A fishery heterogeneous multi-ship cooperative path planning and operation control method under high sea conditions

Abstract

This invention relates to a method for heterogeneous multi-vehicle collaborative path planning and operation control in fisheries under high sea states. The method includes: acquiring meteorological data from the Yellow and Bohai Seas; fitting the wave surface elevation spectrum from the meteorological data to a modified JONSWAP spectral model, while simultaneously decomposing bimodal features to calculate wave encounter frequency; defining a parameterized roll risk function based on the wave encounter frequency to generate a four-dimensional spatiotemporal cost map containing dynamic resonance risk regions; during the harvesting and transfer process, using an improved A algorithm to perform global path planning on the four-dimensional spatiotemporal cost map, generating a zigzag path; and during the live fish handover process in the harvesting and transfer process, establishing a cost function including a relative motion suppression term, and solving the fixed-berth and dynamic docking of multi-vehicle collaboration through rolling optimization. This invention solves the problems of neglecting wave-induced resonance risk and instability in multi-vehicle dynamic docking, significantly improving the safety and success rate of unmanned operations in large-scale offshore aquaculture facilities.

Description

Technical Field

[0001] This invention relates to the interdisciplinary fields of marine engineering, underwater robots and intelligent control, and in particular to a method for collaborative path planning and operation control of heterogeneous multi-vessel fisheries under high sea states. Background Technology

[0002] With the deepening of the national "Deep Blue Fisheries" strategy, marine fisheries are accelerating their transformation from traditional near-shore fishing to deep-sea industrialized aquaculture. According to the "Yellow and Bohai Sea Offshore Engineering Large-Scale Intelligent Aquaculture Model and Equipment" project, more large-scale truss cages (such as the "Deep Blue" series) and engineered fencing facilities have been deployed and are planned for construction in the Yellow and Bohai Seas. To achieve less human intervention or even unmanned operation in the aquaculture process, it is necessary to deploy a heterogeneous fleet of unmanned vessels, including environmental monitoring unmanned vessels, automated harvesting unmanned vessels, live fish transfer unmanned vessels, and feed delivery unmanned vessels, to replace the traditional high-risk manual operation mode.

[0003] However, existing unmanned fishing vessel path planning and control technologies face extremely severe challenges when applied to high sea state environments in the Yellow and Bohai Seas, mainly in the following aspects: 1. Marine environment modeling lacks specificity and ignores the unique wave spectrum characteristics of the sea area: Existing unmanned vessel path planning algorithms (such as traditional A / B / C) lack specificity and ignore the unique wave spectrum characteristics of the sea area. Dijkstra's and RRT algorithms are typically based on static environmental assumptions, or simply introduce a basic wind and wave resistance model as a passage cost. Their core is mostly based on standard wave spectra (such as the standard Jonswap spectrum, peak elevation factor). The parameters are typically taken as 3.3 or a single significant wave height (SWH). However, marine hydrological studies show that the Yellow and Bohai Seas, as semi-enclosed shallow waters, are significantly nonlinear and non-stationary due to the combined influence of seafloor topography and monsoon climate. During the monsoon season, this sea area often exhibits a "bimodal spectrum" characteristic with both wind waves and swells, and under the shallow water effect, its actual peak elevation factor (…) is significantly higher. The energy density of unmanned vessels is often higher than the standard value for open ocean, exhibiting a more concentrated energy characteristic. If the commonly used standard spectral parameters are used for forecasting, it will lead to a serious deviation in the estimation of the dynamic response (RAO) of unmanned vessels. This means that although the planned path may geometrically avoid the high wave area, it may fall into the dynamic danger zone due to misjudgment of the spectral energy distribution.

[0004] 2. Lack of consideration for nonlinear risks such as "parametric roll" in ship dynamics: In severe sea conditions, especially when the ship is sailing with the waves or in stern-angled waves, if the wave encounter frequency is close to twice the ship's natural roll frequency (i.e., This phenomenon can easily induce parametric rolling. According to research by international classification societies and maritime safety agencies, this is a highly dangerous nonlinear instability that can cause severe rolling or even capsizing of the vessel within a short period. Current path planning algorithms mostly focus on geometric obstacle avoidance and energy optimization, rarely incorporating a dynamic risk assessment mechanism that combines encounter frequency with real-time wave spectra into the cost function. For fishing unmanned vessels equipped with sophisticated sonar or carrying high-value live fish, existing planning methods cannot proactively avoid such dynamic resonance risks.

