Stabilizing control method and system for fuel filling ship at zero navigational speed

Abstract

The invention provides a stabilization control method and system for a fuel filling ship at a zero navigational speed. The method provided by the invention comprises the following steps: acquiring rolling data and wind wave data of a refueling ship under a target sea condition in a fuel refueling process in real time; the rolling data and the wind wave data are processed based on a radial basis function neural network, and the precession angular velocity of the stabilization gyro device under the target sea condition is obtained through calculation; the stabilization gyro device is controlled to rotate at the precession angular speed, the stabilization gyro device generates a gyroscopic effect and generates rolling moment, the rolling moment is converted into stabilizing moment for resisting rolling through precession motion, and the rolling angle of the filling ship is reduced; and the stabilizing moment is opposite to the vibration direction of the rolling. According to the method and system, rolling control of the filling ship at the low speed or the zero navigational speed is achieved, the stabilization torque can be adjusted in real time according to sea wave changes, and the stability and safety of the filling ship in the fuel filling operation process are improved.

Description

Technical Field

[0001] The present application relates to the technical field of ship anti-rolling control, and in particular to a method and system for anti-rolling control of a fuel tanker at zero speed. Background Art

[0002] With the development of maritime traffic and transportation, the demand for ships to carry out fuel bunkering operations at sea is increasing. However, the marine environment is complex and changeable, especially in severe sea conditions, the oil supply ship and the oil receiving ship are easily affected by wind and waves and produce violent roll. This roll not only affects the safety and continuity of the bunkering operation, but also increases the loss of equipment and the risk of personnel. Therefore, reducing the roll of the ship at low speed or static state is of great significance to ensure the safety of bunkering operations.

[0003] Most of the anti-rolling technologies used by existing bunkering ships in severe sea conditions rely on the physical structure of the ship, such as anti-roll fins or water tank anti-rolling. However, anti-roll fins usually require a certain speed to work, and the effect is significantly reduced at low speed or stationary state; water tank anti-rolling requires adjusting the amount of water in the water tank. This process is not only time-consuming, but may also affect the load balance of the ship. In addition, these systems are generally passively controlled and lack the ability to be intelligent and adaptive, making it difficult to cope with complex and changeable sea conditions. In terms of coping with the need for roll control at low or zero speeds, existing technical solutions have not yet provided an effective solution, especially in bunkering ship operation scenarios, there is a lack of an efficient anti-rolling method that can adapt to wind and waves in real time.

[0004] Therefore, there is an urgent need for a method to achieve roll control of the bunkering ship at low or zero speed, and to adjust the anti-roll torque in real time according to the changes in waves to improve the stability and safety of the bunkering ship during the fuel bunkering operation. Summary of the invention

[0005] In view of this, the present application provides a roll reduction control method and system for a fuel bunkering ship at zero speed, which is used to achieve roll control of the bunkering ship at low or zero speed, and can adjust the roll reduction torque in real time according to the changes in waves, thereby improving the stability and safety of the bunkering ship during the fuel bunkering operation.

[0006] Specifically, the present application is implemented through the following technical solutions: The first aspect of the present application provides a method for anti-rolling control of a fuel bunkering ship at zero speed, the method comprising: Real-time collection of rolling data and wind and wave data of the bunkering ship under target sea conditions during the bunkering process; the speed of the bunkering ship under the target sea conditions is 0; Processing the roll data and wind and wave data based on a radial basis function neural network to calculate the precession angular velocity of the anti-roll gyro device under the target sea condition; The radial basis function neural network characterizes the relationship between the roll and disturbance of the bunkering ship. The radial basis function neural network calculates the precession angular velocity of the anti-roll gyro device by analyzing the roll angle and angular velocity under the target sea conditions. Under different sea conditions, the radial basis function neural network adaptively adjusts the weight based on the sea conditions. The radial basis function neural network includes a command filter and a nonlinear disturbance observer. The command filter smoothes the input data and processes multi-order nonlinear dynamics step by step in combination with multi-step recursive design and backstepping control. The nonlinear disturbance observer determines the nonlinear disturbance based on the input data and dynamically compensates the radial basis function neural network based on the nonlinear disturbance. The anti-roll gyro device is controlled to rotate at the precession angular velocity, the anti-roll gyro device generates a gyro effect, generates a roll moment, and converts the roll moment into a stabilizing moment to counteract the roll through the precession motion, thereby reducing the roll angle of the bunkering ship; the stabilizing moment is opposite to the vibration direction of the roll.

[0007] The second aspect of the present application provides a roll stabilization control system for a fuel tanker at zero speed, the system comprising a sensor module, a neural network controller and a roll stabilization gyro device connected in sequence; wherein: The sensor module is used to collect the rolling data and wind and wave data of the bunkering ship in real time under the target sea conditions during the fuel bunkering process; the speed of the bunkering ship under the target sea conditions is 0; The neural network controller is used to process the roll data and wind and wave data to calculate the precession angular velocity of the anti-roll gyro device under the target sea condition; The radial basis function neural network characterizes the relationship between the roll and disturbance of the bunkering ship. The radial basis function neural network calculates the precession angular velocity of the anti-roll gyro device by analyzing the roll angle and angular velocity under the target sea conditions. Under different sea conditions, the radial basis function neural network adaptively adjusts the weight based on the sea conditions. The radial basis function neural network includes a command filter and a nonlinear disturbance observer. The command filter smoothes the input data and processes multi-order nonlinear dynamics step by step in combination with multi-step recursive design and backstepping control. The nonlinear disturbance observer determines the nonlinear disturbance based on the input data and dynamically compensates the radial basis function neural network based on the nonlinear disturbance. The neural network controller is also used to control the anti-roll gyro device to rotate at the precession angular velocity, and the anti-roll gyro device generates a gyro effect and a roll moment, and converts the roll moment into a stabilizing moment to counteract the roll through the precession motion, thereby reducing the roll angle of the bunkering ship; the stabilizing moment is opposite to the vibration direction of the roll.

