A control method for target encirclement by underactuated unmanned vehicle with predefined time
Through the predefined time control method, combined with the Lyapunov function and the adaptive RBF neural network, the problems of actuator failure and input saturation in the unmanned boat target encirclement are solved, and the system stability and error convergence in complex environments are achieved, which expands the The scope of application of the controller.
Patent Information
- Application Number
- CN202410476510.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-19
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-04-19
AI Technical Summary
The existing unmanned boat target encirclement control technology fails to effectively consider actuator failure and input saturation, and the application range of controllers based on finite time theory is limited, making it difficult to maintain system stability and control of error convergence time in complex environments.
Using the predefined time control method, the predefined time assisted power system is constructed by establishing a kinematic and dynamic model of three degrees of freedom under-driven unmanned boats, decoupling the model, and constructing the Lyapunov function and predefined time virtual control law, and considering the execution of the predefined time assisted power system. The device failure and input saturation are designed, and the nonlinear term is approximate to the system stability and error convergence are achieved.
Improves the scope of application and stability of the controller, can achieve error convergence within a predefined time, is not affected by the initial value of the system, enhances tolerance for external disturbances and actuator failures, and ensures that the unmanned boat is surrounded by targets. stability and control performance in.
Smart Images

Figure CN118426469B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of underactuated unmanned boat control, and in particular to a control method for encircling a target by an underactuated unmanned boat within a predefined time. Background Art
[0002] On Earth, ocean area accounts for three quarters of the Earth's total area. Unmanned boats are widely used in ocean exploration and development due to their low cost, high intelligence, and flexible maneuverability.
[0003] According to different motion scenarios, the control of underactuated unmanned boats is mainly divided into path following, trajectory tracking, target tracking, target encirclement, etc. Among them, the target encirclement control technology plays an outstanding role in the detection, identification, capture and tracking of targets. Typical applications include offshore oil spill blockade, marine life or enemy ship capture, etc., and it has become one of the research hotspots that has attracted much attention in recent years. At the same time, underactuated unmanned boats are widely used because of their simple structure and lower cost compared to fully driven unmanned boats. However, due to the high coupling of the model and the lack of control in the lateral direction, the design of the controller of the underactuated unmanned boat becomes relatively complex and challenging.
[0004] In the actual operating environment, underactuated unmanned boats are often affected by complex external environmental disturbances such as rivers, lakes or oceans, including but not limited to water flow, waves, wind, tidal currents, plankton and marine meteorology. In addition, due to the high nonlinearity and coupling of underactuated unmanned boat systems, their modeling also has certain uncertainties. These factors will affect the navigation trajectory, speed, attitude and stability of the unmanned boat, making the design and implementation of the controller more difficult.
[0005] During the operation of the underactuated unmanned boat, due to the influence of the external environment or factors such as actuator wear and aging, actuator failure may occur, resulting in the decline of the attitude stability of the underactuated unmanned boat, failure of heading and speed control, and decline of system stability. In addition to actuator failure, when the input signal of the underactuated unmanned boat system exceeds its tolerable range, in order to prevent the actuator from being damaged and the system from losing control, input saturation will occur. Therefore, the control system may not be able to produce sufficient control, resulting in the system being unable to adjust as expected, which will lead to a decline in control performance and poor stability. Therefore, while considering external environmental disturbances, the actuator failure and input saturation problems of the underactuated unmanned boat should also be considered in the design of the controller.
[0006] In the existing unmanned boat target encirclement control, the convergence form of the error is mostly based on the finite time theory. However, the finite time theory is sensitive to the initial conditions of the system, and the difference between the initial state and the desired state may affect the performance and convergence speed of the controller. Therefore, the fixed time theory was created. Unlike the finite time theory, the fixed time theory is insensitive to the initial conditions, and the convergence time of the error can be changed by adjusting the parameters. However, the controller designed based on the fixed time theory needs to adjust multiple parameters to make the time for the error convergence in the closed-loop system within an ideal range, so the scope of application is limited. Compared with the finite time and fixed time controllers, the controller designed based on the predetermined time theory is not only insensitive to the initial conditions, but also can directly set the upper limit of the error convergence time, so the scope of application is wider.
