Target behavior recognition method and device based on low-orbit communication data and medium

Abstract

The invention discloses a target behavior recognition method and device based on low-orbit communication data and a medium, and relates to the technical field of computers. The communication data comprises sea target position data of a sea area covered by a low-orbit satellite orbit flight path, sea area environment data and service data of a low-orbit satellite network state; combining the communication data with a ship automatic identification system signal and radar data, and performing data processing on the combined data; performing feature extraction on the processed data, and constructing a behavior feature model, a network feature model and an environment feature model used for identifying maritime target features by using data features; establishing a behavior template for identifying marine target behaviors; the behavior template is used for carrying out behavior recognition on the to-be-recognized marine target, the accuracy and practicability of target behavior recognition are improved, the false alarm rate is effectively reduced, early warning, real-time monitoring and post analysis of the marine target are achieved, and the marine safety management and control level is improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and particularly to a method, device, and medium for identifying target behaviors based on low-orbit communication data. Background Art

[0002] With the increasing frequency and diversification of marine activities, the identification of target behaviors based on low-orbit communication data has become an important means to ensure marine safety, promote the development of the marine economy, and protect the marine environment. The marine management department can identify and handle potential threats in real time or conduct retrospective investigations on events that have occurred by judging abnormal behaviors of targets. Current research relies too much on the analysis of AIS (Automatic Identification System) and radar data in terms of feature extraction and conclusion verification. However, both types of data have significant limitations, seriously restricting the real-time performance and accuracy of algorithms. AIS data is based on active reporting, with the risk of being tampered with, and the data update frequency is uneven. Especially in the mid- and far-sea areas, the update frequency often exceeds 15 minutes, which greatly reduces the timeliness of the data. Radar data is limited by the detection range, accuracy, and weather and sea conditions, and is particularly ineffective in identifying the target state attributes. Moreover, current research mainly focuses on spatial data such as tracks and dots. Due to the limitations of data collection methods, the comprehensive analysis of multi-dimensional data such as environmental factors and network status is insufficient. Although some trajectory features can be mined through statistical analysis, machine learning, and model prediction methods, these features can only describe a certain movement trajectory of a marine target that does not meet expectations, and cannot comprehensively reflect the complex environment and situation of the marine target. This results in a significant impact on the accuracy rate when identifying abnormal behaviors. Currently, most of the research on the identification of target behaviors based on low-orbit communication data stays at the theoretical level, and less consideration is given to the complexity and implementation difficulty in marine management practice in algorithm design and verification, which cannot meet the requirement of a low false alarm rate.

[0003] As can be seen from the above, how to improve the accuracy and practicality of the identification of target behaviors based on low-orbit communication data, effectively reduce the false alarm rate, achieve early warning, real-time monitoring, and post-event analysis of marine targets, and improve the level of marine safety control is an issue to be solved in this field. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a method, device, and medium for identifying target behaviors based on low-orbit communication data, which can improve the accuracy and practicality of the identification of target behaviors based on low-orbit communication data, effectively reduce the false alarm rate, achieve early warning, real-time monitoring, and post-event analysis of marine targets, and improve the level of marine safety control. The specific solutions are as follows:

[0005] In a first aspect, the present application discloses a method for identifying target behaviors based on low-orbit communication data, including:

[0006] Obtain the communication data of the low-orbit satellite; the communication data includes the position data of marine targets in the sea area covered by the low-orbit satellite's orbital flight path, the sea area environment data, and the service data of the low-orbit satellite network status;

[0007] Combine the communication data with the signals of the Automatic Identification System (AIS) and radar data, and perform data processing on the combined data to obtain processed data; the data processing includes data classification, data cleaning, and data parsing;

[0008] Extract features from the processed data to obtain data features, and use the data features to construct a behavior feature model, a network feature model, and an environment feature model for identifying the features of marine targets;

[0009] Establish a behavior template for identifying marine target behaviors based on the behavior feature model, the network feature model, and the environment feature model;

[0010] Use the behavior template to identify the behaviors of the marine targets to be identified.

[0011] Optionally, the obtaining of the communication data of the low-orbit satellite includes:

[0012] Obtain the position data of marine targets in the sea area covered by the real-time orbital flight path of the low-orbit satellite, the real-time measured sea area environment data sent by the on-board equipment using a data recorder or a real-time communication system, and the service data of the low-orbit satellite network status; the sea area environment data includes the temperature and salinity of seawater, the speed and direction of water flow, the wave height and wave frequency of the sea surface, and the marine meteorological conditions.

[0013] Optionally, the combining of the communication data with the signals of the Automatic Identification System (AIS) and radar data, and the performing of data processing on the combined data to obtain processed data includes:

[0014] Obtain the signals of the Automatic Identification System (AIS) and radar data; the signals of the Automatic Identification System (AIS) are the real-time dynamic information of ships; the real-time dynamic information of ships includes ship position, ship heading, ship speed, ship name, and ship type;

[0015] Combine the communication data with the signals of the Automatic Identification System (AIS) and radar data, and perform data classification, data cleaning, and data parsing on the combined data respectively to obtain processed data.

