An intelligent positioning, navigation and warning system and method for boats

Abstract

The present invention relates to the technical field of navigation and positioning, and particularly to a marine intelligent positioning navigation and warning system and method. The method includes the following steps: collecting and evaluating the sea conditions and positioning signals according to a multi-source sensor synchronization data set to obtain a sensor reliability score table; analyzing the target motion situation according to the sensor reliability score table to obtain a fused situation vector map; judging the target motion state according to the fused situation vector map to obtain a target motion state determination result; evaluating the multi-target collision risk according to the target motion state determination result to obtain a risk degree evaluation matrix; performing navigation operation performance constraints according to the risk degree evaluation matrix to obtain a route planning operation constraint rule set; and searching for a route trajectory according to the route planning operation constraint rule set to obtain a navigation trajectory plan. The present invention significantly improves the positioning accuracy, navigation safety and warning system practicability of ships in complex environments through positioning navigation and warning technology.

Description

Technical Field

[0001] The present invention relates to the technical field of navigation and positioning, and particularly to a boat intelligent positioning navigation and warning system and method. Background Art

[0002] Existing navigation systems mostly rely on a single signal source. In the face of severe sea conditions or signal interference, the positioning accuracy significantly decreases. Some systems using multi-source fusion often adopt static or empirical weights and cannot flexibly adjust according to real-time environment and signal changes, resulting in distorted fusion results and reducing the reliability and accuracy of positioning.

[0003] Current warning methods mainly conduct trajectory prediction and risk assessment for a single target, ignoring the dynamic interaction and mutual influence between multiple navigation targets, making it difficult to identify group behaviors or complex interaction risks, and prone to "missing warnings" or misjudgments of collisions. At the same time, the means for identifying the motion state of targets are relatively simple and cannot effectively cope with complex dynamic environments.

[0004] Traditional route optimization mostly aims at the shortest path or simple obstacle avoidance, lacking comprehensive consideration of the actual maneuverability of ships (such as turning radius, acceleration and deceleration capabilities) and sea area environment (such as sea conditions, flow fields, restricted areas, etc.), resulting in the generated ship tracks being difficult to execute in actual operations and having potential safety hazards.

[0005] In summary, there are problems in the prior art such as poor self-adaptability of multi-source data fusion, insufficient recognition of target dynamic interaction risks, and disconnection between route planning and actual ship motion constraints. There is an urgent need for a boat intelligent navigation and warning method that can improve positioning accuracy, enhance the ability to identify collision risks, and take into account the feasibility of actual operations. Summary of the Invention

[0006] Based on this, it is necessary to provide a boat intelligent positioning navigation and warning system and method to solve at least one of the above technical problems.

[0007] To achieve the above object, a boat intelligent positioning navigation and warning method includes the following steps:

[0008] Step S1: Collect and perform sea condition perception and positioning signal evaluation based on a multi-source sensor synchronous data set to obtain a sensor reliability scoring table; perform target motion situation analysis based on the sensor reliability scoring table to obtain a fused situation vector diagram;

[0009] Step S2: Judge the target motion state according to the fused situation vector diagram to obtain a target motion state determination result; perform multi-target collision risk assessment according to the target motion state determination result to obtain a risk degree assessment matrix;

[0010] Step S3: Perform navigation maneuverability constraints according to the risk assessment matrix to obtain a set of route planning maneuvering constraint rules; perform route trajectory search according to the set of route planning maneuvering constraint rules to obtain a navigation trajectory plan;

[0011] Step S4: Perform real-time monitoring of navigation deviation according to the navigation trajectory plan to obtain a navigation deviation status assessment report; calculate the early warning distance according to the navigation deviation status assessment report to obtain a hierarchical early warning distance table; generate an intelligent early warning decision table according to the hierarchical early warning distance table and the risk assessment matrix early warning and avoidance strategy.

[0012] Through multi-source sensor fusion, the present invention realizes the comprehensive perception and evaluation of sea conditions and positioning signals. The generation of the sensor reliability scoring table can dynamically reflect the performance of each sensor under different sea conditions, providing a reliable weight basis for subsequent data fusion and enhancing the stability and accuracy of positioning. The analysis of the target motion trend, generating a fused trend vector map, can display the target motion state in real time and comprehensively, providing important basic information for subsequent collision risk assessment and navigation decision-making, and improving the accuracy and real-time performance of situation awareness. By analyzing the fused trend vector map, accurately judge the motion state of the target, providing a reliable basis for collision risk assessment. Multi-target collision risk assessment, generating a risk assessment matrix, can quantify the collision risks between different targets, providing an important risk basis for subsequent navigation decision-making and improving the accuracy and comprehensiveness of risk assessment. Through navigation maneuverability constraints, the feasibility and safety of route planning are ensured. The generation of the set of route planning maneuvering constraint rules takes into account factors such as the maneuverability of the ship and sea conditions, ensuring that the route planning conforms to the actual maneuvering ability of the ship and avoiding navigation risks. Route trajectory search, generating a navigation trajectory plan, can optimize navigation on the premise of ensuring safety, improving the efficiency and economy of navigation. By real-time monitoring of navigation deviation, potential navigation risks can be detected in a timely manner. The navigation deviation status assessment report can timely reflect the degree of deviation of the boat from the predetermined route, providing an important basis for subsequent early warning and avoidance. Early warning distance calculation, generating a hierarchical early warning distance table, can provide differentiated early warning information according to different risk levels, improving the timeliness and effectiveness of early warning. Generating an intelligent early warning decision table according to the hierarchical early warning distance table and the risk assessment matrix realizes the intelligent early warning of potential collision risks and the generation of avoidance strategies, improving the safety of boat navigation. Therefore, the present invention provides a boat intelligent positioning, navigation and early warning method, which effectively solves the existing technical problems and significantly improves the positioning accuracy, navigation safety and practicality of the early warning system of ships in complex environments by introducing dynamic weighted fusion of multi-source heterogeneous data based on real-time sea conditions and signal quality, multi-target trajectory correlation prediction and joint collision risk assessment, and route optimization and adaptive early warning threshold adjustment combined with measured maneuverability and environmental constraints.

[0013] Preferably, the present invention further provides a boat intelligent positioning, navigation and early warning system for implementing the boat intelligent positioning, navigation and early warning method as described above. The boat intelligent positioning, navigation and early warning system includes:

[0014] A signal fusion module, configured to collect and evaluate sea conditions and positioning signals according to a multi-source sensor synchronization data set to obtain a sensor reliability score table; analyze the target motion trend according to the sensor reliability score table to obtain a fused trend vector map;

[0015] A risk assessment module, configured to judge the target motion state according to the fused trend vector map to obtain a target motion state determination result; evaluate the multi-target collision risk according to the target motion state determination result to obtain a danger degree assessment matrix;

[0016] A route optimization module, configured to perform navigation operation performance constraints according to the danger degree assessment matrix to obtain a route planning operation constraint rule set; search for a route trajectory according to the route planning operation constraint rule set to obtain a navigation trajectory plan;

[0017] An early warning decision-making module, configured to monitor the navigation deviation in real time according to the navigation trajectory plan to obtain a navigation deviation state assessment report; calculate the early warning distance according to the navigation deviation state assessment report to obtain a hierarchical early warning distance table; generate an early warning and risk avoidance strategy according to the hierarchical early warning distance table and the danger degree assessment matrix to obtain an intelligent early warning decision-making table.

[0018] Through the collaborative work of the four major modules of signal fusion, risk assessment, route optimization and early warning decision-making, this aspect of the system realizes the comprehensive perception of multi-source information, the accurate judgment of the target motion state, the quantitative assessment of navigation risks, and the safe and efficient planning of navigation trajectories, and can provide intelligent early warning and risk avoidance strategies according to the risk level, thereby comprehensively improving the safety, efficiency and intelligence level of boat navigation, reducing the risk of human operation, and ensuring navigation safety. Brief Description of the Drawings

[0019] Figure 1 It is a schematic diagram of the step flow of a boat intelligent positioning, navigation and early warning method.

[0020] The realization of the object, functional features and advantages of the present invention will be further described with reference to the embodiments and the accompanying drawings. Detailed Embodiments

[0021] The technical method of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0022] In addition, the accompanying drawings are only schematic diagrams of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings represent the same or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0023] It should be understood that although terms such as "first" and "second" may be used here to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit can be called the second unit, and similarly the second unit can be called the first unit. The term "and / or" used here includes any and all combinations of one or more of the listed related items.

[0024] In the embodiments of the present invention, refer to Figure 1 As shown, it is a schematic diagram of the step flow of the intelligent positioning, navigation and warning method for boats and ships of the present invention. In this example, the intelligent positioning, navigation and warning method for boats and ships includes the following steps:

[0025] Step S1: Collect and evaluate the sea condition perception and positioning signals according to the multi-source sensor synchronous data set to obtain a sensor reliability score table; analyze the target motion situation according to the sensor reliability score table to obtain a fused situation vector map;

[0026] In the embodiment of the present invention, a multi-source sensor synchronous data set is collected through a navigation radar, an AIS receiver, a Beidou / GPS satellite receiver, a wave height sensor, and a hull attitude sensor. Then, signal evaluation and sea condition correction are performed on the positioning sensor data, and the influence of the hull attitude is compensated. Finally, a sensor reliability score table is obtained. Next, according to the reliability score table, the adaptive Kalman filter parameters are adjusted for the multi-source sensor synchronous data set, and multi-source data fusion and trajectory smoothing are performed. Finally, a fused situation vector map is generated, which includes information such as target ID, type, relative azimuth, relative distance, relative speed, relative heading, and historical trajectory points. For example, if the radar detects a target at a distance of 1000 meters from the ship and at an azimuth of 0°, and the AIS receiver receives the AIS information of the target and confirms it as a cargo ship, and the Beidou / GPS receiver measures the speed of the target to be 10 knots and the heading to be 90°, then the fused situation vector map will display a cargo ship target at a distance of 1000 meters, an azimuth of 0°, a speed of 10 knots, and a heading of 90°.

[0027] Step S2: Determine the target motion state according to the fused situation vector map to obtain the target motion state determination result; perform a multi-target collision risk assessment according to the target motion state determination result to obtain a risk degree assessment matrix;

[0028] In the embodiment of the present invention, the historical trajectory of the target in the fused situation vector map is extracted and segmented to obtain the target trajectory feature sequence. Then, the heading change feature set and the speed change feature set are extracted, the heading motion mode and the speed motion mode are respectively judged, and special motion modes are identified. Combining with the target historical state sequence, the motion intention is inferred, and finally the target motion state determination result is generated. For example, if the historical trajectory of the target shows that it first sails in a straight line at a constant speed, then makes a slow turn, and finally resumes straight-line sailing at a constant speed, it can be inferred that its motion intention is to adjust the heading. Next, according to the target motion state determination result, the relative motion parameters between the targets are calculated, and a target correlation network graph is constructed. Then, multi-target joint trajectory prediction is performed, and a collision risk assessment is performed on the multi-target predicted trajectory set, and the collision risk coefficient is calculated. Finally, a risk degree assessment matrix is generated. For example, if it is predicted that the DCPA between the ship and the target ship is less than the safe distance and the TCPA is less than the safe time, it is considered that there is a collision risk and it is marked in the risk degree assessment matrix.

[0029] Step S3: Perform navigation maneuverability constraints according to the risk degree assessment matrix to obtain a route planning maneuver constraint rule set; perform route trajectory search according to the route planning maneuver constraint rule set to obtain a navigation trajectory plan;

[0030] In an embodiment of the present invention, the current sea area environment information is obtained, including information such as water depth, seabed topography, obstacles, meteorology, tides, and navigation restrictions, to generate a sea area environment constraint map. Then, a navigation risk cost map is constructed in combination with a risk assessment matrix, where the cost of each grid represents the navigation risk of that grid. Next, according to the ship maneuverability constraints, including steering ability, acceleration and deceleration ability, safe speed limit, and economic speed range, etc., a route planning maneuver constraint rule set is generated. Then, direct search and screening of the route are performed according to the constraint rule set and the navigation risk cost map to obtain a set of route candidate solutions. Finally, a navigation instruction sequence is generated for the set of route candidate solutions to obtain the final navigation trajectory solution, which includes information such as a sequence of waypoints, estimated arrival time, and recommended speed. For example, a route is planned, including three waypoints: A(120.1°, 30.1°), B(120.2°, 30.2°), C(120.3°, 30.3°), with estimated arrival times of 10:00, 10:10, and 10:20 respectively, and recommended speeds of 10 knots, 12 knots, and 10 knots respectively.

[0031] Step S4: Real-time monitoring of the navigation deviation is performed according to the navigation trajectory solution to obtain a navigation deviation status assessment report; the warning distance is calculated according to the navigation deviation status assessment report to obtain a hierarchical warning distance table; an intelligent warning decision table is generated according to the hierarchical warning distance table and the risk assessment matrix warning and avoidance strategy.

