A ship berthing risk identification method and system based on multi-factor coupling and stage perception

Abstract

This invention provides a method and system for identifying ship berthing risks based on multi-factor coupling and stage perception, belonging to the field of intelligent ship autonomous berthing and risk identification. The method includes: constructing a multi-dimensional risk feature vector based on environmental disturbances, inherent static attributes of the ship, and dynamic state; introducing a coupling amplification factor to nonlinearly couple and amplify the environmental disturbance index and the dynamic state index, adjusting the distance-to-shore ratio; calculating a stage division index based on the coupled and adjusted distance-to-shore ratio, dividing the berthing process into five stages: long-distance approach, pre-berthing preparation, deceleration maneuvering, very close-to-shore berthing, and mooring; using a stage-perception gating function combined with a membership function to achieve dynamic weight allocation, integrating expert experience weights and data-driven weights to adaptively adjust the risk weights for different berthing stages; and using fuzzy logic reasoning combined with a weighted approach to perform risk defuzzification processing, outputting a continuous risk score with decision-making reference value.

Description

Technical Field

[0001] This invention relates to the field of intelligent ship autonomous berthing and risk identification technology, and more particularly to a ship berthing risk identification method and system based on multi-factor coupling and stage perception. Background Technology

[0002] Vessel berthing operations are a core component of port operations, and their safety and level of intelligence directly impact shipping efficiency and port operational safety. The development of modern intelligent shipping technologies has placed higher demands on risk assessment during berthing. Berthing involves complex interactions among multiple dimensions of indicators, including environmental factors (wind, current, waves), static vessel parameters (ship type, dimensions), and dynamic states (speed, heading, distance from shore). These factors exhibit nonlinear coupling relationships and show significant time-varying characteristics as the berthing phase progresses. Particularly in the very near-shore phase, the combined effects of spatial constraints and maneuvering behaviors significantly increase the difficulty of risk identification.

[0003] In existing technologies, berthing risk assessment mainly employs fuzzy inference or single-factor static models, which struggle to effectively handle the nonlinear coupling of multi-source data. While data mining-based methods have improved the accuracy of risk identification to some extent, they still have the following limitations: First, existing methods lack the ability to segment and dynamically perceive the entire berthing process, failing to adapt to changes in risk characteristics at different stages; second, existing models are insufficient in their coordinated response to environmental disturbances and the ship's dynamic state, resulting in low sensitivity to warnings of abnormal conditions; finally, traditional risk assessment frameworks are difficult to deploy in a lightweight manner, failing to meet the real-time requirements of intelligent ship systems. These technical deficiencies limit the improvement of the safety performance of intelligent berthing systems.

[0004] Therefore, there is an urgent need to propose a ship berthing risk identification framework based on multi-factor coupling and stage perception. This framework constructs multi-dimensional risk characteristics of environmental disturbances, ship static and dynamic indicators, designs nonlinear coupling factors to enhance risk response capabilities, and introduces berthing stage division and dynamic weight adjustment mechanisms to achieve accurate dynamic assessment of risks throughout the entire berthing process. This framework supports lightweight real-time deployment, possesses strong anomaly identification and risk early warning capabilities, provides effective safety assurance and decision support for intelligent ships' autonomous berthing, and meets the higher requirements of modern ports for intelligent berthing risk management. Summary of the Invention

[0005] To address the shortcomings of existing technologies in ship berthing risk identification and control, this invention provides a method and system for ship berthing risk identification based on multi-factor coupling and stage perception. This invention constructs multi-dimensional risk characteristics of environmental disturbances, ship static and dynamic indicators, designs nonlinear coupling factors to enhance risk response capabilities, and introduces berthing stage division and dynamic weight adjustment mechanisms to achieve accurate dynamic assessment of risks throughout the entire berthing process.

