An automatic water surface vehicle trajectory generation method, device, equipment and storage medium

Abstract

The application discloses an automatic water surface vehicle trajectory generation method, device and equipment and a storage medium, relates to the technical field of mathematics and automation control, and comprises the following steps: a state space model for representing a system state of an automatic water surface vehicle is established based on a unified model, and the state space model is discretized to obtain a state space equation; the system state of the automatic water surface vehicle at different moments is predicted through the state space equation; the minimum value of the system state at each moment is controlled to generate a trajectory generation optimization equation; constraint conditions for representing safe operation of the automatic water surface vehicle and a cost function for representing operation conditions of the automatic water surface vehicle are set for the trajectory generation optimization equation to obtain a trajectory generation optimization problem; and an operation trajectory of the automatic water surface vehicle is generated through an optimal solution of the trajectory generation optimization problem. The state space model considers complex marine environment interference and vehicle kinematics and dynamics constraints, and more accurate trajectory generation can be realized.

Description

Technical Field

[0001] This invention relates to the fields of mathematics and automation control technology, and in particular to an automatic method, apparatus, device and storage medium for generating trajectories of surface vehicles. Background Technology

[0002] Automated surface vehicles (ASVs) are unmanned vessels crucial for ocean surveying, underwater exploration, and water quality monitoring. ASVs face unique challenges in motion planning due to the complex marine environment they operate in, influenced by factors such as currents, tides, and waves. Enhancing the safe and efficient execution of motion planning across various missions is essential. The complexity and uncertainty of the marine environment further complicate motion planning, requiring algorithm design to consider unpredictable factors such as changing ocean currents, hidden islands, and shallow waters to ensure ASVs can adapt to constantly changing surface conditions during mission execution.

[0003] Existing motion planning techniques for automated surface vehicles (ASVs) can be broadly categorized into four types: geometry-based methods, potential field-based methods, and numerical optimization methods. Geometry-based methods extract environmental geometric features and utilize curve features for motion planning. Examples include Voronoi diagrams for generating initial paths and improved RRT ship path generation algorithms based on AIS prior information and DP compression. These works address the path generation problem. However, in the presence of dynamic obstacles, local planning needs to be re-executed. Potential field-based methods construct trajectories by combining attractive and repulsive potential fields. Examples include using a fast potential field model for complex obstacle environments suitable for vector maps and adaptive adjustable fast travel methods. These works also address the path generation problem. However, the modeling of kinematics and dynamics has not been adequately considered. Numerical optimization methods further transform the planning problem into a constrained optimization problem. Examples include backward horizon trajectory planning and time-optimal path generation and tracking control methods based on nonlinear model predictive control and spatial reconstruction.

[0004] In summary, while existing motion planning technologies for autonomous surface vehicles can solve the trajectory generation problem, these methods only consider the impact of obstacles in the marine environment on trajectory generation. The conditions they consider are relatively simple and cannot cope with the complex real marine environment, resulting in low accuracy of the final generated trajectory. Summary of the Invention

[0005] Based on the shortcomings of the existing technology, the present invention provides an automatic water vehicle trajectory generation method, device, equipment and storage medium, which solves the problem that the existing methods only consider the influence of obstacles in the marine environment on trajectory generation, and the conditions considered are relatively simple, which cannot cope with the real complex marine environment, resulting in low accuracy of the final generated trajectory.

[0006] The present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides an automatic method for generating the trajectory of a vehicle on a water surface, comprising the following steps:

[0008] A state-space model for characterizing the system state of an automated surface vehicle is established based on a unified model of its kinematics and dynamics; wherein, the total environmental disturbances in the unified model include wind, waves and ocean current disturbances.

[0009] Discretize the state-space model to obtain the state-space equation. Predict the system state of the automated surface vehicle at different times using the state-space equation. Control the minimum system state at each time and generate trajectory generation optimization equation.

[0010] By setting constraints to characterize the safe operation of the automated surface vehicle and a cost function to characterize the operation of the automated surface vehicle, the trajectory generation optimization problem is obtained.

[0011] Obtain the optimal solution to the trajectory generation optimization problem, and generate the running trajectory of the automated surface vehicle using the optimal solution.