[0032] In an embodiment of the present invention, real-time navigation data of the boat is obtained and compared with the navigation trajectory solution to generate a navigation deviation status assessment report, which includes information such as position deviation, course deviation, and speed deviation. Then, real-time environment data is obtained, and the safety impact of environmental factors is quantified according to the navigation deviation status assessment report to obtain an environmental safety impact coefficient table. Next, the boat performance parameters are obtained, and the real-time maneuverability of the boat is evaluated according to the environmental safety impact coefficient table to obtain real-time maneuverability evaluation indicators. Then, the warning distance is calculated according to the real-time maneuverability evaluation indicators and the navigation deviation status assessment report to obtain a hierarchical warning distance table. Finally, an early warning and avoidance strategy is generated according to the hierarchical warning distance table and the risk assessment matrix to obtain an intelligent warning decision table, which includes information such as warning level, warning information, and recommended avoidance measures. For example, if the distance between the own ship and the target ship is less than the first-level warning distance, a first-level warning of "collision risk" is issued, and it is recommended to take a collision avoidance measure of "immediately turn".

[0033] Preferably, the sea condition perception and positioning signal evaluation in step S1 includes:

[0034] Collect a multi-source sensor synchronization data set through a multi-source sensor array, where the multi-source sensor synchronization data set includes radar echo data, AIS received data stream, satellite receiver data, wave height sensor data, and hull attitude sensor data;

[0035] Conduct positioning sensor signal evaluation on the radar echo data, AIS received data stream, and satellite receiver data to obtain the initial score of the positioning sensor;

[0036] Measure the sea condition complexity of the wave height sensor data and the hull attitude sensor data to obtain the sea condition level evaluation result;

[0037] Obtain the basic attenuation coefficient affected by the sea condition according to the sea condition level evaluation result;

[0038] Perform sea condition influence correction on the radar echo data, AIS received data stream, and satellite receiver data according to the basic attenuation coefficient affected by the sea condition to obtain the corrected value of the positioning sensor;

[0039] Scan the plane with the radar and calculate the offset angle caused by the hull swing, denoted as the radar scan offset angle, where the radar scan offset angle includes the root mean square of the roll angle and the root mean square of the pitch angle;

[0040] Compensate the corrected value of the positioning sensor for the influence of the hull attitude according to the radar scan offset angle to obtain the attitude compensation value of the positioning sensor;

[0041] Conduct a comprehensive reliability evaluation based on the attitude compensation value of the positioning sensor and the initial score of the positioning sensor to obtain the sensor reliability score table.

[0042] In the embodiments of the present invention, a multi-source sensor array collects a multi-source sensor synchronous data set. The multi-source sensor array includes: a navigation radar, an AIS receiver, a Beidou / GPS satellite receiver, a wave height sensor, and a hull attitude sensor. The navigation radar operates in the X-band with a scanning frequency of 24 times per minute. Its echo data includes target distance, azimuth, and echo intensity. The AIS receiver receives AIS messages compliant with the ITU-R M.1371 standard. The AIS received data stream includes vessel MMSI number, position, course, and speed information. The Beidou / GPS satellite receiver adopts the dual-frequency RTK mode to provide high-precision position, course, and speed data, and its data update frequency is 10 Hz. The wave height sensor uses the ultrasonic measurement principle to measure the distance from the sea surface to the sensor and calculates the wave height in combination with the hull draft. Its data update frequency is 5 Hz. The hull attitude sensor uses a combination of MEMS gyroscopes and accelerometers to measure the hull roll angle, pitch angle, and heading angle, and its data update frequency is 20 Hz. Through the data acquisition module, the data of all the above sensors are synchronized according to a unified timestamp to generate a multi-source sensor synchronous data set.

[0043] The evaluation of the positioning sensor signals includes: for the echo data of the navigation radar, calculating the signal-to-noise ratio of each target and dividing the signal quality level according to the signal-to-noise ratio. When the signal-to-noise ratio is greater than 20 dB, the score is 0.9; when the signal-to-noise ratio is between 15 dB and 20 dB, the score is 0.7; when the signal-to-noise ratio is between 10 dB and 15 dB, the score is 0.5; when the signal-to-noise ratio is less than 10 dB, the score is 0.3. For the AIS received data stream, calculating the reception integrity rate of the AIS messages. When the integrity rate is greater than 95%, the score is 0.9; when the integrity rate is between 85% and 95%, the score is 0.7; when the integrity rate is between 70% and 85%, the score is 0.5; when the integrity rate is less than 70%, the score is 0.3, and if there is missing or delayed AIS data, the score is reduced. For the data of the Beidou / GPS satellite receiver, counting the number of effective satellites and the PDOP value. When the number of effective satellites is greater than 8 and the PDOP is less than 2, the score is 0.9; when the number of effective satellites is between 6 and 8 and the PDOP is between 2 and 3, the score is 0.7; when the number of effective satellites is less than 6 or the PDOP is greater than 3, the score is 0.3. Combining the evaluation results of the radar, AIS, and Beidou / GPS, the initial score of the positioning sensor is obtained.

[0044] The measurement of sea state complexity includes: data from wave height sensors, calculating the root mean square value of the wave height. When the root mean square value is less than 0.5 meters, it is rated as calm sea state; when the root mean square value is between 0.5 meters and 1.5 meters, it is rated as slightly rough sea state; when the root mean square value is between 1.5 meters and 2.5 meters, it is rated as moderately rough sea state; when the root mean square value is between 2.5 meters and 4 meters, it is rated as relatively rough sea state; when the root mean square value is greater than 4 meters, it is rated as severe sea state; data from hull attitude sensors, calculating the root mean square values of the roll angle and pitch angle of the hull. When both the root mean square of the roll angle and the root mean square of the pitch angle are less than 3 degrees, it is rated as calm sea state; when both the root mean square of the roll angle and the root mean square of the pitch angle are between 3 degrees and 6 degrees, it is rated as slightly rough sea state; when both the root mean square of the roll angle and the root mean square of the pitch angle are between 6 degrees and 9 degrees, it is rated as moderately rough sea state; when both the root mean square of the roll angle and the root mean square of the pitch angle are between 9 degrees and 12 degrees, it is rated as relatively rough sea state; when both the root mean square of the roll angle and the root mean square of the pitch angle are greater than 12 degrees, it is rated as severe sea state. Based on the comprehensive evaluation results of wave height and hull attitude, the most severe sea state level is selected as the final sea state level evaluation result.

[0045] The sea state influence basic attenuation coefficient is set according to the sea state level evaluation result. For calm sea state, the basic attenuation coefficient is 1.0; for slightly rough sea state, the basic attenuation coefficient is 0.8; for moderately rough sea state, the basic attenuation coefficient is 0.6; for relatively rough sea state, the basic attenuation coefficient is 0.4; for severe sea state, the basic attenuation coefficient is 0.2.

[0046] The sea state influence correction includes: multiplying the signal-to-noise ratio of the radar echo data by the sea state influence basic attenuation coefficient to obtain the corrected signal-to-noise ratio; multiplying the reception integrity rate of the AIS received data stream by the sea state influence basic attenuation coefficient to obtain the corrected reception integrity rate; multiplying the positioning accuracy score of the Beidou / GPS satellite receiver by the sea state influence basic attenuation coefficient to obtain the corrected positioning accuracy score.

[0047] The specific method for calculating the radar scanning offset angle is as follows: First, obtain the roll angle and pitch angle data output by the hull attitude sensor, and calculate the root mean square of the roll angle and the root mean square of the pitch angle. The formula for the root mean square of the roll angle is: √(Σ(roll angle i^2) / N), where roll angle i is the roll angle data and N is the number of roll angle data; the formula for the root mean square of the pitch angle is: √(Σ(pitch angle i^2) / N), where pitch angle i is the pitch angle data and N is the number of pitch angle data.

[0048] The hull attitude influence compensation includes: For radar echo data, according to the root mean square of the roll angle and the root mean square of the pitch angle, correct the target azimuth and distance. The corrected azimuth angle = original azimuth angle - root mean square of the roll angle × correction coefficient, and the corrected distance = original distance + root mean square of the pitch angle × correction coefficient; for the AIS received data stream, no attitude compensation is performed; for the Beidou / GPS satellite receiver data, no attitude compensation is performed.

[0049] The comprehensive evaluation of sensor reliability includes: calculating the final radar score = signal-to-noise ratio after attitude compensation × initial radar score; calculating the final AIS score = corrected reception integrity rate × initial AIS score; calculating the final Beidou / GPS score = corrected positioning accuracy score × initial Beidou / GPS score. Record the final radar score, final AIS score, and final Beidou / GPS score in the sensor reliability score table and use it as the basis for subsequent data fusion.

[0050] Preferably, the analysis of the target motion situation in step S1 includes:

[0051] Adjust the parameters of the adaptive Kalman filter for the multi-source sensor synchronous data set according to the sensor reliability score table to obtain an adaptive filter parameter set;

[0052] Perform multi-source data fusion and trajectory smoothing based on the adaptive filter parameter set and the multi-source sensor synchronous data set to obtain target state fusion data;

[0053] Generate and update the situation vector diagram based on the target state fusion data to obtain the fused situation vector diagram.

[0054] In the embodiments of the present invention, the adaptive Kalman filter parameter adjustment includes: setting different process noise covariance matrices Q and measurement noise covariance matrices R according to the final radar score, the final AIS score, and the final Beidou / GPS score in the sensor reliability score table. For radar data, if the final radar score is greater than 0.8, set Q as [0.01, 0; 0, 0.01] and R as [1, 0; 0, 1]; if the score is between 0.5 and 0.8, set Q as [0.05, 0; 0, 0.05] and R as [4, 0; 0, 4]; if the score is less than 0.5, set Q as [0.1, 0; 0, 0.1] and R as [9, 0; 0, 9]. The units of Q and R are position (meters) × position (meters) and velocity (meters / second) × velocity (meters / second), respectively. For AIS data, if the final AIS score is greater than 0.8, set Q as [0.01, 0; 0, 0.01] and R as [0.5, 0; 0, 0.5]; if the score is between 0.5 and 0.8, set Q as [0.05, 0; 0, 0.05] and R as [2, 0; 0, 2]; if the score is less than 0.5, set Q as [0.1, 0; 0, 0.1] and R as [4.5, 0; 0, 4.5]. For Beidou / GPS data, if the final Beidou / GPS score is greater than 0.8, set Q as [0.01, 0; 0, 0.01] and R as [0.25, 0; 0, 0.25]; if the score is between 0.5 and 0.8, set Q as [0.05, 0; 0, 0.05] and R as [1, 0; 0, 1]; if the score is less than 0.5, set Q as [0.1, 0; 0, 0.1] and R as [2.25, 0; 0, 2.25]. According to the above settings, an adaptive filter parameter set is obtained, which includes the Q and R matrices for radar, AIS, and Beidou / GPS data.

[0055] Multi-source data fusion and trajectory smoothing include: filtering radar data, AIS data, and Beidou / GPS data using the adaptive Kalman filtering algorithm. The state vector of the Kalman filter is defined as [x, y, vx, vy], where x and y are the target position coordinates, and vx and vy are the target velocities. The state transition matrix F is [[1, 0, Δt, 0], [0, 1, 0, Δt], [0, 0, 1, 0], [0, 0, 0, 1]], Δt is the sampling time interval, the control matrix B is [[Δt^2 / 2, 0], [0, Δt^2 / 2], [Δt, 0], [0, Δt]], and the observation matrix H is [[1, 0, 0, 0], [0, 1, 0, 0]]. For the data of each sensor, the corresponding Q and R matrices are used for Kalman filtering. For example, for radar data, the corresponding Q and R of the radar are used for filtering to obtain the filtered target position and velocity. The same filtering process is also performed on AIS data and Beidou / GPS data. Then, a weighted fusion method is adopted to fuse the filtered radar data, AIS data, and Beidou / GPS data. The weights of the weighted fusion are determined according to the final scores in the sensor reliability scoring table. For example, for radar data, the weight is the radar final score / (radar final score + AIS final score + Beidou / GPS final score). The same weight calculation is also performed on AIS data and Beidou / GPS data. The calculation formulas for the fused target position and velocity are: fused position = weight of radar × filtered radar position + weight of AIS × filtered AIS position + weight of Beidou / GPS × filtered Beidou / GPS position, and fused velocity = weight of radar × filtered radar velocity + weight of AIS × filtered AIS velocity + weight of Beidou / GPS × filtered Beidou / GPS velocity. Finally, the sliding average filtering algorithm is used to smooth the fused target trajectory. The length of the sliding average filtering window is 5 seconds to obtain the fused target state data.