[0006] The technical means employed in this invention are as follows:

[0007] A method for identifying ship berthing risks based on multi-factor coupling and stage perception includes: S1. Construct a multi-dimensional risk feature vector based on environmental disturbances, inherent static attributes of ships and dynamic states. Among them, the environmental disturbance index is a normalized weighted combination of wind speed, sea current speed and wave height. The static index is combined with ship length, ship width and ship type correction coefficient. The dynamic index is combined with distance from shore ratio, speed and turning rate and considers the spatial constraint compression effect. S2. Introduce a coupling amplification factor to nonlinearly couple and amplify the environmental disturbance index and the dynamic state index, and adjust the distance from the shoreline ratio accordingly. S3. Based on the coupled adjustment of the distance-to-shore ratio, the stage division index is calculated, and the berthing process is divided into five stages: long-distance approach, pre-berthing preparation, deceleration and maneuvering, very close-to-shore berthing and mooring positioning. S4. Dynamic weight allocation is achieved by using a phase-aware gating function combined with a membership function, integrating expert experience weights and data-driven weights, and adaptively adjusting the risk weights for different berthing stages. S5. Use fuzzy logic reasoning combined with weighted thinking to defuzzify risk and output a continuous risk score with decision-making reference value.

[0008] Further, step S1 includes: S11. Define environmental disturbance indicators The formula is as follows:

[0009] in, Indicates wind speed. This indicates the maximum wind speed. Indicates the speed of seawater flow. Indicates the maximum seawater flow velocity. Indicates the height of the ocean waves. Represents the maximum wave height; , , These are the weighting coefficients for wind speed, seawater current speed, and wave height, respectively. S12. Define the inherent static parameters of a ship. This formula is used to quantify the impact of ship dimensions and hull type on maneuvering inertia.

[0010] in, Indicates the captain, Indicates the maximum length of the ship. Indicates the width of the ship. Indicates the maximum beam of the ship. Indicates the ship type correction factor; , , These are the weighting coefficients for ship length, ship width, and ship type correction factors, respectively. S13. Define dynamic indicators by combining distance-to-shore ratio, speed, and turning rate. This is used to depict the motion state and spatial position relationship of a ship in real time, and the formula is as follows:

[0011] in, Indicates the distance from the shore. Indicates speed, Indicates the maximum speed. Indicates the turning rate. Indicates the maximum steering ratio. , , All of these represent dynamic weighting coefficients.

[0012] Further, step S2 includes: S21. Inherent static indicators of the ship The trapezoidal membership function is used to partition fuzzy subsets, as defined below:

[0013] in, Indicates the inherent static properties of a ship The membership function value of a low-fuzzy subset. This represents the lower boundary threshold of the trapezoidal membership function. The upper boundary threshold of the trapezoidal membership function is represented; S22, Dynamic Indicators and environmental disturbance indicators Gaussian membership functions are then used to simulate the fuzzy perception characteristics of a driver in continuous motion:

[0014] in, This represents the input variables, namely dynamic indicators and environmental disturbance indicators; Indicates input variables The membership function value of a fuzzy subset. Indicates the center of Gauss. A coefficient representing the width of the membership degree; S23. Considering the multi-factor nonlinear coupling characteristics of berthing risk, a coupling amplification factor is introduced to correct the weights, defined as:

[0015] in, This represents the coupling amplification factor, used for nonlinear coupling amplification of environmental disturbances and dynamic state indicators. Indicates the scale of magnification. Represents the coupling strength coefficient. This represents the smoothing factor.

[0016] Further, step S3 includes: S31. To enhance the sensitivity of the distance-to-shore ratio to stage division, a coupling amplification factor is introduced to nonlinearly adjust the distance-to-shore ratio, and the stage division index is calculated as follows:

[0017] in, Indicates the stage division index; S32. Based on the aforementioned stage division index, the berthing process is divided into five stages: long-distance approach, pre-berthing preparation, deceleration maneuvering, very close-to-shore berthing, and mooring positioning, as shown in the following formula:

[0018] in, , , , All of these represent preset risk thresholds, corresponding to the boundaries of different stages.

[0019] Further, step S4 includes: S41. Employ a phase-aware gating function to dynamically adjust the risk weights for five phases: long-distance approach, pre-berthing preparation, deceleration maneuvering, very close-to-shore berthing, and mooring positioning. The expression is as follows:

[0020] in, Indicates the first The membership functions corresponding to each stage are used to achieve a smooth transition using trapezoidal or Gaussian functions. This represents the stage weight adjustment coefficient; S42. Combining rule weights, calculate the dynamic weights, as shown in the following expression:

[0021] in, The ratio of expert experience weighting to data-driven weighting. , The weights are derived from expert knowledge and historical data mining, respectively.

