A Virtual Potential Field-Guided Iterative Learning Path Tracking Control Method for USV

Abstract

This invention discloses an iterative learning path tracking control method for USVs based on virtual potential field guidance, comprising: establishing a nonlinear mathematical model of the USV; obtaining the gravitational model and repulsive model of the virtual potential field; obtaining the guidance model of the USV; obtaining the kinematic controller of the USV system; obtaining the dynamic controller of the USV system, and controlling the USV system. This invention's iterative learning path tracking control method for USVs based on virtual potential field guidance, by establishing the gravitational and repulsive models of the virtual potential field of the USV model, incorporates an obstacle avoidance mechanism that considers the shape of the obstacle itself into the guidance algorithm. This solves the problem that most existing adaptive control obstacle avoidance algorithms treat obstacles as point masses, making the obstacle avoidance mechanism of this invention more reliable and enhancing the autonomous collision avoidance performance of ships.

Description

Technical Field

[0001] This invention relates to the field of ship control engineering and the application technology of automated navigation equipment for ships, and in particular to a USV iterative learning path tracking control method based on virtual potential field guidance. Background Technology

[0002] Path tracking control for USVs has always been a research hotspot. Currently, scholars both domestically and internationally have conducted extensive research on USV path tracking control, addressing issues such as model uncertainties, sudden failures, and actuator saturation. However, these studies are primarily based on USV path tracking control during ocean navigation, resulting in generally low path tracking accuracy. This accuracy is suitable for USVs performing general tasks. However, with the development of USV technology, its applications in marine development and military fields are increasing, leading to more diverse performance requirements for USVs. Tasks such as precision maintenance of marine cables and underwater topographic surveying in shipping lanes demand high tracking accuracy from USVs. Existing adaptive control obstacle avoidance algorithms mostly treat obstacles as point masses, ignoring their geometry, resulting in low performance of the obstacle avoidance mechanism. Summary of the Invention

[0003] This invention provides a USV iterative learning path tracking control method based on virtual potential field guidance to overcome the above-mentioned technical problems.

[0004] To achieve the above objectives, the technical solution of the present invention is as follows:

[0005] A USV iterative learning path tracking control method based on virtual potential field guidance includes the following steps:

[0006] S1: Establish a nonlinear mathematical model for the USV;

[0007] S2: Based on the nonlinear mathematical model of the USV, obtain the gravitational model and the repulsive model of the virtual potential field.

[0008] S3: Obtain the guidance model of the USV based on the gravitational model and repulsive model of the virtual potential field.

[0009] S4: Obtain the kinematic controller of the USV system based on the guidance model of the USV;

[0010] S5: Obtain the dynamic controller of the USV system based on the kinematic controller of the USV system, and control the USV system.

[0011] Furthermore, in S1, the nonlinear mathematical model of USV is established as follows:

[0012]

[0013] In the formula: x k y represents the displacement of the unmanned surface vessel in the forward direction at the k-th iteration; k ψ represents the lateral drift displacement of the unmanned surface vessel in the k-th iteration. k This represents the heading angle of the unmanned surface vessel at the k-th iteration. | represents differentiation; |·| represents absolute value; u k v represents the velocity of the unmanned surface vessel in its forward direction at the k-th iteration. k r represents the velocity of the unmanned surface vessel in the lateral drift direction at the k-th iteration. k f represents the heading angle and angular velocity of the unmanned surface vessel at the k-th iteration. u (v) represents the uncertainty of the hull structure in the direction of travel; f v (v) represents the uncertainty of the hull structure in the lateral drift direction; f r (v) represents the uncertainty of the hull structure in the bow roll direction; m u The mass added to the hull in the direction of travel; m v The mass added to the hull in the direction of travel; m r d represents the hull's added mass in the bow direction; wu The term representing the direction of movement is related to marine environmental disturbances; d wv The term representing marine environmental disturbance in the lateral drift direction; d wr The term represents the marine environmental disturbance term in the bow roll direction; n represents the actual main engine speed of the control input, δ represents the actual rudder angle of the control input, and F... u (·) represents the unknown gain function of the propeller speed input of the actuator; T r (·) represents the unknown gain function of the actuator's rudder angle input;

[0014] in,

[0015]

[0016] In the formula: t p The fraction of the wake around the propeller is represented by ρ, and the density of seawater is represented by D. p Represents the propeller diameter; k T J represents the thrust coefficient. p The forward propulsion coefficient; x R The representation depends on the dimensional parameter a, which is the distance from the ship's center of gravity to the propeller blade along the ship's length. H The wake fraction represents the area around the rudder; x H It depends on the dimensional parameter of the distance from the ship's center of gravity to the propeller blades in the beam direction; Λ represents the aspect ratio of the rudder; A RRepresents the rudder area; L represents the length between the two vertical lines; u R v represents the forward velocity of the rudder relative to the wake swell. R This represents the relative swell speed of the rudder under the influence of the wake.

