A method and system for intelligent ship situation awareness prediction

Abstract

The present invention discloses a method and system for intelligent ship situational awareness and prediction, relating to the technical field of intelligent ship situational awareness and trajectory prediction. The system uses a timestamp alignment method to synchronize the collected ship's operating environment data in time and space to generate a ship situation dataset. Based on the ship situation dataset, the obstacle's heading angle change rate and Euclidean distance relative to the ship are extracted, and a weighted fusion of the dynamic collision probability and the obstacle's approach time is performed to obtain an obstacle threat index. The obstacle threat index is input into a ship trajectory prediction model to output the minimum encounter distance between the ship and each obstacle. The minimum encounter distance is classified into risk levels according to the International Regulations for Preventing Collisions at Sea, a graded warning signal is output, and a navigation strategy is formulated to implement obstacle avoidance and route adjustment. The present invention effectively solves the technical problems existing in the prior art, such as delayed risk perception, single and crude avoidance actions, and insufficient trajectory prediction accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent ship situation perception and trajectory prediction, and in particular to a method and system for intelligent ship situation perception and prediction. Background Art

[0002] Intelligent ship situational awareness technology primarily serves to perceive and predict dynamic changes in the ship and its surroundings in real time to support safe navigation decisions. With the continuous development of intelligent sensors, data fusion, and machine learning, ship situational awareness has gradually evolved from monitoring based on a single data source, such as traditional radar and AIS (Automatic Identification System), to integrating heterogeneous data from multiple sources to achieve comprehensive perception of the environment and the ship's own status. Existing technologies typically assist ship operators in collision avoidance and path optimization decisions through navigation environment perception, obstacle detection, and track prediction, playing a positive role in improving navigation safety and reducing accident risks. However, in the face of complex dynamics, large data volumes, and high uncertainty in the real-world maritime environment, existing technologies still have certain shortcomings in the spatiotemporal processing of ship operating environment data, obstacle threat measurement, accurate prediction of minimum encounter distances, and the development of risk-based warning strategies. Further improvements are urgently needed in the accuracy of perception and prediction, as well as the timeliness of response. In particular, existing methods are insufficient in the refined processing capabilities required to comprehensively consider dynamic indicators such as obstacle relative heading changes and approach times to determine threat levels.

[0003] CN118245756B discloses a situational awareness-based intelligent navigation analysis method and system for ships. This method acquires navigation environment monitoring information and ship operation monitoring information to perform environmental perception and ship status assessment, respectively. It then uses digital twin technology to construct a digital model of the waterway and assess the waterway status. Ultimately, it predicts navigation events and assesses the degree of obstruction, enabling the formulation of navigation strategies for collision avoidance and route adjustments. This method comprehensively perceives the navigation environment, improving the ship's adaptability and safety in complex navigation conditions. However, it fails to carefully consider the integrated calculation of dynamic indicators such as heading angle change rate and approach time in its threat assessment of dynamic obstacles, leaving room for improvement in the accuracy of refined threat prediction and graded warnings.

[0004] CN114627363B discloses a panoramic maritime vessel situational awareness method based on multi-task learning. By establishing a multi-task learning network model, this method achieves vessel situation prediction and visualization based on remote sensing image datasets. This method effectively improves the overall accuracy of maritime vessel situational awareness, addressing the issues of isolation between different perception systems and low detection accuracy. However, due to its reliance on remote sensing imagery as the primary perception method, it lacks a dynamic synchronization mechanism for real-time changes in the close-range operating environment. Furthermore, it fails to conduct comprehensive threat assessment based on multiple factors, such as obstacle approach time and dynamic collision probability. This limits its real-time performance and its ability to predict threat trends in highly dynamic scenarios. Summary of the Invention

[0005] In view of the problems in existing ship collision avoidance and navigation risk assessment methods, such as insufficient time-space synchronization accuracy, imperfect obstacle threat assessment, limited trajectory prediction accuracy, and delayed collision avoidance decision response, the present invention is proposed.

[0006] Therefore, the problem to be solved by the present invention is how to achieve high-precision spatiotemporal alignment and fusion of ship operating environment data, accurately assess the threat level of obstacles and perform dynamic risk prediction, and formulate intelligent navigation adjustment strategies that comply with international maritime collision avoidance regulations based on graded warnings based on minimum encounter distances, thereby effectively improving the intelligence level of collision avoidance decisions and navigation safety during ship navigation.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] In a first aspect, an embodiment of the present invention provides a method for intelligent ship situation awareness prediction, which includes:

[0009] The timestamp alignment method is used to synchronize the collected ship's operating environment data in time and space to generate a ship situation dataset;

[0010] Based on the ship situation dataset, the obstacle's heading angle change rate and Euclidean distance relative to the ship are extracted, and the dynamic collision probability and obstacle approach time are weighted and fused to obtain the obstacle threat index.

