CN116564136B

CN116564136B - AIS-based ship collision prediction method

Info

Publication number: CN116564136B
Application number: CN202310571481.0A
Authority: CN
Inventors: 陈倩; 姚荣彬; 刘峻瑜
Original assignee: Guangxi Comprehensive Transportation Big Data Research Institute
Current assignee: Guangxi Comprehensive Transportation Big Data Research Institute
Priority date: 2023-05-20
Filing date: 2023-05-20
Publication date: 2024-03-19
Anticipated expiration: 2043-05-20
Also published as: CN116564136A

Abstract

The invention relates to the technical field of ship monitoring and discloses a ship collision prediction method based on AIS. The invention comprises the following steps: acquiring ship information of the tam ship, and predicting based on extended Kalman filtering to determine the current momentIs a vessel parameter; according to the current timeThe method comprises the steps of determining predicted latest meeting time and correction time factors between ships according to the parameters of the ship and the ship parameter of the ship; after the minimum inter-ship distance set among ships is determined, the minimum value in the minimum inter-ship distance set is used as the correction nearest meeting distance, the correction nearest meeting time is determined, and collision early warning is carried out. The invention corrects the latest meeting distance and the latest meeting time among ships by combining the ship contours, can be closer to the actual situation of ship collision, improves the accuracy of ship collision position evaluation, can assist a driver in reducing the ship collision risk in a narrow inland waterway, and ensures the inland navigation to operate efficiently.

Description

AIS-based ship collision prediction method

Technical Field

The invention relates to the technical field of ship monitoring, in particular to a ship collision prediction method based on AIS.

Background

With the vigorous development of waterway transportation in China, a large number of ships are sailed in ports and inland water areas, so that the sailing area of a single ship is drastically reduced, the sailing environment is deteriorated, the serious water traffic accidents, particularly the ship collision accidents, are remarkably increased, and the healthy development of waterway transportation is seriously restricted. The ship collision early warning system capable of timely finding collision danger and giving an alarm is an effective means for reducing collision accidents.

At present, more anti-collision early warning is mainly carried out by acquiring data through AIS positioning by offshore automatic marine vessel identification system (Automatic Identification System, AIS) hardware equipment carried on a ship and offshore. Most of the monitoring systems use fishing boat anti-collision models based on the distance, speed and direction of the ship.

In the prior art, a ship automatic identification system is widely accepted by most ship drivers according to a mode of estimating the collision of the ship by using a latest meeting distance DCPA and a latest meeting time TCPA, and is the two most important indexes in estimating the collision risk of the ship. However, the traditional calculation mode of DCPA and TCPA parameters calculated by longitude and latitude coordinates of the ship ignores errors caused by size information of the ship, the errors can be ignored in wide water areas, but in inland water areas, the navigation environment can not divide wide channels, so that the navigation range of the ship is limited, the distance of the ship is relatively close, the collision detection method adopting the DCPA and the TCPA can not provide auxiliary support for safety supervision of the ship on inland water, and collision accidents among the ships still easily occur.

Disclosure of Invention

The invention provides a ship collision prediction method based on AIS, which aims at solving at least one of the technical problems existing in the prior art. Therefore, the ship collision prediction method based on the AIS is provided, the latest meeting distance and the latest meeting time are corrected, the accuracy of ship collision position assessment is improved, a driver can be assisted in a narrow inland waterway to reduce the ship collision risk, and efficient operation of inland navigation is ensured.

The technical scheme of the invention relates to a ship collision prediction method based on AIS, which comprises the following steps:

acquiring ship information of the tam ship, and predicting based on extended Kalman filtering to determine the current momentIs a vessel parameter;

acquiring ship information of the ship, and determining the current momentIs a vessel parameter of the ship;

determining predicted latest meeting time between vessels according to the parameters of the vessels of the vessel and the parameters of the vessels of the vesselCorrection time factor->；

At the moment of timeDetermining a set of the profile points of the ship and a set of the profile points of the ship at corresponding moments according to the parameters of the ship and the parameters of the ship, and determining a set of minimum inter-ship distances between ships;

Taking the minimum value in the minimum inter-ship distance set as a corrected nearest meeting distance, and the corrected nearest meeting distance corresponds to the momentIs->As a correction latest meeting time;

and carrying out collision early warning according to the corrected nearest meeting distance.

The beneficial effects of the invention are as follows: the latest meeting distance and the latest meeting time between ships are corrected by combining the profiles of the tambour and the local ship, so that the actual situation of ship collision can be more closely related, the accuracy of ship collision position assessment is improved, a driver can be assisted in a narrow inland waterway to reduce the collision risk of the ship, and the inland navigation is ensured to operate efficiently; and meanwhile, the navigation position of the ship is predicted based on the extended Kalman filtering, so that the situation that hysteresis exists in the process of receiving the data of the ship can be avoided, the efficiency of collision prediction is improved, the accuracy of predicting the navigation position of the ship is high, the operation efficiency and the safety of navigation of the ship can be further improved, the navigation risk is effectively reduced, and the occurrence of collision accidents is avoided.

Drawings

Fig. 1 is a general flow chart of a method according to the invention.