[0016] Optionally, the performing of data classification, data cleaning, and data parsing on the combined data respectively to obtain processed data includes:

[0017] Classify the combined data according to the preset classification rules and using machine learning algorithms to obtain the classified data; the machine learning algorithms include decision trees and support vector machines;

[0018] Clean the classified data to obtain the cleaned data; the data cleaning includes identifying error data, identifying inconsistent data, identifying missing values, correcting error data, correcting inconsistent data, correcting missing values, removing duplicate records, filling in missing values, standardizing data formats, and handling outliers;

[0019] Use statistical analysis methods, data visualization methods, and machine learning algorithms to analyze the cleaned data to obtain the processed data.

[0020] Optionally, extract features from the processed data to obtain data features, and use the data features to construct a behavior feature model, a network feature model, and an environmental feature model for identifying the features of maritime targets, including:

[0021] Use the density-based spatial clustering algorithm to extract features from the processed data to obtain data features; the density-based spatial clustering algorithm includes the DBSCAN clustering algorithm;

[0022] Use the data features to construct a behavior feature model, a network feature model, and an environmental feature model for identifying the features of maritime targets, and establish a feature model library based on the behavior feature model, the network feature model, and the environmental feature model.

[0023] Optionally, establish a behavior template for identifying the behavior of maritime targets based on the behavior feature model, the network feature model, and the environmental feature model, including:

[0024] Use the environmental feature model as a constraint condition, and use the behavior feature model and the network feature model to establish a behavior template for identifying the behavior of maritime targets; the behavior template includes a navigation and communication system abnormal behavior template and a navigation position abnormal behavior template.

[0025] Optionally, use the navigation and communication system abnormal behavior template to identify the navigation and communication system abnormal behavior of the maritime target to be identified, including:

[0026] Obtain the target communication data, the target Automatic Identification System (AIS) signal, and the target radar data corresponding to the maritime target to be identified, combine and process the target communication data, the target AIS signal, and the target radar data to obtain target data;

[0027] Utilize the abnormal behavior template of the communication and navigation system and determine whether there is a signal disappearance and signal appearance situation based on the signal disappearance time in the target data;

[0028] If there is a signal disappearance and signal appearance situation, then determine whether the signal disappearance and signal appearance situation is caused by network anomalies;

[0029] If the signal disappearance and signal appearance situation is not caused by network anomalies, then determine whether the signal disappearance and signal appearance situation is caused by environmental impacts;

[0030] If the signal disappearance and signal appearance situation is not caused by environmental impacts, then determine whether the signal disappearance and signal appearance situation is caused by a maritime target returning to port;

[0031] If the signal disappearance and signal appearance situation is not caused by a maritime target returning to port, then determine that the maritime target to be identified has abnormal behavior of the communication and navigation system.

[0032] Optionally, utilize the abnormal behavior template of the navigation position to identify the abnormal behavior of the navigation position of the maritime target to be identified, including:

[0033] Utilize the abnormal behavior template of the navigation position and determine whether the maritime target to be identified enters a preset electronic fence area based on the current position information in the target data;

[0034] If the maritime target to be identified does not enter the preset electronic fence area, then predict the driving route area of the maritime target to be identified based on the track, speed, course, and current position information in the target data, and determine whether the driving route area belongs to the area in the abnormal behavior rule of the navigation position. If the driving route area belongs to the area in the abnormal behavior rule of the navigation position, then give an alarm for the abnormal behavior of the navigation position;

[0035] If the maritime target to be identified has entered the preset electronic fence area, then determine whether the maritime target to be identified violates the navigation rules of the preset electronic fence area. If the maritime target to be identified violates the navigation rules of the preset electronic fence area, then give an alarm for the abnormal behavior of the navigation position.

[0036] In a second aspect, the present application discloses an electronic device, including:

[0037] A memory for storing a computer program;

[0038] A processor for executing the computer program to implement the foregoing target behavior recognition method based on low-orbit communication data.

[0039] In a third aspect, the present application discloses a computer storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the steps of the aforementioned disclosed target behavior recognition method based on low-orbit communication data are implemented.

[0040] It can be seen that the present application provides a target behavior recognition method based on low-orbit communication data, including obtaining communication data of low-orbit satellites; the communication data includes maritime target position data, sea area environment data, and service data of the low-orbit satellite network status in the sea area covered by the low-orbit satellite orbital flight path; combining the communication data with the signals of the Automatic Identification System (AIS) of ships and radar data, and performing data processing on the combined data to obtain processed data; the data processing includes data classification, data cleaning, and data parsing; extracting features from the processed data to obtain data features, and using the data features to construct a behavior feature model, a network feature model, and an environment feature model for identifying maritime target features; establishing a behavior template for identifying maritime target behaviors based on the behavior feature model, the network feature model, and the environment feature model; and using the behavior template to perform behavior recognition on the maritime target to be identified. The present application introduces the communication data of low-orbit satellites, combines the communication data with the signals of the Automatic Identification System (AIS) of ships and radar data, and performs data processing on the combined data, which can improve the timeliness and accuracy of the data. Extracting features from the processed data and using the data features to construct a behavior feature model, a network feature model, and an environment feature model for identifying maritime target features, constructing a more comprehensive feature model through multi-dimensional features, establishing a behavior template for identifying maritime target behaviors based on the behavior feature model, the network feature model, and the environment feature model, can accurately identify typical abnormal behaviors such as abnormal communication and navigation systems and abnormal navigation positions, effectively reduce the false alarm rate, use the behavior template to perform behavior recognition on the maritime target to be identified, improve the accuracy and practicality of target behavior recognition based on low-orbit communication data, realize early warning, real-time monitoring, and post-event analysis of maritime targets, and improve the level of marine safety control. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0042] Figure 1 It is a flowchart of a target behavior recognition method based on low-orbit communication data disclosed in the present application;

