CN110706827A - Method and system for extracting water flow information of navigable water area based on ship AIS big data - Google Patents

Method and system for extracting water flow information of navigable water area based on ship AIS big data Download PDF

Info

Publication number
CN110706827A
CN110706827A CN201910919843.4A CN201910919843A CN110706827A CN 110706827 A CN110706827 A CN 110706827A CN 201910919843 A CN201910919843 A CN 201910919843A CN 110706827 A CN110706827 A CN 110706827A
Authority
CN
China
Prior art keywords
data
ship
ais
water
water area
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201910919843.4A
Other languages
Chinese (zh)
Other versions
CN110706827B (en
Inventor
何正伟
王森杰
何帆
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Wuhan University of Technology WUT
Original Assignee
Wuhan University of Technology WUT
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Wuhan University of Technology WUT filed Critical Wuhan University of Technology WUT
Priority to CN201910919843.4A priority Critical patent/CN110706827B/en
Publication of CN110706827A publication Critical patent/CN110706827A/en
Application granted granted Critical
Publication of CN110706827B publication Critical patent/CN110706827B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
    • G16Z99/00—Subject matter not provided for in other main groups of this subclass
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/30—Assessment of water resources

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Traffic Control Systems (AREA)

Abstract

The invention relates to a method and a system for extracting water flow information of a navigable water area based on ship AIS big data, wherein the method comprises the steps of dividing a navigable water area into sub-water areas; extracting AIS track data of different sub-water areas from the historical ship AIS data; preprocessing the AIS original track data; analyzing and processing the AIS track data to obtain ship motion characteristic data; extracting historical water flow observation data matched with the spatial attributes, and matching ship motion characteristic data with water flow data based on the time attributes; and designing a BP neural network, excavating a potential mapping relation between sample data, and realizing extraction from AIS data to water flow information. The invention can extract the water flow information of the ship navigation water area from the AIS data, can provide navigation information for a crew to operate the ship, and reduces the accident rate of waterway traffic.

