Co-simulation method for underwater vehicle flow field and time-varying motion model predictive control

Abstract

The application belongs to the technical field of underwater vehicle motion control, and specifically discloses a kind of underwater vehicle flow field and time-varying motion model predictive control cooperative simulation method, comprising: determining the initial parameters of AUV dynamics model;Based on the initial parameters of AUV dynamics model and AUV navigation state information, the pose of AUV is continuously iterated until the pose deviation is within the preset range;Iteration includes: on-line rolling identification of time-varying parameters of AUV dynamics model, determining the optimal parameters and passing to MPC algorithm to update model parameters;Through ALOS guidance algorithm and MPC algorithm, the rudder angle required at the current time is determined;Through N-S equation, the flow field force and flow field moment are determined, and through CFD multi-degree-of-freedom motion equation, the current pose of AUV is updated;Judge whether the pose deviation is within the preset range or not.Through the application, the accuracy and confidence of motion control simulation can be improved in the dynamically changing marine environment.

Description

Technical Field

[0001] This application belongs to the field of underwater vehicle motion control technology, and more specifically, relates to a co-simulation method for predictive control of underwater vehicle flow field and time-varying motion model. Background Technology

[0002] With the deepening of ocean exploration and the intensification of resource development, the demand for underwater unmanned platforms in ocean detection is growing. Autonomous underwater vehicles (AUVs), as intelligent and unmanned ocean detection equipment, have demonstrated significant advantages in performing complex tasks and improving detection efficiency. However, the motion control accuracy and simulation accuracy of AUVs directly depend on the accuracy of their dynamic models and control algorithms.

[0003] In recent years, Model Predictive Control (MPC) has been increasingly applied to path tracking in underwater robots. By predicting future trajectories, it seeks optimal solutions under multiple constraints, balancing real-time performance with multivariable control, demonstrating strong competitiveness. However, MPC has numerous parameters, and improper settings can lead to performance degradation or even instability. Parameters are typically tuned through simulation, but due to the complexity of the marine environment, its description of complex hydrodynamic phenomena (such as wake eddy currents, turbulence, friction, and boundary layer effects) remains insufficient, resulting in often unsatisfactory simulation results in practice. Furthermore, traditional dynamic models are complex and time-consuming to model. The lack of effective guidance during tuning in real-world marine environments leads to high costs.

[0004] Furthermore, MPC (Multi-Purpose Control) is highly dependent on the accuracy of the dynamic model. Due to variations in the characteristics of the unmanned surface vessel (USV) and the influence of the complex marine environment, its model parameters are uncertain. Even if the initial modeling is relatively accurate, using fixed model parameters as the prediction model for the USV recovery process and even its entire life cycle will still lead to a decrease in prediction accuracy and a weakening of control effectiveness. This uncertainty poses a challenge to the stability and accuracy of AUVs, limiting their application in complex missions.

[0005] Improving the accuracy and simulation precision of AUV dynamic models in dynamically changing marine environments is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] In view of the shortcomings of the existing technology, the purpose of this application is to improve the accuracy of AUV dynamics models and simulation precision in dynamically changing marine environments.

[0007] To achieve the above objectives, in a first aspect, this application provides a method for co-simulation of flow field and time-varying motion model predictive control of underwater vehicles, the method comprising:

[0008] Determine the initial parameters of the dynamic model of the autonomous underwater vehicle (AUV);

[0009] Based on the initial parameters of the AUV dynamics model and the AUV's flight status information, the AUV's pose is continuously iterated until the pose deviation is within a preset range. The AUV's flight status information is obtained through CFD simulation.

[0010] Iterate the pose of the AUV, including:

[0011] Based on the initial parameters of the AUV dynamic model and the AUV's navigation status information, the time-varying parameters of the AUV dynamic model are identified online using an optimization algorithm to determine the optimal parameters of the AUV dynamic model.

[0012] Based on the optimal parameters of the AUV dynamic model and the pose deviation between the current AUV pose and the target pose, the required rudder angle at the current moment is determined by the adaptive line-of-sight ALOS guidance algorithm and the model predictive control (MPC) algorithm. The optimal parameters of the AUV dynamic model are passed to the MPC algorithm to update the model parameters of the MPC algorithm.

[0013] Based on the required rudder angle at the current moment, the flow force and flow torque are determined through the NS (Navier-Stokes) equations, and the flow force and flow torque are transferred to the CFD multi-degree-of-freedom motion equations to update the current pose of the AUV (to obtain the updated AUV pose).

[0014] Based on the target pose and the updated AUV pose, determine whether the pose deviation is within the preset range.

[0015] Understandably, by integrating CFD simulation, the OPRI algorithm, and MPC control, the accuracy of AUV motion control simulation is significantly improved, enabling precise simulation of its motion response in complex environments. This method achieves deep coupling between the control algorithm and fluid dynamics, not only verifying the effectiveness of different control strategies but also supporting parameter tuning and control method optimization in complex environments, providing crucial technical support for practical applications. Furthermore, this application allows for in-depth analysis of complex flow field characteristics during motion, including wave loads, eddies, and nonlinear flow phenomena, providing a reliable guarantee for the efficient and stable navigation of AUVs in complex marine environments.

[0016] In one possible implementation, the above-mentioned online rolling identification of time-varying parameters of the AUV dynamics model through optimization algorithms includes:

[0017] Functions that solve optimization problems using the Particle Swarm Optimization (PSO) algorithm.

[0018] Specifically, the OPRI algorithm dynamically adjusts hydrodynamic coefficients and model parameters by collecting AUV velocity and force data and combining them with dynamic equations to adapt to changes in complex environments. More specifically, the algorithm employs particle swarm optimization (PSO) to perform online rolling identification of time-varying parameters of the AUV dynamic model, enabling dynamic updates and real-time optimization of model parameters.

[0019] In actual operation of AUVs, due to the influence of complex nonlinear hydrodynamics, their dynamic model parameters often exhibit strong time-varying and nonlinear characteristics. Therefore, accurate identification of these parameters is crucial for improving control accuracy and motion performance. The PSO algorithm, by simulating the cooperative behavior of flocks of birds or schools of fish during foraging, leverages its advantages of strong global search capability, simple implementation, and fast convergence speed to solve the nonlinear optimization problem in AUV dynamic model parameter identification.

[0020] The original horizontal dynamic model of the AUV (i.e., the AUV dynamic model mentioned above) is as follows:

[0021] ;

[0022] ;

[0023] ;

[0024] In the formula, m represents the mass of the AUV, and v is the lateral velocity. Here, 'r' is the lateral acceleration, 'u' is the forward velocity, and 'r' is the yaw rate. It is yaw acceleration. , , , , These are hydrodynamic coefficients related to lateral velocity, yaw rate, lateral acceleration, yaw acceleration, and rudder angle. It's the rudder angle. This represents the moment of inertia of the AUV about the z-axis. , , , , It is a torque coefficient related to lateral velocity, yaw rate, lateral acceleration, yaw acceleration, and rudder angle. Finally, ψ is the yaw angle. It is the rate of change of the yaw angle, i.e., the yaw rate.

