A method for determining a channel boundary based on AIS data

Abstract

The application provides a kind of waterway boundary determination method based on AIS data, belongs to AIS data processing technical field.Based on the method of probability distribution fitting, the waterway boundary under 95% guarantee rate is determined, which provides scientific basis for waterway network planning and design, and the steps are as follows:1) obtaining AIS data;2) setting a rectangular area perpendicular to the ship flow direction along the track belt, intercepting traffic flow;3) fitting the probability distribution of ship track;4) determining the position corresponding to 95% quantile;5) determining the waterway boundary in the rectangular area;6) connecting multiple control points to obtain the overall route boundary.The application mainly determines the ship navigation boundary with 95% guarantee rate by fitting the probability distribution of ship track, and then determines the waterway boundary by considering the natural conditions around the waterway.The method is strong in theory, can provide scientific reference for waterway network planning and design, and further improve the digitalization and intelligent level of maritime transportation infrastructure construction.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of AIS data processing, and relates to ship navigation feature analysis based on probability statistics, in particular to a channel boundary determination method. BACKGROUND

[0002] The ship automatic identification system (AIS) data is generated and sent to the shore base station by the ship sailing at sea in real time, is the basic information source for ship identification and target tracking, and is an important auxiliary means for maintaining the safe distance of the ship.

[0003] At present, the ship AIS data has been used in the fields of ship trajectory anomaly detection, channel navigation safety early warning, and ship flow prediction, but the existing researches analyze the ship navigation features from the perspective of a single ship or analyze the traffic flow features of the ships in a region for traffic organization and management. However, the continuous development of the shipping network continuously puts forward higher requirements for the scientificity and rationality of the water traffic infrastructure construction and planning, and the planning of the channel network needs to be based on the actual ship navigation features.

[0004] Therefore, a channel boundary determination method based on the analysis of the ship AIS data is urgently needed to assist the planning and design of the channel network. There is still a lack of feasible technical solutions in the related field. SUMMARY

[0005] The purpose of the application is to determine the waterway route boundary under the 95% guarantee rate based on the probability distribution fitting method according to the AIS historical data of the ships in a certain region, and to provide a scientific basis for the planning and design of the waterway network.

[0006] In order to achieve the above purpose, the technical scheme adopted by the application is:

[0007] A channel boundary determination method based on AIS data analysis, wherein the data analysis is based on the method of probability theory to determine the 95% guaranteed ship navigation boundary. The specific steps are as follows:

[0008] Step 1: Obtain AIS data

[0009] The AIS data is sent by the shipborne equipment, collected by the shore base station facility, and finally the AIS data in a certain region is collected into a unified database. The AIS data field content covers the position, navigation direction and speed, departure port and destination port, ship length, draft and many other information of each ship.

[0010] The present application extracts historical data of the region to be analyzed from the existing ship AIS navigation database, takes the Maritime Mobile Service Identify (MMSI) as the unique identification of the ship, and takes the longitude and latitude information of the ship during navigation for restoring the ship navigation track. The final obtained data needs to contain three fields of MMSI, longitude and latitude.

[0011] Step two: Set a rectangular region perpendicular to the ship flow direction along the track belt to intercept the traffic flow

[0012] According to the ship navigation track, the positions of each channel are determined, and a rectangular region is drawn on the channel to be studied to intercept the ship track in the region for subsequent analysis. Specifically:

[0013] (1) According to the ship AIS historical data obtained in step one, the ship track is drawn;

[0014] (2) Determine the channel section to be analyzed, and draw a cross-section region at the position where the channel boundary needs to be determined to ensure that the cross-section length covers the tracks of all ships in the channel section. Since the ship AIS data acquisition frequency is about 6 minutes, all ship records in the rectangular region with the cross-section as the symmetry axis are considered as passing through the cross-section. The length of the rectangle along the ship flow direction can be taken as 1 / 2 of the length perpendicular to the ship flow direction to avoid the situation that the ship passes through the rectangular region without leaving a track point in the region.

[0015] (3) Record the coordinates of the lower left corner point N of the cross-section boundary and the angle size θ between the route direction and the horizontal direction.

[0016] (4) Calculate the distance of each ship track point in the cross-section from the left side of the rectangle along the ship flow direction:

[0017] X=(x1,x2,…,x n )

[0018] Wherein, the distance is a positive value, which is calculated directly by longitude and latitude value without conversion to meter unit.

[0019] The rectangular region setting method and main elements are shown in the attached Figure 1 .

[0020] Step three: Fitting the ship track probability distribution

[0021] According to the calculation results of step two (4), the frequency distribution of the ship position is obtained. If the distribution is symmetric, it is assumed that the distance of the ship from the center line obeys the normal distribution, and the mean μ and variance σ of the distribution are fitted 2If the distribution is asymmetric, it is assumed that the distance of the ship from the centerline follows a gamma distribution, and the shape parameter a and scale parameter β of the distribution are fitted.

