A method for controlling the safety clearance of a floating body considering bottom suction and first sinking effects.

Abstract

This invention discloses a method for controlling the safe clearance of a floating body considering bottoming and bowing effects, belonging to the fields of marine engineering, shallow water navigation control, and intelligent collision avoidance technology. The invention includes: data acquisition of the construction floating body and the environment; establishing a seabed coordinate system and a set of assessment points; calculating static geometric clearance; calculating bottoming and bowing effects; superimposing dynamic vertical displacements caused by waves and attitude; real-time assessment and short-term prediction of the minimum safe clearance; risk classification and early warning; active collision avoidance control quantity set and constraint modeling; active collision avoidance control strategy generation; result output and log traceability. This method calculates the dynamic clearance between key points of the floating body and the seabed in real time through multi-source data fusion, introduces a bottoming and bowing model and a short-term prediction algorithm, and links the safe clearance constraint closed-loop to active collision avoidance control strategies such as speed, heading, and ballast. This achieves accurate assessment, early warning, and autonomous avoidance of bottoming risks in shallow water operations.

Description

Technical Field

[0001] This invention relates to the fields of marine engineering, shallow water navigation control, and intelligent collision avoidance. Specifically, it relates to a method for controlling the safe clearance of a floating body that takes into account the effects of bottoming and sinking. Background Technology

[0002] Nearshore engineering construction often takes place in harbor basins, channels, cofferdams, and adjacent tidal flats. The water depth is significantly affected by tides, siltation, topographic relief, and construction disturbances. Engineering vessels or floating bodies often have characteristics such as deep draft, large dimensions, obvious blunt body features, and large changes in operating attitude. In shallow water conditions, floating bodies will experience sinking and bow sinkage when navigating or operating. Combined with the heave, pitching, and changes in operating load caused by waves, the minimum safe clearance between the floating body and the seabed may decrease sharply in a short period of time, leading to risks such as bottoming, scraping, or bottom attachments touching the bottom. Existing safety management relies mainly on experience margins or static water depth checks, which are difficult to cover dynamic operating conditions.

[0003] The main methods of existing technologies have the following limitations: (1) Static or quasi-static safety clearance verification; In practice, the "design draft + static water depth + safety margin" is generally used to judge the workability of shallow water or to make decisions by combining tide tables and depth sounding data; This type of method is based on static geometric relationships and is difficult to quantify the dynamic clearance fluctuations caused by bottom suction, first sinking, wave motion response, changes in working load and attitude control actions. It is easy to cause the construction efficiency to be affected when the margin is too large, and there is a risk of underreporting when the margin is insufficient or the environment changes suddenly; (2) Offline evaluation based on hydrodynamic simulation; Existing studies mostly use potential flow or viscous flow numerical methods to obtain the results of floating body motion response and wave load, and analyze the impact of operation in offline working conditions; However, this type of simulation focuses on hydrodynamic response prediction and lacks real-time evaluation operators and online prediction mechanisms for "floating body-seabed safety clearance". It is also difficult to update in real time with on-site data such as tide level, siltation and working attitude, and cannot be directly used for real-time collision avoidance control in the construction process; (3) Conventional collision avoidance and positioning control methods Engineering vessels are often equipped with dynamic positioning or maneuvering control systems, which are mostly aimed at maintaining position, heading and track, and focus on horizontal positioning accuracy. For shallow water bottoming risk, only threshold alarms or manual procedures are used to limit speed / operation. There is a lack of special modeling for vertical risks caused by bottoming and bow sinking. The control quantities such as speed, heading, ballast, pretension and operation actions are not integrated with safety clearance constraints to form a closed loop, making it difficult to provide timely and executable active collision avoidance actions. (4) Insufficient overall performance. The existing technology has the following defects: safety clearance assessment is mainly based on static or empirical margins, and lacks real-time quantification of bottoming, bow sinking and wave dynamic effects. Hydrodynamic simulation is carried out offline, and there is a lack of an integrated algorithm of "clearance calculation-short-term prediction-risk classification" that is updated in real time on site. The existing positioning / maneuvering control does not explicitly incorporate the minimum safety clearance constraint into the control target, and lacks active collision avoidance strategies oriented towards executable actions. There is a lack of traceable data chains and verification basis, making it difficult to provide consistent support in construction management, verification review and accident review.

[0004] Therefore, there is an urgent need for a method that can assess the minimum safe clearance between the bottom of the floating body and the seabed in real time, make short-term predictions, and actively generate collision avoidance control quantities. Summary of the Invention

[0005] The purpose of this invention is to overcome the above-mentioned problems existing in the prior art and to greatly improve its technical effect based on the original technology; to this end, this invention provides a method for controlling the safety gap of a floating body considering the effects of bottom suction and first sinking, the method comprising: Step S100, Data Acquisition of Construction Floating Body and Environment: Collect and synchronize multi-source data including floating body status, sea state, water depth, and seabed to form a unified time series; the floating body status includes: draft H, speed V, heading Ψ, and six-degree-of-freedom attitude vector x(t) = [η1, η2, ..., η6] TAnd key point locations; where t represents time t, T is the transpose of the vector, and η1 to η6 refer to the sway, roll, heave, pitch, pitch and yaw of the buoy, respectively; the sea state includes wave parameters and current velocity U. c The water depth and seabed include the measured tidal level Z. tide (t) and seabed elevation data; the key point locations include: the bow, bottom and auxiliary components of the construction floating body; the unified time series is achieved by timestamp alignment; the timestamp alignment adopts at least one of interpolation method and synchronization method based on common reference clock to achieve the fusion of data from different sensors under the same time reference.