[0012] Preferably, the state-space model for characterizing the system state of the automated surface vehicle is established based on a unified model, and the specific state-space model is as follows:

[0013]

[0014] In the formula, For the state-space model of the automated surface vehicle, R(ψ) is the rotation matrix from the fixed reference frame of the vehicle to the fixed reference frame of the Earth, ν is the heading vector in the fixed reference frame of the Earth, and M and C... A D is the total inertia matrix, τ is the input force and torque of the vehicle propeller, τ d Disturbing the marine environment, C is the derivative of the water flow velocity. RB and M A This is the intermediate matrix.

[0015] Preferably, the state-space model is discretized to obtain the state-space equations, which are shown below:

[0016] x(t+1)=φ(t)x(t)+Γ(t)τ(t)+ε(t);

[0017] In the formula, x is the system state, t is the iteration number, φ(t) is the state matrix of the t-th iteration, Γ(t) is the input matrix of the t-th iteration, ε(t) is the nonlinear matrix of the t-th iteration, and τ(t) is the input force and torque of the vehicle propeller in the t-th iteration.

[0018] Preferably, the minimum system state value at each moment is controlled to generate a trajectory generation optimization equation, which is specifically shown below:

[0019]

[0020] Where min J(k) is the trajectory generation optimization equation. Let $T$ represent the input cost, $Q$ represent the input weighting matrix, $J$ represent the optimization objective function, $k$ represent the step size, and $T$ represent the total number of iterations.

[0021] Preferably, the trajectory generation optimization equation is configured with constraints characterizing the safe operation of the automated surface vehicle and a cost function characterizing the operation of the automated surface vehicle, resulting in a trajectory generation optimization problem, as detailed below:

[0022]

[0023] st‖H(x(t)-x n (t))‖2≥R col ;

[0024] x(0)=x k ;

[0025] x(T)=x end ;

[0026] In the formula, M(x) end Let L be the terminal cost function. c (τ,t) is the comfort cost function, P is the speed and input obstacle function, H is the weight matrix, and x n (t) represents the trajectory of the moving obstacle, R col To avoid a safe distance from obstacles, x k Let x be the system state with a step size of k. end This refers to the terminal state of the system immediately preceding the next time step.

[0027] Preferably, obtaining the optimal solution to the trajectory generation optimization problem includes the following steps:

[0028] The trajectory generation optimization problem can be rewritten as follows:

[0029]

[0030]

[0031] In the formula, τ T The transpose of the force and torque input to the vehicle propeller. For the first variable to be rewritten, The matrix f is the input force and torque to the vehicle propeller. T For the second rewritten variable, To avoid obstacles;

[0032] A recurrent neural network is constructed based on the rewritten trajectory generation optimization problem, as shown below:

[0033]

[0034] In the formula, and ∈ y Let be two time constants, and y be the hidden neuron state of the direction vector. The gradient of the objective function. Let λ be the gradient of the inequality constraint, and λ be the hidden neuron state that handles the inequality.

[0035] The optimal solution is obtained by iterating through the recurrent neural network.

[0036] In a second aspect, the present invention provides an automatic water surface vehicle trajectory generation device, comprising:

[0037] A construction module is used to establish a state-space model representing the system state of an automated surface vehicle based on a unified model of its kinematics and dynamics; wherein the total environmental disturbances in the unified model include wind, waves and ocean current disturbances.

[0038] The discrete module is used to discretize the state-space model to obtain the state-space equation. The system state of the automated surface vehicle at different times is predicted by the state-space equation. The minimum value of the system state at each time is controlled to generate the trajectory generation optimization equation.

[0039] The generation module is used to set constraints to characterize the safe operation of the automated surface vehicle and a cost function to characterize the operation of the automated surface vehicle on the trajectory generation optimization equation, so as to obtain the trajectory generation optimization problem.

[0040] The solution module is used to obtain the optimal solution to the trajectory generation optimization problem, and to generate the running trajectory of the automated surface vehicle based on the optimal solution.

[0041] Thirdly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method for generating automatic water surface vehicle trajectories.