[0005] 3. Poor Continuity and Dynamic Docking Stability in Multi-Vessel Collaborative Operations: Fishery production has strict process constraints and timing requirements. For example, the integrated collaborative operation of "catching-grading-transfer" is explicitly required. This process requires the unmanned fishing vessel (UAV) for catching fish and the unmanned live fish transfer vessel to maintain extremely high-precision relative positions (station keeping) under dynamic sea conditions and to physically transfer materials (live fish). Existing collaborative control algorithms are mostly designed for loose formation navigation in open waters and lack control strategies for strongly coupled "dynamic docking" scenarios. Under wave interference, the heave and pitch motions of the two vessels are often asynchronous, resulting in extreme instability during the docking window and a high risk of collision or breakage of the transmission pipeline. Traditional PID or sliding mode control is difficult to handle this type of multivariable, strongly coupled control problem with complex environmental constraints.

[0006] 4. Obstacle avoidance strategies fail to consider the structural field effects of large-scale aquaculture facilities: Large truss cages and engineered fencing facilities are themselves massive hydraulic structures. Their presence significantly alters the local flow and wave fields, generating diffracted and reflected waves, creating complex "structural field effects." Traditional path planning algorithms typically simplify these facilities as static polygonal obstacles, ignoring the hydrodynamic disturbances around them. When unmanned vessels approach these facilities for feeding or inspection, they often encounter unexpected strong currents or irregular waves, leading to control instability.

[0007] In summary, there is an urgent need for an intelligent path planning and control method that can deeply integrate the wave characteristics of the Yellow and Bohai Seas, proactively avoid the risk of dynamic resonance, and enable close collaborative operations of heterogeneous fleets in complex sea conditions. Summary of the Invention

[0008] To address the problems existing in the prior art, the present invention aims to provide a method for collaborative path planning and operation control of heterogeneous multi-vehicle fisheries in high sea states. Specifically, it relates to a method for collaborative path planning and dynamic control of heterogeneous fisheries unmanned vessel fleets (USV Fleets) that is tailored to the unique shallow-water wave characteristics of the Yellow and Bohai Seas. This method combines improved JONSWAP wave spectrum modeling, parametric roll dynamic risk assessment, multi-task collaborative logic, and nonlinear model predictive control (NMPC). The aim is to solve the problems of existing technologies in high sea states in the Yellow and Bohai Seas, such as the lack of adaptability of fisheries unmanned vessel path planning to specific wave spectrum characteristics, neglect of nonlinear dynamic risks such as parametric roll, and unstable relative motion control and difficulty in capturing the operation window in the "catch-transfer" collaborative operation of heterogeneous fleets.

[0009] To achieve the above objectives, the present invention provides the following solution: A method for heterogeneous multi-vessel cooperative path planning and operation control in fisheries under high sea states, comprising: Meteorological data of the Yellow and Bohai Seas are acquired, and the wave surface elevation spectrum in the meteorological data of the Yellow and Bohai Seas is fitted to the modified JONSWAP spectral model. At the same time, the bimodal features are decomposed to calculate the wave encounter frequency. Based on the wave encounter frequency, a parameterized roll risk function is defined to generate a four-dimensional spatiotemporal cost map containing the dynamic resonance risk region. During the capture and transfer process, the improved A was adopted. The algorithm performs global path planning on the four-dimensional spatiotemporal cost map, generating a zigzag path; the improved A The algorithm utilizes A The heading constraint penalty term is introduced into the heuristic function of the algorithm to obtain the result; During the live fish handover process in the aforementioned catch and transfer mission, a cost function containing a relative motion inhibition term is established, and the fixed-berth and dynamic docking of multiple vessels is solved through rolling optimization.

[0010] Optionally, the modified JONSWAP spectral model includes: ; in, These are dynamic values ​​obtained by fitting real-time sea state data using the nonlinear least squares method. The angular frequency of the wave. Let be the wavefront elevation spectral density function. For energy scale parameters, It is the acceleration due to gravity. The spectral peak angular frequency, For peak shape parameters.

[0011] Optionally, decomposing the bimodal features includes: if coexistence of wind waves and swells is detected, then using the Ochi-Hubble model to decompose the wave spectrum into low-frequency swells and high-frequency wind waves.

[0012] Optionally, calculating the wave encounter frequency includes: ; in, For the frequency of wave encounter, For unmanned ships at different candidate speeds, The angular frequency of the wave. For unmanned vessels in different candidate headings, The main wave direction.