[0022] Further, step S5 includes: A weighted approach is used for defuzzification to obtain a continuous risk score with decision-making reference value. The calculation formula is as follows:

[0023] in, Indicates the number of activated rules. Indicates the weight of the corresponding rule. This indicates the centroid value of the output fuzzy set.

[0024] This invention also provides a ship berthing risk identification system based on the above-mentioned ship berthing risk identification method based on multi-factor coupling and stage perception, comprising a risk feature construction module, a multi-factor nonlinear coupling module, a stage division module, a stage perception weight adjustment module, and a risk assessment module, wherein: The risk feature construction module is used to construct a multi-dimensional risk feature vector based on environmental disturbances, inherent static attributes of the ship, and dynamic state. The environmental disturbance index is a normalized weighted average of wind speed, sea current speed, and wave height. The static index is combined with ship length, ship width, and ship type correction coefficients. The dynamic index is combined with distance-to-shore ratio, speed, and turning rate, and considers the spatial constraint compression effect. A coupling amplification factor is introduced to nonlinearly couple and amplify the environmental disturbance index and the dynamic state index, and to couple and adjust the distance-to-shore ratio. The multi-factor nonlinear coupling module introduces a coupling amplification factor to nonlinearly couple and amplify the environmental disturbance index and the dynamic state index, thereby adjusting the distance from the shore. The phase division module calculates the phase division index based on the coupled and adjusted distance-to-shore ratio, dividing the berthing process into five phases: long-distance approach, pre-berthing preparation, deceleration and maneuvering, very close-to-shore berthing, and mooring. The stage-aware weight adjustment module is used to achieve dynamic weight allocation by combining a stage-aware gating function with a membership function, integrating expert experience weights and data-driven weights, and adaptively adjusting the risk weights for different berthing stages. The risk assessment module is used to defuzzify risks by combining fuzzy logic reasoning with weighted mental methods, and outputs continuous risk scores that have decision-making reference value.

[0025] Compared with the prior art, the present invention has the following advantages: 1. This invention significantly improves the sensitivity and accuracy of risk identification under near-shore low-speed conditions by constructing a multi-dimensional risk feature vector and a nonlinear coupling mechanism. It effectively integrates environmental disturbances, ship static attributes and dynamic state information, enhances the ability to characterize the multi-factor interaction in complex and restricted waters, and improves the comprehensiveness and scientific nature of risk assessment.

[0026] 2. This invention designs a berthing stage division and dynamic weight adjustment mechanism based on the distance-to-shore ratio, achieving spatiotemporal focus and stage adaptability of risk assessment results. By dynamically adjusting risk weights through a stage-aware gating function, risk fluctuations in non-critical stages are effectively suppressed, enhancing the responsiveness to sudden risk changes in critical berthing stages and improving the timeliness and stability of risk identification.

[0027] 3. This invention employs a risk fusion model based on fuzzy reasoning, combining expert knowledge with historical data. It utilizes a multi-rule fuzzy system for nonlinear defuzzification, outputting a continuous and interpretable comprehensive risk score. This mechanism improves the system's sensitivity and robustness in identifying abnormal risks such as power failure and heading jitter, providing reliable risk warnings and auxiliary decision support for intelligent ships' autonomous berthing.

[0028] 4. The system of this invention has a reasonable structure and efficient module collaboration. It supports dynamic risk updates based on real-time AIS trajectory and environmental parameters, and has good real-time response capabilities and online adaptability. This system is suitable for intelligent autonomous berthing risk monitoring of large vessels in complex ports and restricted waters, significantly improving the level of berthing safety assurance, and has broad engineering application value and promotion potential. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a flowchart illustrating the method architecture of the present invention. Detailed Implementation

[0031] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification, claims and accompanying drawings of this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such process, method, product or device.