[0006] Step S200: Establish the seabed coordinate system and evaluation point set: In the world coordinate system, establish the seabed surface function Z based on the seabed elevation data obtained in S100. b (x, y); where (x, y) are the horizontal position coordinates; subsequently, key points on the bottom of the floating body and its auxiliary components are selected to form an evaluation point set P = {P i The position of each evaluation point in the ship's coordinate system is r. i The position of the evaluation point in the world coordinate system is obtained through attitude transformation:

[0007] Where p0(t) is the position of the floating body reference point in the world coordinate system, R is the attitude rotation matrix, and η4(t), η5(t) and η6(t) are the roll, pitch and yaw at time t, respectively.

[0008] The evaluation point set P includes: bow bottom, stern bottom, port and starboard bottoms, bottom of the guide frame, and edge of the moon pool; the attitude rotation matrix R is expressed in at least one of Euler angles and quaternion forms, and is calculated point-by-point based on the time-series values ​​of roll η4(t), pitch η5(t), and bow roll η6(t); the seabed surface function Z b (x, y) uses at least one of regular grid, triangular grid, irregular scatter interpolation model and multibeam point cloud reconstruction model to realize real-time query of seabed elevation corresponding to the evaluation horizontal position.

[0009] Step S300, calculate static geometric gap: based on each evaluation point P i , obtain P i Horizontal position (x) i (t), y i (t)), query the seabed surface function Z b (x, y), and considering the tide level, obtain the instantaneous water surface reference height: Among them, Z w，0The datum constant is a fixed offset between the elevation datum and the absolute reference surface, set by the engineering process.

[0010] Subsequently, the static geometric clearance expression is calculated as follows: Among them, Z bottom，i (t) represents the vertical coordinates of evaluation point i in the world coordinate system, obtained from S2.

[0011] If c 0，i Using the form of water depth + draft, it can also be expressed as:

[0012] Where h(x) i y i H(t) represents the real-time water depth at evaluation point i at time t. i (t) Evaluate the local draft of point i at time t.

[0013] Step S400, Calculate the bottom suction and first sinking effects: Calculate the total bottom suction sinking ΔZ using a calibrable parametric model based on speed, relative water depth, and channel constraints. squat and first sinking volume △Z bow The calibrable parameterized model includes: a total bottom sinking model and a first sinking model; wherein, the total bottom sinking model is expressed as:

[0014] Among them, △Z squat (t) represents the total sinking amount at time t, and k s Here, g is the empirical coefficient for bottom suction, V(t) is the acceleration due to gravity, H(t) is the speed at time t, and h is the draft at time t. eff (t) represents the effective water depth at time t, and m is the empirical index; h eff (t) is determined by a combination of static water depth, tidal correction, and local water depth at the current position of the construction buoy, and is dynamically updated according to the track position in restricted channels, shallow water areas, local erosion and deposition terrain, and seabed undulation areas; parameter k s The value of m is set according to the ship type and water type.

[0015] The initial sinking volume model is expressed as follows:

[0016]

[0017] Where, γ b γ is the first sinking ratio coefficient. b ∈[0,1], used to characterize the distribution ratio of longitudinal settlement caused by bottom suction between the bow and stern; parameter γ bOnline corrections are made based on ship type and longitudinal constraints; △Z bow (t) represents the amount of sinking of the bow at time t, ΔZ stern (t) represents the amount of stern sinking at time t.

[0018] Step S500, superposition of dynamic vertical displacement caused by waves and attitude: Calculate the instantaneous vertical changes of heave, pitch, and roll caused by waves on the evaluation point, achieved in two ways: method A based on real-time attitude measurement and method B based on predictive response; method A based on real-time attitude measurement directly obtains heave η3(t), roll η4(t), and pitch η5(t) through inertial navigation and attitude sensors, and then calculates the vertical displacement of the evaluation point as:

[0019] in, and These are the longitudinal and transverse coordinates of the evaluation point in the ship's coordinate system, respectively.

[0020] The predicted response-based method B uses at least one of the response amplitude operator RAO and the time-domain response model to predict the heave η3(t), roll η4(t), and pitch η5(t) attitudes from sea state parameters, and substitutes the predicted attitudes into ΔZ. motion，i Mapped to point displacement; the predicted heave η3(t), roll η4(t) and pitch η5(t) attitudes are also used for short-term prediction in step S600 below.

[0021] Step S600, Real-time assessment and short-term prediction of minimum safety clearance: Combining static clearance, initial sinking of the suction head, and dynamic attitude changes, the real-time safety clearance at the assessment point is obtained:

[0022] Among them, △Z squat，i △Z is applied to the bow and stern points respectively. bow and △Z stern , △Z margin For structural safety margin; The minimum safe clearance for the entire ship is: , where P is the set of evaluation points.