[0013] Optionally, the parameterized roll risk function is defined as follows: ; in, For parameterized roll risk function, These are coefficients related to the ship's main dimensions, namely length, beam, and draft. For the inherent roll frequency of the unmanned vessel, For the effective wave height, The coordinates are the position coordinates of the unmanned vessel in the global coordinate system.

[0014] Optionally, generating the four-dimensional spatiotemporal cost map includes: Based on the parameterized roll risk function, identify the heading-speed combination that is permissible in wave height but dangerous in frequency, and mark it as the first high-risk area; Based on the ship hydrodynamic coefficient database and the corrected real-time wave spectrum, the pitch amplitude and heave acceleration under a specific heading are predicted. If the prediction results exceed the equipment tolerance threshold, it is marked as the second highest risk area. The first high-risk area and the second high-risk area are combined with the static obstacles, water depth restrictions and flow field interference effects around the aquaculture facilities to generate the four-dimensional spatiotemporal cost map.

[0015] Optionally, establishing the cost function includes: ; in, Let cost function be For the prediction time domain of model predictive control, This is the cost of conventional position tracking errors. For the current discrete control moment, To predict the step number in the time domain, This is the position tracking error vector. This is the cost of conventional position tracking errors. This is the position error weight matrix. To control incremental costs, For the control increment of the actuator, To control the incremental weight matrix, This is the cost term for roll stability. The stability penalty weight coefficient, To parameterize the roll risk index, The phase synchronization penalty term is defined as follows: This forces the heave motion of the two ships in the waves to be in phase. For phase synchronization penalty weights, The heave displacement of the pilot vessel, i.e., the initial heave displacement of the fishing boat. This refers to the helical displacement of the following ship, i.e., the transfer ship.

[0016] Optionally, the rolling optimization solution for multi-ship cooperative fixed-berth and dynamic docking includes: The optimal control sequence for multi-ship coordinated berthing and dynamic docking is solved in each control cycle using a nonlinear programming solver, and only the first control variable is executed to compensate for wind, wave and current disturbances in real time.

[0017] The beneficial effects of this invention are as follows: Significantly improving the safety and survivability of operations in high sea states: This invention introduces a resonance risk assessment based on wave spectrum correction for the Yellow and Bohai Seas, enabling unmanned vessels to actively identify and avoid hidden dynamic risks such as "parametric roll." Simulations show that in sea state 5, this method can reduce the average roll amplitude of the vessel by 30%-40%, effectively preventing capsizing accidents, and its adaptability far exceeds that of traditional algorithms based solely on wave height.

[0018] Ensuring the continuity and success rate of the entire operation: This invention's relative motion suppression NMPC, designed for the "catch-transfer" scenario, solves the problem of misalignment and high collision risk between the two vessels in waves, a common problem in traditional methods. Through a phase synchronization strategy, the relative vertical displacement between the two vessels during the docking window is controlled within a safe threshold (e.g., 0.3 meters), significantly improving the success rate and efficiency of transferring maritime supplies / live fish.

[0019] Improving operational efficiency and extending equipment life: This invention optimizes the encounter frequency through zigzag path planning, reducing violent slapping and rolling of the hull, thereby reducing ineffective energy consumption and mechanical wear of the propulsion system and extending the service life of precision fishery equipment (such as underwater sonar, fish suction pumps, and sensors).