[0042] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the foregoing automatic trajectory generation method for an autonomous surface vehicle is implemented.

[0043] Compared with the prior art, at least one of the above technical solutions adopted by the present invention can achieve the following beneficial effects:

[0044] The present invention proposes an automatic trajectory generation method for an autonomous surface vehicle. Modeling is performed through a unified kinematic and dynamic model of a fully actuated autonomous surface vehicle and a complex marine environment model, wherein the complex marine environment includes wind, wave and ocean current disturbances, so as to obtain a state space model. The state space model takes into account real complex marine environment disturbances and kinematic and dynamic constraints of the vehicle, enabling more accurate trajectory generation. On this basis, the state space model is discretized to obtain a state space equation, a trajectory generation optimization equation is generated based on the characteristics of the state space equation, and constraint conditions for characterizing the safe operation of the autonomous surface vehicle and a cost function for characterizing the operation status of the autonomous surface vehicle are integrated, so as to improve the driving performance and real-time performance of the autonomous surface vehicle. Finally, the trajectory generation problem is converted into a corresponding optimization problem and solved, so that the generated trajectory meets the obstacle avoidance requirements under complex marine environment disturbances, providing theoretical support for further research on actual trajectory generation of autonomous surface vehicles. Description of Drawings

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following briefly introduces the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention, and those skilled in the art can obtain other drawings according to these drawings without creative effort.

[0046] Figure 1 is a flow block diagram of an automatic trajectory generation method for an autonomous surface vehicle of the present invention;

[0047] Figure 2 is a simulation diagram of the present invention;

[0048] wherein: Figure 2 (a): an obstacle avoidance path for two fixed obstacles, Figure 2 (b): an obstacle avoidance path for three fixed obstacles, Figure 2 (c): an obstacle avoidance path for four fixed obstacles, Figure 2 (d): an obstacle avoidance path for a moving obstacle. Detailed Description of Embodiments

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] This invention provides an automatic method for generating the trajectory of a surface vehicle, specifically a method for generating the backward trajectory of a surface vehicle based on a dual-timescale neurodynamics approach. Figure 1 As shown, this embodiment applies the method to a fully driven autonomous surface vehicle. By modeling the autonomous surface vehicle and the complex marine environment, a model suitable for theoretical analysis is obtained. The method includes the following steps:

[0051] Step 1: Construct the kinematic and dynamic models of the fully driven autonomous surface vehicle and the complex marine environment model.

[0052] Since modeling in complex marine environments is difficult, this invention proposes a Gaussian distribution function to estimate the influence equation of complex marine environments on the input of fully driven automated surface vehicles, which is the total environmental disturbance input τ. d Its equation is:

[0053]

[0054] Where α1, α2, and α3 represent the weighting parameters for wind, wave, and ocean current disturbances, respectively. F1(rand), F2(rand), and F3(rand) represent the Gaussian distribution functions of wind, wave, and ocean current disturbances, respectively, and rand represents a random number generated in the range [-1, 1].

[0055] In this embodiment, the corresponding modeling parameters are obtained by randomizing the complex marine environment, as shown in Table 1.

[0056] Table 1 Modeling parameters

[0057]

[0058] Based on the conversion rules between the Earth's fixed reference frame and the machine's fixed reference frame, as well as Newton's second law, a unified kinematic and dynamic model of the vehicle is constructed, which is modeled as follows:

[0059]

[0060] in, η and v represent the position and heading vectors in the Earth's fixed reference frame, respectively. rLet R(ψ) represent the relative velocity of the object in a fixed reference frame, and let R(ψ) represent the rotation matrix from the fixed reference frame of the object to the fixed reference frame of the Earth. M and C... A D represents the total inertia matrix, C RB (v) represents the centrifugal force matrix; τ represents the input force and torque of the vehicle propeller. d This indicates total environmental interference.

[0061] By measuring the automated surface vehicle, the corresponding modeling parameters can be obtained. These parameters are then input into the unified kinematics and dynamics model of the vehicle to obtain a specific model.

[0062] In complex marine environments, consider the water flow velocity ν in a fixed reference frame. c The relationship between water flow velocity, relative velocity, and heading vector is as follows:

[0063] ν c =ν-ν r .