[0011] The obstacle threat index is input into the ship trajectory prediction model, and the minimum encounter distance between the ship and each obstacle is output;

[0012] According to the International Regulations for Preventing Collisions at Sea, the minimum encounter distance is divided into risk levels, a graded warning signal is output, and a navigation strategy is formulated to implement obstacle avoidance and route adjustment.

[0013] As a preferred solution of the intelligent ship situation awareness prediction method of the present invention, the minimum encounter distance is divided into risk levels according to the International Regulations for Preventing Collisions at Sea, and a graded warning signal is displayed through a human-computer interaction interface, including:

[0014] Set a safety distance threshold according to the International Regulations for Preventing Collisions at Sea, compare the minimum encounter distance with the safety distance threshold, and determine the navigation risk level;

[0015] Based on the relative azimuth between own ship and obstacle, the encounter status is divided into overtaking situation, crossing situation and encounter situation.

[0016] Based on the ship's current heading and the obstacle threat index, a new navigation path point sequence is obtained to generate a route adjustment plan, where the navigation path point sequence satisfies the ship's minimum turning radius constraint.

[0017] The adjusted course is re-input into the ship trajectory prediction model to calculate the new minimum encounter distance and determine whether it meets the safe navigation requirements;

[0018] Record warning signals, encounter status and course adjustment data, and establish a situation warning database.

[0019] As a preferred solution of the intelligent ship situation awareness prediction method of the present invention, the determination of the navigation risk level includes:

[0020] When the minimum encounter distance is less than the first threshold, the navigation risk level is high risk. If it is high risk, a red warning signal is output, triggering the emergency collision avoidance decision-making process and adjusting the course according to the current encounter status.

[0021] When the minimum encounter distance is between the first threshold and the second threshold, the navigation risk level is medium risk level; if it is medium risk level, a yellow warning signal is output and the yaw angle correction value is calculated;

[0022] When the minimum encounter distance is greater than the second threshold, the navigation risk level is a low risk level; if it is a low risk level, a green warning signal is output, the original course and speed are maintained, and the changing trend of the minimum encounter distance is continuously monitored.

[0023] As a preferred solution of the ship intelligent situation awareness prediction method of the present invention, wherein:

[0024] Adjust course based on the current encounter status, including:

[0025] If in overtaking state, own ship shall deviate to right by a course not less than the first preset angle;

[0026] If in the crossing state, the ship will deviate to the right by a course not less than the second preset angle;

[0027] If in a head-on encounter situation, the ship will deviate to the right by a course not less than the third preset angle.

[0028] As a preferred solution of the ship intelligent situation awareness prediction method of the present invention, the method for obtaining the minimum encounter distance is:

[0029] Construct a trajectory prediction model based on a combined recurrent neural network;

[0030] Sampling the ship position data and obstacle position data at preset time intervals to form a trajectory sequence of several sampling points, wherein the sampling points include position coordinates, heading angles, and speed information;

[0031] Combining the trajectory sequence with the obstacle threat index to construct a spatiotemporal feature tensor, which is input into the bidirectional long short-term memory network layer;

[0032] Extract the temporal dependencies of the trajectory sequence through forward propagation and backpropagation, and output the hidden state sequence;

[0033] The spatiotemporal attention layer calculates an attention score matrix based on the hidden state sequence and the obstacle threat index;

[0034] The attention score matrix is ​​weightedly summed with the hidden state sequence, mapped to the prediction space through a fully connected layer, and the trajectory prediction point sequence is output;

[0035] Calculate the Euclidean distance between the predicted ship trajectory points and the predicted obstacle trajectory points within the prediction time window to form a relative distance matrix;

[0036] Based on the relative distance matrix, a dynamic programming algorithm is used to calculate the minimum encounter distance between the ship and each obstacle within the prediction time window, wherein the minimum encounter distance includes the turning radius and the acceleration and deceleration capability limit.

[0037] As a preferred solution of the intelligent ship situational awareness prediction method described in the present invention, the trajectory prediction model includes a bidirectional long short-term memory network layer and a spatiotemporal attention layer; the bidirectional long short-term memory network layer extracts the historical trajectory characteristics of the ship and the obstacle; and the spatiotemporal attention layer weights the obstacle threat index.