[0008] The present application provides a method and system for anti-roll control of a fuel filling ship at zero speed. In the first aspect, a radial basis function neural network is used to process roll data and wind and wave data, calculate the precession angular velocity of the anti-roll gyro device, and then control the anti-roll gyro device based on the precession angular velocity to achieve anti-roll control of the fuel filling ship. In the roll control of the fuel filling ship, the radial basis function neural network can accurately model the complex relationship between the roll and disturbance of the fuel filling ship. By learning input variables such as roll angle and wind and wave data, a control signal for adjusting the precession angular velocity of the anti-roll gyro device can be generated in real time. Moreover, as the wind and waves change, the radial basis function neural network can optimize the control performance by adaptively adjusting its weights without the need to accurately linearize the model, thereby improving the accuracy of the anti-roll control of the fuel filling ship, achieving the roll control of the fuel filling ship at low speed or zero speed, and can adjust the anti-roll torque in real time according to the changes in wind and waves, thereby improving the stability and safety of the fuel filling ship during the fuel filling operation. Secondly, in the design process of the radial basis function neural network, the radial basis function neural network is optimized through the command filter and the nonlinear disturbance observer. The command filter avoids the adverse oscillation of the radial basis function neural network by smoothing the reference command. The command filter backstepping control combined with the multi-step recursive design can process multi-order nonlinear dynamics step by step to ensure the stability of the radial basis function neural network under complex sea conditions. The command filter can quickly adjust the precession angular velocity of the anti-roll gyro device to respond to external disturbances, so that the bunkering ship always remains in a stable state. In addition, the disturbance caused by wind and waves is strongly nonlinear and random. The nonlinear disturbance observer can accurately estimate these disturbance values ​​and correct the radial basis function neural network through a compensation mechanism. The disturbance estimation function of the nonlinear disturbance observer reduces the computational burden of the radial basis function neural network and the command filter. Even under the superposition of multiple disturbance sources (such as wind and waves, load changes), the nonlinear disturbance observer can still maintain good observation accuracy and dynamic compensation capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A flow chart of a roll reduction control method for a fuel bunkering ship at zero speed provided in Example 1 of the present application; Figure 2 This is a schematic diagram of the structure of the anti-roll gyro device shown in this application; Figure 3 The working principle diagram of the anti-roll gyro device shown in this application; Figure 4 A schematic diagram of a gyro coordinate system shown in the present application; Figure 5 This is a schematic diagram of the structure of a roll stabilization control system for a fuel bunkering ship at zero speed provided in Example 2 of the present application. DETAILED DESCRIPTION

[0010] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

[0011] The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in this article refers to and includes any or all possible combinations of one or more associated listed items.

[0012] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0013] Specific embodiments are given below to introduce the technical solution of the present application in detail.

[0014] Figure 1 This is a flow chart of the anti-rolling control method of the fuel bunkering ship at zero speed provided in Example 1 of the present application. Figure 1 , the method provided in this embodiment may include: S101. Real-time collection of rolling data and wind and wave data of a bunkering ship under a target sea condition during a fuel bunkering process; the speed of the bunkering ship under the target sea condition is 0.

[0015] Specifically, roll data refers to the data generated during the roll motion of the bunkering ship (the swing motion of the bunkering ship around its own longitudinal axis), and includes roll angle, roll angular velocity, roll angular acceleration, frequency and amplitude, etc. Wind and wave data refers to the characteristics of wind and waves on the sea where the bunkering ship is located, which is related to the sea conditions, and includes wind speed, wind direction, wave height, wave direction, wave spectrum, etc.

[0016] Furthermore, the target sea condition refers to the sea condition where the bunkering ship is at a low speed or zero speed. Since the bunkering ship is usually at a low speed or stationary state during the fuel filling process, this application is mainly aimed at the bunkering ship at a low speed or zero speed.

[0017] In a specific implementation, the real-time collection of the rolling data and wind and wave data of the bunkering ship under the target sea conditions during the fuel bunkering process includes: determining a first installation position of the sensor based on the propagation characteristics of the rolling motion and the structure of the bunkering ship; adjusting the first installation position based on the center of gravity position of the bunkering ship and the distribution characteristics of the sensor to obtain the installation position of the sensor; installing the sensor based on the installation position, and collecting the rolling data and wind and wave data of the bunkering ship under the target sea conditions during the fuel bunkering process in real time based on the sensor.

[0018] Specifically, the sensor at least includes an acceleration sensor, an angular velocity sensor, a wind and wave data sensor, etc. The acceleration sensor collects the roll acceleration of the bunkering ship, the angular velocity sensor collects the roll angular acceleration of the bunkering ship, and the wind and wave data sensor collects the wind and wave data of the bunkering ship.

[0019] In specific implementation, the propagation characteristics of the rolling motion are usually related to the geometric shape of the bunkering ship, the damping characteristics and rolling frequency response of the bunkering ship, and the center of mass position of the bunkering ship. When considering the structure of the bunkering ship, the functional divisions of the bunkering ship (such as the fuel bunkering platform and cargo hold) must be considered. When determining the first installation position of the sensor, priority is given to the area with the greatest impact on the rolling, such as near the rolling frequency node (usually in the center area of ​​the bunkering ship) or near the fuel filling position. Furthermore, since the center of gravity of the bunkering ship is the center point of the rolling, if the first installation position deviates too far from the center point, it will cause inaccurate data, and the distribution of sensors should avoid interference with each other. It is usually determined that the sensors are installed in a vertical position above the center of gravity or in the deck area where the rolling motion is most significant. Based on the installed different types of sensors, the rolling data and wind and wave data of the bunkering ship are collected separately.

[0020] Specifically, a preliminary analysis is performed based on the rolling motion characteristics of the bunkering ship (such as rolling period, amplitude, etc.) combined with structural parameters such as the stiffness and mass distribution of the hull. The degree of influence of the rolling motion on different positions is determined by theoretical calculation or numerical simulation (such as using a ship motion model or a finite element analysis model), and the position with the greatest influence is determined as the first installation position of the sensor. Further, based on the position of the center of gravity of the hull and combined with the mass distribution of the hull, the rolling response of different positions of the hull is determined. Based on the rolling response, the first installation position is adjusted to obtain the installation position of the sensor, and the adjustment scale is related to the size of the rolling response.

[0021] S102: Process the roll data and wind and wave data based on a radial basis function neural network to calculate the precession angular velocity of the anti-roll gyro device under the target sea condition.

