A two-ship navigation transverse replenishment decision-making method considering ship-to-ship effect

Abstract

The application discloses a two-ship navigation transverse replenishment decision-making method considering ship-to-ship effect, and relates to the technical field of ship control. The method comprises the following steps: determining a plurality of expected operation paths of a replenishment ship and a receiving ship according to replenishment working condition parameters by using a path planning method; based on a six-degree-of-freedom motion prediction model of the two ships, using a ship motion controller to control the two ships to perform cooperative motion simulation according to the expected operation paths of the two ships respectively, and obtaining replenishment operation data corresponding to each expected operation path of the two ships; screening the replenishment operation data and constructing a safe operation sample database; training a transverse replenishment operation strategy model by using a deep reinforcement learning algorithm; and using the trained transverse replenishment operation strategy model to make real-time decisions for a current replenishment working condition, and obtaining an operation instruction at the current time. The method considers the influence of ship-to-ship effect on the two ships, gives the best safe replenishment operation strategy according to the ship motion state, and effectively improves the sea state adaptability.

Description

Technical Field

[0001] This application relates to the field of ship control technology, and in particular to a decision-making method for lateral replenishment of two ships under the consideration of inter-ship effects. Background Technology

[0002] Lateral replenishment refers to replenishment where the supply vessel and the receiving vessel sail parallel to each other at the same speed and in the same direction, with replenishment carried out by specialized replenishment equipment spanning between the two vessels. Commonly used replenishment equipment, such as rigging and hoses, offers advantages such as convenient docking and undoing due to their short length, versatility in supplying various types of goods, strong adaptability to sea conditions, and high replenishment efficiency. When using replenishment equipment for lateral replenishment, if the lateral distance is too large, cargo transported via overhead cables may be submerged in the sea; if the lateral distance is too small, the ship-to-ship effect can easily lead to collisions.

[0003] The inter-ship effect is the result of the coupling of numerous influencing factors such as the relative position, hull size, and speed of two ships, causing mutual attraction, repulsion, turning, and wave-like hydrodynamic interference between them. Therefore, maintaining an appropriate relative position under the influence of the inter-ship effect is crucial for the success of lateral replenishment at sea. However, current lateral replenishment strategies rely heavily on the experience and knowledge of commanders and navigators. Furthermore, to avoid replenishment failure, the sea conditions and speeds selected based on human experience during actual operations are often conservative. This leads to low efficiency and accuracy in lateral replenishment, severely limiting maritime logistical support capabilities. Summary of the Invention

[0004] To address the aforementioned problems and technical requirements, this application proposes a two-ship lateral replenishment decision-making method that considers inter-ship effects. The technical solution of this application is as follows: A method for making decisions on lateral replenishment during two-ship navigation that considers inter-ship effects includes the following steps: Based on the replenishment conditions parameters, multiple expected operational paths for the replenishment vessel and the receiving vessel are determined using path planning methods; the replenishment conditions parameters include the motion state parameters of the replenishment vessel and the receiving vessel and the sea state parameters of the target sea area; Based on the six-degree-of-freedom motion prediction model of the supply ship and the receiving ship, a ship motion controller is used to control the two ships to perform cooperative motion simulation according to their respective expected operation paths. The supply operation data corresponding to each expected operation path of the supply ship and the receiving ship is obtained. The supply operation data of each ship includes multiple sea state parameters and the motion state parameters and control commands of each ship at multiple simulation moments under each sea state parameter. The six-degree-of-freedom motion prediction model of each ship is constructed by introducing inter-ship interaction loads into the six-degree-of-freedom motion equations of each ship. The inter-ship interaction loads characterize the interference effect of the inter-ship effect between the supply ship and the receiving ship on the two ships. The ship motion controller of each ship generates control commands and controls the motion state of the ship. The replenishment operation data corresponding to all expected operation paths of the replenishment ship and the receiving ship are filtered and a safe operation sample database is constructed. The sea state parameters corresponding to each simulation time in the safe operation sample database are used as the state, the control commands corresponding to each simulation time are used as the action, and the motion state parameters corresponding to each simulation time are used as the reward. The lateral replenishment control strategy model is trained by deep reinforcement learning algorithm. The trained lateral supply control strategy model is used to make real-time decisions on the current supply conditions and obtain the control commands at the current moment.

[0005] A further technical solution involves determining the motion state parameters, including the ship's position and speed, and identifying the inter-ship loads experienced by any given ship, including: The longitudinal distance along the length of the supply ship's center of gravity and the receiving ship's center of gravity is determined based on their positions. And the lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship along the beam direction. ; When the longitudinal distance between the center of gravity of the supply ship and the center of gravity of the receiving ship along the length of the ship. The longitudinal distance requirement is not met, or the lateral distance between the center of gravity of the supply vessel and the center of gravity of the receiving vessel along the beam direction is not met. If the lateral distance requirement is not met, the inter-ship load is set to 0; otherwise, the longitudinal distance between the center of gravity of the supply ship and the center of gravity of the receiving ship is used. Horizontal distance The speeds of the supply ship and the receiving ship, combined with their dimensional parameters, are used to determine the inter-ship loads.