[0007] At present, there have been many studies on the target encirclement control of unmanned boats, and many effective methods have been proposed. However, they have not considered the situation of actuator failure and input saturation during the target encirclement process of the unmanned boat, and most controllers are designed based on finite time theory, which limits the application scope of the controller. Summary of the invention
[0008] In view of the above-mentioned deficiencies in the prior art, a control method for target encirclement by an under-actuated unmanned boat with a predefined time is provided by the present invention, which solves the problems that external disturbances, modeling uncertainty and actuator failure affect the system stability and control effect when the under-actuated unmanned boat is encircling the target, the error convergence time cannot be directly controlled, and the actuator failure and input saturation are not considered simultaneously during the encirclement process.
[0009] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:
[0010] A control method for encircling a target with an underactuated unmanned vehicle for a predefined time is provided, which comprises the following steps:
[0011] S1. Obtain the motion parameters of the unmanned boat at the previous moment, and establish the kinematic and dynamic models of the three-degree-of-freedom under-actuated unmanned boat; the unmanned boat includes the surrounding unmanned boat and the target unmanned boat, and both the surrounding unmanned boat and the target unmanned boat are under-actuated unmanned boats;
[0012] S2, decoupling the kinematic and dynamic models through coordinate transformation to obtain a decoupled model;
[0013] S3. Based on the decoupled model, the relative sight angle, relative distance, and surrounding angle of the surrounding unmanned boat are obtained between the surrounding unmanned boat and the target unmanned boat, and the surrounding angle error and surrounding radius error of the surrounding unmanned boat are set;
[0014] S4, constructing a Lyapunov function of an encircling angle error and an encircling radius error surrounding the unmanned boat, and obtaining a predefined time virtual control law;
[0015] S5, obtaining a filter output of a predefined time virtual control law by constructing a predefined time filter; and setting a longitudinal velocity error and a bow angular velocity error surrounding the unmanned boat according to the filter output of the predefined time virtual control law;
[0016] S6, constructing a predefined time auxiliary power system based on the saturated control input at the previous moment, the original control input at the previous moment, the longitudinal velocity error surrounding the unmanned boat, and the bow angular velocity error, and obtaining the state of the predefined time auxiliary power system;
[0017] S7, obtaining a comprehensive uncertainty term according to the motion parameters of the unmanned boat at the previous moment, the predefined time virtual control law at the current moment, and the longitudinal velocity error and the bow angular velocity error of the surrounding unmanned boat at the current moment;
[0018] S8, constructing a predefined time controller according to the filtered output of the predefined time virtual control law at the current moment, the comprehensive uncertainty term and the state of the auxiliary power system at the predefined time, and obtaining the original control input corresponding to the current moment;
[0019] S9. Obtain the saturated control input corresponding to the current moment according to the original control input corresponding to the current moment; add the actuator fault parameter to the current saturated control input to obtain the control input after the actuator fault and input saturation, that is, obtain the actual control input at the current moment.
[0020] The beneficial effects of the present invention are:
[0021] 1. While introducing the actuator failure and input saturation models, the present invention constructs an auxiliary power system to compensate for the impact of input saturation and estimates the comprehensive uncertainty caused by actuator failure, modeling uncertainty and external environmental disturbance, thereby improving the applicability and stability of the controller.
[0022] 2. The present invention is based on the predefined time theory, so that the convergence time of the error can be set in advance and is not affected by the initial value of the system, further improving the application range and stability of the controller. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 is a schematic diagram of the process of this method;
[0024] Figure 2 is the control flow chart of this method;
[0025] Figure 3A schematic diagram of a surrounding trajectory surrounding an unmanned boat in an embodiment;
[0026] Figure 4 Schematic diagram of surround angle tracking error in the embodiment;
[0027] Figure 5 Schematic diagram of enclosing radius tracking error in the embodiment;
[0028] Figure 6 Input control τ in the embodiment r Schematic diagram;
[0029] Figure 7 Input control τ in the embodiment u Schematic diagram;
[0030] Figure 8 The adaptive RBF neural network weight estimation bi-norm in the embodiment Schematic diagram of value selection;
[0031] Fig. 9 The adaptive RBF neural network weight estimation bi-norm in the embodiment Schematic diagram of value selection. DETAILED DESCRIPTION
[0032] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.
