A method for calculating the resumption time of autonomous collision avoidance of a ship

Abstract

The application discloses a kind of ship autonomous collision avoidance re-navigation time calculation method, belong to ship intelligent navigation technical field.The existing re-navigation decision is only dependent on fixed distance threshold, dimension single, it is difficult to take into account the defects of navigation safety and operating efficiency, after the collision avoidance action is completed, the method real-time acquisition ship, target ship, route and environmental parameters;Dynamic collision risk decay model is constructed by fusing DCPA, TCPA and relative distance, and real-time collision risk value is output;Establish the maneuvering recovery degree evaluation model, quantitatively evaluate the stable state of the ship maneuvering;Combined with the drift distance, construct the comprehensive deviation cost function containing time and fuel cost;With double threshold as constraint, the minimum comprehensive cost as target, the optimal re-navigation time is solved by rolling horizon optimization, and the re-navigation instruction is output after being verified by multiple sliding windows.The application realizes the accurate quantitative decision of re-navigation time, can be adapted to different encounter scenarios, reduces the risk of secondary collision, reduces voyage delay, and improves the safety and economy of autonomous navigation system.

Description

Technical Field

[0001] This invention relates to the field of intelligent navigation technology for ships, and in particular to a method for calculating the timing of a ship's autonomous collision avoidance and subsequent return to navigation. Background Technology

[0002] With the rapid development of autonomous navigation technology, autonomous collision avoidance has become one of the core functions of intelligent ships. After a ship completes its avoidance maneuvers in accordance with international maritime collision avoidance rules and escapes the imminent danger from the target vessel, it needs to promptly resume its original planned route, i.e., a return-to-course operation. The timing of the return-to-course operation directly affects navigation safety and voyage efficiency: returning to course too early leaves the target vessel with the risk of dynamic turning, easily leading to a secondary collision; returning to course too late will result in excessive deviation from the route, increasing fuel consumption and voyage delays, and may even affect subsequent navigation scheduling.

[0003] In existing technologies, the decision-making process for resuming navigation after autonomous collision avoidance often employs a fixed threshold method. This means that when the distance between the vessel and the target vessel reaches a preset safe distance, a resuming navigation command is directly triggered. This method has the following significant drawbacks: First, relying solely on spatial distance as the only criterion for judgment fails to consider the dynamic changes in the target ship's movement trend, relative speed, and encounter situation. This approach cannot adapt to complex scenarios with different speeds and encounter angles, and is prone to misjudgment of risks, posing safety hazards in high-speed encounters or small-angle intersection scenarios.

[0004] Secondly, failing to consider the ship's own maneuverability and forcibly resuming navigation when the ship is still in an unstable state with large rudder angles and large yaw angles will increase the ship's maneuverability load, cause course oscillations, and even lead to loss of control.

[0005] Third, the economic and time costs of resuming navigation have not been quantified, making it impossible to optimize the efficiency of resuming navigation under safety constraints, and making it difficult to meet the needs of global navigation optimization for intelligent ships.

[0006] In publicly available research and patent solutions, some rerouting decision-making methods introduce TCPA as an auxiliary judgment indicator, but they remain at a simple logical level of single threshold superposition. They fail to establish a continuous attenuation model of collision risk evolution with the encounter situation, and cannot characterize the differences in risk evolution under different encounter patterns such as small-angle crossings, overtaking, and head-on encounters. In scenarios where the target vessel's heading and speed are uncertain, decision lags or advances are likely to occur. Other solutions focus on rerouting trajectory planning, assuming that rerouting control can be initiated immediately after the collision avoidance maneuver is completed. They do not conduct a systematic study on the pre-judgment of the rerouting initiation timing, and ignore the ship's maneuvering inertia and attitude stabilization process after large-angle collision avoidance. This can easily lead to problems such as heading overshoot and frequent rudder movements in the initial stage of rerouting, which reduces navigational stability and exacerbates wear and tear on rudder equipment. At the same time, existing rerouting methods generally lack scenario adaptability, using a uniform judgment standard for different waters such as open ocean, coastal channels, and narrow port channels. They cannot adapt to the differentiated requirements of traffic density, navigation restrictions, and hydrological environment, and cannot balance the safety margin and operational efficiency of different navigation scenarios.