[0038] As a preferred solution of the ship intelligent situation awareness prediction method of the present invention, the method for obtaining the obstacle threat index is:

[0039] Based on the position data of the ship situation data set and the distance data of the surrounding obstacles, the relative coordinate system between the ship and the obstacles is established;

[0040] In the relative coordinate system, the heading angle of the obstacle relative to the own ship is calculated, and a heading angle change rate sequence is obtained by performing a differential operation on the heading angles within a continuous time window;

[0041] Inputting the heading angle change rate sequence into a long short-term memory neural network model, wherein the long short-term memory neural network model includes an input layer, a hidden layer, and an output layer;

[0042] Outputting the dynamic collision probability between the obstacle and the own ship through temporal feature learning of the long short-term memory neural network model;

[0043] At the same time, the Euclidean distance is calculated using the respective position coordinates of the ship and the obstacle;

[0044] Based on the Euclidean distance and the relative speed of the obstacle, a dynamic time window method is used to calculate the obstacle approach time;

[0045] An adaptive weight coefficient is designed to perform weighted fusion on the dynamic collision probability and the approach time to obtain an obstacle threat index.

[0046] In a second aspect, an embodiment of the present invention provides an intelligent ship situation awareness and prediction system, which includes:

[0047] The acquisition module is used to synchronize the collected ship's operating environment data in time and space using a timestamp alignment method to generate a ship situation data set;

[0048] The extraction module extracts the obstacle's heading angle change rate and Euclidean distance relative to the ship based on the ship situation dataset, and performs weighted fusion of the dynamic collision probability and the obstacle's approach time to obtain the obstacle threat index;

[0049] The output module inputs the obstacle threat index into the ship trajectory prediction model and outputs the minimum encounter distance between the ship and each obstacle;

[0050] The early warning module classifies the risk level of the minimum encounter distance according to the International Regulations for Preventing Collisions at Sea, outputs a graded early warning signal, and formulates a navigation strategy to implement obstacle avoidance and route adjustment.

[0051] In a third aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the processor executes the computer program, it implements any step of the above-mentioned ship intelligent situational awareness and prediction method.

[0052] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the above-mentioned ship intelligent situational awareness and prediction method is implemented.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] 1. By using the timestamp alignment method to synchronize the collected ship's operating environment data in time and space, a ship situation dataset is generated, achieving consistent fusion of multi-source environmental information. This effectively avoids information distortion caused by data collection time deviations, and provides a unified and continuous time basis for subsequent data-based collision risk assessment and trajectory prediction.

[0055] 2. By extracting the obstacle's heading angle change rate and Euclidean distance relative to the ship based on the ship situation dataset, and weightedly fusing the dynamic collision probability with the obstacle approach time, the obstacle threat index is obtained. This quantitative representation of the obstacle threat level is achieved. Not only is the spatial position relationship considered, but also the dynamic approach trend is incorporated, forming a comprehensive spatiotemporal feature assessment system that reflects the actual threat changes of obstacles to ship safety, enhances the sensitivity and accuracy of risk identification, and achieves rapid and accurate assessment in complex navigation environments.

[0056] 3. By inputting the obstacle threat index into the ship trajectory prediction model and outputting the minimum encounter distance between the ship and each obstacle, accurate risk prediction for future navigation dynamics is achieved, avoiding the failure of traditional static distance judgment in dynamic environments and improving the accuracy and adaptability of trajectory prediction.

[0057] 4. By classifying the minimum encounter distance into risk levels according to the International Regulations for Preventing Collisions at Sea, outputting graded warning signals, and formulating navigation strategies to implement obstacle avoidance and route adjustments, an intelligent navigation decision-making mechanism that is aligned with international standards has been implemented, giving the system standardized and explainable risk response capabilities. By triggering different levels of route adjustment plans based on risk levels, not only does this improve the rationality and compliance of decision-making, it also effectively balances navigation efficiency and safety, thereby enhancing the safety level of autonomous navigation and reducing the risk of human judgment bias.

[0058] 5. By adjusting the course based on the current encounter status, including setting the right turn yaw angle for collision avoidance decisions in different situations such as overtaking, crossing, and encountering, a differentiated course adjustment strategy based on specific situation characteristics is implemented, overcoming the problems of insufficient or excessive avoidance caused by the one-size-fits-all treatment of traditional collision avoidance algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort. Among them:

[0060] Figure 1 This is a flow chart of the ship intelligent situation awareness prediction method. DETAILED DESCRIPTION

[0061] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0062] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0063] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0064] The present invention is described in detail with reference to schematic diagrams. For ease of illustration, cross-sectional views of device structures may be partially enlarged and not to scale when describing embodiments of the present invention. Furthermore, the schematic diagrams are merely illustrative and should not limit the scope of the present invention. Furthermore, in actual production, the three-dimensional dimensions of length, width, and depth should be included.