[0056] Situation vector map generation and update include: establishing a two-dimensional coordinate system centered on the own ship, converting the target positions in the fused target state data into this coordinate system, and calculating the azimuth, distance, relative velocity, and relative course of the target relative to the own ship. The azimuth angle calculation formula is: atan2(y target - y own ship, x target - x own ship);

[0057] The distance calculation formula is: ;

[0058] The relative velocity calculation formula is: ;

[0059] The relative heading calculation formula is: atan2(vy_target - vy_own_ship, vx_target - vx_own_ship). Classify and identify the targets, distinguishing static targets, dynamic targets, and special targets. If the relative speed of the target is less than 0.5 m / s, it is marked as a static target; if the relative speed of the target is greater than 0.5 m / s, it is marked as a dynamic target; if the target AIS information contains special identifiers (such as distress, out of control, etc.), it is marked as a special target. Set the data update period to 1 second and adjust the update frequency according to the target importance and dynamic characteristics. Construct the data structure of the situation vector diagram, including information such as target ID, target type, relative bearing, relative distance, relative speed, relative heading, and historical trajectory points (the last 5 points). Each time an update occurs, update the latest target state fusion data to the situation vector diagram, and update the historical trajectory points according to the changes in the target state. Finally, obtain the fused situation vector diagram.

[0060] Of particular importance is that the adaptive Kalman filter parameter adjustment is specifically as follows:

[0061] Dynamically allocate sensor weights according to the sensor reliability scoring table to obtain the sensor dynamic weight table; map the sea condition filtering parameters for the multi-source sensor synchronous data set according to the sensor dynamic weight table to obtain the sea condition filtering parameter mapping table; set the processing parameters for radar, AIS, and GPS signals according to the sea condition filtering parameter mapping table to obtain the positioning sensor processing parameter table; formulate data anomaly detection rules according to the positioning sensor processing parameter table; evaluate the consistency and adjust the weights of the multi-source sensor synchronous data according to the data anomaly detection rules to obtain the adaptive filtering parameter set;

[0062] In the embodiments of the present invention, the specific method for dynamically allocating sensor weights is as follows: First, obtain the final radar score, the final AIS score, and the final Beidou / GPS score in the sensor reliability score table. Then, calculate the weights of each sensor based on these scores. Define the weight of the radar as W_radar, the weight of the AIS as W_AIS, and the weight of the Beidou / GPS as W_GPS. The calculation formulas are as follows: W_radar = radar final score / (radar final score + AIS final score + Beidou / GPS final score), W_AIS = AIS final score / (radar final score + AIS final score + Beidou / GPS final score), W_GPS = Beidou / GPS final score / (radar final score + AIS final score + Beidou / GPS final score). Record the calculated W_radar, W_AIS, and W_GPS in the sensor dynamic weight table. For example, if the radar final score is 0.9, the AIS final score is 0.7, and the Beidou / GPS final score is 0.8, then W_radar = 0.9 / (0.9 + 0.7 + 0.8) = 0.375, W_AIS = 0.7 / (0.9 + 0.7 + 0.8) = 0.2917, W_GPS = 0.8 / (0.9 + 0.7 + 0.8) = 0.3333. Map the sea condition filtering parameters for the multi-source sensor synchronization data set according to the sensor dynamic weight table to obtain the sea condition filtering parameter mapping table. The specific method for mapping the sea condition filtering parameters is as follows: First, obtain W_radar, W_AIS, and W_GPS in the sensor dynamic weight table. Then, obtain the sea condition level evaluation result. The sea condition levels are divided into five levels: calm, slight, moderate, large, and severe. Determine the mapping parameters of the process noise covariance matrix Q and the measurement noise covariance matrix R by using a look-up table according to the sea condition level and the sensor weights.For example, if the sea state level is medium and W_radar = 0.375, the value range of Q for the radar is from [0.03, 0; 0, 0.03] to [0.07, 0; 0, 0.07], and the value range of R is from [3, 0; 0, 3] to [5, 0; 0, 5]. The specific values are determined by linear interpolation according to the weight, Q = 0.03 + (0.07 - 0.03) × W_radar / 0.375, R = 3 + (5 - 3) × W_radar / 0.375; if the sea state level is medium and W_AIS = 0.2917, the value range of Q for AIS is from [0.03, 0; 0, 0.03] to [0.07, 0; 0, 0.07], and the value range of R is from [1.5, 0; 0, 1.5] to [2.5, 0; 0, 2.5]. The specific values are determined by linear interpolation according to the weight, Q = 0.03 + (0.07 - 0.03) × W_AIS / 0.2917, R = 1.5 + (2.5 - 1.5) × W_AIS / 0.2917; if the sea state level is medium and W_GPS = 0.3333, the value range of Q for GPS is from [0.03, 0; 0, 0.03] to [0.07, 0; 0, 0.07], and the value range of R is from [0.75, 0; 0, 0.75] to [1.25, 0; 0, 1.25]. The specific values are determined by linear interpolation according to the weight, Q = 0.03 + (0.07 - 0.03) × W_GPS / 0.3333, R = 0.75 + (1.25 - 0.75) × W_GPS / 0.3333. Record the calculated Q and R matrix parameters in the sea state filtering parameter mapping table. Perform processing parameter settings for radar, AIS, and GPS signals according to the sea state filtering parameter mapping table to obtain the positioning sensor processing parameter table. The positioning sensor processing parameter settings include: for radar signals, the parameter settings include the process noise covariance matrix Q_radar and the measurement noise covariance matrix R_radar of the Kalman filter, as well as the radar echo intensity threshold; for AIS signals, the parameter settings include the process noise covariance matrix Q_AIS and the measurement noise covariance matrix R_AIS of the Kalman filter, as well as the AIS data update frequency tolerance range; for GPS signals, the parameter settings include the process noise covariance matrix Q_GPS and the measurement noise covariance matrix R_GPS of the Kalman filter, as well as the number of GPS satellites and the PDOP threshold. The specific setting method is: extract the Q and R matrices corresponding to radar, AIS, and GPS from the sea state filtering parameter mapping table.Set \(Q_{radar}\) as the \(Q\) matrix corresponding to the radar in the sea state filtering parameter mapping table, and \(R_{radar}\) as the \(R\) matrix corresponding to the radar in the sea state filtering parameter mapping table; set \(Q_{AIS}\) as the \(Q\) matrix corresponding to AIS in the sea state filtering parameter mapping table, and \(R_{AIS}\) as the \(R\) matrix corresponding to AIS in the sea state filtering parameter mapping table; set \(Q_{GPS}\) as the \(Q\) matrix corresponding to GPS in the sea state filtering parameter mapping table, and \(R_{GPS}\) as the \(R\) matrix corresponding to GPS in the sea state filtering parameter mapping table. Set the radar echo intensity threshold to -100 dBm, and the echo data below this threshold will be filtered; set the tolerance range of the AIS data update frequency to 1 second, and if no AIS data is received beyond this time interval, the data is considered lost; set the GPS satellite number threshold to 6, and the PDOP threshold to 3, and the GPS data with less than 6 satellites or PDOP greater than 3 will be filtered. Record the above parameters in the positioning sensor processing parameter table. Formulate data anomaly detection rules according to the positioning sensor processing parameter table. The data anomaly detection rules include: for radar data, if the target position jumps by more than 10 meters, or the speed jumps by more than 5 m / s, it is marked as abnormal; for AIS data, if the target position jumps by more than 50 meters, or the heading changes by more than 30 degrees, it is marked as abnormal; for GPS data, if the target position jumps by more than 10 meters, or the speed jumps by more than 5 m / s, it is marked as abnormal. In addition, if the data is marked as abnormal for 5 consecutive times, it is considered that the sensor data has a fault. Conduct consistency evaluation and weight adjustment on the multi-source sensor synchronization data according to the data anomaly detection rules to obtain an adaptive filtering parameter set. The consistency evaluation and weight adjustment include: First, use the data anomaly detection rules to detect anomalies in radar, AIS, and GPS data. Then, adjust the weights according to the anomaly detection results. If the radar data is marked as abnormal, set \(W_{radar}\) to 0.1; if the AIS data is marked as abnormal, set \(W_{AIS}\) to 0.1; if the GPS data is marked as abnormal, set \(W_{GPS}\) to 0.1. If a certain sensor data is marked as abnormal for 5 consecutive times, it is considered that the sensor data has a fault and its weight is set to 0. After adjusting the weights, it is necessary to normalize the weights to ensure that the sum of all weights is 1. For example, if \(W_{radar}=0.1\), \(W_{AIS}=0.1\), \(W_{GPS}=0.8\), then the normalized weights are: \(W_{radar}=0.1 / (0.1 + 0.1 + 0.8)=0.1\), \(W_{AIS}=0.1 / (0.1 + 0.1 + 0.8)=0.1\), \(W_{GPS}=0.8 / (0.1 + 0.1 + 0.8)=0.8\). Finally, recalculate the \(Q\) and \(R\) matrices according to the adjusted weights.For example, if \(W_{radar}\) is adjusted to 0.1 and the sea state level is medium, then \(Q_{radar}\) and \(R_{radar}\) are recalculated as \(Q_{radar}=0.03+(0.07 - 0.03)\times0.1 / 0.375\) and \(R_{radar}=3+(5 - 3)\times0.1 / 0.375\). Record the adjusted \(Q\) and \(R\) matrices and the adjusted weights in the adaptive filtering parameter set.

[0063] Preferably, the target motion state determination in step S2 includes:

[0064] Extract the target historical trajectory of the fused situation vector map; segment the target historical trajectory to obtain the target trajectory feature sequence;

[0065] Extract the heading change feature set and speed change feature set of the target trajectory feature sequence;

[0066] Use the heading change feature set to determine the heading motion mode and obtain the heading mode determination result;

[0067] Use the speed change feature set to judge the speed motion mode and obtain the speed mode determination result;

[0068] Identify the special motion mode according to the heading mode determination result and speed mode determination result to obtain the special mode recognition result;

[0069] Determine the comprehensive motion state according to the heading mode determination result, speed mode determination result and special mode recognition result to obtain the basic motion state determination result;

[0070] Obtain the target historical state sequence and infer the motion intention according to the basic motion state determination result to obtain the motion intention inference result;

[0071] Generate the target motion state determination result according to the basic motion state determination result and motion intention inference result.

[0072] In the embodiments of the present invention, the target historical trajectory is segmented to obtain a target trajectory feature sequence. The extraction of the target historical trajectory from the fused situation vector map includes: extracting the historical position data, heading data, and speed data of each target from the fused situation vector map. The length of the target historical trajectory is the data of the past 5 minutes, and the data update frequency is 1 second. Segmenting the target historical trajectory to obtain a target trajectory feature sequence includes: segmenting the target trajectory according to the historical heading change rate and speed change rate of the target. The specific segmentation rules are as follows: If the heading change rate of three consecutive sampling points is less than 3 degrees / second and the speed change rate is less than 0.1 m / s^2, then this segment of the trajectory is classified as a uniform straight-line navigation segment; if the heading change rate of three consecutive sampling points is greater than 3 degrees / second and the speed change rate is less than 0.1 m / s^2, then this segment of the trajectory is classified as a uniform turning segment; if the speed change rate of three consecutive sampling points is greater than 0.1 m / s^2, then this segment of the trajectory is classified as an acceleration segment; if the speed change rate of three consecutive sampling points is less than -0.1 m / s^2, then this segment of the trajectory is classified as a deceleration segment. For each trajectory segment, calculate its average speed, average heading, heading change rate, and speed change rate, and record the start time and end time of the trajectory segment. Combine the characteristic parameters of each trajectory segment to form a target trajectory feature sequence.

[0073] The extraction of the heading change feature set includes: extracting the heading data within the most recent 10 seconds from the target trajectory feature sequence. Calculate the average change rate of the heading within the most recent 10 seconds. The calculation formula is: (heading_end - heading_start) / time difference, where heading_end is the end heading, heading_start is the start heading, and the time difference is 10 seconds. Calculate the maximum change rate of the heading within the most recent 10 seconds. The calculation formula is: max(|headingi+1 - headingi|) / sampling time interval, where headingi and headingi+1 are two adjacent heading data, and the sampling time interval is 1 second. Take the average heading change rate and the maximum heading change rate as the heading change feature set. The extraction of the speed change feature set includes: extracting the speed data within the most recent 10 seconds from the target trajectory feature sequence. Calculate the average change rate of the speed within the most recent 10 seconds. The calculation formula is: (speed_end - speed_start) / time difference, where speed_end is the end speed, speed_start is the start speed, and the time difference is 10 seconds. Calculate the maximum acceleration within the most recent 10 seconds. The calculation formula is: max(|speedi+1 - speedi|) / sampling time interval, where speedi and speedi+1 are two adjacent speed data, and the sampling time interval is 1 second. Take the average speed change rate and the maximum acceleration as the speed change feature set.