[0022] (1) The distance of the ship from the centerline follows a normal distribution

[0023] The parameters μ and σ 2 The parameters μ and σ are calculated by the following equations

[0024]

[0025]

[0026] The probability density distribution function is:

[0027]

[0028] (2) The distance of the ship from the centerline follows a gamma distribution

[0029] The parameter a is obtained by solving the following equation:

[0030]

[0031] where

[0032]

[0033]

[0034] The parameter β is calculated by the following equation

[0035]

[0036] The probability density distribution function is:

[0037]

[0038] Step 4: Determine the position corresponding to the 95% percentile

[0039] According to the probability density distribution function of the ship's distance from the edge line in the rectangular region, the maximum width of the ship's track with a 95% guarantee rate is determined. Let x1 and x2 represent the left and right boundaries of the track with a 95% guarantee rate, respectively, which can be divided into the following two cases:

[0040] (1) The distance of the ship from the centerline follows a normal distribution

[0041]

[0042]

[0043] (2) The distance of the ship from the edge of the rectangle obeys a gamma distribution

[0044]

[0045]

[0046] Step five: determining the channel boundary in the rectangular area

[0047] The obtained longitude calculation method of the left boundary of the ship route is as follows:

[0048] x = x N + x1cos (θ)

[0049] The obtained longitude calculation method of the right boundary of the ship route is as follows:

[0050] x = x N + x2cos (θ)

[0051] The latitude of the ship route boundary can be determined according to the intersection of the meridian and the cross section.

[0052] After obtaining the ship route boundary, it is taken as the boundary of the channel.

[0053] Step six: connecting multiple control points to obtain the overall route boundary

[0054] For the channel section to be studied, multiple rectangular areas are demarcated at the turning points and the route width change points, and steps one to five are repeated. The obtained channel boundaries of the rectangular areas are connected to obtain the overall boundary of the channel section.

[0055] The effects and benefits of the present application are as follows:

[0056] The present application proposes a method for determining the channel boundary based on ship AIS historical data. The method mainly determines the 95% guaranteed ship navigation boundary by fitting the ship track probability distribution, and then determines the channel boundary by considering the natural conditions around the channel. The method is highly theoretical, can provide scientific reference for waterway network planning and design, and can improve the digitalization and intelligentization level of maritime transportation infrastructure construction. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 is a schematic diagram of ship track rectangular area division. In the figure: 1 represents a track point in the ship AIS data; 2 represents the distance of the track point to the left side of the rectangular area; 3 represents the lower left corner point of the rectangular area, i.e. N in step two (3); 4 represents the left boundary of the ship route; 5 represents the right boundary of the ship route; 6 represents the intersection angle between the track and the horizontal direction, i.e. θ in step two (3); and 7 represents the cross section.

[0058] Figure 2is a schematic diagram of ship track, cross section and corresponding rectangular area. In the figure: 1 represents the left end of cross section 1; 2 represents the left end of cross section 2; 3 represents the track of the ship passing through cross section 1 in the rectangular area near the cross section; 4 represents the track of the ship passing through cross section 2 in the rectangular area near the cross section.

[0059] Figure 3 is a histogram of the frequency distribution of the distance of the ship passing through cross section 1 from the left boundary.

[0060] Figure 4 is a histogram of the frequency distribution of the distance of the ship passing through cross section 2 from the left boundary.

[0061] Figure 5 is the calculation result of the channel boundary. In the figure: 1 represents the left end of cross section 1; 2 represents the left boundary of the channel at cross section 1; 3 represents the right boundary of the channel at cross section 1; 4 represents the left end of cross section 2; 5 represents the left boundary of the channel at cross section 2; 6 represents the right boundary of the channel at cross section 2; 7 represents the left boundary of the channel determined according to the method proposed in the application; 8 represents the right boundary of the channel determined according to the method proposed in the application. DETAILED DESCRIPTION

[0062] The application will be further described below in combination with specific embodiments.

[0063] Step one: Obtain AIS data

[0064] Obtain the AIS data of the ships near Hangzhou Bay from the ship AIS database, the time range is from January 2022 to February 2022, and the spatial range is the rectangular area between the two points (31.00N, 121.59E) and (30.23N, 122.51E).

[0065] Step two: Set a rectangular area along the track belt perpendicular to the ship flow direction, intercept the traffic flow, the channel network in this area is developed, and the ship navigation rule is relatively complex. The ship route is complex between the two points (30.41N, 121.91E) and (30.50N, 122.15E), and the channel boundary is relatively blurred. Therefore, the method proposed in the application is used to determine the boundary of this section of channel.

[0066] The ship AIS track and the selected cross sections are shown in the following figure. The left coordinate of the first cross section is (30.45N, 121.97E), and the right coordinate is (30.40N, 121.99E); the left coordinate of the second cross section is (30.49N, 122.04E), and the right coordinate is (30.44N, 122.07E).

[0067] The results are shown in the attached Figure 2 .

[0068] Step three: Fit the probability distribution of ship track

[0069] The distance between the ship position in the cross-section area and the left point of the cross-section is calculated. The frequency distribution histogram of the calculated distance is shown in Fig. 2. Figure 3 Figure 4 The frequency distribution histogram of the distance between the ship position in the cross-section area and the left point of the cross-section is fitted by using the normal distribution and the gamma distribution, respectively, and the results are as follows.