[0043] Figure 2A flowchart of a DBSCAN clustering algorithm disclosed in this application;

[0044] Figure 3 An example diagram of an abnormal behavior template for a communication and navigation system disclosed in this application;

[0045] Figure 4 An example diagram of an abnormal behavior template for a navigation position disclosed in this application;

[0046] Figure 5 A flowchart for identifying abnormal behaviors of a communication and navigation system disclosed in this application;

[0047] Figure 6 A flowchart for identifying abnormal behaviors of a navigation position disclosed in this application;

[0048] Figure 7 A structural diagram of an electronic device provided by this application. Detailed implementation manners

[0049] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0050] With the increasing frequency and diversification of marine activities, target behavior recognition based on low-earth orbit communication data has become an important means to ensure marine safety, promote marine economic development, and maintain the marine environment. By judging the abnormal behavior of targets, the marine management department can identify and dispose of potential threats in real time, or conduct retrospective investigations on events that have occurred. Current research relies too much on the analysis of AIS and radar data in terms of feature extraction and conclusion verification. However, both types of data have significant limitations, seriously restricting the real-time performance and accuracy of algorithms. AIS data is based on active reporting, posing a risk of being tampered with, and the data update frequency is uneven. Especially in the mid- and far-sea areas, the update frequency often exceeds 15 minutes, greatly reducing the timeliness of the data. Radar data is limited by the detection range, accuracy, and the influence of weather and sea conditions, especially being unable to cope well with the identification of target state attributes. Moreover, current research mainly focuses on spatial data such as tracks and dots. Due to the limitations of data collection means, the comprehensive analysis of multi-dimensional data such as environmental factors and network status is insufficient. Although some trajectory features can be mined through methods such as statistical analysis, machine learning, and model prediction, these features can only describe a certain movement trajectory of marine targets that does not meet expectations, and cannot comprehensively reflect the complex environment and situation of marine targets. This results in a significant impact on the accuracy rate when identifying abnormal behaviors. Currently, most of the research on target behavior recognition based on low-earth orbit communication data remains at the theoretical level. The algorithm design and verification consider less the complexity and implementation difficulty in marine management practice and cannot meet the requirement of a low false alarm rate. As can be seen from the above, how to improve the accuracy and practicality of target behavior recognition based on low-earth orbit communication data, effectively reduce the false alarm rate, achieve early warning, real-time monitoring, and post-event analysis of marine targets, and improve the level of marine safety control is an issue to be solved in this field.

[0051] See Figure 1 As shown, an embodiment of the present invention discloses a method for target behavior recognition based on low-earth orbit communication data, which may specifically include:

[0052] Step S11: Obtain the communication data of the low-earth orbit satellite; the communication data includes the position data of marine targets, the sea area environment data, and the service data of the low-earth orbit satellite network status in the sea area covered by the low-earth orbit satellite's orbital flight path.

[0053] In this embodiment, obtain the position data of marine targets in the sea area covered by the real-time orbital flight path from the low-earth orbit satellite, the real-time measured sea area environment data sent by the shipborne equipment using a data recorder or a real-time communication system, and the service data of the low-earth orbit satellite network status; the sea area environment data includes the temperature and salinity of seawater, the speed and direction of water flow, the wave height and wave frequency of the sea surface, and the marine meteorological conditions.

[0054] Specifically, the LEO satellite covers a specific sea area through its orbital flight path and collects target position data in real time. The process of the shipborne equipment obtaining sea area environmental data mainly relies on the collaborative work of a variety of sensors and instruments. These devices usually include temperature sensors, salinity sensors, current meters, wave sensors, and weather stations, etc. First of all, the temperature sensor and the salinity sensor are installed under the hull and can measure the temperature and salinity of seawater in real time. These data are crucial for understanding the physical characteristics and ecological environment of the ocean. Secondly, the current meter is used to measure the speed and direction of the water flow to help analyze the dynamic changes of ocean currents. The wave sensor can monitor the wave height and wave frequency of the sea surface and provide important information for navigation safety. In addition, the shipborne weather station is equipped with an anemometer, a thermometer, a barometer, and a humidity sensor, and can record meteorological conditions in real time, such as wind speed, wind direction, air temperature, and air pressure, etc. After the data is collected, the shipborne equipment transmits the environmental data to the computer system on the ship through a data recorder or a real-time communication system and then sends it to the LEO satellite. In addition, the LEO satellite can monitor parameters such as the temperature of the ocean surface and wave height through multi-spectral sensors. These sensors can capture the reflected spectra of different bands. The LEO satellite usually transmits the collected data back to the ground station through a satellite communication network. The ground station is responsible for receiving, storing, and processing these data. The IoT terminal of the LEO satellite can also transmit the temperature, humidity, wind direction, wind speed, and other environments of the maritime target.

