A sea-going vessel safety situation awareness system

Abstract

The application discloses a kind of sea-going ship safety situation awareness system, belong to data analysis technical field;For solving the technical problem that existing scheme cannot carry out multi-dimensional data analysis and carry out timely and reliable early warning prompt in view of complex sea conditions, leading to weak complex sea condition response capability;Through the whole process of biological anchoring abnormality suppression, data fusion, dynamic feature extraction, it can break the limitation of traditional ship perception relying on physical sensor, realize the high-precision perception of environmental risk and ship state under complex sea conditions, relative to traditional method, it can identify abnormal sea conditions in advance, and can effectively improve the robustness of data, the detection rate of high-impact events with small probability;Through micro physical mechanism and macro data feature deep coupling, cross-scale modeling is realized;The multi-dimensional risk assessment matrix constructed can not only output risk value, but also provide risk propagation path, which can effectively improve the risk assessment accuracy of ship in complex sea conditions.

Description

Technical Field

[0001] This invention relates to the field of data analysis technology, specifically to a safety situation awareness system for ships at sea. Background Technology

[0002] Safety situation awareness of ships at sea refers to the real-time and comprehensive monitoring, analysis, assessment and early warning of the safety status of ships at sea and their marine environment by using various technical means and information sources, thereby forming an overall understanding and judgment of current and future safety risks. Its core objective is to improve the foresight, initiative and response capabilities to maritime safety risks, and to ensure the safety of ships, crew, cargo and marine environment.

[0003] Existing technical solutions, when implemented, cannot perform multi-dimensional data analysis and provide timely and reliable early warnings for complex sea conditions. They also have a low detection rate for identifying low-probability, high-impact events in advance, resulting in a weak ability to cope with complex sea conditions. Summary of the Invention

[0004] The purpose of this invention is to provide a safety situation awareness system for ships at sea, which solves the technical problem that existing solutions cannot perform multi-dimensional data analysis and provide timely and reliable early warnings for complex sea conditions, resulting in weak response capabilities to complex sea conditions.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A safety situation awareness system for ships at sea, comprising:

[0007] Multi-source heterogeneous data dynamic perception enhancement module: Real-time collection of multi-source heterogeneous data of the ship's internal and external environment, and extraction of time-varying sea state feature vectors and ship dynamic state parameters through outlier dynamic suppression and adaptive weighted fusion;

[0008] Spatiotemporal correlation dynamic risk assessment and prompting module: Based on the acquired time-varying sea state feature vector and ship dynamic state parameters, a spatiotemporal correlation risk model integrating physical constraints and data-driven approaches is constructed, and a multidimensional risk assessment matrix is ​​output in real time. Based on the multidimensional risk assessment matrix, multidimensional risk dynamic prompts for collision risk, capsizing risk, and equipment failure risk are provided.

[0009] Preferably, multi-source heterogeneous data of the ship's internal and external environment are collected synchronously through the ship's integrated sensing system, including marine environmental data, ship body data and bioacoustic data; the collected multi-source heterogeneous data are preprocessed to construct a unified spatiotemporal dataset.

[0010] Preferably, bio-based weighted correction is used to correct eligible sensor data. : ;in, The corrected sensor data; It is the first weight; These are estimates of environmental parameters based on biological benchmarks.

[0011] Preferably, the probability of anomalous wave generation is extracted. Ocean current shear gradient Biological avoidance behavior index hull resonant frequency shift The four core precursor features are combined to obtain the time-varying sea state feature vector. ;

[0012] Obtain the roll acceleration calculated from the hull kinematic model. Spatiotemporal distribution of draft , Propulsion power loss rate By combining these parameters, the ship's dynamic state parameters can be obtained. .

[0013] Preferably, when constructing a spatiotemporal correlation risk model that integrates physical constraints and data-driven approaches, the critical threshold for capsizing risk is calculated based on the ship's motion equations:

[0014] Where GM is the initial stability height; d is the initial value of the roll angle; d is the draft.

[0015] The boundary of the ship's domain is calculated based on AIS trajectory data, and the collision hazard zone is determined by the nearest encounter distance and the nearest encounter time.

