Intelligent ship collision avoidance decision-making method and system
By constructing an adaptive data fusion model and introducing motion prediction methods with environmental factor correction, combined with multi-level risk assessment and optimized collision avoidance parameter calculation strategies, the challenges of existing ship collision avoidance decision-making methods in multi-source data fusion, motion prediction, risk assessment and decision optimization are solved, and high-precision and high-reliability intelligent collision avoidance decision-making decisions are achieved, and the navigation safety level is improved.
Patent Information
- Application Number
- CN202510312180.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-17
- Publication Date
- 2025-06-10
- Estimated Expiration
- 2045-03-17
AI Technical Summary
The existing ship collision avoidance decision-making methods have many challenges in multi-source data fusion, motion prediction, risk assessment and decision optimization, resulting in unstable situational awareness accuracy, inaccurate prediction trajectory, inadequate risk assessment, and difficult to balance safety and economy.
By building an adaptive data fusion model, introducing motion prediction methods for environmental factor correction, designing a multi-level risk assessment mechanism, and optimizing collision avoidance parameter calculation strategies, high-precision and high-reliability intelligent collision avoidance decisions are achieved. Specific methods include: collecting radar data and AIS information, building a dynamic prediction model, using Kalman filtering algorithm to predict the target ship trajectory, establishing a risk assessment matrix, using fuzzy neural network to make risk judgments, and calculating collision avoidance maneuver parameters through an adaptive particle swarm optimization algorithm.
The full process automation from information perception, situation prediction, risk assessment to decision-making execution has been achieved, which has greatly reduced the work burden of navigators, improved the level of navigation safety, and maintained high decision-making efficiency and accuracy in complex and multi-ship scenarios.
Smart Images

Figure CN120126349A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of ship intelligent navigation, and particularly to a ship intelligent collision avoidance decision-making method and system. Background Art
[0002] Traditional ship collision avoidance decision-making mainly relies on the experience judgment of crew members, and has inherent defects such as strong subjectivity and low decision-making efficiency. In recent years, with the continuous upgrading of on-board navigation equipment, intelligent collision avoidance technologies based on radar data and AIS information have gradually developed. Current research on collision avoidance decision-making mainly focuses on rule-based expert systems, deep learning-based situation recognition, and dynamic programming-based path planning. Among them, intelligent collision avoidance systems that integrate multi-source data show significant advantages in improving decision-making reliability, but still face many challenges when dealing with multi-ship encounter situations in complex navigation environments.
[0003] Existing ship collision avoidance decision-making methods generally have the following technical problems: First, the dynamic change characteristics of data reliability are not fully considered in the multi-source data fusion process, resulting in unstable situation awareness accuracy; second, the influence of environmental factors such as wind, waves, and currents on the ship motion prediction model is insufficiently considered, affecting the accuracy of the predicted trajectory; third, existing risk assessment methods often use simple distance threshold judgments, lacking in-depth analysis of the ship motion trend and being difficult to accurately reflect the actual collision avoidance risk level; finally, the relationship between safety and economy cannot be effectively balanced in the collision avoidance decision-making optimization process, resulting in problems of over-avoidance or under-avoidance in collision avoidance operations.
[0004] In view of the above problems, the present invention proposes a ship intelligent collision avoidance decision-making method and system. This method constructs an adaptive data fusion model, introduces a motion prediction method corrected by environmental factors, designs a multi-level risk assessment mechanism, and optimizes the calculation strategy of collision avoidance parameters, in order to achieve more accurate and reliable intelligent collision avoidance decision-making. Summary of the Invention
[0005] In view of the problems existing in the existing ship collision avoidance decision-making methods, the present invention is proposed.
[0006] Therefore, the problem to be solved by the present invention is how to achieve high-precision and high-reliability ship intelligent collision avoidance decision-making and effectively deal with multi-ship encounter situations in complex navigation environments.
[0007] 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 collision avoidance decision-making for ships, which includes collecting radar data and AIS information of the ship itself; constructing a ship dynamic prediction model based on the radar data and AIS information, predicting the future navigation trajectory of the target ship through the Kalman filtering algorithm, and calculating the encounter parameters between the two ships in combination with the navigation parameters of the ship itself; establishing a risk assessment matrix based on the encounter parameters, and using the fuzzy neural network algorithm to judge the risk level of the current navigation situation; based on the risk level judgment result, using the adaptive particle swarm optimization algorithm to calculate the collision avoidance maneuver parameters and output a collision avoidance action plan.