[0017] Furthermore, in S2, the gravitational model of the virtual potential field is established as follows:

[0018]

[0019] In the formula: U att (p,v) represents the potential field gravity, k p Positive design parameters for the virtual potential field representing the position component; η pg (P o ,P g ) represents the distance from the ship to the target point; η vg (P o ,P g ) represents the velocity components of the ship relative to the target point; k v Positive design parameters for the virtual potential field representing the velocity components; x r The position and attitude of the LVS in the forward direction, y r Represents the position and attitude of LVS in the lateral drift direction, u xk u represents the component of the ship's forward speed in the x-direction. yk u represents the component of the ship's forward speed in the y-direction. xr u represents the component of the LVS forward velocity in the x-direction. yr This represents the component of the LVS forward velocity in the y-direction.

[0020] Furthermore, in S2, the repulsive force model of the virtual potential field is established as follows:

[0021]

[0022] In the formula: U rep U represents the repulsive force of the potential field. d Represents the reference speed of P-DVS; l d x represents the distance between the ship and the obstacle avoidance point; b The x-coordinate of the obstacle avoidance point, y b It is the ordinate of the obstacle avoidance point.

[0023] Furthermore, in step S3, the guidance model of the USV is obtained as follows:

[0024]

[0025]

[0026] In the formula: ψr This represents the heading angle position and attitude of the LVS (Logical Virtual Ship); x d y represents the forward direction, position, and attitude of the P-DVS dynamic virtual ship in the potential field. d ψ represents the position and attitude of the P-DVS dynamic virtual ship in the lateral drift direction; d The heading angle position and attitude of P-DVS; u r θ represents the reference velocity of the LVS, r represents the turning angular velocity of the LVS; F It represents the direction of the resultant force of all gravitational and repulsive forces.

[0027] Furthermore, in step S4, the method for obtaining the kinematic controller of the USV system is as follows:

[0028] S41: The kinematic error of the kinematic controller of the USV system is obtained as follows:

[0029]

[0030] In the formula: x e,k y represents the position and attitude error of the actual ship and P-DVS in the forward direction when the number of iterations is k. e,k The position and attitude error of the actual ship and P-DVS in the lateral drift direction when the number of iterations is k. This represents the error between the actual ship's heading and the P-DVS heading when the number of iterations is k; The kinematic control law representing the ship's forward direction when the number of iterations is k; The filtering error is caused by phase shift and input noise interference in the virtual control signal of the actual ship displacement direction when passing through the filter at iteration number k; u e,k This represents the velocity error corresponding to the k-th iteration in the direction of the ship's forward movement; The filtered signal represents the virtual controller for the ship's heading error when the actual iteration number is k; The kinematic control law representing the bow rate of the ship when the number of iterations is k; The filtering error is caused by phase shift and input noise interference in the virtual control signal of the actual ship's bow turning rate when the number of iterations is k; r e,k The r represents the error corresponding to the k-th iteration of the ship's bow turning rate; d Represents the ship's reference bow turning rate; Ψ y Represents the self-designed compression parameters, where, q ψe,k The filtering error, α, represents the error caused by phase shift and input noise in the virtual control signal of the actual ship's heading error when passing through the filter after iteration number k. ψe,k The kinematic control law representing the ship's heading error when the number of iterations is k;

[0031] S42: Construct the first-order DSC filter as follows:

[0032]

[0033] In the formula: The time constant representing the direction of the actual ship displacement being greater than zero in the k-th iteration; The time constant representing the direction in which the ship's heading angle is greater than zero in the k-th iteration; The representative value is the time constant during the k-th iteration where the actual ship's heading error is greater than zero. The kinematic control law representing the ship's heading error; represent Initial value at t=0; represent Initial value at t=0; This represents the virtual control signal indicating the actual ship displacement direction during the k-th iteration. represent Initial value at t=0; represent Initial value at t=0; This is the virtual control signal for the heading angle direction of the actual ship in the kth iteration; represent Initial value at t=0; represent Initial value at t=0; This represents a virtual control signal representing the actual ship's heading error. The filtered signal represents the virtual controller of the ship's forward velocity in the k-th iteration; The filtered signal represents the virtual controller of the ship's rotating box angular velocity in the k-th iteration;

[0034] S43: Obtain the following kinematic controller information for the USV system:

[0035]

[0036]

[0037]

[0038] In the formula: The design parameters representing the positive position and attitude of the virtual controller in the forward direction of the actual ship; The design parameters representing the positive position and attitude of the virtual controller in the lateral drift direction of the actual ship; The design parameters of the virtual controller representing the heading angle of a real ship are as follows: The kinematic control law representing the direction of the ship's movement; The kinematic control law representing the ship's bow turning rate.