[0016] Preferably, the collision risk is obtained by integrating the physical constraints of the nearest encounter distance and the nearest encounter time with the trajectory prediction error, and then normalizing using the Logistic function. :

[0017] Where k and m are different sensitivity coefficients; DCPA and TCPA are the nearest encounter distance and the nearest encounter time, respectively. These are all threshold values ​​for the corresponding data items;

[0018] Combined with roll angle acceleration and the critical threshold of overturning risk Calculate and obtain overturning risk :

[0019] ;in, The critical roll acceleration, g is the acceleration due to gravity;

[0020] Based on propulsion power loss rate Calculate and obtain equipment failure risk :

[0021] ;in, For the Sigmoid function; b is the second weight, obtained by training the XGBoost model on historical fault data; b is the bias. For time step index.

[0022] Preferably, based on collision risk Risk of capsizing Equipment failure risk Constructing a multidimensional risk assessment matrix ; where the off-diagonal elements represent the propagation probability between risk factors.

[0023] Preferably, the risk value and risk propagation path of the ship are obtained based on the multidimensional risk assessment matrix, and multidimensional risk dynamic prompts for collision risk, capsizing risk and equipment failure risk are given based on the risk value and risk propagation path.

[0024] Preferably, the elements of the multidimensional risk assessment matrix are updated based on the newly acquired time-varying sea state feature vector S and the ship dynamic state parameter P, and high-frequency fluctuations are suppressed by exponential moving average.

[0025] Compared to existing solutions, the beneficial effects achieved by this invention are:

[0026] This invention, through a complete process of biological anchoring anomaly suppression, data fusion, and dynamic feature extraction, can break through the limitations of traditional ship perception relying on physical sensors, and achieve high-precision perception of environmental risks and ship status under complex sea conditions. Compared with traditional methods, it can identify abnormal sea conditions in advance and effectively improve the robustness of data and the detection rate of low-probability, high-impact events.

[0027] This invention achieves cross-scale modeling by deeply coupling microscopic physical mechanisms with macroscopic data features; the constructed multidimensional risk assessment matrix can not only output risk values, but also provide risk propagation paths, providing a basis for targeted intervention in the generation of subsequent response strategies. Through the triple mechanism of physical constraints, data-driven approach, and causal reasoning, it can effectively improve the accuracy of risk assessment for ships in complex sea conditions. Attached Figure Description

[0028] The invention will now be further described with reference to the accompanying drawings.

[0029] Figure 1 This is a flowchart illustrating the operation of a maritime vessel safety situation awareness system according to the present invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] like Figure 1 As shown, the present invention is a safety situation awareness system for ships at sea, comprising:

[0032] Multi-source heterogeneous data dynamic perception enhancement module: Real-time acquisition of multi-source heterogeneous data of the ship's internal and external environment; extraction of time-varying sea state feature vectors and ship dynamic state parameters through dynamic outlier suppression and adaptive weighted fusion; specific steps include:

[0033] The ship's integrated sensing system synchronously collects multi-source heterogeneous data on the ship's internal and external environment, including marine environmental data, ship body data, and bioacoustic data; specifically:

[0034] Marine environmental data: Shipborne weather radar collects wave height and wave direction data, Doppler current meter measures ocean current velocity profile, and satellite remote sensing receives sea surface temperature gradient data;

[0035] Ship body data: Fiber optic gyroscope inertial measurement units collect the ship's roll angle φ and pitch angle θ, and shaft torque sensors record changes in propulsion system load;

[0036] Bioacoustic data: A towed single-beam sonar collects scattering signals from planktonic communities, extracts community density fluctuation characteristics through time-frequency analysis of the signals, and outputs a spectrum every 2 seconds;

[0037] Time alignment is achieved using GPS timing, and spatial registration is completed through a coordinate transformation matrix from the ship's coordinate system to the geodetic coordinate system, thus constructing a unified spatiotemporal dataset. ;in, These represent marine environmental data, ship body data, and bioacoustic data, respectively; t is the time variable.