[0033] like Figure 1 and Figure 2 As shown, the control method for encircling a target by an underactuated unmanned boat for a predefined time includes the following steps:
[0034] S1. Obtain the motion parameters of the unmanned boat at the previous moment, and establish the kinematic and dynamic models of the three-degree-of-freedom under-actuated unmanned boat; the unmanned boat includes the surrounding unmanned boat and the target unmanned boat, and both the surrounding unmanned boat and the target unmanned boat are under-actuated unmanned boats;
[0035] S2, decoupling the kinematic and dynamic models through coordinate transformation to obtain a decoupled model;
[0036] S3. Based on the decoupled model, the relative sight angle, relative distance, and surrounding angle of the surrounding unmanned boat are obtained between the surrounding unmanned boat and the target unmanned boat, and the surrounding angle error and surrounding radius error of the surrounding unmanned boat are set;
[0037] S4, constructing a Lyapunov function of an encircling angle error and an encircling radius error surrounding the unmanned boat, and obtaining a predefined time virtual control law;
[0038] S5, obtaining a filter output of a predefined time virtual control law by constructing a predefined time filter; and setting a longitudinal velocity error and a bow angular velocity error surrounding the unmanned boat according to the filter output of the predefined time virtual control law;
[0039] S6, constructing a predefined time auxiliary power system based on the saturated control input at the previous moment, the original control input at the previous moment, the longitudinal velocity error surrounding the unmanned boat, and the bow angular velocity error, and obtaining the state of the predefined time auxiliary power system;
[0040] S7, obtaining a comprehensive uncertainty term according to the motion parameters of the unmanned boat at the previous moment, the predefined time virtual control law at the current moment, and the longitudinal velocity error and the bow angular velocity error of the surrounding unmanned boat at the current moment;
[0041] S8, constructing a predefined time controller according to the filtered output of the predefined time virtual control law at the current moment, the comprehensive uncertainty term and the state of the auxiliary power system at the predefined time, and obtaining the original control input corresponding to the current moment;
[0042] S9. Obtain the saturated control input corresponding to the current moment according to the original control input corresponding to the current moment; add the actuator fault parameter to the current saturated control input to obtain the control input after the actuator fault and input saturation, that is, obtain the actual control input at the current moment.
[0043] In one embodiment of the present invention, the time theory is predefined:
[0044] Consider a nonlinear system Define a continuous positive definite function V(x) and design parameters 0<γ<1, T p >0,0<Ξ<∞ satisfies the following conditions:
[0045]
[0046] System is stable for a practically predefined time, and gives the convergence region as follows:
[0047]
[0048] Where T' p Represents the set convergence time, satisfying T max Represents the upper limit of the convergence time, π is radians, and t represents time.
[0049] Adaptive RBF Neural Network:
[0050] The adaptive RBF neural network can approximate any continuous nonlinear function, and its approximation result can be expressed as follows:
[0051]
[0052] in Represents the approximation error of the adaptive RBF neural network; and L>1 represents the number of neurons; is the ideal weight, which is expressed as follows:
[0053]
[0054] in represents the estimated weight of W; Represents the Gaussian basis function vector, which is expressed as follows:
[0055]
[0056] where b l and c l Represent the width vector and center vector of the Gaussian basis function respectively.
[0057] Assumption 1: The approximation error and ideal weights of the neural network are bounded, that is, in and Is a normal number.
[0058] Related proof lemma:
[0059] Lemma 1. For any n>0, p≥0 and h>0, the following inequality holds:
[0060]
[0061] For any n>1, p>0, h≤p, The following inequality holds:
[0062] h n -p n ≤(ph) n (7)
[0063] Lemma 2. For any The following inequality holds:
[0064]
[0065] Lemma 3. For any and The following inequality holds:
[0066]
[0067] where ∈=0.2785 satisfies ∈=e -(∈+1) , e is a natural constant.