[0039] Furthermore, in step S5, the method for obtaining the dynamic controller of the USV system is as follows:

[0040] definition and Its derivative is as follows:

[0041]

[0042] In the formula: u e,k This represents the error in the k-th iteration of the ship's forward speed; The filtered signal represents the virtual controller signal for the forward velocity of the real ship in the k-th iteration; r e,k This represents the error of the k-th iteration of the ship's bow turning angular velocity; F represents the filtered signal of the virtual controller for the angular velocity of the real ship's rotating container in the kth iteration; u (·) represents the unknown gain function of the propeller speed input of the actuator; N2 is an intermediate parameter, N2=n|n|;

[0043] in,

[0044]

[0045] In the formula: S(v) is a radial basis function with Gaussian properties; A(u) is a function with respect to u. k The designed weight matrix; A(r) is for r k The weight matrix is ​​designed; v is a user-defined symbol; ε u (v) is the approximation error of the radial basis function neural network for the uncertainty term in the forward direction; ε r (v) is the approximation error of the radial basis function neural network for the uncertainty term in the turning direction; b u It is the norm of A(u); It is for u e,k The designed compression function, where, It is to calculate w u intermediate variables; b r =||A(r)|| F It is the norm of A(r); w r =A(r) m r e,k It is for r e,k The designed compression function includes, It is to calculate w r intermediate variables;

[0046] The dynamic controller of the USV system is obtained as follows:

[0047]

[0048]

[0049] In the formula: Dynamic control input representing the engine rotation speed; The adaptive rate representing the displacement direction; ρ u,k The representative design uses adaptive parameters for the ship's forward speed for the k-th iteration; ρ 0 u,k The representative design uses the compression parameters related to the ship's forward speed for the k-th iteration; ρ u,k-1 The z1 represents the adaptive parameters of the ship's forward speed for the (k-1)th iteration; λ represents the positive parameters of the speed adaptive design. u,k-1 zk represents the adaptive parameter of the ship's forward speed control rate for the k-th iteration; z2 represents the positive parameter of the main engine speed control rate; Φ u (·) represents the combined positive parameter in the displacement direction robust neural damping technique; Positive design parameters representing the direction of the second-order dynamic displacement of the actual ship; α δ Dynamic control inputs representing the ship's rudder angle; The adaptive rate representing the displacement direction; ρ r,k The representative design uses adaptive parameters for the ship's bow turning angular velocity for the k-th iteration; ρ 0 r,k The representative design is for the compression parameters of the ship's bow turning angular velocity for the kth iteration; ρ r,k-1 The representative design uses adaptive parameters for the ship's bow turning angular velocity for the (k-1)th iteration; λ r,k-1 z3 represents the adaptive parameter of the ship's bow turning angular velocity control rate for the k-th iteration; z4 represents the positive parameter of the rudder angle adaptive design; z5 represents the positive parameter of the rudder angle control rate design; Φ r (·) represents the combined positive parameter in the heading angle direction robust neural damping technique; The positive design parameter represents the direction of the heading angle in the second-order dynamics of the actual ship.

[0050] Beneficial effects: The USV iterative learning path tracking control method based on virtual potential field guidance of the present invention establishes an attractive force model and a repulsive force model of the virtual potential field of the USV model, and adds an obstacle avoidance mechanism that can take into account the shape of the obstacle itself into the guidance algorithm. This solves the problem that most existing adaptive control obstacle avoidance algorithms treat the obstacle as a point mass, making the obstacle avoidance mechanism of the present invention more reliable and enhancing the autonomous collision avoidance performance of the ship. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 This is a flowchart of the control method of the present invention;

[0053] Figure 2 This is a view of the wind field and wave surface under sea state 5.5 in an embodiment of the present invention;

[0054] Figure 3 This is a simulated ship path tracking trajectory in an embodiment of the present invention;

[0055] Figure 4 This is a comparison chart of the algorithm control effects in the embodiments of the present invention;

[0056] Figure 5 This is a simulation diagram of autonomous obstacle avoidance tracking in an embodiment of the present invention;

[0057] Figure 6 This is a potential field repulsion diagram in an embodiment of the present invention;

[0058] Figure 7 This is a comparison diagram of the ship's reference course and desired course in an embodiment of the present invention;

[0059] Figure 8 This is a comparison chart of control inputs for different iteration numbers in embodiments of the present invention;

[0060] Figure 9 This is a comparison chart of errors for different iteration numbers in embodiments of the present invention;

[0061] Figure 10 This is a quantitative analysis comparison chart from an embodiment of the present invention. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0063] This embodiment provides a USV iterative learning path tracking control method based on virtual potential field guidance, such as... Figure 1As shown, it includes the following steps:

[0064] S1: Establish a nonlinear mathematical model for the USV;

[0065] The nonlinear mathematical model of the USV is established as follows:

[0066]

[0067]