Description

Method and system for extracting water flow information of navigable water area based on ship AIS big data
Technical Field
The invention relates to the technical field of big data mining, in particular to a method and a system for extracting water flow information of a navigable water area based on ship AIS big data.
Background
Obtaining the navigation environment information is an important link for ensuring the navigation safety of the ship, so how to obtain the current water flow situation information of the water area in time is very important. The current water flow measuring technology can be divided into 5 types of floating method, mechanical type, electromagnetic induction type, acoustic type and optical remote sensing type according to the principle. The floating method is to throw self-sinking floating plane detecting buoy into water body and to determine the flow speed and direction of surface flow by tracking the space-time change of buoy. The mechanical water flow meter drives the impeller to rotate by means of driving force generated by water flow, then the rotating speed is sensed by the sensor and finally converted into a flow speed value, and the flow direction is determined by the aid of the built-in magnetic compass. The electromagnetic water flow meter measures the flow speed and direction of water flow by using the induced electromotive force generated by the water flow in the geomagnetic field. The acoustic current meter mainly measures the flow velocity and direction information of water flow through the propagation characteristics of ultrasonic waves in the water flow and a Doppler frequency shift technology. The water flow remote sensing observation technology carries out inversion on water flow information through optical information.
However, the methods have defects in different degrees, so that the methods cannot be applied to ship navigation. If a floating flow measurement technology needs to be specially equipped with a special buoy, the mechanical flow measurement technology is usually used for fixed acquisition point acquisition, and the characteristics and the limitations on the operation mode of the measurement technology are referred to, so that the technology cannot be carried on a ship. The measurement accuracy of the electromagnetic current measuring technology is greatly different along with the distribution of a geomagnetic field, and during measurement, a ship needs to drag a sensor to sail on the sea along a Z shape, so that the current components in two directions are measured respectively, and then the vector magnitude and direction of the current are solved, so that the current conflicts with the navigation purpose of the ship. The measurement modes of the acoustic flow measurement technology and the remote sensing technology meet the ship carrying requirement, but the equipment cost of the two measurement technologies is very high, and the market requirement cannot be met.
Disclosure of Invention
The technical problem to be solved by the present invention is to provide a method and a system for extracting navigable water area water flow information based on ship AIS big data, aiming at the above-mentioned defects of the prior art.
The technical scheme adopted by the invention for solving the technical problems is as follows: the method for extracting the water flow information of the navigable water area based on the AIS big data of the ship is constructed, and comprises the following steps:
s1, dividing the navigation water area into sub-water areas;
s2, extracting the ship AIS original track data of different sub-water areas from the historical ship AIS data;
s3, preprocessing the AIS original track data of the ships in different sub-water areas;
s4, analyzing and processing the AIS original track data of the ship to obtain ship motion characteristic data;
s5, extracting historical water flow observation data matched with the spatial attributes, and matching ship motion characteristic data with water flow data based on the time attributes;
s6, designing a BP neural network, and mining potential mapping relations among sample data to realize extraction of water flow information from AIS data to a water area.
In the method for extracting navigable water area water flow information based on ship AIS big data according to the present invention, in step S1, the step of dividing navigable water sub-areas includes:
calculating ship traffic flow density of a navigation water area by counting historical ship AIS data, dividing the navigation water area into sub-water areas according to the traffic flow density, and ensuring that at least two navigation ships are in the divided area at each moment; the area of the divided region is satisfied
Wherein S is the area of the divided region, N is the number of ships in the region, and rho is the ship statistical density of the water area.
In the method for extracting the water flow information of the navigable water area based on the ship AIS big data, in step S3, the step of preprocessing the AIS track data of different sub-water areas comprises the following steps:
cleaning AIS data and calculating track parameters;
the AIS data cleaning is to delete abnormal data and invalid data from the original AIS data, wherein the abnormal data is AIS abnormal data generated by transmission abnormality and analysis abnormality, and the invalid data is AIS data sent by a non-motor ship;
track ofThe parameter calculation is to calculate the average speed sog of the ship passing through the sub-water areaavgStandard deviation std of speed to groundsogAverage heading to ground cogavgStandard deviation of course to ground stdcogAverage bow direction hdgavgStandard deviation std of fore-aft directionhdg。
In the method for extracting navigable water area water flow information based on ship AIS big data, in step S4, the step of analyzing and processing AIS track data to obtain ship motion characteristic data comprises the following steps:
extracting a group of motion characteristic parameters capable of expressing the motion state of the ship from AIS track information of a stable sailing ship, wherein the stable sailing ship is a ship with stable engine power and steering rate of 0;
wherein the motion parameters are expressed as stable speed to ground, stable course to ground and stable heading direction; the judging method is to compare the track parameter with the threshold value to judge the motion state of the ship, and the formula is as follows:
stdsog<sogt(2)
stdhdg<hdgt(3)
stdcog<cogt(4)
wherein std issogStandard deviation of ship to ground speed, sogtDetermining a threshold, std, for the standard deviation of speed over groundhdgStandard deviation of ship fore direction, hdgtJudging threshold value, std for standard deviation of ship bow directioncogIs the standard deviation, cog, of the ship course to the groundtJudging a threshold value for the course to the ground;
when the conditions are met, the ship is considered to be in a stable sailing state, and the track data of the ship is effective data; the extraction of the ship motion characteristic parameters refers to the extraction of motion state data of a stable ship, including ship-to-ground speed, ship-to-ground course, ship fore direction and differential pressure angle information.
In the method for extracting navigable water area water flow information based on ship AIS big data, in step S5, the steps of extracting historical water flow observation data matched with spatial attributes and matching ship motion characteristic data and water flow data based on time attributes comprise:
extracting water flow observation data corresponding to spatial positions from original water flow historical observation data, matching ship motion characteristic data extracted at different times with the water flow observation data based on time fields to form data pairs corresponding to the water flow data and the ship characteristic data one to one, and collecting different data pairs to form a data set.