[0065] In the description of the present invention, it should be noted that the terms "upper, lower, inner, and outer" and other references to orientations or positional relationships are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate and simplify the description of the present invention. They are not intended to indicate or imply that the devices or components referred to must have, be constructed, or operate in a specific orientation, and therefore should not be construed as limitations on the present invention. Furthermore, the terms "first, second, or third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0066] In this disclosure, unless otherwise specified or limited, the terms "mounted," "connected," and "connected" should be interpreted broadly. For example, they may refer to fixed, removable, or integral connections. They may also refer to mechanical, electrical, or direct connections, indirect connections through an intermediary, or internal communication between two components. Those skilled in the art will understand the specific meanings of these terms in this disclosure.

[0067] Example 1

[0068] Reference Figure 1 , which is the first embodiment of the present invention, provides an intelligent ship situation awareness prediction method, comprising:

[0069] S1: Use the timestamp alignment method to synchronize the collected ship's operating environment data in time and space to generate a ship situation dataset;

[0070] S2: Based on the ship situation dataset, the obstacle's heading angle change rate and Euclidean distance relative to the ship are extracted, and the dynamic collision probability and obstacle approach time are weighted and fused to obtain the obstacle threat index;

[0071] S3: Input the obstacle threat index into the ship trajectory prediction model and output the minimum encounter distance between the ship and each obstacle;

[0072] S4: According to the International Regulations for Preventing Collisions at Sea, the minimum encounter distance is divided into risk levels, a graded warning signal is output, and a navigation strategy is formulated to implement obstacle avoidance and route adjustment.

[0073] In the embodiment of the present application, the above step S1 includes:

[0074] It should be noted that the operating environment data includes location data, surrounding obstacle distance data and ambient wind speed data.

[0075] S1.1: Obtain the ship's location data through the ship's GPS positioning module;

[0076] It should be noted that the location data includes the longitude, latitude and heading angle information of the ship, and the collection frequency is once per second.

[0077] S1.2: Use millimeter-wave radar to collect relative distance data of surrounding obstacles;

[0078] It should be noted that the surrounding obstacle distance data includes the relative distance, relative azimuth and relative speed information to the obstacle, and the collection frequency is 2 times per second;

[0079] S1.3: Obtain current ambient wind speed data through a meteorological sensor, where the ambient wind speed data includes wind speed magnitude and wind direction angle;

[0080] S1.4: Add a unified timestamp to the location data, surrounding obstacle distance data, and ambient wind speed data, and align the data collected at different frequencies based on the timestamp.

[0081] Specifically, taking 1 second as the base time unit, the ambient wind speed data is interpolated to obtain the wind condition information per second, and the surrounding obstacle distance data is downsampled to ensure the synchronization of the three types of data in the time dimension.

[0082] S1.5: Integrate the time-aligned data in chronological order to form a ship situation dataset that includes complete operating environment information.

[0083] In the embodiment of the present application, the above step S2 includes:

[0084] S2.1: Based on the position data of the ship situation data set and the distance data of the surrounding obstacles, establish the relative coordinate system between the ship and the obstacles;

[0085] S2.2: Calculate the heading angle of the obstacle relative to the own ship in the relative coordinate system. Obtain the heading angle rate sequence by performing differential calculations on the heading angles within a continuous time window.

[0086] S2.3: Input the heading angle change rate sequence into a long short-term memory neural network model, where the long short-term memory neural network model includes an input layer, a hidden layer, and an output layer;

[0087] It should be noted that the hidden layer has 64 neurons and uses the tanh activation function.

[0088] S2.4: Output the dynamic collision probability between the obstacle and the ship through temporal feature learning of the long short-term memory neural network model;

[0089] Preferably, the specific formula of dynamic collision probability is as follows:

[0090]

[0091] Among them, h t is the LSTM hidden state; σ is the Sigmoid activation function; W o and b o is the output layer parameter; μ is the heading change rate influence coefficient; γ is the distance influence coefficient; d ref is the reference distance; d(t) is the Euclidean distance.

[0092] S2.5: At the same time, the Euclidean distance is calculated using the respective position coordinates of the ship and the obstacle;

[0093] S2.6: Calculate the obstacle approach time using the dynamic time window method based on the Euclidean distance and the relative velocity of the obstacle;

[0094] Preferably, the specific formula for obstacle approach time is as follows:

[0095]

[0096] Among them, λ is the speed adjustment coefficient; v r (t) refers to the speed of the obstacle relative to the ship; θ is the influence coefficient of the course change rate.

[0097] S2.7: Design an adaptive weight coefficient to perform weighted fusion of the dynamic collision probability and approach time to obtain the obstacle threat index;

[0098] Preferably, the specific formula of the obstacle threat index is as follows:

[0099] TI=ω(t)·P col (t)+(1-ω(t))·φ(t app (t));

[0100] Where TI is the obstacle threat index; ω(t) is the adaptive weight coefficient used to dynamically adjust the collision probability P col (t) and the weight ratio of the proximity time transfer function φ(*).