FIG. 2 is a schematic diagram of the calculation principle of the nearest meeting distance and nearest meeting time between ships;

FIG. 3 is a schematic illustration of acquisition of a ship profile point set;

FIG. 4 is a schematic diagram of the principle of calculation of the minimum distance between vessels;

FIG. 5 is a graph showing the comparative effect after prediction according to step S100 of the present invention;

fig. 6 is an enlarged view of area a of fig. 5;

FIG. 7 is an enlarged view of region B of FIG. 5;

FIG. 8 is a graph showing the comparison of errors after prediction in step S100 according to the present invention;

FIG. 9 is a schematic diagram of experiment II based on the principle simulation of FIG. 4;

FIG. 10 is a schematic diagram of a simulation of experiment two after correction according to the present invention;

FIG. 11 is a schematic illustration of experiment three based on the principle simulation of FIG. 4;

FIG. 12 is a schematic illustration of a simulation of experiment three after correction according to the present invention;

FIG. 13 is a schematic illustration of experiment four based on the principle simulation of FIG. 4;

FIG. 14 is a schematic diagram of a simulation of experiment four after correction according to the present invention;

FIG. 15 is a schematic illustration of experiment five based on the principle simulation of FIG. 4;

FIG. 16 is a schematic illustration of experimental five simulated corrections according to the present invention;

Detailed Description

The conception, specific structure, and technical effects produced by the present invention will be clearly and completely described below with reference to the embodiments and the drawings to fully understand the objects, aspects, and effects of the present invention.

It should be noted that, unless otherwise specified, when a feature is referred to as being "fixed" or "connected" to another feature, it may be directly or indirectly fixed or connected to the other feature. It should be noted that, unless otherwise specified, when a feature is referred to as being "electrically connected" or "electrically connected" with another feature, the two features may be directly connected through pins, or connected through cables, or may be connected through a wireless transmission manner. The specific electrical connection mode belongs to a general mode of a person skilled in the art, and the person skilled in the art can realize connection according to the need. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. Furthermore, unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the description presented herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and/or" as used herein includes any combination of one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in this disclosure to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element of the same type from another. For example, a first element could also be termed a second element, and, similarly, a second element could also be termed a first element, without departing from the scope of the present disclosure. The use of any and all examples, or exemplary language (e.g., "such as") provided herein, is intended merely to better illuminate embodiments of the invention and does not pose a limitation on the scope of the invention unless otherwise claimed.

Referring to fig. 1, an AIS-based ship collision prediction method includes the steps of:

step S100, obtaining the ship information of the tam, and predicting based on the extended Kalman filtering to determine the current momentIs a vessel parameter;

at present, a ship mainly collects, monitors, stores and communicates ship navigation related data based on shipborne AIS equipment, and can transmit relevant navigation information of the ship to other ships or coastal AIS base stations in real time, so that the ship can acquire the ship information of the ship, a certain delay can exist when receiving the real-time data related to the ship due to the influence of environment or equipment hardware, and in a set time interval, if the current latest related data of the ship cannot be received, the ship can be predicted based on the expansion Kalman filtering according to the related data of the ship collected before, so that the current moment of the ship can be estimated The invention is based on extended Kalman filtering to evaluate the current time of the ship>The course and the course of the tambour are set unchanged and remain the same as the course and the course collected at the previous moment, and in the embodiment, the parameters of the tambour vessel at least comprise the profile data of the tambour vessel, the longitude and latitude of the predicted position, the sailing speed and the course;

step S200, acquiring ship information of the ship, and determining the current momentIs a vessel parameter of the ship;

wherein, the ship mainly obtains the current moment of the ship from AIS equipment of the shipIs a vessel of the present boatIn this embodiment, the ship parameters of the ship at least include ship profile data, position longitude and latitude, navigation speed and heading of the ship;

step S300, determining predicted latest meeting time among ships according to the parameters of the vessels and the parameters of the vessels of the shipCorrection time factor->A tam profile set and a native vessel profile set;

wherein, according to the ship profile data of the tam ship, the predicted position longitude and latitude, the sailing speed and course and the ship profile data of the tam ship, the position longitude and latitude, the sailing speed and course, the predicted latest meeting time between the ships can be calculated and determined Correction time factor->A tam profile set and a native vessel profile set;

step S400, at the momentDetermining a set of the profile points of the ship and a set of the profile points of the ship at corresponding moments according to the parameters of the ship and the parameters of the ship, and determining a set of minimum inter-ship distances between ships;

wherein, from the time nodeInitially, based on the set of tambour contour points and the set of own ship contour points, at time +.>In order to determine the minimum inter-ship distance between ships at each moment and at the corresponding moment, and to add the moment +.>The minimum inter-vessel distance between vessels found at each moment is determined as a minimum inter-vessel distance set,

step S500, using the minimum value in the minimum inter-ship distance set as the corrected nearest meeting distance, and the time corresponding to the corrected nearest meeting distanceIs->As a correction latest meeting time;

the minimum value is taken as the correction nearest meeting distance, and the corresponding momentIs->As a correction latest meeting time;

it is known that, by correction, the minimum inter-vessel distance between vessels should be zero, so that at a certain moment, the calculation can be stopped when the minimum inter-vessel distance between vessels is zero, and the moment of time corresponding to the zero minimum inter-vessel distance can be calculated Is->The time difference of (2) is taken as the correction latest meeting time;

step S600, collision early warning is carried out according to the corrected nearest meeting distance; and when the latest meeting distance of the correction is smaller than or equal to the safety threshold, a collision early warning signal is sent out so that a rudder can change the course or the navigational speed in time to avoid collision.