[0068] In the formula: x k y represents the displacement of the unmanned surface vessel in the forward direction at the k-th iteration; k ψ represents the lateral drift displacement of the unmanned surface vessel in the k-th iteration. k This represents the heading angle of the unmanned surface vessel at the k-th iteration. | represents differentiation; |·| represents absolute value; u k v represents the velocity of the unmanned surface vessel in its forward direction at the k-th iteration. k r represents the velocity of the unmanned surface vessel in the lateral drift direction at the k-th iteration. k f represents the heading angle and angular velocity of the unmanned surface vessel at the k-th iteration. u (v) represents the uncertainty of the hull structure in the direction of travel; f v (v) represents the uncertainty of the hull structure in the lateral drift direction; f r (v) represents the uncertainty of the hull structure in the bow roll direction; m u The mass added to the hull in the direction of travel; m v The mass added to the hull in the direction of travel; m r d represents the hull's added mass in the bow direction; wu The term representing the direction of movement is related to marine environmental disturbances; d wv The term representing marine environmental disturbance in the lateral drift direction; d wr The term represents the marine environmental disturbance term in the bow roll direction; n represents the actual main engine speed of the control input, δ represents the actual rudder angle of the control input, and F... u (·) represents the unknown gain function of the propeller speed input of the actuator; T r (·) represents the unknown gain function of the rudder angle input of the actuator; k = 1, 2… represents the number of iterations of the system; due to the USV characteristics, the drift velocity v is passively bounded and stable. The lateral moment of the ship caused by the drift velocity v can be canceled out by the lateral disturbance moment acting on the ship, therefore, in the controller design process, it is assumed that v≈0.

[0069] in,

[0070]

[0071] In the formula: t p The fraction of the wake around the propeller is represented by ρ, and the density of seawater is represented by D. p Represents the propeller diameter; k T J represents the thrust coefficient. p The forward propulsion coefficient; x R The representation depends on the dimensional parameter a, which is the distance from the ship's center of gravity to the propeller blade along the ship's length. H The wake fraction represents the area around the rudder; x H It depends on the dimensional parameter of the distance from the ship's center of gravity to the propeller blades in the beam direction; Λ represents the aspect ratio of the rudder; A R Represents the rudder area; L represents the length between the two vertical lines; u R v represents the forward velocity of the rudder relative to the wake swell. R This represents the relative swell speed of the rudder under the influence of the wake;

[0072] S2: Based on the nonlinear mathematical model of the USV, obtain the gravitational model and the repulsive model of the virtual potential field.

[0073] The gravitational model of the virtual potential field in the guidance based on the virtual potential field is established as follows:

[0074]

[0075] In the formula: U att (p,v) represents the potential field gravity, k p Positive design parameters for the virtual potential field representing the position component; η pg (P o ,P g ) represents the distance from the ship to the target point; η vg (P o ,P g ) represents the velocity components of the ship relative to the target point; k v Positive design parameters for the virtual potential field representing the velocity components; x r The position and attitude of the LVS in the forward direction, y r Represents the position and attitude of LVS in the lateral drift direction, u xk u represents the component of the ship's forward speed in the x-direction. yk u represents the component of the ship's forward speed in the y-direction. xr u represents the component of the LVS forward velocity in the x-direction. yr This represents the component of the LVS forward velocity in the y-direction;

[0076] The repulsive force model of the virtual potential field in the guidance based on the virtual potential field is established as follows:

[0077]

[0078] In the formula: U rep U represents the repulsive force of the potential field. d Represents the reference speed of P-DVS; l d x represents the distance between the ship and the obstacle avoidance point; b The x-coordinate of the obstacle avoidance point, y b It is the ordinate of the obstacle avoidance point;

[0079] Specifically, the coordinate system in this patent is the geodetic coordinate system. This embodiment introduces an artificial potential field by establishing a virtual potential field's gravitational and repulsive force models to increase the rationality of the guidance module and better conform to engineering practice.

[0080] S3: Based on the gravitational and repulsive models of the virtual potential field, obtain the guidance model of the USV; to achieve the expected control objective by using the gravitational force of the potential field to make the P-DVS approach the LVS.

[0081] The guidance model of the USV is obtained as follows:

[0082] The direction of the resultant force of all attractive and repulsive forces is θ. F The resulting guidance equations are (5) and (6).

[0083]

[0084]

[0085] In the formula: ψ r This represents the heading angle position and attitude of the LVS (Logical Virtual Ship); x d y represents the forward direction, position, and attitude of the P-DVS dynamic virtual ship in the potential field. d ψ represents the position and attitude of the P-DVS dynamic virtual ship in the lateral drift direction; d The heading angle position and attitude of P-DVS; u r The value represents the reference speed of the LVS, and r represents the turning angular velocity of the LVS.