[0068] In this embodiment, the kinematic and dynamic models of the three-degree-of-freedom underactuated unmanned vehicle can be described as follows:
[0069]
[0070] In order to solve the strong coupling problem caused by the off-diagonal terms in the mass inertia matrix, the coordinate transformation is introduced as follows:
[0071]
[0072] Substituting (11) into (10) yields:
[0073]
[0074] Therefore, the underactuated unmanned boat model can be transformed into:
[0075]
[0076]
[0077] in (x, y) is the position coordinate of the underactuated unmanned vehicle; is the yaw angle of the underactuated unmanned vehicle, [·] T represents the transpose of a matrix, represents the first derivative of η; Represents the rotation matrix between the earth coordinate system and the ship coordinate system; M represents the mass inertia matrix, m 11 =mX du , m 22 =mY dv , m 23 =mx g -Y dr , m 32 =mx g -Y dr , m 33 =I z -N dr , m is the mass of the underactuated unmanned boat, x g represents the longitudinal position of the center of gravity of the underactuated unmanned vehicle, X du Represents the acceleration coefficient of the longitudinal force about the x-axis direction, Y dv Represents the acceleration coefficient of the lateral force moving along the y-axis, Ydr Represents the acceleration coefficient of the lateral component of force about the z-axis, N dr Represents the acceleration coefficient of the yaw moment about the z-axis, I z represents the moment of inertia of the underactuated unmanned vehicle about the z-axis; υ=[u,v,r] T , u, v and r represent the longitudinal velocity, sway velocity and bow angular velocity of the underactuated unmanned vehicle respectively; is the first-order derivative of υ; C(υ) represents the Coriolis force centripetal matrix; D(υ) represents the hydrodynamic damping coefficient matrix, X u The velocity coefficient of the longitudinal force moving along the x-axis direction, Represents the second-order velocity coefficient of the longitudinal component moving about the x-axis, Represents the third-order velocity coefficient of the longitudinal component moving about the x-axis, Y v represents the velocity coefficient of the lateral force moving about the y-axis, Represents the second-order velocity coefficient of the lateral force moving along the y-axis, Y rv Represents the coupling coefficient of the lateral force about the rotation in the z-axis direction and the motion in the y-axis direction, Y vr Represents the coupling coefficient of the lateral force about the motion in the y-axis direction and the rotation in the z-axis direction, Represents the second-order velocity coefficient of the lateral component of force rotating about the z-axis, N v Represents the velocity coefficient of the yaw moment moving in the y-axis direction, Represents the second-order velocity coefficient of the yaw moment about the y-axis direction, N rv Represents the coupling coefficient of the yaw moment about the rotation in the z-axis direction and the motion in the y-axis direction, N vr Represents the coupling coefficient of the yaw moment about the motion in the y-axis direction and the rotation in the z-axis direction, N r Represents the coupling coefficient of the yaw moment about the motion in the y-axis direction and the rotation in the z-axis direction, Represents the second-order velocity coefficient of the yaw moment about the z-axis direction; τ d =[τ Du ,τ Dv ,τ Dr ] T , τ d represents the unknown external environment disturbance, τ Du , τ Dv and τ Dr They represent the unknown external environmental disturbances corresponding to the longitudinal velocity, the unknown external environmental disturbances corresponding to the sway velocity, and the unknown external environmental disturbances corresponding to the yaw angular velocity, respectively; τ f=[τ fu ,0,τ fr ] T , τ f represents the control input, τ fu represents the control input corresponding to the longitudinal velocity; τ fr represents the control input corresponding to the bow angular velocity; the motion parameters of the unmanned boat at the last moment include x, y, u, v and r. express The first derivative of Indicates the position coordinates of the surrounding unmanned boat; Indicates the sway speed of the surrounding unmanned boat; express The first derivative of; f = [f1,f2,f3] T ; d = [d u ,d v ,d r ] T ; τ'=[ξ u τ u ,0,ξ r τ r ] T ; τ m =[ξ u (τ fu -τ u ),0,ξ r (τ fr -τ r )] T , τ m is the control input affected by actuator fault and input saturation; τ u is the original control input corresponding to the longitudinal velocity direction, τ r is the original control input corresponding to the yaw angular velocity direction; χ is the intermediate parameter.
[0078] Assumption 2: External environmental disturbance d = [d u ,d v ,d r ] T Unknown but bounded.
[0079] The relative sight angle between the surrounding unmanned boat and the target unmanned boat:
[0080]
[0081] The relative distance between the surrounding unmanned boat and the target unmanned boat:
[0082]
[0083] The angle between the longitudinal direction (bow direction) of the encircling unmanned boat and the vertical line of sight is the encirclement angle:
[0084]
[0085] Set the wrapping radius error and wrapping angle error:
[0086]
[0087] where atan2(·) is the inverse tangent function; (x t ,y t ) is the position coordinate of the target unmanned boat. d is the expected orbit radius; η d For the desired wrap angle, the design is as follows:
[0088]
[0089] Where Δ>0 is the forward distance based on the line-of-sight guidance method, For x t The first derivative of for y t The first derivative of .
