Off-site law enforcement method for channel improvement building damage based on multi-source spatio-temporal data

Abstract

The invention belongs to the technical field of channel improvement building damage event prediction methods, and particularly relates to a channel improvement building damage off-site law enforcement method based on multi-source spatio-temporal data. The multi-source spatio-temporal data comprises ship AIS data, ship report data, historical ship report data, a channel improvement building data set, real-time tide level data, ship stability characteristic data, ship radar scanning data, video monitoring and intelligent identification data and meteorological real-time data; building collision risk scenes are improved according to three types of channels including unintentional collision, hydrological influence and illegal navigation, and corresponding off-site law enforcement algorithms are executed respectively; and sequentially operating a first arrival crew identification algorithm, a high-risk water area identification algorithm in the tidal fluctuation and ebb period and a trajectory-based off-site law enforcement algorithm to realize collision risk judgment.

Description

Technical Field

[0001] This invention belongs to the technical field of methods for predicting waterway dredging structure damage events, and particularly relates to a non-site enforcement method for waterway dredging structure damage based on multi-source spatiotemporal data. Background Technology

[0002] In the current field of waterway regulation and construction supervision, traditional monitoring methods generally suffer from problems such as single data sources and low integration, making it difficult to form a comprehensive risk perception capability. Collision avoidance mainly relies on AIS and video for dynamic tracking, which cannot cover AIS signal blind spots and fails to effectively integrate key information such as crew historical operation data, real-time hydrological tide levels, meteorological parameters, and waterway structure attributes. For example, for crew members arriving at the port for the first time or those who have not been to the port for a long time, there is a lack of criteria for judging their familiarity with the waters, making it impossible to provide advance information on water features and structure distribution, which can easily lead to unintentional collisions due to crew misjudgments. At the same time, the lack of correlation analysis between real-time tide levels and the elevation of structure tops makes it difficult to dynamically identify high-risk waters during high and low tides, resulting in a failure to provide timely warnings of collision hazards caused by hydrological changes.

[0003] At the risk assessment algorithm level, existing technologies lack a hierarchical and scenario-based precise assessment mechanism, resulting in insufficient timeliness and specificity of early warnings. For risk prediction of ship navigation trajectories, traditional methods mostly trigger warnings based on fixed distance thresholds, failing to combine real-time ship speed and heading angle to calculate trajectory landing points under different deceleration times, and also failing to set geographical error tolerances, making it prone to misjudgment or omission. For tidal scenarios, there are neither clear safety thresholds to define high-risk structures nor reasonable danger zones to delineate around risky structures, making it impossible to accurately monitor whether ships have entered dangerous areas. In addition, there is a lack of coordination and linkage between algorithms. The identification of first-time arriving crew members, hydrological risk identification, and trajectory prediction are all independent, failing to form a three-dimensional risk assessment closed loop involving crew members, ships, and the environment, resulting in some complex risks not being effectively captured.

[0004] In the enforcement response and evidence chain management stages, traditional models suffer from problems such as inaccurate information transmission, low handling efficiency, and incomplete evidence retention. Enforcement information dissemination often uses standardized scripts, failing to tailor content to different scenarios such as first-time arrivals, high-risk waters, and illegal navigation, resulting in limited effective information received by crew members. The dispatch of enforcement resources such as drones and patrol vessels relies on manual commands, lacking automatic linkage with risk assessment results, leading to delays in emergency response. Furthermore, after a collision risk or incident, the evidence chain must be manually collected and integrated, which is not only inefficient but also prone to data omissions or tampering, affecting subsequent enforcement tracing. In addition, the lack of an edge-cloud collaborative computing architecture results in high latency in real-time abnormal data processing, hindering rapid enforcement response and further reducing the overall effectiveness of risk prevention and control. Summary of the Invention

[0005] The purpose of this invention is to address the aforementioned technical problems by providing a non-site enforcement method for waterway regulation and structure damage based on multi-source spatiotemporal data.

[0006] In view of this, the present invention provides a non-site enforcement method for damage to waterway regulation structures based on multi-source spatiotemporal data. The method collects multi-source spatiotemporal data to construct a three-dimensional monitoring data source. The multi-source spatiotemporal data includes ship AIS data, ship report data, historical ship report data, waterway regulation structure dataset, real-time tide data, ship stability characteristic data, ship radar scan data, video surveillance and intelligent recognition data, and real-time meteorological data. For three types of collision risk scenarios involving waterway improvement structures—unintentional collisions, hydrological impacts, and illegal navigation—corresponding off-site enforcement algorithms are implemented. The algorithm for identifying first-time arriving crew members, the algorithm for identifying high-risk waters during high and low tides, and the trajectory-based off-site enforcement algorithm are run sequentially to determine collision risk. Based on the risk assessment results, non-on-site law enforcement scripts are broadcast via VHF channel and non-on-site law enforcement information is pushed through drones approaching and shouting. If a collision risk is determined or a collision event occurs, an evidence chain containing AIS trajectory slices, collision moment video frames, and drone panoramic video is automatically generated and pushed to the law enforcement terminal, while the nearest patrol vessel is dispatched. The above steps are supported by a four-layer architecture of perception, data, processing, and application. The hardware layer provides support for data acquisition, the data layer builds a spatiotemporal database cluster to store and index data, the processing layer executes algorithms through edge and cloud collaborative computing, the application layer implements business function modules, and the user layer provides multi-terminal services.

[0007] Preferably, the specific content of the multi-source spatiotemporal data includes: Ship AIS data includes the ship's MMSI code set, ship's latitude and longitude coordinates, speed, and heading angle; Ship report data includes the ship's MMSI code, a set of crew identification numbers, and a set of declaration timestamps; Historical vessel report data: includes a set of crew member ID numbers and a set of historical declaration timestamps; Data set of waterway regulation structures: includes a set of structure numbers, the latitude and longitude coordinates of the structures, and the elevation of the top of the structures; Real-time tide data: Measurement accuracy is ±1cm, updated and stored in real time; Ship stability characteristic data: used to assess a ship's maneuverability under complex hydrological and meteorological conditions; Ship radar scan data: used to compensate for monitoring blind spots in ship AIS data; Video surveillance and intelligent identification data: One set of shore-based cameras is deployed every 5 kilometers to obtain waterway images, and intelligent algorithms are used to identify whether ships have entered the electronic fence and whether they have illegally slowed down or turned. Real-time meteorological data: including meteorological parameters such as wind force, visibility, and rainfall that affect ship navigation.

