel directions, which should correspond to one another, are selected as the dist_threshold and angle_threshold. If, for example, a travel direction difference of 10 ° is to correspond to a distance of 1000 m, the limit values are selected as follows: $$\begin{array}{l}\text{dist}\_\text{threshold}=1000\text{m},\\ \text{angle}\_\text{threshold}=10°.\end{array}$$

In one embodiment it is provided that the at least one predetermined limit value is or is selected as a function of the geographical position. This makes it possible to flexibly specify a cluster resolution, that is to say a distance between clusters. In simple terms, this means that the larger the limit value is chosen, the lower the number of clusters per area, that is, the lower the cluster resolution. For example, it can be provided that the at least one limit value in the open sea is selected to be greater, so that the clusters there are at a greater distance from one another. In contrast, near the coast, in particular in areas with heavy traffic, the limit value can be selected to be significantly smaller, so that in these areas a distance between the clusters is smaller and a cluster resolution is increased as a result. In this way, it is possible to reduce the amount of data in the created ship route map, but still enable the generation and provision of very large ship route maps, in particular a global ship route map.

In one embodiment it is provided that a cluster resolution is determined by means of the data processing device as a function of an associated geographical position. This enables a cluster resolution to be specified separately for different geographic regions. In particular, this makes it possible to reduce a resolution even afterwards, for example if ship routes can be recognized in the ship route map, which can also be mapped with a lower cluster resolution in sufficient detail. To reduce the resolution, neighboring clusters are combined, the procedure here being analogous to the already described assignment of data points or geographical positions of the tracking data to individual clusters. The clusters are combined as a function of a distance determined by means of the distance measure and a respectively predetermined limit value which is selected according to the desired cluster resolution.

In one embodiment it is provided that the cluster resolution is established or is established on the basis of a predetermined area map. In particular, it can be provided that areas in which a cluster resolution is reduced are or are defined on the basis of the predetermined area map. The area map defines in particular a cluster resolution for areas of the Surface of the earth. In a simple example of such an area map, provision can be made for a cluster resolution of 1 km to be provided near the coast, i.e. for example at a distance of ≤ 5 km from the coast, whereas at a distance of> 5 km, a cluster resolution of 5 km and a cluster resolution of 10 km at a distance of ≥ 30 km. As a result, the resolution of the ship route map can in particular also be post-processed as a function of the specifications stored in the area map if a cluster resolution is above the respectively specified cluster resolution when it is first created. A quantity of data on the ship route map can thereby be reduced in a targeted manner, without a resolution of the ship route map having to be reduced for relevant areas of the earth's surface. To reduce the resolution, neighboring clusters are combined, the procedure here being analogous to the already described assignment of data points or geographical positions of the tracking data to individual clusters. The clusters are combined as a function of a distance determined by means of the distance measure and a respectively predetermined limit value which is selected according to the desired cluster resolution.

In one embodiment it is provided that a cluster resolution is reduced by means of the data processing device as a function of a change between the respectively assigned directions of travel of clusters linked to one another. A cluster resolution here refers in particular to a number of clusters per route. This makes it possible to reduce a cluster resolution of ship routes in those areas in which the ship route can be described with sufficient accuracy even with a lower cluster resolution. This is particularly the case in areas in which a ship route runs straight ahead over a longer distance. This can be seen by determining a change between the directions of travel that are assigned to neighboring clusters. If the change in the direction of travel is small over several clusters, i.e. if the change in the direction of travel is, for example, below a specified threshold value (e.g. <2 °, <5 ° ... etc.), The clusters concerned can respond to a smaller number Clusters can be reduced without a mapping of the ship route thereby losing accuracy. An amount of data for areas in which ship routes are mostly straight can thereby be reduced without impairing the accuracy. In order to reduce the cluster resolution, neighboring clusters are then combined, the procedure here being analogous to the already described assignment of data points or geographical positions of the tracking data to individual clusters. The clusters are combined as a function of a distance determined by means of the distance measure and a respectively predetermined limit value which is selected according to the desired cluster resolution.

Further features for the configuration of the device emerge from the description of configurations of the method. The advantages of the device are in each case the same as in the embodiments of the method

• 1 eine schematische Darstellung einer Ausführungsform der Vorrichtung zum Erstellen einer Schiffsroutenkarte;
• 2 eine schematische Darstellung einer Schiffsroutenkarte;
• 3 eine schematische Darstellung zur Verdeutlichung der Anbindung einer statischen geographischen Position eines Hafens an Schiffsrouten in der Schiffsroutenkarte.