[0033] like Figure 1 As shown, this invention provides a method for identifying ship berthing risks based on multi-factor coupling and stage perception, including: S1. Construct a multi-dimensional risk feature vector based on environmental disturbances, inherent static attributes of ships and dynamic states. Among them, the environmental disturbance index is a normalized weighted combination of wind speed, sea current speed and wave height. The static index is combined with ship length, ship width and ship type correction coefficient. The dynamic index is combined with distance from shore ratio, speed and turning rate and considers the spatial constraint compression effect. S2. Introduce a coupling amplification factor to nonlinearly couple and amplify the environmental disturbance index and the dynamic state index, and adjust the distance from the shoreline ratio accordingly. S3. Based on the coupled adjustment of the distance-to-shore ratio, the stage division index is calculated, and the berthing process is divided into five stages: long-distance approach, pre-berthing preparation, deceleration and maneuvering, very close-to-shore berthing and mooring positioning. S4. Dynamic weight allocation is achieved by using a phase-aware gating function combined with a membership function, integrating expert experience weights and data-driven weights, and adaptively adjusting the risk weights for different berthing stages. S5. Use fuzzy logic reasoning combined with weighted thinking to defuzzify risk and output a continuous risk score with decision-making reference value.

[0034] In a specific implementation, as a preferred embodiment of the present invention, step S1 includes: S11. Define environmental disturbance indicators The formula is as follows:

[0035] in, Indicates wind speed. This indicates the maximum wind speed. Indicates the speed of seawater flow. Indicates the maximum seawater flow velocity. Indicates the height of the ocean waves. Represents the maximum wave height; , , The weighting coefficients are wind speed, seawater flow speed, and wave height, respectively. In this embodiment, the weighting coefficients of the environmental disturbance index satisfy the following conditions: wind load weight is greater than flow load weight, flow load weight is greater than wave load weight, and the sum of the weights of the three is 1.

[0036] S12. Define the inherent static parameters of a ship. This formula is used to quantify the impact of ship dimensions and hull type on maneuvering inertia.

[0037] in, Indicates the captain, Indicates the maximum length of the ship. Indicates the width of the ship. Indicates the maximum beam of the ship. Indicates the ship type correction factor; , , These are the weighting coefficients for ship length, ship width, and ship type correction factors, respectively. S13. Define dynamic indicators by combining distance-to-shore ratio, speed, and turning rate. This is used to depict the motion state and spatial position relationship of a ship in real time, and the formula is as follows:

[0038] in, Indicates the distance from the shore. Indicates speed, Indicates the maximum speed. Indicates the turning rate. Indicates the maximum steering ratio. , , All represent dynamic weighting coefficients. In this embodiment, the distance-to-shore ratio in the dynamic status index is expressed in an exponentially compressed form to reflect the nonlinear impact of spatial constraints on risk when approaching the dock.

[0039] In a specific implementation, as a preferred embodiment of the present invention, step S2 includes: S21. In order to better reflect the fuzzy characteristics of different physical properties, the inherent static indicators of the ship will be... The trapezoidal membership function is used to partition fuzzy subsets, as defined below:

[0040] in, Indicates the inherent static properties of a ship The membership function value of a low-fuzzy subset. This represents the lower boundary threshold of the trapezoidal membership function. The upper boundary threshold of the trapezoidal membership function is represented; S22, Dynamic Indicators and environmental disturbance indicators Gaussian membership functions are then used to simulate the fuzzy perception characteristics of a driver in continuous motion:

[0041] in, This represents the input variables, namely dynamic indicators and environmental disturbance indicators; Indicates input variables The membership function value of a fuzzy subset. Indicates the center of Gauss. A coefficient representing the width of the membership degree; S23. Considering the multi-factor nonlinear coupling characteristics of berthing risk, a coupling amplification factor is introduced to correct the weights, defined as:

[0042] in, This represents the coupling amplification factor, used for nonlinear coupling amplification of environmental disturbances and dynamic state indicators. Indicates the scale of magnification. Represents the coupling strength coefficient. This represents the smoothing factor. In this embodiment, the coupling amplification factor adopts a hyperbolic tangent function with two parameters, including the coupling strength coefficient and the smoothing factor.

[0043] In a specific implementation, as a preferred embodiment of the present invention, step S3 includes: S31. To enhance the sensitivity of the distance-to-shore ratio to stage division, a coupling amplification factor is introduced to nonlinearly adjust the distance-to-shore ratio, and the stage division index is calculated as follows:

[0044] in, Indicates the stage division index; S32. Based on the aforementioned stage division index, the berthing process is divided into five stages: long-distance approach, pre-berthing preparation, deceleration maneuvering, very close-to-shore berthing, and mooring positioning, as shown in the following formula:

[0045] in, , , , All of these represent preset risk thresholds, corresponding to the boundaries of different stages.