[0009] As a preferred scheme of the method for intelligent collision avoidance decision-making for ships according to the present invention, it includes: establishing a collision avoidance constraint condition library based on the risk level judgment result, and at the same time using the adaptive particle swarm algorithm to establish a collision avoidance parameter calculation model, where the collision avoidance constraint condition library includes avoidance rules and ship maneuverability; inputting the encounter parameters and the future navigation trajectory of the target ship into the collision avoidance parameter calculation model, verifying the combination of collision avoidance maneuver parameters, and outputting the steering angle value and the speed adjustment amount corresponding to the combination of collision avoidance maneuver parameters to form a collision avoidance action plan.
[0010] As a preferred scheme of the method for intelligent collision avoidance decision-making for ships according to the present invention, it includes: the collision avoidance maneuver parameters include the optimal steering angle and the speed adjustment amount; the construction method of the collision avoidance parameter calculation model is to use the collision avoidance constraint condition library as the optimization objective function, construct a search space including the steering angle and the speed adjustment amount, and set the initial position distribution and the initial speed distribution of the particle swarm; use the adaptive weight coefficient to update the position and speed of the particle swarm, and calculate the collision avoidance trajectory corresponding to the particle swarm; based on the collision avoidance trajectory, use the collision avoidance safety margin, the route deviation degree and the maneuvering energy consumption as evaluation indicators, and output the individual optimal position and the global optimal position of the particle swarm; perform iterative calculation on the particle swarm according to the evaluation indicators to obtain a combination of collision avoidance parameters that meets the avoidance rules.
[0011] As a preferred embodiment of the intelligent collision avoidance decision-making method for ships according to the present invention, the following steps are included: based on the encounter parameters, a risk assessment matrix is established, and the fuzzy neural network algorithm is used to judge the risk level of the current navigation situation, including: constructing a risk assessment matrix, and dividing the grading criteria of the risk assessment matrix according to the threshold range of the encounter parameters, where the encounter parameters include encounter time and encounter distance; based on the risk assessment matrix, a risk assessment model is constructed using a fuzzy neural network, where the fuzzy neural network includes an input layer, a fuzzy layer, a rule layer, and an output layer; training the risk assessment model according to historical collision avoidance case data, and optimizing the model weight parameters through the backpropagation algorithm to adjust the membership function; inputting the current navigation situation data into the trained risk assessment model, calculating the membership value of the collision risk, and outputting the risk level of the current situation according to the maximum membership principle; based on the risk level of the current situation, continuously collect navigation situation data, and adjust and update the judgment result of the risk level in real time.
[0012] Regarding the division of the risk assessment matrix in the preferred embodiment of the intelligent collision avoidance decision-making method for ships according to the present invention: when the encounter time is less than the first time threshold and the encounter distance is less than the first distance threshold, it corresponds to the emergency danger level of the risk assessment matrix; when the encounter time is less than the second time threshold and the encounter distance is less than the second distance threshold, it corresponds to the high danger level of the risk assessment matrix; when the encounter time is between the second time threshold and the third time threshold and the encounter distance is between the second distance threshold and the third distance threshold, it corresponds to the medium danger level of the risk assessment matrix; when the encounter time is greater than the third time threshold or the encounter distance is greater than the third distance threshold, it corresponds to the low danger level of the risk assessment matrix; when the encounter time is greater than the fourth time threshold and the encounter distance is greater than the fourth distance threshold, it corresponds to the safety level of the risk assessment matrix.
[0013] As a preferred embodiment of the intelligent collision avoidance decision-making method for ships according to the present invention, the following steps are included: constructing a ship dynamic prediction model based on the radar data and AIS information, predicting the future navigation trajectory of the target ship through the Kalman filter algorithm, and calculating the encounter parameters between the two ships in combination with the navigation parameters of the own ship, including: inputting the radar data and AIS information into the ship dynamic prediction model, and establishing a state vector including the longitudinal speed, lateral speed and heading angular velocity of the ship, wherein the ship dynamic prediction model describes the ship navigation state using a six-degree-of-freedom motion equation; constructing a state transition matrix of the Kalman filter according to the state vector, and at the same time establishing an observation equation to map the measured values of the radar data and AIS information into the state space; using the Kalman filter to iteratively update the state of the target ship, optimizing the state estimate value through a prediction-correction cycle, and outputting the future navigation trajectory of the target ship; based on the future navigation trajectory of the target ship, establishing a relative motion model in combination with the navigation parameters of the own ship, converting the predicted trajectory of the target ship into a relative motion coordinate system, and fitting the relative motion trajectory curve using the least squares method; calculating and outputting the encounter parameters between the two ships according to the relative motion trajectory curve; performing uncertainty analysis on the future navigation trajectory of the target ship and the encounter parameters, and outputting the corresponding confidence interval according to the error propagation characteristics of the state estimate value.