In the method for extracting navigable water area water flow information based on ship AIS big data, in step S6, a BP neural network is designed, potential mapping relation between sample data is mined, and the steps of extracting the water area water flow information from AIS data include:
mining and learning a large amount of collected sample data by adopting a machine learning technology; the BP neural network model modifies the network weight through reverse transfer in each data iterative training, and achieves network convergence through minimizing a cost function, so that data regression from ship motion characteristic data to water flow data is realized; the input of the BP neural network model is ship characteristic parameters, and the output is water flow velocity and flow direction information; the network structure is composed of 3 layers of an input layer, a hidden layer and an output layer.
According to the method for extracting the water flow information of the navigable water area based on the ship AIS big data, the number of the network layers, the number of the hidden layer cells and the activation function of the network structure of the BP neural network model are set, so that effective convergence is achieved on a training data set and a test training set at the same time.
The technical scheme adopted by the invention for solving the technical problems is as follows: constructing an extraction system of navigable water area water flow information based on ship AIS big data, comprising:
the sub-water area dividing unit is used for carrying out appropriate longitude and latitude division on the target water area according to the traffic flow density of the water area;
the real-time data extraction unit is used for extracting real-time ship AIS (automatic identification system) original data of different sub-waters according to the longitude and latitude;
the data preprocessing unit is used for preprocessing the ship AIS original data;
the ship motion characteristic extraction unit is used for extracting ship motion characteristics from the preprocessed ship AIS original track data;
and the water flow information calculation unit is used for calculating the input ship motion characteristic data based on the BP neural network so as to obtain water flow information data.
In the extraction system of the water flow information of the navigable water area based on the ship AIS big data, the sub-water area dividing unit can calculate the ship traffic flow density of the navigable water area by counting the historical ship AIS data, and divide the navigable water area into the sub-water area blocks according to the traffic flow density, so that at least two navigable ships are ensured to be in a divided area at each moment, and the divided area satisfies the following conditions:
Figure BDA0002217232710000041
wherein S is the area of the divided region, N is the number of ships in the region, and rho is the ship statistical density of the water area.
In the method for extracting the water flow information of the navigable water area based on the ship AIS big data, the real-time data extraction unit classifies the AIS data received in real time through longitude and latitude to obtain the AIS data of different sub water areas.
In the method for extracting the water flow information of the navigable water area based on the ship AIS big data, a data preprocessing unit is used for completing the cleaning of the real-time AIS data and the calculation of track parameters; the AIS data cleaning is to delete abnormal data and invalid data from the original AIS data, wherein the abnormal data is AIS abnormal data generated by transmission abnormality and analysis abnormality, and the invalid data is AIS data sent by a non-motorized ship; the calculation of the track parameter means that the average speed sog of the ship passing through the sub-water area is calculatedavgStandard deviation std of speed to groundsogAverage heading to ground cogavgStandard deviation of course to ground stdcogAverage bow direction hdgavgStandard deviation std of fore-aft directionhdg。
In the method for extracting the water flow information of the navigable water area based on the AIS big data of the ship, the process that the ship motion characteristic extraction unit can analyze and process the AIS track data to obtain the ship motion characteristic data comprises the following steps: firstly, analyzing track parameters to judge whether a ship is in a stable sailing state; secondly, extracting a group of motion characteristic data from the effective track data; the stable sailing ship is a ship with stable power of a ship engine and 0 steering speed, the motion parameters of the stable sailing ship are represented by stable speed to ground, stable course to ground and stable heading direction, the judgment method is to compare the track parameters with threshold values to judge the motion state of the ship, and the formula is as follows:
stdsog<sogt(2)
stdhdg<hdgt(3)
stdcog<cogt(4)
wherein std isspeedStandard deviation of ship to ground speed, sogtDetermining a threshold, std, for the standard deviation of speed over groundhdgStandard deviation of ship fore direction, hdgtJudging threshold value, std for standard deviation of ship bow directioncogIs the standard deviation, cog, of the ship course to the groundtJudging a threshold value for the course to the ground; when the above conditions are satisfied, the ship is considered to be in a stable sailing state, and the track data of the ship is valid data. The ship motion characteristic data are ship motion state data extracted from the effective track data, and comprise ship-to-ground speed, ship-to-ground course, ship fore direction and pressure difference angle data.
In the method for extracting the water flow information of the navigable water area based on the ship AIS big data, the water flow calculation unit obtains a BP neural network model through the method for extracting the water flow information of the navigable water area based on the ship AIS big data, the model is a convergence model obtained after training on a historical characteristic data set, and the water flow velocity and flow direction information of the current sub water area are effectively estimated according to the current real-time ship motion characteristic data.
Compared with the prior art, the invention has the beneficial effects that:
the method and the device perform data mining on the ship motion characteristic data based on the AIS data to extract the water flow information of the navigable water area, and have the advantages of low cost, convenience in application to any ship with AIS equipment, and the like. And further, a method for effectively acquiring the water flow information of the underway water area for a crew improves the sailing safety of the ship.
Drawings
The invention will be further described with reference to the accompanying drawings and examples, in which:
FIG. 1 is a flow chart of the steps corresponding to the method of the present invention;
FIG. 2 is a schematic view of the motion characteristics of a vessel according to the present invention;
FIG. 3 is a diagram of the BP neural network architecture according to the present invention;
fig. 4 is a corresponding block diagram of the system of the present invention.
Detailed Description
For a more clear understanding of the technical features, objects and effects of the present invention, embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
The design principle of the invention is as follows: since the international maritime organization regulated the installation of Automatic Identification System (AIS) equipment in 2002, AIS equipment has been installed on an increasing number of different types of ships. The AIS data is divided into two parts of ship dynamic data and static information data. The dynamic information records dynamic information such as Marine Mobile Service Identification (MMSI), longitude and latitude, ground speed, ground course, steering rate, bow direction, receiving time and the like. Based on the AIS dynamic information, ship track information can be constructed, so that ship motion characteristics are extracted. The motion characteristic parameters of the stable sailing ship can be proved to contain current water area current information by calculating the Pearson coefficient and the maximum correlation coefficient. Therefore, a characteristic data set can be formed by processing a large amount of historical AIS data and water flow data, and a BP neural network is designed to carry out data mining on the characteristic data set, so that a navigable water area water flow information mining model is obtained.