[0006] The further technical solution is that the inter-ship loads include inter-ship forces and inter-ship moments; Determine the intership forces acting on any ship for:

[0007] Determine the inter-ship moment acting on any ship for:

[0008] in, L It is the average length of the supply ship and the receiving ship. L F He is the captain of the supply ship. L L He is the captain of the receiving ship. U F It's the speed of the supply ship. U L It is the speed of the receiving ship; , , , , , , , It is the inter-ship force coefficient. b 1. b 2. b 3. b 4. b 5. b 6. b 7. b 8. b 9. b 10 , b 11 It is the inter-ship moment coefficient.

[0009] Its further technical solution is that the motion state parameters include the ship's position, heading angle and six-degree-of-freedom motion speed, the control commands include rudder angle commands and propeller speed commands, and the ship motion controller of each ship includes a heading angle controller and a speed controller. The input to the heading angle controller is the deviation between the expected heading angle value and the actual heading angle value, and the output of the heading angle controller is the rudder angle command; The speed controller's input is the deviation between the expected value and the actual value of the resultant velocity of the six-degree-of-freedom motion, and its output is the propeller speed command.

[0010] A further technical solution involves determining the desired bow angle of the receiving vessel. ,in, It is the expected value of the bow angle of the supply ship. It is the longitudinal component of the actual resultant velocity of the six-degree-of-freedom motion of the receiving ship. It is the actual lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship. It is the expected lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship.

[0011] The further technical solution involves determining the expected value of the resultant velocity of the six degrees of freedom motion of the receiving ship. ,in, This is the expected value of the resultant velocity of the supply ship's six degrees of freedom motion. It is a proportionality coefficient that converts distance into speed. This is the actual longitudinal distance between the sending station on the supply ship and the receiving station on the receiving ship. It is the expected longitudinal distance between the sending station on the supply ship and the receiving station on the receiving ship.

[0012] The further technical solution is that each expected operational path includes an entry and positioning phase, a parallel replenishment phase, and a safe withdrawal phase; the motion state parameters include the ship's position and speed, and the resulting safe operational sample database includes: The replenishment operation data corresponding to each expected operation path of the replenishment ship and the receiving ship are screened to determine several safe operation paths that meet the safety operation constraints during the parallel replenishment phase, as well as the safe speed range corresponding to each safe operation path; the entry and positioning phase is the phase in which the two ships gradually approach each other from the predetermined position, the parallel replenishment phase is the phase in which the two ships sail in parallel to transfer materials, and the safe withdrawal phase is the phase in which the two ships gradually move away from the predetermined position. The resupply operation data for each safe operation path within the safe speed range is used as the safe operation data. A safe operation sample database is constructed using the safe operation data corresponding to each safe operation path. Each sample in the safe operation sample database includes a simulation time step. t Corresponding sea state parameters Simulation time step t Corresponding motion state parameters Simulation time step t Corresponding control commands and simulation time step t +1 corresponds to the sea state parameter .

[0013] The further technical solution is that the safety operation constraints include: under full rudder conditions, the lateral resultant forces on the two ships are in opposite directions, and the bow roll resultant moments on the two ships are in opposite directions, and the pitch angle of each ship does not exceed 2° and the roll angle does not exceed 5°.

[0014] The further technical solution is that each expected operational path includes an entry and positioning phase, a parallel replenishment phase, and a safe evacuation phase. Multiple expected operational paths for any given ship are determined as follows: Based on the replenishment operation parameters, multiple expected operation paths of the replenishment ship are determined using a path planning algorithm, and there are no obstacles in the predetermined area around each expected operation path of the replenishment ship. Based on the replenishment operating parameters, determine the lateral distance between the center of gravity of the replenishment ship and the center of gravity of the receiving ship, as well as the misalignment angle between the replenishment ship and the receiving ship. The misalignment angle is the angle between the line connecting the sending station on the replenishment ship and the receiving station on the receiving ship and the width direction of the replenishment ship, and is positive in the clockwise direction. Based on each expected operational path of the supply ship, a path planning algorithm is used to determine multiple expected operational paths of the receiving ship when the lateral distance between the center of gravity of the receiving ship and the center of gravity of the supply ship is within a preset safe lateral distance range and the misalignment angle between the receiving ship and the supply ship is within a preset safe misalignment angle range during the entry and positioning phase and the safe withdrawal phase. The entry and positioning phase is the phase in which the two ships gradually approach each other from the predetermined position. The parallel supply phase is the phase in which the two ships sail in parallel to transfer materials. The safe withdrawal phase is the phase in which the two ships gradually move away from the predetermined position.

[0015] The further technical solution is that the preset safe lateral distance range for the entry and safe withdrawal phases is: W +5 meters W +40 meters, It is the lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship. W It is the sum of half the width of the supply ship and the receiving ship, in meters; When the receiving vessel is positioned on the right side of the supply vessel, the preset safe offset angle range for entering the positioning phase is: The preset safety misalignment angle range for the safe evacuation phase is: When the receiving vessel is on the left side of the supply vessel, the preset safe offset angle range for entering the positioning phase is: The preset safety misalignment angle range for the safe evacuation phase is: ; It is the misalignment angle between the supply ship and the receiving ship.

[0016] The beneficial technical effects of this application are: This application proposes a two-ship lateral replenishment decision-making method considering inter-ship effects. By co-simulating the two-ship lateral replenishment process under multiple different replenishment conditions, the method matches the predicted ship motion states with the ship trajectories during the simulation using spatiotemporal features, obtaining a safe replenishment operation database containing spatiotemporal characteristics. A lateral replenishment control strategy model is then trained using a deep reinforcement learning algorithm, enabling the generation of control commands for safe replenishment operations under any sea state. This method not only provides a reference scheme for lateral replenishment operations at sea for various types of vessels but also provides the optimal safe replenishment control strategy based on the ship's motion state, thereby effectively improving the sea state adaptability of two-ship lateral replenishment operations.