[0020] Strong adaptability to specific sea areas: The model parameters of this invention are directly calibrated for the wind and wave characteristics of the Yellow and Bohai Seas (such as the JONSWAP spectrum correction coefficient). Compared with general algorithms, the wave force prediction accuracy and control effect in this sea area are better, which directly supports the demonstration application in the Yellow and Bohai Sea regions. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a diagram illustrating the overall system architecture of a heterogeneous multi-vessel cooperative path planning and operation control method for fisheries under high sea states, according to an embodiment of the present invention. Figure 2 This is a schematic diagram comparing the corrected JONSWAP spectrum and the standard spectrum under typical sea conditions in the Yellow and Bohai Seas according to an embodiment of the present invention. Figure 3 This is a conceptual diagram of an embodiment of the present invention and a polar coordinate risk map containing a parameterized roll risk zone generated based on this spectrum; Figure 4 An improved version A based on encounter frequency avoidance in this invention embodiment. A diagram comparing the algorithm-generated zigzag safe path with the traditional shortest path; Figure 5 This is a block diagram of the cooperative NMPC control law for heterogeneous fleets considering relative motion suppression, according to an embodiment of the present invention. Figure 6 This is a comparison curve of the relative vertical displacement between the NMPC method of this invention and the traditional PID method during the docking of two ships in sea state 4, as shown in the simulation experiment of this embodiment of the invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0025] like Figure 1As shown in this embodiment, a method for heterogeneous multi-vehicle collaborative path planning and operation control in fisheries under high sea states is disclosed. The method includes: acquiring meteorological data from the Yellow and Bohai Seas; fitting the wavefront elevation spectrum from the meteorological data to a modified JONSWAP spectral model, and simultaneously decomposing the bimodal features to calculate the wave encounter frequency; defining a parameterized roll risk function based on the wave encounter frequency to generate a four-dimensional spatiotemporal cost map containing the dynamic resonance risk region; and employing an improved A... The algorithm performs global path planning on the four-dimensional spatiotemporal cost map, generating a zigzag path; the improved A The algorithm utilizes A The algorithm's heuristic function is obtained by introducing a heading constraint penalty term; during the live fish handover process in the capture and transfer tasks, a cost function containing a relative motion suppression term is established, and the fixed berthing and dynamic docking of multiple vessels is solved through rolling optimization.

[0026] Specifically, this embodiment discloses a method for cooperative path planning and operation control of heterogeneous multi-vessel fisheries under high sea states, including the following steps: Step S1, Environmental Reconstruction and Spectral Correction: Using shipborne sensors and shore-based ocean forecast data, a dynamic wave field model based on the modified JONSWAP spectrum is constructed to suit the characteristics of the Yellow and Bohai Seas, and the wave encounter frequency of the unmanned vessel is calculated in real time. The modified JONSWAP spectrum model described in Step S1 introduces the variation with significant wave height. and wind speed Dynamically changing peak elevation factor and peak shape parameters To adapt to the mixed double-peak characteristics of wind and swell in the shallow waters of the Yellow and Bohai Seas.

[0027] Step S2, Risk Map Construction: Combining the unmanned vessel hydrodynamic coefficient database, calculate the parameterized roll risk index and generate a four-dimensional spatiotemporal cost map containing the dynamic resonance risk region.

[0028] Step S3, Collaborative Global Planning: For the fishery operation task chain, utilize the improved A on the dynamic cost map. The algorithm generates a globally optimal waypoint sequence that avoids resonant frequencies.

[0029] Step S4, Nonlinear Model Predictive Control: In the local collaborative operation stage, a cost function containing relative motion suppression terms is established, and high-precision berthing and dynamic docking of multiple ships are achieved through rolling optimization solution.

[0030] This embodiment proposes a method for heterogeneous multi-vessel collaborative path planning and operation control in fisheries under high sea states, namely, a path planning and control method based on a four-layer progressive architecture of environment reconstruction, risk quantification, global planning, and local collaboration. This method deeply integrates physical oceanography models with robot control theory, reconstructs the unique sea states of the Yellow and Bohai Seas using a modified wave spectrum model, constructs a four-dimensional spatiotemporal map including dynamic risks, and utilizes nonlinear model predictive control to achieve refined operation of multi-vessel collaboration. The specific technical solution includes the following four core steps: Step S1: Reconstruction of dynamic sea state environment in the Yellow and Bohai Seas based on the modified JONSWAP spectrum: Using the X-band wave measuring radar, inertial navigation system and shore-based meteorological forecast data carried by the unmanned vessel, a real-time wave field model for the characteristics of the Yellow and Bohai Seas is constructed.

[0031] Adaptive correction of wave spectrum parameters: The universally accepted standard JONSWAP spectrum parameters are abandoned in favor of a spectrum model modified for the characteristics of shallow waters in the Yellow and Bohai Seas. Considering the bimodal characteristics of wind-driven waves and swells in this sea area, a model based on wind speed is established. Waikaze distance Spectral peak elevation factor Dynamic correction model: ; ; in, It is no longer a constant 3.3, but a dynamic value obtained by fitting real-time sea state data using a nonlinear least squares method. , For the spectral peak period, This refers to the significant wave height. Studies have shown that the Yellow and Bohai Seas... The mean values ​​are often high (e.g., above 4.0), indicating that the energy is more concentrated and requires special correction. At the same time, for the bimodal spectral characteristics, the Ochi-Hubble six-parameter model is used for superposition processing to ensure that both low-frequency swells (which have a significant impact on large transport ships) and high-frequency wind waves (which have a significant impact on small monitoring ships) are included in the model.