[0023] Short-term forecasting refers to calculating the lower bound of the forecast based on the current trend and sea state response within the forecast window τ∈[0, ΔT]. This lower bound is a conservative estimate of the minimum safe clearance that ships are predicted to reach within a future period. The formula for the lower bound is:

[0024] C i (t+τ) refers to the predicted safety gap at time t+τ in the future.

[0025] If the probabilistic lower bound is expressed in the form of mean and confidence margin, then the probabilistic lower bound is expressed as:

[0026] Where, μ cmin To predict the mean of the gap, α cmin k is the standard deviation. α denoted as the confidence coefficient; by considering tidal change plans, speed change plans, and predicted attitude response, and using Monte Carlo and Kalman filtering methods to quantify prediction uncertainties, the mean μ is obtained. cmin and standard deviation α cmin .

[0027] Step S700, Risk Classification and Early Warning: Based on the minimum implementation gap c min (t) Predicting the lower bound Risk levels are constructed based on the rate of change of intervals. The risk levels include: Safety: and

[0028] alert: and At least one of them is true; Danger: and

[0029] Among them, c safe c warn and r crit The threshold is configurable.

[0030] Risk level is based on minimum gap c min (t) Predicting the lower bound In addition to the gap change rate, a comprehensive judgment is made based on the location of the most dangerous assessment point, the dominant risk source, and the duration of the risk; the dominant risk source includes at least one of the following: bottom suction, wave-dominant, attitude-dominant, seabed abrupt change-dominant, and tidal level change-dominant; the c safe c warn and r crit Configure according to ship type and operation type, and make adjustments online.

[0031] Step S800, Active Collision Avoidance Control Quantity Set and Constraint Modeling: Define the executable control quantity vector u(t):

[0032] The elements in the set, from left to right, represent speed, heading, ballast adjustment, DP thrust distribution, mooring pretension adjustment, and operation mode, respectively. The control vector u(t) may also include one or more of the following: propeller speed, rudder angle, ballast water tank flow rate, and mooring cable length. The operation modes include: normal, slow descent, and pause.

[0033] The minimum gap constraint can be expressed as either a hard constraint or a soft constraint; where the hard constraint takes the form of:

[0034] The soft constraint form is: adding a penalty function term to the optimization objective. , where λ is the penalty coefficient.

[0035] Step S900, Active Collision Avoidance Control Strategy Generation: Upon triggering either a warning state or a dangerous state, an active collision avoidance control strategy is generated based on the real-time minimum clearance, predicted lower bound, the location of the most dangerous assessment point, and the dominant risk source. The active collision avoidance control strategy includes: speed limiting, heading adjustment, ballast leveling, preload adjustment, suspension, and deceleration operations. It is implemented using either an optimization-based or rule-based approach. The optimization-based approach requires solving the objective function in each control cycle.

[0036] Where u0 is the current control variable, λ1 is the gap violation penalty coefficient, and λ2 is the control action change penalty coefficient.

[0037] The implementation of the rule type includes: like If so, prioritize reducing speed V and adjust the heading to reduce longitudinal wave excitation; if a certain local point c i If the minimum is located at the bow, then "deceleration + ballast shift" and suspend forward movement are adopted; if it is caused by rolling, then the heading and DP thrust are adjusted to reduce rolling, and lateral operation is restricted.

[0038] The optimization method uses the bottom-suction effect calculation method, wave response prediction method, and floating body dynamics equation as prediction models to solve the optimal control sequence in the prediction time domain and executes only the action of the first control step.

[0039] Step S1000, Result Output and Log Tracking: The system outputs real-time and predicted gap curves, the most dangerous points, risk levels, triggering causes, suggested control actions, and expected improvement amounts; simultaneously, it records key version information: water depth data version, seabed model version, parameter k. s m and γ bVersion, threshold version, and control action logs; triggering reasons include at least one of bottom-fishing dominance, wave dominance, and attitude dominance; the expected improvement includes the estimated increase in real-time minimum gap after implementing control actions, the estimated increase in predicted lower bound, the decrease in risk level, the estimated time required for risk resolution, and the decomposition results of the contribution of each control action to the bottom-fishing item, attitude item, and total gap item; the log traceability supports engineering review and debriefing, recording version information and the timing, cause, and effect evaluation of control actions.