[0090] Construct a Lyapunov function about the surround angle error and surround radius error surrounding the unmanned boat, so as to design the predefined time virtual control rate. The constructed Lyapunov function is as follows:
[0091]
[0092] Taking the derivative of (21) we can get:
[0093]
[0094] The virtual control law is designed according to (22) as follows:
[0095]
[0096] where α u Represents the virtual control law in the longitudinal velocity direction, α r represents the virtual control law for the yaw angular velocity direction, γ is the predefined time constant, 0<γ<1, T p is the error convergence time, T p >0; k1 and k2 are positive constants; sig(·) t =sign(·)|·| t , sign(·) is the sign function, and the superscript t is a constant.
[0097] In this embodiment, in order to solve the problem of "differential explosion" in the derivation of the virtual control law, a predefined time filter is constructed as follows:
[0098]
[0099] in and All are filtered outputs of predefined time virtual control laws; for The first derivative of for The first derivative of .
[0100] Assumption 3: The derivative of the input signal of the predefined time filter is the derivative of the virtual control law is continuous and bounded, that is Among them A i is an unknown positive constant.
[0101] From (24), the longitudinal velocity error, yaw angular velocity error and filter error are set as follows:
[0102]
[0103] where u e is the longitudinal velocity error of the surrounding unmanned boat, r e is the bow angular velocity error of the surrounding unmanned boat; e fu and e fr All of them are filtering errors surrounding the unmanned boat.
[0104] In this embodiment, the problem of input saturation is solved by constructing a predefined time auxiliary power system. The form of input saturation is as follows:
[0105]
[0106] where τ i is the original control input for the design, and t i are the upper and lower bounds of the control input, τ si is the saturated control input.
[0107] The predefined time auxiliary power system constructed in this embodiment is as follows:
[0108]
[0109] where Λ i Indicates the state of the auxiliary power system at a predefined time; For i The first derivative of k Λi is a constant greater than 0; represents the failure caused by device aging and wear, Δτ i =τ si -τ i , τ si represents the saturated control input; τ i Represents the original control input at the previous moment; μ is a positive constant greater than 0.
[0110] In this embodiment, a Lyapunov function of the longitudinal velocity error and the bow angular velocity error surrounding the unmanned boat is constructed, and the actuator failure is added to design a predefined time controller.
[0111] First, the actuator fault model is introduced as follows:
[0112]
[0113] where τ fi is the control input after actuator failure and input saturation, It is a failure caused by aging, wear and tear of the device. It is an offset fault caused by installation error, displacement, and eccentricity. This means that the actuator cannot work properly and is in a fault state. This indicates that the actuator has a bias fault; when and This indicates that the actuator is working normally.
[0114] From (25), we can see that:
[0115]
[0116] Taking the derivative of (25) and substituting it into systems (14) and (15), we can obtain:
[0117]
[0118]
[0119] The Lyapunov function is selected as follows:
[0120]
[0121] Taking the derivative of (32) and substituting (23), (28), (29) and (30) into it, we obtain:
[0122]
[0123] in and represents the comprehensive uncertainty term caused by modeling uncertainty, external environmental disturbance, and partial actuator failure. From (33), the control law can be obtained as:
[0124]
[0125] Where k3 and k4 are both positive constants; u and Λ r Indicates the state of the auxiliary power system at a predefined time; k Λu and k Λr is a constant greater than 0.
[0126] For the comprehensive uncertainties F1 and F2, an adaptive RBF neural network is used to approximate them, which can be rewritten as follows:
[0127]
[0128] where x1=[u,α u ] T ,x2=[r,α r ] T is the input of the adaptive RBF neural network, then the approximation function of the adaptive RBF neural network is:
[0129]
[0130] Set the weight estimation error to:
[0131]
[0132] set up is the upper bound of the approximation error of the adaptive RBF neural network, satisfy The approximation error It can be expressed as:
[0133]
[0134] in Yes The estimated value of .
[0135] According to (36), (37) and (38), (34) can be rewritten as:
[0136]
[0137] Where λ1 and λ2 are positive constants.