[0074] Course mode determination includes: judging the course motion mode according to the average course change rate and the maximum course change rate in the course change feature set. The specific rules are as follows: If the average course change rate is less than 3 degrees per second and the maximum course change rate is less than 5 degrees per second, the course mode determination result is "straight navigation"; if the average course change rate is greater than or equal to 3 degrees per second and less than 10 degrees per second, or the maximum course change rate is greater than or equal to 5 degrees per second and less than 15 degrees per second, the course mode determination result is "slow turning"; if the average course change rate is greater than or equal to 10 degrees per second, or the maximum course change rate is greater than or equal to 15 degrees per second, the course mode determination result is "rapid turning".

[0075] Speed mode determination includes: judging the speed motion mode according to the average speed change rate and the maximum acceleration in the speed change feature set. The specific rules are as follows: If the absolute value of the average speed change rate is less than 0.1 m / s² and the maximum acceleration is less than 0.2 m / s², the speed mode determination result is "uniform speed navigation"; if the average speed change rate is greater than or equal to 0.1 m / s² and less than 0.5 m / s², or the maximum acceleration is greater than or equal to 0.2 m / s² and less than 0.4 m / s², the speed mode determination result is "slow acceleration"; if the average speed change rate is less than or equal to -0.1 m / s² and greater than -0.5 m / s², or the maximum acceleration is less than or equal to -0.2 m / s² and greater than -0.4 m / s², the speed mode determination result is "slow deceleration"; if the average speed change rate is greater than or equal to 0.5 m / s², or the maximum acceleration is greater than or equal to 0.4 m / s², the speed mode determination result is "rapid acceleration"; if the average speed change rate is less than or equal to -0.5 m / s², or the maximum acceleration is less than or equal to -0.4 m / s², the speed mode determination result is "rapid deceleration".

[0076] Special mode recognition includes: recognizing special motion modes according to the course mode determination result and the speed mode determination result. The specific rules are as follows: If the speed mode determination result is "uniform speed navigation" and the course mode determination result is "rapid turning", the special mode recognition result is "emergency turning"; if the speed mode determination result is "slow deceleration" or "rapid deceleration" and the course mode determination result is "straight navigation", the special mode recognition result is "braking"; if the target speed is less than 1 m / s and the course mode determination result is "rapid turning" or "slow turning", the special mode recognition result is "drifting".

[0077] The comprehensive determination of the motion state includes: comprehensively determining the motion state of the target based on the determination results of the heading mode, speed mode, and special mode recognition results. The specific rules are as follows: If there is a special mode recognition result, the special mode recognition result is preferentially used as the basic motion state determination result; if there is no special mode recognition result, the basic motion state is determined according to the combination of the determination results of the heading mode and speed mode. For example, if the determination result of the heading mode is "straight navigation" and the determination result of the speed mode is "uniform speed navigation", the basic motion state determination result is "uniform speed straight navigation"; if the determination result of the heading mode is "slow turning" and the determination result of the speed mode is "uniform speed navigation", the basic motion state determination result is "uniform speed turning".

[0078] The inference of the motion intention includes: obtaining the historical motion state sequence of the target, and the historical motion state sequence includes the basic motion state determination results of the target within the past 5 minutes. Analyze the historical motion state sequence to infer the motion intention of the target. The specific rules are as follows: If there is a pattern of "uniform speed straight navigation" -> "slow turning" -> "uniform speed straight navigation" in the historical motion state sequence of the target, the motion intention inference result is "heading adjustment"; if there is a pattern of "uniform speed straight navigation" -> "deceleration" -> "stop" in the historical motion state sequence of the target, the motion intention inference result is "planned ship stop"; if the basic motion state determination result of the target is "emergency turning", the motion intention inference result is "collision avoidance".

[0079] Generating the target motion state determination result includes: combining the basic motion state determination result and the motion intention inference result to generate the target motion state determination result. For example, if the basic motion state determination result is "uniform speed turning" and the motion intention inference result is "heading adjustment", the target motion state determination result is "uniform speed turning, heading adjustment". If the basic motion state determination result is "uniform speed straight navigation" and the motion intention inference result is "planned ship stop", the target motion state determination result is "uniform speed straight navigation, planned ship stop". If there is no motion intention inference result, only the basic motion state determination result is used as the target motion state determination result.

[0080] Preferably, the multi-target collision risk assessment in step S2 includes:

[0081] Calculating and analyzing the trajectory correlation according to the target motion state determination result to obtain the target correlation network diagram;

[0082] Performing multi-target joint trajectory prediction according to the target motion state determination result and the target correlation network diagram to obtain a multi-target prediction trajectory set;

[0083] Perform a collision risk assessment on the multi-target prediction trajectory set to obtain a collision risk coefficient; generate a danger level assessment matrix based on the collision risk coefficient.

[0084] In the embodiments of the present invention, the calculation and analysis of trajectory correlation include: calculating the relative motion parameters between every two targets, including relative distance, relative speed, and relative heading.

[0085] The calculation formula for relative distance is: ;

[0086] where (x_target1, y_target1) and (x_target2, y_target2) are the position coordinates of the two targets respectively.

[0087] The calculation formula for relative speed is: ;

[0088] where (vx_target1, vy_target1) and (vx_target2, vy_target2) are the velocity vectors of the two targets respectively.

[0089] The calculation formula for relative heading is: atan2(vy_target1 - vy_target2, vx_target1 - vx_target2).

[0090] According to the determination result of the target motion state, judge whether the motion states between the targets are relevant. If both targets are in the "uniform linear navigation" state and the relative heading is close to 0 degrees or 180 degrees, it is considered that there is a potential collision risk between these two targets. If one target is in the "emergency turning" state and the other target is on its turning route, it is considered that there is an association of emergency collision avoidance between these two targets. Calculate the trajectory correlation coefficient between the targets. The calculation formula for the trajectory correlation coefficient is: correlation coefficient = 1 - (relative distance / maximum distance) × (1 - exp(-relative speed / speed threshold)) × (1 - cos(relative heading)), where the maximum distance refers to the maximum safe distance of ships in the current sea area, and the speed threshold is 2 m / s. If the correlation coefficient is greater than 0.7, it is considered that there is a trajectory correlation between these two targets. Construct a target correlation network diagram. The nodes of the network diagram represent the targets. If there is a trajectory correlation between two targets, connect these two nodes in the network diagram, and the weight of the edge is the trajectory correlation coefficient. For example, if the correlation coefficient between target A and target B is 0.8 and the correlation coefficient between target B and target C is 0.6, then connect A and B in the network diagram with the weight of the edge being 0.8, and connect B and C with the weight of the edge being 0.6.

[0091] Multi - target joint trajectory prediction includes: First, based on the determination result of the target motion state, predict the trajectory of each target within the next 5 minutes. If the target is in the "uniform straight - line navigation" state, predict its future trajectory as a straight line; if the target is in the "uniform turning" state, predict its future trajectory as an arc; if the target is in the "acceleration" or "deceleration" state, predict its future trajectory as a curve. For a target in the "emergency turning" state, predict its trajectory based on its current turning rate. Then, according to the target correlation network graph, adjust the predicted trajectory. If there is a trajectory correlation between two targets, consider their mutual influence. For example, if there is a potential collision risk between target A and target B, then when predicting the trajectories of target A and target B, their collision avoidance behaviors need to be considered. If target A is predicted to turn left to avoid collision, then adjust the predicted trajectory of target B so that it turns right. Generate multiple possible trajectory prediction schemes. For example, for target A, if there are two possible collision avoidance schemes (turn left or turn right), then generate two predicted trajectories. Calculate the probability of each prediction scheme. The probability calculation formula is: Probability = exp(−Risk Cost), where the risk cost is calculated based on factors such as the distance between the predicted trajectory and the potential collision area, and the relative speed between the targets. Integrate each scheme to form the final multi - target joint trajectory prediction result, obtaining a multi - target prediction trajectory set, where the trajectory set contains each target and its predicted trajectory.

[0092] Collision risk assessment includes: Calculate the predicted closest point of approach (DCPA) and the time to the closest point of approach (TCPA) between the own ship and each target. The calculation of DCPA and TCPA is based on the predicted trajectory. For a straight - line trajectory, analytical geometry methods can be used for calculation; for a curve trajectory, numerical iteration methods can be used for calculation. Calculate the collision risk coefficient. The formula for calculating the collision risk coefficient is: Risk Coefficient = 1 / (DCPA + 1)×(1−exp(−TCPA / Time Threshold))×Relative Speed×Target Size, where the time threshold is 30 seconds, and the target size is obtained according to the AIS information of the target. If the target is a cargo ship, the target size is 200 meters; if the target is a fishing boat, the target size is 50 meters. According to the collision risk coefficient, generate a risk assessment matrix. The risk assessment matrix is a two - dimensional matrix, where the rows and columns of the matrix represent targets, and the elements of the matrix represent the collision risk coefficients between two targets. For example, the element (i, j) of the matrix represents the collision risk coefficient between target i and target j. If the collision risk coefficient between target i and target j is greater than 0.5, it is considered that there is a collision risk and it is marked in the matrix. Finally, obtain the risk assessment matrix.

[0093] Of particular importance is that the calculation and analysis of trajectory correlation specifically are:

[0094] Calculate the target relative motion parameter table based on the target motion state determination result and the target trajectory feature sequence; calculate the trajectory correlation coefficient matrix based on the target relative motion parameter table; perform interactive behavior pattern recognition based on the trajectory correlation coefficient matrix and the target relative motion parameter table to obtain the target interactive behavior feature table; construct the target relationship structure diagram based on the target interactive behavior feature table; perform key influencing target screening based on the target relationship structure diagram and the target interactive behavior feature table to obtain the target correlation network diagram;

[0095] In the embodiment of the present invention, the calculation of the target relative motion parameter table includes: extracting information from the target motion state determination result and the target trajectory feature sequence. For each target pair (including the own ship and other targets, targets and other targets), calculate the following relative motion parameters: relative distance, relative speed, relative course, approach rate, and relative distance change rate.

[0096] The calculation formula for the relative distance is: ;

[0097] where (x_target1, y_target1) and (x_target2, y_target2) are the position coordinates of the two targets respectively, and these position coordinates are obtained from the fused situation vector diagram.

[0098] The calculation formula for the relative speed is: ;

[0099] where (vx_target1, vy_target1) and (vx_target2, vy_target2) are the velocity vectors of the two targets respectively, and these velocity vectors are obtained from the fused situation vector diagram.