[0022] Among them, the radial basis function neural network characterizes the relationship between the roll and disturbance of the bunkering ship, and the radial basis function neural network calculates the precession angular velocity of the anti-roll gyro device by analyzing the roll angle and angular velocity under the target sea condition; under different sea conditions, the radial basis function neural network adaptively adjusts the weight based on the sea condition; the radial basis function neural network includes a command filter and a nonlinear disturbance observer, the command filter smoothes the input data, and combines multi-step recursive design and backstepping control to process multi-order nonlinear dynamics step by step; the nonlinear disturbance observer determines the nonlinear disturbance based on the input data, and dynamically compensates the radial basis function neural network based on the nonlinear disturbance.

[0023] Specifically, the radial basis function neural network characterizes the relationship between the roll and disturbance of the bunkering ship, and calculates the precession angular velocity of the anti-roll gyro device by analyzing the roll angle and wind and wave data under different sea conditions. It should be noted that due to the different roll angles and wind and wave data under different sea conditions, the corresponding weight parameters of the radial basis function neural network will also be adjusted accordingly.

[0024] Furthermore, the radial basis function neural network includes a command filter and a nonlinear disturbance observer. The command filter is designed based on the roll data and the wind and wave data, and is used to smooth the roll data and the wind and wave data, and to determine the corresponding smoothing method based on the dynamic characteristics of the roll data and the wind and wave data. When the dynamic characteristics are multi-order nonlinear dynamics, multi-step recursive design and backstepping control are used for step-by-step processing. The nonlinear disturbance observer is used to estimate the disturbance of the roll data and the wind and wave data, determine the disturbance information, and dynamically compensate the radial basis function neural network based on the disturbance information.

[0025] It should be noted that the command filter and nonlinear disturbance observer are used in both the training and use stages of the radial basis function neural network. In the training stage, the main goal of the command filter is to smooth the input data, reduce noise and unnecessary high-frequency fluctuations, make the training data more stable, and help the neural network effectively learn the trend of the data without being disturbed by random noise in the data. In the use stage, the command filter is also used to smooth the input data, but more to solve the noise or irregular fluctuations of the input data in actual operation and maintain the stability of the system.

[0026] Furthermore, during the training phase, the nonlinear disturbance observer is used to estimate and compensate for external disturbances (such as wind and waves, changes in hull loads, etc.), and helps the radial basis function neural network to better adapt to complex environments by analyzing the disturbance factors in the training data and generating corresponding disturbance information. During the use phase, the nonlinear disturbance observer continues to estimate and compensate for external disturbances. In actual environments, factors such as wind and waves, tides, etc. may affect the movement of ships, and the disturbance observer helps the radial basis function neural network deal with these uncertainties.

[0027] Optionally, the command filter is designed based on the roll data and wind and wave data, including: (1) Based on the roll data and the wind and wave data, a ship roll-gyro combined dynamic model is constructed; the ship roll-gyro combined dynamic model describes the roll behavior of the bunkering ship under the control of external disturbance and anti-roll gyro device.

[0028] Specifically, the ship roll-gyro joint dynamic model describes the roll behavior of the bunkering ship under the control of external disturbances (such as wind and waves) and the anti-roll gyro device. The ship roll-gyro joint dynamic model consists of a gyro dynamic model that describes the movement of the anti-roll gyro device and a ship roll dynamic model that describes the ship roll response. Among them, the gyro dynamic model is constructed based on the roll data of the bunkering ship, describing how the anti-roll gyro device uses the rotating gyro to generate a reaction torque, thereby effectively resisting the ship's roll. The reaction torque acts on the hull through the rotation and precession effect of the gyro, thereby controlling the roll amplitude of the ship. The ship's roll dynamic model describes the roll motion of the bunkering ship under the action of external disturbances such as wind and waves. By analyzing the effects of the ship structure and waves, the torque generated by the ship's roll is obtained. These external disturbances work together with the inherent characteristics of the ship (such as mass distribution, stiffness, etc.) to affect the ship's roll response.

[0029] Furthermore, the dynamic model of the gyroscope is combined with the roll dynamic model of the ship through torque coupling. The reaction torque of the gyroscope interacts with the roll torque of the ship to form a dynamic balance, and finally constitutes the roll-gyroscope joint dynamic model of the ship.

[0030] In a specific implementation, the ship roll-gyro combined dynamic model is constructed based on the roll data and the wind and wave data, including: (i) analyzing the roll data and establishing a dynamic model of the gyro in combination with the dynamic motion principle of the gyro; the dynamic model of the gyro describes that the anti-roll gyro device generates a reaction torque for controlling the roll of the bunkering ship based on the roll data by using a rotating gyro.

[0031] Specifically, Figure 2Schematic diagram of the structure of the gyro device shown in this application, Figure 3 This is a working principle diagram of the anti-roll gyro device shown in this application. Figure 2 and Figure 3 , when the bunkering ship is subjected to the disturbance torque of the waves The action occurs around the x-axis with an angular velocity of When the ship is rolling, the anti-roll gyro device connected to the hull will be subjected to the interference torque in the same direction as the roll angular velocity. According to the gyro dynamics theory, when a high-speed rotating gyro is subjected to an external torque in a certain direction, its axis will precess in the direction perpendicular to the external torque and generate a torque in that direction. When the anti-roll gyro device shell generates a precession torque around the y-axis due to the external torque and precession angular velocity , the gyro rotor will output a gyroscopic torque around the x-axis under the influence of the precession torque (external torque) Since the output torque of the anti-roll gyro device is in the opposite direction to the wave disturbance torque, the torque applied to the bunkering ship will produce a restoring effect, thereby reducing the roll effect of the bunkering ship. However, when the wave disturbance torque is large, the anti-roll gyro device will precess at a faster angular velocity and easily enter a singular state, resulting in the inability to output the anti-roll torque. Therefore, a damping device is generally used to apply a damping torque in the precession axis direction. , in order to reduce the precession angular velocity of the anti-roll gyro device, so that it can evenly release the anti-roll torque throughout the entire roll cycle, thereby improving the anti-roll effect.