[0014] As a preferred embodiment of the intelligent collision avoidance decision-making method for ships according to the present invention, the following are included: the radar data includes the echo signal and distance information of the target ship; the AIS information includes the longitude and latitude positions, speed values and heading angles of the own ship and the target ship.
[0015] In a second aspect, an embodiment of the present invention provides an intelligent collision avoidance decision-making system for ships, which includes: a data acquisition module for acquiring the radar data and AIS information of the own ship; a target ship trajectory prediction module for constructing a ship dynamic prediction model based on the radar data and AIS information, predicting the future navigation trajectory of the target ship through the Kalman filter algorithm, and calculating the encounter parameters between the two ships in combination with the navigation parameters of the own ship; a judgment module for establishing a risk assessment matrix based on the encounter parameters and using the fuzzy neural network algorithm to judge the risk level of the current navigation situation; a navigation risk assessment module for calculating the collision avoidance maneuver parameters using the adaptive particle swarm optimization algorithm based on the risk level judgment result and outputting a collision avoidance action plan.
[0016] In a third aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the following is included: when the computer program instructions are executed by the processor, the steps of the intelligent collision avoidance decision-making method for ships as described in the first aspect of the present invention are implemented.
[0017] Fourthly, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, wherein: when the computer program instructions are executed by a processor, the steps of the ship intelligent collision avoidance decision-making method described in the first aspect of the present invention are implemented.
[0018] The beneficial effects of the present invention are as follows: by fusing multi-source data to collect radar and AIS information, a comprehensive and reliable information basis is constructed; a dynamic prediction model is constructed using the Kalman filter algorithm to achieve high-precision prediction of the future trajectory of the target ship; a risk assessment matrix is established based on encounter parameters and combined with the fuzzy neural network algorithm to intelligently judge the risk level of the current navigation situation; the adaptive particle swarm optimization algorithm is used to calculate the collision avoidance maneuver parameters, and an optimal collision avoidance plan that not only meets the International Regulations for Preventing Collisions at Sea but also considers the ship's maneuverability is generated; the present invention realizes the full-process automation from information perception, situation prediction, risk assessment to decision execution, greatly reducing the work burden of navigators and improving the level of navigation safety. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings. Among them:
[0020] Figure 1 It is the overall flowchart of the ship intelligent collision avoidance decision-making method in Embodiment 1. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0021] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the specific embodiments of the present invention in detail with reference to the drawings of the specification.
[0022] In the following description, many specific details are set forth in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.
[0023] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation manner of the present invention. The "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that excludes other embodiments.
[0024] Embodiment 1
[0025] Refer to Figure 1, which is the first embodiment of the present invention. This embodiment provides a method for intelligent collision avoidance decision-making of ships, including
[0026] S1: Collect the radar data and AIS information of the ship itself.
[0027] Specifically, the radar data includes the echo signal and distance information of the target ship. Perform adaptive CFAR processing on the radar data to suppress the influence of sea clutter and extract effective target echo information.
[0028] It should be noted that the shipborne X-band radar scanning system continuously emits electromagnetic wave signals, records the time interval of the reflected echo and converts it into the distance data of the target ship, and at the same time measures the intensity value of the echo signal and performs digital processing; uses the Doppler effect principle to calculate the radial velocity of the target ship, and combines the radar scanning azimuth to obtain the relative azimuth data of the target ship.
[0029] Furthermore, the AIS information includes the longitude and latitude positions, speed values, and course angles of the ship itself and the target ship.
[0030] It should be noted that receive the longitude and latitude coordinate information output by the ship's GPS positioning system, obtain the course data through the gyrocompass, collect the ship's speed value relative to the ground using the Doppler log, and integrate the positioning data, course data, and speed data into the ship's status information; monitor the AIS communication channel, receive the dynamic information message broadcast by the target ship at regular time intervals, parse the message to obtain the longitude and latitude position, speed value, and course angle of the target ship, and establish a dynamic information database of the target ship;
[0031] Even further, perform time synchronization and coordinate unification processing on the collected radar data and AIS information, remove outliers and complete missing data.
[0032] S2: Construct a ship dynamic prediction model based on the radar data and AIS information, predict the future navigation trajectory of the target ship through the Kalman filter algorithm, and calculate the encounter parameters between the two ships in combination with the navigation parameters of the ship itself.