[0007] Therefore, there is an urgent need to propose a multi-dimensional coupled method for calculating the timing of autonomous collision avoidance and return to navigation after a ship has autonomously avoided a collision, which can take into account both navigation safety and navigation efficiency, and improve the decision-making reliability of the autonomous navigation system. Summary of the Invention

[0008] To address the shortcomings of existing technologies, this invention provides a method for calculating the timing of a ship's return to course after autonomous collision avoidance. By coupling the dynamic attenuation characteristics of collision risk, the ship's maneuver recovery state, and the cost of course deviation, it achieves precise quantitative decision-making on the timing of the return to course, solving the technical problems of existing technologies, such as single decision-making dimensions, poor robustness, and inability to balance safety and efficiency.

[0009] To achieve the above objectives, the present invention provides the following technical solution: A method for calculating the timing of a ship's return to navigation after autonomous collision avoidance includes the following steps: S1. Situational data acquisition: After the autonomous collision avoidance maneuver is completed, the motion parameters of the ship itself, the motion parameters of the target ship, the planned route parameters, and the navigation environment parameters are collected in real time. S2. Dynamic Collision Risk Attenuation Calculation: Based on the relative motion relationship between the ship and the target ship, calculate the real-time encounter parameters, construct a collision risk attenuation coefficient model, and output the current collision risk value. S3. Calculation of ship maneuverability recovery: Based on the ship's speed, heading, rudder angle and turning angular velocity parameters, construct a maneuverability recovery evaluation model and output the ship's current maneuverability stability. S4. Route Deviation Cost Calculation: Based on the deviation between the ship's current position and the planned route, and in conjunction with the ship's maneuverability, calculate the time cost and fuel cost of rerouting, and construct a comprehensive deviation cost function. S5. Optimal Re-entry Timing Solution: With the constraints of collision risk being below the safety threshold and handling recovery being above the qualified threshold, and with the optimization objective of minimizing the overall deviation cost, a re-entry timing decision model is constructed, and the theoretical optimal re-entry time is obtained by solving the model. S6. Sliding window verification: The confidence level of the theoretically optimal return time is verified by using continuous multi-frame situational data. When the return constraints are met for N consecutive frames, the final return command is output.

[0010] Furthermore, in step S2, the real-time encounter parameters include relative distance, nearest encounter distance DCPA, and nearest encounter time TCPA; the collision risk attenuation coefficient model is: in, , , The weighting coefficients and ; The membership function of DCPA is 0 when DCPA is greater than the safe meeting distance, and otherwise increases linearly as DCPA decreases. This is the TCPA membership function, which takes the value of 0 when TCPA is greater than the security time threshold, and otherwise increases linearly as TCPA decreases; It is a membership function based on relative distance, which decays exponentially as the relative distance increases.

[0011] Furthermore, the weighting coefficients , , The collision risk safety threshold is dynamically adjusted based on visibility level and water traffic density; the threshold is set in stages according to ship size and water type.

[0012] Furthermore, in step S3, the manipulation recovery evaluation model is: in, , , The weighting coefficients and ; The speed recovery rate is calculated based on the deviation between the current speed and the planned speed. For heading stability, it is calculated based on the ratio of the current turning angular velocity to the rated maximum turning angular velocity; The rudder angle reset degree is calculated based on the ratio of the current rudder angle to the maximum rudder angle; the control recovery degree ranges from [0,1].

[0013] Furthermore, in step S4, the comprehensive deviation cost function is: in, For cost weighting coefficients; The time cost is calculated based on the ratio of the estimated resumption time to the reference time; The fuel cost is calculated based on the ratio of additional fuel consumption during resumption of service to a reference fuel consumption.

[0014] Furthermore, the estimated time for resuming navigation and the additional fuel consumption are calculated geometrically based on the current yaw distance, the ship's turning radius, and the fuel consumption rate per unit distance; the weighting coefficients The schedule is dynamically adjusted based on the time priority of each voyage.

[0015] Furthermore, in step S5, the rolling time-domain optimization method is used to solve for the optimal re-entry time. The situation data is updated and the solution is recalculated once in each sampling period. When all constraints are met at the current moment and the comprehensive deviation cost is in the decreasing range, the current moment is determined to be the theoretically optimal re-entry time.

[0016] Furthermore, in step S6, the length N of the sliding window is dynamically set according to the water area type: N is 3~5 for open water areas and 5~8 for restricted water areas; the sampling period is consistent with the ship sensor data update period.

[0017] Furthermore, in step S1, the motion parameters of the ship itself include speed, heading, rudder angle, turning angular velocity, and position coordinates; the motion parameters of the target ship include speed, heading, position coordinates, and ship dimensions; and the navigation environment parameters include wind speed, wind direction, current speed, current direction, and visibility level.

