Complex water area ship collision risk assessment method based on AIS data

Abstract

The invention discloses a complex water area ship collision risk assessment method based on AIS data, and relates to the technical field of ship collision risk analysis, and the method comprises the steps: collecting ship AIS data of a target area, and carrying out the preprocessing and data conversion of the AIS data; based on the processed data, the ships possibly having the collision risk are clustered, a ship set possibly having collision is output, and a target ship is determined; the nearest meeting time TCPA, the nearest meeting distance DCPA, the relative orientation and the ship domain overlapping degree of the ship and the target ship are calculated respectively, and ship environment safety state evaluation characteristic indexes are determined; and determining a ship collision risk level through the ship environment safety state evaluation characteristic indexes, and outputting corresponding early warning information to related ships. According to the method, the effectiveness of collision risk identification can be improved, and potential collision conflicts possibly generated by ships with similar navigation characteristics can be found as soon as possible.

Description

Technical Field

[0001] The present invention relates to the technical field of ship collision risk analysis, and in particular to a method for evaluating ship collision risk in complex waters based on AIS data. Background Art

[0002] In recent years, maritime traffic has made important contributions to social and economic development, but accidents such as ship collisions, collisions and groundings still occur. In complex waters, multiple ship encounters are complicated, the navigation situation of ships is unclear, the trend of large-scale and fast-moving ships is obvious, and multiple factors influence and restrict each other, which puts forward higher requirements for the effectiveness of ship collision risk identification and the intelligence of maritime safety supervision.

[0003] With the development of radar systems and Automatic Identification Systems (AIS) technology, the information perception, interaction, sharing and processing capabilities between ships have been greatly improved, providing software and hardware support and a rich data research foundation for the assessment of ship collision risks. There have been many studies on ship collision risks based on ship AIS data at home and abroad, but they are mainly based on quantitative calculations in open waters based on the distance to closest point of approach (DCPA) and the time to closest point of approach (TCPA). The collision risk is rarely calculated by considering the spatial and temporal elements. Some algorithms are complex and difficult to implement, making it difficult to be widely used in large-scale AIS data for risk assessment. Summary of the invention

[0004] The purpose of the present invention is to overcome the deficiencies in the above-mentioned prior art and to provide a method for assessing the risk of ship collision in complex waters based on AIS data.

[0005] To achieve the above object, the present invention provides a method for assessing the risk of ship collision in complex waters based on AIS data, comprising:

[0006] Collecting AIS data of ships in the target area and preprocessing the AIS data;

[0007] Based on the processed data, the ships that may have a collision risk are clustered, the set of ships that may collide is output, and the target ship is determined;

[0008] The closest encounter time TCPA, the closest encounter distance DCPA, the relative position and the ship area overlap between the own ship and the target ship are calculated respectively, and the characteristic indicators for evaluating the ship's environmental safety status are determined;

[0009] The ship collision risk level is determined through the ship environmental safety status evaluation characteristic index, and corresponding warning information is output to the relevant ships.

[0010] Preferably, the ship AIS data includes dynamic data and static data, wherein the dynamic data includes longitude, latitude, heading, track and speed, and the static data includes ship name, call sign, MMSI, ship type, length and width.

[0011] Preferably, preprocessing the AIS data includes:

[0012] Delete data with the MMSI number of 0, delete data with obvious format errors, delete abnormal point data beyond the characteristic value range, perform interpolation processing on data with uneven time distribution, use Mercator method to calculate the physical distance between the target ship and the ship, and finally perform data conversion; among which, the data with obvious format errors are data with incorrect time format or date format.

[0013] Preferably, the method of using the Mercator method to calculate the physical distance between the target ship and the own ship is:

[0014]

[0015] Distance=(lat o -lat t )·sec(C)

[0016] Where MP is the latitude growth rate, lat is the latitude of the ship, e is the eccentricity of the earth, C is the direction of the target ship TS relative to the ship OS, log o is the longitude of the ship's OS, log t is the longitude of the target ship TS, MP o is the latitude growth rate of the ship's OS, MP t is the latitude growth rate of the target ship TS, Distance is the physical distance between the target ship and the own ship, lat o is the latitude of the ship's OS, lat t is the latitude of the target ship TS.