[0040] The beneficial effects of this invention are: This invention proposes a method for controlling the safe clearance of floating bodies that considers the effects of bottoming and bow sinking; the advantages of this invention are: A1, Dynamic safe clearance assessment: The safe clearance model incorporates speed-related bottoming and bow sinking amounts and longitudinal bow sinking distribution, and superimposes them with the vertical displacement of key points caused by real-time attitude (heave, pitch, and roll), which can reflect the clearance differences caused by "the same water depth, different speeds, and operating conditions" in shallow water environments, avoiding the overly conservative or missed judgments caused by the traditional method of judging only by static water depth + experience margin, and improving the accuracy and engineering applicability of bottoming risk identification; A2, Key point positioning and interpretable early warning: This invention integrates the bow, stern, sides, and accessories Key components such as the bottom of structural members are used as assessment points. Real-time calculation of clearances at each point and the minimum clearance for the entire ship enables automatic identification of the "most dangerous points" under complex attitude and seabed undulation conditions. This provides an interpretable output, moving from "is it dangerous?" to "where is it dangerous and why?", facilitating targeted on-site measures and significantly improving the operability of early warnings. A3, Short-Term Prediction and Advanced Triggering: In addition to providing instantaneous clearances, a lower bound for prediction is calculated through a short-term prediction window (which can be combined with confidence margin). The clearance change rate is introduced as a trigger condition, enabling early identification of risk trends before tidal changes, sudden sea state changes, speed changes, or operational maneuvers, reducing the lag problem of "almost hitting the bottom before issuing an alarm." Enhance proactive safety assurance capabilities during construction; A4, Closed-loop active collision avoidance control: This invention uses speed, heading, ballast adjustment, positioning thrust, maneuvering, mooring pretension, and operating modes (descent and pause, etc.) as controllable variables. After risk classification triggering, it generates an executable collision avoidance sequence, especially leveraging the sensitivity of the relationship between bottom suction and the square of speed to achieve rapid risk reduction. Compared to traditional methods relying solely on manual speed limits or simple depth alarms, this invention achieves a closed-loop linkage of "assessment-trigger-action," improving the real-time performance and effectiveness in handling complex working conditions; A5, Cost reduction and efficiency improvement: By more reliably constraining the minimum safety clearance, this invention can reduce structural damage caused by bottom scraping and collisions. Component damage and positioning deviations reduce downtime for inspection, repair, and rework costs. Simultaneously, because the evaluation model more closely approximates the actual dynamic process, it avoids wasted window time due to overly conservative static margins, improving workable time and construction efficiency while ensuring safety, resulting in significant safety and economic benefits. A6, Traceability and Review Support: This invention version-records seabed models, water level data, bottom suction, initial sinking parameters, threshold configurations, evaluation results, and control actions, and outputs explanations of risk sources (bottom suction-dominated, wave-dominated, and attitude-dominated, etc.), forming a traceable chain of evidence. This facilitates verification and accountability by construction units, supervisors, and reviewers, improving the standardization of project management. Attached Figure Description

[0041] Figure 1 This is a flowchart of a method for controlling the safety gap of a floating body that takes into account the effects of bottom suction and first sinking, according to the present invention. Detailed Implementation

[0042] The specific embodiments of the present invention will be described in more detail below with reference to the accompanying drawings; it should be understood that the specific embodiments given herein are only for illustration and explanation of the present invention, and any simple modifications based on the present invention are within the protection scope of the present invention.

[0043] like Figure 1 The diagram shows a flowchart of a method for controlling the safe clearance of a floating body considering the effects of bottoming and sinking, according to an embodiment of the present invention. The flowchart includes steps S100 to S1000. Steps S100 to S1000 mainly involve obtaining information such as water depth, tide level, seabed elevation, floating body attitude and speed, waves, and current velocity through multi-source data fusion; constructing a vertical geometric relationship model that includes bottoming and sinking terms; calculating the minimum instantaneous clearance in real time and making short-term predictions; when the lower boundary of the predicted clearance is triggered, adopting an active collision avoidance control strategy with "safety clearance constraint" as the core, and making coordinated adjustments to speed, heading, ballast, DP thrust, maneuvering, mooring pretension, and operational actions (such as lifting descent or pausing), thereby reducing the risk of bottoming and improving operational safety and efficiency.

[0044] In a specific embodiment, step S100 involves collecting and synchronizing multi-source data, including the floating body's state, sea state, water depth, and seabed, to form a unified time series. The floating body's state includes: draft H, speed V, heading Ψ, and a six-degree-of-freedom attitude vector x(t) = [η1, η2, ..., η6]. T And the locations of key points (bow, bottom, and auxiliary components of the floating body); where t represents time t, T is the transpose of the vector, and η1 to η6 refer to the sway, roll, heave, pitch, pitch, and bow of the floating body, respectively; sea state includes wave parameters and current velocity U. c Water depth and seabed, including measured tidal level Z. tide (t) and seabed elevation data; the unified time series is achieved through timestamp alignment; the timestamp alignment employs at least one of interpolation and synchronization methods based on a common reference clock to achieve the fusion of data from different sensors under the same time reference.

[0045] Step S200: Establish the seabed surface function Z based on the seabed elevation data obtained in S100 in the world coordinate system. b (x, y); where (x, y) are the horizontal position coordinates; subsequently, key points on the bottom of the floating body and its auxiliary components are selected to form an evaluation point set P = {P i Let r be the position of each evaluation point in the ship's coordinate system. i The position of the evaluation point in the world coordinate system is obtained through attitude transformation:

[0046] Where p0(t) is the position of the reference point of the floating body in the world coordinate system, and p0(t) can be regarded as the position of the center of gravity of the floating body; R is the attitude rotation matrix, and η4(t), η5(t) and η6(t) are the roll, pitch and yaw at time t, respectively.