[0086] S4: Obtain the kinematic controller of the USV system based on the guidance model of the USV;

[0087] Specifically, in this embodiment, guided by two virtual ships, LVS and P-DVS, the vessel can autonomously calculate the reference path and the desired path based on waypoints. The kinematic controller then guides the real vessel to approach the P-DVS.

[0088] S41: The kinematic error of the kinematic controller of the USV system is obtained as follows:

[0089]

[0090] In the formula: x e,k y represents the position and attitude error of the actual ship and P-DVS in the forward direction when the number of iterations is k. e,k The position and attitude error of the actual ship and P-DVS in the lateral drift direction when the number of iterations is k. This represents the error between the actual ship's heading and the P-DVS heading when the number of iterations is k; The kinematic control law representing the ship's forward direction when the number of iterations is k; The filtering error is caused by phase shift and input noise interference in the virtual control signal of the actual ship displacement direction when passing through the filter at iteration number k; u e,k This represents the velocity error corresponding to the k-th iteration in the direction of the ship's forward movement; The filtered signal represents the virtual controller for the ship's heading error when the actual iteration number is k; The kinematic control law representing the bow rate of the ship when the number of iterations is k; The filtering error is caused by phase shift and input noise interference in the virtual control signal of the actual ship's bow turning rate when the number of iterations is k; r e,k The r represents the error corresponding to the k-th iteration of the ship's bow turning rate; d Represents the ship's reference bow turning rate; Ψ y Represents the self-designed compression parameters. q ψe,k The filtering error, α, represents the error caused by phase shift and input noise in the virtual control signal of the actual ship's heading error when passing through the filter after iteration number k. ψe,k The kinematic control law representing the ship's heading error when the number of iterations is k;

[0091] Where, q uk q rk q ψe,k This is because DSC technology was introduced to avoid the "computational explosion" problem caused by the derivative of the kinematic controller, thus making the dynamic surface β... u ,β ψe,k ,β r Replace α u ,α ψe,k ,α r By stabilizing dynamic surface errors, and stabilizing the forward, lateral, and bow positions, better path tracking control results can be obtained, which are also more in line with engineering requirements.

[0092] S42: Construct the first-order DSC filter as follows:

[0093]

[0094] In the formula: The time constant representing the direction of the actual ship displacement being greater than zero in the k-th iteration; The time constant representing the direction in which the ship's heading angle is greater than zero in the k-th iteration; The representative value is the time constant during the k-th iteration where the actual ship's heading error is greater than zero. The kinematic control law representing the ship's heading error; represent Initial value at t=0; represent Initial value at t=0; This represents the virtual control signal indicating the actual ship displacement direction during the k-th iteration. Represents β rk Initial value at t=0; represent Initial value at t=0; This is the virtual control signal for the heading angle direction of the actual ship in the kth iteration; represent Initial value at t=0; represent Initial value at t=0; This represents a virtual control signal representing the actual ship's heading error.

[0095] S43: The kinematic controller of the USV system is obtained through the backstepping method as follows:

[0096]

[0097] in The design parameters representing the positive position and attitude of the virtual controller in the forward direction of the actual ship; The design parameters representing the positive position and attitude of the virtual controller in the lateral drift direction of the actual ship; The design parameters of the virtual controller representing the heading angle of a real ship are as follows: The kinematic control law representing the direction of the ship's movement; The kinematic control law representing the rate of bow turning of a ship;

[0098] S5: Obtain the dynamic controller of the USV system based on the kinematic controller of the USV system;

[0099] The dynamic error of the dynamic controller of the USV system is obtained as follows:

[0100] definition and The derivative is then taken as follows:

[0101]

[0102] In the formula: u e,kThis represents the error in the k-th iteration of the ship's forward speed; The filtered signal represents the virtual controller signal for the forward velocity of the real ship in the k-th iteration; r e,k This represents the error of the k-th iteration of the ship's bow turning angular velocity; F represents the filtered signal of the virtual controller for the angular velocity of the real ship's rotating container in the kth iteration; u (·) represents the unknown gain function of the propeller speed input of the actuator; N2 is an intermediate parameter, N2=n|n|;

[0103]

[0104] In the formula: S(v) is a radial basis function with Gaussian properties; A(u) is a function with respect to u. k The designed weight matrix; A(r) is for r k The weight matrix is ​​designed; v is a user-defined symbol; ε u (v) is the approximation error of the radial basis function neural network for the uncertainty term in the forward direction; ε r (v) is the approximation error of the radial basis function neural network for the uncertainty term in the turning direction; β v β r b u It is the norm of A(u); It is to calculate w u intermediate variables; It is for u e,k The designed compression function;

[0105] b r =||A(r)|| F It is the norm of A(r); It is to calculate w r intermediate variables; w r =A(r) m r e,k It is for r e,k The designed compression function; It is a calculation intermediate variables; It is to calculate α δ intermediate variables; It is design compression function, Is it designing α δ The compression function is obtained by introducing radial basis function neural networks (RBF-NNs) to approximate the unknown function, and the control rates are (11) and (12).