[0055] Step S12: Combine the communication data with the signals of the Automatic Identification System (AIS) of the ship and the radar data, and perform data processing on the combined data to obtain processed data; the data processing includes data classification, data cleaning, and data parsing.

[0056] In this embodiment, the signals of the Automatic Identification System (AIS) of the ship and the radar data are obtained; the signals of the Automatic Identification System (AIS) of the ship are the real-time dynamic information of the ship; the real-time dynamic information of the ship includes ship position, ship heading, ship speed, ship name, and ship type; the communication data is combined with the signals of the Automatic Identification System (AIS) of the ship and the radar data, and data classification, data cleaning, and data parsing are respectively performed on the combined data to obtain processed data.

[0057] Among them, the data processing process is as follows: classify the combined data according to the preset classification rules and using machine learning algorithms to obtain classified data; the machine learning algorithms include decision trees and support vector machines; perform data cleaning on the classified data to obtain cleaned data; the data cleaning includes identifying error data, identifying inconsistent data, identifying missing values, correcting error data, correcting inconsistent data, correcting missing values, removing duplicate records, filling in missing values, standardizing the data format, and handling outliers; use statistical analysis methods, data visualization methods, and machine learning algorithms to perform data parsing on the cleaned data to obtain processed data.

[0058] In this application, AIS (Automatic Identification System) signals and radar data are obtained through a preset acquisition method. AIS signals are the position information actively sent by ships during navigation. Low-earth orbit satellites can receive these signals to obtain the real-time dynamic information of ships. AIS data usually includes the dynamic information of ships, such as position, course, speed, ship name, ship type, etc. Radar data, on the other hand, provides real-time detection information of targets, including the distance, speed, direction, etc. of the targets. Then, data classification, data cleaning, and data parsing are three important links in the data analysis process, aiming to improve the quality and usability of data and thus provide support for decision-making: First, data classification is the process of grouping data according to specific criteria or characteristics. By using machine learning algorithms (such as decision trees, support vector machines, etc.) or rule-based methods, data can be effectively classified into different categories. This process not only helps to identify the structure and patterns of data but also provides a clear framework for subsequent analysis; Second, data cleaning is a key step to ensure data quality. Data cleaning includes identifying and correcting errors, inconsistencies, and missing values in the data. Common cleaning operations include removing duplicate records, filling in missing values, standardizing data formats, and handling outliers. Through these operations, the accuracy and integrity of the data are improved, thereby reducing biases and errors in the analysis process; Finally, data parsing is to conduct in-depth analysis on the cleaned data to extract valuable information and insights. Data parsing can adopt techniques such as statistical analysis, data visualization, and machine learning.

[0059] Step S13: Extract features from the processed data to obtain data features, and use the data features to construct a behavior feature model, a network feature model, and an environmental feature model for identifying the features of maritime targets.

[0060] In this embodiment, the density-based spatial clustering algorithm is used to extract features from the processed data to obtain data features; the density-based spatial clustering algorithm includes the DBSCAN clustering algorithm; use the data features to construct a behavior feature model, a network feature model, and an environmental feature model for identifying the features of maritime targets, and establish a feature model library based on the behavior feature model, the network feature model, and the environmental feature model.

[0061] In this application, a clustering algorithm model is used for data feature extraction. Clustering analysis is to group data objects according to the information found in the data that describes the objects and their relationships. Its goal is that the objects within a group are similar to each other, while the objects in different groups are not similar, that is, the greater the similarity within the group and the greater the difference between groups, the better the clustering effect. Taking the DBSCAN clustering algorithm as an example, the DBSCAN clustering algorithm process is as Figure 2As shown in the figure, the core idea of the DBSCAN algorithm is to cluster high-density regions of spatio-temporal data into clusters, and it is required that the number of other objects within the given Eps-Neighborhood (radius range) of each object in the cluster must reach a certain set threshold MinPts (Minimum Points). When using the DBSCAN clustering algorithm to cluster spatio-temporal data, all data is first marked as unvisited, and then clustering starts randomly from any point in the point cloud dataset. For any selected point p, first search for the objects in the Eps-neighborhood of point p and mark the point as visited. If the number of objects in the Eps-neighborhood of the point is greater than or equal to MinPts, that is, point p is a core point; then for each unvisited point in the Eps-neighborhood objects of point p, search for its Eps-neighborhood objects in turn, and loop through the above steps until all density-connected points in the cluster are found.

[0062] Step S14: Establish a behavior template for identifying the behaviors of maritime targets based on the behavior feature model, the network feature model, and the environmental feature model.

[0063] In this embodiment, the environmental feature model is used as a constraint condition, and the behavior feature model and the network feature model are used to establish a behavior template for identifying the behaviors of maritime targets; the behavior template includes a navigation and communication system abnormal behavior template and a navigation position abnormal behavior template.