Based on the design principle, as shown in fig. 1, the method for extracting the water flow information of the navigable water area based on the ship AIS big data comprises the following steps:
s1 navigation water areaDividing a sub water area; the method is characterized in that a navigation water area is divided reasonably according to the traffic flow density to improve the accuracy of water flow information space of the water area, and at least two underway ships are guaranteed to be in the divided area at each moment. The area of the divided region is satisfied
Figure BDA0002217232710000071
Wherein S is the area of the divided region, N is the number of ships in the region, and rho is the ship statistical density of the water area. The division result is a longitude and latitude set.
S2, extracting AIS track data of different sub-waters from the historical ship AIS data; further, as a preferred example of the present invention, the AIS dynamic information of different sub-waters is extracted from the ship AIS original database according to the longitude and latitude set obtained in step S1; the information is track point information sent by different ships under the corresponding sub-water areas at different moments; the method comprises the steps of ship sampling point longitude and latitude positions (GPS positions), sampling time, marine mobile communication service identification codes (MMSI), ground Speed (SOG), ground Course (COG), course state, steering speed and the like.
S3, preprocessing the AIS original trajectory data; further, as a preferred example of the present invention, the preprocessing includes AIS data cleaning and trajectory parameter calculation; the AIS data adopts a connectionless User Datagram Protocol (UDP), so that the reliability is poor, the self-checking and error-correcting capability is weak, and the analyzed data has a lot of abnormal information, so that the AIS original track information needs to be subjected to data cleaning, which mainly comprises deleting a record with the MMSI number of 0 in the dynamic information; deleting records with the speed over the ground greater than 50; deleting the record with the heading direction of 511 and the like; the calculation of the track parameter refers to the calculation of the average speed sog of the ship passing through the sub-water areaavgStandard deviation std of speed to groundsogAverage heading to ground cogavgStandard deviation of course to ground stdcogAverage bow direction hdgavgStandard deviation std of fore-aft directionhdg。
S4, analyzing and processing the AIS trajectory data to obtain ship motion characteristic data; further, as a preferred example of the present inventionAnalyzing and processing the AIS trajectory data to obtain ship motion characteristic data comprises: firstly, analyzing track parameters to judge whether the track is valid data; and secondly, extracting a group of motion characteristic data from the effective track data. The validity judgment of the ship track data means that after the abnormal data are deleted, the rest AIS data still cannot be fully utilized, and part of ships are in an acceleration, deceleration or steering state when running through a research area; the motion behavior of the ships is determined by drivers, and water flow information cannot be extracted from the data, so that useless data needs to be screened; the judging method is to compare the track parameters with the threshold value to judge the motion state of the ship. When std issog<sogt、stdhdg<hdgtAnd stdcog<cogtConsidering that the ship is in a stable sailing state, keeping the track, and otherwise, deleting the track data; wherein stdsogStandard deviation of ship to ground speed, sogtDetermining a threshold, std, for the standard deviation of speed over groundhdgStandard deviation of ship fore direction, hdgtJudging threshold value, std for standard deviation of ship bow directioncogIs the standard deviation, cog, of the ship course to the groundtA threshold is determined for the heading to ground. Further, as a preferred example of the present invention, S is takent＝0.3、hdgt＝4、cogt4. Wherein each sub-water area extracts at least two ship track data to form a group of ship characteristic data, and when the extracted ship track data exceeds two, the data are matched pairwise to form a plurality of groups of ship characteristic data
S5, extracting historical water flow observation data matched with the spatial attributes, and matching ship motion characteristic data with water flow data based on the time attributes; further, as a preferred example of the present invention, the matching of the water flow data and the ship characteristic data refers to matching the ship motion characteristic data and the water flow data based on the time attribute, where the matching of the ship motion characteristic data and the water flow data refers to extracting water flow observation data corresponding to a spatial position from the original historical water flow observation data, and then matching the ship motion characteristic data and the water flow observation data extracted at different times based on the time field to form data pairs corresponding to the water flow data and the ship characteristic data one to one, and the different data pairs form a data set.
S6, designing a BP neural network, excavating potential mapping relations among sample data, and realizing extraction from AIS data to water flow information; further, as a preferred embodiment of the present invention, the extracting of the AIS data to the water flow information refers to mining and learning a large amount of collected sample data by using a machine learning technique. The input of the BP network model is ship characteristic parameters, and the output is water flow speed and flow direction information. The network structure is composed of 3 layers of an input layer, a hidden layer and an output layer. The network structure of the system has proper and effective network layer number, hidden layer unit lattice number and activation function, and can achieve effective convergence on a training data set and a test training set simultaneously; a cost function of
Figure BDA0002217232710000081
Wherein y 0 is the actual flow rate, y 1 is the actual flow direction, p 0 is the predicted flow rate, and p 1 is the predicted flow direction; and the network corrects the weight value of the network through a reverse transfer function every iteration to reduce the cost function. By setting a reasonable excitation function, the number of hidden layer units and the network iteration times, the network can reach optimal convergence, and a navigation area water flow information estimation model is obtained. Furthermore, by calculating a Pearson coefficient and a maximum information coefficient between the ship motion characteristic and the water characteristic data and analyzing the correlation between characteristic variables, the method selects two ship behavior characteristics sailing at the same moment as input parameters, and the water flow data corresponding to the moment is used as label data. So the data format is constructed as follows
Label＝[CS CA](7)
The neural network structure adopts a 3-layer full-connection layer structure, and the nonlinear fitting capacity of the network is increased by setting an activation function between different layers. In order to prevent the over-fitting phenomenon, the middle layer learns the network parameters by adopting a dropout strategy. The parameters are updated during the training process using a first order gradient based random objective function optimization algorithm Adam. Wherein the learning rate is set to 0.001, the exponential decay rate of the first order moment estimate is set to 0.9, and the exponential decay rate of the second order moment estimate is set to 0.999, for updating the learning rate after each training cycle.