[0100] The calculation formula for the relative heading is: atan2(vy_target1 - vy_target2, vx_target1 - vx_target2). The calculation formula for the approach rate is: (relative speed × cos(relative heading)), and the approach rate is used to measure the degree of approach between targets. The calculation formula for the relative distance change rate is: (relative distance_t+1 - relative distance_t) / time interval, where relative distance_t+1 and relative distance_t are the relative distances at adjacent time points, and the time interval is 1 second. Record the calculated relative motion parameters in tabular form to form a target relative motion parameter table. Calculate the trajectory correlation coefficient matrix based on the target relative motion parameter table. The calculation of the trajectory correlation coefficient matrix includes: calculating the trajectory correlation coefficient according to the relative motion parameters in the target relative motion parameter table. The trajectory correlation coefficient is used to measure the similarity of the trajectories of two targets and the potential interaction risk. The trajectory correlation coefficient matrix is a symmetric matrix, where the rows and columns of the matrix represent targets, and the matrix elements represent the trajectory correlation coefficients between two targets. The calculation formula for the trajectory correlation coefficient is: correlation coefficient = w1×f(relative distance)+w2×g(relative speed)+w3×h(relative heading)+w4×i(approach rate)+w5×j(relative distance change rate). Among them, w1, w2, w3, w4, w5 are weight coefficients, and w1 + w2 + w3 + w4 + w5 = 1. These weight coefficients are adjusted according to the actual application scenario. For example, in a crowded water area, the weights of w1 and w3 can be increased, while in an open water area, the weights of w2 and w4 can be increased. f(relative distance)=1 / (1 + relative distance / distance threshold), where the distance threshold is set according to the actual scenario, for example, set to 1 nautical mile. g(relative speed)=1 - relative speed / speed threshold, where the speed threshold is set according to the actual scenario, for example, set to 5 m / s. h(relative heading)=(1 - |relative heading| / 180 degrees), and the unit of relative heading is degrees. i(approach rate)=approach rate / approach rate threshold, where the approach rate threshold is set according to the actual scenario, for example, set to 2 m / s. j(relative distance change rate)=1−(relative distance change rate / distance change rate threshold), where the distance change rate threshold is set according to the actual scenario, for example, set to 0.5 m / s. Fill the calculated trajectory correlation coefficients into the corresponding positions of the trajectory correlation coefficient matrix. For example, if the correlation coefficient between target A and target B is 0.8, then the elements at positions (A, B) and (B, A) in the trajectory correlation coefficient matrix are both set to 0.8. Perform interactive behavior pattern recognition based on the trajectory correlation coefficient matrix and the target relative motion parameter table to obtain a target interactive behavior feature table. The interactive behavior pattern recognition includes: identifying the interactive behavior patterns between targets according to the trajectory correlation coefficient matrix and the target relative motion parameter table. The interactive behavior patterns include: head-on encounter, crossing encounter, overtaking, and collision avoidance, etc.Identifying the head-on encounter mode: If the relative headings of two targets are close to 180 degrees, the relative distance continues to decrease, and the trajectory correlation coefficient is greater than 0.7, then it is identified as the head-on encounter mode. Identifying the crossing encounter mode: If the relative headings of two targets are between 30 degrees and 150 degrees, the relative distance continues to decrease, and the trajectory correlation coefficient is greater than 0.7, then it is identified as the crossing encounter mode. Identifying the overtaking mode: If the relative heading of two targets is close to 0 degrees, the speed of the own ship is greater than the speed of the target, the relative distance continues to decrease, and the trajectory correlation coefficient is greater than 0.7, then it is identified as the overtaking mode. Identifying the collision avoidance mode: If one target suddenly changes its heading or speed, and the trajectory correlation coefficient between it and the other target is greater than 0.7, and the relative distance rapidly increases, then it is identified as the collision avoidance mode. For each interaction behavior mode, extract its characteristic parameters, including: interaction type, duration, relative distance change rate, relative heading change rate, and trajectory correlation coefficient. Record it in tabular form to generate the target interaction behavior characteristic table. Construct the target relationship structure diagram according to the target interaction behavior characteristic table. The construction of the target relationship structure diagram includes: According to the target interaction behavior characteristic table, construct a graphical structure to represent the relationship between targets. The nodes of the graph represent targets, and the edges of the graph represent the interaction relationships between targets. If there is an interaction behavior between two targets, then connect these two targets in the graph. The type of the edge represents the type of the interaction behavior, and the weight of the edge represents the intensity of the interaction behavior. The types of edges include: head-on encounter, crossing encounter, overtaking, collision avoidance, etc. The weight of the edge is calculated according to the trajectory correlation coefficient and the characteristic parameters of the interaction behavior. For example, for the head-on encounter mode, the weight of the edge can be set as the trajectory correlation coefficient × (1 - relative distance / distance threshold), and for the collision avoidance mode, the weight of the edge can be set as the trajectory correlation coefficient × relative speed change rate. The target relationship structure diagram can be represented by an adjacency matrix or an adjacency list. For example, if there is a head-on encounter behavior between target A and target B with a weight of 0.8, then the elements at positions (A, B) and (B, A) in the adjacency matrix are set to (head-on encounter, 0.8). Conduct key influencing target screening based on the target relationship structure diagram and the target interaction behavior characteristic table to obtain the target correlation network diagram. The key influencing target screening includes: According to the target relationship structure diagram and the target interaction behavior characteristic table, screen out the key influencing targets that have a greater impact on the navigation of the own ship. The screening method includes: calculating the centrality index of each target. The centrality index includes degree centrality, closeness centrality, betweenness centrality, etc. Degree centrality represents the number of edges connecting a target to other targets. Closeness centrality represents the average distance from a target to all other targets. Betweenness centrality represents the number of times a target appears on the shortest path between other targets. Sort the targets according to the centrality index. Select several targets with the highest centrality index as the key influencing targets. For example, select the 3 targets with the highest degree centrality as the key influencing targets.Construct a target correlation network graph, where the nodes of the network graph represent targets. If there is an interaction behavior between two targets, then connect these two nodes in the network graph. Only retain the key influencing targets and the targets that have interaction relationships with the key influencing targets. The types and weights of the edges are the same as those in the target relationship structure diagram. The types of edges include: head-on encounter, crossing encounter, overtaking, collision avoidance, etc. The weights of the edges are calculated based on the trajectory correlation coefficient and the characteristic parameters of the interaction behavior. Finally, obtain the target correlation network graph.

[0101] Step S31: Obtain the current sea area environment information; generate a sea area environment constraint graph according to the sea area environment information;

[0102] Step S32: Construct a navigation risk cost map according to the sea area environment constraint graph and the risk degree evaluation matrix;

[0103] Step S33: Perform ship maneuverability constraints according to the navigation risk cost map to obtain a route planning maneuvering constraint rule set;

[0104] Step S34: Perform direct search and screening of routes according to the route planning maneuvering constraint rule set and the navigation risk cost map to obtain a set of route candidate plans;

[0105] Step S35: Generate a navigation instruction sequence for the set of route candidate plans to obtain a navigation trajectory plan.

[0106] In the embodiment of the present invention, obtaining the current sea area environment information includes: obtaining the water depth data, seabed terrain data, and obstacle distribution information of the sea area from the nautical chart database. The water depth data is stored in the form of a grid map, the grid resolution is 10 meters × 10 meters, and the data unit is meters. The seabed terrain data includes reefs, shoals, etc., and the obstacle distribution information includes bridges, transmission lines, etc. Obtain real-time meteorological data from the meteorological database, including wind direction, wind speed, wave height, wave direction, and visibility. The data update frequencies of the wind speed and wave height are 10 minutes, and the data units are meters per second and meters respectively. Obtain tidal data from the tidal database, including tidal current velocity and direction, and the data update frequency is 1 hour. Obtain navigation restriction area information from the navigation restriction database, including waterways, anchorages, restricted navigation areas, and military areas, etc. The waterway information is stored in the form of line segments, the anchorage information is stored in the form of polygons, and the restricted navigation areas and military areas are stored in the form of polygons. Perform spatio-temporal unification and format conversion on the above various types of environmental data. The unified coordinate system is WGS84, the unified time is UTC time, and the data unit is unified into the International System of Units. Generate a sea area environment constraint graph, and the sea area environment constraint graph is a two-dimensional grid map, and the grid resolution is the same as that of the water depth data. Each grid contains the following information: water depth, seabed terrain, obstacle information, wind speed, wind direction, wave height, wave direction, tidal current velocity, tidal current direction, and navigation restriction information.

[0107] The construction of the navigation risk cost map includes: dividing the sea area environment constraint map into uniform grids, with the grid resolution being the same as the sea area environment constraint Figure 1 Figure 1 leads to. Establish a discrete model of the navigation space. Calculate the grounding risk coefficient based on the water depth data and the draft of the ship. The calculation formula for the grounding risk coefficient is: Risk coefficient_grounding = 1 / (1 + exp(k × (draft - water depth))), where k is a constant set to 0.5. If the water depth is less than the draft of the ship, the grounding risk coefficient approaches 1; otherwise, it approaches 0. Calculate the collision risk coefficient for each grid according to the risk assessment matrix. The calculation formula for the collision risk coefficient is: Risk coefficient_collision = Σ(Risk coefficient_ij × f(Relative distance_ij)), where Risk coefficient_ij is the collision risk coefficient between the own ship and target j in the risk assessment matrix, f(Relative distance_ij) = 1 / (1 + Relative distance_ij / Safety distance), Relative distance_ij is the distance between the own ship and target j, and the safety distance is the ship's length × 5. Considering the impact of wind and waves on navigation, calculate the meteorological risk coefficient. The calculation formula for the meteorological risk coefficient is: Risk coefficient_meteorological = 1 + a × Wind speed + b × Wave height, where a and b are constants set to 0.1 and 0.2 respectively. Construct a comprehensive risk cost function. The calculation formula for the comprehensive risk cost function is: Cost = w_grounding × Risk coefficient_grounding + w_collision × Risk coefficient_collision + w_meteorological × Risk coefficient_meteorological, where w_grounding, w_collision, and w_meteorological are weight coefficients, and w_grounding + w_collision + w_meteorological = 1. The weight coefficients are set according to the actual application scenario. For example, in shallow water areas, the weight of w_grounding can be increased. Assign the calculated cost to each grid to generate the navigation risk cost map. The navigation risk cost map is also a two-dimensional grid map, and each grid contains cost information.

[0108] Constraints on ship maneuverability include: Query the measured turning radius at different speeds according to the ship's turning ability query table. The ship's turning ability query table is generated according to sub-step 3.3.2. Obtain the measured turning radius at different speeds. For example, when the speed is 10 knots, the measured turning radius is 100 meters. According to the ship's acceleration and deceleration ability query table, obtain the time and distance required to reach the target speed from the current speed. The ship's acceleration and deceleration ability query table is generated according to sub-step 3.3.4. For example, the time required to accelerate from 10 knots to 15 knots is 30 seconds, and the distance is 200 meters. According to the safe speed limit table, determine the maximum safe speed under different sea conditions. The safe speed limit table is generated according to sub-step 3.3.6. For example, in light sea conditions, the maximum safe speed is 0.9 times the designed maximum speed. According to the economic speed range table, determine the speed range with the best fuel economy. The economic speed range table is generated according to sub-step 3.3.7. For example, the economic speed range is 70% to 85% of the designed speed. Convert the above constraints into hard limits for route planning. The hard limits include: Set the minimum turning radius limit. The distance between adjacent waypoints must be greater than or equal to the minimum turning radius. Set the speed limit. According to the sea conditions and the safe speed limit table, limit the speed of the waypoints. Set the acceleration and deceleration limits. The speed change between adjacent waypoints cannot exceed the maximum acceleration or deceleration. Mark the infeasible areas on the navigation risk cost map. Create a set of route planning maneuvering constraint rules. The set of route planning maneuvering constraint rules includes: Constraints such as minimum turning radius, maximum speed, maximum acceleration, and maximum deceleration.

[0109] Direct route search and screening include: Mark the starting point, ending point, and intermediate mandatory points directly on the nautical chart to determine the basic requirements of the navigation task. Adopt a hierarchical search strategy. Use the A* algorithm for route search. Set the route evaluation criteria. The route evaluation criteria include: Safety first, considering the shortest voyage under the premise of ensuring safety. The evaluation index for safety is: The maximum cost on the route. The evaluation index for the voyage is: The total length of the route. According to the A* algorithm for route search, calculate the cost and voyage of the route, and select the route that meets the safety constraints and has the shortest voyage. Generate 3 candidate routes. These 3 candidate routes all meet the safety constraints and have similar voyages. Conduct a comprehensive score for the 3 candidate routes. The formula for the comprehensive score is: Score = α×(1−Cost / Maximum Cost)+β×(1−Voyage / Maximum Voyage), where α and β are weight coefficients, and α+β = 1. The weight coefficients are set according to the actual application scenario. For example, in a high-risk area, the weight of α can be increased. Select the route with the highest comprehensive score as the recommended plan. Generate a set of route candidate plans, and the set of route candidate plans contains 3 candidate routes.

[0110] Generate a navigation instruction sequence for the set of candidate route plans to obtain a navigation trajectory plan. The generation of the navigation instruction sequence includes: smoothing the optimal route (i.e., the route with the highest comprehensive score) to eliminate unnecessary heading changes. Interpolating the optimal route to generate route points. Extracting key heading change points. A key heading change point refers to a route point where the heading changes significantly. Calculate the recommended speed and estimated time of arrival for each key point. The calculation method for the recommended speed is: determine the recommended speed according to the safe speed limit and the economic speed range. The calculation method for the estimated time of arrival is: calculate the estimated time of arrival based on the voyage distance and the recommended speed. Generate a standardized navigation instruction sequence, including heading, speed, and time information. For example, the instruction sequence can be expressed as: [heading 1, speed 1, time 1], [heading 2, speed 2, time 2],... Simulate and verify the generated navigation instructions. The simulation verification includes: using a ship maneuvering simulator to simulate the ship sailing according to the generated navigation instructions and verifying whether the navigation trajectory meets the safety and executability requirements. If the simulation results do not meet the requirements, readjust the navigation instructions until the requirements are met. Generate an optimal navigation trajectory plan, including a sequence of heading points, estimated time of arrival, and navigation instructions for the recommended speed.

[0111] Preferably, the ship maneuvering performance constraints in step S33 include:

[0112] Collect and construct a steering ability query table based on the measured data of the boat's steering ability;

[0113] Collect and construct an acceleration / deceleration ability query table based on the acceleration / deceleration performance test data set of the boat;

[0114] Determine the sea condition influence coefficient based on the steering ability query table and the acceleration / deceleration ability query table to obtain a sea condition correction coefficient table for maneuvering performance;

[0115] Determine the safe speed limit table and economic speed range table of the boat;

[0116] Generate a ship maneuvering performance constraint table based on the steering ability query table, the acceleration / deceleration ability query table, the sea condition correction coefficient table for maneuvering performance, the safe speed limit table, and the economic speed range table;

[0117] Perform a hard constraint conversion on the ship maneuvering performance constraint table and the navigation risk cost map to obtain a set of navigation planning maneuvering constraint rules.