[0017] Considering the impact of ship-to-ship effects on both the supply and receiving ships, a six-degree-of-freedom motion prediction model for the two ships is constructed by introducing ship-to-ship interaction loads into the six-degree-of-freedom equations of motion. This effectively improves the accuracy of the simulation results of the ship motion state during the lateral replenishment process. Furthermore, the ship-to-ship interaction load calculation formula proposed in this application comprehensively considers the influence of the shape, speed, and relative position of the supply and receiving ships, accurately calculating the magnitude and direction of the ship-to-ship effect. This overcomes the shortcomings of existing ship-to-ship effect prediction methods, which are not applicable to deep water, same speed, and parallel states, effectively improving the reliability of safety decisions for lateral replenishment. Simultaneously, a constructed ship motion controller monitors the ship motion state in real time throughout the entire lateral replenishment process and continuously adjusts control commands to optimize the ship's motion state, achieving a full-process simulation of lateral replenishment in waves with integrated spatiotemporal control.

[0018] By planning the operational paths of the supply ship and the receiving ship under different supply conditions, this application takes into account the influence of the inter-ship effect compared with the traditional path planning method. In addition to meeting the collision-free condition, the lateral distance and misalignment angle of the two ships are also restricted when planning the path, which restricts the navigation trajectory of the two ships, avoids wasting too much exploration time, and can effectively improve the safety and efficiency of the lateral supply operation of the two ships.

[0019] In addition, when screening samples for safe replenishment operations, this application takes into account the impact of the ship's roll and pitch motion in the waves on the replenishment operation, as well as the inter-ship loads on the two ships. By using safety constraints, it effectively screens learning samples for replenishment safety decisions. The lateral replenishment control strategy model trained in this way can adaptively adjust replenishment control decisions according to changes in the operation object and replenishment environment, thereby enhancing the safety decision-making capability of replenishment operations under complex sea conditions. Attached Figure Description

[0020] Figure 1 This is a flowchart of the decision-making process for lateral replenishment during navigation between two ships.

[0021] Figure 2 This is a schematic diagram of the two ships' lateral replenishment operation process.

[0022] Figure 3 This is a schematic diagram of the structure of a ship motion controller.

[0023] Figure 4 This is a flowchart of constructing a lateral supply control strategy model in one embodiment. Detailed Implementation

[0024] The specific embodiments of this application will be further described below with reference to the accompanying drawings.

[0025] This application proposes a two-ship lateral replenishment decision-making method that considers inter-ship effects. Please refer to [reference needed]. Figure 1 The flowchart shown illustrates the specific steps of this method: Step 1: Based on the replenishment operational parameters, multiple expected operational paths for the replenishment and receiving vessels are determined using path planning methods. These parameters include the motion state parameters of the replenishment and receiving vessels and the sea state parameters of the target sea area. The target sea area for the replenishment operation is pre-determined based on the replenishment mission, combined with meteorological, hydrological, and traffic density data. Various sea state parameters, including wind, wave, and current information, are then determined based on historical data from this target sea area. The replenishment mission is pre-determined according to the vessel type, including operational information such as the type of replenishment material, replenishment quantity, replenishment speed, and replenishment station location (sending station, receiving station). The replenishment operational parameters also include the motion state parameters of the replenishment and receiving vessels, including navigation parameters such as position, heading, and speed, as well as the vessels' six degrees of freedom motion response parameters.

[0026] The path planning method is used to determine each expected operational path for the supply ship and the receiving ship, including the entry and positioning phase, the parallel supply phase, and the safe withdrawal phase. Multiple expected operational paths for any given ship are determined, including: (1) Based on the replenishment working condition parameters, multiple expected operation paths of the replenishment ship are determined using a path planning algorithm. There are no obstacles in the predetermined area around each expected operation path of the replenishment ship.

[0027] During actual replenishment operations, the replenishment vessel sails towards the predetermined target point in the target sea area. The receiving vessel, which needs to receive replenishment, gradually approaches the replenishment vessel. After entering the positioning phase and entering the parallel replenishment phase to complete the replenishment operation, the receiving vessel departs from the replenishment vessel after a safe withdrawal phase and returns to its original navigation route. Throughout the entire replenishment operation, the replenishment vessel's navigation route remains unchanged; it only needs to navigate towards the predetermined target point while avoiding obstacles. Therefore, it is first necessary to determine the expected operating path of the replenishment vessel. The path planning algorithm used can be a path planning method of existing technology, such as the line of sight (LOS) method. The specific implementation steps will not be elaborated in this application.

[0028] (2) Based on the replenishment operating parameters, determine the lateral distance between the center of gravity of the replenishment vessel and the center of gravity of the receiving vessel, as well as the misalignment angle between the replenishment vessel and the receiving vessel, such as... Figure 2 As shown, the misalignment angle is the angle between the line connecting the sending station on the supply ship and the receiving station on the receiving ship and the beam direction of the supply ship, with clockwise rotation being positive. Figure 2 middle It is the misalignment angle between the supply ship and the receiving ship. It is the lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship.