[0018] Compared with the prior art, the present invention has the following beneficial effects: 1. Enhanced safety: A dynamic collision risk attenuation model is constructed, integrating DCPA, TCPA, and relative motion trends to replace the traditional fixed distance threshold. This model can adapt to different encounter situations and flight speed scenarios, accurately identify the risk attenuation process, avoid secondary collision risks, and significantly improve the safety of rerouting decisions.

[0019] 2. Adaptation to maneuvering dimensions: Introduce a ship maneuvering recovery evaluation, combine the ship's real-time motion status to determine the feasibility of resuming navigation, avoid the ship entering the resuming navigation under unstable maneuvering conditions, reduce heading oscillation and maneuvering load, and improve the engineering robustness of decision-making.

[0020] 3. Efficiency dimension optimization: Establish a quantitative route deviation cost function, transform the rerouting opportunity into a constrained optimization problem, minimize the overall cost within the safety boundary, balance navigation safety and voyage efficiency, and adapt to voyage needs of different priorities.

[0021] 4. Enhanced reliability: The multi-frame sliding window verification mechanism filters out false triggers caused by sensor data fluctuations and instantaneous maneuvers of the target vessel, adapting to different scenarios such as open water and restricted water, and improving the stability of navigation decisions in complex environments. Attached Figure Description

[0022] Figure 1 is a schematic diagram of the overall process of the method for calculating the timing of resuming navigation according to the present invention; Figure 2 is a logical framework diagram of the decision-making model for resuming flights according to the present invention; Figure 3 is a schematic diagram showing the change of the collision risk attenuation coefficient of the present invention over time. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0024] The present invention proposes a method for calculating the timing of a ship's return to navigation after autonomous collision avoidance, comprising the following steps: S1. Situational Data Acquisition: After the autonomous collision avoidance maneuver is completed, the motion parameters of the ship itself, the target ship, the planned route, and the navigation environment are collected in real time. After the data is collected, the data is first preprocessed. Kalman filtering is used to fuse and correct the AIS and radar target data, remove abnormal jump points and noise interference, and uniformly convert them to the geodetic coordinate system for relative motion calculation. At the same time, the environmental parameters are classified and quantified, and visibility, wind level, and current speed are mapped to the corresponding environmental impact coefficients to provide input for the subsequent dynamic adjustment of model parameters. S2, Dynamic Collision Risk Attenuation Calculation: Based on the relative motion relationship between the ship and the target ship, real-time encounter parameters are calculated, a collision risk attenuation coefficient model is constructed, and the current collision risk value is output. For multi-target ship encounter scenarios, the collision risk value between the ship and each target ship is calculated separately, and a maximum value fusion strategy is adopted to obtain the global collision risk, ensuring that the most dangerous target is always used as a decision constraint. The weight coefficients are calibrated by combining expert experience and navigation scenario through fuzzy hierarchical analysis. The weights are adjusted in real time according to the environmental impact coefficient under the basic weights to achieve scenario adaptability of the risk model. S3. Calculation of Ship Maneuvering Restoration: Based on the ship's speed, heading, rudder angle, and turning angular velocity parameters, a maneuverability recovery evaluation model is constructed to output the ship's current maneuverability stability. The model further introduces the ship's roll and bow angle fluctuation amplitudes as auxiliary evaluation indicators to supplement and correct the ship's attitude stability after large-angle avoidance. At the same time, environmental force feedforward compensation is introduced to calculate the equivalent rudder angle disturbance based on wind and current forces, correct the measured rudder angle and turning angular velocity, and eliminate misjudgments of maneuverability stability caused by environmental disturbances. S4. Calculation of route deviation costs: Based on the deviation of the ship's current position from the planned route, and in conjunction with the ship's maneuverability, the time cost and fuel cost of rerouting are calculated, and a comprehensive deviation cost function is constructed. The function additionally introduces a route smoothness penalty term to apply an incremental cost to situations where the curvature of the rerouting turning trajectory exceeds the standard due to excessive deviation distance, thereby avoiding sharp turns caused by late rerouting timing. Both the time cost and fuel cost are processed using dimensionless normalization to eliminate the impact of dimensional differences on weight allocation. The weight coefficients can be adjusted online through priority instructions issued by the voyage management system. S5. Optimal Re-entry Timing Solution: With collision risk below the safety threshold and handling recovery above the acceptable threshold as constraints, and minimizing the overall deviation cost as the optimization objective, a re-entry timing decision model is constructed to obtain the theoretically optimal re-entry time. A rolling time-domain optimization method is used, with the prediction window set to 5-10 future sampling periods. Within each period, all candidate times within the prediction window are traversed, constraints are verified one by one, and the overall cost is calculated. The earliest time with the minimum cost and satisfying all constraints is selected as the theoretically optimal solution. When no time satisfies the constraints within the prediction window, a waiting instruction is output, and the next iteration begins. S6. Sliding window validation: The confidence level of the theoretically optimal retracing time is verified using continuous multi-frame situational data. When the retracing constraints are met for N consecutive frames, the final retracing command is output. The sliding window length adopts an adaptive adjustment mechanism, dynamically increasing or decreasing the window length according to the frequency of target ship maneuvers. When the target ship's heading and speed fluctuations are less than a set threshold, the window is shortened to improve response speed. When the target ship has continuous maneuvers, the window is extended to enhance anti-interference capability. The retracing command output after verification includes a suggested retracing heading and a recommended rudder angle, providing a pre-input for the lower-level trajectory tracking control.