[0070] The data obtained in the first cross-section is fitted by using the normal distribution, in which the parameter μ is 2.6220e-2 and the parameter σ is 5.7311e-5. The data obtained in the second cross-section is fitted by using the gamma distribution, in which the parameter α is 2.4144e2 and the parameter β is 6.3469e-4. 2

[0071] Step four: determining the position corresponding to the 95% quantile point

[0072] The left boundary and the right boundary of the track band with a 95% guarantee rate in the two cross-sections are calculated according to the method described in step four of the summary, and the results are shown in Fig. 3. Figure 5

[0073] In the first cross-section, the left boundary of the channel is about 1.32 km away from the left end point of the cross-section, and the right boundary is about 3.70 km away from the left interface.

[0074] In the second cross-section, the left boundary of the channel is about 1.33 km away from the left end point of the cross-section, and the right boundary is about 4.35 km away from the left interface.

[0075] Step five: determining the channel boundary

[0076] According to the method described in step five of the summary, the channel boundaries in the two cross-sections are as follows.

[0077] In the first cross-section, the left boundary of the channel is (30.4373N, 121.9721E), and the right boundary is (30.4150N, 121.9831E).

[0078] In the second cross-section, the left boundary of the channel is (30.4784N, 122.0472E), and the right boundary is (30.4507N, 122.0642E).

[0079] Step six: connecting multiple control points to obtain the overall route boundary

[0080] The control points of the two cross-sections are connected to form the channel boundary, as shown in Fig. 4. Figure 5

[0081] ​​​​The above embodiments only express the implementation ways of the present application, and cannot be understood as the limitation to the scope of the present application patent. It should be pointed out that, for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application.

Claims

1. An AIS data based fairway boundary determination method for determining a 95% guaranteed rate of ship navigation boundary, characterized in that, Comprising the following steps: Step one: Obtain AIS data AIS data is sent by shipboard equipment, collected by shore base facilities, and ultimately AIS data in a certain area is collected into a unified database; AIS data field content covers the location, direction and speed of each ship, the port of origin and destination, the captain, the draft, or other information; From the existing ship AIS navigation database, extract the historical data of the area to be analyzed, take the ship's mobile communication service identification code MMSI as the unique identification of the ship, and take the longitude and latitude information during the ship's navigation process to restore the ship's navigation track. The final data needs to include MMSI, longitude and latitude information; Step two: Set a rectangular area perpendicular to the ship flow direction along the track belt to intercept the traffic flow; According to the ship's navigation track, determine the location of each channel, and draw a rectangular area on the channel to be studied to intercept the ship's track in the area for subsequent analysis: (1) According to the ship AIS historical data obtained in step one, draw the ship's track; (2) Determine the channel section to be analyzed, and draw a cross-sectional area at the position where the channel boundary needs to be determined to ensure that the cross-sectional length covers the tracks of all ships in the channel section; All ship records in the rectangular area with the cross section as the axis are considered to pass through the cross section; The length of the rectangle along the ship flow direction can be 1 / 2 of the length perpendicular to the ship flow direction to avoid the situation that the ship passes through the rectangular area without leaving a track point in the area; (3) record the coordinates of the lower left corner point N of the section boundary, and the angle between the heading direction and the horizontal direction ; (4) Calculate the distance of each ship track point in the cross section from the left side of the rectangle along the ship flow direction: ； Wherein, the distance is a positive value, which is calculated directly by longitude and latitude values without conversion to meters; Step three: Fit the ship track probability distribution Based on the calculation results of step (4) in step two, the frequency distribution of the ship's position is obtained; if the distribution is symmetrical, it is assumed that the distance of the ship from the centerline follows a normal distribution, and the mean of the distribution is obtained by fitting the data. and variance If the distribution is asymmetric, then assume that the distance of the ship from the centerline follows a gamma distribution, and fit the shape parameters of this distribution. and scale parameters ; Step four: Determine the position corresponding to the 95% quantile According to the probability density distribution function of the ship distance from the boundary value in the delimited rectangular area, the maximum width of the ship track with 95% guarantee rate is determined; and and respectively represent the left boundary and the right boundary of the track band with 95% guarantee rate. Step five: Determine the channel boundary in the rectangular area The calculation method of the left boundary longitude of the ship route is: ； The calculation method of the right boundary longitude of the ship route is: ； The latitude of the ship route boundary can be determined according to the intersection of the meridian and the cross section; After obtaining the ship route boundary, it is used as the boundary of the channel; Step six: Connect multiple control points to obtain the overall route boundary For the channel section to be studied, multiple rectangular areas are drawn at the turning points and changes in route width, and steps one to five are repeated; Connect the channel boundaries of the rectangular areas obtained to obtain the overall boundary of the channel section.

2. The method of claim 1, wherein, The fourth step is divided into two cases: (1) The distance of the ship from the rectangular boundary line follows a normal distribution; ； ； (2) The distance of the ship from the rectangular boundary line follows a gamma distribution; ； 。

Images