[0064] In this embodiment, a complete behavior template should include content such as the target behavior intention, the participating units and the relationships between the participating units, the target attributes, the actions and the relationships between the actions, etc. For the abnormal behaviors of the navigation and communication systems and the position abnormal behaviors of maritime targets studied in the present invention, the following behavior templates are established:

[0065] (1) Navigation and communication system abnormal behavior template. The abnormal behavior of the navigation and communication system of a maritime target refers to the abnormal disappearance or abnormal appearance behavior of the ship's navigation and communication system. Based on this, a navigation and communication system abnormal behavior template is constructed as Figure 3 shown:

[0066] Participants: Ships;

[0067] Attributes (features): Navigation and communication signal disappearance / appearance time: According to the disappearance time length of the track point positions, network communication data, and the speed information before the point trace disappears, it is possible to comprehensively judge whether the navigation and communication signal of the target disappears; this data can be obtained through the point trace reporting time of the AIS data source and the satellite communication data source;

[0068] Speed: Through the activity area and speed characteristics, it is possible to comprehensively judge whether the target is in a moored state; this characteristic can be obtained comprehensively through the point traces and time reported by the AIS data source;

[0069] Meteorological data: This feature can comprehensively determine whether the disappearance of communication and navigation signals is affected by weather based on the ambient temperature, wind direction, wave height, etc. around the ship reported by the ship meteorological instrument data source;

[0070] Network communication: According to the network status data and terminal location data of the low-earth orbit satellite network terminal, it can be comprehensively determined whether the disappearance of communication and navigation signals is an abnormal disappearance; this data can be transmitted back by the low-earth orbit satellite network terminal;

[0071] Note: When the disappearance time of the communication and navigation signals reaches a certain threshold and there is no behavior such as untimely signal transmission due to environmental interference, this event can be considered to approximately occur.

[0072] (2) Abnormal behavior template for navigation position. Abnormal navigation position refers to the abnormal behavior that occurs when the ship's navigation position is in the area of concern. The abnormal behavior template for navigation position is constructed as Figure 4 shown below:

[0073] Participants: Ship;

[0074] Attributes (features): Spatial relationship with the area of concern: The ship's entry into the area of concern is one of the important bases for determining abnormal behavior;

[0075] Communication terminal location: Based on the location data of the low-earth orbit satellite network terminal, AIS-reported data, and radar data, comprehensively determine the abnormal behavior of the navigation position;

[0076] Track: Through the track after fusion processing, the course and initial position features can be extracted, which is beneficial to determining the navigation position information;

[0077] Track features can be obtained by fusing data points from AIS, radar, etc.;

[0078] Note: Through the position information of the ship and the area of concern, use relevant algorithms for judgment.

[0079] Step S15: Use the behavior template to perform behavior recognition on the sea target to be recognized.

[0080] In this embodiment, the specific process of using the abnormal behavior template of the communication and navigation system to identify the abnormal behavior of the communication and navigation system of the sea target to be identified is as follows: Obtain the target communication data, the target Automatic Identification System (AIS) signal, and the target radar data corresponding to the sea target to be identified, perform data combination and data processing on the target communication data, the target AIS signal, and the target radar data to obtain target data; Use the abnormal behavior template of the communication and navigation system and determine whether there are situations of signal disappearance and signal appearance according to the signal disappearance time in the target data; If there are situations of signal disappearance and signal appearance, determine whether the situations of signal disappearance and signal appearance are caused by network anomalies; If the situations of signal disappearance and signal appearance are not caused by network anomalies, determine whether the situations of signal disappearance and signal appearance are caused by environmental impacts; If the situations of signal disappearance and signal appearance are not caused by environmental impacts, determine whether the situations of signal disappearance and signal appearance are caused by the sea target returning to the port; If the situations of signal disappearance and signal appearance are not caused by the sea target returning to the port, determine that the sea target to be identified has abnormal behavior of the communication and navigation system.

[0081] In this embodiment, the specific process of using the abnormal behavior template of the navigation position to identify the abnormal behavior of the navigation position of the sea target to be identified is as follows: Use the abnormal behavior template of the navigation position and determine whether the sea target to be identified enters the preset electronic fence area according to the current position information in the target data; If the sea target to be identified does not enter the preset electronic fence area, predict the driving route area of the sea target to be identified according to the track, speed, course, and current position information in the target data, and determine whether the driving route area belongs to the area in the abnormal behavior rules of the navigation position. If the driving route area belongs to the area in the abnormal behavior rules of the navigation position, issue an alarm for abnormal behavior of the navigation position; If the sea target to be identified has entered the preset electronic fence area, determine whether the sea target to be identified violates the navigation rules of the preset electronic fence area. If the sea target to be identified violates the navigation rules of the preset electronic fence area, issue an alarm for abnormal behavior of the navigation position.

