Ship behavior identification system, method and platform

Abstract

A ship behavior identification system, the identification system includes a data layer, a service layer and an application layer. The data layer includes a data warehouse; the service layer includes data services, algorithm services, and interface services; the application layer includes application display; the data warehouse is used to provide storage services including GIS data; the data service is used to process AIS data collected from multiple parties; the algorithm service is used to analyze and identify the behavior and status data of the ship to obtain a prediction of the behavior or status of the ship; the interface service is used to provide real-time or historical behavior and status queries for the entire life cycle of the ship; and the application display is used to provide users with a visual display of query results.

Description

Technical Field

[0001] The present invention belongs to the field of ship intelligent informatization, and in particular relates to a ship full life cycle behavior recognition system, method and platform. Background Art

[0002] With the acceleration of the process of social informatization, especially the rapid development of the Internet, the world has entered the era of big data, and the information economy has gradually been integrated into the industrial economy, driving and leading the development trend. Using advanced big data technology to accelerate the research and development and application of "Internet of Things" and "smart ships" is the main way to achieve technological progress in the industry, the main means to promote "information networking, refined management, and intelligent operation", and the main means to enhance the overall competitiveness of the industry.

[0003] With the popularization of the Automatic Identification System (AIS), it has become possible to analyze various shipping and waterway information through AIS big data mining. AIS trajectory data currently contains millions of ships in the world and is an important basis and information source for studying shipping data. Complete AIS information includes ship information (ship name, call sign, MMSI, ship number, ship type, ship length, ship width, etc.) and navigation dynamic information (longitude, latitude, ship heading, heading, speed, cargo, draft, destination, planned arrival time, etc.). Based on LBS (location-based services), AIS (automatic identification system) and other basic geographic information and vehicle location and status data, the shipping industry integrates business data, conducts big data analysis and mining, AI applications, and data visualization analysis to support the "digitalization" transformation of enterprises. It has become an industry consensus and is showing an increasingly strong demand. Summary of the invention

[0004] One embodiment of the present invention provides a ship behavior identification system, the identification system comprising a data layer, a service layer and an application layer, wherein:

[0005] The data layer includes the data warehouse,

[0006] The service layer includes data services, algorithm services, and interface services.

[0007] The application layer includes application display,

[0008] The data warehouse is used to provide storage services including GIS data.

[0009] Data service is used to process AIS data collected from multiple parties.

[0010] The algorithm service is used to analyze and identify the behavior and status data of the ship to obtain a prediction of the behavior or status of the ship.

[0011] Interface service, used to provide real-time or historical behavior and status query of the ship throughout its life cycle,

[0012] Application display is used to provide users with a visual display of query results.

[0013] The data service includes preprocessing of AIS data, and the preprocessing includes eliminating abnormal data. The sliding window method is used to eliminate the abnormal AIS data. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the detailed description below with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, in which:

[0015] Figure 1 Architecture diagram of a ship life cycle behavior recognition system according to one embodiment of the present invention.

[0016] Figure 2 Flowchart of a method for identifying ship life cycle behavior based on AIS big data according to one embodiment of the present invention. DETAILED DESCRIPTION

[0017] The current problems with the AIS system include:

[0018] (1) AIS information data has many defects that need to be resolved.

[0019] AIS information consists of two parts. One part is instrument parameters such as speed information, coordinate information, time information, etc., which have a high accuracy rate. The other part is based on manually filled information, such as destination, draft, navigation status and other information, which has poor accuracy and timeliness. Therefore, AIS information has the following problems: a large amount of ship characteristic information such as ship name is missing, and the timeliness and accuracy of destination ETA, draft and navigation status are unreliable. AIS also has problems such as missing within a certain time period and large drift at certain moments. In response to this problem, the present invention conducts relevant data analysis and mining and algorithm research based on latitude coordinates, speed and heading, and assists in the use of information such as destination, draft, navigation status, etc., so as to achieve better results and have practical application significance.

[0020] (2) Currently, the mining and utilization of AIS information data is still relatively rudimentary.