[0047] Specifically, the evaluation point set P includes: bow bottom, stern bottom, port and starboard bottoms, bottom of the guide frame, and edge of the moon pool, etc.; the attitude rotation matrix R is expressed in at least one of Euler angles and quaternion forms, and is calculated point by point based on the time-series values ​​of roll η4(t), pitch η5(t), and bow roll η6(t); the seabed surface function Z b (x, y) uses at least one of regular grid, triangular grid, irregular scatter interpolation model and multibeam point cloud reconstruction model to realize real-time query of seabed elevation corresponding to the evaluation horizontal position.

[0048] It should be noted that the world coordinate system can also be replaced with the chart coordinate system. The coordinates of various points on the floating body and the seabed can be established through the chart coordinate system for the implementation of subsequent technical solutions.

[0049] Step S300 then proceeds based on each evaluation point P i , obtain P i Horizontal position (x) i (t), y i (t)), query the seabed surface function Z b (x, y), and considering the tide level, obtain the instantaneous water surface reference height: Among them, Z w，0 The datum constant is a fixed offset between the elevation datum and the absolute reference surface, set by the engineering process.

[0050] Subsequently, the static geometric clearance expression is calculated as follows: Among them, Z bottom，i (t) represents the vertical coordinates of evaluation point i in the world coordinate system, obtained from the world coordinate system of S2.

[0051] Furthermore, c 0，i It can also be expressed in the form of water depth + draft, which can then be represented as:

[0052] Where h(x) i y i H(t) represents the real-time water depth at evaluation point i at time t. i (t) Evaluate the local draft of point i at time t.

[0053] Step S400, Calculate the bottom suction and first sinking effects: Calculate the total bottom suction sinking ΔZ using a calibrable parametric model based on speed, relative water depth, and channel constraints. squat and first sinking volume △Z bow The calibrable parameterized model includes: a total bottom sinking model and a first sinking model; wherein, the total bottom sinking model is expressed as:

[0054] Among them, △Z squat (t) represents the total sinking amount at time t, and k s Here, g is the empirical coefficient for bottom suction, V(t) is the acceleration due to gravity, H(t) is the speed at time t, and h is the draft at time t. eff (t) represents the effective water depth at time t, h eff (t) is determined by a combination of static water depth, tidal correction, and local water depth at the current location of the construction buoy, and is dynamically updated according to the track position in restricted channels, shallow water areas, local erosion and deposition terrain, and seabed undulation areas; m is an empirical index; parameter k s The value of m is set according to the ship type and water type.

[0055] The initial sinking volume model is expressed as follows:

[0056]

[0057] Where, γ b γ is the first sinking ratio coefficient. b ∈[0,1], used to characterize the distribution ratio of longitudinal settlement caused by bottom suction between the bow and stern, which is corrected online according to the ship type and the degree of longitudinal constraint; △Z bow (t) represents the amount of sinking of the bow at time t, ΔZ stern (t) represents the amount of stern sinking at time t.

[0058] Step S500, superposition of dynamic vertical displacement caused by waves and attitude: Calculate the instantaneous vertical changes of heave, pitch, and roll caused by waves on the evaluation point, achieved in two ways: method A based on real-time attitude measurement and method B based on predictive response; method A based on real-time attitude measurement directly obtains heave η3(t), roll η4(t), and pitch η5(t) through inertial navigation and attitude sensors, and then calculates the vertical displacement of the evaluation point as:

[0059] in, and These are the longitudinal and transverse coordinates of the evaluation point in the ship's coordinate system, respectively.

[0060] Among them, method B based on predictive response uses at least one of the response amplitude operator RAO and time-domain response model to predict the heave η3(t), roll η4(t), and pitch η5(t) attitudes from sea state parameters, and substitutes the predicted attitudes into ΔZ. motion，i Mapped to point displacement; in addition, the heave η3(t), roll η4(t), and pitch η5(t) attitudes predicted from sea state parameters are also used for short-term prediction in step S600.

[0061] Step S600, Real-time assessment and short-term prediction of minimum safety clearance: Combining static clearance, initial sinking of the suction head, and dynamic attitude changes, the real-time safety clearance at the assessment point is obtained:

[0062] Among them, △Z squat，i △Z is applied to the bow and stern points respectively. bow and △Z stern , △Z margin This is for structural safety margins (depth sounding error, positioning error, and seabed uncertainty, etc.).

[0063] The minimum safe clearance for the entire ship is: , where P is the set of evaluation points.

[0064] Short-term forecasting refers to calculating the lower bound of the forecast based on the current trend and sea state response within the forecast window τ∈[0, ΔT]. This lower bound is a conservative estimate of the minimum safe clearance that ships are predicted to reach within a future period. The formula for the lower bound is:

[0065] C i (t+τ) refers to the predicted safety gap at time t+τ in the future.

[0066] In addition, step S600 provides a probabilistic lower bound expressed in the form of mean and confidence margin, wherein the probabilistic lower bound is expressed as:

[0067] Where, μ cmin To predict the mean of the gap, α cmin k is the standard deviation. α The confidence coefficient (e.g., with a 90% lower bound) is used. By considering tidal change plans, speed change plans, and expected attitude responses, and quantifying the prediction uncertainty using Monte Carlo and Kalman filtering methods, the mean μ is obtained. cmin and standard deviation α cmin .