[0082] This application discovers two types of abnormal behaviors through a template abnormal behavior reasoning method for ocean management. First, based on the ship feature database established by the clustering algorithm, a diagnosis is made by comprehensively considering the established typical target abnormal behavior templates, and a data structure for typical abnormal behavior hypotheses is established. The data structure of the specific abnormal behavior established by the system has the format of a pattern, including slots, constraint relationships, and hierarchical relationships with other data structures. It is a hypothetical model of the observed target activities. Then, the reasoning process calculates the matching degree between the actions of the observed target and the specific template structure. During the matching process, the observed behavior characteristics are compared with the constraint relationship slots of the specific template. If the behavior characteristics well conform to these constraint relationships, the slot is filled. Each filled slot will increase the evaluation value of the overall matching degree between the situation activity and the specific template structure. The matching process is carried out whenever a new behavior activity is detected, resulting in the continuous increase of the matching degree of certain specific behavior patterns over time.

[0083] Among them, the process of identifying abnormal behaviors of the communication and navigation system for ocean management is as Figure 5 shown, and the steps are as follows:

[0084] In the first step, discover the maritime target to be identified, perform information fusion processing, and extract relevant information such as the terminal location information and network communication data information in the network communication device, the track, signal disappearance time, disappearance location, etc. in the AIS device, and the environmental information such as temperature, humidity, sea conditions, and wind speed transmitted back by the Internet of Things devices;

[0085] In the second step, judge whether there is a signal disappearance and reappearance pair according to the signal disappearance time of the target communication and navigation system (that is, judge whether there is a situation of signal disappearance and signal reappearance). If a signal disappearance and reappearance pair appears, enter the third step. If a signal disappearance pair appears, enter the fourth step. If neither appears, store it in the target behavior database;

[0086] In the third step, comprehensively judge whether the communication and navigation signal disappearance and reappearance event occurs due to network anomalies based on the network communication data. If network communication data is continuously received and the network link is normal, enter the fourth step. If no network communication data is received, it means that the communication and navigation signal is temporarily abnormal due to satellite network problems and is stored in the target behavior database;

[0087] In the fourth step, carry out the determination of abnormal behavior events by comprehensively considering the environmental data, extract the environmental data information, and analyze whether the communication and navigation signal temporarily disappears due to the influence of the environmental data. If there is an environmental influence, wait for the communication and navigation signal to appear and end the process. If there is no environmental influence, enter the fifth step;

[0088] Step 5: Determine whether the sea target to be identified is returning to port by comprehensively considering port location information at home and abroad. If the communication and navigation signals disappear due to the target returning to port, store it in the target behavior database and end the process. If the sea target to be identified does not return to port, determine that the target has abnormal behavior, store it in the abnormal behavior database, and end the process.

[0089] Furthermore, the process of identifying abnormal behavior of navigation positions for ocean management is as Figure 6 shown below. The steps are as follows:

[0090] Step 1: Detect the sea target to be identified, perform information fusion processing, and extract the speed, course, track from the AIS device, and the terminal location information from the network communication device.

[0091] Step 2: Determine whether the current location information of the sea target to be identified has entered the preset electronic fence area (such as an electronic fence area with custom rules, etc.). If it has entered the preset electronic fence area, go to Step 4; otherwise, go to Step 3.

[0092] Step 3: Determine the area where the sea target to be identified is heading. Infer the possible areas it may pass through in the future based on information such as the target's track, speed, course, and location. If it may pass through the area of concern, combine the nature of the electronic fence area to decide whether to give a warning message.

[0093] Step 4: The sea target to be identified has entered the preset electronic fence area. If it has entered the electronic fence area, combine information such as the area supervision type (speed limit area, course limit area, no-go area, etc.), speed, track, dot track, and course to determine whether it violates the electronic fence area type and whether an alarm message should be given.

[0094] This application mainly uses low-Earth orbit satellite communication data, supplemented by AIS and radar data, to build a research data foundation. Low-Earth orbit satellites have the characteristics of low orbit, fast operating speed, and global service. Their communication data has high transmission rate, low latency, and can cover the mid and far sea areas. Compared with traditional AIS and radar data, low-Earth orbit satellite communication data has higher real-time performance and accuracy. Moreover, by introducing characteristic factors of the marine environment and network status, a multi-dimensional algorithm model is constructed, significantly improving the real-time performance and accuracy of the data. By integrating multi-dimensional features such as the marine environment and network status, a more comprehensive abnormal behavior recognition model is constructed. The present invention intends to expand the data dimension and deeply mine on the basis of traditional maritime target track analysis. On the one hand, analyze environmental factors such as temperature, humidity, wind direction, and wind speed of maritime targets transmitted back by low-Earth orbit satellite Internet of Things terminals, and establish a model to identify their abnormal changes or abnormal values. On the other hand, analyze data such as the network service status and network terminal location of low-Earth orbit satellites, and construct a "maritime electronic fence". Use the DBSCAN clustering algorithm to establish a feature model library, and combine it with the template abnormal behavior reasoning method for ocean management, which can accurately identify typical abnormal behaviors such as abnormal navigation and communication systems and abnormal navigation positions, effectively reducing the false alarm rate. In addition, this application combines data feature analysis with template reasoning to improve the adaptability of the algorithm to business management. On the basis of deeply understanding the internal laws and patterns of business requirements, study the template abnormal behavior reasoning method for ocean management, construct a business model combining behavior features, environmental features, and network features, and verify it with actual data for actual scenarios. It provides a new idea for abnormal behavior recognition for ocean management departments, can realize early warning, real-time monitoring, and post-event analysis of maritime targets, improve the level of ocean security control, and at the same time promote the application and popularization of low-Earth orbit satellites in the ocean field. It takes the lead in combining the new network form of low-Earth orbit satellites with behavior feature analysis, opening up a new technical path for maritime target recognition research, and has important innovative significance.