[0008] Preferably, the specific execution process of the first-arrival crew member identification algorithm is as follows: The geographical boundary of the port is defined by longitude intervals. and latitude range The time threshold is defined as 180 days. The port entry determination function determines whether the ship's latitude and longitude coordinates fall within the set port geographical boundary. If the determination result is 1, the subsequent steps are executed. Using the identity matching operator, the historical ship report database is retrieved using the ID numbers of the current ship's crew members; If the search results are empty, or the difference between the most recent historical declaration time and the current time exceeds 180 days, off-site enforcement will be triggered, and port waterway characteristic information and waterway improvement structure distribution information will be pushed to the crew of the vessel.

[0009] Preferably, the specific execution process of the high-risk water area identification algorithm during high and low tides is as follows: Set a safety threshold of 6 meters and a buffer zone radius of 50 meters, and obtain real-time tide level data with a measurement accuracy of ±1 cm and the elevation of the top of the channel regulation structure; By comparing the real-time tide level with the altitude of the top of the building using the risk building identification logic, if the difference between the two is less than 6 meters, the building is determined to be in a high-risk state. A 50-meter radius area is designated as a danger zone, centered on buildings in a high-risk condition. The system monitors whether a vessel's latitude and longitude coordinates fall into a dangerous area by entering the zone. If the vessel is confirmed to have entered, the system sends off-site enforcement information to the vessel and reminds the crew to adjust their navigation status.

[0010] Preferably, the specific execution process of the trajectory-based off-site law enforcement algorithm is as follows: Set the deceleration distance parameters to 30 minutes and 10 minutes, and set the geographical error tolerance to 5 minutes; Based on the ship's real-time speed, calculate the distance the ship may travel in 30 minutes and 10 minutes respectively; Based on the ship's current heading angle, the predicted endpoint coordinates of the ship after traveling the corresponding distance under the above two deceleration times are calculated and integrated to form a prediction point set; The spatial intersection determination method is used to check whether there are prediction points in the prediction point set that are close to the waterway regulation structure; If present, a tiered off-site enforcement mechanism is triggered. Level 1 off-site enforcement is triggered when the predicted point corresponding to the 10-minute deceleration distance parameter approaches the building, and Level 2 off-site enforcement is triggered when the predicted point corresponding to the 30-minute deceleration distance parameter approaches the building. At the same time, the linkage mechanism between video surveillance and drone patrol is activated.

[0011] Preferably, the specific method for pushing law enforcement information and generating evidence chains is as follows: Customized off-site law enforcement scripts are broadcast for different risk scenarios. For the first arrival of a vessel, the scripts provide information on tidal range, no-navigation zones around structures, and recommended speeds. For high-risk waters, the scripts provide information on the height of the top of structures above the water and the risk of the vessel's draft. The drone was dispatched to fly within 100 meters of the ship to broadcast additional law enforcement information and film the ship's speed and course adjustments, with the footage being simultaneously transmitted back to the law enforcement terminal. When a collision risk or collision event occurs, the system automatically integrates AIS trajectory slices, shore-based camera video frames, drone panoramic video, and VHF voice recordings to form a complete chain of evidence, which is then pushed to the law enforcement terminal. While the chain of evidence was being pushed out, the nearest patrol vessel was dispatched to handle the situation.

[0012] Preferably, the specific content of the four-layer architecture and the user layer is as follows: This includes shore-based cameras (one set every 5 kilometers), AIS base stations with a coverage radius of ≥20 kilometers, tide level sensors with a measurement accuracy of ±1 cm, and drone base stations with a control radius of ≥10 kilometers, forming a three-dimensional monitoring network. It includes a ship dynamic database with a storage period of 1 year and a query response of ≤100ms, a remediation building attribute database with a quarterly update frequency, a ship report database storing current crew data, a historical ship report database storing historical crew information, and a tide database with a query response of ≤200ms. It uses PostGIS spatial extension to implement latitude and longitude indexing. An edge-cloud collaborative computing architecture is adopted, with edge nodes processing real-time abnormal data with a latency of ≤50ms, and cloud nodes used for ship trajectory mining and algorithm model training with a computing power of ≥10PFlops. It includes a real-time monitoring and management module, a collaborative law enforcement management module, and a decision analysis and handling management module, enabling spatiotemporal dynamic data fusion, intelligent analysis and decision-making, and multi-departmental data sharing; We developed desktop terminals, mobile terminals, and a public WeChat mini-program to provide waterway management personnel with risk statistics and law enforcement resource deployment functions, waterway law enforcement personnel with early warning reception and navigation dispatch functions, and the public with waterway safety information query functions.

[0013] Preferred options also include off-site law enforcement business processes, specifically: Upon entering the jurisdiction, vessels are automatically verified and classified, and waterway information is pushed to vessels arriving at the port for the first time. When a vessel enters a Level 3 electronic fence, it broadcasts non-on-site law enforcement messages via VHF. When it enters a Level 2 electronic fence, it activates a high-definition camera to track the vessel in real time and uses a drone to approach and shout messages within 100 meters. When it enters a Level 1 electronic fence or a collision occurs, it automatically generates a chain of evidence and pushes it to the law enforcement terminal and dispatches patrol vessels. The system enables blockchain-based evidence storage and payment of fines for vessels involved in incidents, online retrieval of law enforcement data by third parties, and random public disclosure of cases.

[0014] Preferably, multi-source spatiotemporal data are applied using dynamic linkage and deep fusion methods: When the AIS data of a vessel arriving at port for the first time shows that it is approaching a channel improvement structure with a low safety margin, the vessel's speed is confirmed by the vessel's radar scan data, and the impact of wind on the vessel's handling is judged by combining real-time meteorological data. This forms a closed loop of linkage between data perception, risk assessment, multi-level intervention, and evidence preservation, covering the blind spots and lag problems of traditional monitoring.

[0015] Preferably, the transmission of law enforcement information is optimized for special weather scenarios: when visibility is less than 1.5 kilometers, the broadcast frequency of the VHF channel is increased and the visibility warning message is supplemented, and drones turn on their strobe lights to guide the direction of ships; When rainfall reaches the level of a rainstorm, the changes in water flow velocity are predicted by combining real-time tide data, and ships are notified via VHF to adjust their speed to cope with the impact of the water flow on their course.

[0016] The beneficial effects of this invention are: By deeply integrating multi-source spatiotemporal data and constructing a three-dimensional monitoring network, the problems of single data sources and numerous monitoring blind spots in traditional waterway supervision have been effectively solved. The collaborative application of ship AIS data, radar scan data, video surveillance data, and real-time tide and meteorological data not only fills the monitoring gaps caused by AIS signal interference or shutdown, but also accurately locates risk sources by combining waterway regulation structure datasets. The implementation of algorithms for identifying first-time arriving crew members and high-risk waters during high and low tides can identify collision hazards caused by crew members' unfamiliarity with the waters and hydrological changes in advance, improving the timeliness of risk prediction for unintentional collisions and hydrological influences to 10-30 minutes before collisions, significantly reducing the incidence of illegal navigation and accidents.