Referring to fig. 4, the present invention also provides a real-time navigation water flow information extraction system based on the AIS big data of a ship, comprising
The sub-water area dividing unit can divide the target water area into proper longitude and latitude according to the traffic flow density of the water area;
the real-time data extraction unit can extract real-time ship AIS (automatic identification system) original data of different sub-waters according to the longitude and latitude;
a data preprocessing unit capable of preprocessing the real-time AIS data;
the ship motion characteristic extraction unit can extract ship motion characteristics from effective ship track data;
the water flow information calculation unit can calculate the input ship motion characteristic data based on the BP neural network so as to obtain water flow information data;
further, as a preferable embodiment of the present invention,
the sub-water area dividing unit can calculate the ship traffic flow density of the navigation water area by counting the historical ship AIS data, divide the navigation water area into sub-water area blocks according to the traffic flow density, and ensure that the divided area at each moment at least has two underway ships. The area of the divided region is satisfied
Figure BDA0002217232710000101
The method comprises the following steps that S is the area of a divided region, N is the number of ships in the region, and rho is the ship statistical density of a water area;
further, as a preferable embodiment of the present invention,
the real-time data extraction unit can extract AIS data of different sub-water areas from the real-time received AIS original data. The method is characterized in that longitude and latitude data and boundary values of the sub-waters are judged, and therefore the AIS data are classified and extracted.
Further, as a preferable embodiment of the present invention,
the data preprocessing unit can complete real-time AIS data cleaning and track parameter calculation; further, as a preferred example of the present invention, the preprocessing includes AIS data cleaning and trajectory parameter calculation; the AIS data adopts a connectionless User Datagram Protocol (UDP), so that the reliability is poor, the self-checking and error-correcting capability is weak, and the analyzed data has a lot of abnormal information, so that the AIS original track information needs to be subjected to data cleaning, which mainly comprises deleting a record with the MMSI number of 0 in the dynamic information; deleting records with the speed over the ground greater than 50; deleting the record with the heading direction of 511 and the like; the calculation of the track parameter refers to the calculation of the average speed sog of the ship passing through the sub-water areaavgStandard deviation std of speed to groundsogAverage heading to ground cogavgStandard deviation of course to ground stdcogAverage bow direction hdgavgStandard deviation std of fore-aft directionhdg。
The ship motion characteristic extraction unit can analyze and process the AIS trajectory data to obtain ship motion characteristic data. The whole extraction process comprises the following steps: firstly, analyzing track parameters to judge whether a ship is in a stable sailing state; and secondly, extracting a group of motion characteristic data from the effective track data. The stable sailing ship is a ship with stable power of a ship engine and 0 steering speed. The motion parameters are expressed as stable speed to ground, stable course to ground and stable heading direction. When std issog<sogt、stdhdg<hdgtAnd stdcog<cogtConsidering that the ship is in a stable sailing state, keeping the track, and otherwise, deleting the track data; wherein stdsogStandard deviation of ship to ground speed, sogtDetermining a threshold, std, for the standard deviation of speed over groundhdgStandard deviation of ship fore direction, hdgtJudging threshold value, std for standard deviation of ship bow directioncogIs the standard deviation, cog, of the ship course to the groundtA threshold is determined for the heading to ground. Further, as a preferred example of the present invention, S is takent＝0.3、hdgt＝4、cogt4. At least two ship track data are extracted from each sub-water area to form a group of ship characteristic data, and when the extracted ship track data exceeds two, the data are matched pairwise to form a plurality of groups of ship characteristic data. The water flow calculation unit is a BP neural network model obtained by implementing the method for extracting the water flow information of the navigable water area based on the AIS big data of the ship, and the model is a convergence model obtained after training on a historical characteristic data set, and not only the network structure design. The model can effectively estimate the current water flow velocity and flow direction information of the current sub-water area according to the current real-time ship motion characteristic data.
The following is a demonstration with specific examples as examples: taking the right channel of the shrimp door as an example, the navigation density of the shrimp door is about 0.75 ship per square kilometer. In order to ensure that at least two ship data are extracted from each sub-water area, the water area is divided into 2.57 kilometers wide and 2.76 kilometers long. The ship motion characteristic extraction step is implemented by taking one of the sub-water areas as an example, and other sub-water area data processing methods are implemented according to the same steps. 15 o' clock and 20 min at 1 month and 1 day of 2018, wherein the sub water area comprises 6 underway ships, AIS track information of 5 minutes is analyzed and processed, and track parameters are calculated as follows:
Figure BDA0002217232710000111
two underway ships are found not to be in a stable sailing state, so that the two tracks are deleted, and finally 4 ship motion characteristic data are obtained as follows:
Figure BDA0002217232710000112
Figure BDA0002217232710000121
by combining the data two by two, and converting the angle into radian, the speed is converted into meter per second to obtain the following 6 sets of characteristic data:
through a water flow information extraction model, 6 groups of flow velocity and flow direction estimated values of the sub-water area are obtained:
estimated flow velocity (m/s) Flow direction estimation (radian) True flow rate (meters/second) True flow direction (arc)
1.1973379 2.6543428 0.79 2.461
0.79704463 2.8735384 0.79 2.461
0.8271495 2.2365875 0.79 2.461
0.77975297 2.9356466 0.79 2.461
0.8228705 2.9466735 0.79 2.461
0.795712 2.7243428 0.79 2.461
After equalization treatment, the estimation result of the water flow of the sub-water area 2018-01-0115: 20:00 is as follows: flow direction 156, flow rate 0.87. Compared with the actual flow direction 141, the flow speed is 0.79, the flow speed error is 0.08m/s, and the flow direction error is 15 degrees.
Compared with the prior art, the invention has the beneficial effects that:
the method and the device perform data mining on the ship motion characteristic data based on the AIS data to extract the water flow information of the navigable water area, and have the advantages of low cost, convenience in application to any ship with AIS equipment, and the like. And further, a method for effectively acquiring the water flow information of the underway water area for a crew improves the sailing safety of the ship.
While the present invention has been described with reference to the embodiments shown in the drawings, the present invention is not limited to the embodiments, which are illustrative and not restrictive, and it will be apparent to those skilled in the art that various changes and modifications can be made therein without departing from the spirit and scope of the invention as defined in the appended claims.