[0064] Define x = [η, ν] T By combining the above vehicle models, we can obtain the following state-space model of the vehicle:

[0065]

[0066] in,

[0067]

[0068] By further rewriting, the following simplified state-space model can be obtained:

[0069]

[0070] Step 2: Based on the given state-space model, discretize the constructed vehicle model and construct the discretized state-space equations.

[0071] Based on the characteristics of computer processing of discretized equations, the above state-space model needs to be discretized to construct state-space equations.

[0072] This state-space model is a nonlinear model, which is discretized using the fourth-order Runge-Kutta method. This method provides high accuracy in solving nonlinear differential equations and can guarantee the accuracy of the discretized model. For nonlinear functions... The fourth-order Runge-Kutta method is used to solve for the function value at the next time step k+1, as shown below:

[0073]

[0074] Where h represents the time interval between time steps k and k+1, and K1, K2, K3, and K4 are defined as follows:

[0075] K1 = hf(x,τ);

[0076]

[0077] K4 = hf(x + K3, τ).

[0078] Based on the above expression, the discretized state-space equation is as follows:

[0079] x(t+1)=φ(t)x(t)+Γ(t)τ(t)+ε(t);

[0080] in,

[0081]

[0082] In the formula, t represents the iteration number, φ(t) is the state matrix of the t-th iteration, Γ(t) is the input matrix of the t-th iteration, ε(t) is the nonlinear matrix of the t-th iteration, and τ(t) is the input force and torque of the vehicle propeller in the t-th iteration.

[0083] Step 3: First, based on the backward horizon rule, the system state at the next moment can be predicted using the state-space equation. By controlling the minimum predicted value at each moment, the fully driven automatic water vehicle can generate a safe trajectory. Therefore, based on the existing research, this problem has been transformed into an optimization problem.

[0084] set up and Where T>1 represents the user-defined prediction time for trajectory generation. In trajectory generation, τ(0) represents the control of the current time k.

[0085] By simultaneously solving the state-space equations, we obtain the following affine constraints:

[0086]

[0087] in,

[0088]

[0089] This leads to an iterative model based on the backward horizon rule.

[0090] By using this iterative model and controlling the minimization of the input, a trajectory generation optimization equation can be obtained as follows:

[0091]

[0092] in, Let represent the input cost, and Q represent the input weighting matrix.

[0093] Step 4: Secondly, the generated trajectory must be a safe trajectory, that is, it must be able to avoid obstacles and take into account the interference of complex marine environment, so that the fully driven autonomous surface vehicle can maintain its own stability and reach the target point within a reasonable time. Therefore, this purpose is achieved by setting a series of constraints and obstacle functions.

[0094] The constraints defining moving or fixed obstacles must satisfy the following conditions:

[0095] ‖H(x(t)-x n (t))‖2≥R col ;

[0096] Among them, R col H represents the safe distance to avoid collisions with obstacles, where H is the weight matrix and x represents the distance to avoid collisions with obstacles. n (t) represents the trajectory of the moving obstacle.

[0097] Define the velocity and input obstacle functions. First, define the velocity constraints and input constraints, as shown below:

[0098] v min ≤ν(t)≤ν max ;

[0099] τ min ≤τ(t)≤τ max ;

[0100] Among them, v min and v max The lower and upper limits of velocity, τ min and τ max These are the lower and upper bounds of the input. Then, the constraint is written as g. i The velocity and input barrier functions are defined in the form of (τ(t)) as follows:

[0101]

[0102] Where i represents the number of constraints, and μ represents the weight coefficient of the obstacle function.

[0103] A terminal cost function and a comfort cost function are defined. To enable the fully driven autonomous surface vehicle to quickly move towards the target point, a terminal cost function is designed. Simultaneously, to maintain vehicle stability during travel and improve load-bearing capacity, a comfort cost function is designed. The terminal cost function and comfort cost function are as follows:

[0104]

[0105] Where W and Q represent the terminal weighting matrix and the comfort weighting matrix, respectively.