In some embodiments of the invention, the vessel information of the tambour includes: the ship profile data, position longitude and latitude, navigation speed, heading and positioning time of the tambour;

the ship information of the ship comprises: the ship profile data, the position longitude and latitude, the navigation speed, the heading and the positioning time of the ship;

the tambour parameters include: the tambour is at the current momentThe profile data of the ship, the longitude and latitude of the predicted position, the sailing speed and the course;

the ship parameters of the ship comprise: the ship is at the current momentLower ship profile data, position longitude and latitude, sailing speed and course.

In some embodiments of the invention, the step 100 includes a step S110, a step S120, a step S130, a step S140, a step S150, and a step S160.

Step 110, acquiring ship information of the tam through shipborne AIS equipment, and establishing a discrete nonlinear system based on the extended Kalman filtering;

Wherein, according to the ship information of the tambour, defining the state vector and the input vector of the tambour at the k moment are respectively:

（1），

the discrete nonlinear system is:

（2）；

is the state vector at the time of said tam k +.>Is the longitude and latitude coordinates at the time of the tam, the +.>Is the input vector at the time of the said tam, specifically, including the heading and speed of the tam,/->For the speed and +.>Heading at the time of the tam; />Is a nonlinear function representing the relationship between state variables; />Representing the system transfer process noise, assuming zero mean Gaussian noise +.>Is the observation vector at the time of said tam k,/>It is described how observations can be derived from state variables, usually observations can be measured by ship sensors,/-or%>Is a nonlinear function describing the relationship between state variables and observations; />Representing noise in the observation process, and assuming zero-mean Gaussian noise;

step 120, determining a state transfer function of the tambour according to the ship information of the tambour;

in practical application, the observation vector is determined by the stored data of the AIS equipment of the tambour, and the ship track dynamics variance of the tambour can be used as a state transfer function of longitude and latitude, wherein the state transfer function is as follows:

（3），

（4），

Wherein,the heading of the tam at time k-1 is obtained in the on-board AIS equipment, and (2)>The expression is clockwise rotated by plus degrees with 0 degrees on the y-axis, +.>For the heading of the tambour at time k-1,/>The expression is 0 degrees on the x-axis, positive degrees rotated counterclockwise, +.>For the predicted time and the time interval of the previous time; c is a conversion constant, ">-a latitude of the tam at time k-1;

in this embodiment, the unit of the input speed of the AIS device is meter per second, and the unit of the sea speed 1 obtained from the AIS device is saved by 0.514444 meters per second; meanwhile, the calculation of the trigonometric function in the computer program generally uses an radian system, the representation of the angle direction is that the x-axis is 0 degree, the anticlockwise rotation is positive degree, the heading obtained in the AIS is that the y-axis is 0 degree, the clockwise rotation is positive degree, and the heading conversion is required by using a formula (4); c is a conversion constant of about 0.000009009, representing the number of degrees of latitude represented by 1 meter, which is dependent on the latitude of the place where it is located, since the length of 1 degree of longitude is not fixed, and thus the latitudeThe number of warp represented at 1 meter is +.>；

Step 130, taking a Jacobian matrix obtained by expanding the state transfer function by Taylor as a state transfer matrix of the tambour;

The state transition matrix is as follows:

（5）；

step 140, predicting a first state vector and a first covariance matrix at the next moment, and determining a Kalman gain matrix of each time node according to the first covariance matrix;

setting initial state value according to amount history data in ship information of tambourInitial covariance matrix->According to the step of the extended Kalman algorithm, the first state vector of the next moment is predicted by using a system model as follows:

（6），

the first covariance matrix is:

（7），

wherein the method comprises the steps ofIs->Process noise covariance moment of (2)The Kalman gain matrix is:

（8），

wherein,for measuring the matrix +.>The method comprises the steps of carrying out a first treatment on the surface of the Can be determined from the ship information of the tambour,/->Is->Is a covariance matrix of (a);

step 150, determining the current time of the tam according to the Kalman gain matrix, the discrete nonlinear system, the first state vector, and the first covariance matrixA lower second state vector and a second covariance matrix;

the second state vector at k moment can be obtained by combining the observation residual and the Kalman gain as follows:

（9），

the second covariance matrix is:

（10），

wherein the second state vector represents the optimal state vector estimation at the k moment, and the second covariance matrix is the updated state covariance matrix, and the current moment Second state ofThe vector and the second covariance matrix can be obtained according to the formula (9) and the formula (10);

furthermore, the initial covariance matrixRepresenting a covariance matrix of the system state estimation, wherein the initial setting can be performed according to the empirical knowledge of the system model, and the second covariance matrix is->Then the covariance matrix at the current time needs to be updated according to equation (10).

A covariance matrix representing system noise, which describes the effect of process noise on state in a system model. In the design of Kalman filtering, +.>Typically set based on empirical or measured data during the system modeling phase and remain unchanged during the filtering process. Those skilled in the art can know the noise characteristics of the system>Preliminary setting is performed on the initial value of (2). For example, if the noise is Gaussian-distributed and has a known variance, the variance may be set to +.>Is defined by the diagonal of the (c). />Covariance matrix representing measurement noise, +.>The effect of observed noise in a measurement model on the measured value is described. Similar to->，/>Is typically predetermined in the filter design and remains unchanged during the filtering process.