[0106]

[0107]

[0108] In the formula: Dynamic control input representing the engine rotation speed; The adaptive rate representing the displacement direction; ρ u,k The representative design uses adaptive parameters for the ship's forward speed for the k-th iteration; ρ 0 u,k The representative design uses the compression parameters related to the ship's forward speed for the k-th iteration; ρ u,k-1 The z1 represents the adaptive parameters of the ship's forward speed for the (k-1)th iteration; λ represents the positive parameters of the speed adaptive design. u,k-1 zk represents the adaptive parameter of the ship's forward speed control rate for the k-th iteration; z2 represents the positive parameter of the main engine speed control rate; Φ u (·) represents the combined positive parameter in the displacement direction robust neural damping technique; Positive design parameters representing the direction of the second-order dynamic displacement of the actual ship; α δ Dynamic control inputs representing the ship's rudder angle; The adaptive rate representing the displacement direction; ρ r,k The representative design uses adaptive parameters for the ship's bow turning angular velocity for the k-th iteration; ρ 0 r,k The representative design is for the compression parameters of the ship's bow turning angular velocity for the kth iteration; ρ r,k-1 The representative design uses adaptive parameters for the ship's bow turning angular velocity for the (k-1)th iteration; λ r,k-1 z3 represents the adaptive parameter of the ship's bow turning angular velocity control rate for the k-th iteration; z4 represents the positive parameter of the rudder angle adaptive design; z5 represents the positive parameter of the rudder angle control rate design; Φ r (·) represents the combined positive parameter in the heading angle direction robust neural damping technique; Positive design parameters representing the direction of the heading angle in the second-order dynamics of a real ship;

[0109] in,

[0110] This embodiment controls the USV using a dynamic controller designed based on the iterative control concept described in this application. This controller not only exhibits strong robustness but also increases control precision by increasing the number of iterations, thus achieving superior control performance—a requirement in current engineering practice. (The number of iterations is user-selectable and can be set during programming. In engineering practice, operators can pre-set the number of iterations to balance computational load and control precision.)

[0111] To verify the effectiveness of the control method proposed in this invention, computer simulation experiments were conducted using MATLAB. A 38m boat was used as the controlled object (ship parameters and potential field parameters are referenced in the literature), and simulations were performed in a simulated marine environment (e.g., Figure 2 The initial state of the ship is set as [x(0),y(0),ψ(0),u(0),v(0),r(0)]=[0m,150m,0deg,,0deg,6m / s,0m / s].

[0112] Figures 3-10 The simulation results of the proposed algorithm and the comparison algorithm under simulated marine environmental disturbances are presented. Figure 3 The reference trajectory generated by waypoint-based LVS guidance and the tracking trajectory of two algorithms are described. Figure 4 A comparison chart of the control results of the two algorithms is shown. Figure 5 This represents the path tracking results of different iteration numbers in the iterative algorithm under the potential field-based DVS guidance method. Figure 6 , 7 The diagram shows the effect of the repulsive potential field in the guidance algorithm. Figure 8 Figure 9 The graph shows a comparison of errors and control inputs at different iteration numbers. Figure 10 The graph represents the quantitative analysis results of different algorithms.

[0113] pass Figure 3-4 It can be seen that this algorithm has a significant advantage in path tracking performance. From Figure 5-7 As can be seen from the above, the algorithm proposed in this invention can improve the controller performance by increasing the number of iterations. Simultaneously, the obstacle avoidance method proposed in this invention has excellent effectiveness, automatically planning a safer route when the ship's distance from shore is too small. Comparisons in 8-9 show that the algorithm proposed in this invention can effectively improve chattering, resulting in a smoother input control rate. From... Figure 10 It can be observed that increasing the number of iterations can significantly reduce the tracking error of path tracing, and the increased computational cost is acceptable.

[0114] This embodiment has the following main advantages in marine engineering applications:

[0115] To improve guidance performance, this embodiment proposes a DVS guidance technology based on an artificial potential field, specifically a USV iterative learning path tracking control method based on virtual potential field guidance. This method can enhance the autonomous collision avoidance performance of ships and optimize the desired path. In the proposed potential field-based guidance strategy, LVS is used to plan a smoother reference path based on waypoints, thereby eliminating abnormalities such as chattering in actuator signals at turning points. Based on the planned reference path, P-DVS plans a desired path according to the gravitational and repulsive forces acting on the actual ship. The actual ship can navigate more safely on this path.

[0116] Based on maritime practice, this embodiment proposes an iterative learning control scheme with main engine speed and rudder angle as control inputs. This scheme features high precision, enabling ships to complete high-precision path tracking tasks under acceptable computational loads. It achieves path tracking control for ships in high-precision scenarios and is suitable for engineering tasks such as waterway inspection and precision maintenance of submarine cables. Simulation experiments have verified that this scheme effectively improves control accuracy and is suitable for ship path control in high-precision scenarios. Furthermore, by incorporating techniques such as DSC and RBF-NNs into the iterative learning control, the computational load of the algorithm is effectively reduced.