[0138] The adaptive law of the adaptive RBF neural network is designed as follows:
[0139]
[0140] in Is a normal number.
[0141] Based on the virtual control law (23), predefined time filter (24), predefined time auxiliary power system (27), control law (39), neural network adaptive law (40) and (41) proposed above, the encircling unmanned boat can complete the encirclement task of the target unmanned boat within the predefined time in the case of actuator failure and input saturation.
[0142] The proof is as follows:
[0143] The Lyapunov function is selected as follows:
[0144]
[0145] Taking the derivative of (42) and substituting it into (33), we can obtain:
[0146]
[0147] The proof process only considers |Λ i |≥μ, that is, Δτ i ≠0, |Λ i The case of |<μ is similar. Substituting (23), (24), (27), (35), (39), (40), (41) into (43), we can obtain:
[0148]
[0149] According to Young's inequality:
[0150]
[0151] Substituting (45) into (44) yields:
[0152]
[0153] Choose appropriate k1 and k2 so that According to Lemma 3, we can get:
[0154]
[0155] Substituting (40) and (41) into (47), we obtain:
[0156]
[0157] According to Lemma 1, we can get:
[0158]
[0159] According to Young's inequality:
[0160]
[0161] Substituting (49) and (50) into (48), we obtain:
[0162]
[0163] in
[0164] Solving the differential equation (51) yields:
[0165]
[0166] where C is a constant.
[0167] From this we can see that V3 is bounded, that is, all variables in V3 are also uniformly bounded, so we can further assume that there is a positive constant Ω i Make Then (51) can be rewritten as:
[0168]
[0169] in:
[0170]
[0171] Therefore, according to the predefined time theory, all signals in the closed-loop system will be Converges to the set The proof is complete.
[0172] In the specific implementation process, in order to verify the effectiveness of the method of the present invention, a simulation experiment was carried out, as follows:
[0173] For the target unmanned boat, its longitudinal speed u t =0.1m / s, yaw angular velocity r t =0.01sin(0.02t)rad / s, initial position [x t (0),y t (0)] = [20, 20]. The initial position of the surrounding unmanned boat is at the origin of the coordinate system; the environmental disturbance is set to: τ Du =-10+4sin(0.5t)cos(0.5t)-6cos(t)cos(0.5t), τ Dv =5sin(0.1t),τ Dr =8cos(0.3t)sin(1.1t); the actuator fault setting is:
[0174]
[0175]
[0176] Input saturation related parameters are: τ u =-800N, τ r =-500N·m, μ=0.01, Λ u (0) = Λ r (0) = 0, k Λu =k Λr =0.01. The relevant parameters of the adaptive RBF neural network are: F1=0.01,F2=0.01,the number of neurons is 43,and its center vector c i ,i=1,2 distribution range is [-4.2,4.2], width b i , i=1,2 is 25. The model parameters are shown in Table 1, and the control system parameters are shown in Table 2.
[0177] Table 1: Model parameters
[0178]
[0179]
[0180] Table 2: Control system parameters
[0181]
[0182] from Figure 3 It can be seen that the encirclement unmanned boat can well achieve the encirclement mission of the target unmanned boat, and the motion trajectory is relatively smooth. Figure 4 and Figure 5 It can be seen that the tracking error of the surround angle and the tracking error of the surround radius converge before the set time (15s), and the error fluctuation is very small, with high accuracy. Figure 6 and Figure 7 It can be seen that under the compensation of the auxiliary power system, the control input is constrained within the specified boundary. After 50 seconds, the actuator begins to fail, but the control input quickly returns to stability and still converges, indicating that the controller has high stability. Figure 8 and Fig. 9 It can be seen that the adaptive RBF neural network weight estimation bi-norm converges after a period of time, and the method using the adaptive RBF neural network is effective.
[0183] In summary, the present invention is based on the predefined time theory, estimates the combined impact of external environmental disturbances and actuator failures, takes actuator failures into consideration in controller design, and constructs a predefined time auxiliary power system to solve the impact of input saturation.