[0021] At present, there is a navigation saturation model based on ship AIS data. By analyzing a large amount of historical ship AIS data, the ship traffic flow, navigation density, and throughput capacity are calculated. The research is in the basic stage of construction, and there are problems such as inconsistent data formats, difficult to ensure accuracy, and limited analysis and mining. At present, the vast majority of applications are still in the analysis of shipping channels, and no research has been seen on information such as ship berthing. In response to this problem, the present invention proposes a time series hierarchical processing algorithm based on the latitude and longitude coordinates and speed information in the ship's AIS data. Through layer-by-layer mining of data, the ship's arrival, anchoring, berthing, leaving the berth and departure data information are finally processed, and the actual effect is very good.

[0022] According to one or more embodiments, Figure 1 As shown in the figure, the ship life cycle behavior recognition system based on AIS big data includes three levels, including data warehouse, data service, algorithm service, interface service and application display. Among them, regarding data service, it mainly collects multiple AIS signal sources, converts the format, and then stores them in the warehouse. The data warehouse uses Alibaba PolarDB, which can better support large amounts of data, and PostGis can provide fast geographic location calculation. The interface service mainly provides the front end with the query of the analysis results of the ship's entire life cycle, and supports mobile and Web terminals, providing real-time dynamics, historical dynamics, navigation tracks and other queries. The service is implemented by java. Application display mainly provides users with a visible display terminal for easy query and use, and the service is implemented by Vue.

[0023] In the algorithm service, the ship speed and longitude and latitude coordinates involved are relatively accurate data in AIS. The main basis for the ship's arrival and departure behavior is also the speed, longitude and latitude coordinates and status data of AIS. Since it is a big data processing of AIS of hundreds of thousands of ships, the algorithm needs to meet the following requirements:

[0024] (1) Satisfy the requirements of real-time processing and historical data at the same time;

[0025] (2) In order to process large amounts of data in a timely manner, it is necessary to consider the impact of algorithm complexity on computational efficiency;

[0026] (3) Breakpoint resuming calculation mechanism;

[0027] (4) To try to predict ship behaviors that have not occurred and to try to correct ship behaviors that have occurred;

[0028] Data services include data preprocessing. Because the speed is directly read from the ship's speed table, it is highly accurate and does not require preprocessing. The longitude and latitude coordinate information comes from the collection, and there may be drift or missing. For the drift of longitude and latitude coordinate points, the main thing is to detect and process the abnormal points, and detect and remove the abnormal data points.

[0029] As shown in Table 1, the coordinate point numbered 6 has obvious drift. It is impossible for the ship to sail to such a far position within the interval of Ais. The following algorithm is needed for detection and elimination.

[0030] Table 1 Missing longitude and latitude of a certain ship's coordinate point

[0031]

[0032] Assume that there is a ship AIS time series S = {P1, P2, P3, P4, P5, ..., P i ,…,P N}, where P i =(lon i ,lat i ,t i ,status i ),lon i is the longitude, lat i is the latitude, t i is the time. P is calculated i With P i+1 The distance between (i,i+1) and time difference ΔT (i,i+1) The maximum speed a ship can reach is V max When V max *ΔT (i,i+1) >ΔD (i,i+1) , it is considered that drift points have been generated and should be removed.

[0033] The sliding window method is used in the actual processing process. The first step is to determine the starting reference point. If i = 1, calculate the distance ΔD between the P1 and P2 points. (1,2) and time difference ΔT (1,2) , when V max *ΔT (1,2) >ΔD (1,2) When , it indicates that there is a drift point. Recalculate the distance ΔD between P1 and P3 (1,3) and time difference ΔT (1,3) Continuously process 5 points. When the drift point no longer appears, determine the starting reference point as P1. If there are drift points all the time, discard P1 and repeat the above calculation process with P2 as the reference until the starting reference point P is determined. i .

[0034] The second step is based on the current reference point P i When V max *ΔT (i,i+1) >ΔD (i,i+1) When the drift point P is generated, i+1, remove P i+1 , and then calculate P i and P i+2 Until P is calculated i and P i+k When drifting stops, set the current reference point to P. i+k , proceed.

[0035] In the algorithm service, the process of forming the ship status is described as follows.