The invention is explained in more detail below on the basis of preferred exemplary embodiments with reference to the figures. Here show:

• 1 a schematic representation of an embodiment of the device for creating a ship route map;
• 2 a schematic representation of a ship route map;
• 3 a schematic representation to illustrate the connection of a static geographical position of a port to ship routes in the ship route map.

In 1 Figure 3 is a schematic representation of one embodiment of the apparatus 1 for creating a ship route map 20th shown. The device 1 comprises an input device 2 , a data processing device 3 , and an output device 4th . The entrance facility 2 and the output device 4th are designed, for example, as interfaces for data exchange.

The entrance facility 2 receives or records tracking data 10 of ships. The tracking data 10 comprise for each of the ships at least one ordered set of geographical positions with respectively associated directions of travel in the form of data points (data tuples), i.e. each geographical position is also assigned a direction of travel which indicates the direction in which this geographical position was traversed by the ship . In particular, tracking data is used 10 received or recorded by a large number of ships that are in particular underway around the world. The tracking data 10 are based, for example, on historical AIS data from the ships that were collected and the input facility 2 are fed.

The data processing device 3 summarizes the tracking data 10 of the ships together to form clusters as a function of a distance determined on the basis of a distance measure and at least one predetermined limit value. The grouping takes place, for example, using the BIRCH method. The distance measure takes into account both a position difference between the individual geographical positions and a travel direction difference between the respective associated travel directions. Each cluster is or is assigned a geographical position and a direction of travel (cf. 3 ), which, for example, based on the respectively assigned tracking data 10 is determined, for example by averaging. The clusters are created by the data processing device 3 linked via directed edges (cf. 3 ).

The number of clusters linked via the directed edges generated in this way defines the ship route map 20th . The exit facility 4th provides the ship route map 20th as a ship route map data set 21 ready. In particular, there is the output device 4th the ship route map data set 21 as a data stream or file.

The in the form of the shipping route record 21 provided ship route map 20th or the ship routes defined therein via the linked clusters can then be used, for example, to estimate an arrival time of a ship.

Provision can be made for a weighting value to be assigned or assigned to the directed edges between the clusters, the weighting value between two clusters connected to one another for each based on tracking data 10 additional unidirectional linkage between these clusters is increased, and the weighting value between two linked clusters for each based on tracking data 10 additional linkage between these clusters, which runs in the opposite direction, is reduced.

Furthermore, it can be provided that a weighted sum between a respective position difference and the associated travel direction difference is formed as a distance measure.

It can further be provided that the distance measure is selected such that a travel direction difference in the determined distance is taken into account disproportionately as a function of the determined travel direction difference.

In 2 is a schematic representation of a means of the method or the device 1 created ship route map 20th shown. The 2 shows an example of an excerpt from the ship route map 20th . Ship routes are shown 30th in a coastal area 31 around a mainland 32 around. The ship routes 30th were identified as “one-way streets” as part of the disclosed procedure, as they are on these ship routes 30th is only driven in one direction of travel (each indicated schematically by the arrows). To the ship routes 30th To generate, the disclosed method was carried out, that is, data points of tracking data became individual clusters 40 assigned (for the sake of clarity only some of the clusters are 40 marked with a reference symbol).

It can be provided that the at least one predetermined limit value is selected or is selected as a function of the geographical position.

It can further be provided that by means of the data processing device 2 a cluster resolution is determined depending on an associated geographical position. In particular, it can be provided that by means of the data processing device 2 a cluster resolution is reduced as a function of an associated geographical position. A result of such a dependency is shown schematically in FIG 2 shown. Within the coastal area 31 is a distance between the clusters 40 smaller than outside the coastal area 31 in an offshore area 34 .

It can be provided that the cluster resolution is set or is set on the basis of a predetermined area map. In particular, it can be provided that areas 31 , 34 in which a cluster resolution is decreased are set or are set based on the predetermined area map. Such an area map would be in the example of 2 in particular a maximum cluster resolution for the coastal area 31 and a different maximum cluster resolution for the high seas 34 specified, with a specification for the cluster resolution of the coastal area 31 is selected to be greater than the cluster resolution of the high seas area 34 .

It can be provided that by means of the data processing device 2 ( 1 ) a cluster resolution as a function of a change between the respectively assigned driving directions of linked clusters 40 is decreased. In particular, this enables cluster resolution for sections of ship routes 30th be reduced in which the ship route 30th runs predominantly straight ahead and can therefore also be mapped with a lower cluster resolution.