[0101] It should be noted that the weight coefficient is dynamically adjusted according to the complexity of the navigation environment.

[0102] In the embodiment of the present application, the above step S3 includes:

[0103] S3.1: Construct a trajectory prediction model based on a combined recurrent neural network;

[0104] In an optional embodiment, the trajectory prediction model includes a bidirectional long short-term memory network layer and a spatiotemporal attention layer; the bidirectional long short-term memory network layer extracts the historical trajectory characteristics of the ship and the obstacle; the spatiotemporal attention layer assigns weights to the obstacle threat index; the trajectory prediction model also includes an input layer, a first-layer gated recurrent unit network and a second-layer gated recurrent unit network; the input layer receives state parameters such as the obstacle threat index, the current heading and speed of the ship, the first-layer gated recurrent unit network contains 128 hidden neurons for extracting the temporal characteristics of the ship's motion; the second-layer gated recurrent unit network contains 64 hidden neurons for predicting the trajectory point sequence of the ship's future navigation.

[0105] S3.2: Sampling the ship position data and obstacle position data at preset time intervals to form a trajectory sequence of several sampling points, where the sampling points include position coordinates, heading angles, and speed information;

[0106] S3.3: Combine the trajectory sequence with the obstacle threat index to construct a spatiotemporal feature tensor, which is then fed into the bidirectional long short-term memory network layer.

[0107] S3.4: Extract the temporal dependencies of the trajectory sequence through forward propagation and backpropagation, and output the hidden state sequence;

[0108] S3.5: The spatiotemporal attention layer calculates the attention score matrix based on the hidden state sequence and the obstacle threat index;

[0109] Preferably, the specific formula of the attention score matrix is ​​as follows:

[0110]

[0111] Among them, S ij is the attention score of the i-th time step to the j-th time step, is the permutation of the hidden state vector at the i-th time step; h j is the hidden state vector at the jth time step; W q is the query matrix weight; is the permutation of the key matrix weights; d is the hidden state dimension; n is the time series length; η is the threat adjustment coefficient; TI is the obstacle threat index.

[0112] It should be noted that the attention score matrix represents the importance of different spatiotemporal positions.

[0113] S3.6: Take the weighted sum of the attention score matrix and the hidden state sequence, map it to the prediction space through the fully connected layer, and output the trajectory prediction point sequence;

[0114] S3.7: Calculate the Euclidean distances between the predicted ship trajectory points and the predicted obstacle trajectory points within the prediction time window to form a relative distance matrix;

[0115] Preferably, the specific formula of the relative distance matrix is ​​as follows:

[0116]

[0117] Among them, R ij is the relative distance matrix element between time point i and time point j; D base is the basic Euclidean distance; is the ship's maneuverability coefficient; is the relative heading angle change rate from time point i to time point j; ∈ is the standard deviation of the heading angle change rate; α is the heading influence factor, with a value range of [0,1]; Δψ ij is the relative heading difference between the ship and the obstacle.

[0118] S3.8: Based on the relative distance matrix, a dynamic programming algorithm is used to calculate the minimum encounter distance between the ship and each obstacle within the prediction time window, where the minimum encounter distance includes the turning radius and acceleration and deceleration capacity limitations.

[0119] Preferably, the specific formula of the minimum encounter distance is as follows:

[0120]

[0121] Where D[i][j] is the dynamic programming state value from time step i to j; R ij is the relative distance matrix element between time point i and time point j; c1, c2 and c3 are all path cost coefficients; η is the threat adjustment coefficient; TI is the obstacle threat index; Y is the length of the prediction time window.

[0122] In the embodiment of the present application, the above step S4 includes:

[0123] S4.1: Set a safety distance threshold according to the International Regulations for Preventing Collisions at Sea, compare the minimum encounter distance with the safety distance threshold, and determine the navigation risk level;

[0124] Specifically include:

[0125] When the minimum encounter distance is less than the first threshold, the navigation risk level is high risk. If it is high risk, a red warning signal is output, triggering the emergency collision avoidance decision-making process and adjusting the course according to the current encounter status.