[0046] In a specific implementation, as a preferred embodiment of the present invention, step S4 includes: S41. Employ a phase-aware gating function to dynamically adjust the risk weights for five phases: long-distance approach, pre-berthing preparation, deceleration maneuvering, very close-to-shore berthing, and mooring positioning. The expression is as follows:

[0047] in, Indicates the first The membership functions corresponding to each stage are used to achieve a smooth transition using trapezoidal or Gaussian functions. This represents the stage weight adjustment coefficient. In this embodiment, the stage-aware gating function is implemented using a trapezoidal or Gaussian membership function, and the weight adjustment coefficient is determined based on expert experience and historical data.

[0048] S42. Combining rule weights, calculate the dynamic weights, as shown in the following expression:

[0049] in, The ratio of expert experience weighting to data-driven weighting. , The weights are derived from expert knowledge and historical data mining, respectively. In this embodiment, dynamic weight fusion uses a linear weighting method to fuse expert experience weights and data-driven weights, and the fusion ratio can be dynamically adjusted.

[0050] In a specific implementation, as a preferred embodiment of the present invention, step S5 includes: A weighted approach is used for defuzzification to obtain a continuous risk score with decision-making reference value. The calculation formula is as follows:

[0051] in, Indicates the number of activated rules. Indicates the weight of the corresponding rule. This represents the centroid value of the output fuzzy set. In this embodiment, the risk scoring uses a weighted centroid defuzzification method based on activation rule weights and the centroid of the output fuzzy set to achieve continuous quantitative output of risk.

[0052] Corresponding to the ship berthing risk identification method based on multi-factor coupling and stage awareness in this application, this application also provides a ship berthing risk identification system based on multi-factor coupling and stage awareness, including a risk feature construction module, a multi-factor nonlinear coupling module, a stage division module, a stage awareness weight adjustment module, and a risk assessment module, wherein: The risk feature construction module is used to construct a multi-dimensional risk feature vector based on environmental disturbances, inherent static attributes of the ship, and dynamic state. The environmental disturbance index is a normalized weighted average of wind speed, sea current speed, and wave height. The static index is combined with ship length, ship width, and ship type correction coefficients. The dynamic index is combined with distance-to-shore ratio, speed, and turning rate, and considers the spatial constraint compression effect. A coupling amplification factor is introduced to nonlinearly couple and amplify the environmental disturbance index and the dynamic state index, and to couple and adjust the distance-to-shore ratio. The multi-factor nonlinear coupling module introduces a coupling amplification factor to nonlinearly couple and amplify the environmental disturbance index and the dynamic state index, and adjust the distance from the shore. In this embodiment, the system is designed with a multi-factor nonlinear coupling module, which uses a coupling amplification factor to enhance the interaction between environmental disturbance and ship dynamic state, and significantly improves the risk identification sensitivity under near-shore and low-speed conditions.

[0053] The phase division module calculates a phase division index based on the coupled and adjusted distance-to-shore ratio, dividing the berthing process into five phases: long-distance approach, pre-berthing preparation, deceleration maneuvering, very close-to-shore berthing, and mooring positioning. In this embodiment, based on the distance-to-shore ratio and coupling characteristics, the system divides the berthing process into five key phases: long-distance approach, pre-berthing preparation, deceleration maneuvering, very close-to-shore berthing, and mooring positioning, constructing a phase division module to provide clear spatiotemporal phase support for risk assessment.

[0054] The stage-aware weight adjustment module is used to achieve dynamic weight allocation by combining a stage-aware gating function with a membership function, integrating expert experience weights and data-driven weights, and adaptively adjusting the risk weights of different berthing stages. In this embodiment, in order to dynamically adjust the risk weights, the system introduces a stage-aware weight adjustment module, which uses gating functions and membership functions to achieve adaptive adjustment of stage-related risk weights, suppress risk fluctuations in non-critical stages, and enhance the risk response capability of critical stages in the berthing process.

[0055] The risk assessment module utilizes fuzzy logic reasoning combined with a weighted mindset to defuzzify risks, outputting a continuous risk score with decision-making reference value. In this embodiment, the risk assessment module combines expert knowledge and historical data, employing a multi-rule fuzzy system and a weighted mindset to achieve nonlinear defuzzification of multi-factor coupled risks, outputting a continuous and interpretable comprehensive risk score.