In the coastal area 31 there is also a port 33 . The harbor 33 was made by specifying a static geographical position 11 with the nearby shipping routes 30th connected. For this purpose it is provided that the static geographic position 11 by means of the input device 2 ( 1 ) is received, for the static geographical position 11 by means of the data processing device 3 two clusters 40 be generated.

This is shown schematically in the 3 made clear. The two for the static geographic position 11 generated clusters 40 (for the sake of clarity, these are shown slightly offset) are clusters 40 in a given radius 50 (e.g. 2 km, 3 km, ...) around the static geographical position 11 linked around, with the associated directed edges 43 , 44 these links for the two clusters 40 have different directions.

In the 3 are also shown for the clusters 40 the geographical positions in each case 41 and each of the clusters 40 assigned directions of travel 42 .

List of reference symbols

1

contraption

2

Entrance facility

3

Data processing device

4th

Exit facility

10

Tracking data

11

static geographic position (port)

20th

Ship route map

21

Ship route record

30th

Ship route

31

Coastal area

32

Mainland

33

port

34

Offshore area

40

Cluster

41

geographical position

42

Direction of travel

43

directed edge (one direction)

44

directed edge (other direction)

50

Radius

QUOTES INCLUDED IN THE DESCRIPTION

This list of the documents listed by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Non-patent literature cited

• T. Zhang et al., BIRCH: an efficient data clustering method for very large databases, ACM Sigmod Record, 1996 [0004]

Claims (10)

Method for creating a ship route map (20), tracking data (10) from ships being received or recorded by means of an input device (2), the tracking data (10) for each of the ships each having at least one ordered set of geographical positions (41) each with associated driving directions (42) include, wherein the tracking data (10) of the ships are combined into clusters (40) by means of a data processing device (3) depending on a distance determined on the basis of a distance measure and at least one predetermined limit value, the distance measure being both a position difference between the individual geographical positions (41 ) and a travel direction difference between the respective associated travel directions (42) are taken into account, and each cluster (40) is assigned a geographical position (41) and a travel direction (42), and wherein the clusters (40) are linked to one another or are linked via directed edges (43, 44), and wherein the quantity of clusters (40) linked via the directed edges (43, 43) generated in this way is provided as a ship route map data set (41) by means of an output device (4). Procedure according to Claim 1 , characterized in that a weighting value is or is assigned to the directed edges (43, 44) between the clusters (40), the weighting value between two interconnected clusters (40) for each unidirectional link between these clusters (40) is increased, and wherein the weighting value between two linked clusters (40) is reduced for each link between these clusters (40) that is added on the basis of tracking data (10) and runs in the opposite direction. Procedure according to Claim 1 or 2 , characterized in that at least one static geographical position (11) is received by means of the input device (2), two clusters (40) being generated for the at least one static geographical position (11) by means of the data processing device (3), and wherein the two generated clusters (40) are linked with clusters (40) which lie in a predetermined radius (50) around the at least one static geographical position (11), the associated directed edges (43, 44) of these links for the two Clusters (40) have different directions. Method according to one of the preceding claims, characterized in that a weighted sum between a respective position difference and the associated travel direction difference is formed as the distance measure. Method according to one of the preceding claims, characterized in that the distance measure is selected in such a way that a travel direction difference in the determined distance is taken into account disproportionately as a function of the determined travel direction difference. Method according to one of the preceding claims, characterized in that the at least one predetermined limit value is or is selected as a function of the geographical position (41). Method according to one of the preceding claims, characterized in that a cluster resolution is determined by means of the data processing device (3) as a function of an associated geographical position (41). Method according to one of the preceding claims, characterized in that the cluster resolution is established or is established on the basis of a predetermined area map. Method according to one of the preceding claims, characterized in that by means of the data processing device (3) a cluster resolution is reduced as a function of a change between the respectively assigned driving directions of linked clusters (40). Device (1) for creating a ship route map (20), comprising: an input device (2), a data processing device (3), and an output device (4), the input device (2) being set up to send tracking data (10) from ships receive or detect, the tracking data (10) for each of the ships each comprising at least one ordered set of geographical positions (41) with respectively associated directions of travel (42), and the data processing device (3) being set up to combine the tracking data (10) of the ships into clusters (40) as a function of a distance determined on the basis of a distance measure and at least one predetermined limit value, the distance measure being both a position difference between the individual geographical positions (41) and a travel direction difference between the respective associated travel directions (42) are taken into account, and each cluster (40) is assigned or is assigned a geographical position (41) and a travel direction (42), and the clusters (40) via directed To link edges (43,44) with one another, and wherein the output device (4) is set up to provide the set of clusters (40) linked via the directed edges (43,44) as a ship route map data set (21) .

Images