[0032] Encounter Frequency Mapping Calculation: The aforementioned "adaptive correction of wave spectrum parameters" provides a crucial reference input for the calculation of encounter frequencies. Specifically, the modified JONSWAP spectrum model can accurately calculate the absolute circular frequency (i.e., spectral peak angular frequency) of the wave with the most concentrated energy under the current sea state. (And high-energy frequency bands). The system calculates the unmanned surface vessel in real time under different candidate headings. and speed The frequency of the waves encountered below : ; in, Main wave direction, This is the wave's circular frequency. This step forms the physical basis for subsequent risk assessments.

[0033] Step S2: Construct a "Wave-Resonance Risk Map" based on dynamic response: Unlike traditional static obstacle grid maps, this invention constructs a risk map that includes "hidden dynamic obstacles," quantifying the potential threat of sea conditions to the ship's hull into a cost value.

[0034] Parametric Roll Risk Index (PRRI) Quantification: Defining a parametric roll risk function Based on the instability principle of the Mathieu equation, when encountering a frequency... Falling into the inherent rolling frequency of the unmanned vessel The resonance range (usually) (nearby frequency bands) and significant wave height When the threshold is exceeded, the risk index increases exponentially: ; in, These are coefficients related to the ship's main dimensions (length, beam, draft). This function can identify course-speed combinations that are "permissible in wave height but dangerous in frequency." At this point, the risk index increases significantly, and this high penalty is the path planning A discussed later. The "heading constraint penalty" introduced in the algorithm's heuristic function will force the planning algorithm to abandon high-risk straight routes and instead generate a zigzag path that changes course to avoid roll resonance.

[0035] Pitch and Heave Response Prediction: Based on the Ship Hydrodynamics Database (RAO) and combined with the corrected real-time wave spectrum, the pitch amplitude under a specific heading is predicted. and heave acceleration For monitoring vessels equipped with sophisticated sensors, if the predicted acceleration exceeds the equipment's tolerance threshold (e.g., 0.5g), the area is marked as high-risk.

[0036] Comprehensive Risk Cost Field Generation: By superimposing the aforementioned dynamic risks with static obstacles (cages, islands), water depth limitations, and flow field disturbances around aquaculture facilities, a four-dimensional spatiotemporal risk cost field is generated. .

[0037] Step S3, Task-driven collaborative global path planning for heterogeneous fleets: For complex task chains such as "capture-transfer", an improved multi-objective A / B algorithm is adopted. The algorithm generates a global waypoint sequence on the risk map.

[0038] Task chain temporal decoupling: The complex fishery operation process is decomposed into an orderly sequence of sub-tasks: environmental monitoring vessel cruise scanning, fishing vessel approaching the cages for operation, transfer vessel dynamically rendezvous, and coordinated return voyage.

[0039] Frequency avoidance heuristic function: In the total cost function of the traditional A algorithm Based on this, a "heading constraint penalty" based on roll risk is introduced. The specific acquisition and calculation methods are as follows: The algorithm expands the search from the current node to candidate nodes. When determining the path, first calculate the candidate connecting headings for that segment. Subsequently, the course was changed. The risk index for this leg of the journey is obtained by substituting the wave frequency from the current speed U, the corrected JONSWAP spectrum output, into the parameterized roll risk function. Finally, the heading constraint penalty term is defined as: ; in, This represents the penalty weight coefficient. The improved A algorithm's total cost evaluation function is updated as follows: ; The heuristic cost is the Euclidean distance from the candidate node to the target point. If the candidate node's course is in the high-risk range with the waves and the frequency matches, It will output an extremely high cost value, thus being automatically eliminated by the algorithm when selecting nodes.

[0040] Zig-Zag Roll Reduction Planning: Based on the aforementioned penalty mechanism, the algorithm automatically plans a zig-shaped path. By periodically changing its course, the unmanned vessel switches between the left and right leading waves, always avoiding direct head-on waves or dangerous tail-angle waves, thus achieving "safety optimization" rather than simply "shortest distance" at the dynamic level.

[0041] Step S4, Local Coordination of Nonlinear Model Predictive Control (NMPC) Based on Relative Motion Suppression: During the "dynamic docking" stage of live fish handover between the fishing vessel and the transfer vessel, the accuracy of global planning no longer meets the requirements. Therefore, this invention switches to the NMPC strategy for refined control.