Claims (10)

1. The method for extracting the water flow information of the navigable water area based on the real-time ship AIS big data is characterized by comprising the following steps of:
s1, dividing the navigation water area into sub-water areas;
s2, extracting the ship AIS original track data of different sub-water areas from the historical ship AIS data;
s3, preprocessing the AIS original track data of the ships in different sub-water areas;
s4, analyzing and processing the AIS original track data of the ship to obtain ship motion characteristic data;
s5, extracting historical water flow observation data matched with the spatial attributes, and matching ship motion characteristic data with water flow data based on the time attributes;
s6, designing a BP neural network, and mining potential mapping relations among sample data to realize extraction of water flow information from AIS data to a water area.
2. The method for extracting navigable water area flow information based on real-time ship AIS big data according to claim 1, wherein in step S1, the step of navigable water area sub-water area division comprises:
calculating ship traffic flow density of a navigation water area by counting historical ship AIS data, dividing the navigation water area into sub-water areas according to the traffic flow density, and ensuring that at least two navigation ships are in the divided area at each moment; the area of the divided region is satisfied
Figure FDA0002217232700000011
Wherein S is the area of the divided region, N is the number of ships in the region, and rho is the ship statistical density of the water area.
3. The method for extracting navigable water area flow information based on real-time ship AIS big data according to claim 1, wherein the step of preprocessing AIS trajectory data of different sub-waters in step S3 comprises:
cleaning AIS data and calculating track parameters;
the AIS data cleaning is to delete abnormal data and invalid data from the original AIS data, wherein the abnormal data is AIS abnormal data generated by transmission abnormality and analysis abnormality, and the invalid data is AIS data sent by a non-motor ship;
the track parameter is calculated by calculating the average speed sog of the ship passing through the sub-water areaavgStandard deviation std of speed to groundsogAverage heading to ground cogavgStandard deviation of course to ground stdcogAverage bow direction hdgavgStandard deviation std of fore-aft directionhdg。
4. The method for extracting navigable water area flow information based on real-time ship AIS big data according to claim 1, wherein in step S4, the step of analyzing and processing AIS trajectory data to obtain ship motion characteristic data comprises:
extracting a group of motion characteristic parameters capable of expressing the motion state of the ship from AIS track information of a stable sailing ship, wherein the stable sailing ship is a ship with stable engine power and steering rate of 0;
wherein the motion parameters are expressed as stable speed to ground, stable course to ground and stable heading direction; the judging method is to compare the track parameter with the threshold value to judge the motion state of the ship, and the formula is as follows:
stdsog＜sogt(2)
stdhdg＜hdgt(3)
stdcog＜cogt(4)
wherein std issogStandard deviation of ship to ground speed, sogtDetermining a threshold, std, for the standard deviation of speed over groundhdgStandard deviation of ship fore direction, hdgtJudging threshold value, std for standard deviation of ship bow directioncogJudging a threshold value for the standard deviation of the ship to the ground course and cogt as the ground course;
when the conditions are met, the ship is considered to be in a stable sailing state, and the track data of the ship is effective data; the extraction of the ship motion characteristic parameters refers to the extraction of motion state data of a stable ship, including ship-to-ground speed, ship-to-ground course, ship fore direction and differential pressure angle information.
5. The method for extracting navigable water area flow information based on real-time ship AIS big data according to claim 1, wherein the step of extracting historical water flow observation data matched with spatial attributes in step S5, and the step of matching ship motion characteristic data and water flow data based on time attributes comprises:
extracting water flow observation data corresponding to spatial positions from original water flow historical observation data, matching ship motion characteristic data extracted at different times with the water flow observation data based on time fields to form data pairs corresponding to the water flow data and the ship characteristic data one to one, and collecting different data pairs to form a data set.
6. The method for extracting navigable water flow information based on real-time ship AIS big data according to claim 5, wherein in step S6, a BP neural network is designed, potential mapping relations between sample data are mined, and the step of extracting the water flow information from AIS data comprises:
mining and learning a large amount of collected sample data by adopting a machine learning technology; the BP neural network model modifies the network weight through reverse transfer in each data iterative training, and achieves network convergence through minimizing a cost function, so that data regression from ship motion characteristic data to water flow data is realized; the input of the BP neural network model is ship characteristic parameters, and the output is water flow velocity and flow direction information; the network structure is composed of 3 layers of an input layer, a hidden layer and an output layer.
7. The method for extracting navigable water area flow information based on real-time ship AIS big data according to claim 6, wherein the method comprises the following steps:
the number of network layers, the number of hidden layer cells and an activation function of a network structure of the BP neural network model are set, so that effective convergence is achieved on a training data set and a test training set at the same time.
8. The utility model provides an extraction system of navigation waters rivers information based on boats and ships AIS big data which characterized in that includes:
the sub-water area dividing unit is used for carrying out appropriate longitude and latitude division on the target water area according to the traffic flow density of the water area;
the real-time data extraction unit is used for extracting real-time ship AIS (automatic identification system) original data of different sub-waters according to the longitude and latitude;
the data preprocessing unit is used for preprocessing the ship AIS original data;
the ship motion characteristic extraction unit is used for extracting ship motion characteristics from the preprocessed ship AIS original track data;
and the water flow information calculation unit is used for calculating the input ship motion characteristic data based on the BP neural network so as to obtain water flow information data.
9. The extraction system of navigable water area flow information based on ship AIS big data according to claim 8, wherein the sub-water area division unit is capable of calculating ship traffic flow density of navigable water area by counting historical ship AIS data, dividing navigable water area into sub-water area blocks according to the traffic flow density, ensuring that at least two navigable ships are owned in each time division area, and the division area satisfies:
wherein S is the area of the divided region, N is the number of ships in the region, and rho is the ship statistical density of the water area.
10. The extraction system of navigable water area water flow information based on ship AIS big data according to claim 8, characterized in that the real-time data extraction unit classifies AIS data received in real time by longitude and latitude to obtain AIS data of different sub-water areas.
CN201910919843.4A 2019-09-26 2019-09-26 Method and system for extracting navigable water area water flow information based on ship AIS (automatic identification System) big data Active CN110706827B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910919843.4A CN110706827B (en) 2019-09-26 2019-09-26 Method and system for extracting navigable water area water flow information based on ship AIS (automatic identification System) big data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910919843.4A CN110706827B (en) 2019-09-26 2019-09-26 Method and system for extracting navigable water area water flow information based on ship AIS (automatic identification System) big data