[0106] In summary, the final optimization problem can be obtained as follows:

[0107]

[0108] st‖H(x(t)-x n (t))‖2≥R col ;

[0109] x(0)=x k ;

[0110] x(T)=x end ;

[0111] Where, x k This refers to the system state at time k, x end This indicates the terminal state of the system immediately preceding the next time step.

[0112] Step 5: Based on the optimization problem proposed above, a dual-timescale neurodynamics method was designed. Specifically, a recurrent neural network for online trajectory generation of fully driven automated surface vehicles was proposed, and the effectiveness of the method was demonstrated through simulation experiments.

[0113] Construct an optimization problem for trajectory generation and solve it by building a recurrent neural network.

[0114] The final optimization problem can be rewritten in the following form:

[0115]

[0116] in, This indicates obstacle avoidance constraints.

[0117] The recurrent neural network used for online trajectory generation of fully driven automated surface vehicles is shown below:

[0118]

[0119] Table 2 Algorithm 1

[0120]

[0121] The above problem can be solved using Algorithm 1 from Table 2, where... and ∈ y There are two time constants, where y is the hidden neuron state of the direction vector. This represents the gradient of the objective function. Let represent the gradient of the inequality constraint, and λ be the hidden neuron state that handles the inequality. It is a symmetric positive definite matrix used to approximate the Lagrange-Hessian matrix in optimization problems. If If the result is positive definite, then the output of the recurrent neural network converges globally to the optimal solution.

[0122] First, to achieve more accurate trajectory generation, this invention establishes a complex environmental disturbance model and a nonlinear model for automated surface vehicles (ASVs). Second, based on this model, a comfort evaluation cost function and a series of obstacle functions are proposed to improve the driving performance and real-time performance of ASVs. Finally, corresponding algorithms are designed using a dual-timescale neurodynamic optimization method and a backward trajectory generation method. Compared with traditional global planning methods, the complexity of the optimization problem is effectively reduced by integrating a backward time-domain optimization mechanism. The fast convergence capability of the dual-timescale neurodynamic method in optimization tasks solves the trajectory generation challenge of ASVs equipped with full propulsion configurations under meteorological disturbance conditions. Simulation results verify the effectiveness of the algorithm, providing theoretical support for further research on trajectory generation for practical ASVs.

[0123] Simulation diagram of the automatic water surface vehicle trajectory generation method of the present invention is shown below. Figure 2 As shown, stable trajectories can be generated under conditions with different numbers of fixed and moving obstacles, and the present invention greatly improves the accuracy of trajectory generation.

[0124] Based on the same concept, the present invention also provides an automatic water surface vehicle trajectory generation device, including a construction module, a discrete module, a generation module and a solution module.

[0125] The building module is used to establish a state-space model to characterize the system state of an automated surface vehicle based on a unified model of its kinematics and dynamics; the total environmental disturbances in the unified model include wind, wave and ocean current disturbances.

[0126] The discrete module is used to discretize the state-space model to obtain the state-space equation. The state-space equation is used to predict the system state of the automated surface vehicle at different times, control the minimum system state at each time, and generate trajectory generation optimization equations.

[0127] The generation module is used to set constraints to characterize the safe operation of the automated surface vehicle and a cost function to characterize the operation of the automated surface vehicle, thus obtaining the trajectory generation optimization problem.

[0128] The solution module is used to obtain the optimal solution to the trajectory generation optimization problem, and to generate the running trajectory of the automated surface vehicle using the optimal solution.

[0129] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method for generating automatic water vehicle trajectories.

[0130] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described method for generating automatic water surface vehicle trajectories.

[0131] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0132] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for automatically generating the trajectory of a vehicle on a water surface, characterized in that, Includes the following steps: A state-space model for characterizing the system state of an automated surface vehicle is established based on a unified model of its kinematics and dynamics; wherein, the total environmental disturbances in the unified model include wind, waves and ocean current disturbances. The state-space model is discretized to obtain the state-space equation. The system state of the automated surface vehicle at different times is predicted by the state-space equation. The minimum value of the system state at each time is controlled to generate the trajectory generation optimization equation. By setting constraints to characterize the safe operation of the automated surface vehicle and a cost function to characterize the operation of the automated surface vehicle, the trajectory generation optimization problem is obtained. Obtain the optimal solution to the trajectory generation optimization problem, and generate the running trajectory of the automated surface vehicle using the optimal solution.