[0017] Relying on a hierarchical algorithm system and an automated law enforcement linkage mechanism, the transformation of waterway law enforcement from passive response to proactive prevention has been achieved. The trajectory-based off-site law enforcement algorithm, through 30-minute and 10-minute deceleration distance parameter settings and hierarchical responses, can accurately match enforcement intensity according to the predicted vessel trajectory, avoiding the drawbacks of traditional one-size-fits-all enforcement. The linkage of customized VHF scripts, drone-based close-range broadcasts within 100 meters, and automatic dispatch of patrol vessels improves the efficiency of law enforcement information transmission by more than 50%. Furthermore, when collision risks or incidents occur, it can automatically integrate AIS trajectory slices, shore-based video frames, and drone panoramic video to form a complete evidence chain, solving the problems of slow and easily overlooked evidence collection in traditional law enforcement, and providing reliable data support for law enforcement tracing.

[0018] By leveraging a four-layer architecture encompassing perception, data, processing, and application, along with a multi-terminal service system, the system ensures efficient operation throughout the entire process of risk prevention and control and regulatory services. The perception layer's three-dimensional monitoring network and the data layer's spatiotemporal database ensure real-time acquisition and rapid retrieval of multi-source data; edge-cloud collaborative computing supports efficient algorithm execution, enabling real-time anomaly data processing and trajectory mining; and the application layer's functional modules, combined with desktop terminals, mobile devices, and public mini-programs, not only meet the professional needs of waterway management and law enforcement personnel but also provide safety information services to the public. This constructs a closed-loop regulatory model encompassing monitoring, analysis, response, and service, comprehensively improving the intelligence and precision of waterway management and construction damage prevention and control. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the non-site enforcement method for waterway regulation and structure damage based on multi-source spatiotemporal data according to the present invention. Detailed Implementation

[0020] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1 This is a flowchart illustrating the non-site enforcement method for waterway regulation and structure damage based on multi-source spatiotemporal data, as per the present invention. The following is a detailed introduction to this non-site enforcement method for waterway regulation and structure damage based on multi-source spatiotemporal data.

[0021] Step S110: Collect ship AIS data, ship report data, historical ship report data, waterway regulation structure dataset, real-time tide data, ship stability characteristic data, ship radar scan data, video surveillance and intelligent identification data, and real-time meteorological data to construct a three-dimensional monitoring data source; In this embodiment, in the Yangtze River Estuary's river-sea confluence channel scenario, multi-source data, through dynamic linkage and deep fusion, constructs a three-dimensional monitoring network covering all elements of ships, crew, structures, hydrology, and meteorology. Specific applications can span the entire process of daily channel supervision, risk analysis, and emergency response. In the complex environment of the Yangtze River Estuary with an average of 200 ships passing through daily, ship AIS data transmits each ship's MMSI code, latitude and longitude, speed, and heading angle in real time. The system tracks the dynamic trajectory of ships based on this data. If a cargo ship's AIS shows a heading angle continuously deviating from the planned channel and approaching the Xinliu River sandbar, preliminary location data can be immediately triggered for off-site enforcement. Ship report data and historical ship report data focus on the crew's familiarity with the waterway. When a vessel enters the jurisdiction, the system extracts the crew's ID number from the vessel's report data and compares it with historical report data. If the system finds that the crew has no record of declaring for this waterway, or that the most recent declaration was made more than 180 days ago, it can determine that the crew is either arriving at port for the first time or has not been to port for a long time. The system then pushes key information such as the distribution of dredging structures and the location of shoals in the waterway to prevent unintentional collisions caused by crew unfamiliarity with the environment. For example, it can send a targeted message to such crew members that there are three underwater dredging structures in the Xinliuhe Sand Area, and that a speed of ≤8 knots should be maintained during high tide. The channel dredging structure dataset provides a basis for risk positioning. The system can quickly retrieve the corresponding latitude, longitude, and elevation of the top of the structure by its number, and calculate the safety margin by combining it with real-time tide data. Vessel stability characteristic data is used to assess the vessel's maneuverability under complex hydrological and meteorological conditions. When real-time meteorological data shows winds reaching level 7 and turbulent currents, the system, combined with the stability data of a passenger ship, will, if it determines that the ship's stability is insufficient, warn the ship in advance to avoid areas with dense dredging structures and suggest adjusting its course to a sheltered channel to avoid collisions with structures due to loss of control. Ship radar scan data can compensate for blind spots in AIS data. In foggy winter conditions at the Yangtze River estuary, some ships' AIS signals may be interfered with or deliberately turned off. Radar scanning can capture the position and movement trends of these invisible ships. If a short-haul transport ship is found to have failed to report to AIS and shows signs of crossing the channel towards a shunting structure, the system can continuously track its trajectory via radar and link video surveillance to confirm ship details. Video surveillance and intelligent recognition data enable visual verification and evidence preservation. Shore-based cameras, arranged every 5 kilometers, can capture real-time channel images. Intelligent algorithms can automatically identify whether ships have entered the electronic fence and whether they have illegally slowed down or changed direction. When a ship enters the first-level electronic fence of a shunting structure, the camera will zoom in to track the distance between the ship's waterline and the structure, while capturing video frames as evidence. In the event of a collision, the system can also record the moment of collision, forming a complete chain of evidence together with AIS trajectory slices and drone panoramic video.Real-time meteorological data is also used in conjunction with other data to optimize off-site enforcement strategies. For example, when visibility is less than 1.5 kilometers, the system will increase the AIS data refresh frequency, expand the radar monitoring range, and simultaneously broadcast encrypted off-site enforcement messages via VHF channels to remind ships that visibility is currently low and that the area within 50 meters of the remediation buildings in the Xinliuhe Sand Area is a danger zone, advising them not to approach. When rainfall reaches the level of a rainstorm, the system will also combine tide data to predict changes in water flow speed and inform ships in advance to adjust their speed to cope with the impact of water flow on their course, thus comprehensively reducing the risk of collision. These data are not used in isolation but form a closed-loop linkage: if the AIS of a ship arriving at the port for the first time shows that it is approaching a remediation building with a low safety margin, radar and video confirm that the ship's speed is too high, and meteorological data shows that the wind is increasing, the system will integrate this information, first pushing water area information, then broadcasting off-site enforcement messages via VHF, and finally activating drones to approach and issue warnings, while simultaneously dispatching the nearest patrol vessels to stand by. This achieves three-dimensional supervision from data perception, risk assessment, multi-layered intervention, and evidence preservation, completely covering the blind spots and lag problems of traditional monitoring.