[0095] In this embodiment, communication data of a low-Earth orbit satellite is obtained; the communication data includes maritime target position data, maritime environment data, and service data of the low-Earth orbit satellite network status in the sea areas covered by the low-Earth orbit satellite's orbital flight path; the communication data is combined with the signals of the Automatic Identification System (AIS) of ships and radar data, and the combined data is processed to obtain processed data; the data processing includes data classification, data cleaning, and data parsing; feature extraction is performed on the processed data to obtain data features, and the data features are used to construct a behavior feature model, a network feature model, and an environment feature model for identifying the features of maritime targets; a behavior template for identifying the behaviors of maritime targets is established based on the behavior feature model, the network feature model, and the environment feature model; the behavior template is used to perform behavior identification on the maritime targets to be identified. This application introduces the communication data of low-Earth orbit satellites, combines the communication data with the signals of the Automatic Identification System (AIS) of ships and radar data, processes the combined data, which can improve the timeliness and accuracy of the data. Feature extraction is performed on the processed data, and the data features are used to construct a behavior feature model, a network feature model, and an environment feature model for identifying the features of maritime targets. Through multi-dimensional features, a more comprehensive feature model is constructed. A behavior template for identifying the behaviors of maritime targets is established based on the behavior feature model, the network feature model, and the environment feature model, which can accurately identify typical abnormal behaviors such as abnormal communication and navigation systems and abnormal navigation positions, effectively reduce the false alarm rate, and use the behavior template to perform behavior identification on the maritime targets to be identified, improving the accuracy and practicality of target behavior identification based on low-Earth orbit communication data, realizing early warning, real-time monitoring, and post-event analysis of maritime targets, and enhancing the level of maritime safety control.

[0096] Figure 7 FIG. 4 is a schematic structural diagram of an electronic device provided by an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. Among them, the memory 22 is used to store a computer program, and the computer program is loaded and executed by the processor 21 to implement the relevant steps in the method for identifying target behaviors based on low-Earth orbit communication data executed by the electronic device disclosed in any of the foregoing embodiments.

[0097] In this embodiment, the power supply 23 is used to provide working voltages for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows is any communication protocol applicable to the technical solution of the present application, and no specific limitation is imposed thereon here; the input / output interface 25 is used to obtain external input data or output data to the outside, and its specific interface type can be selected according to specific application requirements, and no specific limitation is made here.

[0098] In addition, as a carrier for storing resources, the memory 22 can be a read-only memory, a random access memory, a magnetic disk, an optical disk, etc. The resources stored thereon include an operating system 221, a computer program 222, data 223, etc. The storage method can be temporary storage or permanent storage.

[0099] Among them, the operating system 221 is used to manage and control each hardware device on the electronic device 20 and the computer program 222, so as to enable the processor 21 to perform operations and processing on the data 223 in the memory 22. It can be Windows, Unix, Linux, etc. In addition to the computer program that can be used to complete the target behavior recognition method based on low-orbit communication data executed by the electronic device 20 disclosed in any of the foregoing embodiments, the computer program 222 can further include computer programs that can be used to complete other specific tasks. In addition to the data that can include the data transmitted by external devices received by the target behavior recognition device based on low-orbit communication data, the data 223 can also include the data collected by its own input / output interface 25, etc.

[0100] The steps of the method or algorithm described in combination with the embodiments disclosed in this document can be directly implemented by hardware, software modules executed by a processor, or a combination of both. The software module can be placed in a random access memory (RAM), memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, register, hard disk, removable disk, CD-ROM, or any other form of storage medium well known in the technical field.

[0101] Furthermore, the embodiments of the present application also disclose a computer-readable storage medium. When the computer program stored in the storage medium is loaded and executed by a processor, the steps of the target behavior recognition method based on low-orbit communication data disclosed in any of the foregoing embodiments are implemented.

[0102] Finally, it should also be noted that in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of another identical element in the process, method, article or device including the element.

[0103] The above has introduced in detail a method, device, and storage medium for target behavior recognition based on low-orbit communication data. Specific examples are used in this article to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A target behavior recognition method based on low-orbit communication data, characterized in that: include: Acquire communication data of the low-orbit satellite; the communication data includes sea target location data, sea environment data, and service data of the low-orbit satellite network status in the sea area covered by the low-orbit satellite orbital flight path; Combining the communication data with the ship automatic identification system signal and the radar data, and performing data processing on the combined data to obtain processed data; the data processing includes data classification, data cleaning and data analysis; Extracting features from the processed data to obtain data features, and using the data features to construct a behavior feature model, a network feature model, and an environment feature model for identifying features of maritime targets; Establishing a behavior template for identifying maritime target behavior based on the behavior feature model, the network feature model, and the environment feature model; The behavior template is used to perform behavior recognition on the sea target to be recognized.