The debugging method comprises the following steps: optimizing through experimentation and debuggingIs the initial value of (a). The Q matrix can be set to be a larger value, so that the filter is more sensitive to process noise, and then the observation result of an actual system is debugged, so that the degree of the observation result is gradually reduced>Until a satisfactory filtering performance is achieved.

Step 160, determining the parameters of the ship according to the second state vector and the second covariance matrix;

at the time of confirmationIn the second state vector below, the current time can be confirmed>Predicted longitude and latitude of next-time ship, combined with the present embodiment, current time +.>The speed and heading of the lower-speed ship are consistent with those of the last moment, so that it can be determined that the current parameters of the lower-speed ship at least comprise +.>The ship profile data, the predicted position longitude and latitude, the sailing speed and the course.

The calculation principle of the nearest meeting distance and the nearest meeting time between traditional ships is as follows:

the principle is as shown in figure 2, and the position coordinate of the ship at the moment t is assumed to beSpeed of navigation is +.>Heading of +.>At the same time the position coordinates of the tambour are +.>Speed of navigation is +.>Heading of +.>。

The relative distance D of the two vessels is:

（11），

Speed of the tambour relative to the hostThe vector size is:

（12），

wherein:the velocity vector sum of the two vessels on the x axis; />Is the sum of velocity vectors of the two vessels on the y-axis.

Speed of the tambour relative to the shipIs +.>：

（13），

Wherein:in order of right->=90°；/>When it is negative, it is added>=270°；

True azimuth of the tambour relative to the host shipThe method comprises the following steps:

（14），

wherein: and (V) _y Representing the difference between the ship longitudinal coordinates; and (V) _x Representing the difference between the two ship abscissas; and (V) _x In order to be positive in this respect,=90°；△ _x when it is negative, it is added>=270°。

Assuming that at time t the two vessels continue to travel at the current heading and speed, the minimum distance between their relative motion trajectories DCPA (t) and the time to reach this point TCPA (t) can be calculated by the following formula:

（15），

（16）。

DCPA (t) and TCPA (t) are two important collision parameters between ships, and the collision danger and the occurrence time of the danger of the ships can be accurately identified. DCPA (t) refers to the shortest distance between vessels at the nearest meeting point, and if the value of DCPA (t) is smaller than a preset safe distance threshold, collision risk is considered to exist, and in this case, the vessels need to take collision prevention measures, such as adjusting the navigational speed or changing the heading direction, so as to increase the value of DCPA (t) and ensure that the safe distance between the vessels is maintained. A TCPA (t) greater than zero indicates how long the two vessels still reach the nearest meeting point, and by combining with the DCPA (t), whether collision danger exists between the vessels can be judged, wherein a TCPA (t) smaller than zero means that the two vessels have driven through the yielding request, and the situation of collision danger does not exist any more.

In some embodiments of the present invention, step S300 includes step S310, step S320, step S330 and step S340:

step S310, according to the current timeThe position longitude and latitude, the navigation speed and the heading of the ship and the predicted position longitude and latitude, the navigation speed and the heading of the ship are described below to determine the current moment +.>The relative distance, relative speed and relative azimuth of the ship and the tam ship;

according to the received ship information of the ship, confirming the predicted position longitude and latitude, the sailing speed and the course of the ship, and confirming the position longitude and latitude, the sailing speed and the course of the ship; then it can be confirmed that the current time of the ship and the host shipThe following coordinates and the current time can be determined according to formula (11)>The relative distance D between the lower-distance ship and the own ship, and then rootThe current time +.>Speed of the lower tambour relative to the own ship>The vector magnitude, i.e. the above relative velocity, can then be given the current moment +.>The true orientation of the lower part is the relative orientation;

step S320, determining the predicted latest meeting time according to the relative distance, the relative speed and the relative orientation Judging the state among ships;

specifically, at the current time of dayAfter the relative distance D between the host ship and the target ship, the relative velocity, and the relative direction, the corresponding latest time of chance can be obtained according to the formula (16), and the obtained latest time of chance is the predicted latest time of chance because the host ship is the predicted position at the current time>Meanwhile, according to the relative orientation and by combining the conventional technical means in the field, the system can automatically judge that the ship is in a meeting state or a overtaking state;

step S330, when the ship is in a meeting state, correcting the time factorThe maximum value of the time required for the ship or the tam to navigate the ship according to the current navigational speed;

step S340, when the ship is in the overtaking state, correcting the time factorThe time required for the ship to travel by itself at the relative speed is required for the ship to be tracked.

In addition, the principle of ship contour point set generation is as follows:

it is known that the ship profile dimensions obtained from on-board AIS equipment are typically expressed in planar coordinates, such as meters or feet, while latitude and longitude coordinates are based on the surface of the earth's ellipsoid. In order to combine ship size data with the actual geographical position of the ship, longitude and latitude coordinates need to be converted into a planar coordinate system with the same outline size, and a map projection method can achieve the conversion. Map projection is a technology which has undergone long-term development, and various projection methods such as mercator projection, cone projection and gaussian projection have been widely used in the fields of navigation, map making, engineering measurement and the like. The Mokatuo projection has small distortion in low and medium latitude areas, the characteristics of angles and directions are reserved, so that the Mokatuo projection is very useful in navigation and map making, and meanwhile, the calculation efficiency is greatly improved by a relatively simple mathematical formula.