[0068] Step S700, Risk Classification and Early Warning: Based on the minimum implementation gap cmin (t) Predicting the lower bound Risk levels are constructed based on the rate of change of intervals. The risk levels include: Safety: and

[0069] alert: and At least one of them is true.

[0070] Danger: and

[0071] Among them, c safe c warn and r crit The threshold is configurable.

[0072] Risk level is based on minimum gap c min (t) Predicting the lower bound In addition to the gap change rate, a comprehensive judgment is made based on the location of the most dangerous assessment point, the dominant risk source, and the duration of the risk; the dominant risk source includes at least one of the following: bottom suction, wave-dominant, attitude-dominant, seabed abrupt change-dominant, and tidal level change-dominant; the aforementioned c safe c warn and r crit Configure according to ship type and operation type, and adjust online to form configurable threshold conditions.

[0073] Step S800, Active Collision Avoidance Control Quantity Set and Constraint Modeling: Define the executable control quantity vector u(t):

[0074] The elements in the set, from left to right, represent speed, heading, ballast adjustment, DP thrust distribution, mooring pretension adjustment, and operating mode, respectively; the control vector u(t) also includes propeller speed, rudder angle, ballast tank flow rate, and mooring cable length; the operating modes include: normal, slow descent, and pause.

[0075] The minimum gap constraint can be expressed as either a hard constraint or a soft constraint; where the hard constraint takes the form of:

[0076] The soft constraint form is: adding a penalty function term to the optimization objective. , where λ is the penalty coefficient.

[0077] Step S900, Active Collision Avoidance Control Strategy Generation: Upon triggering either a warning state or a dangerous state, an active collision avoidance control strategy is generated based on the real-time minimum clearance, predicted lower bound, the location of the most dangerous assessment point, and the dominant risk source. The active collision avoidance control strategy includes: speed limiting, heading adjustment, ballast leveling, preload adjustment, suspension, and deceleration operations. It is implemented using either an optimization-based or rule-based approach. The optimization-based approach requires solving the objective function in each control cycle.

[0078] Where u0 is the current control variable, λ1 is the gap violation penalty coefficient, and λ2 is the control action change penalty coefficient; the optimization method uses the bottoming effect calculation method, wave response prediction method, and floating body dynamic equation as the prediction model, solves the optimal control sequence in the prediction time domain, and only executes the action of the first control step.

[0079] The implementation of the rule type includes: like Then, prioritize reducing speed V (due to bottom suction term and V). 2 (strong correlation), and adjust the heading to reduce longitudinal wave excitation; if a certain local point c i If the minimum is located at the bow, then "deceleration + ballast shift" and suspend forward movement are adopted; if it is caused by rolling, then the heading and DP thrust are adjusted to reduce rolling, and lateral operation is restricted.

[0080] In the above embodiments, the formula for the optimization strategy includes minimizing gap violations and limiting the control action amplitude; wherein the first half of the plus sign inside the integral sign is for minimizing gap violations, which serves as a collision avoidance penalty term; the second half is for limiting the control action amplitude; in minimizing gap violations, when the predicted gap c... min (t + τ; u) is greater than or equal to the safety threshold c safe When max() = 0, it indicates no penalty; when the prediction gap c min (t+τ;u) is lower than c safe When the penalty is proportional to the square of the violation, the penalty increases with the severity of the violation. This factor can drive the optimizer to select the control variable u so that the gap within the entire prediction window is as close as possible to the safety threshold, thereby avoiding hitting the bottom.

[0081] In addition, the limit control action amplitude term represents the change amplitude of the penalty control quantity u relative to the current control quantity u0; it is used to prevent the control action from being too violent (such as sudden deceleration and sharp steering), to ensure the system is stable and executable, and to reduce mechanical wear.

[0082] It should be noted that the penalty coefficient λ1 is much larger than λ2, which reflects that the priority of balancing collision avoidance is greater than the cost of controlling the magnitude of the action, ensuring safety first.

[0083] Step S1000, Result Output and Log Tracking: The system outputs real-time and predicted gap curves, the most dangerous points, risk levels, triggering causes, suggested control actions, and expected improvement amounts; simultaneously, it records key version information: water depth data version, seabed model version, parameter k. s m and γ b Version, threshold version, and control action logs; triggering reasons include at least one of bottom-fishing dominance, wave dominance, and attitude dominance; expected improvement includes the estimated increase in real-time minimum gap after implementing control actions, the estimated increase in predicted lower bound, the decrease in risk level, the estimated time required for risk resolution, and the decomposition results of the contribution of each control action to the bottom-fishing item, attitude item, and total gap item; log traceability supports engineering review and debriefing, recording version information and the timing, cause, and effect evaluation of control actions.