[0032] In specific implementation, the initial rotation angular velocity of the anti-roll gyro device is set according to the roll response of the bunkering ship and the target sea conditions. When subjected to the wave interference torque, the bunkering ship rolls and generates a roll angular velocity. The anti-roll gyro device outputs a torque in the reverse direction according to the roll angular velocity to suppress the roll. Under the action of the roll angular velocity, the rotor of the anti-roll gyro device begins to precess. According to the principle of gyrodynamics, a reverse torque is generated under the change of the precession angle and the precession angular velocity to reduce the roll. Further, the precession angular velocity is compared with the maximum safe precession angular velocity. Based on the difference between the precession angular velocity and the safety margin, the initial rotation angular velocity or the precession angular velocity is adjusted. If the precession angular velocity is too large, the anti-roll effect may be reduced or a singular state may be entered. When the precession angular velocity exceeds the safety margin, the initial rotation angular velocity or the precession angular velocity is adjusted according to the safety margin to make the gyro enter a stable state and avoid excessive precession.

[0033] In a specific implementation, the rolling data is analyzed, and a dynamic model of the gyro is established in combination with the dynamic motion principle of the gyro, including: establishing an inertial coordinate system with the rotor center of the gyro device as the origin, the longitudinal direction of the bunkering ship as the X-axis, the transverse direction of the bunkering ship as the Y-axis, and the vertical direction as the Z-axis; rotating along the vertical axis direction of the inertial coordinate system to obtain the gyro coordinate system; rotating the rotor of the gyro device along the vertical axis direction of the gyro coordinate system, rotating the frame of the gyro device along the longitudinal axis direction of the gyro coordinate system, performing dynamic analysis on the rotor motion, and calculating the precession angle and precession angular velocity of the gyro device and the roll angular velocity of the bunkering ship in combination with the external torque and the angular momentum change of the gyro device; and determining the stabilizing torque of the gyro device in the inertial coordinate system based on the precession angle and precession angular velocity of the gyro device and the roll angular velocity of the bunkering ship.

[0034] Specifically, Figure 4 This is a schematic diagram of the gyro coordinate system shown in this application. Figure 4 , with the rotor center of the gyro device Establish an inertial coordinate system for the origin ; Along the vertical axis direction of the inertial coordinate system Rotation Angle , get the gyro coordinate system ; The rotor of the anti-roll gyro device is moved along the vertical axis direction of the anti-roll gyro coordinate system At a constant angular velocity Rotate the frame of the gyro device along the longitudinal axis of the gyro coordinate system Rotation. Since the anti-roll gyro device is an axisymmetric rigid structure, the angular momentum is , we can get the anti-roll gyro device along the precession axis The precession angle , precession angular velocity , moment of inertia and the rolling angular velocity of the bunkering vessel ; Determine the position of the anti-roll gyro device in the inertial coordinate system based on the precession angle and precession angular velocity of the anti-roll gyro device and the roll angular velocity of the bunkering ship The stabilizing moment under is: ; Among them, the is the stabilizing moment of the gyro device in three directions in the inertial coordinate system; is the angular momentum; is the precession angle; is the precession angular velocity; is the moment of inertia; is the rolling angular velocity of the bunkering vessel.

[0035] Furthermore, the electromagnetic torque of the torque motor is: ; Among them, the is the electromagnetic torque output by the motor; is the electromagnetic torque coefficient; is the armature current; is the total inductance of the armature circuit; is the armature terminal voltage; is the total resistance of the armature circuit; is the mechanical angular velocity.

[0036] It should be noted that since the motor output shaft is directly connected to the precession shaft of the anti-roll gyro device, By setting the damping coefficient , and control the armature terminal voltage in real time according to the precession angular velocity , the electromagnetic torque can be controlled.

[0037] (ii) analyzing the roll data and wind and wave data, and establishing a roll dynamics model of the ship in combination with the structure of the bunkering ship; the roll dynamics model of the ship describes the roll motion of the bunkering ship under the action of wind and waves.

[0038] Specifically, under the disturbance of random waves, the bunkering ship will produce rolling motion in multiple directions. In order to simplify the calculation, only the force analysis is carried out on the rolling motion of the bunkering ship under the disturbance of waves. The bunkering ship has a center of buoyancy, a center of gravity and a metacenter. Among them, the distance between the center of gravity and the metacenter of the bunkering ship is the initial stability height, which is one of the important parameters for measuring the wave resistance of the bunkering ship. When the hull is stationary, the center of buoyancy, the center of gravity and the metacenter, and the baseline of the bottom of the ship are located on the same vertical line. The buoyancy and gravity are equal in magnitude and opposite in direction, that is, the bunkering ship is in a positive buoyant equilibrium state. According to the ship wave resistance theory, when the bunkering ship rolls under the interference of wave moment, its center of buoyancy will deflect relative to the hull. Since the buoyancy and gravity are in opposite directions, a force arm will be generated on the tilted side of the hull, thereby generating a restoring moment. Secondly, when the bunkering ship tilts to one side, the volume of seawater displaced on both sides will also change. Due to the effects of water viscosity and wave impact, friction damping, wave damping and vortex damping will be generated on the bunkering ship, forming a synthetic damping torque. In addition, due to the mass and load of the bunkering vessel itself, there is an inertia moment in the rolling direction. .

[0039] Furthermore, under the interference of waves, each part of the moment constitutes the resultant moment in the rolling direction of the bunkering ship, that is, the rolling dynamics model of the ship is: ; Among them, the is the moment of inertia of the bunkering vessel itself; is the additional mass moment of inertia; is the rolling angular acceleration of the bunkering ship; is the rolling angular velocity of the bunkering ship; is the rolling angle of the bunkering ship; , is the damping coefficient; , , is the restoring moment coefficient; is the displacement of the bunkering vessel; The initial stability is high; is the effective wave inclination angle.

[0040] (iii) Based on the principle of dynamic balance of torque and dynamic coupling, the reaction torque generated by the dynamic model of the gyroscope is introduced into the roll dynamic model of the ship, and the reaction torque is coupled with the roll torque of the ship to obtain a ship roll-gyroscope combined dynamic model.