[0017] Preferably, clustering of ships with possible collision risks is performed based on the processed data, including:

[0018] Normalize the processed data to remove the influence of differences between different data attributes;

[0019] By using the normalized data, clustering is performed according to the geographical distance converted from the AIS latitude and longitude coordinates, and the set of ships that may collide is output;

[0020] The set of ships that may collide is:

[0021]

[0022] Where dist is the Minkowski distance, lon is t is the longitude of the target ship TS, lon o is the longitude of the ship's OS, lat t is the latitude of the target ship TS, lat o is the latitude of the ship's OS, sog t is the speed of the target ship TS, sog o is the OS speed of the ship, cog t is the heading of the target ship TS, cog o The heading of own ship's OS.

[0023] Preferably, the methods for calculating the closest encounter time TCPA and the closest encounter distance DCPA between the own ship and the target ship are respectively:

[0024]

[0025] Where Distance is the physical distance between the target ship and the own ship, α t is the actual azimuth between own ship OS and target ship TS, is the relative heading between own ship OS and target ship TS, ν r is the relative speed between own ship OS and target ship TS, DCPA is the closest approach distance between own ship OS and target ship TS, TCPA is the closest approach time between own ship OS and target ship TS.

[0026] Preferably, the method for calculating the relative position of the own ship and the target ship is:

[0027]

[0028] Where R is the relative azimuth, and α is the actual azimuth between the own ship and the target ship;

[0029] Based on the fuzzy four-element ship domain model, the ship domain overlap between the own ship and the target ship is calculated; wherein the fuzzy four-element ship domain model is:

[0030]

[0031] In the formula, R fore , R aft are the longitudinal front and rear radii of the quaternary ship domain, R port , R starbare the left and right radii of the quaternary ship area, L is the length of the ship, v is the speed of the ship, k AD is the coefficient of the initial diameter of the cycle, k DT is the coefficient of ship advance;

[0032] The minimum safe encounter distance between the own ship and the target ship is:

[0033]

[0034] Wherein, the minimum safe encounter distance r1 is the distance from the boundary of the four-element ship domain to the OS of the own ship, α is the actual azimuth between the own ship and the target ship, y is the latitude, and x is the longitude.

[0035] Preferably, determining the characteristic index for evaluating the ship environment safety status includes:

[0036] The influence of multi-factor coupling effect on the overall collision risk is quantified through the TCPA, DCPA, relative position and ship area overlap between the own ship and the target ship;

[0037] Calculate and weight the spatial and temporal collision risks to obtain the relationship between the spatial DCPA collision risk, the temporal TCPA collision risk, and DCPA and TCPA;

[0038] Based on the concept of the arena, the collision risk of relative orientation is obtained;

[0039] Based on the concept of four-dimensional ship domain, when the target ship is outside the ship domain of the own ship, the collision risk of the overlapping degree of the ship domain is obtained;

[0040] By integrating the DCPA collision risk in space, the TCPA collision risk in time, the collision risk in relative direction and the collision risk of the degree of overlap of ship fields, a comprehensive expression of collision risk is obtained, namely, the characteristic index for evaluating the ship environmental safety status.

[0041] Preferably, the characteristic index for evaluating the ship environment safety status is:

[0042] CRI=k1CRI d +k2CRI t +k3CRI RB +k4CRI SD

[0043] In the formula, CRI is the characteristic index for evaluating the ship's environmental safety status. d For DCPA collision risk in space, CRI t For the TCPA collision risk in terms of time, CRI RB is the collision risk of relative orientation, CRISD is the collision risk of the overlapping degree of ship fields, and k1, k2, k3, and k4 are the coefficients of the characteristic indicators of each safety status evaluation.

[0044] Preferably, determining the ship collision risk level includes:

[0045] When TCPA is less than 0, there is no collision risk between the target ship and the own ship;

[0046] When the physical distance between the target ship and the own ship is less than r1, the actual distance between the target ship and the own ship is less than the minimum safe encounter distance, that is, the target ship is within the ship area of ​​the own ship and there is a risk of collision. In this case, it is recommended to slow down or stop the ship.

[0047] If it is other cases, the ship collision risk level is judged according to the value of the ship environmental safety status evaluation characteristic index CRI, where CRI is a value between 0 and 1. The closer the CRI is to 1, the greater the collision risk between the target ship and the ship; the closer the CRI is to 0, the smaller the collision risk between the target ship and the ship. When the CRI value is between 0.5 and 1, the ship is considered to have a potential collision risk.