[0042] High-fidelity modeling of relative motion: Establishing the equations of state for the relative motion of the navigator (fishing vessel) and the follower (transfer vessel). This model not only includes the positional error on the horizontal plane, but also explicitly introduces the relative motion on the vertical plane induced by waves (heave difference, pitch difference).

[0043] Design of cost function for collaborative multi-objective optimization: Constructing an optimization objective function that includes a relative motion suppression term. : ; in, This is the cost of conventional position tracking errors. The phase synchronization penalty is defined as follows: This mechanism forces the heave motions of the two vessels in the waves to be in phase. That is, when one vessel is at the crest of a wave, the controller adjusts the position of the other vessel to also be at the crest of a wave, thereby minimizing the relative height difference between the decks of the two vessels and ensuring the safe connection of the fish passage pipe.

[0044] Rolling optimization solution: Using a nonlinear programming solver (such as IPOPT or SQP), the optimal control sequence (thrust speed, rudder angle) is solved in real time within each control cycle (such as 100ms), and only the first control variable is executed to achieve real-time compensation for wind, wave and flow disturbances.

[0045] In one embodiment, the present invention provides a specific implementation plan to address the operation and maintenance needs of large-scale offshore engineered aquaculture facilities in the Yellow and Bohai Seas. The implementation process of the present invention is described in detail below with reference to the accompanying drawings and mathematical models.

[0046] Environmental Perception & Reconstruction: An unmanned fleet (including a fishing vessel USV-A and a transfer vessel USV-B, both belonging to "intelligent production equipment for large-scale offshore aquaculture facilities") enters the target operating area.

[0047] Data Acquisition and Fusion: The X-band wave measurement radar on board the USV-A acquires sea surface echo images, which are combined with heave displacement data provided by the shipborne inertial navigation system (INS) and macro-meteorological data (wind speed, wind direction) released by shore-based meteorological stations.

[0048] Yellow and Bohai Sea Characteristic Parameter Identification: The system reads current meteorological data for the Yellow and Bohai Seas (e.g., wind speed). (Wind direction southeast). Using the nonlinear least squares method, the measured wavefront elevation spectrum is fitted to the modified JONSWAP spectral model, such as... Figure 2 As shown.

[0049] Spectral peak enhancement factor Dynamic correction: Based on the shallow water effect and monsoon characteristics of the Yellow and Bohai Seas, the following settings are made: The initial value was 3.3, but it was dynamically corrected based on the measured spectral width parameter. According to historical data, under high sea states in this area, The values ​​are often too large. In this embodiment, the corrected formula is used to calculate the current actual value. This indicates that the wave energy is more concentrated, resulting in a stronger impact and destructive force on the ship's hull.

[0050] Bimodal Spectrum Decomposition: If a bimodal feature (coexistence of wind waves and swells) is detected, the system uses the Ochi-Hubble model to decompose the wave spectrum into a low-frequency swell component and a high-frequency wind wave component, and calculates their characteristic parameters separately to ensure that no frequency band that may cause resonance is missed.

[0051] Real-time encounter frequency calculation: The system calculates the frequency based on the current speed. and heading Combined with the main wave direction Real-time calculation of encounter frequency For example, when a ship is sailing against the waves at a speed of 10 knots, It will increase significantly, possibly moving out of the roll resonance zone but into the violent pitch zone.

[0052] Dynamic Risk Map Construction Based on Dynamics: The system loads the hydrodynamic coefficient database (RAO) of this type of unmanned vessel, which is obtained through simulation using the "Fan Li Large Model".

[0053] Resonance risk zone marker: The inherent roll period of the unmanned vessel is known. , The system calculates the frequency range of major hazards based on the parameterized roll risk formula. (Right now (nearby), such as Figure 3 As shown.

[0054] Polar coordinate risk field generation: In the raster map, in addition to marking large cages, anchor lines, and reefs as static obstacles, all objects that could cause... A "heading-speed combination" falling into a danger zone is marked as a high-risk area. For example, when sailing with the waves and the wavelength is close to the ship's length ( When ), the parameterized roll risk index The peak has been reached. In the visualized polar risk map, this area is rendered as a red "no-go sector." This means that even if there are no physical obstacles in that direction, the unmanned vessel is prohibited from traveling at that course and speed.