2. The automatic water surface vehicle trajectory generation method as described in claim 1, characterized in that, The state-space model for characterizing the system state of the automated surface vehicle is established based on a unified model. The specific state-space model is shown below: In the formula, For the state-space model of the automated surface vehicle, R(ψ) is the rotation matrix from the fixed reference frame of the vehicle to the fixed reference frame of the Earth, v is the heading vector in the fixed reference frame of the Earth, and M and C... A D is the total inertia matrix, τ is the input force and torque of the vehicle propeller, τ d Disturbing the marine environment, C is the derivative of the water flow velocity. RB and M A This is the intermediate matrix.

3. The automatic water surface vehicle trajectory generation method as described in claim 1, characterized in that, The state-space model is discretized to obtain the state-space equations, which are shown below: x(t+1)=φ(t)x(t)+Γ(t)τ(t)+ε(t); In the formula, x is the system state, t is the iteration number, φ(t) is the state matrix of the t-th iteration, Γ(t) is the input matrix of the t-th iteration, ε(t) is the nonlinear matrix of the t-th iteration, and τ(t) is the input force and torque of the vehicle propeller in the t-th iteration.

4. The automatic water surface vehicle trajectory generation method as described in claim 3, characterized in that, The system state minimum value at each moment is controlled to generate a trajectory generation optimization equation, which is shown in detail below: Where minJ(k) is the trajectory generation optimization equation. Let $T$ represent the input cost, $Q$ represent the input weighting matrix, $J$ represent the optimization objective function, $k$ represent the step size, and $T$ represent the total number of iterations.

5. The automatic water surface vehicle trajectory generation method as described in claim 4, characterized in that, The trajectory generation optimization equation is configured with constraints to characterize the safe operation of the automated surface vehicle and a cost function to characterize the operation of the automated surface vehicle, resulting in the trajectory generation optimization problem, which is shown below: In the formula, M(X) end Let L be the terminal cost function. c (τ, t) is the comfort cost function, P is the speed and input obstacle function, H is the weight matrix, and x n (t) represents the trajectory of the moving obstacle, R col To avoid a safe distance from obstacles, x k Let x be the system state with a step size of k. end This refers to the terminal state of the system immediately preceding the next time step.

6. The automatic water surface vehicle trajectory generation method as described in claim 5, characterized in that, Obtaining the optimal solution to the trajectory generation optimization problem includes the following steps: The trajectory generation optimization problem can be rewritten as follows: In the formula, τ T The transpose of the force and torque input to the vehicle propeller. For the first variable to be rewritten, The matrix f is the input force and torque to the vehicle propeller. T For the second rewritten variable, To avoid obstacles; A recurrent neural network is constructed based on the rewritten trajectory generation optimization problem, as shown below: In the formula, and ∈ y Let be two time constants, and y be the hidden neuron state of the direction vector. The gradient of the objective function. Let λ be the gradient of the inequality constraint, and λ be the hidden neuron state that handles the inequality. The optimal solution is obtained by iterating through the recurrent neural network.

7. An automatic water surface vehicle trajectory generation device, characterized in that, include: A construction module is used to establish a state-space model representing the system state of an automated surface vehicle based on a unified model of its kinematics and dynamics; wherein the total environmental disturbances in the unified model include wind, waves and ocean current disturbances. The discrete module is used to discretize the state-space model to obtain the state-space equation. The system state of the automated surface vehicle at different times is predicted by the state-space equation. The minimum value of the system state at each time is controlled to generate the trajectory generation optimization equation. The generation module is used to set constraints to characterize the safe operation of the automated surface vehicle and a cost function to characterize the operation of the automated surface vehicle on the trajectory generation optimization equation, so as to obtain the trajectory generation optimization problem. The solution module is used to obtain the optimal solution to the trajectory generation optimization problem, and to generate the running trajectory of the automated surface vehicle based on the optimal solution.

8. A computer device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the automatic water vehicle trajectory generation method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the automatic water vehicle trajectory generation method according to any one of claims 1-6.

Images