[0033] Specifically, input the radar data and AIS information into the ship dynamic prediction model, and establish a state vector including the longitudinal speed, lateral speed, and heading angular velocity of the ship, where the ship dynamic prediction model uses the six-degree-of-freedom motion equation to describe the ship's navigation state.
[0034] It should be noted that the encounter parameters include the encounter time and encounter distance.
[0035] Preferably, the specific formula of the ship dynamic prediction model is as follows:
[0036]
[0037] Among them, is the eastward coordinate of the ship, is the northward coordinate of the ship, is the heading angle of the ship, u is the longitudinal speed of the ship, v is the lateral speed of the ship, r is the angular velocity of the ship's turning, and V w is the wind speed, and θ w is the wind direction angle, β is the correction coefficient, and θ c is the sea current direction.
[0038] Furthermore, construct the state transition matrix of the Kalman filter according to the state vector, and at the same time establish an observation equation to map the measured values of radar data and AIS information to the state space.
[0039] Preferably, the specific formula of the observation equation is as follows:
[0040]
[0041] Among them, z k is the observation vector, x k is the eastward coordinate at time k, y k is the northward coordinate at time k, ψ k is the heading angle at time k, α 1 is the eastward measurement error coefficient of the position, α 2 is the northward measurement error coefficient of the position, α 3 is the measurement error coefficient of the heading angle, R k is the target distance, and w k is the observation noise.
[0042] Even further, use the Kalman filter to iteratively update the state of the target ship, optimize the state estimate value through the prediction-correction cycle, and output the future navigation trajectory of the target ship; based on the future navigation trajectory of the target ship, establish a relative motion model in combination with the navigation parameters of the own ship, convert the predicted trajectory of the target ship to the relative motion coordinate system, and use the least squares method to fit the relative motion trajectory curve.
[0043] It should be noted that the relative motion model is based on.
[0044] Specifically, calculate and output the encounter parameters of the two ships according to the relative motion trajectory curve; conduct uncertainty analysis on the future navigation trajectory and encounter parameters of the target ship, and output the corresponding confidence interval according to the error propagation characteristics of the state estimate value.
[0045] S3: Based on the encounter parameters, establish a risk assessment matrix, and use the fuzzy neural network algorithm to judge the risk level of the current navigation situation;
[0046] Specifically, construct a risk assessment matrix, and divide the grade standard of the risk assessment matrix according to the threshold range of the encounter parameters.
[0047] Preferably, the risk assessment matrix is divided as follows: when the encounter time is less than the first time threshold and the encounter distance is less than the first distance threshold, it corresponds to the emergency danger level of the risk assessment matrix; when the encounter time is less than the second time threshold and the encounter distance is less than the second distance threshold, it corresponds to the high danger level of the risk assessment matrix; when the encounter time is between the second time threshold and the third time threshold and the encounter distance is between the second distance threshold and the third distance threshold, it corresponds to the medium danger level of the risk assessment matrix; when the encounter time is greater than the third time threshold or the encounter distance is greater than the third distance threshold, it corresponds to the low danger level of the risk assessment matrix; when the encounter time is greater than the fourth time threshold and the encounter distance is greater than the fourth distance threshold, it corresponds to the safety level of the risk assessment matrix.
[0048] It should be noted that if the membership degree value of the risk level corresponds to the emergency danger level, the emergency collision avoidance plan is activated and an alarm signal is sent to the driver; if the membership degree value of the risk level corresponds to the high danger level, a collision avoidance suggestion is generated and a collision avoidance countdown is displayed to the driver; if the membership degree value of the risk level corresponds to the medium danger level, the change of encounter parameters is continuously monitored and a collision avoidance plan is prepared; if the membership degree value of the risk level corresponds to the low danger level or the safety level, the navigation situation monitoring is maintained.
[0049] Furthermore, based on the risk assessment matrix, a risk assessment model is constructed using a fuzzy neural network, where the fuzzy neural network includes an input layer, a fuzzy layer, a rule layer, and an output layer.
[0050] It should be noted that the membership function of the encounter parameters is set in the fuzzy layer; the expert knowledge rule base is embedded in the rule layer.
[0051] Even further, the risk assessment model is trained according to the historical collision avoidance case data, and the model weight parameters are optimized through the backpropagation algorithm to adjust the membership function; the current navigation situation data is input into the trained risk assessment model, the membership degree value of the collision risk is calculated, and the risk level of the current situation is output according to the maximum membership degree principle.