[0036] After data preprocessing, based on the processed data, data analysis shows that the ship may have a value less than V when it is anchored or berthed. low knots, so V i <V low The speed of knots is regarded as the moored state, based on which the ship status is continuously monitored.

[0037] Analyze the continuity between AIS. Assume that the state formation threshold is the number of continuous states N min , N max and the maximum time threshold T max When the ship is in berth, the behavior status of the ship is obtained through comprehensive analysis based on the status field and t timestamp in AIS.

[0038] Assume that there is a ship AIS time series S = {P1, P2, P3, P4, P5, ..., P i ,…,P N}, where P i =(lon i ,lat i ,t i ,status i ),lon i is the longitude, lat i is the latitude, t i It's time, status i is a state. For example, from P i Start calculating P i Get with P i Sequence P with the same tatus value i+1 ,P i+2 ,…,P i+k , obtained from P i Start to accumulate the number k of nodes with the same status value and the time difference ΔT (i,i+k) When ΔT is satisfied (i,i+k) ≥T min &&k≥N min Or satisfy ΔT (i,i+k) ≥T maxWhen the ship status is considered to be formed, it is recorded and the following status is formed:

[0039] P i1 Anchor status = 1, P i1+1 Anchor status = 1, ..., P i1+j1 Anchor status = 1

[0040] P i2 Berth status = 5, P i2+1 Berth status = 5, ..., P i2+j2 Berth status = 5

[0041] P i3 Navigation status = 0, P i3+1 Navigation status = 0, ..., P 31+j3 Navigation status = 0

[0042] P i4 Stranded status = 6, P i4+1 Stranded status = 6, ..., P i4+j4 Stranded status = 6. So far, the AIS points are continuously detected, and the status mark is formed according to the continuous N points. The next step is to form the start and end of the status according to the status accumulation, as follows:

[0043] P i1 Anchoring starts, P i1+j1 Anchoring ended, P i2 Berthing begins, P i2+j2 Berthing completed, P i3 The voyage begins, 31+j3 The voyage ends, P i4 The stranding begins, P i4+1 The grounding was completed. Thus, the full status of the ship was obtained.

[0044] In the algorithm service, the arrival and departure behavior recognition process is described as follows.

[0045] The full status of the ship obtained in the previous article needs to be further processed to match the AIS point location, so as to obtain the port of berthing or anchoring, etc.

[0046] In the berthing or anchoring state, when the AIS coordinate P i Distance from a port Smaller than anchorage area D berth When the AIS coordinate point is marked with a port mark, each AIS point is detected and matched with a port. In order to reduce the actual amount of calculation, a step-by-step approach is adopted.

[0047] Assume that the ship behavior is obtained: P i1 Anchoring starts, Pi1+j1 Anchoring ended, P i2 Berthing begins, P i2+j2 Berthing completed, P i3 The voyage begins, 31+j3 The voyage ends. i2 =(lon i ,lat i ,t i ,status i ) point, port list coordinate point T = {T1, T2, T3, T4, T5, ..., T i ,…,T N}, T m =(lon m ,lat n ),.

[0048] The first step is box calculation, screening ports with longitude and latitude differences within a certain range. According to the even distribution of the earth's latitude, the first step is to compare the latitude Δlat and Δlat min , and then relax the comparison of Δlon and Δlon min When Δlat <= Δlat min &&Δlon<=Δlon min , enter the next step of calculation, if not satisfied, skip.

[0049] The second step is distance calculation. The coordinates of the ship's position and the port's longitude and latitude are converted into radians. The ship's position P i2 The longitude and latitude of the second point is port T m The longitude and latitude of the point are LonB, LatB). The distance D is calculated according to the following formula. Its weight R is the radius of the earth.

[0050]

[0051] D=R*Arccos(C)*pi / 180

[0052] If D <D berth , then it is considered that the ship is berthed at the current port T m At this time, the anchor point P is calculated back i1 The distance from the port, if It is considered that the ship is at P i1 Anchored in the port m , otherwise it is regarded as an intermediate anchorage.

[0053] Thus, we can get the arrival and departure behavior of ships. i1 Anchorage PortT m , P i1+j1 Anchoring ended, P i2 Berth Port Tm , P i2+j2 Berthing completed, P i3 Start sailing out of port.

[0054] The process of identifying the refueling and repair behaviors of ships involved in the algorithm service is described as follows.

[0055] According to the ship's arrival and departure behavior, the ship's refueling and repair behavior can be further analyzed. When a ship anchors at a port midway, the anchorage time T anchor ≥T gas , and when the port is a refueling port, the ship is refueling.

[0056] If the ship is berthed at the port T m , then in PostGis based on the current position P i2 , further calculate the berth and determine whether the berth is a repair berth. If it is a repair berth, the ship is under repair.

[0057] According to one or more embodiments, the composition and deployment of the ship life cycle behavior recognition system based on AIS big data are as follows:

[0058] Serial number system resource type describe 1 Data Warehouse PolarDB database Providing big data storage services 2 Data Warehouse PostGIS database Provide GIS storage and computing 3 Data Services BI Services Tools Convert AIS and store it in the warehouse 4 Algorithm Service Computing Services Program Core algorithm implementation 5 Interface Service Program Services Program Provide data query interface 6 Application Showcase Front-end services Program Provide data display 7 server server Hardware For deploying services

[0059] According to one or more embodiments, Figure 2 As shown, a ship behavior identification method is characterized in that it includes the steps of obtaining AIS data of a ship, preprocessing the obtained AIS data, and eliminating drift data; judging the state of the ship according to the AIS data, which states include berthing, sailing or anchoring; for a berthed ship, judging whether it is in a repair state; for an anchored ship, judging whether it is in a refueling state.

[0060] The present invention involves realizing the information collection, cleaning, analysis and processing of ship AIS and the display of results, and realizes the behavior recognition of the whole life cycle of ships based on AIS big data, which provides strong support for further analysis of ships, ports and routes.

[0061] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, or it can be an electrical, mechanical or other form of connection.

[0062] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0063] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0064] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A ship behavior recognition system, characterized in that: The identification system includes a data layer, a service layer and an application layer, wherein: The data layer includes the data warehouse, The service layer includes data services, algorithm services, and interface services. The application layer includes application display, The data warehouse is used to provide storage services including GIS data. Data service is used to process AIS data collected from multiple parties. The algorithm service is used to analyze and identify the behavior and status data of the ship to obtain a prediction of the behavior or status of the ship. Interface service, used to provide real-time or historical behavior and status query of the ship throughout its life cycle, Application display, used to provide users with a visual display of query results. The data service includes preprocessing of AIS data, which includes removing abnormal AIS data using a sliding window method, and this step further includes: Suppose a ship AIS time series S = {P1, P2, P3, P4, P5, ..., P i ,…,P N }, where P i =(lon i ,lat i ,t i ,status i ),lon i is the longitude, lat i is the latitude, t i is the time, and P is calculated i With P i+1 The distance between (i,i+1) and time difference ΔT (i,i+1) The maximum speed that a ship can reach is V max , The first step is to determine the starting reference point. If i=1, calculate the distance ΔD between points P1 and P2. (1,2) and time difference ΔT (1,2) , when V max *ΔT (1,2) >ΔD (1,2) When , it indicates that there is a drift point. Recalculate the distance ΔD between P1 and P3 (1,3) and time difference ΔT (1,3) And compare, process 5 points continuously, when the drift point no longer appears, determine the starting reference point as P1, If there are drift points all the time, discard point P1 and repeat the above calculation process with P2 as the reference until the starting reference point P is determined. i , The second step is based on the current reference point P i When V max *ΔT (i,i+1) >ΔD (i,i+1) When the drift point P is generated, i+1 , remove P i+1 , and then calculate P i and P i+2 Until P is calculated i and P i+k No more drift. When drift no longer occurs, set the current reference point to P i+k , proceed, The algorithm service obtains the status of the ship based on the AIS data analysis, After data preprocessing, based on the processed data, data analysis shows that the ship may have a value less than V when it is anchored or berthed. low knots, so V i <V low The speed of knots is regarded as the moored state, and the ship status is continuously monitored based on this. Assume that the state formation threshold is the number of continuous states N min , N max and the maximum time threshold T max When the ship is at anchor, the behavior status of the ship is obtained through comprehensive analysis based on the status field and t timestamp in AIS. Suppose a ship AIS time series S = {P1, P2, P3, P4, P5, ..., P i ,…,P N }, where P i =(lon i ,lat i ,t i ,status i ),lon i is the longitude, lat i is the latitude, t i It's time, status i It's the state. Assume that from P i Start calculating P i To and P i Sequence P with the same tatus value i+1 ,P i+2 ,…,P i+k , obtained from P i Start to accumulate the number k of nodes with the same status value and the time difference ΔT (i,i+k) , when ΔT is satisfied (i,i+k) ≥T min &&k≥N min Or satisfy ΔT (i,i+k) ≥T max When the ship status is considered to be formed, it is recorded and the following status is formed: P i1 Anchor status = 1, P i1+1 Anchor status = 1, ..., P i1+j1 Anchor status = 1 P i2 Berth status = 5, P i2+1 Berth status = 5, ..., P i2+j2 Berth status = 5 P i3 Navigation status = 0, P i3+1 Navigation status = 0, ..., P 31+j3 Navigation status = 0 P i4 Stranded status = 6, P i4+1 Stranded status = 6, ..., P i4+j4 Stranded status = 6 So far, the AIS points are detected continuously, and the status mark is formed according to the continuous N points. The next step is to accumulate the states and form the start and end of the states, as follows: P i1 Anchoring starts, P i1+j1 Anchoring ended, P i2 Berthing begins, P i2+j2 Berthing completed, P i3 The voyage begins, 31+j3 The voyage ends, P i4 The stranding begins, P i4+1 The grounding was completed, and the full status of the ship was obtained. The algorithm service identifies the arrival and departure behavior of the ship according to the status of the ship and the berth position, specifically including: In the berthing or anchoring state, when the AIS coordinate P i The distance D from a port is less than the anchorage range D berth When the AIS coordinate point is marked with a port mark, each AIS point is detected and matched with the port. In order to reduce the actual amount of calculation, a step-by-step approach is adopted. Assume that the ship behavior is obtained: P i1 Anchoring starts, P i1+j1 Anchoring ended, P i2 Berthing begins, P i2+j2 Berthing completed, P i3 The voyage begins, i3+j3 The voyage is over, for P i2 =(lon i2 ,lat i2 ,t i2 ,status i2 ) point, port list coordinate point T = {T1, T2, T3, T4, T5, ..., T i ,…,T N }, T m =(lon m ,lat m ), The first step is box calculation, screening ports with longitude and latitude differences within a certain range. According to the even distribution of the earth's latitude, the first step is to compare the latitude Δlat and Δlat min , and then relax the comparison of Δlon and Δlon min , When Δlat <= Δlat min &&Δlon<=Δlon min When , enter the next step of calculation, if not satisfied, skip, The second step is distance calculation. The coordinates of the ship's position and the longitude and latitude of the port are converted into radians. The ship's position P i2 The longitude and latitude of the second point is port T m The longitude and latitude of the According to the following formula, the distance D is calculated, where R is the radius of the earth. C=sin(MLatA)*sin(MlatB)*cos(MlonA-MlonB)+cos(MlatA)*cos(MlatB) D=R*Arccos(C)*pi / 180 If D <D berth , then it is considered that the ship is berthed at the current port T m , then the back calculation anchor point P i1 The distance from the port, if D <D anchor , then it is considered that the ship is at P i1 Anchored in the port m , otherwise it is regarded as an intermediate anchorage. Thus, we can get the arrival and departure behavior of ships. i1 Anchorage PortT m , P i1+j1 Anchoring ended, P i2 Berth Port T m , P i2+j2 Berthing completed, P i3 Start sailing out of the port, Further identifying the refueling and repairing behavior of the ship based on the arrival and departure behavior of the ship, including: When a ship anchors at a port midway, the anchorage time is T anchor ≥T gas , and if the port is a refueling port, the ship has refueled. If the ship is berthed at the port T m , then in PostGis based on the current position P i2 , further calculate the berth and determine whether the berth is a repair berth. If it is a repair berth, the ship is under repair.

Images