[0025] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0026] As shown in Figure 1, the specific implementation steps of the method for calculating the timing of a ship's autonomous collision avoidance and subsequent return to navigation according to the present invention are as follows: S1, Situational Data Acquisition After the autonomous collision avoidance system determines that the collision avoidance action has been completed and the ship enters the cruise phase after collision avoidance, the process for calculating the timing of returning to the original route is started; four types of data are collected in real time through the ship's integrated navigation system, AIS, radar and environmental sensors: Own ship motion parameters: including own ship speed , course , rudder angle , turning angular velocity , latitude and longitude position coordinates; Target ship motion parameters: including target ship speed , course , position coordinates, ship length and width and other dimension information; Planned route parameters: including the coordinates of the start and end points of the planned leg, planned speed, route yaw limit threshold; Navigation environment parameters: including wind speed, wind direction, flow velocity, flow direction and visibility level; The data sampling cycle is consistent with the update cycle of the ship's sensors, usually 1 to 10 seconds, which can be adjusted according to the traffic density of the water area.

[0027] S2. Dynamic collision risk attenuation calculation First, calculate real-time encounter parameters based on the principle of relative motion: the relative velocity vector is synthesized from the velocity vectors of the own ship and the target ship, and the relative distance is the straight-line distance between the positions of the two ships; the calculation formulas for the Distance to Closest Point of Approach (DCPA) and the Time to Closest Point of Approach (TCPA) are: ; ; Wherein, is the relative bearing angle, is the magnitude of the relative speed, are the position coordinates of the own ship and the target ship respectively.

[0028] Construct a collision risk attenuation coefficient model: In the formula: , , are weight coefficients, satisfying , which can be dynamically adjusted according to visibility and water area level: increase the distance weight when visibility is good, and increase the TCPA weight when visibility is poor; is the DCPA membership function: when DCPA ≥ the safe encounter distance DSD, ; when DCPA < DSD, ; TCPA membership function: when TCPA ≥ security time threshold hour, When TCPA < hour, ; The relative distance membership function: ,in The attenuation coefficient decreases exponentially with increasing relative distance.

[0029] Set collision risk safety threshold ,when At that time, the situation is deemed to meet the safety conditions for resuming navigation. The safety threshold is set according to the size classification of the vessel: the threshold is higher for large vessels and lower for small vessels.

[0030] S3, Calculation of Ship Maneuvering Restoration Maneuverability recovery (MR) characterizes the degree to which the vessel recovers from an avoidance maneuver to a stable straight-ahead state. Its value ranges from [0,1], with higher values ​​indicating more stable maneuverability. The evaluation model is as follows: ; In the formula: , , For the weighting coefficients, satisfying The default values ​​are 0.3, 0.4, and 0.3 respectively. For speed recovery: ,in For the planned speed, To allow the maximum speed deviation; For heading stability: ,in The current steering angular velocity, The rated maximum turning angular velocity of the vessel; For rudder angle reset degree: ,in The current rudder angle, This is the maximum rudder angle of the ship.

[0031] Set the acceptable threshold for manipulation recovery. The default value is 0.75. When the vessel's maneuvering status is deemed to meet the conditions for resuming navigation, the threshold can be appropriately increased when navigating in restricted waters to ensure sufficient maneuvering margin for resuming navigation.