[0052] Preferably, the specific formula for the membership degree value is as follows:
[0053]
[0054] where μ is the risk membership degree value, γ 1 is the DCPA influence weight coefficient, DCPA is the closest distance of approach, DCPA s is the safety threshold distance, γ 2 is the speed influence weight coefficient, V r is the relative speed, V c is the critical safety speed.
[0055] Specifically, based on the risk level of the current situation, continuously collect navigation situation data and adjust and update the judgment result of the risk level in real time.
[0056] S4: Based on the risk level judgment result, use the adaptive particle swarm optimization algorithm to calculate the collision avoidance maneuver parameters and output the collision avoidance action plan.
[0057] Specifically, establish a collision avoidance constraint condition library according to the risk level judgment result, and at the same time use the adaptive particle swarm algorithm to establish a collision avoidance parameter calculation model, where the collision avoidance constraint condition library includes avoidance rules and ship maneuvering performance.
[0058] Furthermore, the construction method of the collision avoidance parameter calculation model is as follows: taking the collision avoidance constraint condition library as the optimization objective function, construct a search space including the steering angle and the speed adjustment amount, and set the initial position distribution and initial speed distribution of the particle swarm; use the adaptive weight coefficient to update the position and speed of the particle swarm, and calculate the collision avoidance trajectory corresponding to the particle swarm; based on the collision avoidance trajectory, take the collision avoidance safety margin, the route deviation degree and the maneuvering energy consumption as evaluation indicators, and output the individual optimal position and the global optimal position of the particle swarm; perform iterative calculation on the particle swarm according to the evaluation indicators to obtain a collision avoidance parameter combination that meets the avoidance rules.
[0059] Furthermore, input the encounter parameters and the future navigation trajectory of the target ship into the collision avoidance parameter calculation model, verify the collision avoidance maneuver parameter combination, and output the steering angle value and the speed adjustment amount corresponding to the collision avoidance maneuver parameter combination to form a collision avoidance action plan.
[0060] It should be noted that the collision avoidance maneuver parameters include the optimal steering angle and the speed adjustment amount.
[0061] Specifically, the verification of the collision avoidance maneuver parameters includes: when the steering angle exceeds the maximum steering angle of the ship, the steering angle is limited within the maximum steering angle range of the ship; when the speed adjustment amount exceeds the maximum deceleration or acceleration range of the ship, the speed adjustment amount is limited within the range allowed by the ship's maneuvering performance; when the minimum distance between the calculated collision avoidance trajectory and the predicted trajectory of the target ship is less than the safety distance, re-perform the collision avoidance parameter optimization calculation; when the collision avoidance trajectory deviates too much from the original planned route, select a collision avoidance parameter combination with a smaller deviation degree on the premise of ensuring safety; when the collision avoidance maneuver parameters meet the ship's maneuvering performance constraints and can ensure collision avoidance safety, output the corresponding steering angle value and speed adjustment amount as the final collision avoidance action plan.
[0062] Furthermore, this embodiment also provides a ship intelligent collision avoidance decision-making system, including: a data acquisition module for collecting radar data and AIS information of the ship itself; a target ship trajectory prediction module for constructing a ship dynamic prediction model based on the radar data and AIS information, predicting the future navigation trajectory of the target ship through the Kalman filter algorithm, and calculating the encounter parameters between the two ships in combination with the navigation parameters of the ship itself; a judgment module for establishing a risk assessment matrix based on the encounter parameters and using the fuzzy neural network algorithm to judge the risk level of the current navigation situation; a navigation risk assessment module for calculating collision avoidance maneuver parameters using the adaptive particle swarm optimization algorithm based on the risk level judgment result and outputting a collision avoidance action plan.
[0063] This embodiment also provides a computer device applicable to the situation of the ship intelligent collision avoidance decision-making method, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the ship intelligent collision avoidance decision-making method proposed in the above embodiment.
[0064] This computer device can be a terminal, and this computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of this computer device is used to provide computing and control capabilities. The memory of this 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 this computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of this computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of this computer device can be a touch layer covered on the display screen, or a button, a trackball, or a touchpad provided on the outer shell of the computer device, or an external keyboard, touchpad, or mouse, etc.