[0056] This invention's system can adjust the parameters of each module in real time according to the actual berthing environment, including weighting coefficients, membership function parameters, and coupling factors, to improve the system's adaptability and robustness. The system supports dynamic risk updates based on real-time AIS trajectories and environmental parameters, possessing strong capabilities for identifying and warning of abnormal risks, effectively covering typical abnormal states such as power failure and heading turbulence. Through port-based measured data and verification in abnormal scenarios, the system demonstrates superior dynamic risk capture capabilities and robustness, significantly improving the safety assurance level of intelligent ships' autonomous berthing.

[0057] The system of this invention has a reasonable structure and efficient module collaboration. It has good real-time response capability and online adaptability. It is suitable for safety risk monitoring and auxiliary decision-making of large ships berthing in complex and restricted waters, and provides reliable technical support and engineering application foundation for intelligent ship berthing risk management.

[0058] In summary, this embodiment achieves dynamic monitoring and precise early warning of risks throughout the entire ship berthing process by constructing a multi-factor, nonlinearly coupled, stage-aware risk identification framework. This method not only enhances the ability to perceive risks under extremely near-shore, low-speed conditions but also improves the response speed to risk changes in key stages. Experimental verification shows that the model can effectively identify typical abnormal risks, providing strong safety assurance and decision support for intelligent ships' autonomous berthing.

[0059] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying ship berthing risks based on multi-factor coupling and stage perception, characterized in that, include: S1. Construct a multi-dimensional risk feature vector based on environmental disturbances, inherent static attributes of ships and dynamic states. Among them, the environmental disturbance index is a normalized weighted combination of wind speed, sea current speed and wave height. The static index is combined with ship length, ship width and ship type correction coefficient. The dynamic index is combined with distance from shore ratio, speed and turning rate and considers the spatial constraint compression effect. S2. Introduce a coupling amplification factor to nonlinearly couple and amplify the environmental disturbance index and the dynamic state index, and adjust the distance from the shoreline ratio accordingly. S3. Based on the coupled adjustment of the distance-to-shore ratio, the stage division index is calculated, and the berthing process is divided into five stages: long-distance approach, pre-berthing preparation, deceleration and maneuvering, very close-to-shore berthing and mooring positioning. S4. Dynamic weight allocation is achieved by using a phase-aware gating function combined with a membership function, integrating expert experience weights and data-driven weights, and adaptively adjusting the risk weights for different berthing stages. S5. Use fuzzy logic reasoning combined with weighted thinking to defuzzify risk and output a continuous risk score with decision-making reference value.

2. The ship berthing risk identification method based on multi-factor coupling and stage perception according to claim 1, characterized in that, Step S1 includes: S11. Define environmental disturbance indicators The formula is as follows: in, Indicates wind speed. This indicates the maximum wind speed. Indicates the speed of seawater flow. Indicates the maximum seawater flow velocity. Indicates the height of the ocean waves. Represents the maximum wave height; , , These are the weighting coefficients for wind speed, seawater current speed, and wave height, respectively. S12. Define the inherent static parameters of a ship. This formula is used to quantify the impact of ship dimensions and hull type on maneuvering inertia. in, Indicates the captain, Indicates the maximum length of the ship. Indicates the width of the ship. Indicates the maximum beam of the ship. Indicates the ship type correction factor; , , These are the weighting coefficients for ship length, ship width, and ship type correction factors, respectively. S13. Define dynamic indicators by combining distance-to-shore ratio, speed, and turning rate. This is used to depict the motion state and spatial position relationship of a ship in real time, and the formula is as follows: in, Indicates the distance from the shore. Indicates speed, Indicates the maximum speed. Indicates the turning rate. Indicates the maximum steering ratio. , , All of these represent dynamic weighting coefficients.