[0041] Specifically, in the ship roll-gyro joint dynamics model, the total moment of the bunkering ship is equal to the sum of the roll moment generated by the bunkering ship roll and the reaction moment generated by the anti-roll gyro device. Therefore, the combined ship roll-shedding dynamics model can be expressed as: ; Among them, the is the moment of inertia of the bunkering vessel itself; is the additional mass moment of inertia; is the rolling angular acceleration of the bunkering ship; is the rolling angular velocity of the bunkering ship; is the rolling angle of the bunkering ship; , is the damping coefficient; , , is the restoring moment coefficient; is the displacement of the bunkering vessel; The initial stability is high; is the effective wave inclination angle; is the angular momentum; is the precession angle; is the precession angular velocity; is the moment of inertia; is the electromagnetic torque coefficient; is the armature current; is the total inductance of the armature circuit; is the armature terminal voltage; is the total resistance of the armature circuit; is the mechanical angular velocity; is the damping coefficient; is the voltage gain coefficient.

[0042] (2) Based on the ship roll-gyro combined dynamic model, a state equation is obtained; the state equation at least includes the roll angle.

[0043] Specifically, the state equation includes at least a roll angle and a roll angular velocity.

[0044] In a specific implementation, the state equation is obtained based on the ship roll-gyro combined dynamic model, including: deforming the ship roll-gyro combined dynamic model to obtain a unified mathematical model; converting the unified mathematical model into a lower triangular state equation.

[0045] Specifically, the ship roll-gyro joint dynamics model is transformed and rewritten as follows: ; ; ; ; ; ; ; ; Among them, the is the rolling angle of the bunkering ship; is the rolling angular velocity of the bunkering ship; is the rolling angular acceleration of the bunkering ship; is the effective wave inclination angle; is the moment of inertia of the bunkering vessel itself; is the additional mass moment of inertia; , is the damping coefficient; , , is the restoring moment coefficient; is the displacement of the bunkering vessel; The initial stability is high; is the angular momentum; is the precession angle.

[0046] Further, the above formula is converted into the state equation of the lower triangular form, and , state variables It can be expressed as: .

[0047] Among them, the is a state variable; is the roll angle; Roll angular velocity; is the rolling angle of the bunkering vessel; is the rolling angular velocity of the bunkering ship.

[0048] Setting the control input for: , wherein the For the control input, the is the precession angular velocity.

[0049] Setting external unknown disturbance for: , wherein the is an external unknown disturbance; , is the displacement of the bunkering vessel; For high initial stability, the is the moment of inertia of the bunkering vessel itself; is the additional mass moment of inertia; is the effective wave inclination angle.

[0050] Output for: , wherein the For output; is the roll angle.

[0051] Furthermore, the above formula is sorted out to obtain the state equation: ; ; Among them, the is the roll angle; Roll angular velocity; is the control input; is an external unknown disturbance; For output; For control input gain.

[0052] (3) Based on the state equation, determine the relevant parameters of the command filter.

[0053] Specifically, the relevant parameters of the command filter include error variable, control rate and adaptive rate. Among them, the error variable is the deviation between the current state and the desired target state, which is used to adjust the output of the command filter. The control rate determines the speed and amplitude of the command filter's response to the error variable. The adaptive rate is the speed at which the command filter adjusts the control rate or error variable. It includes three adaptive rates, which are used to determine the weight of the radial basis function neural network, the unknown control gain and its inverse.

[0054] In a specific implementation, the related parameters of the command filter are determined based on the state equation, including: determining the error variable based on the roll angle and the roll angular velocity in the state equation; analyzing the error variable based on the command filter backstepping method to determine the control rate of the command filter; and determining the adaptive law of the command filter based on the control rate and the error variable.

[0055] Specifically, the error variable is based on the roll angle in the state equation and the rolling angular velocity Determined, the error variable can be expressed as: ; Among them, the is the error variable; is the roll angle; is the rolling angular velocity; The status of the command filter.

[0056] Furthermore, the command filter can be defined as: ; Among them, the , is the state of the instruction filter; , is a design parameter and its value is greater than 0. .

[0057] Furthermore, the speed and amplitude of the error variable are analyzed by the command filter backstepping method to determine the control rate: ; ; ; Among them, the , is the control rate; is a weight vector; is the error variable; is the state of the instruction filter; is the design parameter; is a constant; For control gain.

[0058] Furthermore, the adaptation rate is determined based on the error variable and the control rate: ; ; ; ; Among them, the , , is the adaptation rate; is a weight vector; is the error variable to be compensated; is a constant; , is the control rate; For control The gain of is the design parameter; , , For the forgetting factor.

[0059] Optionally, the processing of the roll data and the wind and wave data based on the radial basis function neural network to calculate the precession angular velocity of the anti-roll gyro device under the target sea condition includes: the command filter analyzing the roll data and the wind and wave data, determining a corresponding smoothing method based on dynamic characteristics, and smoothing the reference commands in the roll data and the wind and wave data based on the smoothing method to obtain processed input data; the dynamic characteristics characterize the relationship between input and output; the nonlinear disturbance observer analyzing the roll data and the wind and wave data, estimating external disturbances in real time, and determining disturbance information; and inputting the processed input data and the disturbance information into the radial basis function neural network, and the radial basis function neural network calculating the precession angular velocity of the anti-roll gyro device based on the relationship between the processed input data and the disturbance information.

[0060] Optionally, the instruction filter analyzes the roll data and wind and wave data, determines a corresponding smoothing method based on dynamic characteristics, and smoothes reference instructions in the roll data and wind and wave data based on the smoothing method to obtain processed input data, including: the instruction filter performs spectrum analysis and time response analysis on the roll data and wind and wave data to determine the dynamic characteristics of the roll data and wind and wave data; when the dynamic characteristics are multi-order nonlinear dynamics, the instruction filter gradually processes the first nonlinear dynamics through recursive filtering, and performs stability adjustment on the second nonlinear dynamics through backstepping control in each recursion to obtain processed input data; the order of the first nonlinear dynamics is smaller than that of the second nonlinear dynamics; when the dynamic characteristics are not multi-order nonlinear dynamics, the instruction filter uses a filter to smooth the roll data and wind and wave data to obtain processed input data.

[0061] Specifically, the roll data and wind and wave data of the bunkering ship under the target sea conditions are collected in real time as input data of the radial basis function neural network. The command filter is used to perform spectrum analysis and time response analysis on the roll data and wind and wave data to identify their dynamic characteristics. Based on the identified dynamic characteristics, a suitable smoothing method (such as recursive filtering or backstepping control) is selected to smooth the reference instructions in the input data. The smoothed data is sent to the radial basis function neural network as processed input data. Furthermore, a nonlinear disturbance observer is used to perform real-time analysis on the roll data and wind and wave data, estimate external disturbances (such as wind and wave changes, hull instability, etc.) and determine disturbance information, calculate the amplitude and frequency of the disturbance, and send it as additional information to the radial basis function neural network. After receiving the processed input data and disturbance information, the radial basis function neural network outputs the precession angular velocity of the anti-roll gyro device by learning the relationship between the input data and the disturbance information.