Claims

1. A method for controlling the safety clearance of a floating body considering the effects of bottom suction and initial sinking, characterized in that, The method includes: Step S100, Data Acquisition of Construction Floating Body and Environment: Collect and synchronize multi-source data including floating body status, sea state, water depth, and seabed to form a unified time series; the floating body status includes: draft H, speed V, heading Ψ, and six-degree-of-freedom attitude vector x(t) = [η1, η2, ..., η6] T And key point locations; where t represents time t, T is the transpose of the vector, and η1 to η6 refer to the sway, roll, heave, pitch, pitch and yaw of the buoy, respectively; the sea state includes wave parameters and current velocity U. c The water depth and seabed include the measured tidal level Z. tide (t) and seabed elevation data; Step S200: Establish the seabed coordinate system and evaluation point set: In the world coordinate system, establish the seabed surface function Z based on the seabed elevation data obtained in S100. b (x, y); where (x, y) are the horizontal position coordinates; subsequently, key points on the bottom of the floating body and its auxiliary components are selected to form an evaluation point set P = {P i The position of each evaluation point in the ship's coordinate system is r. i The position of the evaluation point in the world coordinate system is obtained through attitude transformation: Where p0(t) is the position of the floating body reference point in the world coordinate system, R is the attitude rotation matrix, and η4(t), η5(t) and η6(t) are the roll, pitch and yaw at time t, respectively. Step S300, calculate static geometric gap: based on each evaluation point P i , obtain P i Horizontal position (x) i (t), y i (t)), query the seabed surface function Z b (x, y), and considering the tide level, obtain the instantaneous water surface reference height: Among them, Z w，0 The datum constant is a fixed offset between the elevation datum and the absolute reference surface, set by the engineering process. Subsequently, the static geometric clearance expression is calculated as follows: Among them, Z bottom，i (t) represents the vertical coordinates of evaluation point i in the world coordinate system, obtained from S2; Step S400, Calculate the bottom suction and first sinking effects: Calculate the total bottom suction sinking ΔZ using a calibrable parametric model based on speed, relative water depth, and channel constraints. squat and first sinking volume △Z bow The calibrable parameterized model includes: a total bottom sinking model and a first sinking model; wherein, the total bottom sinking model is expressed as: Among them, △Z squat (t) represents the total sinking amount at time t, and k s Here, g is the empirical coefficient for bottom suction, V(t) is the acceleration due to gravity, H(t) is the speed at time t, and h is the draft at time t. eff (t) represents the effective water depth at time t, and m is the empirical index; The initial sinking volume model is expressed as follows: Where, γ b γ is the first sinking ratio coefficient. b ∈[0,1], used to characterize the distribution ratio of longitudinal settlement caused by bottom suction between the bow and stern; △Z bow (t) represents the amount of sinking of the bow at time t, ΔZ stern (t) represents the amount of stern sinking at time t; Step S500, superposition of dynamic vertical displacement caused by waves and attitude: Calculate the instantaneous vertical changes of heave, pitch, and roll caused by waves on the evaluation point, achieved in two ways: method A based on real-time attitude measurement and method B based on predictive response; method A based on real-time attitude measurement directly obtains heave η3(t), roll η4(t), and pitch η5(t) through inertial navigation and attitude sensors, and then calculates the vertical displacement of the evaluation point as: in, and These are the longitudinal and transverse coordinates of the evaluation point in the ship's coordinate system, respectively. The predicted response-based method B uses at least one of the response amplitude operator RAO and the time-domain response model to predict the heave η3(t), roll η4(t), and pitch η5(t) attitudes from sea state parameters, and substitutes the predicted attitudes into ΔZ. motion，i Mapped to point displacement; Step S600, Real-time assessment and short-term prediction of minimum safety clearance: Combining static clearance, initial sinking of the suction head, and dynamic attitude changes, the real-time safety clearance at the assessment point is obtained: Among them, △Z squat，i △Z is applied to the bow and stern points respectively. bow and △Z stern , △Z margin For structural safety margin; The minimum safe clearance for the entire ship is: Where P is the set of evaluation points Short-term forecasting refers to calculating the lower bound of the forecast based on the current trend and sea state response within the forecast window τ∈[0, ΔT]. This lower bound is a conservative estimate of the minimum safe clearance that ships are predicted to reach within a future period. The formula for the lower bound is: C i (t+τ) refers to the predicted safety gap at time t+τ in the future; Step S700, Risk Classification and Early Warning: Based on the minimum implementation gap c min (t) Predicting the lower bound Risk levels are constructed based on the rate of change of intervals. The risk levels include: Safety: and ; alert: and At least one of them is true; Danger: and ; Among them, c safe c warn and r crit The threshold is configurable; Step S800, Active Collision Avoidance Control Quantity Set and Constraint Modeling: Define the executable control quantity vector u(t): The elements in the set, from left to right, represent speed, heading, ballast adjustment, DP thrust distribution, mooring pretension adjustment, and operating mode, respectively. The minimum gap constraint can be expressed as either a hard constraint or a soft constraint; where the hard constraint takes the form of: The soft constraint form is: adding a penalty function term to the optimization objective. , where λ is the penalty coefficient; Step S900, Active Collision Avoidance Control Strategy Generation: Upon triggering either a warning state or a dangerous state, an active collision avoidance control strategy is generated based on the real-time minimum clearance, predicted lower bound, the location of the most dangerous assessment point, and the dominant risk source. The active collision avoidance control strategy includes: speed limiting, heading adjustment, ballast leveling, preload adjustment, suspension, and deceleration operations. It is implemented using either an optimization-based or rule-based approach. The optimization-based approach requires solving the objective function in each control cycle. Where u0 is the current control quantity, λ1 is the gap violation penalty coefficient, and λ2 is the control action change penalty coefficient; The implementation of the rule type includes: like If so, prioritize reducing speed V and adjust the heading to reduce longitudinal wave excitation; if a certain local point c i If the minimum is located at the bow, then "deceleration + ballast shift" and suspend forward movement are adopted; if it is caused by rolling, then the heading and DP thrust are adjusted to reduce rolling, and lateral operation is restricted. Step S1000, Result Output and Log Tracking: The system outputs real-time and predicted gap curves, the most dangerous points, risk levels, triggering causes, suggested control actions, and expected improvement amounts; simultaneously, it records key version information: water depth data version, seabed model version, parameter k. s m and γ b Version, threshold version, and control action logs.