[0126] For example, a high-risk level (minimum approach distance <0.5 nautical miles) is used as an example: In a specific navigation scenario, the own vessel and a preceding vessel are in an intersecting encounter. The minimum approach distance calculated from the vessel situation dataset is 0.3 nautical miles, which is lower than the preset first threshold (0.5 nautical miles). The current navigation risk level is determined to be high, and a red warning signal is immediately triggered. The warning information is highlighted on the bridge alarm and electronic chart. Simultaneously, the emergency collision avoidance decision-making process is initiated, analyzing the current encounter status in accordance with the International Regulations for Preventing Collisions at Sea (COLREGs). Because the two vessels are in an intersecting encounter and the own vessel is the give-way vessel, the system automatically generates a course adjustment instruction, requiring the own vessel to turn starboard by no less than 30 degrees. After the turn, the system inputs the new course data into the vessel trajectory prediction model and recalculates the minimum approach distance until it is confirmed to be within a safe range (≥0.5 nautical miles). If the minimum approach distance remains below the threshold after the adjustment, the steering angle will be increased or speed reduction will be recommended to ensure the effectiveness of the collision avoidance measures.

[0127] Preferably, if in the overtaking state, the ship deviates to the right by a course not less than the first preset angle; if in the crossing state, the ship deviates to the right by a course not less than the second preset angle; if in the encounter state, the ship deviates to the right by a course not less than the third preset angle.

[0128] When the minimum encounter distance is between the first threshold and the second threshold, the navigation risk level is medium risk level; if it is medium risk level, a yellow warning signal is output and the yaw angle correction value is calculated;

[0129] For example, at the medium risk level (0.5 nautical miles ≤ minimum approach distance < 1.0 nautical miles): when the ship approaches a vessel on the port side, the system calculates the minimum approach distance to be 0.7 nautical miles, which is between the first threshold (0.5 nautical miles) and the second threshold (1.0 nautical miles). The navigation risk level is determined to be medium risk, and a yellow warning signal is output. A yellow mark is used on the electronic nautical chart to alert the driver to the potential risk. At this time, emergency collision avoidance action will not be triggered. Instead, a gentle yaw angle correction (such as a 10° adjustment to the right) is calculated based on the obstacle's heading angle change rate and approach speed to gradually increase the minimum approach distance. The driver can manually confirm or adjust the correction. The trend of the corrected minimum approach distance is continuously monitored. If the distance is reduced to the high-risk range, the warning is immediately upgraded and the emergency collision avoidance process is initiated. If the minimum approach distance stabilizes or increases, the current strategy is maintained until the risk is resolved.

[0130] When the minimum encounter distance is greater than the second threshold, the navigation risk level is a low risk level; if it is a low risk level, a green warning signal is output, the original course and speed are maintained, and the changing trend of the minimum encounter distance is continuously monitored.

[0131] It should be noted that the safety distance threshold includes a first threshold and a second threshold; the first threshold is determined based on the minimum safe encounter distance required for a ship under emergency collision avoidance conditions; the second threshold is determined based on the minimum safe encounter distance required for a ship to maintain a navigation safety margin under conventional avoidance conditions.

[0132] For example, the low-risk level (minimum encounter distance ≥ 1.0 nautical miles): when sailing in open waters, the system detects that the minimum encounter distance of the nearest obstacle to the ship is 1.2 nautical miles, which exceeds the second threshold (1.0 nautical miles). Therefore, the navigation risk level is judged to be low risk, and a green warning signal is output, indicating that the current navigation status is safe and no collision avoidance measures are required. The electronic nautical chart normally displays the trajectory of the ship and surrounding ships, and continuously monitors the dynamic changes of the minimum encounter distance. If the minimum encounter distance is reduced to the medium-risk range due to subsequent acceleration or change of course of the obstacle, it will switch to the yellow warning state in time, and recommend fine-tuning of the course in advance to maintain a safe distance; the driver can always grasp the changes in the situation and make efficient and safe navigation decisions with the assistance of the system.

[0133] S4.2: Based on the relative bearing between own ship and obstacle, the encounter situation is divided into overtaking situation, crossing situation and encounter situation;

[0134] Preferably, when the obstacle is located at the stern relative to the ship and has a slower speed than the ship, it is judged as an overtaking situation; when the obstacle is located within a preset range on the port or starboard side of the ship relative to the ship and has a convergence trend, it is judged as a crossing situation; when the obstacle is located directly in front of the ship and has an opposite or nearly opposite heading, it is judged as an encounter situation.

[0135] S4.3: Based on the current course of the ship and the obstacle threat index, a new navigation path point sequence is obtained and a route adjustment plan is generated, where the navigation path point sequence satisfies the minimum turning radius constraint of the ship;

[0136] S4.4: Re-input the adjusted course into the ship trajectory prediction model, calculate the new minimum encounter distance, and determine whether it meets the safe navigation requirements;

[0137] Preferably, when the new minimum encounter distance is still less than the first threshold, it is determined that the emergency collision avoidance strategy does not meet the safety requirements, and the secondary collision avoidance decision process is triggered, and the heading adjustment angle is further increased (such as adding a 5° deflection) or coordinated speed reduction measures are taken; when the new minimum encounter distance is increased to between the first threshold and the second threshold, it is determined that the risk level is reduced to medium risk, the current adjusted heading is maintained, the yellow warning signal is continuously output and the dynamic threat index is monitored; when the new minimum encounter distance is greater than the second threshold, it is determined that the risk level is reduced to low risk, a green warning signal is output, the current heading is locked as a safe heading, and the collision avoidance action is terminated.