[0062] In specific implementation, the command filter uses the spectrum analysis method to process the roll data and wind and wave data, analyzes its frequency components, and identifies the periodic changes and mutations in the data. At the same time, the data is analyzed for time response, the trend of the data changing over time is checked, and the transient and steady-state characteristics of the data are identified. The dynamic characteristics of the roll data and wind and wave data are determined by combining the results of the spectrum analysis and the time response analysis. Determine whether it is a multi-order nonlinear dynamic based on the dynamic characteristics. If so, continue to perform recursive filtering; otherwise, use simple smoothing. In the case of determining that it is a multi-order nonlinear dynamic, the command filter gradually processes the first nonlinear dynamic (lower order) through recursive filtering. In each round of recursive filtering, the data is smoothed and the input data is updated. In each recursive processing, the second nonlinear dynamic (higher order) is adjusted for stability in combination with the backstepping control method. Backstepping control ensures system stability and avoids instability by compensating for higher-order nonlinear dynamics. After recursive filtering and backstepping control processing, the smoothed input data is obtained. When it is determined to be non-multi-order nonlinear dynamics (ie, linear or low-order nonlinear), the command filter directly uses a conventional filter to smooth the roll data and wind and wave data to obtain processed input data.

[0063] It should be noted that the nonlinear disturbance observer is used to estimate and compensate for the nonlinear disturbance and uncertainty in the radial basis function neural network to ensure the convergence of the radial basis function neural network. After compensating the external disturbance through the nonlinear disturbance observer, the convergence of the roll angle of the bunkering ship is analyzed in order to make the roll angle converge to the preset angle (usually 0).

[0064] In specific implementation, the nonlinear disturbance observer and radial basis function neural network are combined to obtain the difference between the roll angle of the bunkering ship and the preset angle in real time, and the system disturbance is estimated by the nonlinear disturbance observer. The radial basis function neural network dynamically adjusts the output control quantity according to the difference and the system disturbance, and the output control quantity is input into the anti-roll gyro device to adjust the attitude of the bunkering ship to ensure that the roll angle of the bunkering ship tends to 0.

[0065] Optionally, the process of determining the nonlinear disturbance observer includes: determining the nonlinear disturbance observer based on the state equation; and selecting a corresponding form of Lyapunov function based on the convergence requirement of the roll angle.

[0066] Specifically, the nonlinear disturbance observer is determined based on the state equation and the relevant parameters of the command filter: ; Among them, the is the displacement of the bunkering vessel; is the design parameter; is a weight vector; is the control rate; For control The gain of is the rolling angular velocity; is a constant.

[0067] Furthermore, after reorganizing the above formula, we get: ; ; Among them, the is the error variable to be compensated; is a weight vector; is a constant; For control The gain of is the control rate; is the error variable to be compensated; is the design parameter.

[0068] Furthermore, the convergence of the roll angle requires the selection of the following Lyapunov function: ; ; ; Among them, the , is the Lyapunov function; , is the error variable to be compensated; is a weight vector; is a constant; To control the gain.

[0069] Further, yes Taking the derivative and rearranging it, we get: ; ; ; ; Among them, the is the Lyapunov function; , is the design parameter; , is the error variable to be compensated; is a weight vector; is the design parameter; is a constant; is the control gain; , , is the forgetting factor; For the control rate.

[0070] S103, controlling the anti-roll gyro device to rotate at the precession angular velocity, the anti-roll gyro device generates a gyro effect, generates a roll moment, converts the roll moment into a stabilizing moment to counteract the roll through the precession motion, and reduces the roll angle of the bunkering ship; the stabilizing moment is opposite to the vibration direction of the roll.

[0071] Specifically, the direction of the stabilizing moment is opposite to the vibration direction of the rolling of the bunkering ship, and the magnitude of the stabilizing moment is not less than the magnitude of the rolling moment.

[0072] In specific implementation, the gyro effect refers to the fact that when the anti-roll gyro device is subjected to a roll torque during its rotation, it will redistribute the roll torque through precession motion (the anti-roll gyro device slowly rotates around a certain axis under the action of the roll torque), generating a stabilizing torque (usually in the vertical direction) to resist the roll motion and reduce the roll. By controlling the precession angular velocity of the anti-roll gyro device, the magnitude of the generated stabilizing torque can be controlled. When the roll angle increases due to external disturbances, the stabilizing torque generated by the precession of the anti-roll gyro device automatically counteracts the roll. The stabilizing torque is always in the opposite direction to the roll vibration, causing the roll angle to gradually decrease and approach zero.