[0117] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A USV iterative learning path tracking control method based on virtual potential field guidance, characterized in that, The method comprises the following steps: S1: establishing a nonlinear mathematical model of the USV; S2: obtaining an attractive model of a virtual potential field and a repulsive model of the virtual potential field according to the nonlinear mathematical model of the USV; S3: obtaining a guidance model of the USV according to the attractive model of the virtual potential field and the repulsive model of the virtual potential field; S4: obtaining a kinematic controller of the USV system according to the guidance model of the USV; In the S4, the method for obtaining the kinematic controller of the USV system is as follows: S41: obtaining a kinematic error of the kinematic controller of the USV system as follows: (7) wherein: represents the position and attitude error of the real ship and the P-DVS in the forward direction at the kth iteration; represents the position and attitude error of the real ship and the P-DVS in the cross drift direction at the kth iteration; represents the error of the real ship and the P-DVS in the heading direction at the kth iteration; represents the kinematic control rate of the ship in the forward direction at the kth iteration; represents the filter error of the virtual control signal of the real ship displacement direction at the kth iteration, which is disturbed by the phase shift and the input noise when passing through the filter; represents the error of the ship in the forward direction corresponding to the kth iteration; represents the filter signal of the virtual controller of the real ship heading error at the kth iteration; represents the kinematic control rate of the ship in the turning angle rate at the kth iteration; represents the filter error of the virtual control signal of the real ship turning angle rate at the kth iteration, which is disturbed by the phase shift and the input noise when passing through the filter; represents the error of the ship in the turning angle rate corresponding to the kth iteration; represents the reference turning angle rate of the ship; represents the designed compression parameter, wherein, ; represents the filter error of the virtual control signal of the real ship heading error at the kth iteration, which is disturbed by the phase shift and the input noise when passing through the filter; represents the kinematic control rate of the ship in the heading error at the kth iteration; S42: establishing a first-order filter of the DSC as follows: (8) wherein: represents the time constant of the real ship displacement direction greater than zero at the kth iteration; represents the time constant of the real ship heading angle greater than zero at the kth iteration; represents the time constant of the real ship heading error greater than zero at the kth iteration; represents the kinematic control rate of the ship heading error; represents at the initial value of the time; represents at the initial value of the time; represents at the initial value of the time; represents the virtual control signal of the kth iteration of the real ship displacement direction; represents at the initial value of the time; represents at the initial value of the time; represents at the initial value of the time; represents the virtual control signal of the kth iteration of the real ship heading angle direction; represents at the initial value of the time; represents at the initial value of the time; represents at the initial value of the time; represents the virtual control signal of the real ship heading error; represents the filter signal of the kth iteration of the real ship forward direction speed virtual controller; represents the filter signal of the kth iteration of the real ship turning box angle speed virtual controller; S43: obtaining the kinematic controller of the USV system as follows: (9) In the formulae: represents the positive design parameter of the virtual controller of the position and posture of the forward direction of the actual ship; represents the positive design parameter of the virtual controller of the position and posture of the transverse direction of the actual ship; represents the positive design parameter of the virtual controller of the heading angle direction of the actual ship; represents the kinematic control rate of the forward direction of the ship; represents the kinematic control rate of the turning angle rate of the ship; S5: obtaining a dynamic controller of the USV system according to the kinematic controller of the USV system, and controlling the USV system.

2. The USV iterative learning path tracking control method based on virtual potential field guidance according to claim 1, characterized in that, In the S1, the nonlinear mathematical model of the USV is established as follows: (1) wherein: represents the displacement of the unmanned surface vehicle in the forward direction of the vehicle at the kth iteration; represents the displacement of the unmanned surface vehicle in the cross drift direction of the vehicle at the kth iteration; represents the heading angle of the unmanned surface vehicle at the kth iteration; represents the derivative operation; represents the absolute value operation; represents the velocity of the unmanned surface vehicle in the forward direction at the kth iteration; represents the velocity of the unmanned surface vehicle in the cross drift direction at the kth iteration; represents the angular velocity of the heading angle of the unmanned surface vehicle at the kth iteration; represents the forward direction hull structure uncertainty term; represents the cross drift direction hull structure uncertainty term; represents the yaw direction hull structure uncertainty term; represents the forward direction hull added mass; represents the forward direction hull added mass; represents the yaw direction hull added mass; represents the forward direction ocean environment disturbance term; represents the cross drift direction ocean environment disturbance term; represents the yaw direction ocean environment disturbance term; represents the actual control input of the main engine speed, represents the actual control input of the rudder angle, represents the unknown gain function of the propeller speed input of the actuator; represents the unknown gain function of the rudder angle input of the actuator; Wherein, (2) wherein: represents the fraction of the wake around the propeller, represents the density of the sea water; represents the diameter of the propeller; represents the thrust coefficient; represents the advance coefficient of the propeller; represents a dimension parameter depending on the distance of the center of gravity of the ship to the propeller blade in the direction of the ship length, represents the fraction of the wake around the rudder; is a dimension parameter depending on the distance of the center of gravity of the ship to the propeller blade in the direction of the ship width; represents the aspect ratio of the rudder; represents the area of the rudder; represents the length between two perpendiculars; represents the advance velocity of the rudder relative to the wave under the influence of the wake, represents the sway velocity of the rudder relative to the wave under the influence of the wake.