[0048] Compared with the prior art, the present invention has the following advantages and technical effects:

[0049] The present invention obtains a comprehensive expression of collision risk, namely, a characteristic index for evaluating the environmental safety status of ships, by studying the relationship between the comprehensive index method based on DCPA and TCPA, the ship field and the collision risk. This method can improve the effectiveness of collision risk identification, and early detect potential collision conflicts that may occur between ships with similar navigation characteristics, ultimately providing accurate data support for safety management at sea. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0051] Figure 1 A flow chart of a method for assessing ship collision risk in complex waters based on AIS data according to an embodiment of the present invention;

[0052] Figure 2 AIS data preprocessing and clustering steps flow chart of an embodiment of the present invention;

[0053] Figure 3 A flowchart of the steps for quantifying the risk of ship collision according to an embodiment of the present invention;

[0054] Figure 4Schematic diagram of calculating DCPA and TCPA according to an embodiment of the present invention;

[0055] Figure 5 A schematic diagram of calculating a relative bearing RB according to an embodiment of the present invention;

[0056] Figure 6 It is a schematic diagram of a four-dimensional ship field according to an embodiment of the present invention. DETAILED DESCRIPTION

[0057] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0058] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0059] The present invention proposes a method for assessing the risk of ship collision in complex waters based on AIS data. Figure 1 ,include:

[0060] Collect AIS data of ships in the target area and perform preprocessing and data conversion on the AIS data.

[0061] Specifically, the ship AIS data includes dynamic data and static data, wherein the dynamic data includes longitude, latitude, heading, track and speed, and the static data includes ship name, call sign, MMSI, ship type, length and width.

[0062] Preprocess AIS data, including:

[0063] Delete the data with the MMSI number of 0, delete the data with obvious format errors, delete the abnormal point data that obviously exceeds the characteristic value range, interpolate the data with uneven time distribution, use the Mercator method to calculate the physical distance between the target ship and the ship, and finally convert the data. Among them, the values ​​with obvious format errors refer to the wrong data such as time format and date format. The values ​​outside the characteristic range refer to the heading of the ship, which should be between 0 and 360 degrees, which is mainly caused by the aging of AIS equipment or abnormal data exchange of sensors.

[0064] Furthermore, the method for data conversion is:

[0065]

[0066] Distance=(lato -lat t )·sec(C)

[0067] Where MP is the latitude growth rate, lat is the latitude of the ship, e is the eccentricity of the earth, C is the direction of the target ship TS relative to the ship OS, log o is the longitude of the ship's OS, log t is the longitude of the target ship TS, MP o is the latitude growth rate of the ship's OS, MP t is the latitude growth rate of the target ship TS, Distance is the physical distance between the target ship TS and the ship OS, lat o is the latitude of the ship's OS, lat t is the latitude of the target ship TS.

[0068] Based on the processed data, the ships that may have collision risks are clustered, the set of ships that may collide is output, and the target ship is determined.

[0069] Specifically, the traditional K-means clustering method performs clustering based on the geographical distance converted from the AIS longitude and latitude coordinates, but this ignores the characteristic quantities such as speed SOG and heading COG contained in the AIS dynamic data.

[0070] Longitude, latitude, speed SOG and heading COG have different magnitudes and units, which will affect the accuracy of data analysis. In order to remove the influence of differences between different attributes of input data and reduce the prediction error of the model, the input data needs to be normalized.

[0071]

[0072] Among them, max(x) is the maximum value of the sample data on this attribute, min(x) is the minimum value of the sample data on this attribute, x is the original data; X' is the converted data.

[0073] After normalization, the differences between the attributes of the data can be removed so that all dimensions of the data are in the same value range, limited to [0, 1].

[0074] The set of ships that may collide is:

[0075]

[0076] Where dist is the Minkowski distance, lon is t is the longitude of the target ship TS, lon o is the longitude of the ship's OS, lat t is the latitude of the target ship TS, lat ois the latitude of the ship's OS, sog t is the speed of the target ship TS, sog o is the OS speed of the ship, cog t is the heading of the target ship TS, cog o The heading of own ship's OS.