Claims
1. A control method for encircling a target with an underactuated unmanned vehicle for a predefined time, characterized in that: The following steps are involved: S1. Obtain the motion parameters of the unmanned boat at the previous moment, and establish the kinematic and dynamic models of the three-degree-of-freedom underactuated unmanned boat; the unmanned boat includes the surrounding unmanned boat and the target unmanned boat, and both the surrounding unmanned boat and the target unmanned boat are underactuated unmanned boats; S2, decoupling the kinematic and dynamic models through coordinate transformation to obtain a decoupled model; S3. Based on the decoupled model, the relative sight angle, relative distance, and surrounding angle of the surrounding unmanned boat are obtained between the surrounding unmanned boat and the target unmanned boat, and the surrounding angle error and surrounding radius error of the surrounding unmanned boat are set; S4, constructing a Lyapunov function of an encircling angle error and an encircling radius error surrounding the unmanned boat, and obtaining a predefined time virtual control law; S5, obtaining a filter output of a predefined time virtual control law by constructing a predefined time filter; and setting a longitudinal velocity error and a bow angular velocity error surrounding the unmanned boat according to the filter output of the predefined time virtual control law; S6, constructing a predefined time auxiliary power system based on the saturated control input at the previous moment, the original control input at the previous moment, the longitudinal velocity error surrounding the unmanned boat, and the bow angular velocity error, and obtaining the state of the predefined time auxiliary power system; S7, obtaining a comprehensive uncertainty term according to the motion parameters of the unmanned boat at the previous moment, the predefined time virtual control law at the current moment, and the longitudinal velocity error and the bow angular velocity error of the surrounding unmanned boat at the current moment; S8, constructing a predefined time controller according to the filtered output of the predefined time virtual control law at the current moment, the comprehensive uncertainty term and the state of the auxiliary power system at the predefined time, and obtaining the original control input corresponding to the current moment; S9, obtaining the saturated control input corresponding to the current moment according to the original control input corresponding to the current moment; adding the actuator fault parameter to the current saturated control input to obtain the control input after the actuator fault and input saturation, that is, obtaining the actual control input at the current moment; The expression of the predefined time auxiliary power system in step S6 is: where Λ i Indicates the state of the auxiliary power system at a predefined time; For i The first derivative of k Λi is a constant greater than 0; Indicates a fault caused by aging and wear of the device. When , it indicates that the actuator is in a fault state; Δτ i =τ si -τ i , τ si represents the saturated control input; τ i represents the original control input at the previous moment; μ is a positive constant greater than 0; u e is the longitudinal velocity error of the surrounding unmanned boat, r e is the bow angular velocity error of the surrounding unmanned boat; m 11 =mX du , m 22 =mY dv , m 23 =mx g -Y dr , m 32 =mx g -Y dr , m 33 =I z -N dr , m is the mass of the underactuated unmanned boat, x g represents the longitudinal position of the center of gravity of the underactuated unmanned vehicle, X du Represents the acceleration coefficient of the longitudinal force about the x-axis direction, Y dv Represents the acceleration coefficient of the lateral force moving along the y-axis, Y dr Represents the acceleration coefficient of the lateral component of force about the z-axis, N dr Represents the acceleration coefficient of the yaw moment about the z-axis, I z represents the inertia moment of the underactuated unmanned vehicle about the z-axis; π is radians; γ is the predefined time constant of the design, 0<γ<1; T p is the error convergence time, T p >0;sig(·) t =sign(·)|·| t , sign(·) is the sign function, and the superscript t is a constant.