According to the application scene and the data requirement, the universal transverse axis mercator projection (Universal Transverse Mercator, UTM) based on the ellipsoid model in the mercator projection is selected for coordinate transformation. The accuracy error compared to a web mercator projection based on a sphere model is typically within one meter, with computation times being approximately 1 to 3 times that of a web mercator projection, with the continued improvement of modern computing devices and optimized algorithms, single-computation differences being on the order of milliseconds.

UTM is divided into orthographic projection, which converts longitude and latitude coordinates into UTM plane coordinates, and back projection, which restores UTM plane coordinates to longitude and latitude coordinates, which use the same ellipsoid parameters, projection zone and central meridian.

After orthographic projection of a ship positioning point is converted into two-dimensional coordinates, ship contour points under a two-dimensional plane can be obtained through ship size information sampling, and a ship contour longitude and latitude coordinate set is obtained after back projection of the contour points and is used for calculating the minimum distance between ships.

Receiving ship AIS information, and acquiring the position of the ship under a rectangular coordinate system by adopting a general transverse-axis cutterhead projection modeCourse->And the ship dimension A, B, C, D (the relative distance between the AIS equipment installation position and the upper, lower, left and right sides of the ship profile) at time +. >Lower, wherein the ship position is +.>Heading of +.>The navigational speed isShip size data is->The position of the tambour is +.>Heading of +.>Ship size data is->。

FIG. 3 shows a process of obtaining contour coordinates of a ship by taking and calculating vertex coordinates on a rectangular ship contour based on ship position and size data, setting a proper taking step length, and taking points sequentially from top to bottom and from left to right according to ship length and ship width to obtain a contour point setWhere n is the number of all contour points, define the first fourThe four points are four vertexes of the rectangular outline of the ship. Since there is no heading information in AIS data sent by a plurality of ships in the inland navigation process and the error between heading and heading degree is small in the normal navigation process, the heading information is directly used for replacing heading, and the heading information is used for replacing heading according to the heading of the ship>The rotated contour point set is +.>The rotated contour point coordinates are obtained by the following coordinate transformation formula:

（17），

（18），

the contour point set of the similarly available tambour is. Therefore, when the corresponding ship coordinates are confirmed at the corresponding time, the corresponding contour point set can be obtained by combining the principle.

The calculation principle of the minimum distance between ships is as follows:

before calculating the minimum distance between the vessels, it is necessary to determine whether or not the two vessels have crashed in the present case. And respectively judging whether four vertexes of the rectangular outline of the ship are in the outline of the ship and whether the four vertexes of the rectangular outline of the ship are in the outline of the ship, if so, indicating that two ships collide, otherwise, calculating the shortest distance of the two ships. As shown in FIG. 4, the method of vector inner product can be used for judging the left front vertex of the tambour Whether or not to invade the outline of the ship, when->At the position of、/>Between and->At->、/>Description of the time between->Has invaded the ship.

At this time、/>、/>、/>All are acute angles, expressed as the vector inner product:

（19），

when the conditions are not met, the two vessels are not collided, and the minimum inter-vessel distance between the two vessels is calculated by adopting the contour point set. The method comprises the following steps of:

（20），

wherein the method comprises the steps ofThe method comprises the steps of obtaining the distance between two points in two ship contour point sets, wherein n and k are the number of the two ship contour points, firstly converting coordinates in a rectangular coordinate system into longitude and latitude coordinates through universal transverse-axis ink-card support back projection, and then calculating by adopting a semi-normal vector formula to obtain the three-dimensional coordinate system:

(21) Where r is the earth radius, 6378000m, < > -in this embodiment>For longitude and latitude coordinates of the ship, +.>Is the longitude and latitude coordinates of the tambour.

In some embodiments of the present invention, step S400 includes the steps of:

at the moment of timeAnd when the minimum inter-ship distance at the corresponding moment is determined by calculation according to the following formula: />

（22），

Wherein,for time->The ship contour point set below, < >>For time->Said set of tambour points below,/->Indicating time->The distance between two points in the lower two ship contour point sets, n and k are two ship contour point numbers, r is the earth radius, and +. >For time->Longitude and latitude coordinates of the ship contour point set are described below, namely +.>For time->The longitude and latitude coordinates of the tambour are described below;

period of time is setAll minimum inter-ship distances determined at the corresponding moments are taken as the set of minimum inter-ship distances +.>。

In the present embodiment, the time is confirmedThe navigation parameters and coordinates of the ship under the ship, at the moment +.>In the method, when the minimum distance between the vessels is calculated at the corresponding time, the speed and the course of the default vessel are unchanged, the speed and the course of the vessel are also unchanged, the coordinates of the vessel are linearly changed according to the corresponding speed and course,the coordinates of the vessel are also, so that the conventional mathematical principle can be combined, thereby at the momentTo determine the coordinates of the corresponding host vessel and the coordinates of the tambour.

In some embodiments of the present invention, the minimum value in the minimum inter-ship distance set is taken as a corrected nearest meeting distance, and the corrected nearest meeting distance corresponds to the timeIs->The calculation formula for correcting the latest meeting time is as follows:

（24），

（25），

wherein,for said correction the nearest meeting distance, +.>Is the contour point set of the ship,For said set of tam profile points.