[0077] Calculate the closest encounter time TCPA, closest encounter distance DCPA, relative position and ship area overlap between the ship and the target ship respectively, and determine the characteristic indicators for ship environmental safety status evaluation;

[0078] Specifically, the methods for calculating the TCPA and DCPA of the own ship and the target ship are:

[0079]

[0080] In the formula, Distance is the physical distance between the target ship TS and the own ship OS, α t is the actual azimuth between own ship OS and target ship TS, is the relative heading between own ship OS and target ship TS, ν r is the relative speed between own ship OS and target ship TS, DCPA is the closest approach distance between own ship OS and target ship TS, TCPA is the closest approach time between own ship OS and target ship TS.

[0081] The method for calculating the relative position of own ship and target ship is:

[0082]

[0083] Where R is the relative azimuth, and α is the actual azimuth between the own ship and the target ship;

[0084] Based on the fuzzy four-element ship domain model, the ship domain overlap between the own ship and the target ship is calculated; wherein the fuzzy four-element ship domain model is:

[0085]

[0086]

[0087] In the formula, R fore , R aft are the longitudinal front and rear radii of the quaternary ship domain, R port , R starb are the left and right radii of the quaternary ship area, L is the length of the ship, v is the speed of the ship, k AD is the coefficient of the initial diameter of the cycle, k DT is the coefficient of ship advance;

[0088] The minimum safe encounter distance between this ship and the target ship is:

[0089]

[0090] Wherein, the minimum safe encounter distance r1 is the distance from the boundary of the four-element ship domain to the OS of the own ship, α is the actual azimuth between the own ship and the target ship, y is the latitude, and x is the longitude.

[0091] By evaluating the characteristic indicators of the ship's environmental safety status, the ship collision risk level is determined, and corresponding warning information is output to relevant ships.

[0092] Specifically, the impact of multi-factor coupling effects on the overall collision risk is quantified through the TCPA, DCPA, relative orientation and ship field overlap between the own ship and the target ship;

[0093] Calculate and weight the spatial and temporal collision risks to obtain the relationship between the spatial DCPA collision risk, the temporal TCPA collision risk, and DCPA and TCPA;

[0094] Based on the concept of the arena, CRI is obtained RB , reflects the collision risk of relative orientation;

[0095] Based on the concept of four-dimensional ship domain, when the target ship is outside the ship domain of the own ship, the CRI is obtained. SD , which reflects the collision risk of the degree of overlap in the ship domains;

[0096] By integrating the DCPA collision risk in space, the TCPA collision risk in time, the collision risk in relative direction and the collision risk of the degree of overlap of ship fields, a comprehensive expression of collision risk is obtained, namely, the characteristic index for evaluating the ship environmental safety status.

[0097] The characteristic indicators for evaluating the ship's environmental safety status are:

[0098] CRI=k1CRI d +k2CRI t +k3CRI RB +k4CRI SD

[0099] In the formula, CRI is the characteristic index for evaluating the ship's environmental safety status. d For DCPA collision risk in space, CRI t For the TCPA collision risk in terms of time, CRI RB is the collision risk of relative orientation, CRI SDis the collision risk of the overlapping degree of ship fields, and k1, k2, k3, and k4 are the coefficients of the characteristic indicators of each safety status evaluation.

[0100] Furthermore, the ship collision risk level is determined, including:

[0101] When TCPA is less than 0, there is no collision risk between the target ship and the own ship;

[0102] When Distance<r1, the actual distance between the two ships is less than the minimum safe encounter distance, that is, the target ship is within the ship area of ​​the own ship, and there is a high risk of collision. In this case, it is recommended that the target ship slow down or stop.

[0103] In other cases, the collision risk level of the ships is determined based on the CRI value. CRI is a value between 0 and 1. The closer the CRI is to 1, the greater the collision risk between the two ships. The closer the CRI is to 0, the smaller the collision risk between the two ships. Generally, when the CRI value is between 0.5 and 1, it is considered that the ship has a potential collision risk and should take appropriate collision avoidance measures as soon as possible, such as changing the speed and course.

[0104] In order to more clearly express the technical solution of the present invention, the following specific embodiments are provided to introduce the solution:

[0105] like Figure 1 , a method for assessing the risk of ship collision in complex waters based on AIS data, specifically comprising the following steps:

[0106] Step 1: Collect AIS data of ships in the target area and pre-process and convert the AIS data;

[0107] like Figure 2 ,The preprocessing of the ship AIS data in the target area mainly includes: deleting the data with the MMSI number of 0, deleting the data with obvious ,format errors, deleting the abnormal point data that obviously exceeds the ,characteristic value range, and performing interpolation processing on the ,data with uneven time distribution.