[0118] In the embodiments of the present invention, the construction of the steering ability lookup table includes: conducting a standard Z-type steering test. Collect the standard Z-type steering test data of the boat at 5 typical speed points (the lowest speed, 1 / 4 design speed, 1 / 2 design speed, 3 / 4 design speed, and the highest design speed). Repeat the test 3 times at each speed point and record the test data. For each speed point, record the following parameters: the time difference between the moment of rudder angle input and the moment when the boat starts to turn, and calculate the "steering response delay time"; the angle change of the course per second during the steering process, and calculate the "course change rate"; the actual turning radius in the steady steering state, denoted as the "measured turning radius". For example, at the highest design speed, the recorded steering response delay time is 3 seconds, the course change rate at the maximum rudder angle is 5 degrees / second, and the measured turning radius is 120 meters. Eliminate the obviously abnormal data points to ensure the consistency and reliability of the data. Take the average value of the multiple test data at the same speed. Organize and record: the speed value (knots), the corresponding steering response delay time (seconds), the course change rate at the maximum rudder angle (degrees / second), and the measured turning radius (meters) at the standard rudder angle (15 degrees). For example, the organized data is: 10 knots, 3 seconds, 5 degrees / second, 120 meters; 15 knots, 2.5 seconds, 6 degrees / second, 150 meters. Use the piecewise linear interpolation method to calculate the steering performance parameter values between the test points. Construct a "steering ability lookup table" with a speed interval of 1 knot, including the steering performance parameters of all integer speed points within the speed range. For example, at 12 knots, according to linear interpolation, the steering response delay time is 2.75 seconds, the course change rate at the maximum rudder angle is 5.5 degrees / second, and the measured turning radius is 135 meters. Add data reliability marks to the lookup table, mark the test point data as "measured data", and mark the interpolated data as "estimated data".

[0119] The construction of the acceleration and deceleration ability query table includes: conducting acceleration and deceleration performance tests. Collecting the standard acceleration test data of the boat at different initial speeds: accelerating from the stationary state to each target speed; accelerating from multiple intermediate speeds to the maximum designed speed. Collecting the standard deceleration test data of the boat at different initial speeds: decelerating from each initial speed to the stationary state, including two working conditions of normal deceleration and emergency stop. For each group of acceleration tests, record the time and sailing distance required from the initial speed to the target speed. For each group of deceleration tests, record the time and sailing distance required to decelerate from the initial speed to a stop. Calculate the average acceleration and deceleration under each working condition, denoted as "measured acceleration performance" and "measured deceleration performance". For example, when accelerating from 10 knots to 15 knots, the required time is 30 seconds, the sailing distance is 200 meters, and the average acceleration is 0.17 knots / second; when decelerating from 15 knots to a stop emergently, the required time is 60 seconds, the sailing distance is 300 meters, and the average deceleration is -0.25 knots / second. Construct an "acceleration ability query table", with the row being the initial speed and the column being the target speed, and the table content including: acceleration required time (seconds), sailing distance during acceleration (meters), average acceleration (knots / second). Construct a "deceleration ability query table", with the row being the initial speed and the column being the deceleration type (normal / emergency), and the table content including: time required to decelerate to a stop (seconds), sailing distance during deceleration (meters), average deceleration (knots / second). Use the two-dimensional interpolation method to supplement the data between the test points to form a complete query table. Check the key points in the query table to ensure the consistency and rationality of the data.

[0120] The construction of the seakeeping correction coefficient table for maneuverability performance includes: collecting the data of the changes in the maneuverability performance of the boat under different sea state levels. The sea state levels are divided into: calm sea state (level 0 - 2), slight sea state (level 3 - 4), moderate sea state (level 5 - 6), large sea state (level 7 - 8), and severe sea state (above level 9). For example, under the slight sea state, the measured turning radius increases by 20%, the acceleration time increases by 10%, and the deceleration time increases by 30%. Calculate the change ratio of the steering performance under each sea state level to obtain the "seakeeping correction coefficient for steering performance": calm sea state (level 0 - 2): steering correction coefficient = 1.0; slight sea state (level 3 - 4): steering correction coefficient = 1.2; moderate sea state (level 5 - 6): steering correction coefficient = 1.5; large sea state (level 7 - 8): steering correction coefficient = 2.0; severe sea state (above level 9): steering correction coefficient = 3.0. Calculate the change ratio of the acceleration performance under each sea state level to obtain the "seakeeping correction coefficient for acceleration performance": calm sea state (level 0 - 2): acceleration correction coefficient = 1.0; slight sea state (level 3 - 4): acceleration correction coefficient = 1.1; moderate sea state (level 5 - 6): acceleration correction coefficient = 1.3; large sea state (level 7 - 8): acceleration correction coefficient = 1.6; severe sea state (above level 9): acceleration correction coefficient = 2.0. Calculate the change ratio of the deceleration performance under each sea state level to obtain the "seakeeping correction coefficient for deceleration performance": calm sea state (level 0 - 2): deceleration correction coefficient = 1.0; slight sea state (level 3 - 4): deceleration correction coefficient = 1.3; moderate sea state (level 5 - 6): deceleration correction coefficient = 1.7; large sea state (level 7 - 8): deceleration correction coefficient = 2.2; severe sea state (above level 9): deceleration correction coefficient = 3.0.

[0121] The construction of the safe speed limit table includes: extracting key stability parameters from the boat stability calculation book, including: initial metacentric height (m), maximum righting lever (m), angle of vanishing stability (degrees). Combining the actual navigation test data, analyze the motion response of the boat under different sea conditions, including: maximum roll angle (degrees), maximum pitch angle (degrees), deck wetting frequency (times / hour). Based on the safety threshold standard, determine the maximum safe speed coefficient for each sea condition. Calm sea conditions (level 0 - 2): safe speed coefficient = 1.0 (can reach the designed maximum speed); light sea conditions (level 3 - 4): safe speed coefficient = 0.9; moderate sea conditions (level 5 - 6): safe speed coefficient = 0.7; rough sea conditions (level 7 - 8): safe speed coefficient = 0.5; severe sea conditions (above level 9): safe speed coefficient = 0.3. Calculate the specific maximum safe speed value for each sea condition: maximum safe speed = designed maximum speed × safe speed coefficient. The construction of the economic speed range table includes: analyzing the three-dimensional relationship curve of engine power - speed - fuel consumption. Calculate the fuel consumption rate per unit voyage to obtain the "speed - fuel efficiency curve". Find the lowest point of the fuel consumption rate to determine the "optimal economic speed". Determine the speed range where the fuel consumption rate does not exceed 10% of the lowest point to obtain the "economic speed range". Verify that the economic speed range usually falls within the range of 70% - 85% of the designed speed. Consider the variation of the economic speed under different displacements and establish a comparison table of "displacement - economic speed correction coefficient".

[0122] The generation of the ship maneuverability constraint table includes: integrating various maneuverability parameters to construct a "basic maneuverability data set". Design the data structure of the maneuverability constraint table, which includes the following main parts: speed-related constraints (minimum speed, maximum safe speed, economic speed range), steering-related constraints (steering response time, minimum turning radius, maximum steering rate), acceleration and deceleration-related constraints (maximum acceleration, maximum deceleration, emergency stopping distance), sea condition influence correction (correction coefficients of each performance parameter under different sea conditions). According to the current sea condition level, automatically select the corresponding correction coefficient. Calculate the corrected values of each maneuverability parameter. Generate a complete ship maneuverability constraint table, including the original data and the actually available data after sea condition correction. For example, in light sea conditions, the maximum speed is 0.9 times the designed speed, the minimum turning radius increases by 20%, the acceleration time increases by 10%, and the deceleration time increases by 30%.

[0123] Route planning hard constraint conversion includes: converting maneuverability constraints into hard constraint conditions for the route planning algorithm. Calculate the minimum distance requirement between adjacent route points based on the minimum turning radius. Calculate the maximum course change limit between adjacent route segments based on the maximum turning rate. Calculate the minimum distance requirement between speed change points based on the acceleration and deceleration capabilities. Mark the infeasible areas on the navigation risk cost map. Based on the minimum turning radius, mark the areas with insufficient turning radius on the navigation risk cost map, and the grids in these areas are marked as impassable. Based on the safe speed limit and economic speed range, mark the speed limit areas on the navigation risk cost map. For example, in shallow water areas or narrow channels, limit the speed, and the corresponding grids are marked as speed limit areas. Based on the acceleration and deceleration capabilities, mark the acceleration and deceleration limited areas on the navigation risk cost map. For example, in the deceleration area near the port, limit the acceleration and deceleration, and the corresponding grids are marked as acceleration and deceleration limited areas. Set the maneuverability constraint check rules for the route planning algorithm: Course change check: The course change between adjacent route segments does not exceed the maximum allowable turning angle. The formula for calculating the maximum allowable turning angle is: arcsin(turning radius / distance between adjacent route points). Turning space check: Ensure that there is enough space around the turning point to perform the turning operation. The condition for the turning space check is that the grids around the turning point are all passable. Speed change check: Ensure that there is enough distance between speed change points to perform the acceleration and deceleration operations. According to the acceleration and deceleration capability query table, calculate the distance required for acceleration or deceleration, and ensure that the distance between adjacent route points is greater than this distance. Generate the "maneuverability constraint rule set" that can be directly used by the route planning algorithm. The route planning maneuver constraint rule set includes: minimum turning radius, maximum speed, maximum acceleration, maximum deceleration, as well as course change constraints, turning space constraints, acceleration and deceleration distance constraints, etc. These constraint rules are used to check and adjust the route during the route planning process to ensure that the generated route meets the maneuverability requirements of the ship. For example, during the route search process, if the distance between the current route point and the next route point is less than the minimum turning radius, the route is considered infeasible and needs to be adjusted. If the speed of the route point exceeds the safe speed limit, the speed needs to be adjusted, or the route needs to be replanned.

[0124] Preferably, step S4 includes the following steps:

[0125] Step S41: Obtain the real-time navigation data of the boat and perform real-time monitoring of the navigation deviation according to the navigation trajectory plan to obtain the navigation deviation status evaluation report;

[0126] Step S42: Obtain the real-time environmental data and quantify the safety impact of environmental factors according to the navigation deviation status evaluation report to obtain the environmental safety impact coefficient table;

[0127] Step S43: Obtain the boat performance parameters, and conduct real-time evaluation of the boat's maneuverability according to the environmental safety impact factor table to obtain the real-time maneuverability evaluation index;

[0128] Step S44: Calculate the warning distance according to the real-time maneuverability evaluation index and the navigation deviation status evaluation report to obtain the graded warning distance table;

[0129] Step S45: Generate the warning and avoidance strategy according to the graded warning distance table and the risk assessment matrix to obtain the intelligent warning decision table.

[0130] In the embodiments of the present invention, obtaining the real-time navigation data of a boat includes: obtaining the real-time navigation status data such as the current position, heading, and speed of the boat. The real-time navigation status data is obtained through on-board sensors, and the data update frequency is 1 second. Obtain the optimal navigation trajectory plan, which is the navigation trajectory plan generated in step S35 and includes navigation instructions such as a sequence of heading points, estimated arrival time, and recommended speed. Real-time monitoring of navigation deviation according to the navigation trajectory plan includes: calculating the position deviation, heading deviation, and speed deviation between the actual navigation status and the optimal navigation trajectory plan. The calculation formula for the position deviation is: √((x_actual - x_trajectory)^2 + (y_actual - y_trajectory)^2), where (x_actual, y_actual) is the actual position of the boat, and (x_trajectory, y_trajectory) is the position at the corresponding moment in the optimal navigation trajectory plan. The calculation formula for the heading deviation is: |heading_actual - heading_trajectory|, where heading_actual is the actual heading of the boat, and heading_trajectory is the heading at the corresponding moment in the optimal navigation trajectory plan. The calculation formula for the speed deviation is: |speed_actual - speed_trajectory|, where speed_actual is the actual speed of the boat, and speed_trajectory is the recommended speed at the corresponding moment in the optimal navigation trajectory plan. Analyze the changing trend of the deviation to identify whether the deviation is increasing or decreasing. If the position deviation, heading deviation, and speed deviation all show an increasing trend in 5 consecutive data, it is considered that the deviation is increasing. If the deviation shows a decreasing trend in 5 consecutive data, it is considered that the deviation is decreasing. Calculate the time and space margins required for deviation correction. The time required for deviation correction is calculated based on the heading deviation and speed deviation. For example, if the heading deviation is too large and the heading needs to be adjusted, the time required to adjust the heading is calculated according to the steering ability lookup table. The space margin required for deviation correction is calculated based on the position deviation and heading deviation. For example, if the position deviation is too large and the heading and speed need to be adjusted, the space required to adjust the heading and speed is calculated according to the acceleration / deceleration ability lookup table and the steering ability lookup table. Evaluate the degree of impact of the current deviation on navigation safety. If the deviation is too large and exceeds the safety threshold, it is considered to have a greater impact on navigation safety. The safety threshold can be set according to factors such as boat type and navigation environment. Generate a navigation deviation status assessment report. The navigation deviation status assessment report includes information such as position deviation, heading deviation, speed deviation, deviation changing trend, time and space margins required for deviation correction, and the degree of impact on navigation safety.