[0138] S4.5: Record warning signals, encounter status, and course adjustment data, and establish a situation warning database.

[0139] In summary, the present invention uses a timestamp alignment method to synchronize the ship operating environment data in time and space, ensures the consistency and timeliness of various types of data, and improves the accuracy and reliability of the situation data set; by extracting the relative heading angle change rate and Euclidean distance of the obstacle, and weightedly fusing the dynamic collision probability and approach time, a threat index is constructed, which accurately reflects the real-time threat level of the obstacle to the ship's navigation safety; by inputting the threat index into the trajectory prediction model and outputting the minimum encounter distance, the potential risk interval between the ship and the obstacle can be dynamically evaluated, and dangerous situations can be identified in advance; by dividing the minimum encounter distance into risk levels according to the International Regulations for Preventing Collisions at Sea and outputting graded warning signals, and formulating navigation strategies at the same time, differentiated handling and timely avoidance of different risk scenarios are achieved; the graded warning results are displayed in real time through the human-computer interaction interface, and the warning signals and adjustment data are recorded to establish a situation warning database, which provides historical support for the continuous optimization of collision avoidance decisions and navigation path adjustments, and significantly improves the accuracy of ship intelligent situational awareness, the timeliness of navigation decisions and navigation safety as a whole.

[0140] Furthermore, this embodiment also provides an intelligent ship situation awareness and prediction system, including:

[0141] The acquisition module is used to synchronize the collected ship's operating environment data in time and space using a timestamp alignment method to generate a ship situation data set;

[0142] The extraction module extracts the obstacle's heading angle change rate and Euclidean distance relative to the ship based on the ship situation dataset, and performs weighted fusion of the dynamic collision probability and the obstacle's approach time to obtain the obstacle threat index;

[0143] The output module inputs the obstacle threat index into the ship trajectory prediction model and outputs the minimum encounter distance between the ship and each obstacle;

[0144] The early warning module classifies the minimum encounter distance into risk levels according to the International Regulations for Preventing Collisions at Sea, outputs graded early warning signals, and formulates navigation strategies to implement obstacle avoidance and route adjustments.

[0145] This embodiment also provides an electronic device, which includes a processor, memory, a communication interface, a display screen, and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a multi-task edge computing resource scheduling method is implemented. The display screen of the computer device can be a liquid crystal display or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad provided on the computer device housing, or an external keyboard, touchpad or mouse.

[0146] This embodiment further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the method proposed in the above embodiment is implemented.

[0147] The storage medium proposed in this embodiment and the method proposed in the above embodiment belong to the same inventive concept. For technical details not fully described in this embodiment, please refer to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0148] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented with the help of software and necessary general-purpose hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the method of the embodiment of the present invention.

[0149] 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 the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0150] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The scheme in the embodiment of the present application can be implemented in various computer languages.