[0038] The LSTM network is used to predict the trajectory sequence of planktonic organisms and output the community movement trajectory for the next 2 seconds. Calculate the measured trajectory Deviation from predicted trajectory ,Will As a biological benchmark for environmental stability;

[0039] Among them, LSTM (Long Short-Term Memory) network is a special type of recurrent neural network (RNN) specifically designed to solve the problem that standard RNNs have difficulty learning long-term dependencies when processing long sequence data. By introducing gating mechanisms (forget gate, input gate, output gate) and cell states, it effectively solves the long-term dependency problem of standard RNNs, and can selectively remember or forget information, thus performing well when processing long sequence data. It is one of the important tools for processing sequence tasks in deep learning.

[0040] In this embodiment of the invention, the movement trajectory of the planktonic community is used as the environmental benchmark, and the position of the biological community in the next 2 seconds is predicted by an LSTM network to construct a planktonic trajectory sequence.

[0041] Sensor data ,like , For dynamic thresholds, , As the baseline value, Sea state classification is based on the International Sea State Classification (ISC) standards (levels 1-5); biological basis weighted correction is used.

[0042] ;in, As the first weight, The weight of the measured value is dynamically reduced as the deviation increases. These are estimates of environmental parameters based on biological benchmarks, such as wave height derived from plankton density.

[0043] It should be noted that by utilizing the sensitivity of plankton to environmental changes and dynamically filtering sensor noise and fault data, the problem of physical sensors being susceptible to interference under high sea states can be solved. This effectively improves the accuracy of outlier identification, the data signal-to-noise ratio, and reduces the false alarm rate of sensor faults, providing highly reliable input data for subsequent fusion.

[0044] In adaptive weighted fusion, a confidence factor is assigned to each data source. By fusing historical accuracy data from sensors And the consistency coefficient of biological signals Implementation:

[0045] Where i represents different data sources, j is the summation index, indicating that all n data sources carried by the ship are traversed; when the wave height H > 5m, the weight of the bioacoustic data is automatically increased, i.e. ×1.5, to compensate for radar measurement errors under high sea states;

[0046] Extracting the probability of anomalous wave generation Ocean current shear gradient Biological avoidance behavior index hull resonant frequency shift The four core precursor features are combined to obtain the time-varying sea state feature vector. ;

[0047] Among them, the probability of abnormal wave generation By fitting the wave height distribution using a Gaussian mixture model, the probability of the wave height exceeding 3 times the root mean square is calculated as follows:

[0048] ;in, The weight of the k-th Gaussian component; k is the index of the Gaussian component, k=1,2,3; The cumulative function is the standard normal distribution; H is the anomaly threshold. Let be the mean of the k-th Gaussian component; Let $\frac{k}{k}$ be the standard deviation of the $k$-th Gaussian component.

[0049] Ocean current shear gradient Calculation of vertical gradient based on ADCP velocity profile Combined with seawater density and drag coefficient The calculation yielded:

[0050] ;in, The absolute value of the ocean current velocity; vertical gradient. denoted by z, the rate of change of ocean current velocity in the vertical direction, i.e., the change in velocity per unit depth; z represents depth.

[0051] Biological Avoidance Behavior Index The error was calculated using an exponential function of the plankton trajectory prediction error:

[0052] Where N is the prediction window size, which takes the value 20; exp() is the exponential function; and t is the time index. This is a predicted value for planktonic trajectory; These are measured values ​​of planktonic organism tracks; Mean absolute prediction error;

[0053] Hull resonant frequency shift The measured dominant frequency was calculated by analyzing the ship's vibration signal using FFT. With design value The deviation is obtained as follows:

[0054] ;

[0055] Obtain the roll acceleration calculated from the hull kinematic model. Spatiotemporal distribution of draft , Propulsion power loss rate By combining these parameters, the ship's dynamic state parameters can be obtained. ;

[0056] Among them, roll angular acceleration The roll angle data collected by the fiber optic gyroscope inertial measurement unit is obtained by second derivative:

[0057] ;in, The sampling interval is 0.01.

[0058] Spatiotemporal distribution of draft The sensor array data was obtained by bicubic B-spline interpolation.