[0055] When constructing risk maps, the red sectors in the polar coordinate system visually represent dangerous course directions. For example, when the main wave direction is southeast (135°), if the ship's bow is northwest (315°) and the speed is high, the encounter frequency may trigger resonance, and this direction is marked as high risk. This risk quantification based on physical mechanisms transforms the "dead" static nautical chart into a "living" dynamic map.

[0056] Task chain-driven collaborative global planning: Task instructions: The harvesting vessel USV-A needs to go from the mother ship to cage No. 3 to carry out harvesting operations, and then the transfer vessel USV-B needs to go to the cage to meet and transfer the fish.

[0057] Global path generation: Improved A When searching for a path, the algorithm predicts the heading of each node. If a direct route (shortest distance) would cause the unmanned surface vessel to sail with the waves for an extended period and encounter frequencies falling into the resonance zone, the algorithm will automatically plan a zig-zag path, such as... Figure 4 As shown.

[0058] Path characteristics: Although the total distance of the zigzag path increases by about 15%, the frequency of encounters is always avoided by periodically changing the course (e.g., switching between the left and right leading waves). The danger zone.

[0059] Cost assessment: In the weighted total cost function, the stability cost is greatly reduced because the violent shaking caused by resonance is avoided, making this path with a longer physical distance the optimal solution in the overall score.

[0060] Cooperative rendezvous point selection: The planner also considers the flow field effect around the cage. Based on measured ocean current data (velocity 0.8 m / s, flow direction 270°), the algorithm selects a low-velocity area downstream of the cage as the rendezvous point to reduce energy consumption of the two vessels when stationary.

[0061] NMPC Local Dynamic Docking Control (NMPC Colic Control): When the distance between the catching vessel and the transfer vessel is less than 50 meters, the "catching-transfer" collaborative operation phase begins, and control is transferred to the NMPC controller. Figure 5 As shown.

[0062] State prediction and relative motion modeling: The NMPC controller predicts the next 20 time steps based on the currently reconstructed wave spectrum (Horizon). The model describes the motion of the two ships within the area. The relative motion equations under wave disturbance are explicitly included in the model.

[0063] Cost function solution and phase synchronization: in the cost function In China, significantly improved (Relative motion weights) and The weight of (synchronous penalty item).

[0064] Scenario simulation: The solver found that if USV-B only keeps its geographical location unchanged (in the traditional dynamic positioning DP mode), due to the spatial phase difference of the waves, when USV-A rises with the waves, USV-B may be exactly at the trough of the wave, causing the two ships to have a relative vertical displacement of up to 1.5 meters, which poses a risk of collision or breaking of the fish transport pipe.

[0065] Control strategy optimization: The NMPC solver outputs a set of optimized control commands through nonlinear programming. It directs the USV-B to actively perform small longitudinal acceleration and heading adjustments, so that its phase of wave motion tends to synchronize with that of the USV-A (Phase Synchronization).

[0066] Result: Through this active phase adjustment, the relative vertical displacement between the two ships was reduced to less than 0.3 meters, meeting the safety docking requirements of the flexible fish passage pipe.

[0067] Command execution: The underlying thrust allocation algorithm (ControlAllocation) distributes the generalized force commands (longitudinal force, lateral force, and yaw moment) output by NMPC to the azimuth thrusters and side thrusters to execute precise maneuvers.

[0068] This process was verified in simulation experiments. Compared with traditional control methods, the relative motion suppression effect of the present invention is significant.

[0069] like Figure 6 As shown, simulation verification and data analysis: Based on the "Fan Li Large Model" digital twin platform, the above method was verified through simulation. The simulation environment was set as typical sea state 4 in the Yellow and Bohai Seas (significant wave height 1.5m, JONSWAP spectrum). ).

[0070] Enhanced safety: The unmanned vessel using the path planning algorithm of this invention reduced its roll amplitude by an average of 40%, and no dangerous roll exceeding 15° occurred during the entire voyage, completely eliminating the parametric roll phenomenon.

[0071] Docking stability: In a dual-ship collaborative transport scenario, the phase synchronization control of NMPC improved the proportion of time during which the relative vertical displacement of the two ships was controlled within 0.3 meters during the docking window by 60%. In contrast, the relative displacement of traditional PID controllers fluctuates wildly and often exceeds the safety threshold.