[0073] The anti-roll control method of a fuel filling ship at zero speed provided in this embodiment, in the first aspect, uses a radial basis function neural network to process roll data and wind and wave data, calculates the precession angular velocity of the anti-roll gyro device, and then controls the anti-roll gyro device based on the precession angular velocity to achieve anti-roll control of the fuel filling ship. In the roll control of the fuel filling ship, the radial basis function neural network can accurately model the complex relationship between the roll and disturbance of the fuel filling ship, and by learning input variables such as roll angle and wind and wave data, a control signal for adjusting the precession angular velocity of the anti-roll gyro device can be generated in real time. Moreover, as the wind and waves change, the radial basis function neural network can optimize the control performance by adaptively adjusting its weights, without the need to accurately linearize the model, thereby improving the accuracy of the anti-roll control of the fuel filling ship, achieving the roll control of the fuel filling ship at low speed or zero speed, and can adjust the anti-roll torque in real time according to the changes in wind and waves, thereby improving the stability and safety of the fuel filling ship during the fuel filling operation. Secondly, in the design process of the radial basis function neural network, the radial basis function neural network is optimized through the command filter and the nonlinear disturbance observer. The command filter avoids the adverse oscillation of the radial basis function neural network by smoothing the reference command. The command filter backstepping control combined with the multi-step recursive design can process multi-order nonlinear dynamics step by step to ensure the stability of the radial basis function neural network under complex sea conditions. The command filter can quickly adjust the precession angular velocity of the anti-roll gyro device to respond to external disturbances, so that the bunkering ship always remains in a stable state. In addition, the disturbance caused by wind and waves is strongly nonlinear and random. The nonlinear disturbance observer can accurately estimate these disturbance values ​​and correct the radial basis function neural network through a compensation mechanism. The disturbance estimation function of the nonlinear disturbance observer reduces the computational burden of the radial basis function neural network and the command filter. Even under the superposition of multiple disturbance sources (such as wind and waves, load changes), the nonlinear disturbance observer can still maintain good observation accuracy and dynamic compensation capabilities. Thirdly, by using anti-roll gyro devices in bunkering ships, the gyroscopic effect can effectively offset the rolling moment caused by wind and waves, especially at low speed or at anchor. The anti-roll gyro device significantly improves the stability of the ship, ensures the accuracy and efficiency of the fuel filling operation, while improving the safety and comfort of the crew, reducing equipment loss and maintenance costs. Compared with traditional anti-roll fins or water tanks, anti-roll gyro devices can adapt to complex sea conditions, have outstanding environmental performance, and have higher flexibility. They are ideal solutions to ensure the safe and efficient operation of bunkering ships. At the same time, it reduces downtime during the fuel filling process, reduces equipment loss, saves operating costs, and the anti-roll gyro device is not only suitable for fuel bunkering ships, but can also be promoted to other ships that need to reduce roll at low speed or stationary state, such as rescue ships and oil recovery ships, and has a wide range of applicability.

[0074] Corresponding to the aforementioned embodiment of a roll stabilization control method for a fuel bunkering ship at zero speed, the present application also provides an embodiment of a roll stabilization control system for a fuel bunkering ship at zero speed.

[0075] Figure 5 This is a schematic diagram of the structure of the anti-roll control system of the fuel filling ship at zero speed provided in Example 2 of this application. Figure 5 The system provided in this embodiment includes a sensor module, a neural network controller and a gyro device connected in sequence; wherein, The sensor module is used to collect the rolling data and wind and wave data of the bunkering ship in real time under the target sea conditions during the fuel bunkering process; the speed of the bunkering ship under the target sea conditions is 0; The neural network controller is used to process the roll data and wind and wave data to calculate the precession angular velocity of the anti-roll gyro device under the target sea condition; The radial basis function neural network characterizes the relationship between the roll and disturbance of the bunkering ship. The radial basis function neural network calculates the precession angular velocity of the anti-roll gyro device by analyzing the roll angle and angular velocity under the target sea conditions. Under different sea conditions, the radial basis function neural network adaptively adjusts the weight based on the sea conditions. The radial basis function neural network includes a command filter and a nonlinear disturbance observer. The command filter smoothes the input data and processes multi-order nonlinear dynamics step by step in combination with multi-step recursive design and backstepping control. The nonlinear disturbance observer determines the nonlinear disturbance based on the input data and dynamically compensates the radial basis function neural network based on the nonlinear disturbance. The neural network controller is also used to control the anti-roll gyro device to rotate at the precession angular velocity, and the anti-roll gyro device generates a gyro effect and a roll moment, and converts the roll moment into a stabilizing moment to counteract the roll through the precession motion, thereby reducing the roll angle of the bunkering ship; the stabilizing moment is opposite to the vibration direction of the roll.

[0076] Optionally, the anti-roll gyro device comprises an outer frame, a high-speed rotor, a rotor rotation axis and a frame axis; the outer frame is located at the outermost layer of the anti-roll gyro device, and is used to support and protect the internal structure, so as to provide a stable external structure for the anti-roll gyro device; The high-speed rotor is located at the center of the anti-roll gyro device and is the core component of the anti-roll gyro device. The high-speed rotor generates angular momentum through rotation to achieve anti-roll function. The rotor rotation shaft is connected to the high-speed rotor and is used to control the rotation of the high-speed rotor; The frame shaft connects the outer frame and the high-speed rotor and is used to control the high-speed rotor to rotate within the outer frame. The frame shaft is used to connect and support the outer frame and limit the relative movement between the high-speed rotor and the outer frame.

[0077] The system of this embodiment can be used to perform Figure 1 The steps, specific implementation principles and implementation processes of the method embodiment shown are similar and will not be repeated here.

[0078] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.

[0079] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application scheme. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0080] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for anti-rolling control of a fuel bunkering ship at zero speed, characterized in that: The method comprises: Real-time collection of rolling data and wind and wave data of the bunkering ship under target sea conditions during the bunkering process; the speed of the bunkering ship under the target sea conditions is 0; The roll data and wind and wave data are processed based on a radial basis function neural network to calculate the precession angular velocity of the anti-roll gyro device under the target sea condition; The radial basis function neural network characterizes the relationship between the roll and disturbance of the bunkering ship. The radial basis function neural network calculates the precession angular velocity of the anti-roll gyro device by analyzing the roll angle and angular velocity under the target sea conditions. Under different sea conditions, the radial basis function neural network adaptively adjusts the weight based on the sea conditions. The radial basis function neural network includes a command filter and a nonlinear disturbance observer. The command filter smoothes the input data and processes multi-order nonlinear dynamics step by step in combination with multi-step recursive design and backstepping control. The nonlinear disturbance observer determines the nonlinear disturbance based on the input data and dynamically compensates the radial basis function neural network based on the nonlinear disturbance. The anti-roll gyro device is controlled to rotate at the precession angular velocity, the anti-roll gyro device generates a gyro effect, generates a roll moment, and converts the roll moment into a stabilizing moment to counteract the roll through the precession motion, thereby reducing the roll angle of the bunkering ship; the stabilizing moment is opposite to the vibration direction of the roll.

2. The method according to claim 1, characterized in that The command filter is designed based on the roll data and wind and wave data, and includes: Based on the roll data and the wind and wave data, a ship roll-gyro joint dynamics model is constructed; the ship roll-gyro joint dynamics model describes the roll behavior of the bunkering ship under the control of external disturbance and anti-roll gyro device; Based on the ship roll-gyro joint dynamics model, a state equation is obtained; the state equation at least includes a roll angle; Based on the state equation, relevant parameters of the command filter are determined.