[0108] Since the data source for calculation is AIS data, its location information is recorded by longitude and latitude, so coordinate projection is required here.

[0109] In this embodiment, the Mercator method is used to calculate the physical distance between the target ship and the own ship, and then data conversion is performed, wherein the own ship is identified as OS, and the state parameter of the own ship is (lon o ,lat o ,sog o ,cog o ), the target ship is identified as TS i , the target ship state parameter is (lont ,lat t ,sog t ,cog t );

[0110] The formula for calculating the ship distance is:

[0111] Distance=(lat o -lat t )·sec(C)

[0112] Where Distance is the physical distance between the own ship and the target ship, lat o is the latitude of the ship's OS, lat t is the latitude of the target ship TS, and C is the azimuth of the target ship TS relative to the own ship OS.

[0113] Step 2: Cluster the ships that may have collision risks based on the processed data, output the set of ships that may collide, and determine the target ship;

[0114] The traditional K-means clustering method clusters according to the geographical distance converted from the AIS longitude and latitude coordinates, but this ignores the characteristic quantities such as speed SOG and heading COG contained in the AIS dynamic data.

[0115] Longitude, latitude, speed SOG and heading COG have different magnitudes and units, which will affect the accuracy of data analysis. In order to remove the influence of differences between different attributes of input data and reduce the prediction error of the model, the input data needs to be normalized, specifically:

[0116]

[0117] Among them, max(x) is the maximum value of the sample data on this attribute, min(x) is the minimum value of the sample data on this attribute, x is the original data; X' is the converted data.

[0118] After normalization, the differences between the attributes of the data can be removed so that all dimensions of the data are in the same value range, limited to [0, 1].

[0119] Step 3: Based on TCPA, DCPA, relative orientation and ship domain overlap, determine the characteristic indicators for ship environmental safety status evaluation.

[0120] like Figure 4 After coordinate transformation, the calculation formulas of ship collision risk indexes DPCA and TPCA in the geodetic fixed coordinate system are:

[0121]

[0122] Distance is the physical distance between the target ship and the own ship, α t is the actual azimuth between own ship OS and target ship TS, is the relative heading between own ship OS and target ship TS, ν r is the relative speed between own ship OS and target ship TS, DCPA is the closest approach distance between own ship OS and target ship TS, TCPA is the closest approach time between own ship OS and target ship TS.

[0123] Due to the International Regulations for Preventing Collisions at Sea, the OS of the ship and the TS of the target ship i Different responsibilities of the target ship, and ships approaching from different directions pose different risks to the operating system. i The relationship between the direction and the risk of collision was proposed by British scholar Davis. The specific data of the Arena model was obtained through a large number of surveys and statistics. Figure 5 After smoothing the boundary, the Arena model is a circle with a radius of 2.7n mile. The OS is 1.7n mile away from the center of the circle and the direction is 199°, which is the safest, and its collision risk is set to 0. The OS is the most dangerous when it is 19° away from the center, and its collision risk is set to 1.

[0124] Relative position:

[0125] Previous studies on ships mostly established circular or elliptical ship shapes without considering factors such as ship length and speed, resulting in large errors.

[0126] like Figure 6 , this embodiment adopts a fuzzy four-element ship domain model that comprehensively considers various influencing factors. The specific calculation formula is as follows:

[0127]

[0128] In the formula, R fore , R aft are the longitudinal front and rear radii of the quaternary ship domain, R port , R starb are the left and right radii of the quaternary ship area, L is the length of the ship, v is the speed of the ship, k AD is the coefficient of the initial diameter of the cycle, k DT is the coefficient of ship advance.

[0129]

[0130] Wherein, the minimum safe encounter distance r1 is the distance from the boundary of the four-element ship domain to the OS of the own ship, α is the actual azimuth between the own ship and the target ship, y is the latitude, and x is the longitude.

[0131] Step 4: Quantify the impact of multi-factor coupling effects on the overall collision risk.

[0132] Based on a series of studies on establishing collision risk models and collision risk influencing factor models based on the distance between ships, it is proved that the negative exponential function can be used to describe the nonlinear relationship between the distance between ships and the collision risk of ships. As the DCPA and TCPA values ​​between ships increase, the collision risk between ships decreases nonlinearly. Therefore, the spatial and temporal collision risks are weighted to obtain the CRI. d 、CRI t Relationship with DCPA and TCPA:

[0133] CRI d =a d exp(b d DCPA)

[0134] CRI t =a t exp(b t TCPA)

[0135] Where, CRI d For DCPA collision risk in space, CRI t is the TCPA collision risk in terms of time, a d , b d 、a t , b t are parameters that quantify spatial collision risk and temporal collision risk, respectively.