[0025] The change takes the following form:

[0026] ;

[0027] ;

[0028] ;

[0029] ;

[0030] In the formula, This represents the force acting on the hull along the y-axis. This represents the torque acting on the hull along the z-axis. This represents the force acting on the rudder in the y-axis direction; This represents the torque acting on the rudder in the z-axis direction.

[0031] In the model, These parameters may change with time and operating conditions, therefore, it is necessary to identify these time-varying parameters in real time to improve the accuracy of the model. The parameter identification problem of the AUV dynamic model can be reduced to a nonlinear optimization problem, the goal of which is to continuously adjust the parameters. , so that the model predicts the value Compared with actual measured value The error is minimized. The mathematical expression of the optimization problem is as follows:

[0032] The function for the optimization problem is as follows:

[0033] ;

[0034] in, It is the parameter vector of the AUV dynamic model. Indicates a fixed parameter. The parameters to be optimized are represented by . The initial solution of the particle swarm optimization algorithm uses the initial parameters of the AUV dynamics model. It is the objective function. This represents the measurement value obtained through CFD simulation. This represents the model prediction value obtained through the AUV dynamics model. , express The first in item, , express The first in item, The value is 4. This represents the force acting on the hull along the y-axis (for the three axes in a coordinate system, the x-axis is the horizontal axis, the y-axis is the vertical axis, and the z-axis is the vertical axis). This represents the torque acting on the hull along the z-axis (vertical axis). This represents the force acting on the rudder in the y-axis direction; This represents the torque acting on the rudder in the z-axis direction. It is the input vector of the AUV dynamics model. This represents the lateral velocity obtained through CFD simulation. This represents the yaw rate obtained through CFD simulation. This represents the lateral acceleration obtained through CFD simulation. This represents the yaw acceleration obtained through CFD simulation. This represents the rudder angle obtained through CFD simulation. This represents the error weighting matrix for forces and moments under each degree of freedom.

[0035] In one possible implementation, the above-mentioned adaptive line-of-sight (ALOS) guidance algorithm and model predictive control (MPC) algorithm are used to determine the required rudder angle at the current moment, including:

[0036] Based on the pose deviation between the current pose of the AUV and the target pose, the desired heading angle is obtained through the ALOS guidance algorithm;

[0037] The optimal parameters of the AUV dynamics model are passed to the MPC algorithm to update the model parameters of the MPC algorithm. Based on the desired heading angle, AUV speed information and AUV current pose, the required rudder angle at the current moment is obtained through the MPC algorithm after updating the model parameters.

[0038] Specifically, for the ALOS (Adaptive Line of Sight) guidance algorithm, the ALOS algorithm predicts the motion state at multiple future moments and optimizes the output rudder angle by combining pose deviation with the AUV's dynamic model and constraints. The adaptive LOS guidance algorithm transforms the path tracking problem into a heading control problem, improving convergence speed and tracking stability by dynamically adjusting the line-of-sight distance Δ. When the lateral offset is large, Δ is set to a smaller value to improve convergence speed; when approaching the target path, Δ is set to a larger value to prevent overshoot. Furthermore, the accept circle radius is adaptively adjusted based on the difference θ between the path angles of adjacent paths to ensure attitude adjustment during path switching.

[0039] Model Predictive Control (MPC) is widely used in the motion control of underwater vehicles (AUVs) because it can effectively handle system constraints. MPC predicts the system state over a future period using a mathematical model of the system at each sampling time, and optimizes the control input accordingly. Its core components include a predictive model, a rolling optimization strategy, and a feedback correction scheme. For AUV path tracking, MPC constructs an optimization objective function to minimize the tracking error and the change in input rudder angle, while considering the constraints of rudder angle and rudder speed. It solves a quadratic programming problem using an efficient set algorithm to obtain the optimal rudder angle increment, gradually achieving rolling optimization control for AUV path tracking.

[0040] In one possible implementation, the aforementioned flow field forces and moments are determined by the following formulas:

[0041] ;

[0042] ;

[0043] In the formula, For flow field force, For AUV surfaces, The local pressure on the surface of the AUV, Let be the normal vector of the AUV surface. The shear stress on the surface of the AUV. For the flow field torque, Let be the distance vector from the centroid of the AUV to the center of the surface.

[0044] Solving the Navier-Stokes equations includes simulating wave loads, eddies, and nonlinear flow phenomena.

[0045] Based on the required rudder angle at the current moment, the result can be obtained by solving the Navier-Stokes equations. and .

[0046] In one possible implementation, the local pressure on the surface of the aforementioned AUV It is determined by the following formula:

[0047] ;

[0048] in, These are the micro-elements (tiny facets) on the surface of the AUV. It is the sum of the static and dynamic pressures of the fluid, obtained based on the following formula;

[0049] ;

[0050] ;

[0051] ;

[0052] ;

[0053] In the formula, , and Let x, y, and y represent the x, y, and y axes of the Cartesian coordinate system, respectively. For dynamic viscosity, For fluid density, For the velocity vector field of the fluid, , and These represent the velocity components of the fluid in the horizontal, vertical, and longitudinal directions, respectively. , and These are the mass force components of the fluid in the horizontal, vertical, and longitudinal directions, respectively.

[0054] In one possible implementation, the above-mentioned transfer of flow field forces and moments to the CFD multi-degree-of-freedom motion equations, updating the current pose of the AUV, includes:

[0055] Based on the flow field force and flow field torque, the angular velocity and the velocity of the AUV's center of mass are obtained through CFD multi-degree-of-freedom motion equations;

[0056] Update the current pose of the AUV based on its angular velocity and the velocity of its center of mass.

[0057] The CFD multi-degree-of-freedom motion equations are as follows:

[0058] ;

[0059] ;

[0060] In the formula, This is the sum of the mass of the AUV and the added mass of the AUV. The velocity of the AUV's center of mass. For flow field force, External forces generated by multibody constraints or joints To control force, For the previous moment, For the current moment, Let the moment of inertia tensor be... The angular velocity of the AUV, For the flow field torque, External torque generated by multibody constraints or joints To control the torque.

[0061] In one possible implementation, the AUV's current pose is updated based on its angular velocity and the velocity of its center of mass. Specifically, this involves updating the AUV's current pose using the following formula:

[0062] ;

[0063] ;

[0064] ;

[0065] ;

[0066] ;

[0067] ;

[0068] in, , and Let be the position components of the AUV pose along the horizontal, vertical, and longitudinal axes, respectively. For the next moment, For the current moment, , and Let be the velocity components of the AUV's center of mass in the horizontal, vertical, and longitudinal directions, respectively. , and Let be the rotation angle components of the AUV pose in the horizontal, vertical, and longitudinal directions, respectively. , and Let be the angular velocity components of the AUV in the horizontal, vertical, and longitudinal directions, respectively. The interval between the next moment and the current moment;

[0069] The updated AUV pose includes , , , , and .

[0070] Secondly, this application provides a co-simulation device for predictive control of flow field and time-varying motion model of underwater vehicle, comprising:

[0071] The initial parameter determination module is used to determine the initial parameters of the autonomous underwater vehicle (AUV) dynamic model.