[0131] Obtaining real-time environmental data includes: obtaining real-time data on environmental factors such as the current sea state, visibility, and wind force. The sea state data is obtained through wave height sensors and hull attitude sensors, the visibility data is obtained through visibility sensors, and the wind force data is obtained through wind speed and direction sensors. The data update frequency is 10 seconds. Analyze the degree of influence of each environmental factor on the boat handling performance. The influence of the sea state on the boat handling performance is evaluated according to the sea state level and the sea state correction coefficient table for handling performance. The influence of visibility on the boat handling performance is that the lower the visibility, the greater the influence on the radar detection range and the AIS signal reception quality. The influence of wind force on the boat handling performance is that the greater the wind force, the greater the influence on the boat's course and speed. Calculate the safe stopping distance and turning radius under different environmental conditions. The calculation formula for the safe stopping distance is: braking distance = braking distance in the braking distance reference table × sea state correction coefficient × (1 + visibility influence coefficient), where the braking distance reference table is generated in sub-step 4.4.1, the sea state correction coefficient is obtained according to the sea state correction coefficient table for handling performance, and the visibility influence coefficient is calculated according to the visibility. If the visibility is less than 0.5 nautical miles, the visibility influence coefficient is 0.5. If the visibility is between 0.5 - 1 nautical mile, the visibility influence coefficient is 0.2, otherwise it is 0. The calculation formula for the turning radius is: turning radius = turning radius in the turning ability query table × sea state correction coefficient, where the turning ability query table is generated in sub-step 3.3.2, and the sea state correction coefficient is obtained according to the sea state correction coefficient table for handling performance. Evaluate the influence of environmental factors on the radar detection range and the AIS signal reception quality. The radar detection range is affected by the sea state and visibility. The lower the visibility, the shorter the detection range. The AIS signal reception quality is affected by the sea state and visibility. The worse the sea state and the lower the visibility, the worse the reception quality. Considering the influence of all environmental factors, generate an environmental safety coefficient. The calculation formula for the environmental safety coefficient is: safety coefficient = α × sea state influence coefficient + β × visibility influence coefficient + γ × wind force influence coefficient, where α, β, and γ are weight coefficients, and α + β + γ = 1. The weight coefficients are set according to the actual application scenario. Generate an environmental safety influence coefficient table. The environmental safety influence coefficient table includes information such as sea state influence coefficient, visibility influence coefficient, wind force influence coefficient, and comprehensive safety coefficient.

[0132] Obtaining boat performance parameters includes: obtaining the current power state, rudder effectiveness, and propeller operating state of the boat. The power state includes engine speed, power, etc., the rudder effectiveness includes rudder angle, rudder speed, etc., and the propeller operating state includes pitch, speed, etc. These parameters are obtained through onboard sensors, and the data update frequency is 1 second. Consider the impact of the current load condition on the maneuverability of the boat. The load of the boat affects the draft of the boat, thereby affecting the navigation performance of the boat. The load information can be obtained through onboard load sensors or manually input. Combine the environmental safety impact factor to calculate the actual available maneuverability parameters. The actual available maneuverability parameters include: maximum speed, minimum turning radius, maximum acceleration, maximum deceleration, etc. According to the environmental safety impact factor table, correct the boat performance parameters. For example, in rough sea conditions, the maximum speed needs to be reduced and the minimum turning radius needs to be increased. Evaluate the minimum reaction time and operating space required for emergency avoidance. Based on the current speed, heading, and environmental safety factor of the boat, as well as the ship maneuverability constraint table, evaluate the minimum reaction time and operating space required for emergency avoidance. Generate a comprehensive evaluation result reflecting the actual maneuvering ability of the boat under current conditions. The comprehensive evaluation result includes information such as maximum speed, minimum turning radius, maximum acceleration, maximum deceleration, minimum reaction time required for emergency avoidance, and operating space. For example, the evaluation result can be expressed as: maximum speed 15 knots, minimum turning radius 150 meters, emergency avoidance reaction time 5 seconds, and operating space required for emergency avoidance 200 meters. Generate real-time maneuverability evaluation indicators. The real-time maneuverability evaluation indicators include information such as maximum speed, minimum turning radius, maximum acceleration, maximum deceleration, minimum reaction time required for emergency avoidance, and operating space.

[0133] Early warning distance calculation includes: obtaining the basic braking distance at the current speed according to the real-time maneuverability evaluation index. The basic braking distance is obtained from the speed braking distance reference table (sub-step 4.4.1). Adjust the safety factor according to the sea state level. The sea state level is obtained from the environmental safety impact factor table. For example, in calm sea conditions, the safety factor is 1.0; in slightly rough sea conditions, the safety factor is 1.2; in moderate sea conditions, the safety factor is 1.5; in relatively rough sea conditions, the safety factor is 2.0; in severe sea conditions, the safety factor is 2.5. Calculate the distances for each early warning level. The early warning levels include: Caution, Warning, Danger, Emergency. The early warning distance for the Caution level = basic braking distance × safety factor × 3. The early warning distance for the Warning level = basic braking distance × safety factor × 2. The early warning distance for the Danger level = basic braking distance × safety factor × 1.5. The early warning distance for the Emergency level = basic braking distance × safety factor × 1. Extract the current risk type from the navigation deviation status evaluation report. The risk types include: approaching head-on, crossing, overtaking, etc. Adjust the early warning distance according to the risk type. For the approaching head-on risk type, increase the early warning distances for all levels by 20%. For the crossing risk type, keep the early warning distance unchanged. For the overtaking risk type, reduce the early warning distances for all levels by 10%. If the navigation deviation is greater than the preset threshold, an additional 15% safety margin is added to all early warning distances. Generate a hierarchical early warning distance table. The hierarchical early warning distance table includes information such as the early warning distance for the Caution level, the early warning distance for the Warning level, the early warning distance for the Danger level, the early warning distance for the Emergency level, etc.

[0134] Early warning and risk avoidance strategy generation includes: comparing the real-time risk situation with the hierarchical early warning distance table to determine the current early warning level. Determine the early warning level according to the relative distance between the own ship and the target. If the relative distance is less than the emergency level early warning distance, compare the real-time risk situation with the hierarchical early warning distance table to determine the current early warning level. Determine the early warning level according to the relative distance between the own ship and the target. If the relative distance is less than the emergency level early warning distance, the early warning level is "emergency". If the relative distance is between the danger level early warning distance and the emergency level early warning distance, the early warning level is "danger". If the relative distance is between the warning level early warning distance and the danger level early warning distance, the early warning level is "warning". If the relative distance is between the attention level early warning distance and the warning level early warning distance, the early warning level is "attention". If the relative distance is greater than the attention level early warning distance, the early warning level is "safe". Determine the collision risk between the targets according to the risk degree assessment matrix. The risk degree assessment matrix comes from step S2. If the collision risk coefficient between two targets is greater than 0.5, it is considered that there is a collision risk. For different early warning levels, formulate corresponding early warning methods. The early warning methods include: sound and light prompts, screen displays, etc. For the "emergency" and "danger" level early warnings, use sound and light prompts and display the detailed information of the target on the screen, including the target position, course, speed, relative distance, predicted closest point of approach distance (DCPA), and time to closest point of approach (TCPA). For the "warning" and "attention" level early warnings, use screen displays and display the brief information of the target on the screen. Based on the current risk type and early warning level, generate targeted risk avoidance suggestions. The risk avoidance suggestions are determined according to the early warning level and risk type. For example, for the head-on risk at the "emergency" level, the risk avoidance suggestion is "immediately turn right and decelerate"; for the crossing risk at the "danger" level, the risk avoidance suggestion is "carefully adjust the course to avoid crossing the target's course"; for the overtaking risk at the "warning" level, the risk avoidance suggestion is "maintain the course and speed and pay attention to the target's dynamics". Consider the current state of the boat and environmental conditions to optimize the execution order and timing of the risk avoidance operations. The execution order of the risk avoidance operations needs to consider the maneuverability of the boat. For example, first adjust the course and then adjust the speed. The timing of the risk avoidance operations needs to consider the motion state and predicted trajectory of the target. For example, take risk avoidance measures before the target is about to enter the collision area. Form a complete early warning strategy and risk avoidance suggestions and set the early warning priority. The early warning priority is determined according to the early warning level and risk type. For example, the risk avoidance suggestions at the "emergency" level have the highest priority and need to be executed immediately. The risk avoidance suggestions at the "danger" and "warning" levels have lower priorities and can be adjusted according to the situation. The risk avoidance suggestions at the "attention" level have the lowest priority and are only for reference. Generate an intelligent early warning decision table. The intelligent early warning decision table contains information such as the early warning level, early warning method, risk avoidance suggestions, and early warning priority.For example, the intelligent early warning decision table can be expressed as: Early warning level: Urgent; Early warning method: Acoustic and light prompt, screen display; Risk avoidance suggestion: Immediately turn right and decelerate; Early warning priority: High.

[0135] Preferably, the early warning distance calculation in step S44 includes:

[0136] Construct a speed braking distance reference table according to the real-time maneuverability evaluation index;

[0137] Obtain the real-time speed of the boat and determine the current speed-based safety distance according to the speed braking distance reference table;

[0138] Extract the current sea state level from the environmental safety impact coefficient table and adjust the sea state safety coefficient according to the current speed-based safety distance to obtain the sea state-adjusted safety distance;

[0139] Set a basic early warning level distance table according to the sea state-adjusted safety distance;

[0140] Extract the current risk type from the navigation deviation state evaluation report and perform risk type differential adjustment according to the basic early warning level distance table to obtain a hierarchical early warning distance table.