3. The method according to claim 2, characterized in that The constructing of a ship roll-gyro combined dynamics model based on the roll data and the wind and wave data comprises: The roll data is analyzed, and a dynamic model of the gyro is established in combination with the dynamic motion principle of the gyro; the dynamic model of the gyro describes that the anti-roll gyro device generates a reaction torque for controlling the roll of the bunkering ship based on the roll data by using the rotating gyro; The roll data and wind and wave data are analyzed, and a roll dynamics model of the ship is established in combination with the structure of the bunkering ship; the roll dynamics model of the ship describes the roll motion of the bunkering ship under the action of wind and waves; Based on the principle of dynamic balance of moment and dynamic coupling, the reaction moment generated by the dynamic model of the gyroscope is introduced into the roll dynamic model of the ship, and the reaction moment is coupled with the roll moment of the ship to obtain a ship roll-gyroscope joint dynamic model.

4. The method according to claim 2, characterized in that: The state equation is obtained based on the ship roll-gyro joint dynamic model, including: The ship roll-gyro combined dynamic model is deformed to obtain a unified mathematical model; The mathematical model in unified form is converted into the state equation in lower triangular form.

5. The method according to claim 2, characterized in that: Determining relevant parameters of the instruction filter based on the state equation includes: Determining an error variable based on the roll angle and the roll angular velocity in the state equation; Analyzing the error variable based on the command filter backstepping method to determine the control rate of the command filter; An adaptive law of a command filter is determined based on the control rate and the error variable.

6. The method according to claim 1, characterized in that The processing of the roll data and the wind and wave data based on the radial basis function neural network to calculate the precession angular velocity of the anti-roll gyro device under the target sea condition includes: The command filter analyzes the roll data and the wind and wave data, determines a corresponding smoothing method based on dynamic characteristics, and performs smoothing processing on the reference commands in the roll data and the wind and wave data based on the smoothing method to obtain processed input data; the dynamic characteristics represent the relationship between input and output; The nonlinear disturbance observer analyzes the roll data and wind and wave data, estimates the external disturbance in real time, and determines the disturbance information; The processed input data and the disturbance information are input into the radial basis function neural network, and the radial basis function neural network calculates the precession angular velocity of the anti-roll gyro device based on the relationship between the processed input data and the disturbance information.

7. The method according to claim 6, characterized in that The command filter analyzes the roll data and the wind and wave data, determines a corresponding smoothing method based on dynamic characteristics, and performs smoothing processing on reference commands in the roll data and the wind and wave data based on the smoothing method to obtain processed input data, including: The command filter performs spectrum analysis and time response analysis on the roll data and wind and wave data to determine the dynamic characteristics of the roll data and wind and wave data; When the dynamic characteristic is a multi-order nonlinear dynamic, the instruction filter gradually processes the first nonlinear dynamic by recursive filtering, and performs stability adjustment on the second nonlinear dynamic by backstepping control in each recursion to obtain processed input data; the order of the first nonlinear dynamic is smaller than that of the second nonlinear dynamic; When the dynamic characteristic is not a multi-order nonlinear dynamic characteristic, the command filter uses a filter to smooth the roll data and the wind and wave data to obtain processed input data.

8. The method according to claim 3, characterized in that The step of analyzing the roll data and establishing a dynamic model of the gyroscope in combination with the dynamic motion principle of the gyroscope includes: An inertial coordinate system is established with the rotor center of the anti-roll gyro device as the origin, the longitudinal direction of the bunkering ship as the X-axis, the transverse direction of the bunkering ship as the Y-axis, and the vertical direction as the Z-axis; Rotating along the vertical axis direction of the inertial coordinate system to obtain a gyroscopic coordinate system; The rotor of the gyro device is rotated along the vertical axis direction of the gyro coordinate system, and the frame of the gyro device is rotated along the longitudinal axis direction of the gyro coordinate system, and the rotor motion is dynamically analyzed, and the precession angle and precession angular velocity of the gyro device and the roll angular velocity of the bunkering ship are calculated by combining the external torque and the angular momentum change of the gyro device; The stabilizing moment of the gyro device in the inertial coordinate system is determined based on the precession angle and the precession angular velocity of the gyro device and the roll angular velocity of the bunkering ship.

9. The method according to claim 1, characterized in that: The real-time collection of rolling data and wind and wave data of the bunkering ship under target sea conditions during the bunkering process includes: Determining a first installation position of the sensor based on propagation characteristics of the rolling motion and the structure of the bunkering vessel; The first installation position is adjusted based on the center of gravity position of the bunkering ship and the distribution characteristics of the sensor to obtain the installation position of the sensor; A sensor is installed based on the installation position, and the rolling data and wind and wave data of the bunkering ship under the target sea conditions during the fuel bunkering process are collected in real time based on the sensor.

10. A roll reduction control system for a fuel bunkering ship at zero speed, characterized in that: The system comprises a sensor module, a neural network controller and a gyro device connected in sequence; wherein, The sensor module is used to collect the rolling data and wind and wave data of the bunkering ship in real time under the target sea conditions during the fuel bunkering process; the speed of the bunkering ship under the target sea conditions is 0; The neural network controller is used to process the roll data and wind and wave data to calculate the precession angular velocity of the anti-roll gyro device under the target sea condition; The neural network controller characterizes the relationship between the roll and disturbance of the bunkering ship. The neural network controller calculates the precession angular velocity of the anti-roll gyro device by analyzing the roll angle and angular velocity under the target sea conditions. Under different sea conditions, the neural network controller adaptively adjusts the weight based on the sea conditions. The neural network controller includes a command filter and a nonlinear disturbance observer. The command filter smoothes the input data and processes multi-order nonlinear dynamics step by step in combination with multi-step recursive design and backstepping control. The nonlinear disturbance observer determines the nonlinear disturbance based on the input data and dynamically compensates the neural network controller based on the nonlinear disturbance. The neural network controller is also used to control the anti-roll gyro device to rotate at the precession angular velocity, and the anti-roll gyro device generates a gyro effect and a roll moment, and converts the roll moment into a stabilizing moment to counteract the roll through the precession motion, thereby reducing the roll angle of the bunkering ship; the stabilizing moment is opposite to the vibration direction of the roll.

Images