2. The method for controlling the safety clearance of a floating body considering the effects of bottom suction and initial sinking, as described in claim 1, is characterized in that... The key locations of step S100 include: the bow, bottom, and auxiliary components of the construction floating hull; the unified time series is achieved through a timestamp alignment method; the timestamp alignment adopts at least one of interpolation method and synchronization method based on a common reference clock to achieve the fusion of data from different sensors under the same time reference.

3. The method for controlling the safety clearance of a floating body considering the effects of bottom suction and initial sinking, as described in claim 1, is characterized in that... The evaluation point set P includes: bow bottom, stern bottom, port and starboard bottoms, bottom of the guide frame, and edge of the moon pool; the attitude rotation matrix R is expressed in at least one of Euler angles and quaternion forms, and is calculated point-by-point based on the time-series values ​​of roll η4(t), pitch η5(t), and bow roll η6(t); the seabed surface function Z b (x, y) uses at least one of regular grid, triangular grid, irregular scatter interpolation model and multibeam point cloud reconstruction model to realize real-time query of seabed elevation corresponding to the evaluation horizontal position.

4. The method for controlling the safety clearance of a floating body considering the effects of bottom suction and initial sinking, as described in claim 1, is characterized in that... c in step S300 0，i Using the form of water depth + draft, it can also be expressed as: Where h(x) i y i H(t) represents the real-time water depth at evaluation point i at time t. i (t) Evaluate the local draft of point i at time t.

5. The method for controlling the safety clearance of a floating body considering the effects of bottom suction and initial sinking, as described in claim 1, is characterized in that... h in step S400 eff (t) is determined by a combination of static water depth, tidal correction, and local water depth at the current position of the construction buoy, and is dynamically updated according to the track position in restricted channels, shallow water areas, local erosion and deposition terrain, and seabed undulation areas; parameter k s And m is set according to the ship type and water type, parameter γ b The ship type and longitudinal constraints are adjusted online; the heave η3(t), roll η4(t), and pitch η5(t) attitudes predicted by the sea state parameters in step S500 are also used for short-term prediction in step S600.

6. The method for controlling the safety clearance of a floating body considering the effects of bottom suction and initial sinking, as described in claim 1, is characterized in that... Step S600 further includes: a probabilistic lower bound expressed in the form of mean and confidence margin, wherein the probabilistic lower bound is expressed as: Where, μ cmin To predict the mean of the gap, α cmin k represents the standard deviation. α denoted as the confidence coefficient; by considering tidal change plans, speed change plans, and predicted attitude response, and using Monte Carlo and Kalman filtering methods to quantify prediction uncertainties, the mean μ is obtained. cmin and standard deviation α cmin .

7. A method for controlling the safety clearance of a floating body considering bottom suction and initial sinking effects according to claim 1, characterized in that, The risk level in step S700 is based on the minimum gap c. min (t) Predicting the lower bound In addition to the gap change rate, a comprehensive judgment is made based on the location of the most dangerous assessment point, the dominant risk source, and the duration of the risk; the dominant risk source includes at least one of the following: bottom suction, wave-dominant, attitude-dominant, seabed abrupt change-dominant, and tidal level change-dominant; the c safe c warn and r crit Configure according to ship type and operation type, and make adjustments online.

8. A method for controlling the safety clearance of a floating body considering bottom suction and initial sinking effects according to claim 1, characterized in that, The control vector u(t) in step S800 also includes the thruster speed, rudder angle, ballast water tank flow rate and mooring cable length. The operating modes include: normal, slow descent and pause.

9. A method for controlling the safety clearance of a floating body considering bottom suction and initial sinking effects according to claim 1, characterized in that, In step S900, the optimization method uses the bottoming effect calculation method, wave response prediction method, and floating body dynamics equation as prediction models to solve the optimal control sequence in the prediction time domain and executes only the action of the first control step.

10. A method for controlling the safety clearance of a floating body considering bottom suction and initial sinking effects according to claim 1, characterized in that, The triggering cause of step S1000 includes at least one of bottom-fishing dominance, wave dominance, and attitude dominance; the expected improvement includes the estimated increase in the real-time minimum gap after implementing control actions, the estimated increase in the predicted lower bound, the decrease in risk level, the estimated value of the time required for risk resolution, and the decomposition results of the contribution of each control action to the bottom-fishing item, attitude item, and total gap item; the log traceability supports engineering review and debriefing, and records the information of each version and the timing, cause, and effect evaluation of control actions.

Images