[0151] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0152] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0153] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0154] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0155] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A method for intelligent ship situation awareness prediction, characterized by: include, The timestamp alignment method is used to synchronize the collected ship's operating environment data in time and space to generate a ship situation dataset; Based on the ship situation dataset, the obstacle's heading angle change rate and Euclidean distance relative to the ship are extracted, and the dynamic collision probability and obstacle approach time are weighted and fused to obtain the obstacle threat index. The obstacle threat index is input into the ship trajectory prediction model, and the minimum encounter distance between the ship and each obstacle is output; According to the International Regulations for Preventing Collisions at Sea, the minimum encounter distance is classified into risk levels, a graded warning signal is output, and a navigation strategy is formulated to implement obstacle avoidance and route adjustment; The method for obtaining the minimum encounter distance is: Construct a trajectory prediction model based on a combined recurrent neural network; Sampling the ship position data and obstacle position data at preset time intervals to form a trajectory sequence of several sampling points, wherein the sampling points include position coordinates, heading angles, and speed information; Combining the trajectory sequence with the obstacle threat index to construct a spatiotemporal feature tensor, which is input into a bidirectional long short-term memory network layer; Extract the temporal dependencies of the trajectory sequence through forward propagation and backpropagation, and output the hidden state sequence; The spatiotemporal attention layer calculates an attention score matrix based on the hidden state sequence and the obstacle threat index; The attention score matrix is ​​weightedly summed with the hidden state sequence, mapped to the prediction space through a fully connected layer, and the trajectory prediction point sequence is output; Calculate the Euclidean distance between the predicted ship trajectory points and the predicted obstacle trajectory points within the prediction time window to form a relative distance matrix; Based on the relative distance matrix, a dynamic programming algorithm is used to calculate the minimum encounter distance between the ship and each obstacle within the prediction time window, wherein the minimum encounter distance includes turning radius and acceleration and deceleration capability limitations; The trajectory prediction model includes a bidirectional long short-term memory network layer and a spatiotemporal attention layer; the bidirectional long short-term memory network layer extracts the historical trajectory features of the ship and the obstacle; the spatiotemporal attention layer assigns weights to the obstacle threat index; The method for obtaining the obstacle threat index is: Based on the position data of the ship situation data set and the distance data of the surrounding obstacles, the relative coordinate system between the ship and the obstacles is established; In the relative coordinate system, the heading angle of the obstacle relative to the own ship is calculated, and a heading angle change rate sequence is obtained by performing a differential operation on the heading angles within a continuous time window; Inputting the heading angle change rate sequence into a long short-term memory neural network model, wherein the long short-term memory neural network model includes an input layer, a hidden layer, and an output layer; Outputting the dynamic collision probability between the obstacle and the own ship through temporal feature learning of the long short-term memory neural network model; At the same time, the Euclidean distance is calculated using the respective position coordinates of the ship and the obstacle; Based on the Euclidean distance and the relative speed of the obstacle, a dynamic time window method is used to calculate the obstacle approach time; An adaptive weight coefficient is designed to perform weighted fusion on the dynamic collision probability and the approach time to obtain an obstacle threat index.

2. The method for intelligent ship situation awareness and prediction according to claim 1, wherein: The minimum encounter distance is divided into risk levels according to the International Regulations for Preventing Collisions at Sea, and graded warning signals are displayed through the human-computer interaction interface, including: Set a safety distance threshold according to the International Regulations for Preventing Collisions at Sea, compare the minimum encounter distance with the safety distance threshold, and determine the navigation risk level; Based on the relative azimuth between own ship and obstacle, the encounter status is divided into overtaking situation, crossing situation and encounter situation. Based on the ship's current heading and the obstacle threat index, a new navigation path point sequence is obtained to generate a route adjustment plan, where the navigation path point sequence satisfies the ship's minimum turning radius constraint. The adjusted course is re-input into the ship trajectory prediction model to calculate the new minimum encounter distance and determine whether it meets the safe navigation requirements; Record warning signals, encounter status and course adjustment data, and establish a situation warning database.

3. The method for intelligent ship situation awareness and prediction according to claim 2, characterized in that: The determination of the navigation risk level includes: When the minimum encounter distance is less than the first threshold, the navigation risk level is high risk. If it is high risk, a red warning signal is output, triggering the emergency collision avoidance decision-making process and adjusting the course according to the current encounter status. When the minimum encounter distance is between the first threshold and the second threshold, the navigation risk level is medium risk level; if it is medium risk level, a yellow warning signal is output and the yaw angle correction value is calculated; When the minimum encounter distance is greater than the second threshold, the navigation risk level is a low risk level; if it is a low risk level, a green warning signal is output, the original course and speed are maintained, and the changing trend of the minimum encounter distance is continuously monitored.

4. The method for intelligent ship situation awareness and prediction according to claim 3, characterized in that: Adjust course based on the current encounter status, including: If in overtaking state, own ship shall deviate to right by a course not less than the first preset angle; If in the crossing state, the ship will deviate to the right by a course not less than the second preset angle; If in a head-on encounter situation, the ship will deviate to the right by a course not less than the third preset angle.

5. A ship intelligent situation awareness prediction system, based on the ship intelligent situation awareness prediction method according to any one of claims 1 to 4, characterized in that: include, The acquisition module is used to synchronize the collected ship's operating environment data in time and space using a timestamp alignment method to generate a ship situation data set; The extraction module extracts the obstacle's heading angle change rate and Euclidean distance relative to the ship based on the ship situation dataset, and performs weighted fusion of the dynamic collision probability and the obstacle's approach time to obtain the obstacle threat index; The output module inputs the obstacle threat index into the ship trajectory prediction model and outputs the minimum encounter distance between the ship and each obstacle; The early warning module classifies the risk level of the minimum encounter distance according to the International Regulations for Preventing Collisions at Sea, outputs a graded early warning signal, and formulates a navigation strategy to implement obstacle avoidance and route adjustment.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the ship intelligent situation awareness prediction method according to any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the ship intelligent situation awareness prediction method according to any one of claims 1 to 4 are implemented.

Images