[0022] Step S111: The multi-source spatiotemporal data mentioned in Step 1 specifically includes ship AIS data: where ship AIS data includes the ship MMSI code set S, ship latitude and longitude coordinates, and speed. and heading angle Ship report data: includes the ship's MMSI code, a set of crew ID numbers P, and a set of declaration timestamps T; Historical ship report data: includes a set of crew ID numbers P and a set of historical declaration timestamps T; Channel improvement structure dataset: includes a set of structure numbers B, the latitude and longitude coordinates of the structures, and the elevation of the top of the structures. Real-time tide data: measurement accuracy ±1cm, updated and stored in real time.

[0023] Step S120: For the three types of waterway management structure collision risk scenarios—unintentional collision, hydrological impact, and illegal navigation—the corresponding off-site enforcement algorithms are executed respectively; In this embodiment, within the Yangtze River Estuary waterway supervision scenario, to address the risk of unintentional collisions caused by crew members' unfamiliarity with the waters, non-site enforcement is implemented using a first-time arrival crew member identification algorithm. First, the geographical boundaries of the port are clearly defined, and a 180-day time threshold is set. The system first determines whether a vessel has entered the port area using AIS data. Once a cargo ship is confirmed to have entered, the system extracts the crew members' ID numbers from the vessel report data and then searches the historical vessel report database. If the search reveals that a crew member has no historical declaration record for that waterway, or that the most recent declaration was made more than 180 days ago, the non-site enforcement process is triggered. The system immediately sends the crew member characteristic information of the Yangtze River Estuary, such as water flow characteristics and shoal distribution, and details the specific locations of waterway regulation structures, such as the specific area of ​​the Xinliu River sandbar, to prevent crew members from making operational errors due to unfamiliarity with the water environment and causing collisions.

[0024] To address the collision risks caused by the hydrological impacts of high and low tides, a high-risk water area identification algorithm will be used for off-site enforcement. The system will pre-set a safety threshold of 6 meters and a buffer zone radius of 50 meters. Real-time tide level data will be obtained through tide level sensors with an accuracy of ±1 cm, and then calculated in conjunction with the elevation of the top of the channel regulation structures. When the difference between the real-time tide level and the elevation of the top of the structure is less than 6 meters, the structure will be determined to be in a high-risk state, and a danger zone with a radius of 50 meters will be demarcated centered on the structure. If the latitude and longitude coordinates of a bulk carrier are detected by ship AIS data falling within this danger zone, the system will immediately push off-site enforcement information to the ship, and simultaneously broadcast a warning message on the VHF channel to remind the crew to pay attention to the status of underwater structures, slow down, and activate nearby shore-based cameras to track the ship's movement in real time to ensure that the ship stays away from the danger zone.

[0025] To address the collision risks associated with vessels illegally crossing waterways to save time, a trajectory-based off-site enforcement algorithm will be used for monitoring. The system will set two deceleration distance parameters: 30 minutes and 10 minutes. First, based on the vessel's real-time speed from its AIS data, the system will calculate the distance the vessel may travel within these two time periods. Then, combined with the vessel's heading angle, the system will predict the vessel's future trajectory and destination coordinates, forming a set of prediction points. Then, by determining whether these predicted points are close to the channel regulation structures, if a point close to a structure is found in the predicted trajectory of a short-haul transport ship, a tiered non-site enforcement mechanism will be triggered: if the point corresponds to a trajectory with a 10-minute deceleration distance, a level two enforcement mechanism will be activated, immediately turning on high-definition cameras to track the ship in real time, and simultaneously dispatching drones to approach within 100 meters to issue a warning, requiring the ship to adjust its course; if the point corresponds to a 30-minute deceleration distance and is closer to the structure, a level one enforcement mechanism will be triggered, automatically initiating the linkage between video surveillance and drone patrols, and simultaneously generating a preliminary record containing ship AIS trajectory slices and real-time video frames. Once the ship continues to approach, a complete chain of evidence will be immediately generated and pushed to the enforcement terminal, and the nearest patrol vessel will be dispatched to handle the situation to avoid a collision.

[0026] S130: The algorithm for identifying first-time arriving crew members, the algorithm for identifying high-risk waters during high and low tides, and the trajectory-based off-site enforcement algorithm are run sequentially to determine the risk. In this embodiment, during routine monitoring of the Yangtze River Estuary waterway, the system sequentially executes three types of algorithms to gradually complete risk assessment, forming a closed loop for end-to-end risk prevention and control. First, the system activates the initial arrival crew identification algorithm: when an unfamiliar cargo ship fully loaded with goods enters the Yangtze River Estuary port jurisdiction, the system first confirms that its latitude and longitude fall within the preset port geographical boundaries using the ship's AIS data. Then, it extracts the ID numbers of all crew members from the ship's report data and compares them with the historical ship report database. If it finds that multiple crew members have never had a declaration record for the Yangtze River Estuary waterway, or that the most recent declaration was more than 180 days ago, the system immediately determines that the ship poses a risk of unintentional collision due to crew members' unfamiliarity with the water environment. The system then sends the crew a map of the Yangtze River Estuary waterway improvement structures and navigation precautions to proactively eliminate potential risks caused by crew members' unfamiliarity with the environment.

[0027] After completing the initial risk assessment of arriving crew members, the system simultaneously runs a high-risk water area identification algorithm during high and low tides. This involves using tide level sensors deployed along the channel with a measurement accuracy of ±1cm to acquire real-time tide level data of the Yangtze River estuary, and then calculating the data by combining this data with the elevation of the top of each structure in the channel regulation structure dataset. When the difference between the real-time tide level and the elevation of the top of a certain regulation structure in a section of the channel is less than 6 meters, the system determines that the waters surrounding that structure have entered a high-risk state due to hydrological changes, and then delineates a danger zone with a radius of 50 meters centered on that structure. If the AIS track of the aforementioned cargo ship shows that it is heading towards this danger zone, the system will trigger a risk assessment again, confirming that the ship faces a collision risk due to hydrological influences. It will immediately broadcast a warning message to the ship via VHF, reminding the crew that a regulation structure ahead has decreased in height due to high and low tides, indicating it has entered a high-risk zone, and requesting them to slow down and detour. Simultaneously, it will activate shore-based cameras located every 5 kilometers to track the ship's movements in real time, ensuring it stays away from the danger zone.