[0136] By taking into account the opinions of the navigation watch officers and maritime experts, common assumptions are set for the calculation of spatial collision risk and temporal collision risk, and the values ​​of the parameters can be estimated.

[0137] Based on the concept of the arena, the collision risk function reflecting the relative orientation is proposed as:

[0138]

[0139] Where, CRI RB is the collision risk in relative orientation.

[0140] Based on the concept of four-dimensional ship domain, when the target ship is outside the ship domain of the own ship, the collision risk function reflecting the degree of overlap of the ship domain is proposed as follows:

[0141]

[0142] Where, CRI SD To reflect the collision risk of the degree of overlap between ship territories.

[0143] In summary, the comprehensive calculation expression of collision risk is obtained:

[0144] CRI=k1CRI d +k2CRI t +k3CRI RB +k4CRI SD

[0145] k1, k2, k3, and k4 are the coefficients of each safety status evaluation characteristic index.

[0146] Step 5: Determine the ship collision risk level and output corresponding warning information to the relevant ships.

[0147] like Figure 3 , when TCPA<0, there is no collision risk between the target ship and own ship;

[0148] When Distance<r1, the actual distance between the two ships is less than the minimum safe encounter distance, that is, the target ship is within the ship area of ​​the own ship, and there is a high risk of collision. In this case, it is recommended that the target ship slow down or stop.

[0149] In other cases, the collision risk level of the ships is determined based on the CRI value. CRI is a value between 0 and 1. The closer the CRI is to 1, the greater the collision risk between the two ships. The closer the CRI is to 0, the smaller the collision risk between the two ships. Generally, when the CRI value is between 0.5 and 1, it is considered that the ship has a potential collision risk and should take appropriate collision avoidance measures as soon as possible, such as changing the speed and course.

[0150] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for assessing the risk of ship collision in complex waters based on AIS data, characterized in that: include: Collecting AIS data of ships in the target area and preprocessing the AIS data; Based on the processed data, the ships that may have a collision risk are clustered, the set of ships that may collide is output, and the target ship is determined; The closest encounter time TCPA, the closest encounter distance DCPA, the relative position and the ship area overlap between the own ship and the target ship are calculated respectively, and the characteristic indicators for evaluating the ship's environmental safety status are determined; The ship collision risk level is determined through the ship environmental safety status evaluation characteristic index, and corresponding warning information is output to the relevant ships.

2. The method for assessing ship collision risk in complex waters based on AIS data according to claim 1, characterized in that: The ship AIS data includes dynamic data and static data, wherein the dynamic data includes longitude, latitude, heading, track and speed, and the static data includes ship name, call sign, MMSI, ship type, length and width.

3. The method for assessing ship collision risk in complex waters based on AIS data according to claim 2, characterized in that: Preprocessing the AIS data includes: Delete data with the MMSI number of 0, delete data with obvious format errors, delete abnormal point data beyond the characteristic value range, perform interpolation processing on data with uneven time distribution, use Mercator method to calculate the physical distance between the target ship and the ship, and finally perform data conversion; among which, the data with obvious format errors are data with incorrect time format or date format.

4. The method for assessing the risk of ship collision in complex waters based on AIS data according to claim 3 is characterized in that: The method of using the Mercator method to calculate the physical distance between the target ship and the own ship is: Distance=(lato-latt)·sec(C) Where MP is the latitude growth rate, lat is the latitude of the ship, e is the eccentricity of the earth, C is the direction of the target ship TS relative to the ship OS, log o is the longitude of the ship's OS, log t is the longitude of the target ship TS, MP o is the latitude growth rate of the ship's OS, MP t is the latitude growth rate of the target ship TS, Distance is the physical distance between the target ship and the own ship, lat o is the latitude of the ship's OS, lat t is the latitude of the target ship TS.