[0029] (3) Based on each expected operational path of the supply ship, a path planning algorithm is used to determine multiple expected operational paths of the receiving ship when the lateral distance between the center of gravity of the receiving ship and the center of gravity of the supply ship is within a preset safe lateral distance range and the misalignment angle between the receiving ship and the supply ship is within a preset safe misalignment angle range during the entry and positioning phases and the safe withdrawal phase. The entry and positioning phase is the phase in which the two ships gradually approach each other from the predetermined position; the parallel supply phase is the phase in which the two ships sail in parallel to transfer materials; and the safe withdrawal phase is the phase in which the two ships gradually move away from the predetermined position. Similarly, the path planning algorithm for planning the expected operational path of the receiving ship can adopt existing path planning methods.

[0030] When planning the expected operational path of the receiving vessel, it is essential to ensure both the efficiency and safety of the replenishment operation. This is achieved by monitoring the motion parameters of the replenishment and receiving vessels in real time to constrain the misalignment angle and the lateral distance between their centers of gravity. Since the receiving and replenishment vessels must maintain the same direction and speed during the parallel replenishment phase, the lateral distance between their centers of gravity depends on the position and heading angle of the receiving vessel during the entry and safe withdrawal phases. Therefore, it is necessary to constrain the lateral distance and misalignment angle between the two vessels' centers of gravity during the entry and safe withdrawal phases to ensure the safe and efficient conduct of the replenishment operation.

[0031] In one embodiment, the preset safe lateral distance range for the entry and safe evacuation phases is: W +5 meters W +40 meters, It is the lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship. W It is the sum of half the width of the supply ship and the receiving ship, in meters.

[0032] Since the misalignment angle between the supply ship and the receiving ship is positive in the clockwise direction, the misalignment angle between the two ships will be different when the receiving ship is on different sides of the supply ship. When the receiving ship is on the right side of the supply ship, the preset safe misalignment angle range for entering the positioning phase is: The preset safety misalignment angle range for the safe evacuation phase is: When the receiving vessel is on the left side of the supply vessel, the preset safe offset angle range for entering the positioning phase is: The preset safety misalignment angle range for the safe evacuation phase is: ; This refers to the offset angle between the supply vessel and the receiving vessel. The specific values ​​of the preset safe lateral distance range and the preset safe offset angle range are set based on human experience and are preset constraint ranges used to narrow the expected operation path range when planning the expected operation path of the receiving vessel.

[0033] In one example, the receiving vessel is a container ship with an overall length of 203 meters, a beam of 35.2 meters, a depth of 17.5 meters, a draft of 9.2 meters, and a displacement of 50,000 tons. The receiving station is located 25.2 meters forward of the center of gravity. The supply vessel has an overall length of 166 meters, a beam of 17.6 meters, a depth of 13.8 meters, a draft of 8.0 meters, and a displacement of 21,000 tons. The dispatch station is located 19.3 meters forward of the center of gravity. The supply is dry cargo, with a quantity of 2 tons and a supply speed of 1.5 meters per second. The target sea area for the supply operation is the coastal waters of the Shandong Peninsula, with a sea state of Class 2 and sea state parameters of: wind speed of 10.7 meters per second, current speed of 0.9 meters per second, significant wave height of 2 meters, wave period of 3.0 seconds, and wave direction of 180°. Under these resupply conditions, the preset safe lateral distance range for the entry and safe withdrawal phases is 31.4 meters to 66.4 meters.

[0034] Step 2: Based on the six-degree-of-freedom motion prediction model of the supply ship and the receiving ship, the ship motion controller is used to control the two ships to perform cooperative motion simulation according to their respective expected operation paths, and to obtain the supply operation data corresponding to each expected operation path of the supply ship and the receiving ship. The supply operation data of each ship includes multiple sea state parameters and the motion state parameters and control commands of each ship at multiple simulation times under each sea state parameter. The six-degree-of-freedom motion prediction model of each ship is constructed by introducing inter-ship interaction loads into the six-degree-of-freedom motion equations of each ship. The inter-ship interaction loads characterize the interference effect of the inter-ship effect of the supply ship and the receiving ship on the two ships. The ship motion controller of each ship generates control commands and controls the motion state of the ship.

[0035] This application considers the influence of inter-ship loads on the motion states of the two ships during replenishment operations. It introduces inter-ship loads into the traditional six-degree-of-freedom equations of motion to construct six-degree-of-freedom motion prediction models for the replenishment and receiving ships, respectively, to predict the motion of each ship in each degree of freedom. The six-degree-of-freedom motion prediction model for each ship can be simplified as follows: (1) in, m It is the quality of the ship. m x It is the added mass of the ship in the longitudinal direction. m y It is the ship's added mass in the lateral direction. u It is the longitudinal speed of the ship. v It is the ship's lateral speed. p It is the ship's roll speed. r It is the ship's turning speed. I xx It is the moment of inertia of the ship's hull about the x-axis. I zz It is the moment of inertia of the ship's hull about the z-axis.J xx It is the additional moment of inertia of the hull about the x-axis. J zz It is the additional moment of inertia of the hull about the z-axis. Z G It is the amount of heave motion. θ It is the pitch angle; b and c are all hydrodynamic derivatives. The subscript zz indicates the z-axis, the subscript yy indicates the y-axis, the subscript yz indicates the coupling between the y-axis and the z-axis, and the subscript zy indicates the coupling between the z-axis and the y-axis. X It is the longitudinal force acting on the hull. Y It is the lateral force acting on the hull. K It is the rolling moment acting on the hull. N It is the bow roll moment acting on the hull, subscript H Indicates the hydrodynamic forces acting on the ship's hull, subscript P Indicates propeller thrust, subscript R Indicates rudder force, subscript Indicates wind load, subscript Indicates wave load, subscript L Indicates the load applied by the overhead cable, subscript SS Indicates the loads acting between ships. F z Indicates the force of the wave in the direction of heave. M y This indicates the forced torque of the wave in the direction of pitching.