[0072] The iteration module is used to continuously iterate the AUV's pose based on the initial parameters of the AUV dynamic model and the AUV's flight status information until the pose deviation is within a preset range. The AUV's flight status information is obtained through CFD simulation.

[0073] The iterative module includes: a model optimal parameter determination unit, a rudder angle determination unit, an AUV pose update unit, and a pose deviation judgment unit, wherein:

[0074] The optimal parameter determination unit is used to determine the optimal parameters of the AUV dynamic model by performing online rolling identification of the time-varying parameters of the AUV dynamic model based on the initial parameters of the AUV dynamic model and the AUV navigation state information through optimization algorithms.

[0075] The rudder angle determination unit is used to determine the required rudder angle at the current moment based on the optimal parameters of the AUV dynamic model and the pose deviation between the current pose of the AUV and the target pose, through the adaptive line-of-sight ALOS guidance algorithm and the model predictive control (MPC) algorithm. The optimal parameters of the AUV dynamic model are passed to the MPC algorithm to update the model parameters of the MPC algorithm.

[0076] The AUV pose update unit is used to determine the flow field force and flow field torque based on the required rudder angle at the current moment through the NS equation, and to transmit the flow field force and flow field torque to the CFD multi-degree-of-freedom motion equation to update the current pose of the AUV.

[0077] The pose deviation judgment unit is used to determine whether the pose deviation is within a preset range based on the target pose and the updated AUV pose.

[0078] Thirdly, this application provides an electronic device, comprising: at least one memory for storing a program; and at least one processor for executing the program stored in the memory, wherein when the program stored in the memory is executed, the processor is configured to execute the method described in the first aspect or any possible implementation thereof.

[0079] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when run on a processor, causes the processor to perform the method described in the first aspect or any possible implementation thereof.

[0080] It is understood that the beneficial effects of the second to fourth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here.

[0081] Overall, the technical solutions conceived in this application have the following beneficial effects compared with the prior art:

[0082] The co-simulation method for predictive control of underwater vehicle flow field and time-varying motion model provided in this application integrates CFD, OPRI, and MPC. CFD simulation can accurately simulate complex fluid dynamic phenomena, providing high-precision dynamic model support for MPC. Through OPRI technology, the system can dynamically identify and update model parameters, ensuring that the control strategy remains optimal in complex flow field environments. The deep coupling of CFD simulation and dynamic control algorithm (OPRI-MPC) enables real-time adjustment and optimization of AUV dynamic parameters, improving the accuracy of AUV dynamic models and simulation precision in dynamically changing marine environments.

[0083] This simulation method can not only simulate various complex underwater conditions in a laboratory environment, but also adjust and tune parameters before deployment, significantly reducing experimental time. By exploring the control boundary performance under extreme conditions, researchers can optimize control parameters in advance, improving the efficiency of actual deployment experiments. This not only enhances the mission execution capabilities of AUVs, but also significantly reduces experimental costs and improves their adaptability and reliability in complex marine environments.

[0084] In addition, the platform helps analyze and evaluate the performance limits of AUVs under extreme environmental conditions, ensuring that the vehicle can maintain optimal operational safety and reliability when facing challenging missions. Attached Figure Description

[0085] Figure 1 This is a flowchart illustrating the co-simulation method for predictive control of flow field and time-varying motion model of underwater vehicle provided in the embodiments of this application;

[0086] Figure 2 This is a schematic diagram of the LOS guidance algorithm provided in the embodiments of this application;

[0087] Figure 3 This is a schematic diagram illustrating the basic principle of MPC provided in the embodiments of this application;

[0088] Figure 4 This is a schematic diagram of the multi-waypoint test trajectory curve provided in the embodiments of this application;

[0089] Figure 5 This is a schematic diagram of the multi-waypoint test offset curve provided in the embodiments of this application;

[0090] Figure 6 This is a schematic diagram of the multi-waypoint test heading curve provided in the embodiments of this application;

[0091] Figure 7 This is a schematic diagram of the rudder curve for multi-waypoint testing provided in an embodiment of this application;

[0092] Figure 8 This is a schematic diagram of typical time position distribution under AUV path tracking provided in the embodiments of this application;

[0093] Figure 9 This is a spatial distribution diagram of AUV surface pressure and nearby flow velocity at four time points (a, b, c, d) provided in the embodiments of this application;

[0094] Figure 10 This is a spatial distribution diagram of AUV surface pressure and nearby flow velocity at four time points (e, f, g, h) provided in the embodiments of this application;

[0095] Figure 11This is a schematic diagram of the geodetic coordinate system and the carrier coordinate system provided in the embodiments of this application;

[0096] Figure 12 This is a schematic diagram of the structure of the underwater vehicle flow field and time-varying motion model prediction and control co-simulation device provided in the embodiments of this application;

[0097] Figure 13 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0098] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0099] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0100] In the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more, for example, multiple processing units means two or more processing units, multiple elements means two or more elements, etc.

[0101] The embodiments of this application are described below with reference to the accompanying drawings.

[0102] Figure 1 This is a flowchart illustrating the co-simulation method for predictive control of flow field and time-varying motion model of underwater vehicle provided in this application embodiment, as shown below. Figure 1 As shown, the method includes the following steps S101 and S102.

[0103] Step S101: Determine the initial parameters of the autonomous underwater vehicle (AUV) dynamic model;

[0104] Step S102: Based on the initial parameters of the AUV dynamic model and the AUV navigation status information (AUV navigation status information includes: AUV velocity information, AUV force information, AUV position information, and AUV attitude information), the AUV pose is continuously iterated until the pose deviation is within a preset range. The AUV navigation status information is obtained through CFD simulation.

[0105] The pose of the iterative AUV in step S102 specifically includes:

[0106] Based on the initial parameters of the AUV dynamic model and the AUV's navigation status information, the time-varying parameters of the AUV dynamic model are identified online using an optimization algorithm to determine the optimal parameters of the AUV dynamic model.

[0107] Based on the optimal parameters of the AUV dynamic model and the pose deviation between the current AUV pose and the target pose, the required rudder angle at the current moment is determined by the adaptive line-of-sight ALOS guidance algorithm and the model predictive control (MPC) algorithm. The optimal parameters of the AUV dynamic model are passed to the MPC algorithm to update the model parameters of the MPC algorithm.

[0108] Based on the required rudder angle at the current moment, the flow field force and flow field torque are determined through the NS equation, and the flow field force and flow field torque are transmitted to the CFD multi-degree-of-freedom motion equation to update the current pose of the AUV.

[0109] Based on the target pose and the updated AUV pose, determine whether the pose deviation is within the preset range.

[0110] Understandably, the co-simulation method for predictive control of underwater vehicle flow field and time-varying motion model provided in this application can integrate CFD, OPRI, and MPC. CFD simulation can accurately simulate complex fluid dynamic phenomena, providing high-precision dynamic model support for MPC. Through OPRI technology, the system can dynamically identify and update model parameters, ensuring that the control strategy remains optimal in complex flow field environments. The deep coupling of CFD simulation and dynamic control algorithm (OPRI-MPC) enables real-time adjustment and optimization of AUV dynamic parameters, improving the accuracy of AUV dynamic models and simulation precision in dynamically changing marine environments.