[0059] ;in, The control vertex coefficients for the B-spline; All are B-spline basis functions; These are normalized parameter coordinates; All are indices of the B-spline basis functions, with values ​​ranging from 0 to 3. a corresponds to the basis function index along the ship's length direction, which is the u-parameter direction; b corresponds to the basis function index along the ship's width direction, which is the v-parameter direction.

[0060] Driven power loss rate Calculated based on data from shaft torque sensors:

[0061] ;in, n is the measured torque of the shaft system; n is the measured speed of the propulsion shaft. To increase the rated power of the system.

[0062] It should be noted that by transforming the fused data into low-dimensional feature vectors that characterize environmental risks and ship status, reliable data support can be provided for subsequent causal risk chain mining.

[0063] In this embodiment of the invention, the entire process of biological anchoring anomaly suppression, data fusion, and dynamic feature extraction can break the limitations of traditional ship perception relying on physical sensors, and achieve high-precision perception of environmental risks and ship status under complex sea conditions. Compared with traditional methods, it can identify abnormal sea conditions in advance and effectively improve the robustness of data and the detection rate of low-probability, high-impact events.

[0064] Spatiotemporal correlation dynamic risk assessment and alert module: Based on the acquired time-varying sea state feature vector and ship dynamic state parameters, a spatiotemporal correlation risk model integrating physical constraints and data-driven approaches is constructed. A multi-dimensional risk assessment matrix is ​​output in real time, and dynamic multi-dimensional risk alerts for collision risk, capsizing risk, and equipment failure risk are provided based on the multi-dimensional risk assessment matrix. Specific steps include:

[0065] When constructing a spatiotemporal correlation risk model that integrates physical constraints and data-driven approaches, the critical threshold for capsizing risk is calculated based on the ship's motion equations:

[0066] Where GM is the initial stability height; d is the initial value of the roll angle; d is the draft.

[0067] The boundary of the vessel's domain is calculated based on AIS trajectory data, and the collision hazard zone is determined using the nearest encounter distance and the nearest encounter time.

[0068] If the nearest encounter distance is less than 2L and the nearest encounter time is less than 300s, a collision risk warning will be triggered; L is the length of the hull.

[0069] The acquired time-varying sea state feature vector S and ship dynamic state parameters P are input into the depth spatiotemporal network (DSTN) to achieve data-driven feature enhancement.

[0070] The deep spatiotemporal network includes a temporal attention module and a spatial interaction module.

[0071] Time Attention Module: Assigns dynamic weights to the time-varying sea state feature vector S and ship dynamic state parameter P over the past 5 minutes, for example, recent data has higher weights;

[0072] Spatial Interaction Module: Employs a graph attention mechanism to capture the trajectory correlations of surrounding vessels. Node features include speed, heading angle, and biological avoidance behavior index. ;

[0073] It should be noted that by integrating physical laws and implicit data features—physical laws such as ship stability and collision avoidance rules, and implicit data features such as biological precursors and sudden changes in sea state—we can provide reliable mechanistic data-driven input features for risk assessment. Compared with traditional pure data-driven models, physical constraints can effectively reduce the error in risk factor extraction.

[0074] When outputting the multidimensional risk assessment matrix R, the collision risk is obtained by integrating the nearest encounter distance, the nearest encounter time physical constraint, and the trajectory prediction error, and then normalizing using the Logistic function. :

[0075] Where k and m are different sensitivity coefficients, taking values ​​of 0.5 and 0.02 respectively; DCPA and TCPA are the nearest encounter distance and the nearest encounter time, respectively. These are the threshold values ​​for the corresponding data items, with values ​​of 2L and 300s respectively;

[0076] Combined with roll angle acceleration and the critical threshold of overturning risk Calculate and obtain overturning risk :

[0077] ;in, The critical roll acceleration, g is the acceleration due to gravity;

[0078] Based on propulsion power loss rate Calculate and obtain equipment failure risk :

[0079] ;in, For the Sigmoid function; b is the second weight, obtained by training the XGBoost model on historical fault data; b is the bias. Indexed by time step;

[0080] Based on collision risk Risk of capsizing Equipment failure risk Constructing a multidimensional risk assessment matrix ; where the off-diagonal elements represent the propagation probability between risk factors;

[0081] The risk value and risk propagation path of the ship are obtained based on the multidimensional risk assessment matrix, and the multidimensional risk dynamic prompts of collision risk, capsizing risk and equipment failure risk are provided based on the risk value and risk propagation path.