[0036] The hydrodynamic loads, wind loads, wave loads, propeller thrust, and rudder force acting on the hull can be determined using conventional methods, and will not be elaborated upon in this application. The loads acting on the overhead cables are determined using the method described in the applicant's previously published patent (CN118348826 A). Since the inter-ship loads have a particularly important impact on the safety of replenishment operations, and existing methods for determining inter-ship loads require model tests and viscous flow calculations for each replenishment condition parameter, the workload is large and the computational efficiency is low, failing to meet the needs of real-time forecasting. To achieve efficient and accurate calculation of inter-ship loads, this application analyzes various influencing factors related to inter-ship loads and obtains the inter-ship loads in real time by monitoring the motion state parameters of the replenishment vessel and the receiving vessel.

[0037] In one embodiment, the motion state parameters include the ship's position and speed, and determining the inter-ship loads on any ship includes: The longitudinal distance along the length of the supply ship's center of gravity and the receiving ship's center of gravity is determined based on their positions. And the lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship along the beam direction. ; When the longitudinal distance between the center of gravity of the supply ship and the center of gravity of the receiving ship along the length of the ship. The longitudinal distance requirement is not met, or the lateral distance between the center of gravity of the supply vessel and the center of gravity of the receiving vessel along the beam direction is not met. If the lateral distance requirement is not met, the inter-ship load is set to 0; otherwise, the longitudinal distance between the center of gravity of the supply ship and the center of gravity of the receiving ship is used. Horizontal distance The speeds of the supply ship and the receiving ship, combined with their dimensional parameters, are used to determine the inter-ship loads. The longitudinal and lateral distance requirements are determined based on the ship's dimensional parameters and actual replenishment operation experience. This application specifies that the longitudinal distance requirement between the center of gravity of the supply ship and the center of gravity of the receiving ship should not exceed 174.28 meters, and the lateral distance requirement should not exceed 143.9 meters.

[0038] Inter-ship loads only need to be considered when the two ships are within the aforementioned distance range. This application takes into account the influence of lateral distance, longitudinal distance, and the size and speed of the two ships on the inter-ship loads, and constructs a calculation formula for the inter-ship loads. Inter-ship loads include inter-ship forces and inter-ship moments; Determine the intership forces acting on any ship for: (2) Determine the inter-ship moment acting on any ship for: (3) in, L It is the average length of the supply ship and the receiving ship. L F He is the captain of the supply ship. L L He is the captain of the receiving ship. U F It's the speed of the supply ship. U L It is the speed of the receiving ship; , , , , , , , It is the inter-ship force coefficient. b 1. b 2. b 3. b 4. b 5. b 6. b 7. b 8. b 9.b 10 , b 11 These are the inter-ship moment coefficients. The inter-ship force coefficient and inter-ship moment coefficient are obtained beforehand by using CFD or model testing methods to determine the inter-ship forces and moments experienced by the ships under specified replenishment conditions. These are then fitted using the formulas for calculating inter-ship forces and moments. When the lateral force experienced by the receiving ship in still water exceeds 120% of the lateral force under single-ship conditions, the inter-ship effect is considered to be in effect. The inter-ship force coefficient and inter-ship moment coefficient are then obtained by fitting the inter-ship forces and moments under specified replenishment conditions where the inter-ship effect exists. This method only requires prior model testing or numerical simulation under specified replenishment conditions. During subsequent real-time forecasting, the inter-ship forces and moments under any replenishment conditions can be directly calculated using the formulas, effectively improving the real-time performance of the motion forecasting model.

[0039] The intership forces and intership moments calculated according to formulas (2) and (3) are the resultant forces and resultant moments acting on the ship. The longitudinal forces and transverse forces acting on the hull can be obtained by decomposing the ship into its motion coordinate system.

[0040] After calculating all the external excitation forces on the ship, the high and low frequency motions of each ship in each degree of freedom at the current simulation moment can be obtained by inputting them into the six-degree-of-freedom motion prediction model of formula (1). Then, the sway, roll, and bow roll are calculated and the ship's position and heading angle are updated in time, and fed back to the ship motion controller to calculate the control command for the next simulation moment.

[0041] In one embodiment, the motion state parameters include the ship's position, heading angle, and six-degree-of-freedom motion speed, and the control commands include rudder angle commands and propeller speed commands. The ship motion controller for each ship includes a heading angle controller and a speed controller. The input to the bow angle controller is the deviation between the expected and actual bow angle values, and the output is the rudder angle command. Any controller model can be used to construct the ship's motion controller. In one example, a ship motion controller is constructed based on a PID controller, such as... Figure 3 As shown, by monitoring the ship's heading angle in real time and adjusting the rudder angle command, the ship's lateral and bow roll motions can be changed, thereby effectively avoiding collisions and capsizing accidents. Taking the heading angle controller of the receiving ship as an example, the specific formula is expressed as follows: (4) in, It is the heading angle deviation. This is the actual value of the bow angle of the receiving vessel. It is the expected value of the bow angle of the receiving ship. It is a rudder angle command. It is the proportional coefficient of the heading angle controller. It is the integral coefficient of the heading angle controller. These are the differential coefficients of the heading angle controller. k It is a simulation moment. N It is the maximum value at the simulation time.