[0111] This method achieves deep coupling of guidance, model predictive control, and CFD algorithms through a multidisciplinary collaborative simulation framework. It can accurately simulate the motion response of AUVs in real-world environments, verify the effectiveness of different model predictive control parameters and strategies, and deeply analyze the complex flow field characteristics during motion, including wave loads, eddies, and nonlinear flow phenomena. This application significantly improves the accuracy and confidence of motion control simulation, providing crucial technical support for AUV navigation control optimization, parameter tuning, control algorithm verification, and control boundary performance exploration in complex environments.

[0112] The following examples illustrate the co-simulation method for predictive control of underwater vehicle flow field and time-varying motion model provided in this application.

[0113] The research object of this application is a small autonomous underwater vehicle (AUV) developed in the laboratory. Similar to conventional propeller-driven AUVs, the main propeller at the stern is the primary power source, and the stern rudder and stern elevator form a cruciform rudder as the AUV's control mechanism. All three are the main control inputs of the system to achieve three-dimensional motion control of the AUV. Therefore, the research object has underactuated characteristics and cannot achieve horizontal lateral movement. Due to the strong coupling of the AUV's motion, the difficulty of motion control is increased.

[0114] The AUV has a symmetrical structure, basically symmetrical in all directions. To ensure the vehicle's hydrodynamic performance, reduce power consumption during navigation, and improve endurance, the AUV's main body adopts the most common Myring type shape. The AUV hull is divided into three parts: the bow section, the midsection, and the stern section. The bow section is 0.11 m long, the midsection is cylindrical and 0.74 m long, and the stern section, including the propeller, is 0.22 m long. The total hull length is 1.07 m, and the hull diameter is 0.14 m. The AUV's dry weight is 8.25 kg. The main propulsion engine is a waterproof propeller with a maximum thrust of 70 N. The AUV's maximum speed can reach 2 m / s, and the designed cruising speed is 1 m / s. At cruising speed, the endurance is up to 4 hours.

[0115] Both the rudder and elevator adopt the standard NACA0021 airfoil. The span of a single rudder blade is 0.08 m, the chord length is 0.05 m, the aspect ratio is 1.6, the maximum thickness is 0.01 m, the center distance of the rudder blade from the AUV's central axis is 0.1 m, and the center distance of the rudder blade from the AUV's fore-end is 0.89 m. The relevant technical parameters of the AUV are shown in Figure 1.

[0116] Table 1 Basic Parameters of AUV

[0117]

[0118] The motion of underwater robots (AUVs) can be considered as rigid body motion in six degrees of freedom. To facilitate the study of the relevant laws of AUV motion and to better describe the motion and force states of AUVs, the definitions of the coordinate system, related terminology, and symbol rules must take into account the conventions of fluid mechanics and rigid body kinematics, as well as the simplicity of description. This paper adopts the terminology system recommended by the International Tank Conference (ITTC) and the Society of Naval Architects and Marine Engineers (SNAME), establishing two right-handed Cartesian coordinate systems. For example... Figure 11 As shown, one type is the geodetic coordinate system. One is used to describe the motion trajectory and attitude of an AUV; the other is the carrier coordinate system. , is used to describe the external forces that an AUV experiences during motion.

[0119] When an AUV navigates underwater, it involves motion in six degrees of freedom. The motion state of each degree of freedom, including displacement, velocity, and force, needs to be represented by mathematical symbols. The physical meaning and symbol definition of the motion state of each degree of freedom of an AUV are shown in Table 2.

[0120] Table 2. Definition of AUV Coordinate System Symbols

[0121]

[0122] When using the MPC algorithm to solve the path tracking problem, the model is usually transformed into a state-space form for computational convenience. To simplify the model, it is assumed that the effects of heave, roll, and pitch motions can be ignored when the AUV moves on the horizontal plane. Since the path tracking problem does not impose time requirements, the propeller speed is taken as a constant value, the forward velocity remains basically constant, and the cruising speed is approximately 0.9 m / s.

[0123] The horizontal motion model of the AUV is shown in formulas (1) to (6).

[0124] Formula 1 is as follows:

[0125] ;

[0126] ;

[0127] ;

[0128] In the formula, Let Z be the moment of inertia of the AUV. Parameters for the AUV rudder angle dynamic model: ; ; ; ; -7.840; ; ; ; ; .

[0129] Take the lateral velocity angular velocity of the bow and bow angle For state variables, i.e. The system's matrix form equation of motion is as follows (this reflects the relationship between the parameters of the MPC and AUV dynamics model). Formula (2) is as follows:

[0130] ;

[0131] The AUV horizontal plane motion model is written in standard state space form, as shown in formula (3):

[0132] ;

[0133] ;

[0134] Formula (4) is as follows:

[0135] ;

[0136] Formula (5) is as follows:

[0137] ;

[0138] Formula (6) is as follows:

[0139] .

[0140] Solving optimization problems with multiple constraints online is difficult to achieve using logic circuits to calculate the optimal solution; it generally requires the aid of a digital computer. To facilitate computer processing, the Euler method is used to discretize the continuous state-space equations of the AUV (Automatic Virtual Unit). The discretized system matrix is ​​as follows:

[0141] ;

[0142] ;

[0143] ;

[0144] in, For discrete time steps, It is an identity matrix.

[0145] Discretized system state-space equations:

[0146] ;

[0147] ;

[0148] Considering the errors in the modeling process and the noise interference in the control process, to eliminate the error in steady state, an augmented form of the state-space equation is adopted in model predictive control based on the state-space equation. Using the rate of change of denoted as the control input as the control input quantity, the system can achieve zero steady-state error control. Then, by performing difference operations on both sides of the state-space equation, we obtain:

[0149] ;

[0150] ;

[0151] The following symbols are introduced:

[0152] ;

[0153] ;

[0154] The augmented state-space equations can be obtained as follows:

[0155] ;

[0156] ;

[0157] Right now:

[0158] ;

[0159] ;

[0160] The numerical simulation for computational fluid dynamics (CFD) was constructed with a background domain measuring 80 meters in length, 80 meters in width, and 12 meters in height, including an 8-meter underwater portion. The downstream boundary was defined as the pressure outlet boundary, while all other boundaries were defined as velocity inlet boundaries. The densified region within the background domain, determined by the AUV's target path, was covered and densified with a cylindrical region of 1.2 meters in diameter.