Wherein it can be known that at each instant in timeThe minimum inter-ship distance at the corresponding time can be found according to formula (22), and then the time period +.>In, thenThe minimum distance between ships is compared and confirmed, and according to the minimum distance calculation principle between ships, the correction factor T is combined to correct that the nearest meeting distance is zero, so that the minimum distance between ships is +.>When searching for the minimum value, if the corrected latest meeting distance is equal to zero, the confirmation can be stopped.

In some embodiments of the present invention, step S600 includes:

step S610, determining a minimum safe distance of the ship and a minimum safe distance of the ship according to the parameters of the ship and the parameters of the ship;

step S620, taking the maximum value of the minimum safety distance of the tambour and the minimum safety distance of the tambour as a safety threshold;

step 630, when the corrected latest meeting distance is smaller than or equal to the safety threshold, a collision early warning signal is sent.

The calculation of the minimum safe distance of the ship collision belongs to the conventional technical means in the field, so that the specific calculation process is not repeated in this time; when collision prediction is carried out among different ships, the minimum safe distances among the two ships are compared, and the minimum value is selected as a safe threshold value.

Specifically, step S100 in the above AIS-based ship collision prediction method is verified in a specific embodiment.

Experiment one-verification based on extended Kalman Filter prediction

And taking a section of GUIHUA0068 ship measured AIS data acquired in Guangxi river basin as a true value, wherein the selected AIS data comprises 88 track points and corresponding time, speed and course. The initial value of the state is the information at the first track point,(initial state covariance matrix)>(System noise covariance matrix)>The setting of the matrix (observed noise covariance matrix) is shown in a formula (26), wherein np. Diag is used for creating a diagonal matrix, four input parameters respectively represent uncertainty of longitude, latitude, speed and course from left to right, error_lon and error_lat are randomly generated longitude and latitude noise errors, and the value is [ -0.00003,0.00003]Random numbers in (a) and (b). />、/>The matrix is empirically set up->The setting of the matrix is determined by the accuracy of the designation in the AIS equipment.

（26）

The state transition matrix is determined by the formula (3), the observation value is AIS data with an error term added at the next moment, the AIS data is obtained according to the prediction and update steps of the extended Kalman filtering in a circulating way, namely, the results obtained according to the steps S110, S120, S130, S140, S150 and S160 are shown in the figures 5 to 7, wherein the figure 5 is an overall prediction effect diagram, and the figures 6 and 7 are effect diagrams obtained by amplifying the bent parts of two sections of channels in the figure 5. Compared with the predicted value obtained by the ship track dynamics model and the AIS observed value added with the error term, the position after the extended Kalman filtering is closer to the real position according to the step S100, and the filtering effect is good.

Through the above steps S110 to S160, the distance errors between 87 estimated points can be obtained by calculating the distances between the predicted value, the observed value, the filtered value and the actual value, respectively, as shown in fig. 8. The dots represent positioning errors in AIS data, the positioning errors are randomly generated according to positioning accuracy (3 meters) given by ship-borne AIS equipment, the crosses represent position errors predicted by a ship track dynamics model, and the stars are position errors obtained by fusion of the AIS data and the AIS data through an extended Kalman filtering algorithm. As can be seen in connection with fig. 5 to 7, the use of the model of the dynamics of the ship track in the curved leg is less effective in prediction, which is caused by the lack of consideration of changes in heading and speed, in particular changes in heading, in the prediction. The results obtained by averaging all the position errors are shown in table 1, and compared with the predicted value and the observed value, the average position error after filtering is reduced by 65.23% and 19.68% respectively, and the position error after filtering is obviously reduced, which indicates that the position error processing method based on the extended kalman filter algorithm in the step S100 has better effect in the aspect of processing the position error, can improve the accuracy of position estimation, and has obvious advantages in reducing the positioning error.

TABLE 1 average error results

The following simulation tests are performed on steps S300 to S500 in the AIS-based ship collision prediction method of the present invention: in order to verify the effectiveness of the ship collision prediction method based on AIS provided herein, two most common meeting situation meeting states and overtaking states in inland environments are respectively set, simulation experiments are carried out by using pyrm software through collected AIS historical data and two ship data at the same moment, input data are shown in table 2 and are divided into two ship dynamic and static information under four different scenes, wherein A, B, C, D is the distance from a ship locating point to a bow, a stern, a ship port and a ship starboard respectively.

Table 2 simulation experiment input data

Experiment two-encounter status

The predicted collision situation in the opposite state is shown in fig. 9 and 10, and the two vessels meet with each other to form an opposite situation, wherein the point O and the point T respectively represent the current positions of the ship and the tambour, and the point O 'and the point T' represent the calculated nearest opposite points.

Based on the parameters of table 2, and corrected by the correction method proposed in steps S300 to S500 of the present invention, the calculation result of the predicted collision in the encountered state is as follows:

the DCPA and TCPA obtained by the conventional calculation method can be known to reach the nearest meeting point after 38.41s according to the current course speed, and the nearest meeting distance between the two ship locating points is 6.36m, so that the two ships can be found to collide after the ship profile is generated, as shown in fig. 9.

After the correction in steps S300 to S500, as shown in fig. 10, the two vessels collide after 19.61S, so the corrected latest meeting time TCPA 'is 19.61S, and the minimum distance 0m between the two vessels is used as the corrected latest meeting distance DCPA'.