[0082] Among them, the risk value, represented by the main diagonal elements, directly indicates the quantified value of each type of risk; the risk propagation path, for example... This indicates the probability of a collision triggering a rollover, forming a path. ;

[0083] Every 500ms, the multidimensional risk assessment matrix elements are updated based on the newly acquired time-varying sea state feature vector S and ship dynamic state parameters P, and high-frequency fluctuations are suppressed by exponential moving average:

[0084] ;in, This is the smoothing coefficient, with a value of 0.2. This is the smoothed risk assessment matrix from the previous time step t−1; This is the original risk assessment matrix at the current time t.

[0085] It should be noted that by integrating physical constraints, data-driven predictions, and causal propagation laws into a structured risk matrix, a comprehensive assessment of collisions, capsizing, and equipment failures can be achieved. In addition, multidimensional risk coupling assessment can effectively reduce the false alarm rate, and the dynamic matrix update mechanism can reduce the response delay to sudden sea conditions.

[0086] In this embodiment of the invention, cross-scale modeling is achieved through deep coupling of microscopic physical mechanisms and macroscopic data features; the constructed multidimensional risk assessment matrix can not only output risk values, but also provide risk propagation paths, providing a basis for targeted intervention in the generation of subsequent response strategies. Through the triple mechanism of physical constraints, data-driven approach, and causal reasoning, the accuracy of risk assessment for ships in complex sea conditions can be effectively improved.

[0087] Adaptive Response Strategy Generation Module: Based on the acquired multidimensional risk assessment matrix R and combined with ship maneuvering performance parameters, this module dynamically generates a set of coordinated response strategies, including route correction, speed adjustment, and emergency operations. Specific steps include:

[0088] Obtain the ship's static maneuvering performance parameter database and combine it with the ship's dynamic state parameters. Real-time correction of the upper speed limit and minimum turning radius in the ship's static maneuvering performance parameter database;

[0089] Among them, when When the g value is greater than 0.1g, g represents the acceleration due to gravity. The maximum speed can be reduced using the following formula:

[0090] ;in, The roll effect coefficient is set to 0.5. The preset base speed limit; To adjust the speed limit;

[0091] Based on the spatiotemporal distribution of draft d(x,y), the minimum safe turning radius in shallow water is calculated using bicubic interpolation:

[0092] ;in, This refers to the real-time minimum draft. For design draft; This is the preset minimum safe turning radius in shallow water areas; To correct the minimum safe turning radius in shallow water areas;

[0093] By sorting and combining the corrected upper speed limit and the corrected minimum safe turning radius in shallow water, the ship handling performance correction parameters are obtained.

[0094] Based on the generated multidimensional risk assessment matrix R, the causal propagation chain between risk factors is extracted, for example: main path 1: → → The corresponding abnormal wave → increased rolling → risk of capsizing;

[0095] Main path 2: → → This corresponds to ocean current shearing → collision risk → equipment overload failure;

[0096] For each transmission path's source risk factors, a targeted intervention strategy is matched from a pre-defined strategy library:

[0097] like A value greater than 0.6 corresponds to a high risk of abnormal waves, triggering a combined strategy of route deviation and speed reduction. The adjustment parameters for route deviation and speed reduction are determined based on the existing strategy, and the specific details are not limited.

[0098] like A value greater than 0.7 indicates a high collision risk; a heading angle correction is generated using AIS trajectory prediction. The expression involved is: ;in, Current speed;

[0099] Atomic strategies that satisfy physical feasibility are generated based on constraints; the types of atomic strategies include speed adjustment and heading correction.