[0161] Based on the overlapping mesh theory, the fluid region is divided into a background region, overlapping region 1 (hull), overlapping region 2 (upper rudder), overlapping region 3 (lower rudder), overlapping region 4 (left rudder), and overlapping region 5 (right rudder). The overlapping region mesh is 1.5 meters long, 0.5 meters wide, and 0.5 meters high. The simulation uses a hexagonal cut-body mesh as the primary mesh type, and a prism layer mesh is used to model the boundary layer of the hull surface. The mesh near the wavefront is refined along the wave height direction to ensure the accuracy of wave generation and propagation. Furthermore, mesh refinement is performed at the boundaries between the background mesh and each overlapping region to ensure the stability and accuracy of the dynamic overlapping mesh algorithm; mesh refinement is also performed in the stern rudder region of the AUV to ensure the accuracy of the local flow field and facilitate the calculation of the disturbance effect after the rudder surface rotates.

[0162] Ignoring the compressibility and temperature effects of seawater, the Reynolds-averaged Navier-Stokes (RANS) method and the Volume of Fluid (VOF) wave model are adopted. The governing equations include the mass conservation equation and the Navier-Stokes (NS) equation. The SST k-ω model is selected as the turbulence model to accurately predict the fluid motion states in the near-wall region and the far field. The NS governing equation is as follows:

[0163] ;

[0164] ;

[0165] ;

[0166] ;

[0167] In the formula, x, y, and z represent the Cartesian coordinate system, and p is the pressure. For dynamic viscosity, For fluid density, , , X, Y, and Z are the velocity components along the x, y, and z axes, respectively, while X, Y, and Z are the mass force components along the x, y, and z axes, respectively.

[0168] The Online Parameter Rolling Identification (OPRI) algorithm is constructed, as the dynamic model of an AUV is a crucial foundation for describing its motion characteristics and force behavior. Due to the complex nonlinear hydrodynamic effects experienced by AUVs in actual operation, their dynamic model parameters often exhibit strong time-varying and nonlinear characteristics. Therefore, accurately identifying the parameters of the AUV dynamic model is of great significance for improving the control accuracy and motion performance of the AUV. This section employs the Particle Swarm Optimization (PSO) algorithm to perform online rolling identification of the time-varying parameters of the AUV dynamic model, enabling dynamic updates and real-time optimization of the model parameters. The PSO algorithm, by simulating swarm intelligence behavior, possesses advantages such as strong global search capability, simple implementation, and fast convergence speed, making it highly suitable for solving nonlinear optimization problems in AUV dynamic model parameter identification.

[0169] From the following formula, we can see that:

[0170] ;

[0171] ;

[0172] ;

[0173] ;

[0174] In the formula, This represents the force acting on the hull along the y-axis. This represents the torque acting on the hull along the z-axis. This represents the force acting on the rudder in the y-axis direction; This represents the torque acting on the rudder in the z-axis direction.

[0175] In the model, These parameters may change with time and operating conditions, therefore, it is necessary to identify these time-varying parameters in real time to improve the accuracy of the model. The parameter identification problem of the AUV dynamic model can be reduced to a nonlinear optimization problem, the goal of which is to continuously adjust the parameters. , so that the model predicts the value Compared with actual measured value The error is minimized. The mathematical expression of the optimization problem is as follows:

[0176] ;

[0177] in, It is the vector of model parameters to be identified. This indicates a fixed parameter that does not participate in optimization; its value is 0.5 times the initial parameter. This represents the parameter to be optimized, whose initial value is randomly generated within the range of plus or minus 0.3; It is the objective function, representing the error between the model's predicted values ​​and the actual measured values; These are external forces and moments calculated by CFD; The forces and moments are calculated using a dynamic model; This is the input vector of the AUV, where the acceleration term is calculated from the dynamic model, and the velocity and rudder angle terms are provided by the CFD. This is the error weighting matrix for the forces and moments of each degree of freedom.

[0178] Particle Swarm Optimization (PSO) is a global optimization algorithm based on swarm intelligence. Its basic idea is to find the global optimum in the solution space by simulating the cooperative behavior of flocks of birds or schools of fish during foraging. The core of the algorithm lies in approximating the optimal solution through the cooperative search of a swarm of particles. Each particle represents a feasible solution, and its position is determined by the parameters of the problem to be optimized. Particles continuously adjust their velocity and position to find the optimal solution; the update formulas for velocity and position are:

[0179] ;

[0180] in, and Let c1 and c2 represent the velocity and position of particle i in the t-th iteration, respectively. w is the inertia weight, used to balance global and local search capabilities. c1 and c2 are learning factors that control the particle's movement toward its optimal position. and global optimal position The movement degree, r1 and r2 are random numbers ranging from [0,1], used to increase the randomness of the search. In this way, the particle can continuously update its position in the solution space, gradually approaching the global optimum.

[0181] In the identification of AUV dynamic model parameters, the specific implementation steps of the particle swarm optimization algorithm include: first, initializing the position and velocity of the particle swarm, where the position of each particle represents a set of model parameters to be identified. Then, the fitness value of each particle is calculated, which is the objective function. , representing the sum of the absolute values ​​of the errors between the model's predicted values ​​and the actual measured values. Then, the particle's velocity and position are adjusted according to the update formula, and the individual optimal position of each particle is updated. and global optimal position Repeat the above process until the objective function value meets the accuracy requirement or the maximum number of iterations is reached, and finally output the optimal parameters. .

[0182] The ALOS-MPC control algorithm is constructed for the AUV studied in this paper, which has a propeller control structure. Its motion control system exhibits underactuated characteristics, requiring more degrees of freedom to be controlled than the number of actuators. Using LOS guidance to transform the path tracking problem into a heading control or pitch control problem can overcome the underactuated nature of the AUV. The LOS guidance algorithm, as a classic guidance method, has advantages such as computational simplicity and accurate guidance, and has been widely used in fields such as autonomous driving tracking, missile guidance and control, and AUV motion control. It can not only track straight lines but also transform curved path problems into straight path tracking problems.

[0183] The principle of the LOS horizontal plane guidance algorithm is as follows: Figure 2 As shown, waypoints , and Connecting the lines forms the target path; the current target path is a straight line. The current waypoint is The offset of the AUV's centroid from the current target path is... The control objective of path tracking is to converge the offset to 0. The guidance strategy employed by the LOS algorithm is to determine a line-of-sight guidance point on the current target path. This simplifies line tracking to point tracking. (Line-of-sight guide point) There are two methods for selecting the point. The first method is to draw a circle with a certain radius centered on the AUV's centroid, and select the point that is closer to the next waypoint among the intersections with the current target path. The second method involves drawing a perpendicular line from the AUV particle to the current target path, with a fixed distance from the foot of the perpendicular along the current target path. The point is The first method requires calculating the intersection point and may result in no intersection. The second method is prone to oscillations and cannot converge under large offsets and high speeds. Since AUVs have relatively low speeds, the second method is chosen. Therefore, the line-of-sight angle... for:

[0184] ;

[0185] To ensure the AUV is aligned with the line-of-sight guide point To get closer, you need to adjust your heading. The expected heading angle is:

[0186] ;

[0187] Due to drift angle Because of its existence, the heading angle and the yaw angle are not equal; the yaw angle is:

[0188] ;

[0189] From the above equation, we can see that the desired heading angle is:

[0190] ;

[0191] In traditional LOS guidance algorithms, line-of-sight distance... The lateral offset is fixed at the beginning of path tracking. The distance is too large, requiring the AUV to quickly approach the current target path, but due to the fixed line-of-sight distance... Due to limitations, the hull cannot turn quickly, resulting in slow convergence speed. As the AUV approaches the current target path, stable tracking is required, necessitating a shorter line-of-sight distance. This can easily lead to overshoot and oscillation problems, therefore this section uses a method based on lateral offset. Size, adjust line of sight in real time Length, variable line of sight distance The length is determined by the following formula.