[0042] Since the supply ship always sails along the predetermined target point, its expected heading angle is a known quantity. The expected heading angle of the receiving ship is determined based on the expected heading angle of the supply ship, as well as the speeds and distances between the two ships. ,in, It is the expected value of the bow angle of the supply ship. It is the longitudinal component of the actual resultant velocity of the six-degree-of-freedom motion of the receiving ship. It is the actual lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship. It is the expected lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship.

[0043] The speed controller's input is the deviation between the expected and actual results of the six-degree-of-freedom motion velocity, and its output is the propeller speed command. By monitoring the ship's six-degree-of-freedom motion velocity in real time and adjusting the propeller speed command, the ship's motion state is controlled, thereby ensuring its stability. Taking the speed controller of a receiving ship as an example, the specific formula is as follows: (5) in, It's a speed deviation. It receives the actual value of the resultant velocity of the ship's six degrees of freedom motion. It is the expected value of the resultant velocity of the receiving ship's six degrees of freedom motion. It is a propeller speed command. It is the proportional coefficient of the speed controller. It is the integral coefficient of the speed controller. These are the differential coefficients of the speed controller.

[0044] Similarly, the expected value of the resultant velocity of the receiving ship's six degrees of freedom motion is determined based on the expected value of the resultant velocity of the supply ship's six degrees of freedom motion and the distance between the two ships. ,in, This is the expected value of the resultant velocity of the supply ship's six degrees of freedom motion. It is a proportionality coefficient that converts distance into speed. This is the actual longitudinal distance between the sending station on the supply ship and the receiving station on the receiving ship. It is the expected longitudinal distance between the sending station on the supply ship and the receiving station on the receiving ship.

[0045] By changing the replenishment parameters such as the speed, sea state, and wave direction of the two ships during the parallel replenishment phase, and performing cooperative motion simulations of the replenishment and receiving ships along various expected operational paths, replenishment operation data corresponding to each expected operational path can be obtained. For each expected operational path, this application sets the speed of the two ships within the range of 10 knots to 18 knots; the wave direction is set with 180° as the initial wave direction, 0° as the minimum wave direction, and 5° intervals until the wave direction range of 0° to 360° is traversed; the initial sea state is level 2, and the maximum sea state is level 6, traversing sea states 2, 3, 4, 5, and 6 to complete the lateral replenishment simulation.

[0046] Step 3: Filter the replenishment operation data corresponding to all expected operation paths of the replenishment ship and the receiving ship, and construct a safe operation sample database.

[0047] Since both the supply vessel and the receiving vessel have multiple planned operational paths, and each planned operational path corresponds to various supply conditions, in order to subsequently train the optimal lateral supply control strategy model, it is necessary to filter all the obtained supply operation data to obtain safe operation data. Each planned operational path includes an entry and positioning phase, a parallel supply phase, and a safe withdrawal phase; motion state parameters include the vessel's position and speed. The resulting safe operation sample database includes: The replenishment operation data corresponding to each expected operation path of the replenishment ship and the receiving ship are screened to determine several safe operation paths that meet the safety operation constraints during the parallel replenishment phase, as well as the safe speed range corresponding to each safe operation path. The replenishment operation data of each safe operation path within the safe speed range are used as safe operation data, and a safe operation sample database is constructed using the safe operation data corresponding to each safe operation path. Each sample in the safe operation sample database includes a simulation time step. t Corresponding sea state parameters Simulation time step t Corresponding motion state parameters Simulation time step t Corresponding control commands and simulation time step t +1 corresponds to the sea state parameter Each safe operation sample contains time information, motion state information that changes over time, and control command information. The rudder and propeller control commands corresponding to the safe operation path and safe speed are used as state information. These are matched with the predicted six degrees of freedom motion and trajectory of the ship during the simulation to form a safe operation sample containing both time and space characteristics.

[0048] The safe operating constraints are used to ensure the safe and stable replenishment operation between the two vessels, requiring that the lateral forces generated within the steerable range balance the forces and moments acting on the two vessels. In one embodiment, the safe operating constraints include: under full rudder conditions, the resultant lateral forces acting on the two vessels are in opposite directions, and the resultant bow moments acting on the two vessels are in opposite directions, with each vessel's pitch angle not exceeding 2° and roll angle not exceeding 5°.

[0049] Specifically, such as Figure 2 As shown, when the supply ship is on the left side of the receiving ship, the resultant lateral force on the supply ship satisfies... The resultant bow roll moment experienced by the supply ship satisfies The resultant lateral force on the receiving ship satisfies The resultant bow roll moment experienced by the receiving ship satisfies ,in, It is the lateral force generated at a full rudder angle of 35°. It is the lateral moment generated under a full rudder angle of 35°; the ship's motion coordinate system has the ship's center of gravity as the origin, the starboard direction as the positive y-axis, and the bow direction as the positive x-axis.