[0028] Finally, the trajectory-based off-site enforcement algorithm is run to complete the final risk assessment: The system first obtains the real-time speed and heading angle of the cargo ship, calculates the distance the ship may travel in these two time periods according to the deceleration time parameters of 30 minutes and 10 minutes, and then combines the heading angle to deduce the future trajectory and destination coordinates of the ship, forming two sets of prediction points. By comparing the positional relationship between the predicted point set and the channel improvement structures, if it is found that a vessel is traveling at its current course and speed and may approach a certain improvement structure in 10 minutes, the system determines that the vessel has a collision risk due to illegal navigation or operational deviation. It immediately triggers a graded response, activates high-definition cameras to continuously track the vessel's trajectory, and dispatches drones from drone base stations with a control radius of ≥10 kilometers to approach within 100 meters of the vessel to issue a warning, requesting a change of course. If the prediction shows that the vessel will directly enter the danger zone around the structure in 30 minutes, the system upgrades the risk assessment result, simultaneously activates the linkage recording of video surveillance and drone patrols, automatically saves the vessel's trajectory slices, and immediately generates a preliminary evidence chain containing AIS trajectory and drone panoramic video if the vessel fails to adjust according to the warning. This chain is then pushed to the law enforcement terminal and the nearest patrol vessel is dispatched to stand by, completely preventing the collision risk.

[0029] Step S131: The algorithm for identifying first-arriving crew members specifically includes first defining the geographical boundary of the port, which is defined by specific latitude and longitude intervals, i.e., longitude within... to Between and latitude to The system first defines the area between the designated port and the surrounding area; then, it sets a 180-day time threshold; next, it uses an entry determination function to determine whether a vessel has entered the defined port area. When the function returns a result of 1 for a particular vessel, it indicates that the vessel's latitude and longitude coordinates fall within the defined port geographical boundary, and only then does it proceed to the next step. Then, using an identity matching operator, it searches the historical vessel report database using the current vessel's crew member's ID number. If the search result is empty, meaning no record of the crew member is found in the historical report data; or if the crew member's historical record is found, but the difference between the crew member's most recent historical declaration time and the current time exceeds the previously set 180-day time threshold, either of these two conditions will trigger off-site enforcement. Finally, it pushes the characteristic information of the port waters and the distribution information of waterway improvement structures to the crew members on the vessel that triggered off-site enforcement.

[0030] Step S132: The algorithm for identifying high-risk waters during high and low tides specifically includes: First, setting relevant parameters, including a 6-meter safety threshold and a 50-meter buffer radius, while acquiring real-time tide level data with a measurement accuracy of ±1cm. Next, the risk status of each channel regulation structure is determined through a risk structure identification logic. Specifically, this involves comparing the real-time tide level with the elevation of the top of the structure. When the difference is less than the set 6-meter safety threshold, the structure is determined to be in a high-risk state, and subsequent steps are executed. Then, a 50-meter radius area is designated as a danger zone centered on this high-risk structure. Afterwards, the vessel's position is monitored through an entry-into-danger zone determination logic, i.e., determining whether the vessel's latitude and longitude coordinates fall within the designated danger zone. When it is confirmed that a vessel has entered the danger zone, off-site enforcement information is pushed to the vessel, reminding the crew to be aware of the risks and adjust their navigation status.

[0031] Step S133: The trajectory-based off-site law enforcement algorithm specifically includes: First, setting two deceleration distance parameters, 30 minutes and 10 minutes respectively, and setting a 5-minute geographical error tolerance. Next, based on the ship's real-time speed, calculating the distance the ship may travel within the time corresponding to these two deceleration distance parameters. Then, combining the ship's current heading angle, calculating the predicted endpoint coordinates of the ship after traveling the corresponding distance under the two deceleration times, and integrating these predicted endpoint coordinates to form a predicted point set.

[0032] Then, by determining the spatial intersection, the system checks whether any predicted points in the predicted point set are close to channel regulation structures. If such predicted points exist, tiered off-site enforcement is triggered. Tier 1 off-site enforcement corresponds to the presence of predicted points calculated from the 10-minute deceleration distance parameter, and these predicted points are close to structures; Tier 2 off-site enforcement corresponds to the presence of predicted points calculated from the 30-minute deceleration distance parameter, and these predicted points are close to structures. Simultaneously with triggering tiered off-site enforcement, a linkage mechanism between video surveillance and drone patrols is activated to track vessel movements in real time and retain relevant information.

[0033] Step S140: Based on the algorithm's judgment result, non-on-site law enforcement information is pushed through the VHF channel by broadcasting non-on-site law enforcement scripts and by drones approaching and shouting. If a collision risk or collision event occurs, an evidence chain containing AIS trajectory slices, collision moment video frames, and drone panoramic video is automatically generated and pushed to the law enforcement terminal to dispatch the nearest patrol vessel.