[0192] ;

[0193] As can be seen from the formula above, the line of sight distance... Length varies with lateral offset Decrease and increase and These are the maximum and minimum values ​​of the line-of-sight distance. This is the line-of-sight convergence coefficient. Based on experience, the maximum and minimum line-of-sight distances should be 2 to 5 times the length of the submarine. The value needs to be adjusted based on the AUV's handling performance. When the lateral offset is large, Taking a smaller value can improve the convergence speed. As the AUV approaches the target path... Taking a larger value ensures convergence stability and effectively prevents overshoot.

[0194] When switching target paths, it is usually determined whether the current AUV has entered the acceptance circle of the current target path's endpoint, such as... Figure 2 As shown. If the difference in waypoints between adjacent paths is large, that is... When the radius is large, the AUV needs to turn earlier to allow sufficient time to adjust its attitude. If the receiving circle radius is too small, the turning lag will lead to a large overshoot. Conversely, when... When the size is small, turning too early can lead to incomplete tracking of the current path, therefore a method is designed based on... A method for adaptively adjusting the radius of the receiving circle. The formula for the radius of the receiving circle is as follows:

[0195] ;

[0196] in and These are parameters to be determined. Related to the handling performance of the AUV, it is generally taken as 1 to 2 times the full-steering turning radius. Indicates the angle at the bend in the target path.

[0197] MPC has excellent constraint handling capabilities and strong engineering application potential, making its application in the field of underwater vehicle motion control highly significant. The basic principles of MPC are as follows: Figure 3 As shown. At each sampling time, the established system mathematical model is used to calculate the current time... The measurement output predicts the system state over a future period of time, which is called the prediction time domain. To predict the time domain The error between the predicted output and the target trajectory is used as the objective function to solve for the prediction time domain. The predictive control input series that minimizes the objective function is obtained, and the first element of the obtained predictive control input series is taken as the current time step. The system control input is obtained by repeating the above process at the next time step. This rolling calculation yields the locally optimal control input at each time step.

[0198] As can be seen from the above process, the basic framework of MPC includes three aspects: a predictive model for predicting the future state of the system, a rolling optimization strategy for solving the objective function, and a feedback correction scheme.

[0199] The model prediction algorithm is based on the AUV dynamics model; therefore, a prediction model can be built, and the motion state of the AUV at any future time can be predicted simply by knowing the control input and the initial state. Assuming the control input sequence in the prediction time domain is known, the prediction time domain can be predicted from the current state. Internal system output. Future The system state prediction equation is as follows:

[0200] ;

[0201] Output matrix Substituting into the system state prediction equation, we obtain the system output prediction equation as follows:

[0202] ;

[0203] use and The coefficient matrix representing the above equation can be obtained as follows:

[0204] ;

[0205] in Indicates the current Time prediction AUV system status at all times Indicates the current The system output sequence in the prediction time domain for time-time prediction. Indicates the current The system input sequence at time t.

[0206] As a type of optimal control, the most important step in model predictive control is constructing the objective function. For the AUV path tracking problem, the first priority is to minimize the tracking error in the prediction time domain. The energy consumption issue, i.e., minimizing the change in the input rudder angle, must also be considered. Therefore, the objective function is constructed as follows:

[0207] ;

[0208] in To output the tracking error weighting matrix, Input a weighting matrix for the change in rudder angle.

[0209] Substituting the prediction equation into the optimization objective function yields:

[0210] ;

[0211] The first term in the formula is a constant and can be ignored. The remaining part can be written in standard quadratic programming form as follows:

[0212] ;

[0213] in , , The desired heading angle sequence.

[0214] MPC (Multi-Purpose Control) is widely recognized for its ability to handle system constraints effectively. In practice, AUV actuators are constrained by finite rates of change and saturation states. To achieve better control performance, these system constraints need to be taken into account. (The text then abruptly shifts to a seemingly unrelated topic: "by rudder angle change...") The rudder angle in the time domain of the rudder prediction can be obtained as shown in the following formula.

[0215] ;

[0216] The coefficient matrix in the above formula adopts and Therefore, the saturation constraint of the rudder is:

[0217] ;

[0218] ;

[0219] in, , These are the upper and lower limits of the rudder amplitude, respectively.

[0220] The rudder speed constraint expression is as follows:

[0221] ;

[0222] in, and These are the upper and lower limits of the rudder speed sequence.

[0223] The objective function is combined with the control input constraints, presenting a quadratic programming problem. The effective set is a classic algorithm for solving quadratic programming problems, particularly suitable for problems with linear equality and inequality constraints; this paper employs it for optimization.

[0224] Obtain the optimal rudder angle increment sequence After that, take The rudder angle command at the current moment is obtained by adding the first rudder angle increment to the rudder angle at the next moment. The above steps are repeated at the next moment, and so on, rolling optimization is used to achieve predictive control for AUV path tracking.

[0225] The multi-degree-of-freedom CFD motion equations for the underwater vehicle are constructed, and the flow field calculation output terms are embedded into them, as shown in the following form:

[0226] ;

[0227] ;

[0228] In the formula, For body mass, For the velocity of the center of mass, This indicates the external force generated by the rudder. It represents the forces in the flow field (including flow, wave forces, viscous forces, etc.). Let the moment of inertia tensor be... Let ω be the angular velocity of the underwater vehicle. This indicates the external torque generated by the rudder. It represents the torque of the flow field (including flow, wave force, viscous force, etc.).

[0229] The high-precision motion control co-simulation framework is realized through the collaborative simulation of CFD and MATLAB, ignoring the compressibility of water and the influence of water temperature, and adopting the Reynolds-averaged Navier-Stokes method (RANS) and the VOF (Volume of Fluid) wave model; MATLAB is mainly responsible for the AUV adaptive model predictive controller based on online rolling identification.

[0230] The framework's operation mainly involves the following steps:

[0231] (1) The rudder angle output by the controller at time t is input into the N-S equation to calculate the rudder force and rudder torque. Combined with the current flow field force and torque, the motion response of the AUV (including position, velocity and attitude) is calculated. Then the motion state at time t+1 is fed back to the CFD module to update the flow field boundary conditions.

[0232] (2) The motion state information of the AUV at time t+1 (including velocity, attitude and force) is imported into the online rolling parameter identification module to update the dynamic model parameters of the current motion state of the AUV.

[0233] (3) Then the dynamic model at time t+1 and the motion response of the AUV are imported into the ALOS-MPC controller to calculate the rudder angle required at time t+1.

[0234] This forms a closed-loop feedback control, which is performed once per control cycle. To improve simulation efficiency, parallel computing technology is used to accelerate the collaborative computation of CFD and MPC.