[0032] S4. Calculation of Route Deviation Costs The overall deviation cost is a weighted average of time and fuel costs, quantifying the efficiency loss caused by the delay in resuming operations: ; In the formula: This is the cost weighting coefficient, with a value range of [0,1], which can be adjusted according to the time priority of the voyage: voyages prioritizing timeliness increase the weighting. Cost-priority voyages reduced ; Cost of time: ,in The estimated time to resume the planned route from the current location. Reference time in units; Cost of fuel: ,in Additional voyage distance for resumption of service This refers to the fuel consumption rate per unit distance for ships. Reference fuel quantity is provided in units.

[0033] Among them, the estimated time for resumption of flights With additional sailing distance Based on the current yaw distance tactical turning radius of ships Calculated using geometric relationships: ; / Yaw distance This represents the vertical distance from the current ship position to the planned voyage segment.

[0034] S5. Solving for the optimal resumption time. Construct a constrained optimization model to transform the timing of the return voyage into a constrained optimization problem: Optimization goal: ; Constraints: 1. ; 2. ; 3. ( (The moment the collision avoidance maneuver is completed). The rolling time-domain optimization method is used to solve the problem. The situation data is updated once in each sampling period and the optimal solution is solved again. Since the yaw distance continues to increase with time, the overall cost function shows a monotonically increasing trend. Therefore, the moment when both constraints are satisfied at the same time for the first time is the theoretically optimal time to resume navigation. If the target ship has maneuvers that cause the risk to rise, we wait for the constraints to be satisfied again and select the time to achieve the target with the minimum cost.

[0035] S6, Sliding Window Validation To avoid false triggering caused by sensor data noise and instantaneous maneuvering of the target vessel, a sliding time window is set for confidence verification. The window length is N sampling periods. When N consecutive frames of data meet the collision risk and maneuverability recovery constraints and the optimal return time converges within the current interval, the final return command is output to trigger the automatic return control of the vessel. If a frame that does not meet the constraints appears within the window, the window is reset and monitoring continues. The window length N is dynamically set according to the type of water area: for open water areas with low traffic density, N is set to 3~5; for restricted water areas and waterways with high traffic density, N is set to 5~8 to improve the conservatism of decision-making.

[0036] Example 1: This embodiment uses a scenario of two bodies of water meeting at an intersection as an example to explain the implementation process of this method in detail: Assuming the vessel is a bulk carrier with a planned speed of 12 knots and a planned heading of 090°, and the target vessel is a container ship with a speed of 15 knots and a heading of 000°, the two vessels will encounter each other. After the vessel turns to starboard to avoid the target vessel, the avoidance maneuver will be completed when the relative distance reaches 2 nautical miles, and the timing for restarting the vessel will be calculated.

[0037] 1. Data Acquisition: The sampling period was 10 seconds. The data collected showed that the ship's speed was 11.5 knots, heading was 120°, rudder angle was 5°, and turning angular velocity was 0.2° / s. The target ship's position was continuously updated. The environmental visibility was good, and the wind force was level 3. 2. Collision risk calculation: Set DSD=1 nautical mile. =10min, weight =0.4、 =0.3、 =0.3, safety threshold =0.3; Initial time =0.32, slightly higher than the threshold; as the two ships move further away, at t=2min It dropped to 0.28, meeting the safety constraints for the first time; 3. Manipulation recovery calculation: set up =0.75, calculated as follows =0.75、 =0.8、 ≈0.857, overall MR≈0.802, satisfying manipulation constraints; 4. Cost Calculation: The current yaw distance is 0.6 nautical miles, the turning radius is 0.3 nautical miles, the calculated additional travel distance is 0.25 nautical miles, and the recovery time is approximately 4.5 minutes; [Settings] =0.5, overall cost =0.3, and the cost continues to increase over time; 5. Finding the optimal solution: When the time is 2 minutes, both constraints are satisfied simultaneously for the first time, and the cost is at its minimum, which is determined to be the theoretically optimal time for rerouting. 6. Sliding validation: set up =3, if the data for 3 consecutive sampling periods all meet the constraint, the final resumption command is output, and the ship begins to resume its original route.

[0038] In this embodiment, compared with the traditional fixed 2-nautical-mile re-entry threshold, this method accurately captures the risk threshold and triggers the re-entry approximately 1 minute in advance, reducing voyage delays and fuel consumption, while ensuring that the collision risk is within a safe range.