[0141] In the embodiment of the present invention, to construct a reference table for the braking distance at ship speed, first, obtain the real-time maneuverability evaluation index, which comes from step S43 and includes the maximum deceleration of the boat under the current sea conditions. Secondly, establish a relationship model between the ship speed and the braking distance according to the design performance of the boat. This model is based on braking physics and takes into account the differences in braking performance of the boat at different ship speeds. The model formula is: braking distance = 0.5×ship speed^2 / maximum deceleration. Wherein, the unit of ship speed is meters per second, the unit of maximum deceleration is meters per second^2, and the unit of braking distance is meters. Then, determine the discrete points of ship speed. For example, set the ship speed interval to 2 knots, and generate a series of ship speed points from the lowest ship speed to the highest designed ship speed. For each ship speed point, substitute it into the formula to calculate the corresponding braking distance. For example, when the ship speed of the boat is 10 knots (about 5.14 meters per second) and the maximum deceleration is 0.5 meters per second^2 (from the real-time maneuverability evaluation index), then the braking distance = 0.5×(5.14)^2 / 0.5 = 26.42 meters. Organize the calculation results in tabular form, that is, the reference table for the braking distance at ship speed. This table contains two columns: ship speed (knots) and braking distance (meters). For example: 10 knots, 26.42 meters; 12 knots, 38.05 meters; 14 knots, 52.99 meters; 16 knots, 71.68 meters. Obtain the real-time ship speed of the boat, and this ship speed data is obtained from step S41. Determine the basic safety distance at the current ship speed according to the reference table for the braking distance at ship speed by using the method of looking up the table. First, determine the real-time ship speed of the boat. Then, find the ship speed value in the reference table for the braking distance at ship speed that is closest to the real-time ship speed. If the real-time ship speed is exactly the same as the ship speed value in the reference table, directly extract the corresponding braking distance as the basic safety distance. If the real-time ship speed is between two ship speed values in the reference table, use the linear interpolation method to calculate the basic safety distance. The linear interpolation formula is: basic safety distance = braking distance1 + (real-time ship speed - ship speed1)×(braking distance2 - braking distance1) / (ship speed2 - ship speed1). Wherein, braking distance1 and braking distance2 are the braking distances corresponding to two adjacent ship speeds in the reference table, and ship speed1 and ship speed2 are two adjacent ship speeds in the reference table. For example, if the real-time ship speed is 13 knots, in the reference table for the braking distance at ship speed, the braking distance corresponding to 12 knots is 38.05 meters, and the braking distance corresponding to 14 knots is 52.99 meters, then the basic safety distance = 38.05 + (13 - 12)×(52.99 - 38.05) / (14 - 12) = 45.52 meters. Extract the current sea condition level from the environmental safety impact coefficient table, and this table comes from step S42. Extract the current sea condition level, for example, the sea condition level is medium sea condition. Adjust the sea condition safety coefficient according to the basic safety distance at the current ship speed by using multiplication operation. First, determine the sea condition safety coefficient. The sea condition safety coefficient is set according to the sea condition level, and the sea condition safety coefficient is positively correlated with the sea condition level. Set the sea condition safety coefficient table, and this table contains two columns: sea condition level and sea condition safety coefficient.For example: calm sea conditions, 1.0; slight sea conditions, 1.2; moderate sea conditions, 1.5; relatively large sea conditions, 2.0; severe sea conditions, 3.0. According to the current sea condition level, extract the corresponding sea condition safety factor from the sea condition safety factor table. For example, the sea condition safety factor corresponding to moderate sea conditions is 1.5. Then, calculate the sea condition adjusted safety distance. The calculation formula is: sea condition adjusted safety distance = basic safety distance × sea condition safety factor. For example, if the basic safety distance is 45.52 meters and the sea condition safety factor is 1.5, then the sea condition adjusted safety distance = 45.52 × 1.5 = 68.28 meters. Set the basic warning level distance table, which defines the distance thresholds for different warning levels. The basic warning level distance table contains two columns: warning level and distance threshold (meters). The warning levels are divided into three grades: first-level warning, second-level warning, and third-level warning. The warning distance is positively correlated with the sea condition adjusted safety distance. The warning distance is set according to experience. For example: first-level warning distance = sea condition adjusted safety distance × 2; second-level warning distance = sea condition adjusted safety distance × 1.5; third-level warning distance = sea condition adjusted safety distance. For example, if the sea condition adjusted safety distance is 68.28 meters, then: first-level warning distance = 68.28 × 2 = 136.56 meters; second-level warning distance = 68.28 × 1.5 = 102.42 meters; third-level warning distance = 68.28 meters. Then the basic warning level distance table is: first-level warning, 136.56 meters; second-level warning, 102.42 meters; third-level warning, 68.28 meters. Extract the current risk type from the navigation deviation status assessment report, which comes from step S41. The risk types include: position deviation, course deviation, speed deviation, collision risk, grounding risk, etc. Make differential adjustments to the warning distance threshold according to the basic warning level distance table for different risk types. Set the risk type adjustment coefficient table. This table contains two columns: risk type and adjustment coefficient. For example: position deviation, 1.0; course deviation, 1.2; speed deviation, 1.1; collision risk, 1.5; grounding risk, 2.0. According to the current risk type, extract the corresponding adjustment coefficient from the risk type adjustment coefficient table. For example, if the current risk type is collision risk, the adjustment coefficient is 1.5. Adjust the warning distance threshold. The adjustment formula is: adjusted warning distance = basic warning level distance × adjustment coefficient. Generate a graded warning distance table according to the adjusted warning distance. For example, the basic warning level distance table is: first-level warning 136.56 meters, second-level warning 102.42 meters, third-level warning 68.28 meters; the current risk type is collision risk, and the adjustment coefficient is 1.5. Then: first-level warning distance = 136.56 × 1.5 = 204.84 meters; second-level warning distance = 102.42 × 1.5 = 153.63 meters; third-level warning distance = 68.28 × 1.5 = 102.42 meters. The graded warning distance table is: first-level warning, 204.84 meters; second-level warning, 153.63 meters; third-level warning, 102.42 meters.

[0142] Therefore, from any perspective, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Accordingly, all changes that fall within the meaning and scope of the equivalent elements of the application documents are intended to be encompassed within the present invention.

[0143] The above description is only a specific implementation manner of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features invented herein.

Claims

1. A method for intelligent positioning, navigation and warning of a boat, characterized in that, It includes the following steps: Step S1: Collect and evaluate the sea state perception and positioning signals based on the multi-source sensor synchronization dataset to obtain the sensor reliability score table; Analyze the target motion situation according to the sensor reliability score table to obtain the fused situation vector diagram; Step S2: Judge the target motion state according to the fused situation vector diagram to obtain the target motion state determination result; Evaluate the multi-target collision risk according to the target motion state determination result to obtain the risk degree evaluation matrix. The multi-target collision risk evaluation in Step S2 includes: Calculate and analyze the trajectory correlation according to the target motion state determination result, including: calculating the target relative motion parameter table according to the target motion state determination result and the target trajectory feature sequence; calculating the trajectory correlation coefficient matrix according to the target relative motion parameter table; identifying the interactive behavior pattern according to the trajectory correlation coefficient matrix and the target relative motion parameter table to obtain the target interactive behavior feature table; constructing the target relationship structure diagram according to the target interactive behavior feature table; screening the key influencing targets according to the target relationship structure diagram and the target interactive behavior feature table to obtain the target correlation network diagram; Perform multi-target joint trajectory prediction according to the target motion state determination result and the target correlation network diagram to obtain the multi-target prediction trajectory set; Evaluate the collision risk of the multi-target prediction trajectory set to obtain the collision risk coefficient; generate the risk degree evaluation matrix according to the collision risk coefficient; Step S3: Constrain the navigation operation performance according to the risk degree evaluation matrix to obtain the route planning operation constraint rule set; search for the route trajectory according to the route planning operation constraint rule set to obtain the navigation trajectory plan; Step S4: Monitor the navigation deviation in real time according to the navigation trajectory plan to obtain the navigation deviation state evaluation report; calculate the warning distance according to the navigation deviation state evaluation report to obtain the hierarchical warning distance table; generate the warning and avoidance strategy according to the hierarchical warning distance table and the risk degree evaluation matrix to obtain the intelligent warning decision table.

2. The boat intelligent positioning, navigation and warning method according to claim 1, characterized in that The sea state perception and positioning signal evaluation in Step S1 includes: Collect the multi-source sensor synchronization dataset through the multi-source sensor array, where the multi-source sensor synchronization dataset includes radar echo data, AIS received data stream, satellite receiver data, wave height sensor data, and hull attitude sensor data; Evaluate the positioning sensor signals of the radar echo data, AIS received data stream, and satellite receiver data to obtain the initial score of the positioning sensor; Measure the sea state complexity of the wave height sensor data and the hull attitude sensor data to obtain the sea state level evaluation result; Obtain the basic attenuation coefficient of the sea state influence according to the sea state level evaluation result; Correct the radar echo data, AIS received data stream, and satellite receiver data according to the basic attenuation coefficient of the sea state influence to obtain the corrected value of the positioning sensor; Scan the radar plane and calculate the offset angle caused by the hull sway, denoted as the radar scan offset angle, where the radar scan offset angle includes the root mean square of the roll angle and the root mean square of the pitch angle; Compensate the corrected value of the positioning sensor according to the radar scan offset angle for the hull attitude influence to obtain the attitude compensation value of the positioning sensor; Conduct a comprehensive reliability assessment based on the attitude compensation value of the positioning sensor and the initial score of the positioning sensor to obtain the sensor reliability score table.

3. The boat intelligent positioning, navigation and warning method according to claim 1, characterized in that, The target motion situation analysis in step S1 includes: Adjust the adaptive Kalman filter parameters for the multi-source sensor synchronous data set according to the sensor reliability score table to obtain the adaptive filter parameter set; Perform multi-source data fusion and trajectory smoothing based on the adaptive filter parameter set and the multi-source sensor synchronous data set to obtain the target state fusion data; Generate and update the situation vector diagram based on the target state fusion data to obtain the fused situation vector diagram.

4. The boat intelligent positioning, navigation and warning method according to claim 1, characterized in that The target motion state judgment in step S2 includes: Extract the target historical trajectory of the fused situation vector diagram; segment the target historical trajectory to obtain the target trajectory feature sequence; Extract the heading change feature set and speed change feature set of the target trajectory feature sequence; Use the heading change feature set to determine the heading motion mode to obtain the heading mode determination result; Use the speed change feature set to judge the speed motion mode to obtain the speed mode determination result; Identify the special motion mode according to the heading mode determination result and the speed mode determination result to obtain the special mode recognition result; Judge the comprehensive motion state according to the heading mode determination result, the speed mode determination result and the special mode recognition result to obtain the basic motion state determination result; Obtain the target historical state sequence, and infer the motion intention according to the basic motion state determination result to obtain the motion intention inference result; Generate the target motion state determination result according to the basic motion state determination result and the motion intention inference result.

5. The boat intelligent positioning, navigation and warning method according to claim 1, wherein Step S3 includes the following steps: Step S31: Obtain the current sea area environment information; generate the sea area environment constraint diagram according to the sea area environment information; Step S32: Construct the navigation risk cost map according to the sea area environment constraint diagram and the risk degree evaluation matrix; Step S33: Perform ship maneuverability constraints according to the navigation risk cost map to obtain the route planning maneuvering constraint rule set; Step S34: Conduct direct search and screening of the route according to the route planning maneuvering constraint rule set and the navigation risk cost map to obtain the route candidate plan set; Step S35: Generate the navigation instruction sequence for the route candidate plan set to obtain the navigation trajectory plan.

6. The boat intelligent positioning, navigation and warning method according to claim 5, characterized in that, The ship maneuverability constraints in step S33 include: Collect and construct the steering ability query table according to the measured data of the boat's steering ability; Collect and construct the acceleration and deceleration ability query table according to the acceleration and deceleration performance test data set of the boat; Determine the sea condition influence coefficient according to the steering ability query table and the acceleration and deceleration ability query table to obtain the maneuverability sea condition correction coefficient table; Determine the safe speed limit table and economic speed interval table of the boat; Generate the ship maneuverability constraint table according to the steering ability query table, the acceleration and deceleration ability query table, the maneuverability sea condition correction coefficient table, the safe speed limit table and the economic speed interval table; Perform hard constraint conversion of the route planning for the ship maneuverability constraint table and the navigation risk cost map to obtain the route planning maneuvering constraint rule set.

7. The boat intelligent positioning navigation warning method according to claim 1, wherein Step S4 includes the following steps: Step S41: Obtain the real-time navigation data of the boat, and conduct real-time monitoring of the navigation deviation according to the navigation trajectory plan to obtain the navigation deviation state assessment report; Step S42: Obtain real-time environmental data, and quantify the safety impact of environmental factors according to the navigation deviation status assessment report to obtain an environmental safety impact coefficient table; Step S43: Obtain the boat performance parameters, and conduct real-time assessment of the boat's maneuverability according to the environmental safety impact coefficient table to obtain real-time maneuverability assessment indicators; Step S44: Calculate the warning distance according to the real-time maneuverability assessment indicators and the navigation deviation status assessment report to obtain a hierarchical warning distance table; Step S45: Generate an early warning and avoidance strategy according to the hierarchical warning distance table and the risk degree assessment matrix to obtain an intelligent early warning decision table.

8. The boat intelligent positioning, navigation and warning method according to claim 7, characterized in that, The warning distance calculation in Step S44 includes: Construct a reference table for the speed braking distance based on the real-time maneuverability assessment indicators; Obtain the real-time speed of the boat, and determine the basic safety distance for the current speed according to the reference table for the speed braking distance; Extract the current sea state level from the environmental safety impact coefficient table, and adjust the sea state safety coefficient according to the basic safety distance for the current speed to obtain the sea state adjusted safety distance; Set a reference table for the basic warning level distance according to the sea state adjusted safety distance; Extract the current risk type from the navigation deviation status assessment report, and conduct differential adjustment according to the reference table for the basic warning level distance for the risk type to obtain a hierarchical warning distance table.

9. A boat intelligent positioning, navigation and warning system, characterized in that, For implementing the boat intelligent positioning, navigation and warning method as described in Claim 1, the boat intelligent positioning, navigation and warning system includes: A signal fusion module, configured to collect and evaluate the sea state perception and positioning signals according to the multi-source sensor synchronous data set to obtain a sensor reliability score table; analyze the target motion situation according to the sensor reliability score table to obtain a fused situation vector map; A risk assessment module, configured to judge the target motion state according to the fused situation vector map to obtain a target motion state determination result; conduct a multi-target collision risk assessment according to the target motion state determination result to obtain a risk degree assessment matrix; A route optimization module, configured to impose constraints on the navigation maneuverability according to the risk degree assessment matrix to obtain a set of route planning maneuvering constraint rules; search for the route trajectory according to the set of route planning maneuvering constraint rules to obtain a navigation trajectory plan; An early warning decision module, configured to conduct real-time monitoring of the navigation deviation according to the navigation trajectory plan to obtain a navigation deviation status assessment report; calculate the warning distance according to the navigation deviation status assessment report to obtain a hierarchical warning distance table; generate an early warning and avoidance strategy according to the hierarchical warning distance table and the risk degree assessment matrix to obtain an intelligent early warning decision table.

Images