[0100] The constraint is: speed , rudder angle ;

[0101] Speed ​​adjustment: Generate a candidate set of ΔV that satisfies the speed constraint. ;

[0102] Heading correction: Generate a candidate set of Δθ that satisfies the rudder angle constraint. ;

[0103] Based on the adapted ship maneuvering performance correction parameters, detect whether there are conflicting terms in the atomic strategy; for example:

[0104] The speed commands "accelerate to avoid" (ΔV=+2 knots) and "reduce speed" (ΔV=−3 knots) conflict.

[0105] "Turn left" =+15°) and "Right Turn" ( A heading instruction conflict occurred at (-10°);

[0106] Conflicting items are removed, and non-conflicting strategies are retained and combined to obtain a composite strategy set;

[0107] For each strategy in the composite strategy set, calculate its risk reduction rate J1 and execution cost J2 to quantify strategy performance:

[0108] ;

[0109] ;

[0110] in, This is a safety threshold matrix; The penalty for emergency operations is 10; trace is the matrix trace, which is the sum of the elements on the main diagonal, and comprehensively reflects the overall risk reduction ratio. For speed adjustment;

[0111] A Pareto optimal strategy is selected using a multi-objective optimization algorithm to form a cooperative strategy set C.

[0112] Among them, multi-objective optimization algorithms, such as NSGA-III, are advanced evolutionary algorithms for solving multi-objective optimization problems (MOOPs). NSGA-III is an improved version of the NSGA-II algorithm. By introducing a reference point mechanism to replace the traditional crowding degree calculation, it can effectively solve the problem of uneven distribution of solution sets in high-dimensional multi-objective optimization problems. While maintaining the convergence of solutions, it significantly improves the diversity and distribution of the population, making it an important tool for handling complex multi-objective optimization problems. The specific implementation steps of NSGA-III will not be elaborated here, as it is an existing conventional technical solution.

[0113] Pareto optimality is defined as follows: if strategy c satisfies "there is no other strategy c such that J1(c)≥J1(c*) and J2(c)≤J2(c*)", then c* is the Pareto optimal solution, meaning that it is impossible to reduce the cost without reducing the risk reduction rate, or to increase the risk reduction rate without increasing the cost.

[0114] Select the final cooperative strategy set C from the Pareto optimal solution to ensure that the strategies are conflict-free and cover different optimization directions;

[0115] Among them, the principles for strategy selection include the principle of diversity and the principle of no redundancy;

[0116] The principle of diversity: retain strategies with different trade-offs between J1 and J2, such as "high J1-high J2", "medium J1-medium J2", and "low J1-low J2".

[0117] No redundancy principle: Eliminate strategies with highly similar effects, such as those with a risk reduction rate difference of less than 5% between sections ΔV=-2 and ΔV=-1.8; The implementation of the strategy selection principle can be determined based on the application requirements of the actual application scenario.

[0118] It should be noted that by resolving strategy conflicts, a Pareto optimal strategy set that balances risk reduction and execution cost can be generated. By applying Pareto optimal strategies, the strategy conflict rate can be effectively reduced.

[0119] In this embodiment of the invention, by targeting the source of risk propagation, the efficiency of risk reduction under complex sea conditions can be significantly improved; at the same time, by combining the dynamic adaptation strategy of real-time ship maneuvering performance, conflict-free execution is ensured, and the synergistic optimization of ship safety and energy efficiency is achieved.

[0120] In the several embodiments provided by this invention, it should be understood that the disclosed system can be implemented in other ways. For example, the embodiments of the invention described above are merely illustrative; for example, the division of modules is only a logical functional division, and there may be other division methods in actual implementation.

[0121] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0122] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated module can be implemented in hardware or in the form of hardware plus software functional modules.

[0123] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the essential characteristics of the present invention.