[0034] In this embodiment, in a navigation scenario around the Xinliuhe Sandbar Embankment in the Yangtze River Estuary, when the system, through its first-time arrival crew identification algorithm, detects that a cargo ship from another region has no Yangtze River Estuary declaration record for the past 180 days and determines that there is a risk of unintentional collision, it will first broadcast a non-on-site enforcement message via VHF channel: "Attention cargo ship MMSI code XXXX, your crew is arriving in the Yangtze River Estuary for the first time. The maximum tidal range in this area is 4 meters. The Xinliuhe Sandbar Embankment is currently in a semi-submerged state, and the area within 50 meters of the embankment is a no-navigation zone. It is recommended that your ship reduce its speed to below 5 knots and sail along the north side of the channel marker. If you have any questions, you can call back via VHF channel 16." If no response is received from the ship within 5 minutes, the system will dispatch the nearest drone base station to send a drone to fly within 100 meters of the cargo ship to make an approaching call, supplementing the VHF message with: "Cargo ship MMSI code XXXX, the drone has arrived at your ship's starboard side. We are now sending you an electronic distribution map of the waterway management structures. Please confirm receipt and do not deviate from the recommended route to avoid collision with the sandbar embankment." Simultaneously, the drone activates its high-definition camera to record the ship's speed and course adjustments in real time, transmitting the footage back to the law enforcement terminal. If the ship begins to decelerate and follows the recommended route, the drone maintains a 500-meter distance, flying alongside it to a safe area before returning. During this time, it continuously broadcasts via VHF that the current tide level is in the low tide phase, and requests that the ship pay attention to subsequent tide level changes. When the system, through its high-risk water area identification algorithm during high and low tides, detects that a section of the Yangtze River estuary has a real-time tide level of 6.5 meters and the top of a structure is 1.8 meters above sea level (a difference of 4.7 meters less than the 6-meter safety threshold), and demarcates a 50-meter danger zone, if a bulk carrier is detected approaching this area, it will immediately broadcast via VHF: "MMSI code YYYY bulk carrier, please note that your ship is approaching the danger zone of the B005 channel improvement structure. The current tide level is 6.5 meters, the top of the structure is only 4.7 meters above the water, and your ship's draft is 5.5 meters. There is a risk of grounding and collision. Please immediately turn left 30 degrees and leave the current area, maintaining a speed of less than 3 knots." Simultaneously, a drone is dispatched from a base station 10 kilometers away, arriving near the bulk carrier within 12 minutes. Upon approaching within 100 meters, it broadcasts: "MMSI code YYYY bulk carrier, the drone has confirmed your waterline. Continuing forward will result in contact with structure B005. Turn immediately! After turning, please report via VHF channel 16." The drone simultaneously captures images of the vessel's waterline, course, and surrounding waters, integrating the video frames with AIS trajectory data in real time. If the vessel turns and leaves within 3 minutes, the system notifies via VHF that it has left the danger zone; please pay attention to subsequent alerts from the navigation channel's electronic fence. If the vessel does not turn, the drone continues to approach and broadcast, automatically generating a preliminary evidence chain including VHF broadcast recordings, drone video, and AIS trajectory data, which is then pushed to the law enforcement terminal, dispatching the nearest patrol vessel to handle the situation.When the system, using a trajectory-based off-site enforcement algorithm, determines that a short-haul transport vessel is traveling at an excessive speed of 13 knots and is predicted to approach the Xinliu River sandbar in 10 minutes, it first broadcasts an urgent message via VHF channel: "Attention, MMSI code ZZZZ transport vessel! Your current heading is 315 degrees, speed is 13 knots, and you are predicted to approach the Xinliu River sandbar in 10 minutes. This behavior constitutes an illegal crossing of the waterway and poses a collision risk. Please immediately reduce your speed to 5 knots, turn 45 degrees to the right, and correct your course to the main waterway." Immediately afterward, shore-based cameras along the route are activated to track the vessel's trajectory. Simultaneously, a drone takes off from the nearest base station and approaches within 100 meters of the transport vessel within 14 minutes, issuing a supplementary announcement: "MMSI code ZZZZ transport vessel, the drone has recorded your vessel's illegal course. You are now required to immediately reduce your speed and turn, otherwise, a Level 1 off-site enforcement action will be triggered, and patrol vessels will arrive at the scene within 18 minutes." If the vessel begins to decelerate and turn, the drone will record the adjustment process and then return to port. Subsequent VHF broadcasts will confirm that your vessel has corrected its course; please maintain your speed and comply with navigation rules. If the vessel continues to maintain its original course and speed, the system will be upgraded to Level 1 non-site enforcement. The VHF message will change to: "MMSI code ZZZZ transport vessel, your vessel is refusing to cooperate with enforcement and has entered a Level 1 electronic fence. The patrol vessel has departed, and the drone will continue to record your vessel's behavior. You will be responsible for any collision." The drone will begin panoramic recording, integrating the video with AIS trajectory slices and VHF broadcast recordings into a complete chain of evidence, which will be pushed to the enforcement terminal to ensure sufficient evidence for subsequent case investigations. In foggy weather, if the system determines that a vessel is at risk using any algorithm, it will increase the frequency of broadcasts on the VHF channel, adding a message that visibility is currently low and urging passengers to observe nearby vessels and building signs. The drone will activate its strobe lights to guide your vessel. When the drone approaches and broadcasts messages, it will simultaneously activate its high-power loudspeaker to ensure that the crew can clearly receive the message. At the same time, it will use flashing lights to convey simple instructions such as slowing down and turning, avoiding information transmission failure due to poor visibility and further reducing the risk of collision.

[0035] Step S141: Off-site enforcement and enforcement coordination also include off-site enforcement business processes, specifically: Upon entering the jurisdiction, vessels are automatically verified and classified. For vessels arriving for the first time, water information is pushed to the system. When a vessel enters a Level 3 electronic fence, non-on-site enforcement messages are broadcast via VHF. When a vessel enters a Level 2 electronic fence, high-definition cameras are activated for real-time tracking, and drones are used to approach and broadcast messages within 100 meters. When a vessel enters a Level 1 electronic fence or a collision occurs, an evidence chain is automatically generated and pushed to the enforcement terminal, dispatching patrol vessels. This enables the use of blockchain for evidence storage and fine payment for vessels involved, online data retrieval by third parties, and random public disclosure of cases.

[0036] Step S150: Adopt a four-layer architecture of perception, data, processing, and application. The hardware infrastructure layer provides data acquisition support, the data layer builds a spatiotemporal database cluster to store and index data, the processing layer executes algorithms through edge-cloud collaborative computing, the application layer implements business function modules, and the user layer provides multi-terminal services.

[0037] In this embodiment, in the daily navigation monitoring scenario around the Xinliuhe Sandbar Embankment in the Yangtze River Estuary, the four-layer architecture and the user layer work together to form a closed loop for full-process risk prevention and enforcement. The hardware facilities of the perception layer first construct a three-dimensional monitoring network: one set of shore-based cameras is deployed every 5 kilometers to capture the navigation attitude of ships in the channel in real time; AIS base stations with a coverage radius of ≥20 kilometers continuously receive MMSI codes, latitude and longitude, speed and heading data of passing ships; tide level sensors with a measurement accuracy of ±1cm are fixed in the waters near the embankment and update the real-time tide level every 10 seconds; UAV base stations with a control radius of ≥10 kilometers are set up around the Baoshan Renewable Resources Utilization Center, ready to perform patrol and broadcasting tasks at any time. When a cargo ship from another region enters the waters, the AIS base station first captures its dynamic data, the shore-based cameras simultaneously capture details of the ship's hull, and the tide level sensor records the current tide level as 7.2 meters. This data is transmitted to the data layer in real time. The spatiotemporal database cluster in the data layer then stores and indexes multi-source data: the ship dynamics database archives the cargo ship's AIS data according to a 1-year storage cycle and a ≤100ms query response standard, facilitating subsequent trajectory tracing; the remediation building attribute database retrieves the attribute information of the Xinliuhe sandbank protection dike—number B012, latitude and longitude range, and top elevation of 1.5 meters; the ship report database extracts the set of crew ID numbers declared by the cargo ship, while the historical ship report database prepares historical crew declaration records; the tide database stores the current tide level data of 7.2 meters with a response speed of ≤200ms, and indexes the ship's latitude and longitude and building's latitude and longitude through PostGIS spatial extension, ensuring rapid retrieval by subsequent algorithms. When the system needs to determine whether a crew member is arriving at the port for the first time, the data layer can complete the comparison between the current crew member's ID number and the historical report database within 100ms, providing data support for risk assessment. The processing layer executes algorithms through edge-cloud collaborative computing: edge nodes prioritize processing real-time abnormal data. When AIS data shows that a cargo ship's course deviates from the main channel and tends to approach the breakwater, the edge node completes preliminary calculations of the trajectory-based off-site enforcement algorithm within 50ms, determining that the ship may enter the 50-meter danger zone of the breakwater within the next 10 minutes, and immediately triggers an early warning signal. Simultaneously, cloud nodes utilize computing power of ≥10 PFlops to batch analyze ship trajectory data in this waterway over the past 3 months, optimizing the safety threshold of the high-risk water area identification algorithm during high and low tides. Combining the current tide level of 7.2 meters with the breakwater elevation of 1.5 meters, it confirms that the 6-meter safety threshold is still applicable, and synchronizes the optimization results to the edge nodes. In addition, the cloud also mines historical data of crew members arriving at port for the first time, statistically analyzing common navigation deviation areas of ships arriving at the Yangtze River Estuary for the first time in the past six months, providing a basis for subsequent enforcement priorities.The application layer's business function modules translate algorithm results into actual law enforcement actions: The real-time monitoring and management module integrates AIS data, camera footage, and tide data to mark the cargo ship's location, breakwater range, and danger zone on an electronic map, using flashing red to warn of risks; The coordinated law enforcement management module triggers a tiered response—first, it broadcasts a non-site law enforcement message via VHF channel: "Attention, cargo ship MMSI code XXXX, your ship is approaching the danger zone of breakwater B012. The current tide level is 7.2 meters, and the top of the breakwater is 5.7 meters above the water surface. Please immediately adjust your course to 330 degrees and reduce speed to below 5 knots. At the same time, the drone base station will dispatch a drone to approach within 100 meters of the cargo ship within 12 minutes, supplementing the message with: 'The drone has confirmed your ship's position. Do not deviate from the recommended route, otherwise, the dispatch of patrol vessels will be initiated.'" If the cargo ship still does not adjust, the decision analysis and handling management module automatically generates an evidence chain containing AIS trajectory slices, camera video frames, and drone panoramic video, and pushes it to the law enforcement terminal. The user layer caters to the needs of different roles through multi-terminal services: waterway law enforcement personnel use a mobile app to receive early warning information, view evidence chains, remotely dispatch the nearest patrol vessels, and track the progress of handling in real time; waterway management personnel log in to the system via desktop terminals, view risk warning statistics for the waterway over the past week in the decision analysis module, analyze high-risk periods, and adjust the deployment of law enforcement resources; the public accesses the Yangtze River Estuary Waterway Safety Service section via a WeChat mini-program to check the real-time location, current tide level, and recommended routes of the Xinliuhe Sandbar, avoiding unauthorized entry into dangerous areas. Throughout the process, the four-layer architecture seamlessly integrates with the user layer, achieving both real-time risk prevention and control and providing support for long-term law enforcement optimization and public services.