Publications (2)

Publication Number Publication Date
CN110706827A true CN110706827A (en) 2020-01-17
CN110706827B CN110706827B (en) 2022-06-03

Family

ID=69196536

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910919843.4A Active CN110706827B (en) 2019-09-26 2019-09-26 Method and system for extracting navigable water area water flow information based on ship AIS (automatic identification System) big data

Country Status (1)

Country Link
CN (1) CN110706827B (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111968406A (en) * 2020-08-18 2020-11-20 江苏航运职业技术学院 Chaos theory-based fuzzy prediction method for short-term traffic flow of ship
CN112162558A (en) * 2020-09-29 2021-01-01 广州海事科技有限公司 Ship safe navigational speed control method, system, computer equipment and storage medium
CN112373642A (en) * 2020-10-30 2021-02-19 东南大学 Inland ship overtaking behavior detection and tracking method based on ship field
CN113129645A (en) * 2021-03-26 2021-07-16 亿海蓝（北京）数据技术股份公司 Monitoring system for AIS equipment specification
CN113870620A (en) * 2021-10-19 2021-12-31 遨海科技有限公司 Ship identification method for simultaneously starting multiple AIS (automatic identification system) devices
CN114005303A (en) * 2021-09-29 2022-02-01 交通运输部天津水运工程科学研究所 Intelligent forecasting method and system for navigation safety conditions of water area of bridge area
CN114066303A (en) * 2021-11-25 2022-02-18 交通运输部规划研究院 AIS-based port planning adaptability analysis method and system
CN115273553A (en) * 2022-10-08 2022-11-01 交通运输部规划研究院 Space dividing method for ship restricted navigation and no-navigation area