[0039] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for calculating the timing of a ship's return to navigation after autonomous collision avoidance, characterized in that, Includes the following steps: S1. Situational data acquisition: After the autonomous collision avoidance maneuver is completed, the motion parameters of the ship itself, the motion parameters of the target ship, the planned route parameters, and the navigation environment parameters are collected in real time. S2. Dynamic Collision Risk Attenuation Calculation: Based on the relative motion relationship between the ship and the target ship, calculate the real-time encounter parameters, construct a collision risk attenuation coefficient model, and output the current collision risk value. S3. Calculation of ship maneuverability recovery: Based on the ship's speed, heading, rudder angle and turning angular velocity parameters, construct a maneuverability recovery evaluation model and output the ship's current maneuverability stability. S4. Route Deviation Cost Calculation: Based on the deviation between the ship's current position and the planned route, and in conjunction with the ship's maneuverability, calculate the time cost and fuel cost of rerouting, and construct a comprehensive deviation cost function. S5. Optimal Re-entry Timing Solution: With the constraints of collision risk being below the safety threshold and handling recovery being above the qualified threshold, and with the optimization objective of minimizing the overall deviation cost, a re-entry timing decision model is constructed, and the theoretical optimal re-entry time is obtained by solving the model. S6. Sliding window verification: The confidence level of the theoretically optimal return time is verified by using continuous multi-frame situational data. When the return constraints are met for N consecutive frames, the final return command is output.

2. The method for calculating the timing of a ship's return to navigation after autonomous collision avoidance according to claim 1, characterized in that, In step S2, the real-time encounter parameters include relative distance, nearest encounter distance (DCPA), and nearest encounter time (TCPA); the collision risk attenuation coefficient model is: in, , , The weighting coefficients and ; The membership function of DCPA is 0 when DCPA is greater than the safe meeting distance, and otherwise increases linearly as DCPA decreases. This is the TCPA membership function, which takes the value of 0 when TCPA is greater than the security time threshold, and otherwise increases linearly as TCPA decreases; It is a membership function based on relative distance, which decays exponentially as the relative distance increases.

3. The method for calculating the timing of a ship's return to navigation after autonomous collision avoidance according to claim 2, characterized in that, The weighting coefficient, , Adjustments will be made dynamically based on visibility levels and waterway traffic density. The collision risk safety threshold is set in stages based on the size of the vessel and the type of water area.

4. The method for calculating the timing of a ship's return to navigation after autonomous collision avoidance according to claim 1, characterized in that, In step S3, the manipulation recovery evaluation model is: in, , , The weighting coefficients and ; The speed recovery rate is calculated based on the deviation between the current speed and the planned speed. For heading stability, it is calculated based on the ratio of the current turning angular velocity to the rated maximum turning angular velocity; The rudder angle reset degree is calculated based on the ratio of the current rudder angle to the maximum rudder angle; the control recovery degree ranges from [0,1].

5. The method for calculating the timing of a ship's return to navigation after autonomous collision avoidance according to claim 1, characterized in that, In step S4, the comprehensive deviation cost function is: in, Cost weighting coefficient; The time cost is calculated based on the ratio of the estimated resumption time to the reference time; The fuel cost is calculated based on the ratio of additional fuel consumption during resumption of service to a reference fuel consumption.

6. The method for calculating the timing of a ship's return to navigation after autonomous collision avoidance according to claim 5, characterized in that, The estimated time for resumption of navigation and the additional fuel consumption are calculated geometrically based on the current yaw distance, the ship's turning radius, and the fuel consumption rate per unit distance; the weighting coefficients... The schedule is dynamically adjusted based on the time priority of each voyage.

7. The method for calculating the timing of a ship's return to navigation after autonomous collision avoidance according to claim 1, characterized in that, In step S5, the rolling time-domain optimization method is used to solve for the optimal re-entry time. The situation data is updated and the solution is recalculated once every sampling period. When all constraints are met at the current time and the comprehensive deviation cost is in the decreasing range, the current time is determined to be the theoretically optimal re-entry time.

8. The method for calculating the timing of a ship's return to navigation after autonomous collision avoidance according to claim 1, characterized in that, In step S6, the length N of the sliding window is dynamically set according to the water area type: N is 3~5 for open water areas and 5~8 for restricted water areas; the sampling period is consistent with the ship sensor data update period.

9. The method for calculating the timing of a ship's return to navigation after autonomous collision avoidance according to claim 1, characterized in that, In step S1, the motion parameters of the ship itself include speed, heading, rudder angle, turning angular velocity, and position coordinates; the motion parameters of the target ship include speed, heading, position coordinates, and ship dimensions; and the navigation environment parameters include wind speed, wind direction, current speed, current direction, and visibility level.

Images