[0038] Step S151: The hardware infrastructure layer specifically includes: shore-based cameras in groups of 5 kilometers, AIS base stations with a coverage radius of ≥20 kilometers, tide level sensors with a measurement accuracy of ±1 cm, and UAV base stations with a control radius of ≥10 kilometers, forming a three-dimensional monitoring network.

[0039] Step S152: The data layer specifically includes a ship dynamic database with a storage period of 1 year and a query response of ≤100ms, a building attribute database with a quarterly update frequency, a ship report database storing current crew data, a historical ship report database storing historical crew information, and a tide database with a query response of ≤200ms. Latitude and longitude indexing is implemented using PostGIS spatial extension.

[0040] Step S153: The processing layer adopts an edge-cloud collaborative computing architecture. Edge nodes handle real-time anomalies with a latency of ≤50ms, while cloud nodes are used for trajectory mining and model training with a computing power of ≥10PFlops.

[0041] Step S154: The application layer includes real-time monitoring and management, collaborative law enforcement management, and decision analysis and disposal management modules to achieve spatiotemporal dynamic data fusion, intelligent analysis and decision-making, and multi-department data sharing; the user layer develops desktop and mobile terminals to provide services to waterway law enforcement personnel, other waterway management personnel, and the public, respectively.

[0042] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A non-site enforcement method for waterway regulation and structure damage based on multi-source spatiotemporal data, characterized by: Multi-source spatiotemporal data is collected to construct a three-dimensional monitoring data source. The multi-source spatiotemporal data includes ship AIS data, ship report data, historical ship report data, waterway regulation structure dataset, real-time tide data, ship stability characteristic data, ship radar scan data, video surveillance and intelligent recognition data, and real-time meteorological data. For three types of collision risk scenarios involving waterway improvement structures—unintentional collisions, hydrological impacts, and illegal navigation—corresponding off-site enforcement algorithms are implemented. The algorithm for identifying first-time arriving crew members, the algorithm for identifying high-risk waters during high and low tides, and the trajectory-based off-site enforcement algorithm are run sequentially to determine collision risk. Based on the risk assessment results, non-on-site law enforcement scripts are broadcast via VHF channel and non-on-site law enforcement information is pushed out by drones approaching and broadcasting. If a collision risk is determined or a collision event occurs, an evidence chain containing AIS trajectory slices, collision moment video frames, and drone panoramic video is automatically generated and pushed to the law enforcement terminal, while the nearest patrol vessel is dispatched. The above steps are supported by a four-layer architecture of perception, data, processing, and application. The hardware layer provides support for data acquisition, the data layer builds a spatiotemporal database cluster to store and index data, the processing layer executes algorithms through edge and cloud collaborative computing, the application layer implements business function modules, and the user layer provides multi-terminal services.

2. The non-site enforcement method for waterway regulation and structure damage based on multi-source spatiotemporal data according to claim 1, characterized in that: The specific content of the multi-source spatiotemporal data includes: Ship AIS data includes the ship's MMSI code set, ship's latitude and longitude coordinates, speed, and heading angle; Ship report data includes the ship's MMSI code, a set of crew identification numbers, and a set of declaration timestamps; Historical vessel report data: includes a set of crew member ID numbers and a set of historical declaration timestamps; Data set of waterway regulation structures: includes a set of structure numbers, the latitude and longitude coordinates of the structures, and the elevation of the top of the structures; Real-time tide data: Measurement accuracy is ±1cm, updated and stored in real time; Ship stability characteristic data: used to assess a ship's maneuverability under complex hydrological and meteorological conditions; Ship radar scan data: used to compensate for monitoring blind spots in ship AIS data; Video surveillance and intelligent identification data: One set of shore-based cameras is deployed every 5 kilometers to obtain waterway images, and intelligent algorithms are used to identify whether ships have entered the electronic fence and whether they have illegally slowed down or turned. Real-time meteorological data: including meteorological parameters such as wind force, visibility, and rainfall that affect ship navigation.