5. The method for assessing ship collision risk in complex waters based on AIS data according to claim 1, characterized in that: Clustering of ships with potential collision risks based on the processed data, including: Normalize the processed data to remove the influence of differences between different data attributes; By using the normalized data, clustering is performed according to the geographical distance converted from the AIS latitude and longitude coordinates, and the set of ships that may collide is output; The set of ships that may collide is: Where dist is the Minkowski distance, lon is t is the longitude of the target ship TS, lon o is the longitude of the ship's OS, lat t is the latitude of the target ship TS, lat o is the latitude of the ship's OS, sog t is the speed of the target ship TS, sog o is the OS speed of the ship, cog t is the heading of the target ship TS, cog o The heading of own ship's OS.

6. The method for assessing ship collision risk in complex waters based on AIS data according to claim 1, characterized in that: The methods for calculating the closest encounter time TCPA and the closest encounter distance DCPA between the own ship and the target ship are: Where Distance is the physical distance between the target ship and the own ship, α t is the actual azimuth between own ship OS and target ship TS, is the relative heading between own ship OS and target ship TS, ν r is the relative speed between own ship OS and target ship TS, DCPA is the closest approach distance between own ship OS and target ship TS, TCPA is the closest approach time between own ship OS and target ship TS.

7. The method for assessing the risk of ship collision in complex waters based on AIS data according to claim 6 is characterized in that: The method for calculating the relative position of the own ship and the target ship is: Where R is the relative azimuth, and α is the actual azimuth between the own ship and the target ship; Based on the fuzzy four-element ship domain model, the ship domain overlap between the own ship and the target ship is calculated; wherein the fuzzy four-element ship domain model is: In the formula, R fore , R aft are the longitudinal front and rear radii of the quaternary ship domain, R port , R starb are the left and right radii of the quaternary ship domain, L is the length of the ship, v is the speed of the ship, k AD is the coefficient of the initial diameter of the cycle, k DT is the coefficient of ship advance; The minimum safe encounter distance between the own ship and the target ship is: Wherein, the minimum safe encounter distance r1 is the distance from the boundary of the four-element ship domain to the OS of the own ship, α is the actual azimuth between the own ship and the target ship, y is the latitude, and x is the longitude.

8. The method for assessing the risk of ship collision in complex waters based on AIS data according to claim 7 is characterized in that: Determine the characteristic indicators of the ship environmental safety status evaluation, including: The influence of multi-factor coupling effect on the overall collision risk is quantified through the TCPA, DCPA, relative position and ship area overlap between the own ship and the target ship; Calculate and weight the spatial and temporal collision risks to obtain the relationship between the spatial DCPA collision risk, the temporal TCPA collision risk, and DCPA and TCPA; Based on the concept of the arena, the collision risk of relative orientation is obtained; Based on the concept of four-dimensional ship domain, when the target ship is outside the ship domain of the own ship, the collision risk of the overlapping degree of the ship domain is obtained; By integrating the DCPA collision risk in space, the TCPA collision risk in time, the collision risk in relative direction and the collision risk of the degree of overlap of ship fields, a comprehensive expression of collision risk is obtained, namely, the characteristic index for evaluating the ship environmental safety status.

9. The method for assessing ship collision risk in complex waters based on AIS data according to claim 8, characterized in that: The characteristic indicators for evaluating the ship's environmental safety status are: CRI=k1CRI d +k2CRI t +k3CRI RB +k4CRI SD In the formula, CRI is the characteristic index for evaluating the ship's environmental safety status. d For DCPA collision risk in space, CRI t For the TCPA collision risk in terms of time, CRI RB is the collision risk of relative orientation, CRI SD is the collision risk of the overlapping degree of ship fields, and k1, k2, k3, and k4 are the coefficients of the characteristic indicators of each safety status evaluation.

10. The method for assessing ship collision risk in complex waters based on AIS data according to claim 9, characterized in that: Determine the ship collision risk level, including: When TCPA is less than 0, there is no collision risk between the target ship and the own ship; When the physical distance between the target ship and the own ship is less than r1, the actual distance between the target ship and the own ship is less than the minimum safe encounter distance, that is, the target ship is within the ship area of ​​the own ship and there is a risk of collision. In this case, it is recommended to slow down or stop the ship. If it is other cases, the ship collision risk level is judged according to the value of the ship environmental safety status evaluation characteristic index CRI, where CRI is a value between 0 and 1. The closer the CRI is to 1, the greater the collision risk between the target ship and the ship; the closer the CRI is to 0, the smaller the collision risk between the target ship and the ship. When the CRI value is between 0.5 and 1, the ship is considered to have a potential collision risk.

Images