2. The target behavior recognition method based on low-orbit communication data according to claim 1 is characterized in that: The obtaining of communication data of the low-orbit satellite comprises: The sea target position data of the sea area covered by the real-time orbital flight path, the real-time measured sea environment data sent by shipborne equipment using data recorders or real-time communication systems, and the service data of the low-orbit satellite network status are obtained from low-orbit satellites; the sea environment data include seawater temperature and salinity, water flow speed and direction, sea surface wave height and wave frequency, and marine meteorological conditions.

3. The target behavior recognition method based on low-orbit communication data according to claim 1 is characterized in that: The step of combining the communication data with the ship automatic identification system signal and the radar data, and performing data processing on the combined data to obtain processed data, includes: Acquire ship automatic identification system signals and radar data; the ship automatic identification system signals are real-time dynamic information of the ship; the real-time dynamic information of the ship includes the ship's position, ship's heading, ship's speed, ship's name, and ship's type; The communication data is combined with the ship automatic identification system signal and the radar data, and the combined data are respectively classified, cleaned and analyzed to obtain processed data.

4. The target behavior recognition method based on low-orbit communication data according to claim 3 is characterized in that: The combined data are classified, cleaned and analyzed to obtain processed data, including: Classifying the combined data according to preset classification rules and using a machine learning algorithm to obtain classified data; the machine learning algorithm includes a decision tree and a support vector machine; Performing data cleaning on the classified data to obtain cleaned data; the data cleaning includes identifying erroneous data, identifying inconsistent data, identifying missing values, correcting erroneous data, correcting inconsistent data, correcting missing values, removing duplicate records, filling missing values, standardizing data formats, and processing outliers; Statistical analysis, data visualization, and machine learning algorithms are used to analyze the cleaned data to obtain processed data.

5. The target behavior recognition method based on low-orbit communication data according to claim 1 is characterized in that: The feature extraction of the processed data to obtain data features, and the use of the data features to construct a behavior feature model, a network feature model, and an environment feature model for identifying maritime target features, include: Using a density-based spatial clustering algorithm to extract features from the processed data to obtain data features; the density-based spatial clustering algorithm includes a DBSCAN clustering algorithm; A behavior feature model, a network feature model and an environment feature model for identifying maritime target features are constructed using data features, and a feature model library is established based on the behavior feature model, the network feature model and the environment feature model.

6. The target behavior recognition method based on low-orbit communication data according to any one of claims 1 to 5, characterized in that: The establishing of a behavior template for identifying maritime target behavior based on the behavior feature model, the network feature model, and the environment feature model includes: The environmental feature model is used as a constraint condition, and the behavior feature model and the network feature model are used to establish a behavior template for identifying maritime target behavior; the behavior template includes an abnormal behavior template of a communication and navigation system and an abnormal behavior template of a navigation position.

7. The target behavior recognition method based on low-orbit communication data according to claim 6 is characterized in that: The abnormal behavior template of the communication and navigation system is used to identify the abnormal behavior of the communication and navigation system of the maritime target to be identified, including: Acquire target communication data, target ship automatic identification system signal and target radar data corresponding to the marine target to be identified, combine and process the target communication data, the target ship automatic identification system signal and the target radar data to obtain target data; Using the abnormal behavior template of the communication and navigation system and judging whether there is a signal disappearance or signal appearance according to the signal disappearance time in the target data; If there is a situation where the signal disappears and the signal appears, determine whether the situation where the signal disappears and the signal appears is caused by a network abnormality; If the signal disappearance and signal appearance are not caused by network abnormalities, determine whether the signal disappearance and signal appearance are caused by environmental influences; If the signal disappearance and signal appearance are not caused by environmental influences, it is determined whether the signal disappearance and signal appearance are caused by the return of the target at sea; If the signal disappearance and signal appearance are not caused by the return of the maritime target to the port, it is determined that the maritime target to be identified has abnormal behavior in the communication and navigation system.

8. The target behavior recognition method based on low-orbit communication data according to claim 7 is characterized in that: The navigation position abnormal behavior template is used to identify the abnormal navigation position behavior of the maritime target to be identified, including: Using the navigation position abnormal behavior template and according to the current position information in the target data, determining whether the sea target to be identified has entered the preset electronic fence area; If the sea target to be identified does not enter the preset electronic fence area, the driving path area of ​​the sea target to be identified is predicted according to the track, speed, heading, and current position information in the target data, and it is determined whether the driving path area belongs to the area in the navigation position abnormal behavior rule. If the driving path area belongs to the area in the navigation position abnormal behavior rule, an abnormal navigation position behavior alarm is issued; If the sea target to be identified has entered the preset electronic fence area, it is determined whether the sea target to be identified violates the navigation rules of the preset electronic fence area. If the sea target to be identified violates the navigation rules of the preset electronic fence area, an abnormal navigation position behavior alarm is issued.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the target behavior recognition method based on low-orbit communication data as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: Used to store computer programs; wherein, when the computer program is executed by a processor, the target behavior recognition method based on low-orbit communication data as described in any one of claims 1 to 8 is implemented.

Images