[0124] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A safety situation awareness system for ships at sea, characterized in that, include: Multi-source heterogeneous data dynamic perception enhancement module: Real-time collection of multi-source heterogeneous data of the ship's internal and external environment, and extraction of time-varying sea state feature vectors and ship dynamic state parameters through outlier dynamic suppression and adaptive weighted fusion; Among them, multi-source heterogeneous data of the ship's internal and external environment are collected synchronously through the ship's integrated sensing system, including marine environmental data, ship body data and bioacoustic data; the collected multi-source heterogeneous data are preprocessed to construct a unified spatiotemporal dataset; Using the movement trajectory of planktonic communities as an environmental benchmark, the location of the biological community in the next 2 seconds is predicted through an LSTM network, and a planktonic trajectory sequence is constructed. This is then applied to sensor data. ,like , The deviation between the measured trajectory and the predicted trajectory. For dynamic thresholds, a biological benchmark weighted correction is used: ;in, As the first weight, These are estimates of environmental parameters based on biological benchmarks. Spatiotemporal correlation dynamic risk assessment and prompting module: Based on the acquired time-varying sea state feature vector and ship dynamic state parameters, a spatiotemporal correlation risk model integrating physical constraints and data-driven approaches is constructed, and a multidimensional risk assessment matrix is ​​output in real time. Based on the multidimensional risk assessment matrix, multidimensional risk dynamic prompts for collision risk, capsizing risk, and equipment failure risk are provided. In constructing a spatiotemporal correlation risk model that integrates physical constraints and data-driven approaches, the critical threshold for capsizing risk is calculated based on the ship's motion equations. Where GM is the initial stability height; d is the initial value of the roll angle; d is the draft. The boundary of the ship's domain is calculated based on AIS trajectory data, and the collision hazard zone is determined by the nearest encounter distance and the nearest encounter time. The acquired time-varying sea state feature vector S and ship dynamic state parameters P are input into a depth-spatiotemporal network to achieve data-driven feature enhancement. The deep spatiotemporal network includes a temporal attention module and a spatial interaction module. Time Attention Module: Assigns dynamic weights to the time-varying sea state feature vector S and the ship dynamic state parameter P over the past 5 minutes; Spatial interaction module: Employs graph attention mechanism to capture the trajectory correlation of surrounding ships; When outputting the multidimensional risk assessment matrix R, the collision risk is obtained by integrating the nearest encounter distance, the nearest encounter time physical constraint, and the trajectory prediction error, and then normalizing using the Logistic function. : Where k and m are different sensitivity coefficients; DCPA and TCPA are the nearest encounter distance and the nearest encounter time, respectively. These are all threshold values ​​for the corresponding data items; Combined with roll angle acceleration and the critical threshold of overturning risk Calculate and obtain overturning risk : ;in, The critical roll acceleration, g is the acceleration due to gravity; Based on propulsion power loss rate Calculate and obtain equipment failure risk : ;in, For the Sigmoid function; b is the second weight, obtained by training the XGBoost model on historical fault data; b is the bias. Indexed by time step; Based on collision risk Risk of capsizing Equipment failure risk Constructing a multidimensional risk assessment matrix ; where the off-diagonal elements represent the propagation probability between risk factors.

2. The maritime vessel safety situation awareness system according to claim 1, characterized in that, When extracting time-varying sea state feature vectors and ship dynamic state parameters, extract the probability of anomalous wave generation. Ocean current shear gradient Biological avoidance behavior index hull resonant frequency shift By combining these features, we obtain the time-varying sea state feature vector. ; Among them, the biological avoidance behavior index The error was calculated using an exponential function of the plankton trajectory prediction error: Where N is the prediction window size; exp() is the exponential function; and t is the time index. This is a predicted value for planktonic trajectory; These are measured values ​​of planktonic organism tracks; Mean absolute prediction error; Obtain the roll acceleration calculated from the hull kinematic model. Spatiotemporal distribution of draft , Propulsion power loss rate By combining these parameters, the ship's dynamic state parameters can be obtained. .

3. The maritime vessel safety situation awareness system according to claim 1, characterized in that, The risk value and risk propagation path of the ship are obtained by using a multi-dimensional risk assessment matrix, and multi-dimensional risk dynamic prompts are made based on the risk value and risk propagation path for collision risk, capsizing risk and equipment failure risk.

4. The maritime vessel safety situation awareness system according to claim 3, characterized in that, The elements of the multidimensional risk assessment matrix are updated based on the newly acquired time-varying sea state feature vector S and ship dynamic state parameters P, and high-frequency fluctuations are suppressed by exponential moving average.

Images