[0072] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for cooperative path planning and operation control of heterogeneous multi-vessel fisheries under high sea states, characterized in that, include: Meteorological data of the Yellow and Bohai Seas are acquired, and the wave surface elevation spectrum in the meteorological data of the Yellow and Bohai Seas is fitted to the modified JONSWAP spectral model. At the same time, the bimodal features are decomposed to calculate the wave encounter frequency. Based on the wave encounter frequency, a parameterized roll risk function is defined to generate a four-dimensional spatiotemporal cost map containing the dynamic resonance risk region. During the capture and transfer process, the improved A was adopted. The algorithm performs global path planning on the four-dimensional spatiotemporal cost map, generating a zigzag path; the improved A The algorithm utilizes A The heading constraint penalty term is introduced into the heuristic function of the algorithm to obtain the result; During the live fish handover process in the aforementioned catch and transfer mission, a cost function containing a relative motion inhibition term is established, and the fixed-berth and dynamic docking of multiple vessels is solved through rolling optimization.

2. The method for cooperative path planning and operation control of heterogeneous multi-vessel fisheries under high sea states as described in claim 1, characterized in that, The modified JONSWAP spectral model includes: ； in, These are dynamic values ​​obtained by fitting real-time sea state data using the nonlinear least squares method. The angular frequency of the wave. Let be the wavefront elevation spectral density function. For energy scale parameters, It is the acceleration due to gravity. The spectral peak angular frequency, For peak shape parameters.

3. The method for cooperative path planning and operation control of heterogeneous multi-vessel fisheries under high sea states as described in claim 1, characterized in that, The decomposition of the bimodal features includes: if the coexistence of wind waves and swells is detected, the wave spectrum is decomposed into low-frequency swells and high-frequency wind waves using the Ochi-Hubble model.

4. The method for cooperative path planning and operation control of heterogeneous multi-vessel fisheries under high sea states according to claim 1, characterized in that, Calculating the wave encounter frequency includes: ； in, For the frequency of wave encounter, For unmanned ships at different candidate speeds, The angular frequency of the wave. For unmanned vessels in different candidate headings, The main wave direction.

5. The method for cooperative path planning and operation control of heterogeneous multi-vessel fisheries under high sea states according to claim 1, characterized in that, The parameterized roll risk function is defined as follows: ； in, For parameterized roll risk function, These are coefficients related to the ship's main dimensions, namely length, beam, and draft. For the inherent roll frequency of the unmanned vessel, For the effective wave height, The coordinates are the position coordinates of the unmanned vessel in the global coordinate system.

6. The method for cooperative path planning and operation control of heterogeneous multi-vessel fisheries under high sea states according to claim 1, characterized in that, Generating the four-dimensional spatiotemporal cost map includes: Based on the parameterized roll risk function, identify the heading-speed combination that is permissible in wave height but dangerous in frequency, and mark it as the first high-risk area; Based on the ship hydrodynamic coefficient database and the corrected real-time wave spectrum, the pitch amplitude and heave acceleration under a specific heading are predicted. If the prediction results exceed the equipment tolerance threshold, it is marked as the second highest risk area. The first high-risk area and the second high-risk area are combined with the static obstacles, water depth restrictions and flow field interference effects around the aquaculture facilities to generate the four-dimensional spatiotemporal cost map.

7. The method for cooperative path planning and operation control of heterogeneous multi-vessel fisheries under high sea states according to claim 1, characterized in that, Establishing the cost function includes: ； in, Let cost function be For the prediction time domain of model predictive control, This is the cost of conventional position tracking errors. For the current discrete control moment, To predict the step number in the time domain, This is the position tracking error vector. This is the cost of conventional position tracking errors. This is the position error weight matrix. To control incremental costs, For the control increment of the actuator, To control the incremental weight matrix, This is the cost term for roll stability. The stability penalty weight coefficient, To parameterize the roll risk index, The phase synchronization penalty term is defined as follows: This forces the heave motion of the two ships in the waves to be in phase. For phase synchronization penalty weights, The heave displacement of the pilot vessel, i.e., the initial heave displacement of the fishing boat. This refers to the helical displacement of the following ship, i.e., the transfer ship.

8. The method for cooperative path planning and operation control of heterogeneous multi-vessel fisheries under high sea states according to claim 1, characterized in that, The rolling optimization method for solving the fixed-berthing and dynamic docking of multiple vessels includes: The optimal control sequence for multi-ship coordinated berthing and dynamic docking is solved in each control cycle using a nonlinear programming solver, and only the first control variable is executed to compensate for wind, wave and current disturbances in real time.

Images