[0065] In summary, the present invention collects radar and AIS information through multi-source data fusion to build a comprehensive and reliable information basis; uses the Kalman filter algorithm to build a dynamic prediction model to achieve high-precision prediction of the future trajectory of the target ship; establishes a risk assessment matrix based on the encounter parameters and combines the fuzzy neural network algorithm to intelligently judge the risk level of the current navigation situation; uses the adaptive particle swarm optimization algorithm to calculate collision avoidance maneuver parameters and generate an optimal collision avoidance plan that not only meets the International Regulations for Preventing Collisions at Sea but also considers the ship's maneuverability; the present invention realizes the full-process automation from information perception, situation prediction, risk assessment to decision execution, greatly reducing the workload of seafarers and improving the level of navigation safety.
[0066] Embodiment 2
[0067] Referring to Table 1, this is the second embodiment of the present invention. This embodiment provides an intelligent collision avoidance decision-making method for ships. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.
[0068] Specifically, in this experiment, a 5,000-ton container ship is selected as the own ship, and multiple target ship scenarios are set in the sea area; the experimental equipment includes an X-band radar system (model: XR-420, detection range 24 nautical miles), an AIS receiving device (model: AIS-M7, receiving range 30 nautical miles), and an independently developed ship collision avoidance decision-making system (ICADS-V2.0); the ship collision avoidance decision-making system is initialized and configured, including parameter settings of the radar and AIS data acquisition modules, adjustment of Kalman filter parameters, determination of the risk assessment matrix threshold, and optimization of particle swarm algorithm parameters.
[0069] Furthermore, in the data acquisition and processing stage, the experimental system collects radar data and AIS information at a frequency of 2 Hz. The radar data includes the echo signal and distance information of the target ship, and the AIS information includes the longitude and latitude position, speed value, and heading angle of the ship; an adaptive noise reduction algorithm is used to process the radar echo signal, and the integrity of the AIS data is checked to eliminate outliers; in the dynamic prediction model construction stage, a ship dynamic prediction model is constructed based on the six-degree-of-freedom motion equation, which considers state variables such as the longitudinal speed, lateral speed, and heading angular velocity of the ship; in the experiment, the state transition matrix of the Kalman filter is designed as a time-varying matrix to adapt to the dynamic changes of the navigation environment; the observation noise covariance matrix is determined through multiple experiments and optimized. Through the prediction-correction cycle, the system iteratively updates the state of the target ship, and the prediction time window is set to 10 minutes, and the sampling interval is 30 seconds. In the risk assessment stage, a risk assessment matrix is constructed based on the encounter parameters, and five risk levels are set: emergency danger, high danger, medium danger, low danger, and safety. The threshold range of the risk assessment matrix is: the first time threshold is 3 minutes, and the first distance threshold is 0.5 nautical miles; the second time threshold is 6 minutes, and the second distance threshold is 1 nautical mile; the third time threshold is 12 minutes, and the third distance threshold is 2 nautical miles; the fourth time threshold is 20 minutes, and the fourth distance threshold is 3 nautical miles. In the experiment, the fuzzy neural network adopts a four-layer structure, including 10 input neurons, 25 fuzzy neurons, 30 rule neurons, and 5 output neurons. The model is trained with 500 sets of historical collision avoidance case data, the learning rate is set to 0.05, and the number of training iterations is 1,000 times. Finally, the classification accuracy of the model reaches 94.3%.
[0070] Furthermore, in the stage of collision avoidance decision-making, the experimental system established a collision avoidance constraint library according to the risk level, including rules such as crossing situation, head-on situation and overtaking situation in the International Regulations for Preventing Collisions at Sea, as well as the maneuvering performance parameters of the own ship. The parameters of the adaptive particle swarm optimization algorithm were set as follows: the number of particles was 50, the maximum number of iterations was 100, and the range of inertia weight was [0.4, 0.9]; a search space including steering angles [-60°, 60°] and speed adjustment amounts [-5 kn, +2 kn] was constructed, and the collision avoidance safety margin (weight 0.5), course deviation (weight 0.3) and maneuvering energy consumption (weight 0.2) were used as evaluation indexes; through adaptive weight adjustment and iterative optimization, the optimal collision avoidance action plan meeting the collision avoidance requirements was finally output.
[0071] Specifically, as shown in Table 1, through the analysis of experimental data, the method of the present invention shows significant technical advantages; in terms of decision timeliness, the average collision avoidance decision time is 13.87 seconds, which is improved by about 65%, and a relatively high decision-making efficiency can still be maintained in complex multi-ship encounter scenarios; in terms of collision avoidance safety, the minimum encounter distance in all scenarios is greater than the safety distance (0.5 nautical miles) specified in the COLREGs, and the average minimum encounter distance reaches 0.95 nautical miles, fully ensuring the navigation safety.