[0235] Due to sensor limitations, the underwater positioning accuracy of the AUV was insufficient during the lake trial. To facilitate location information acquisition, the AUV was set to a slightly positive buoyancy state, keeping it near the water surface for GPS positioning to obtain its location. On the day of the test, there was a northwest wind of approximately 2 m / s on the lake surface. Horizontal path tracking, being near the water surface, was affected by wind and wave disturbances, allowing for better testing of the controller's tracking performance. The CFD simulation settings were consistent with the actual experiment, including wind speed and direction.

[0236] Multi-waypoint path tracking test method: Maintaining a constant AUV propeller speed (and thrust during simulation), with a cruising speed of approximately 0.9 m / s, the AUV is initially remotely controlled to the vicinity of starting point A. It is then switched to autonomous tracking mode. The AUV sequentially passes through points B, C, and D, reaching the vicinity of point E to conclude the test. The control cycle is 0.2 s. The simulation ends when the AUV enters the receiving circle of the last pathpoint. Control parameters remain consistent with the simulation, as follows:

[0237] The performance of the ALOS+MPC control based on the time-varying parameter model proposed in this paper is compared with that of the control group based on the fixed parameter model, as shown in the figure. Figures 4 to 7 As shown. Figure 4 A multi-waypoint test trajectory curve; Figure 5 Offset curve for multi-waypoint tests; Figure 6 For multi-waypoint test heading curves Figure 7 This is a rudder curve diagram for multi-waypoint testing.

[0238] Four performance indicators were selected to conduct a quantitative comparative analysis of the multi-waypoint path tracking performance of the three control algorithms: the average tracking offset, the average overshoot of the heading during three turns, the average time required for the three turns, and the total change in rudder angle. The four performance indicators obtained are shown in Table 3.

[0239] Table 3 Performance Indicators of Multi-Flight Test

[0240]

[0241] Through detailed data analysis and simulation results, this application demonstrates that the proposed high-fidelity simulation framework closely approximates the actual experiment, and that the proposed ALOS+ORPI+MPC control algorithm has smaller tracking offset error and less tracking overshoot.

[0242] The advantage of the co-simulation framework is that it can clearly analyze the distribution of flow field, pressure, etc. at various times and in various regions during the movement of the AUV. In order to further analyze the response of the UUV to trajectory deviation in Case 3 (ALOS+ORPI+MPC, CFD) in Table 3, the flow field at 8 typical time points is analyzed. Figure 8 This reflects the positional distribution of the AUV during path tracking at these eight time points. For details of the flow field, see [link to flow field diagram]. Figure 9 , Figure 10 .

[0243] This study addresses the issues of dynamic modeling, control algorithms, and simulation in AUV motion control. It proposes a high-precision motion control co-simulation framework integrating CFD and MPC, aiming to solve the problem of "simulation without realism" in autonomous AUV navigation under complex flow fields. Simultaneously, within this framework, real-time AUV navigation data provided by CFD is used to perform online rolling parameter identification of the dynamic model in MPC, improving the control accuracy of MPC. This provides theoretical support for high-precision control and simulation of autonomous AUV navigation, possessing significant engineering application value and academic significance.

[0244] Figure 12 This is a schematic diagram of the structure of the underwater vehicle flow field and time-varying motion model predictive control co-simulation device provided in the embodiments of this application, as shown below. Figure 12 As shown, the device includes: an initial parameter determination module 10 and an iteration module 20. Wherein:

[0245] Initial parameter determination module 10 is used to determine the initial parameters of the autonomous underwater vehicle (AUV) dynamic model;

[0246] The iteration module 20 is used to continuously iterate the AUV's pose based on the initial parameters of the AUV dynamic model and the AUV's navigation status information until the pose deviation is within a preset range. The AUV's navigation status information is obtained through CFD simulation.

[0247] It is understood that the detailed functional implementation of each of the above units / modules can be found in the description in the aforementioned method embodiments, and will not be repeated here.

[0248] It should be understood that the above-described device is used to execute the methods in the above embodiments. The implementation principle and technical effect of the corresponding program modules in the device are similar to those described in the above methods. The working process of the device can be referred to the corresponding process in the above methods, and will not be repeated here.

[0249] Based on the methods in the above embodiments, this application provides an electronic device. Figure 13 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application, such as... Figure 13 As shown, the electronic device may include a processor 810, a communications interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communications interface 820, and the memory 830 communicate with each other through the communication bus 840. The processor 810 can call logical instructions in the memory 830 to execute the methods in the above embodiments.

[0250] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0251] Based on the methods in the above embodiments, this application provides a computer-readable storage medium storing a computer program that, when run on a processor, causes the processor to execute the methods in the above embodiments.

[0252] Based on the methods in the above embodiments, this application provides a computer program product that, when run on a processor, causes the processor to execute the methods in the above embodiments.

[0253] It is understood that the processor in the embodiments of this application can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. A general-purpose processor can be a microprocessor or any conventional processor.

[0254] The method steps in this application embodiment can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can reside in an ASIC.

[0255] It is understood that the various numerical designations used in the embodiments of this application are merely for the convenience of description and are not intended to limit the scope of the embodiments of this application.

[0256] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for co-simulation of flow field and time-varying motion model predictive control of an underwater vehicle, characterized in that, include: Determine the initial parameters of the dynamic model of the autonomous underwater vehicle (AUV); Based on the initial parameters of the AUV dynamics model and the AUV's flight status information, the AUV's pose is continuously iterated until the pose deviation is within a preset range. The AUV's flight status information is obtained through CFD simulation. Iterate the pose of the AUV, including: Based on the initial parameters of the AUV dynamic model and the AUV's navigation status information, the time-varying parameters of the AUV dynamic model are identified online using an optimization algorithm to determine the optimal parameters of the AUV dynamic model. Based on the optimal parameters of the AUV dynamic model and the pose deviation between the current AUV pose and the target pose, the required rudder angle at the current moment is determined by the adaptive line-of-sight ALOS guidance algorithm and the model predictive control (MPC) algorithm. The optimal parameters of the AUV dynamic model are passed to the MPC algorithm to update the model parameters of the MPC algorithm. Based on the required rudder angle at the current moment, the flow field force and flow field torque are determined through the NS equation, and the flow field force and flow field torque are transmitted to the CFD multi-degree-of-freedom motion equation to update the current pose of the AUV. Based on the target pose and the updated AUV pose, determine whether the pose deviation is within the preset range; The online rolling identification of time-varying parameters of the AUV dynamics model through optimization algorithms includes: The function used to solve the optimization problem is the particle swarm optimization algorithm. The function for the optimization problem is as follows: ； in, It is the parameter vector of the AUV dynamic model. Indicates a fixed parameter. The parameters to be optimized are represented by . The initial solution of the particle swarm optimization algorithm uses the initial parameters of the AUV dynamics model. It is the objective function. This represents the measurement value obtained through CFD simulation. This represents the model prediction value obtained through the AUV dynamics model. , express The first in item, , express The first in item, The value is 4. This represents the force acting on the hull along the y-axis. This represents the torque acting on the hull along the z-axis. This represents the force acting on the rudder in the y-axis direction; This represents the torque acting on the rudder in the z-axis direction. It is the input vector of the AUV dynamics model. This represents the lateral velocity obtained through CFD simulation. This represents the yaw rate obtained through CFD simulation. This represents the lateral acceleration obtained through CFD simulation. This represents the yaw acceleration obtained through CFD simulation. This represents the rudder angle obtained through CFD simulation. This represents the error weighting matrix for forces and moments under each degree of freedom; The process of determining the required rudder angle at the current moment using the adaptive line-of-sight (ALOS) guidance algorithm and the model predictive control (MPC) algorithm includes: Based on the pose deviation between the current pose of the AUV and the target pose, the desired heading angle is obtained through the ALOS guidance algorithm; The optimal parameters of the AUV dynamics model are passed to the MPC algorithm to update the model parameters of the MPC algorithm. Based on the desired heading angle, AUV speed information and AUV current pose, the required rudder angle at the current moment is obtained through the MPC algorithm after updating the model parameters.