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107657144A (en) * 2017-08-30 2018-02-02 天津大学 A kind of offshore flow field inversion method based on watercraft AIS and log data
CN109855688A (en) * 2019-02-28 2019-06-07 武汉理工大学 A kind of inland harbour marine exhaust discharge Measurement Method
CN110009937A (en) * 2019-03-21 2019-07-12 武汉理工大学 A kind of ship collision risk analysis method based on AIS data

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107657144A (en) * 2017-08-30 2018-02-02 天津大学 A kind of offshore flow field inversion method based on watercraft AIS and log data
CN109855688A (en) * 2019-02-28 2019-06-07 武汉理工大学 A kind of inland harbour marine exhaust discharge Measurement Method
CN110009937A (en) * 2019-03-21 2019-07-12 武汉理工大学 A kind of ship collision risk analysis method based on AIS data

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
杨帆等: "基于深度网络和船舶交通流的航道水深预测方法研究", 《武汉理工大学学报》 *

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111968406A (en) * 2020-08-18 2020-11-20 江苏航运职业技术学院 Chaos theory-based fuzzy prediction method for short-term traffic flow of ship
CN111968406B (en) * 2020-08-18 2021-06-25 江苏航运职业技术学院 Chaos theory-based fuzzy prediction method for short-term traffic flow of ship
CN112162558A (en) * 2020-09-29 2021-01-01 广州海事科技有限公司 Ship safe navigational speed control method, system, computer equipment and storage medium
CN112373642A (en) * 2020-10-30 2021-02-19 东南大学 Inland ship overtaking behavior detection and tracking method based on ship field
CN112373642B (en) * 2020-10-30 2022-02-01 东南大学 Inland ship overtaking behavior detection and tracking method based on ship field
CN113129645A (en) * 2021-03-26 2021-07-16 亿海蓝（北京）数据技术股份公司 Monitoring system for AIS equipment specification
CN114005303A (en) * 2021-09-29 2022-02-01 交通运输部天津水运工程科学研究所 Intelligent forecasting method and system for navigation safety conditions of water area of bridge area
CN113870620A (en) * 2021-10-19 2021-12-31 遨海科技有限公司 Ship identification method for simultaneously starting multiple AIS (automatic identification system) devices
CN113870620B (en) * 2021-10-19 2023-07-21 遨海科技有限公司 Ship identification method for simultaneously opening multiple AIS devices
CN114066303A (en) * 2021-11-25 2022-02-18 交通运输部规划研究院 AIS-based port planning adaptability analysis method and system
CN115273553A (en) * 2022-10-08 2022-11-01 交通运输部规划研究院 Space dividing method for ship restricted navigation and no-navigation area

Also Published As

Publication number Publication date
CN110706827B (en) 2022-06-03

Similar Documents

Publication Publication Date Title
CN110706827B (en) Method and system for extracting navigable water area water flow information based on ship AIS (automatic identification System) big data
Lermusiaux et al. A future for intelligent autonomous ocean observing systems
Zhang et al. Subsea pipeline leak inspection by autonomous underwater vehicle
CN104390646B (en) The location matching method of underwater hiding-machine terrain aided inertial navigation system
CN109901205B (en) Underwater robot multi-sensor fusion and motion trajectory prediction method
CN111695299B (en) Mesoscale vortex trajectory prediction method
CN113221449A (en) Ship track real-time prediction method and system based on optimal strategy learning
CN109239709A (en) A kind of autonomous construction method of local environment map of unmanned boat
CN110954132A (en) Method for carrying out navigation fault identification through GRNN (generalized regression neural network) assisted adaptive Kalman filtering
CN110135559A (en) A method of it is thrown the net based on ship position data acquisition and nets position and fishing effort
CN115015911B (en) Method and system for manufacturing and using navigation map based on radar image
CN113516626B (en) Side-scan sonar submarine sand wave detection method based on multi-scale convolution and pooling strategy
CN108362289A (en) A kind of mobile intelligent terminal PDR localization methods based on Multi-sensor Fusion
CN107064865A (en) The passive co-located method of polar coordinates Dynamic Programming clustered based on depth
Lam et al. At-sea real-time coupled four-dimensional oceanographic and acoustic forecasts during Battlespace Preparation 2007
CN115050214B (en) AIS data-based ship collision risk prediction method
CN117198330A (en) Sound source identification method and system and electronic equipment
Yang et al. An approach to ship behavior prediction based on AIS and RNN optimization model
CN104180801B (en) Forecasting Methodology and system based on ADS B system track points
CN113238208B (en) Method for calculating forward acoustic scattering Doppler frequency shift of moving target in irregular track water
CN116660996B (en) Drifting type shallow sea local earth sound parameter prediction method based on deep learning
CN114565176B (en) Long-term ship track prediction method
CN115690602A (en) Mesoscale vortex identification method and system
Divya et al. Ocean internal wave detection from SAR images using particle swarm optimization
Ooi et al. Multi-vehicle oceanographic feature exploration

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant