Mmsi renumbering detection method, device and system based on connected components

Abstract

The present application relates to ship automatic identification system data processing technology, especially to a kind of MMSI re-number detection method, device and system based on connected component, the method includes: obtaining the AIS message sequence of MMSI to be detected, local message segment is intercepted using sliding window mechanism, constructs space-time reachability adjacency matrix, determines the number of connected components of trajectory node in the sliding window and judges whether MMSI exists long-distance re-number behavior accordingly;The minimum deviation angle between the actual displacement direction between adjacent trajectory nodes and the ground course in the message is calculated and the logical abnormality rate of course consistency is calculated according to the minimum deviation angle, and whether MMSI exists short-distance re-number behavior is judged according to the minimum deviation angle.The present application realizes the full coverage detection of re-number in long-distance and short-distance two scenes through the double complement of physical constraint and motion logic, can improve the accuracy and adaptability of re-number detection, and can identify the behavior of more than two ships re-number in long-distance re-number determination.

Description

Technical Field

[0001] This invention relates to data processing technology for Automatic Identification System (AIS) of ships, and more particularly to a method, apparatus and system for detecting duplicate MMSI numbers based on connected components. Background Technology

[0002] AIS (Automatic Identification System) is a core system for ensuring maritime navigation safety and implementing Vessel Traffic Service (VTS). It facilitates information exchange between ships and between ships and shore through automatically broadcast messages. The Maritime Mobile Service Identifier (MMSI) is a unique identifier designated by the International Maritime Organization (IMO) and the International Telecommunication Union (ITU) for ships, coastal radio stations, search and rescue aircraft, and other similar entities. In practice, the phenomenon of "one code for multiple ships" frequently occurs, leading to misjudgments by the VTS system and impacting search and rescue efficiency. This duplicate code problem not only seriously threatens navigation safety but also poses significant challenges to maritime emergency search and rescue and shipping supervision.

[0003] Several detection methods have been proposed to address the issue of duplicate MMSI numbers, which can be mainly categorized as follows:

[0004] The first type is based on manual verification. This method relies on manual comparison of ship static information, which is prone to misjudgment and cannot handle complex scenarios where multiple ships report information alternately.

[0005] The second category is detection methods based on the frequency of AIS message broadcasts. For example, the method disclosed in Chinese Patent Publication No. CN121357524A utilizes the stability characteristics of time slot intervals in the SOTDMA (Self-Organizing Time Division Multiple Access) communication mechanism. It calculates the average deviation of the time intervals between adjacent messages with the same MMSI and compares it with the standard time slot length. If the deviation exceeds a threshold, it is determined to be a duplicate number. This type of method handles interference such as message loss and frequency switching, but its core detection logic relies solely on the statistical characteristics of the time interval. When two ships alternate reporting at the same or similar frequencies, the time interval may exhibit a pseudo-stable state, leading to missed detections. More importantly, this method completely fails to identify close-range duplicate number behavior.

[0006] The third category is detection methods based on the dynamic information of AIS messages. For example, the method disclosed in Chinese Patent Publication No. CN121644518A utilizes the principle of physical continuity of ship tracks. It calculates the actual displacement between adjacent messages and compares it with the theoretical maximum displacement (based on ground speed and time difference). If the actual displacement significantly exceeds the theoretical maximum displacement, a spatial jump is determined, and thus it is identified as a duplicate message. This type of method has a certain ability to identify duplicate messages with long-distance, large displacement jumps, but it has significant drawbacks: firstly, this method relies solely on displacement constraints and cannot identify duplicate messages at close range; secondly, this method uses single-point sequential comparison, which is overly dependent on data continuity and is susceptible to packet loss and noise interference. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide a method, apparatus and system for detecting duplicate MMSI numbers based on connected components, which improves the accuracy of MMSI duplicate detection and takes into account scenarios of long-distance spatial jumps and short-distance spoofing.

[0008] To solve the above problems, the technical solution adopted by the present invention is as follows:

[0009] In a first aspect, the present invention provides an MMSI duplicate number detection method based on connected components, the key technology of which includes the following steps:

[0010] Step S1: Obtain the AIS message sequence of the MMSI to be detected, where each message represents a trajectory node. Includes time ,latitude ,longitude and ground heading The message sequence is cleaned to remove invalid noise points that cross the latitude and longitude boundaries, and a sliding window mechanism is used to extract local message segments from the AIS message sequence.

[0011] Step S2: Within the sliding window, at the ship's maximum physical speed Assuming spatiotemporal constraints, a spatiotemporal reachability adjacency matrix is ​​constructed, and the number of connected components of the trajectory nodes within the sliding window is determined using a connected component extraction algorithm in graph theory.

[0012] Step S3: If the number of connected components is greater than 1 and the connected components overlap in time, it is determined that the MMSI has long-distance duplicate number behavior; otherwise, the data is passed to step S4 for short-distance duplicate number detection.

[0013] Step S4: Within the sliding window, calculate the minimum deviation angle between the actual displacement direction of adjacent trajectory nodes and the ground heading in the message. The system calculates the logical anomaly rate of heading consistency. If the logical anomaly rate exceeds a preset threshold, it determines that the MMSI has a close-range duplicate number behavior.

[0014] Furthermore, in step S1, the method for removing invalid noise points that cross the latitude and longitude boundaries is as follows:

[0015] Perform latitude and longitude validity checks, and mark messages with longitudes outside the range of [-180°, 180°] or latitudes outside the range of [-90°, 90°] as invalid noise and discard them;

[0016] Perform missing value cleaning to remove messages with missing or abnormal latitude and longitude coordinate values.

[0017] Furthermore, in step S2, the method for constructing the spatiotemporal reachability adjacency matrix is ​​as follows:

[0018] Calculate any two trajectory nodes within the sliding window and spherical distance between ;

[0019] Calculate the necessary speed based on the time difference between the two trajectory nodes. :

[0020] ;

[0021] like If the two trajectory nodes are physically reachable, the corresponding position in the adjacency matrix is ​​set to 1; otherwise, it is set to 0.

[0022] Among them, the maximum physical speed of the ship The value is 50-70 sections.

[0023] Furthermore, the spherical distance Calculated using the Haversine formula:

[0024] ;

[0025] ;

[0026] in, ; ;

[0027] For trajectory nodes and The central angle of the great circle at the center of the Earth; For trajectory nodes and latitude and The difference, For trajectory nodes and longitude and The difference, This is the average radius of the Earth.

[0028] Furthermore, in step S2, the step of determining the number of connected components using the connected component extraction algorithm in graph theory is as follows:

[0029] Initialize the access marker array, iterate through all nodes within the sliding window, and for each unvisited node... ,by Starting from the node, add it to the search queue. Pop nodes from the queue sequentially and add all adjacent nodes that meet the connectivity condition to the queue, until the queue is empty. A complete physically connected trajectory has been extracted. Repeat the above steps until all nodes in the window have been visited, ultimately dividing the trajectory point set T into k disjoint subsets. .

[0030] Furthermore, in step S4, the minimum deviation angle The calculation method is as follows:

[0031] Based on adjacent trajectory nodes and Calculate the actual displacement vector angle using the latitude and longitude. :

[0032] ;

[0033] Where X represents the northward component after spherical correction, and Y represents the eastward component after spherical correction.

[0034] ,

[0035] ,

[0036] ,

[0037] , and , Trajectory nodes and latitude and longitude For trajectory nodes and longitude and The difference;

[0038] ;

[0039] in The actual displacement azimuth angle is measured clockwise from true north, and its value ranges from [0, 2π].

[0040] Furthermore, in step S4, the statistical method for the logic anomaly rate is as follows:

[0041] Let the size of the sliding window be... The minimum deviation angle between all adjacent points within the statistical window. Greater than the deviation threshold Number of ;

[0042] Logical anomaly rate ;

[0043] like If so, it is determined that there is a close-range MMSI duplicate number behavior;

[0044] in, This represents the number of messages, with a value ranging from 30 to 100. This is the logic exception rate threshold.

[0045] Furthermore, the deviation threshold The angle is 30° to 50°, and the logic exception rate threshold τ is 20% to 40%.

[0046] Secondly, the present invention provides an MMSI duplicate number detection device based on connected components, comprising:

[0047] The data preprocessing unit is used to acquire the AIS message sequence of the MMSI to be detected, perform cleaning processing, and use a sliding window mechanism to extract local message fragments.

[0048] The spatiotemporal reachability analysis unit is used to determine the ship's maximum physical speed within a sliding window. Given the spatiotemporal constraints, calculate the spherical distance between any two message points within the window and construct the spatiotemporal reachability adjacency matrix;

[0049] The connected component extraction unit is used to extract connected components from the adjacency matrix using a connected component extraction algorithm in graph theory, and cluster the trajectory points into a set of k connected components.

[0050] The long-distance duplicate number determination unit is used to determine whether the MMSI has long-distance duplicate number behavior based on whether the number of connected components k is greater than 1 and whether the time intervals of each connected component overlap.

[0051] The motion logic consistency analysis unit is used to calculate the minimum deviation angle between the actual displacement direction between adjacent trajectory nodes and the ground heading in the message, and to count the logic anomaly rate of heading consistency within the sliding window.

[0052] The near-distance duplicate number determination unit is used to determine whether the MMSI has near-distance duplicate number behavior based on whether the logic anomaly rate exceeds a preset threshold.

[0053] Thirdly, the present invention provides an MMSI duplicate number detection system based on connected components, the system comprising:

[0054] One or more memories for storing instructions; and

[0055] One or more processors are configured to retrieve and execute the instructions from the memory, performing the method as described in the first aspect.

[0056] The beneficial effects of adopting the above technical solution are as follows:

[0057] 1. The present invention provides a MMSI duplicate number detection method based on connected components. Through a two-layer cascaded filtering detection of physical limit constraint analysis and motion logic analysis of ships, the method can simultaneously achieve full coverage detection of duplicate numbers in both long-distance and short-distance scenarios, which can improve the accuracy and adaptability of duplicate number detection. Furthermore, in the determination of duplicate numbers at long distances, it can simultaneously identify the behavior of duplicate numbers of two or more ships.

[0058] 2. This invention first constructs a spatiotemporal reachability matrix based on the physical motion limits of ships. Utilizing a connected component extraction algorithm from graph theory, the trajectory points of the MMSI to be detected are clustered into one or more independent connected components. By jointly judging the number of components and temporal overlap, trajectory breaks caused by long-distance spatial jumps can be effectively identified. Building upon this, for concealed duplicate MMSI numbers that have not triggered speed constraints, a motion logic consistency verification mechanism is further introduced between the reported heading-to-ground (COG) and the actual displacement direction. By statistically analyzing the logic anomaly rate within a window, close-range "stand-in" deception is accurately identified. These two complementary approaches achieve coverage of different scenarios of MMSI duplicate behavior, from long-distance to close-range.

[0059] 3. This invention employs a sliding window mechanism for local correlation analysis and utilizes a connected component algorithm to determine the trajectory based on its overall topological structure. This avoids the excessive reliance on data continuity inherent in traditional single-point sequential comparison methods and effectively overcomes the negative impacts of AIS data packet loss, latency, and noise interference on the detection results. Simultaneously, motion logic consistency verification, through decoupling analysis of the reported heading relative to the ground and the monitored actual displacement, identifies false trajectories from a kinematic perspective and further identifies near-distance duplicate trajectories, significantly reducing the probability of misjudgment. Attached Figure Description

[0060] Figure 1 This is a flowchart of the detection method of the present invention.

[0061] Figure 2 This is a simulation result diagram of connected component clustering based on spatiotemporal constraints according to the present invention; wherein, Figure 2 (a) is a schematic diagram of geospatial clustering of AIS data trajectories after processing by the connected component extraction algorithm. Figure 2 (b) is a schematic diagram of the spatiotemporal distribution and time overlap detection results.

[0062] Figure 3 This is a simulation result diagram of the close-range duplicate number detection based on heading consistency according to the present invention; wherein, Figure 3 (a) is a schematic diagram of the motion vector consistency analysis results. Figure 3 (b) is a schematic diagram of the statistical results of heading deviation. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of the present invention clearer, the invention will be described clearly and completely below in conjunction with specific embodiments.

[0064] The essence of the MMSI duplicate number detection method described in this invention is to filter out abnormal features that violate the laws of physical motion and navigation logic based on AIS message information. This method analyzes and judges the phenomenon of multiple ships sharing the same code based on the spatiotemporal reachability constraints and motion vector consistency of ship motion, from two dimensions: long-distance trajectory continuity and short-distance motion logic. Spatiotemporal reachability refers to the limitation on the range of ship movement within a finite time window during MMSI duplicate number detection, which is restricted by the maximum physical speed and dynamic characteristics. If the ship is at time... lie in Its maximum physical speed is Then at the next moment If the ship's effective physical space (i.e., the ship's current position) Departure, based on maximum physical speed At the next moment The theoretically reachable geographical area) in Centered on, with radius as Within the reachable neighborhood, it is called spatiotemporally reachable, that is:

[0065] .

[0066] If subsequent points in the AIS message sequence fall outside this neighborhood, it means that the necessary average speed between the two points exceeds the maximum physical speed. This violates the principle of trajectory continuity and is termed spatiotemporal disconnection. Therefore, the normal trajectory of a single ship should constitute a fully connected subgraph, but long-distance spatial jumps caused by MMSI duplicate numbers will disrupt its topological connectivity, causing the trajectory to split into two or more independent connected components. Based on the above phenomenon, this invention uses a clustering algorithm to identify the phenomenon of long-distance duplicate numbers.

[0067] For concealed overlapping behaviors that occur in close proximity and do not trigger the maximum physical speed limit, simple speed constraints fail, necessitating the introduction of kinematic logic constraints. The core principle is the coupling between the ship's attitude vector and its displacement observation vector. Under normal navigation conditions, the ship, as a nonholonomic constraint system, has a specific navigation direction (i.e., the ship's heading towards the ground). The direction of the instantaneous velocity is determined, that is, the direction of the actual displacement. Should be in accordance with the ship's heading towards the ground Statistically, they maintain a high degree of consistency. However, in a multi-ship scenario with one yard, the spurious displacement vectors generated by alternating message reporting are completely independent of the ship's actual course. When the observed displacement direction differs from the reported heading relative to the ground... Significant decoupling occurs, i.e., the following conditions are met:

[0068] ;

[0069] in If the deviation threshold is set, it can be determined that the trajectory is a pseudo-trajectory formed by alternating reports from different ships.

[0070] The two mechanisms mentioned above complement each other from the two dimensions of physical accessibility and motion logic consistency, overcoming AIS data packet loss and noise interference while realizing MMSI duplicate number detection in two different scenarios: long-distance jumping and close-range deception.

[0071] Example 1

[0072] like Figure 1 As shown, to address the diversity of MMSI duplicate number behavior, this invention provides an MMSI duplicate number detection method based on connected components, the method comprising the following steps:

[0073] Step S1: Obtain the AIS message sequence of the MMSI to be detected, where each message represents a trajectory node. Includes time ,latitude ,longitude and ground heading The message sequence is cleaned to remove invalid noise points that cross the latitude and longitude boundaries, and a sliding window mechanism is used to extract local message segments from the AIS message sequence.

[0074] Continuously receive AIS messages in a time series, parse the received dynamic messages, and obtain MMSI, message ID, and timestamp. ,latitude ,longitude and ground heading The AIS messages for each MMSI to be detected are stored using the MMSI as a identifier, so that each MMSI to be detected corresponds to a message sequence. Each trajectory point It contains the above information.

[0075] For any MMSI to be detected, the stored message sequence is preprocessed to remove invalid noise points with out-of-bounds latitude and longitude coordinates. The method for removing invalid noise points with out-of-bounds latitude and longitude coordinates is as follows:

[0076] Perform latitude and longitude validity checks, and mark messages with longitudes outside the range of [-180°, 180°] or latitudes outside the range of [-90°, 90°] as invalid noise and discard them;

[0077] Missing value cleaning is performed to remove messages with missing or abnormal latitude and longitude coordinate values. Specifically, messages with null or zero latitude and longitude coordinate values ​​are removed. Abnormal messages refer to messages whose parsed fields do not conform to the standard message format (refer to the standard: International Telecommunication Union. Technical characteristics of Automatic Identification System for Time Division Multiple Access in VHF Waterborne Mobile Band: ITU-R M.1371-5[S]. 2014.).

[0078] A sliding window mechanism is used to extract partial message segments from the AIS message sequence. In this embodiment, the sliding window size is... Set the number of messages to 30-100 to ensure the real-time performance and statistical reliability of local feature analysis.

[0079] Step S2: Within the sliding window, at the ship's maximum physical speed Given the spatiotemporal constraints, a spatiotemporal reachability adjacency matrix is ​​constructed, and the number of connected components of the trajectory nodes within the sliding window is determined using a connected component extraction algorithm in graph theory. This involves performing spatiotemporal reachability analysis and calculating the number of connected components.

[0080] This step is based on the ship's physical motion limits (i.e., maximum physical speed). Construct a spatiotemporal reachability adjacency matrix, and use a connected component extraction algorithm from graph theory to determine whether the trajectory is broken in physical space, i.e., to determine the number of connected components within the window. The specific implementation is as follows:

[0081] S2.1 Construction of the Spatiotemporal Reachability Matrix

[0082] For any two points within the sliding window and ( Regardless of whether two points are adjacent in the sequence, their physical reachability is calculated. First, the spherical distance between the two points is calculated using the Haversine formula. :

[0083] ;

[0084] ;

[0085] In the formula This is the Earth's average radius (approximately 6371 km). For trajectory nodes and The central angle of the great circle at the center of the Earth;

[0086] , These are the latitude values ​​of the i-th point and the j-th point, and the latitude difference, respectively. ; , These are the longitude values ​​of the i-th point and the j-th point, respectively, and the longitude difference. .

[0087] Next, calculate the necessary speed between the two points. Based on maximum physical speed Constructing an adjacency matrix ,like If two points are physically reachable, they are defined as connected. ;

[0088] ;

[0089] The above parameters can be configured according to the actual application scenario. In this embodiment, Taking 60 knots as the theoretical upper limit of a ship's maximum physical speed can cover the vast majority of commercial and high-speed ships.

[0090] S2.2 Connected Component Extraction

[0091] Adjacency matrix After the matrix is ​​constructed, a breadth-first search (BFS) algorithm is used to traverse the matrix and cluster the trajectory points into k independent connected components. The specific steps are as follows:

[0092] Initialize the access marker array, iterate through all nodes within the sliding window, and for each unvisited node... ,by Starting from the node, add it to the search queue. Pop nodes from the queue sequentially and add all adjacent nodes that meet the connectivity condition to the queue, until the queue is empty. A complete physically connected trajectory has been extracted. Repeat the above steps until all nodes in the window have been visited, ultimately dividing the trajectory point set T into k disjoint subsets. .

[0093] Step S3: Perform long-distance duplicate number discrimination: If the number of connected components is greater than 1 and the connected components overlap in time, determine that the current MMSI has long-distance duplicate number behavior and output the result; otherwise, pass the data to step S4 for short-distance duplicate number detection.

[0094] Specifically, for the original set of trajectories Number of subsets in Analysis and judgment:

[0095] like This indicates that all message nodes have a physically connected path, meaning that the ship trajectory corresponding to this MMSI is continuous on the trajectory, that is, there is no problem of duplicate numbers that are far apart. Then the data is passed to step S4 for close-range duplicate number detection.

[0096] like This indicates that there are two or more independent connected components within the window, meaning that the message nodes within the window are physically broken into multiple segments. In this case, the ship corresponding to this MMSI may have a duplicate number, and it is necessary to further determine whether the times of each connected component overlap.

[0097] set up and For connected components The start and end times, and For connected components The start and end times. If they exist. and Make That is, two connected components and If there is overlap in time, it is determined that the MMSI has long-distance duplicate number behavior; if there is no overlap in time among the connected components, it may be due to the interruption of normal trajectory caused by the ship's engine start-up and shutdown, and the data is transferred to step S4 for short-distance duplicate number detection.

[0098] Step S4, Near-distance duplicate detection based on heading consistency: Within the sliding window, calculate the actual displacement direction between each pair of adjacent trajectory nodes. With the ground heading in the message Minimum deviation angle between Based on this, the logical anomaly rate of heading consistency is calculated. If the logic exception rate exceeds a preset threshold If so, it is determined that the MMSI has a close-range duplicate number behavior.

[0099] Specifically, in step S3, when k=1, it only indicates that the necessary speed between all trajectory nodes has not exceeded the maximum physical speed, but it cannot rule out the possibility of deception by alternating reports at close range. This step identifies this by verifying the logical consistency between the ship's sailing direction and its actual displacement direction.

[0100] S4.1 Calculation of Actual Displacement Direction

[0101] Calculate neighboring points based on the received AIS message information. and Actual displacement vector angle Considering the influence of Earth's curvature, this embodiment uses the following formula to calculate the azimuth angle:

[0102] ;

[0103] ;

[0104] ;

[0105] in, It is a two-parameter arctangent function. X represents the northward component after spherical correction. Y represents the eastward component after spherical correction.

[0106] S4.2 Calculation of Minimum Deviation Angle

[0107] Obtain the ground heading from the message Calculate the minimum deviation angle between it and the actual displacement direction. :

[0108] .

[0109] in The actual displacement azimuth angle is measured clockwise from true north, and its value ranges from [0, 2π].

[0110] S4.3 Logical Error Rate Statistics

[0111] Statistical error rate within the sliding window :

[0112] ;

[0113] ;

[0114] Where W is the size of the sliding window, which is 50 in this embodiment; The deviation threshold is set to 45° in this embodiment; This embodiment uses a preset threshold value. Take 30%. If greater than the preset anomaly rate threshold If the condition is met, it is determined that the MMSI has a close-range duplicate number behavior, and the result is output.

[0115] However, those skilled in the art will understand that the above parameters can be adjusted within a reasonable range based on actual application scenarios and long-term statistical data results, for example... An angle of 30° to 50° can be used, and τ can be 20% to 40%. The number of sections can be 50 to 70, and the number of W can be 30 to 100, all of which can achieve the technical effect of the present invention.

[0116] Step S5: Output of detection results

[0117] MMSIs identified as having duplicate numbers at long or short distances are reported to the VDL system for further processing by relevant personnel or the system. For MMSIs identified as normal, a sliding window mechanism is used to extract the next segment of the message for cyclical detection, continuously monitoring for duplicate number behavior.

[0118] In this embodiment, the deviation threshold It is mainly used to capture the geometrical changes in trajectory (zigzag / Z-shaped) caused by two or more ships alternately sending AIS messages when there are multiple ships in a single yard, such as... Figure 3 As shown. To address different sea areas and navigation conditions, deviation thresholds can be set according to the actual situation. The value of .

[0119] For example, in ports, bends, or when ships are making sharp evasive maneuvers or drifting in large waves, the ship's actual course can change drastically. In such environments, if the deviation exceeds a certain threshold... If the value is too small (e.g., <30°), these normal maneuvers can easily be misjudged as trajectory anomalies. Therefore, in situations involving high maneuverability and narrow waterways, the deviation threshold... A 50° angle is acceptable, providing sufficient margin for error during normal ship turning and greatly reducing the false alarm rate in busy waters.

[0120] On open ocean or straight shipping routes, ships typically maintain a stable course. In such situations, if two ships with similar names are close together, the angle jump caused by alternating messages may be relatively gradual (less than 45°). If the deviation threshold... Larger values ​​(e.g., >30°) can easily lead to missed detections. In this case, the deviation threshold... Using 30° can significantly improve the sensitivity to small angle jumps and reduce the false negative rate.

[0121] Deviation threshold The range of values ​​for the logic anomaly rate threshold τ can be adaptively adjusted based on the actual detection accuracy.

[0122] Simulation Verification and Analysis

[0123] To verify the effectiveness and reliability of the detection method proposed in this invention, a simulation environment was built in this embodiment to test both long-distance and short-distance duplicate number scenarios. The simulation results are as follows: Figure 2 and Figure 3 As shown.

[0124] 1. Simulation of Connected Component Clustering Based on Spatiotemporal Constraints

[0125] The scenario simulates two physically separated vessels illegally using the same MMSI. Vessel A is a normal vessel, positioned in a certain sea area at approximately 20.0°N, 110.0°E, sailing at a constant speed over a long distance of 45° northeast; Vessel B is a vessel with an additional registration number, positioned in a certain sea area at approximately 30.0°N, 122.0°E, simulating being stationary or engaging in small-scale activities.

[0126] like Figure 2 As shown in (a), after processing by the connected component extraction algorithm, the AIS data trajectory is divided into two independent clusters: Cluster 1 (blue dots) represents normal vessel A, and Cluster 2 (green squares) represents vessel B with duplicate registration numbers. To avoid misjudgment of duplicate registration numbers caused by instantaneous power on / off of vessels, time overlap is assessed.

[0127] like Figure 2 As shown in (b), this figure illustrates the relationship between latitude and time. Within the same time period, the same MMSI appears both near latitude 30° and near latitude 20°, showing significant time overlap. Therefore, the algorithm determines that this MMSI exhibits long-distance duplicate number behavior.

[0128] 2. Simulation of close-range duplicate number detection based on heading consistency

[0129] The scenario simulates two ships sailing parallel to each other in the same sea area. Ship A is a normal vessel, and Ship B is a vessel with a duplicate registration number. Both are moving at a speed of 10 knots in a due north direction (heading 0°). The two ships differ in longitude by only 0.008°, with an actual distance of approximately 0.4 nautical miles. The two ships share the same MMSI and alternately transmit AIS messages at 1-minute intervals, forming... , , , The interwoven sequence, in which, For the track nodes reported by vessel A, This refers to the trajectory nodes reported by vessel B.

[0130] like Figure 3As shown in (a), the red arrow represents the reported heading. Due to AIS equipment errors and data fluctuations, the reported heading deviates slightly from the actual heading, but the overall direction is close to true north. The blue arrow represents the actual displacement (displacement detected by the method of this invention) vector, which jumps at most times. The actual heading deviates significantly from the reported heading, i.e., the angle is greater than 45°. Figure 3 (b) The angle between the actual heading and the reported heading was statistically analyzed, with a threshold of 45°. The results showed an anomaly rate of 88.2%, far exceeding the preset anomaly rate threshold of 30%. Therefore, the algorithm determined that the MMSI exhibited close-range duplicate number behavior.

[0131] 3. Simulation Conclusions

[0132] The simulation results show that the detection method proposed in this invention can effectively identify long-distance spatial jumps and short-distance motion logic anomalies caused by duplicate MMSI numbers. It has high detection accuracy and low false judgment rate, providing maritime regulatory authorities with an efficient and reliable automated identification technology for duplicate MMSI numbers.

[0133] Example 2

[0134] This embodiment provides an MMSI duplicate number detection device based on connected components, the device comprising:

[0135] The data preprocessing unit acquires the AIS message sequence of the MMSI to be detected, performs cleaning processing, removes invalid noise points that cross the latitude and longitude boundaries, and uses a sliding window mechanism to extract local message fragments.

[0136] The spatiotemporal reachability analysis unit is used to determine the ship's maximum physical speed within a sliding window. Given the spatiotemporal constraints, calculate the spherical distance between any two message points within the window and construct the spatiotemporal reachability adjacency matrix;

[0137] The connected component extraction unit is used to extract connected components from the adjacency matrix using a connected component extraction algorithm in graph theory, and cluster the trajectory points into a set of k connected components.

[0138] The long-distance duplicate number determination unit is used to determine whether the MMSI has long-distance duplicate number behavior based on whether the number of connected components k is greater than 1 and whether the time intervals of each connected component overlap.

[0139] The motion logic consistency analysis unit is used to calculate the minimum deviation angle between the actual displacement direction between adjacent trajectory nodes and the ground heading in the message, and to count the logic anomaly rate of heading consistency within the sliding window.

[0140] The near-distance duplicate number determination unit is used to determine whether the MMSI has near-distance duplicate number behavior based on whether the logic anomaly rate exceeds a preset threshold.

[0141] In one example, the unit in the above-described device may be one or more integrated circuits configured to implement the above methods, such as one or more application-specific integrated circuits (ASICs), or one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs), or a combination of at least two of these integrated circuit forms.

[0142] For example, when the units in the device can be implemented through a processing element scheduler, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processor capable of calling programs. Alternatively, these units can be integrated together to form a system-on-a-chip (SoC).

[0143] Example 3

[0144] This embodiment provides an MMSI duplicate number detection system based on connected components, the system comprising:

[0145] One or more memories for storing instructions; and

[0146] One or more processors are configured to call and execute the instructions from the memory, performing the method as described in Example 1.

[0147] The processor described above can be a CPU, microprocessor, ASIC, or one or more integrated circuits used to control the execution of the above methods. The memory is used to store necessary program instructions and data. The processor and the memory can be decoupled and located on different devices, connected via wired or wireless means to support the chip system in implementing the above functions; alternatively, the processor and the memory can be coupled to the same device. Optionally, the memory can be a storage unit within the chip, such as a register or cache; the memory can also be a storage unit located outside the chip within the terminal, such as ROM or other types of static storage devices capable of storing static information and instructions, RAM, etc.

[0148] Example 4

[0149] This embodiment provides a computer-readable storage medium, the computer-readable storage medium comprising:

[0150] When the program is run by the processor, it executes the MMSI duplicate number detection method based on connected components as described in Embodiment 1 of the present invention.

[0151] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting duplicate MMSI numbers based on connected components, characterized in that: Includes the following steps: Step S1: Obtain the AIS message sequence of the MMSI to be detected, where each message represents a trajectory node. Includes time ,latitude ,longitude and ground heading The message sequence is cleaned to remove invalid noise points that cross the latitude and longitude boundaries, and a sliding window mechanism is used to extract local message segments from the AIS message sequence. Step S2: Within the sliding window, at the ship's maximum physical speed Assuming spatiotemporal constraints, a spatiotemporal reachability adjacency matrix is ​​constructed, and the number of connected components of the trajectory nodes within the sliding window is determined using a connected component extraction algorithm in graph theory. Step S3: If the number of connected components is greater than 1 and the connected components overlap in time, it is determined that the MMSI has long-distance duplicate number behavior; otherwise, the data is passed to step S4 for short-distance duplicate number detection. Step S4: Within the sliding window, calculate the minimum deviation angle between the actual displacement direction of adjacent trajectory nodes and the ground heading in the message. The system calculates the logical anomaly rate of heading consistency. If the logical anomaly rate exceeds a preset threshold, it determines that the MMSI has a close-range duplicate number behavior.

2. The MMSI duplicate number detection method based on connected components according to claim 1, characterized in that: In step S1, the method for removing invalid noise points that cross the latitude and longitude boundaries is as follows: Perform latitude and longitude validity checks, and mark messages with longitudes outside the range of [-180°, 180°] or latitudes outside the range of [-90°, 90°] as invalid noise and discard them; Perform missing value cleaning to remove messages with missing or abnormal latitude and longitude coordinate values.

3. The MMSI duplicate number detection method based on connected components according to claim 1, characterized in that: In step S2, the method for constructing the spatiotemporal reachability adjacency matrix is ​​as follows: Calculate any two trajectory nodes within the sliding window and spherical distance between ; Calculate the necessary speed based on the time difference between the two trajectory nodes. : ； like If the two trajectory nodes are physically reachable, the corresponding position in the adjacency matrix is ​​set to 1; otherwise, it is set to 0. Among them, the maximum physical speed of the ship The value is between 50 and 70 sections.

4. The MMSI duplicate number detection method based on connected components according to claim 3, characterized in that: The spherical distance Calculated using the Haversine formula: ； ； in, ; ; For trajectory nodes and The central angle of the great circle at the center of the Earth; For trajectory nodes and latitude and The difference, For trajectory nodes and longitude and The difference, This is the average radius of the Earth.

5. The MMSI duplicate number detection method based on connected components according to claim 1, characterized in that: In step S2, the step of determining the number of connected components using the connected component extraction algorithm in graph theory is as follows: Initialize the access marker array, iterate through all nodes within the sliding window, and for each unvisited node... ,by Starting from the node, add it to the search queue. Pop the nodes in the queue one by one and add all adjacent nodes that meet the connectivity conditions to the queue until the queue is empty. A complete physical connectivity trajectory is then extracted. Repeat the above steps until all nodes within the window have been visited, ultimately dividing the trajectory point set T into k disjoint subsets. .

6. The MMSI duplicate number detection method based on connected components according to claim 1, characterized in that: In step S4, the minimum deviation angle The calculation method is as follows: Based on adjacent trajectory nodes and Calculate the actual displacement vector angle using the latitude and longitude. : ； Where X represents the northward component after spherical correction, and Y represents the eastward component after spherical correction. ， ， ， , and , Trajectory nodes and latitude and longitude For trajectory nodes and longitude and The difference; ； in The actual displacement azimuth angle is measured clockwise from true north, and its value ranges from [0, 2π].

7. The MMSI duplicate number detection method based on connected components according to claim 1, characterized in that: In step S4, the statistical method for the logic anomaly rate is as follows: Let the size of the sliding window be... The minimum deviation angle between all adjacent points within the statistical window. Greater than the deviation threshold Number of ; Logical anomaly rate ; like If so, it is determined that there is a close-range MMSI duplicate number behavior; in, This represents the number of messages, with a value ranging from 30 to 100. This is the logic exception rate threshold.

8. The MMSI duplicate number detection method based on connected components according to claim 7, characterized in that: The deviation threshold The angle is 30° to 50°, and the logic exception rate threshold τ is 20% to 40%.

9. A device for detecting duplicate MMSI numbers based on connected components, characterized in that, include: The data preprocessing unit is used to acquire the AIS message sequence of the MMSI to be detected, perform cleaning processing, and use a sliding window mechanism to extract local message fragments. The spatiotemporal reachability analysis unit is used to determine the ship's maximum physical speed within a sliding window. Given the spatiotemporal constraints, calculate the spherical distance between any two message points within the window and construct the spatiotemporal reachability adjacency matrix; The connected component extraction unit is used to extract connected components from the adjacency matrix using a connected component extraction algorithm in graph theory, and cluster the trajectory points into a set of k connected components. The long-distance duplicate number determination unit is used to determine whether the MMSI has long-distance duplicate number behavior based on whether the number of connected components k is greater than 1 and whether the time intervals of each connected component overlap. The motion logic consistency analysis unit is used to calculate the minimum deviation angle between the actual displacement direction between adjacent trajectory nodes and the ground heading in the message, and to count the logic anomaly rate of heading consistency within the sliding window. The near-distance duplicate number determination unit is used to determine whether the MMSI has near-distance duplicate number behavior based on whether the logic anomaly rate exceeds a preset threshold.

10. A system for detecting duplicate MMSI numbers based on connected components, characterized in that, The system includes: One or more memories for storing instructions; and One or more processors are configured to retrieve and execute the instructions from the memory to perform the method as described in any one of claims 1 to 8.

Images