Experiment three-Condition

The collision-free situation is expected in the opposite situation as shown in fig. 11 and 12, where the two vessels meet head-on to form an opposite situation.

Based on the parameters of table 2, and corrected by the correction method proposed in steps S300 to S500 of the present invention, the calculation result of the expected collision-free in the encountered state is as follows:

from the calculated DCPA and TCPA, the running speed according to the current course will reach the nearest meeting point after 38.17s, the nearest meeting distance is 21.46m, and the minimum distance between ships is 6.31m, as shown in FIG. 11.

After the correction in steps S300 to S500 of the present invention, as shown in fig. 12, the two vessels will reach the nearest meeting point after 27.97S, so TCPA 'is 27.97S, and the minimum distance between vessels at this time is 3.96m as the corrected nearest meeting distance DCPA'.

Experimental four-chase state

The expected collision situation in the case of the pursuit situation is shown in fig. 13 and 14, and the pursuit situation is constituted by two vessels, one after the other, with the heading being approximately the same and the speed of the rear vessel being greater than that of the front vessel.

Based on the parameters of table 2, after being corrected by the correction method proposed in steps S300 to S500, the following estimated collision related parameters and calculation results are obtained:

the calculated DCPA and TCPA show that the ship can reach the nearest meeting point after 247.09s according to the current course speed, the nearest meeting distance between two ship locating points is 7.78m, and the collision of the two ships is found after the ship profile is generated.

After the correction in steps S300 to S500 of the present invention, as shown in fig. 14, two vessels collide after 19.61S, so the corrected latest meeting time TCPA 'is 19.61S, and the minimum distance 0m between the two vessels is used as the corrected latest meeting distance DCPA'.

Experimental five-chase state

The collision-free situation is expected in the case of the pursuit, as shown in fig. 15 and 16, and the two ships form the pursuit situation.

Based on the parameters of table 2, after being corrected by the correction method proposed in steps S300 to S500 of the present invention, the following collision-free related parameters are estimated and calculated as follows:

it is known from the calculated DCPA and TCPA that the running according to the current heading speed reaches the nearest meeting point after 125.99s, and the DCPA is 79.73m and the minimum distance between ships is 50.53m as shown in FIG. 15.

After the correction in steps S300 to S500 of the present invention, as shown in fig. 16, the two vessels will reach the nearest meeting point after 150.99S, so TCPA 'is 27.97S, and the minimum distance between vessels at this time is 43.91m as the corrected nearest meeting distance DCPA'.

The results from the experiment two to the experiment five show that the ship collision prediction method based on the AIS provided by the invention has a good prediction effect under the collision and collision-free situations of the opposite and overtaking situations, and a powerful theoretical basis is provided for the actual inland ship collision early warning.

It should be appreciated that the method steps in embodiments of the present invention may be implemented or carried out by computer hardware, a combination of hardware and software, or by computer instructions stored in non-transitory computer-readable memory. The AIS based ship collision prediction method may use standard programming techniques. Each program may be implemented in a high level procedural or object oriented programming language to communicate with a computer system. However, the program(s) can be implemented in assembly or machine language, if desired. In any case, the language may be a compiled or interpreted language. Furthermore, the program can be run on a programmed application specific integrated circuit for this purpose.

Furthermore, the operations of the processes described herein may be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The processes (or variations and/or combinations thereof) described herein may be performed under control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications), by hardware, or combinations thereof, collectively executing on one or more processors. The computer program includes a plurality of instructions executable by one or more processors.

Further, the method may be implemented in any type of computing platform operatively connected to a suitable computing platform, including, but not limited to, a personal computer, mini-computer, mainframe, workstation, network or distributed computing environment, separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, and so forth. Aspects of the invention may be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, optical read and/or write storage medium, RAM, ROM, etc., such that it is readable by a programmable computer, which when read by a computer, is operable to configure and operate the computer to perform the processes described herein. Further, the machine readable code, or portions thereof, may be transmitted over a wired or wireless network. When such media includes instructions or programs that, in conjunction with a microprocessor or other data processor, implement the steps described above, the invention described herein includes these and other different types of non-transitory computer-readable storage media. The invention may also include the computer itself when programmed according to the methods and techniques of the present invention.

The computer program can be applied to the input data to perform the functions described herein, thereby converting the input data to generate output data that is stored to the non-volatile memory. The output information may also be applied to one or more output devices such as a display. In a preferred embodiment of the invention, the transformed data represents physical and tangible objects, including specific visual depictions of physical and tangible objects produced on a display.

The present invention is not limited to the above embodiments, but can be modified, equivalent, improved, etc. by the same means to achieve the technical effects of the present invention, which are included in the spirit and principle of the present invention. Various modifications and variations are possible in the technical solution and/or in the embodiments within the scope of the invention.