[0050] Under these safety constraints, taking a replenishment speed of 12 knots as an example, the safe lateral distance that the receiving ship should actually meet when transitioning from the entry and positioning phase to the parallel replenishment phase is 34.5 meters to 66.2 meters, and the safe offset angle range is 5.6° to 28.4°. When transitioning from the parallel replenishment phase to the safe withdrawal phase, the safe lateral distance that should actually meet is 34.5 meters to 66.4 meters, and the safe offset angle range is -25.6° to 5.9°. All replenishment operation data that meet these conditions are considered safe operation data, which can then be used to construct a safe operation sample database, laying an important foundation for subsequent model training.

[0051] Step 4: Using the sea state parameters corresponding to each simulation moment in the safety operation sample database as the state, the control commands corresponding to each simulation moment as the action, and the motion state parameters corresponding to each simulation moment as the reward, a lateral resupply control strategy model is trained using a deep reinforcement learning algorithm.

[0052] Deep reinforcement learning algorithms can employ any existing algorithm. In one example, the evaluation network and target network are generated using the Deep Q-Network (DQN) algorithm. The specific construction process is as follows: Figure 4 As shown, based on the current state (Simulation time step) t (Sea state parameters such as wave height and direction) are used to select actions using a greedy strategy. (Control commands such as rudder angle and propeller speed at simulation time step t), interact with the environment to obtain rewards. (Simulation time step)t (The following are motion state parameters: the position of the supply ship, the position of the receiving ship, etc.) and the state at the next moment. (Simulation time step) t+ The transfer sample consists of sea state parameters such as wave height and wave direction (as shown in Figure 1). The data is stored in the experience replay pool. Training begins when the data in the experience replay pool is greater than or equal to the training samples in each round. The predicted value and temporal difference target value are calculated using the current network and the target network. The mean squared error loss function is calculated, and the current network parameters are updated using the gradient descent algorithm. Simultaneously, every [time period missing]... The target network parameters are updated step by step, and the optimal maneuvering strategy for lateral replenishment of two ships under any wave environment is obtained through training.

[0053] Step 5: Use the trained lateral supply control strategy model to make real-time decisions on the current supply conditions and obtain the control commands at the current moment.

[0054] The lateral replenishment control strategy model trained by the method of this application can not only provide a reference scheme for lateral replenishment operations at sea for various types of vessels, but also provide the best safe replenishment control strategy according to the vessel's motion state, thereby effectively improving sea state adaptability.

[0055] The above descriptions are merely preferred embodiments of this application, and this application is not limited to the above embodiments. It is understood that other improvements and variations that can be directly derived or conceived by those skilled in the art without departing from the spirit and concept of this application should be considered to be included within the protection scope of this application.

Claims

1. A two-ship lateral replenishment decision-making method considering inter-ship effects, characterized in that, The decision-making method for lateral replenishment during navigation of the two ships includes: Multiple expected operational paths for the supply vessel and the receiving vessel are determined using path planning methods based on replenishment operational parameters; the replenishment operational parameters include the motion state parameters of the supply vessel and the receiving vessel and the sea state parameters of the target sea area; Based on the six-degree-of-freedom motion prediction model of the supply ship and the receiving ship, a ship motion controller is used to control the two ships to perform cooperative motion simulation according to their respective expected operation paths. The supply operation data corresponding to each expected operation path of the supply ship and the receiving ship is obtained. The supply operation data of each ship includes multiple sea state parameters and the motion state parameters and control commands of each ship at multiple simulation moments under each sea state parameter. The six-degree-of-freedom motion prediction model of each ship is constructed by introducing inter-ship interaction loads into the six-degree-of-freedom motion equations of each ship. The inter-ship interaction loads characterize the interference effect of the inter-ship effect between the supply ship and the receiving ship on the two ships. The ship motion controller of each ship generates control commands and controls the motion state of the ship. The replenishment operation data corresponding to all expected operation paths of the replenishment ship and the receiving ship are filtered, and a safe operation sample database is constructed. The sea state parameters corresponding to each simulation time in the safe operation sample database are used as the state, the control commands corresponding to each simulation time are used as the action, and the motion state parameters corresponding to each simulation time are used as the reward. A lateral replenishment control strategy model is trained using a deep reinforcement learning algorithm. The trained lateral supply control strategy model is used to make real-time decisions on the current supply conditions and obtain the control commands at the current moment.

2. The two-ship lateral replenishment decision-making method according to claim 1, characterized in that, The motion state parameters include the ship's position and speed, and the inter-ship loads on any ship include: The longitudinal distance along the length of the supply ship's center of gravity and the receiving ship's center of gravity is determined based on their positions. And the lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship along the beam direction. ; When the longitudinal distance between the center of gravity of the supply ship and the center of gravity of the receiving ship along the length of the ship. The longitudinal distance requirement is not met, or the lateral distance between the center of gravity of the supply vessel and the center of gravity of the receiving vessel along the beam direction is not met. If the lateral distance requirement is not met, the inter-ship load is set to 0; otherwise, the longitudinal distance between the center of gravity of the supply ship and the center of gravity of the receiving ship is used. Horizontal distance The speeds of the supply ship and the receiving ship, combined with their dimensional parameters, are used to determine the inter-ship loads.