3. The non-site enforcement method for waterway regulation and structure damage based on multi-source spatiotemporal data according to claim 2, characterized in that: The specific execution process of the algorithm for identifying first-time arriving crew members is as follows: The geographical boundary of the port is defined by longitude intervals. and latitude range The time threshold is defined as 180 days. The port entry determination function determines whether the ship's latitude and longitude coordinates fall within the set port geographical boundary. If the determination result is 1, the subsequent steps are executed. Using the identity matching operator, the historical ship report database is retrieved using the ID numbers of the current ship's crew members; If the search results are empty, or the difference between the most recent historical declaration time and the current time exceeds 180 days, off-site enforcement will be triggered, and port waterway characteristic information and waterway improvement structure distribution information will be pushed to the crew of the vessel.

4. The non-site enforcement method for waterway regulation and structure damage based on multi-source spatiotemporal data according to claim 3, characterized in that: The specific execution process of the algorithm for identifying high-risk water areas during high and low tides is as follows: Set a safety threshold of 6 meters and a buffer zone radius of 50 meters, and obtain real-time tide level data with a measurement accuracy of ±1 cm and the elevation of the top of the channel regulation structure; By comparing the real-time tide level with the altitude of the top of the building using the risk building identification logic, if the difference between the two is less than 6 meters, the building is determined to be in a high-risk state. A 50-meter radius area is designated as a danger zone, centered on buildings in a high-risk condition. The system monitors whether a vessel's latitude and longitude coordinates fall into a dangerous area by entering the zone. If the vessel is confirmed to have entered, the system sends off-site enforcement information to the vessel and reminds the crew to adjust their navigation status.

5. The non-site enforcement method for waterway regulation and structure damage based on multi-source spatiotemporal data according to claim 4, characterized in that: The specific execution process of the trajectory-based off-site law enforcement algorithm is as follows: Set the deceleration distance parameters to 30 minutes and 10 minutes, and set the geographical error tolerance to 5 minutes; Based on the ship's real-time speed, calculate the distance the ship may travel in 30 minutes and 10 minutes respectively; Based on the ship's current heading angle, the predicted endpoint coordinates of the ship after traveling the corresponding distance under the above two deceleration times are calculated and integrated to form a prediction point set; The spatial intersection determination method is used to check whether there are prediction points in the prediction point set that are close to the waterway regulation structure; If present, a tiered off-site enforcement mechanism is triggered. Level 1 off-site enforcement is triggered when the predicted point corresponding to the 10-minute deceleration distance parameter approaches the building, and Level 2 off-site enforcement is triggered when the predicted point corresponding to the 30-minute deceleration distance parameter approaches the building. At the same time, the linkage mechanism between video surveillance and drone patrol is activated.

6. The non-site enforcement method for waterway regulation structure damage based on multi-source spatiotemporal data according to claim 5, characterized in that: The specific methods for pushing law enforcement information and generating evidence chains are as follows: Customized off-site law enforcement scripts are broadcast for different risk scenarios. For the first arrival of a vessel, the scripts provide information on tidal range, no-navigation zones around structures, and recommended speeds. For high-risk waters, the scripts provide information on the height of the top of structures above the water and the risk of the vessel's draft. The drone was dispatched to fly within 100 meters of the ship to broadcast additional law enforcement information and film the ship's speed and course adjustments, with the footage being simultaneously transmitted back to the law enforcement terminal. When a collision risk or collision event occurs, the system automatically integrates AIS trajectory slices, shore-based camera video frames, drone panoramic video, and VHF voice recordings to form a complete chain of evidence, which is then pushed to the law enforcement terminal. While the chain of evidence was being pushed out, the nearest patrol vessel was dispatched to handle the situation.

7. The non-site enforcement method for waterway regulation structure damage based on multi-source spatiotemporal data according to claim 6, characterized in that: The specific content of the four-layer architecture and the user layer is as follows: This includes shore-based cameras (one set every 5 kilometers), AIS base stations with a coverage radius of ≥20 kilometers, tide level sensors with a measurement accuracy of ±1 cm, and drone base stations with a control radius of ≥10 kilometers, forming a three-dimensional monitoring network. It includes a ship dynamic database with a storage period of 1 year and a query response of ≤100ms, a remediation building attribute database with a quarterly update frequency, a ship report database storing current crew data, a historical ship report database storing historical crew information, and a tide database with a query response of ≤200ms. It uses PostGIS spatial extension to implement latitude and longitude indexing. It adopts an edge and cloud collaborative computing architecture. Edge nodes process real-time abnormal data with a latency of ≤50ms, while cloud nodes are used for ship trajectory mining and algorithm model training with a computing power of ≥10PFlops. It includes a real-time monitoring and management module, a collaborative law enforcement management module, and a decision analysis and handling management module, enabling spatiotemporal dynamic data fusion, intelligent analysis and decision-making, and multi-departmental data sharing; We developed desktop terminals, mobile terminals, and a public WeChat mini-program to provide waterway management personnel with risk statistics and law enforcement resource deployment functions, waterway law enforcement personnel with early warning reception and navigation dispatch functions, and the public with waterway safety information query functions.

8. The non-site enforcement method for waterway regulation structure damage based on multi-source spatiotemporal data according to claim 7, characterized in that: It also includes off-site law enforcement business processes, specifically: Upon entering the jurisdiction, vessels are automatically verified and classified, and waterway information is pushed to vessels arriving at the port for the first time. When a vessel enters a Level 3 electronic fence, it broadcasts non-on-site law enforcement messages via VHF. When it enters a Level 2 electronic fence, it activates a high-definition camera to track the vessel in real time and uses a drone to approach and shout messages within 100 meters. When it enters a Level 1 electronic fence or a collision occurs, it automatically generates a chain of evidence and pushes it to the law enforcement terminal and dispatches patrol vessels. The system enables blockchain-based evidence storage and payment of fines for vessels involved in incidents, online retrieval of law enforcement data by third parties, and random public disclosure of cases.

9. The non-site enforcement method for waterway regulation structure damage based on multi-source spatiotemporal data according to claim 8, characterized in that: Multi-source spatiotemporal data are applied using dynamic linkage and deep fusion methods: When the AIS data of a vessel arriving at port for the first time shows that it is approaching a channel improvement structure with a low safety margin, the vessel's speed is confirmed by the vessel's radar scan data, and the impact of wind on the vessel's handling is judged by combining real-time meteorological data. This forms a closed loop of linkage between data perception, risk assessment, multi-level intervention, and evidence preservation, covering the blind spots and lag problems of traditional monitoring.

10. The non-site enforcement method for waterway regulation structure damage based on multi-source spatiotemporal data according to claim 9, characterized in that: Optimize law enforcement information transmission for special weather scenarios: when visibility is less than 1.5 kilometers, increase the broadcast frequency of the VHF channel and supplement the visibility warning message; drones turn on their strobe lights to guide ships. When rainfall reaches the level of a rainstorm, the changes in water flow velocity are predicted by combining real-time tide data, and ships are notified via VHF to adjust their speed to cope with the impact of the water flow on their course.

Images