[0072] Table 1. Test data table
[0073] Meeting type Initial meeting distance Relative speed Collision avoidance time Minimum meeting distance Course deviation Decision accuracy Crossing meeting 2.5 nautical miles 15 knots 12.3s 1.2 nautical miles 5.8% 96.5% Overtaking meeting 1.8 nautical miles 8 knots 15.6s 0.8 nautical miles 4.2% 97.8% Head-on meeting 2.2 nautical miles 12 knots 13.8s 1.1 nautical miles 6.1% 95.9% Quarter crossing 1.5 nautical miles 18 knots 10.2s 0.9 nautical miles 7.3% 98.2% Oblique meeting 2.0 nautical miles 14 knots 14.5s 1.0 nautical miles 5.5% 96.8% Multiple ships meeting 1.7 nautical miles 16 knots 16.8s 0.7 nautical miles 8.2% 94.5%
[0074] Further, in terms of course deviation, the method of the present invention optimizes the collision avoidance parameters through the adaptive particle swarm optimization algorithm, so that the average course deviation is only 6.18%, which is reduced by about 65%. Especially in the overtaking encounter scenario, only 4.2% of course deviation is required to complete safe avoidance, significantly improving the navigation efficiency; in terms of decision accuracy, the average accuracy of six typical encounter scenarios reaches 96.62%, and even in the most complex multi-ship encounter scenario, a high accuracy of 94.5% is still maintained, reflecting the robustness and reliability of the method in complex navigation environments.
[0075] 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 them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.
Claims
1. A ship intelligent collision avoidance decision method, characterized by: include, Collect radar data and AIS information of the ship; A ship dynamic prediction model is constructed based on the radar data and AIS information, the future navigation trajectory of the target ship is predicted by a Kalman filter algorithm, and the encounter parameters of the two ships are calculated in combination with the navigation parameters of the own ship; Based on the encounter parameters, a risk assessment matrix is established, and a fuzzy neural network algorithm is used to determine the risk level of the current navigation situation; Based on the risk level judgment results, the collision avoidance maneuver parameters are calculated using the adaptive particle swarm optimization algorithm, and the collision avoidance action plan is output.
2. The intelligent ship collision avoidance decision method according to claim 1, characterized in that: Based on the risk level judgment results, the collision avoidance maneuver parameters are calculated using the adaptive particle swarm optimization algorithm, and the collision avoidance action plan is output, including: Establishing a collision avoidance constraint condition library according to the risk level judgment result, and establishing a collision avoidance parameter calculation model by using an adaptive particle swarm algorithm, wherein the collision avoidance constraint condition library includes avoidance rules and ship maneuverability; The encounter parameters and the future navigation trajectory of the target ship are input into the collision avoidance parameter calculation model, the collision avoidance maneuver parameter combination is verified, and the steering angle value and speed adjustment amount corresponding to the collision avoidance maneuver parameter combination are output to form a collision avoidance action plan.
3. The intelligent ship collision avoidance decision method according to claim 2, characterized in that: The collision avoidance maneuver parameters include the optimal steering angle and the speed adjustment amount; the construction method of the collision avoidance parameter calculation model is: Taking the collision avoidance constraint condition library as the optimization objective function, constructing a search space including a steering angle and a speed adjustment amount, and setting an initial position distribution and an initial speed distribution of the particle swarm; Adaptive weight coefficients are used to update the position and velocity of the particle swarm, and the corresponding collision avoidance trajectory of the particle swarm is calculated; Based on the collision avoidance trajectory, the collision avoidance safety margin, the route deviation and the control energy consumption are used as evaluation indicators to output the individual optimal position and the global optimal position of the particle swarm; The particle swarm is iteratively calculated according to the evaluation index to obtain a collision avoidance parameter combination that satisfies the avoidance rule.
4. The intelligent ship collision avoidance decision method according to claim 3, characterized in that: Based on the encounter parameters, a risk assessment matrix is established, and the fuzzy neural network algorithm is used to judge the risk level of the current navigation situation, including: Constructing a risk assessment matrix, and dividing the level standards of the risk assessment matrix according to the threshold range of the encounter parameters, wherein the encounter parameters include the encounter time and the encounter distance; Based on the risk assessment matrix, a risk assessment model is constructed using a fuzzy neural network, wherein the fuzzy neural network includes an input layer, a fuzzy layer, a rule layer and an output layer; The risk assessment model is trained based on historical collision avoidance case data, and the model weight parameters are optimized and the membership function is adjusted through a back propagation algorithm; Input the current navigation situation data into the trained risk assessment model, calculate the membership value of the collision risk, and output the risk level of the current situation according to the maximum membership principle; Based on the risk level of the current situation, navigation situation data is continuously collected, and the risk level determination results are adjusted and updated in real time.