3. The USV iterative learning path tracking control method based on virtual potential field guidance according to claim 1, characterized in that, In the S2, the attractive model of the virtual potential field is established as follows: (3) wherein: represents the potential field attraction, represents the positive design parameter of the virtual potential field of the position component; represents the distance of the ship to the target point; represents the velocity component of the ship to the target point; represents the positive design parameter of the virtual potential field of the velocity component; represents the LVS forward direction position attitude, represents the LVS lateral drift direction position attitude, represents the component of the ship forward velocity in the direction, represents the component of the ship forward velocity in the direction, represents the component of the LVS forward velocity in the direction, represents the component of the LVS forward velocity in the direction.

4. The USV iterative learning path tracking control method based on virtual potential field guidance according to claim 1, characterized in that, In the S2, the repulsive model of the virtual potential field is established as follows: (4) In the formula: represents the repulsive force of the potential field, represents the reference speed of P-DVS; represents the distance between the ship and the obstacle point; represents the horizontal coordinate of the obstacle point, is the vertical coordinate of the obstacle point.

5. The USV iterative learning path tracking control method based on virtual potential field guidance according to claim 1, characterized in that, In the S3, the guidance model of the USV is obtained as follows: (5) (6) wherein: represents the heading angle position attitude of the LVS, i.e. the logical virtual ship; represents the forward direction position attitude of the P-DVS, i.e. the potential field dynamic virtual ship, represents the lateral direction position attitude of the P-DVS, i.e. the potential field dynamic virtual ship; represents the heading angle position attitude of the P-DVS; represents the reference velocity of the LVS, represents the yaw rate of the LVS; represents the resultant direction of all attractive and repulsive forces.

6. The USV iterative learning path tracking control method based on virtual potential field guidance according to claim 1, characterized in that, In the S5, the method for obtaining the dynamic controller of the USV system is as follows: Definitions and The derivative of which is taken as follows: (10) wherein: represents the error of the kth iteration of the ship's forward speed; represents the filtered signal of the kth iteration of the real ship forward direction speed virtual controller; represents the error of the kth iteration of the ship's turning angle speed; represents the filtered signal of the kth iteration of the real ship turning angle speed virtual controller; represents the unknown gain function of the propeller speed input of the actuator; is an intermediate parameter, ; Wherein, ， wherein: is a radial basis function with Gaussian function property; is a weight matrix designed for ; is a weight matrix designed for ; is a self-defined symbol; is the approximation error of the radial basis neural network for the uncertainty term of the forward direction; is the approximation error of the radial basis neural network for the uncertainty term of the turning direction; is the norm of ; is a compression function designed for , wherein is an intermediate variable for calculating ; is the norm of ; is a compression function designed for , wherein is an intermediate variable for calculating ; The dynamic controller of the USV system is obtained as follows: (11) (12) wherein: represents the dynamics control input of the main engine rotation speed; represents the adaptive rate in the displacement direction; represents the adaptive parameter designed for the kth iteration regarding the ship forward speed; represents the compression parameter designed for the kth iteration regarding the ship forward speed; represents the adaptive parameter designed for the k-1th iteration regarding the ship forward speed; represents the positive parameter of the speed adaptive design; represents the adaptive parameter designed for the kth iteration regarding the ship forward speed control rate; represents the positive parameter of the design of the main engine rotation speed control rate; represents the positive parameter incorporated in the displacement direction robust neural damping technique; represents the positive parameter of the design of the real ship second order dynamics displacement direction; represents the dynamics control input of the ship rudder angle; represents the adaptive rate in the displacement direction; represents the adaptive parameter designed for the kth iteration regarding the ship yaw rate; represents the compression parameter designed for the kth iteration regarding the ship yaw rate; represents the adaptive parameter designed for the k-1th iteration regarding the ship yaw rate; represents the adaptive parameter designed for the kth iteration regarding the ship yaw rate control rate; represents the positive parameter of the yaw adaptive design; represents the positive parameter of the design of the yaw control rate; represents the positive parameter incorporated in the heading angle direction robust neural damping technique; represents the positive parameter of the design of the real ship second order dynamics heading angle direction.

Images