3. The ship berthing risk identification method based on multi-factor coupling and stage perception according to claim 1, characterized in that, Step S2 includes: S21. Inherent static indicators of the ship The trapezoidal membership function is used to partition fuzzy subsets, as defined below: in, Indicates the inherent static properties of a ship The membership function value of a low-fuzzy subset. This represents the lower boundary threshold of the trapezoidal membership function. The upper boundary threshold of the trapezoidal membership function is represented; S22, Dynamic Indicators and environmental disturbance indicators Gaussian membership functions are then used to simulate the fuzzy perception characteristics of a driver in continuous motion: in, This represents the input variables, namely dynamic indicators and environmental disturbance indicators; Indicates input variables The membership function value of a fuzzy subset. Indicates the center of Gauss. A coefficient representing the width of the membership degree; S23. Considering the multi-factor nonlinear coupling characteristics of berthing risk, a coupling amplification factor is introduced to correct the weights, defined as: in, This represents the coupling amplification factor, used for nonlinear coupling amplification of environmental disturbances and dynamic state indicators. Indicates the scale of magnification. Represents the coupling strength coefficient. This represents the smoothing factor.

4. The ship berthing risk identification method based on multi-factor coupling and stage perception according to claim 1, characterized in that, Step S3 includes: S31. To enhance the sensitivity of the distance-to-shore ratio to stage division, a coupling amplification factor is introduced to nonlinearly adjust the distance-to-shore ratio, and the stage division index is calculated as follows: in, Indicates the stage division index; S32. Based on the aforementioned stage division index, the berthing process is divided into five stages: long-distance approach, pre-berthing preparation, deceleration maneuvering, very close-to-shore berthing, and mooring positioning, as shown in the following formula: in, , , , All of these represent preset risk thresholds, corresponding to the boundaries of different stages.

5. The ship berthing risk identification method based on multi-factor coupling and stage perception according to claim 1, characterized in that, Step S4 includes: S41. Employ a phase-aware gating function to dynamically adjust the risk weights for five phases: long-distance approach, pre-berthing preparation, deceleration maneuvering, very close-to-shore berthing, and mooring positioning. The expression is as follows: in, Indicates the first The membership functions corresponding to each stage are used to achieve a smooth transition using trapezoidal or Gaussian functions. This represents the stage weight adjustment coefficient; S42. Combining rule weights, calculate the dynamic weights, as shown in the following expression: in, The ratio of expert experience weighting to data-driven weighting. , The weights are derived from expert knowledge and historical data mining, respectively.

6. The ship berthing risk identification method based on multi-factor coupling and stage perception according to claim 1, characterized in that, Step S5 includes: A weighted approach is used for defuzzification to obtain a continuous risk score with decision-making reference value. The calculation formula is as follows: in, Indicates the number of activated rules. Indicates the weight of the corresponding rule. This indicates the centroid value of the output fuzzy set.

7. A ship berthing risk identification system based on multi-factor coupling and stage perception, implemented according to the ship berthing risk identification method based on multi-factor coupling and stage perception as described in any one of claims 1-6, characterized in that, It includes a risk feature construction module, a multi-factor nonlinear coupling module, a stage division module, a stage perception weight adjustment module, and a risk assessment module, among which: The risk feature construction module is used to construct a multi-dimensional risk feature vector based on environmental disturbances, inherent static attributes of the ship, and dynamic state. The environmental disturbance index is a normalized weighted average of wind speed, sea current speed, and wave height. The static index is combined with ship length, ship width, and ship type correction coefficients. The dynamic index is combined with distance-to-shore ratio, speed, and turning rate, and considers the spatial constraint compression effect. A coupling amplification factor is introduced to nonlinearly couple and amplify the environmental disturbance index and the dynamic state index, and to couple and adjust the distance-to-shore ratio. The multi-factor nonlinear coupling module introduces a coupling amplification factor to nonlinearly couple and amplify the environmental disturbance index and the dynamic state index, thereby adjusting the distance from the shore. The phase division module calculates the phase division index based on the coupled and adjusted distance-to-shore ratio, dividing the berthing process into five phases: long-distance approach, pre-berthing preparation, deceleration and maneuvering, very close-to-shore berthing, and mooring. The stage-aware weight adjustment module is used to achieve dynamic weight allocation by combining a stage-aware gating function with a membership function, integrating expert experience weights and data-driven weights, and adaptively adjusting the risk weights for different berthing stages. The risk assessment module is used to defuzzify risks by combining fuzzy logic reasoning with weighted mental methods, and outputs continuous risk scores that have decision-making reference value.

Images