3. The two-ship lateral replenishment decision-making method according to claim 2, characterized in that, The inter-ship loads include inter-ship forces and inter-ship moments; Determine the intership forces acting on any ship for: Determine the inter-ship moment acting on any ship for: in, L It is the average length of the supply ship and the receiving ship. L F He is the captain of the supply ship. L L He is the captain of the receiving ship. U F It's the speed of the supply ship. U L It is the speed of the receiving ship; , , , , , , , It is the inter-ship force coefficient. b 1. b 2. b 3. b 4. b 5. b 6. b 7. b 8. b 9. b 10 , b 11 It is the inter-ship moment coefficient.

4. The two-ship lateral replenishment decision-making method according to claim 1, characterized in that, The motion state parameters include the ship's position, heading angle, and six-degree-of-freedom motion speed; the control commands include rudder angle commands and propeller speed commands; and each ship's motion controller includes a heading angle controller and a speed controller. The input to the heading angle controller is the deviation between the expected heading angle value and the actual heading angle value, and the output of the heading angle controller is the rudder angle command; The input to the speed controller is the deviation between the expected value and the actual value of the resultant velocity of the six degrees of freedom motion, and the output of the speed controller is the propeller speed command.

5. The two-ship lateral replenishment decision-making method according to claim 4, characterized in that, Determine the expected bow angle of the receiving vessel. ,in, It is the expected value of the bow angle of the supply ship. It is the longitudinal component of the actual resultant velocity of the six-degree-of-freedom motion of the receiving ship. It is the actual lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship. It is the expected lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship.

6. The two-ship lateral replenishment decision-making method according to claim 4, characterized in that, Determine the expected value of the resultant velocity of the six-degree-of-freedom motion of the receiving ship. ,in, This is the expected value of the resultant velocity of the supply ship's six degrees of freedom motion. It is a proportionality coefficient that converts distance into speed. This is the actual longitudinal distance between the sending station on the supply ship and the receiving station on the receiving ship. It is the expected longitudinal distance between the sending station on the supply ship and the receiving station on the receiving ship.

7. The two-ship lateral replenishment decision-making method according to claim 1, characterized in that, Each planned operational path includes an entry and positioning phase, a parallel replenishment phase, and a safe evacuation phase; the motion state parameters include the ship's position and speed, and the constructed safe operational sample database includes: The replenishment operation data corresponding to each expected operation path of the replenishment ship and the receiving ship are screened to determine several safe operation paths that meet the safety operation constraints during the parallel replenishment phase, as well as the safe speed range corresponding to each safe operation path; the entry and positioning phase is the phase in which the two ships gradually approach each other from the predetermined position, the parallel replenishment phase is the phase in which the two ships sail in parallel to transfer materials, and the safe withdrawal phase is the phase in which the two ships gradually move away from the predetermined position. The resupply operation data for each safe operation path within the safe speed range is used as the safe operation data. A safe operation sample database is constructed using the safe operation data corresponding to each safe operation path. Each sample in the safe operation sample database includes a simulation time step. t Corresponding sea state parameters Simulation time step t Corresponding motion state parameters Simulation time step t Corresponding control commands and simulation time step t +1 corresponds to the sea state parameter .

8. The two-ship lateral replenishment decision-making method according to claim 7, characterized in that, The safety operating constraints include: under full rudder conditions, the resultant lateral forces on the two vessels are in opposite directions, and the resultant bow moments on the two vessels are in opposite directions, and the pitch angle of each vessel does not exceed 2° and the roll angle does not exceed 5°.

9. The two-ship lateral replenishment decision-making method according to claim 1, characterized in that, Each anticipated operational path includes an entry and positioning phase, a parallel replenishment phase, and a safe evacuation phase. Multiple anticipated operational paths for any given vessel include: Based on the replenishment operation parameters, multiple expected operation paths of the replenishment ship are determined using a path planning algorithm, and there are no obstacles in the predetermined area around each expected operation path of the replenishment ship. Based on the replenishment operating parameters, determine the lateral distance between the center of gravity of the replenishment ship and the center of gravity of the receiving ship, as well as the misalignment angle between the replenishment ship and the receiving ship. The misalignment angle is the angle between the line connecting the sending station on the replenishment ship and the receiving station on the receiving ship and the width direction of the replenishment ship, and is positive in the clockwise direction. Based on each expected operational path of the supply ship, a path planning algorithm is used to determine multiple expected operational paths of the receiving ship when the lateral distance between the center of gravity of the receiving ship and the center of gravity of the supply ship is within a preset safe lateral distance range and the misalignment angle between the receiving ship and the supply ship is within a preset safe misalignment angle range during the entry and positioning phase and the safe withdrawal phase. The entry and positioning phase is the phase in which the two ships gradually approach each other from the predetermined position. The parallel supply phase is the phase in which the two ships sail in parallel to transfer materials. The safe withdrawal phase is the phase in which the two ships gradually move away from the predetermined position.

10. The two-ship lateral replenishment decision-making method according to claim 9, characterized in that, The preset safe lateral distance ranges for entering the positioning phase and the safe evacuation phase are: W +5 meters W +40 meters, It is the lateral distance between the center of gravity of the supply ship and the center of gravity of the receiving ship. W It is the sum of half the width of the supply ship and the receiving ship, in meters; When the receiving vessel is positioned on the right side of the supply vessel, the preset safe offset angle range for entering the positioning phase is: The preset safety misalignment angle range for the safe evacuation phase is: When the receiving vessel is on the left side of the supply vessel, the preset safe offset angle range for entering the positioning phase is: The preset safety misalignment angle range for the safe evacuation phase is: ; It is the offset angle between the supply ship and the receiving ship.

Images