Claims

1. An AIS-based ship collision prediction method, characterized in that it comprises the following steps:

acquiring ship information of the ship, and determining the current moment Is a vessel parameter of the ship;

performing collision early warning according to the corrected nearest meeting distance;

the predicted latest meeting time between ships is determined according to the parameters of the ship and the ship parameter of the shipCorrection time factor->The method comprises the following steps:

according to the current timeThe position longitude and latitude, the navigation speed and the heading of the ship and the predicted position longitude and latitude, the navigation speed and the heading of the ship are described below to determine the current moment +.>The relative distance, relative speed and relative azimuth of the ship and the tam ship;

determining the predicted latest time of chance based on the relative distance, the relative speed, and the relative orientation Judging the state among ships;

when the ships are in a meeting state, the time factor is correctedThe maximum value of the time required for the ship or the tam to navigate the ship according to the current navigational speed;

when the ship is in a state of overtaking, the time factor is correctedThe time required for the ship to travel by itself at the relative speed is required for the ship to be tracked.

2. The AIS based marine collision prediction method of claim 1, wherein:

the ship information of the tambour includes: the ship profile data, position longitude and latitude, navigation speed, heading and positioning time of the tambour;

3. The AIS-based ship collision prediction method according to claim 1 or 2, wherein the obtaining of the information of the ship and the prediction are performed based on extended kalman filtering to determine the current time The steps of the tambour parameters include:

acquiring ship information of the tam through on-board AIS equipment, and establishing a discrete nonlinear system based on the extended Kalman filtering;

determining a state transfer function of the tambour according to the ship information of the tambour;

taking a Jacobian matrix obtained by expanding the state transfer function by Taylor as a state transfer matrix of the tambour;

predicting a first state vector and a first covariance matrix at the next moment, and determining a Kalman gain matrix of each time node according to the first covariance matrix;

determining the current time of the tam according to the Kalman gain matrix, the discrete nonlinear system, the first state vector and the first covariance matrixA lower second state vector and a second covariance matrix;

and determining the parameters of the tambour according to the second state vector and the second covariance matrix.

4. The AIS based marine collision prediction method of claim 3, wherein:

defining a state vector and an input vector of the tambour at the time of k according to the ship information of the tambour as follows:

，

the discrete nonlinear system is:

；

Is the state vector at the time of said tam k +.>Is the longitude and latitude coordinates at the time of the tambour k,is the input at the time of the tamVector (S)>For the speed and +.>Heading at the time of the tam; />Is a nonlinear function representing the relationship between state variables; />Representing the system transfer process noise, assuming zero mean Gaussian noise +.>Is the observation vector at the time of said tam k,/>Is a nonlinear function describing the relationship between state variables and observations; />Representing noise during observation, assuming zero-mean gaussian noise.

5. The AIS based marine collision prediction method of claim 4, wherein:

the state transfer function is:

，

wherein,the heading of the tam at time k-1 is obtained in the on-board AIS equipment, and (2)>The expression is clockwise rotated by plus degrees with 0 degrees on the y-axis, +.>For the heading of the tambour at time k-1,/>The expression is 0 degrees on the x-axis, positive degrees rotated counterclockwise, +.>For the predicted time and the time interval of the previous time; c is a conversion constant and is used to convert the light into electric light,for the latitude of the tambour at time k-1.

6. The AIS based marine collision prediction method of claim 5, wherein:

The state transition matrix is as follows:

；

the first state vector is:

，

the first covariance matrix is:

，

the Kalman gain matrix is:

，

the second state vector is:

，

the second covariance matrix is:

，

wherein,is an initial state value set according to the ship information of the tambour>For the initial covariance matrix +.>Is->Process noise covariance matrix of>For measuring the matrix +.>The method comprises the steps of carrying out a first treatment on the surface of the Can be determined from the ship information of the tambour,/->Is->Is a covariance matrix of (a).

7. The AIS based marine vessel collision prediction method according to claim 1, wherein the time instant isAnd determining a set of the profile points of the ship and a set of the profile points of the ship at corresponding moments according to the parameters of the ship and the parameters of the ship, and determining a set of minimum inter-ship distances among ships, wherein the steps are as follows:

at the moment of timeWhen in use, the coordinates of the ship and the coordinates of the ship at corresponding moments are determined according to the parameters of the ship and the parameters of the ship, so as to further determine the contour point set of the ship and the contour point set of the ship at the corresponding moments, and simultaneously calculate according to the following formulas to determine the minimum inter-ship distance at the corresponding moments,

，

Wherein,for time->The ship contour point set below, < >>For time->Said set of tambour points below,/->Indicating time->The distance between two points in the lower two ship contour point sets, n and k are two ship contour point numbers, r is the earth radius, and +.>For time->Longitude and latitude coordinates of the ship contour point set are described below, namely +.>For time->The longitude and latitude coordinates of the tambour are described below;

8. The AIS-based ship collision prediction method according to claim 7, wherein the minimum value in the minimum inter-ship distance set is used as a corrected nearest meeting distance, and the corrected nearest meeting distance corresponds to a time pointIs->The calculation formula for correcting the latest meeting time is as follows:

；

wherein,for said correction the nearest meeting distance, +.>For time->The ship contour point set below, < >>For time->The set of the tambour points below.

9. The AIS-based ship collision prediction method according to claim 1, wherein the step of performing collision pre-warning according to the corrected nearest meeting distance comprises:

Determining a minimum safe distance of the ship and a minimum safe distance of the ship according to the parameters of the ship and the ship parameter;

taking the maximum value of the minimum safety distance of the tam and the minimum safety distance of the local ship as a safety threshold value;

and when the latest meeting distance of the correction is smaller than or equal to the safety threshold value, a collision early warning signal is sent out.