5. The intelligent ship collision avoidance decision method according to claim 4, characterized in that: The risk assessment matrix is divided into: When the encounter time is less than the first time threshold and the encounter distance is less than the first distance threshold, the emergency danger level of the corresponding risk assessment matrix; When the encounter time is less than the second time threshold and the encounter distance is less than the second distance threshold, it corresponds to a high risk level of the risk assessment matrix; When the encounter time is between the second time threshold and the third time threshold and the encounter distance is between the second distance threshold and the third distance threshold, it corresponds to the medium danger level of the risk assessment matrix; When the encounter time is greater than the third time threshold or the encounter distance is greater than the third distance threshold, it corresponds to the low danger level of the risk assessment matrix; When the encounter time is greater than the fourth time threshold and the encounter distance is greater than the fourth distance threshold, it corresponds to the safety level of the risk assessment matrix.
6. The intelligent ship collision avoidance decision method according to claim 5, characterized in that: A ship dynamic prediction model is constructed based on the radar data and AIS information, the future navigation trajectory of the target ship is predicted by the Kalman filter algorithm, and the encounter parameters of the two ships are calculated in combination with the navigation parameters of the own ship, including: Inputting radar data and AIS information into a ship dynamic prediction model, and establishing a state vector including the longitudinal velocity, lateral velocity and bow angular velocity of the ship, wherein the ship dynamic prediction model uses a six-degree-of-freedom motion equation to describe the navigation state of the ship; Constructing a state transfer matrix of a Kalman filter according to the state vector, and establishing an observation equation to map the measured values of radar data and AIS information to the state space; The Kalman filter is used to iteratively update the state of the target ship, and the state estimation value is optimized through a prediction-correction cycle to output the future navigation trajectory of the target ship; Based on the future navigation trajectory of the target ship and the navigation parameters of the own ship, a relative motion model is established, the predicted trajectory of the target ship is converted into a relative motion coordinate system, and the relative motion trajectory curve is fitted using the least squares method; Calculate and output the encounter parameters of the two ships according to the relative motion trajectory curve; Uncertainty analysis is performed on the future navigation trajectory of the target ship and the encounter parameters, and a corresponding confidence interval is output according to the error propagation characteristics of the state estimation value.
7. The intelligent ship collision avoidance decision method according to claim 6, characterized in that: The radar data includes the echo signal and distance information of the target ship; the AIS information includes the latitude and longitude positions, speed values and heading angles of the own ship and the target ship.
8. An intelligent ship collision avoidance decision system, based on the intelligent ship collision avoidance decision method according to any one of claims 1 to 7, characterized in that: include, Data acquisition module, used to collect radar data and AIS information of the ship; The target ship trajectory prediction module is used to construct a ship dynamic prediction model based on the radar data and AIS information, predict the future navigation trajectory of the target ship through the Kalman filter algorithm, and calculate the encounter parameters of the two ships in combination with the navigation parameters of the own ship; A judgment module, used to establish a risk assessment matrix based on the encounter parameters, and use a fuzzy neural network algorithm to judge the risk level of the current navigation situation; The navigation risk assessment module calculates the collision avoidance maneuver parameters based on the risk level judgment results using the adaptive particle swarm optimization algorithm and outputs a collision avoidance action plan.
9. 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 intelligent ship collision avoidance decision-making method according to any one of claims 1 to 7 are implemented.
10. 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 intelligent ship collision avoidance decision-making method according to any one of claims 1 to 7 are implemented.
Citation Information
Patent Citations
Ship collision avoidance assisted decision-making method and system based on track fusion and track prediction
CN105390029A
Pedestrian tracking method based on least square locus prediction and intelligent obstacle avoidance model
CN106023244A
Ship collision risk degree calculation method based on accident data mining
CN113177713A
Automatic ship collision avoidance method based on improved particle swarm optimization
CN113538973A
Collision early warning method and system based on cloud side-end cooperation
CN115691223A
Cited By
Intelligent decision-making method based on maritime collision avoidance rule
CN120823729A
Intelligent ship visual monitoring system
CN121037536A
Trailing suction dredger sailing collision risk dynamic assessment and collision avoidance decision-making method
CN121122074A
Dynamic assessment of collision risk and collision avoidance decision-making method for trailing suction hopper dredger
CN121122074B
Collision risk early warning method and system based on foresight camera of unmanned ship
CN121392517A