2. The co-simulation method for predictive control of flow field and time-varying motion model of underwater vehicle according to claim 1, characterized in that, The flow field force and the flow field torque are determined by the following formulas: ； ； In the formula, For flow field force, For AUV surfaces, The local pressure on the surface of the AUV, Let be the normal vector of the AUV surface. The shear stress on the surface of the AUV. For the flow field torque, Let be the distance vector from the centroid of the AUV to the center of the surface.

3. The underwater vehicle flow field and time-varying motion model predictive control co-simulation method according to claim 2, characterized in that, Local pressure on the surface of an AUV It is determined by the following formula: ； in, For micro-elements on the surface of AUV, It is the sum of the static and dynamic pressures of the fluid.

4. The co-simulation method for predictive control of flow field and time-varying motion model of underwater vehicle according to claim 1, characterized in that, The process of transferring flow field forces and torques to the CFD multi-degree-of-freedom motion equations and updating the current pose of the AUV includes: Based on the flow field force and flow field torque, the angular velocity and the velocity of the AUV's center of mass are obtained through CFD multi-degree-of-freedom motion equations; Update the current pose of the AUV based on its angular velocity and the velocity of its center of mass. The CFD multi-degree-of-freedom motion equations are as follows: ； ； In the formula, This is the sum of the mass of the AUV and the added mass of the AUV. The velocity of the AUV's center of mass. For flow field force, External forces generated by multibody constraints or joints To control force, For the previous moment, For the current moment, Let the moment of inertia tensor be... The angular velocity of the AUV, For the flow field torque, External torque generated by multibody constraints or joints To control the torque.

5. The co-simulation method for predictive control of flow field and time-varying motion model of underwater vehicle according to claim 4, characterized in that, The method of updating the AUV's current pose based on its angular velocity and center-of-mass velocity specifically includes updating the AUV's current pose using the following formula: ； ； ； ； ； ； in, , and Let be the position components of the AUV pose along the horizontal, vertical, and longitudinal axes, respectively. For the next moment, For the current moment, , and Let be the velocity components of the AUV's center of mass in the horizontal, vertical, and longitudinal directions, respectively. , and Let be the rotation angle components of the AUV pose in the horizontal, vertical, and longitudinal directions, respectively. , and Let be the angular velocity components of the AUV in the horizontal, vertical, and longitudinal directions, respectively. The interval between the next moment and the current moment; The updated AUV pose includes , , , , and .

6. A co-simulation device for predictive control of flow field and time-varying motion model of an underwater vehicle, characterized in that, include: The initial parameter determination module is used to determine the initial parameters of the autonomous underwater vehicle (AUV) dynamic model. The iteration module is used to continuously iterate the AUV's pose based on the initial parameters of the AUV dynamic model and the AUV's flight status information until the pose deviation is within a preset range. The AUV's flight status information is obtained through CFD simulation. The iterative module includes: a model optimal parameter determination unit, a rudder angle determination unit, an AUV pose update unit, and a pose deviation judgment unit, wherein: The optimal parameter determination unit is used to determine the optimal parameters of the AUV dynamic model by performing online rolling identification of the time-varying parameters of the AUV dynamic model based on the initial parameters of the AUV dynamic model and the AUV navigation state information through optimization algorithms. The rudder angle determination unit is used to determine the required rudder angle at the current moment based on the optimal parameters of the AUV dynamic model and the pose deviation between the current pose of the AUV and the target pose, through the adaptive line-of-sight ALOS guidance algorithm and the model predictive control (MPC) algorithm. The optimal parameters of the AUV dynamic model are passed to the MPC algorithm to update the model parameters of the MPC algorithm. The AUV pose update unit is used to determine the flow field force and flow field torque based on the required rudder angle at the current moment through the NS equation, and to transmit the flow field force and flow field torque to the CFD multi-degree-of-freedom motion equation to update the current pose of the AUV. The pose deviation judgment unit is used to determine whether the pose deviation is within a preset range based on the target pose and the updated AUV pose. The online rolling identification of time-varying parameters of the AUV dynamics model through optimization algorithms includes: The function used to solve the optimization problem is the particle swarm optimization algorithm. The function for the optimization problem is as follows: ； in, It is the parameter vector of the AUV dynamic model. Indicates a fixed parameter. The parameters to be optimized are represented by . The initial solution of the particle swarm optimization algorithm uses the initial parameters of the AUV dynamics model. It is the objective function. This represents the measurement value obtained through CFD simulation. This represents the model prediction value obtained through the AUV dynamics model. , express The first in item, , express The first in item, The value is 4. This represents the force acting on the hull along the y-axis. This represents the torque acting on the hull along the z-axis. This represents the force acting on the rudder in the y-axis direction; This represents the torque acting on the rudder in the z-axis direction. It is the input vector of the AUV dynamics model. This represents the lateral velocity obtained through CFD simulation. This represents the yaw rate obtained through CFD simulation. This represents the lateral acceleration obtained through CFD simulation. This represents the yaw acceleration obtained through CFD simulation. This represents the rudder angle obtained through CFD simulation. This represents the error weighting matrix for forces and moments under each degree of freedom; The process of determining the required rudder angle at the current moment using the adaptive line-of-sight (ALOS) guidance algorithm and the model predictive control (MPC) algorithm includes: Based on the pose deviation between the current pose of the AUV and the target pose, the desired heading angle is obtained through the ALOS guidance algorithm; The optimal parameters of the AUV dynamics model are passed to the MPC algorithm to update the model parameters of the MPC algorithm. Based on the desired heading angle, AUV speed information and AUV current pose, the required rudder angle at the current moment is obtained through the MPC algorithm after updating the model parameters.

7. An electronic device, characterized in that, include: At least one memory for storing computer programs; At least one processor is configured to execute a program stored in the memory, wherein when the program stored in the memory is executed, the processor is configured to perform the method as described in any one of claims 1-5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is run on the processor, it causes the processor to perform the method as described in any one of claims 1-5.

Images