2. The control method for encircling a target with an underactuated unmanned vehicle for a predefined time according to claim 1, characterized in that: The kinematic and dynamic model expressions of the three-degree-of-freedom underactuated unmanned boat in step S1 are: in (x, y) is the position coordinate of the underactuated unmanned vehicle; is the yaw angle of the underactuated unmanned vehicle, [·] T represents the transpose of a matrix, represents the first derivative of η; Represents the rotation matrix between the earth coordinate system and the ship coordinate system; M represents the mass inertia matrix, m 11 =mX du , m 22 =mY dv , m 23 =mx g -Y dr , m 32 =mx g -Y dr , m 33 =I z -N dr , m is the mass of the underactuated unmanned boat, x g represents the longitudinal position of the center of gravity of the underactuated unmanned vehicle, X du Represents the acceleration coefficient of the longitudinal force about the x-axis direction, Y dv Represents the acceleration coefficient of the lateral force moving along the y-axis, Y dr Represents the acceleration coefficient of the lateral component of force about the z-axis, N dr Represents the acceleration coefficient of the yaw moment about the z-axis, I z represents the moment of inertia of the underactuated unmanned vehicle about the z-axis; υ=[u,v,r] T , u, v and r represent the longitudinal velocity, sway velocity and bow angular velocity of the underactuated unmanned vehicle respectively; is the first-order derivative of υ; C(υ) represents the Coriolis force centripetal matrix; D(υ) represents the hydrodynamic damping coefficient matrix, X u The velocity coefficient of the longitudinal force moving along the x-axis direction, Represents the second-order velocity coefficient of the longitudinal component moving about the x-axis, Represents the third-order velocity coefficient of the longitudinal component moving about the x-axis, Y v represents the velocity coefficient of the lateral force moving about the y-axis, Represents the second-order velocity coefficient of the lateral force moving along the y-axis, Y rv Represents the coupling coefficient of the lateral force about the rotation in the z-axis direction and the motion in the y-axis direction, Y vr Represents the coupling coefficient of the lateral force about the motion in the y-axis direction and the rotation in the z-axis direction, Represents the second-order velocity coefficient of the lateral component of force rotating about the z-axis, N v Represents the velocity coefficient of the yaw moment moving in the y-axis direction, Represents the second-order velocity coefficient of the yaw moment about the y-axis direction, N rv Represents the coupling coefficient of the yaw moment about the rotation in the z-axis direction and the motion in the y-axis direction, N vr Represents the coupling coefficient of the yaw moment about the motion in the y-axis direction and the rotation in the z-axis direction, N r Represents the coupling coefficient of the yaw moment about the motion in the y-axis direction and the rotation in the z-axis direction, Represents the second-order velocity coefficient of the yaw moment about the z-axis direction; τ d =[τ Du ,τ Dv ,τ Dr ] T , τ d represents the unknown external environment disturbance, τ Du , τ Dv and τ Dr They represent the unknown external environmental disturbances corresponding to the longitudinal velocity, the unknown external environmental disturbances corresponding to the sway velocity, and the unknown external environmental disturbances corresponding to the yaw angular velocity, respectively; τ f =[τ fu ,0,τ fr ] T , τ f represents the control input, τ fu represents the control input corresponding to the longitudinal velocity; τ fr represents the control input corresponding to the bow angular velocity; the motion parameters of the unmanned boat at the last moment include x, y, u, v and r.
3. The control method for encircling a target with an underactuated unmanned vehicle for a predefined time according to claim 2, characterized in that: The expression of the decoupled model in step S2 is: in express The first derivative of Indicates the position coordinates of the surrounding unmanned boat; Indicates the sway speed of the surrounding unmanned boat; express The first derivative of; f = [f1,f2,f3] T ; d = [d u ,d v ,d r ] T ; τ'=[ξ u τ u ,0,ξ r τ r ] T ; τ m =[ξ u (τ fu -τ u ),0,ξ r (τ fr -τ r )] T , τ m is the control input affected by actuator fault and input saturation; τ u is the original control input corresponding to the longitudinal velocity direction, τ r is the original control input corresponding to the yaw angular velocity direction; χ is the intermediate parameter; 4. The control method for encircling a target with an underactuated unmanned vehicle for a predefined time according to claim 3, characterized in that: The relative sight angle β between the surrounding unmanned boat and the target unmanned boat in step S3 s The expression is: where atan2(·) is the inverse tangent function; (x t ,y t ) is the position coordinate of the target unmanned boat; The relative distance between the surrounding unmanned boat and the target unmanned boat ρ s The expression is: The surrounding angle η surrounding the unmanned boat s The expression is: Where π is radians; The surrounding angle error ρ of the unmanned boat e and the enclosing radius error η e The expressions are: where ρ d is the expected orbit radius; η d is the desired wrap angle, Δ is the forward distance based on the line-of-sight guidance method, Δ>0, For x t The first derivative of for y t The first derivative of .
5. The control method for encircling a target with an underactuated unmanned vehicle for a predefined time according to claim 4, characterized in that: The expression of the predefined time virtual control law in step S4 is: where α u Represents the virtual control law in the longitudinal velocity direction, α r represents the virtual control law for the yaw angular velocity direction, γ is the predefined time constant, 0<γ<1, T p is the error convergence time, T p >0; k1 and k2 are positive constants; sig(·) t =sign(·)|·| t , sign(·) is the sign function, and the superscript t is a constant.
6. The control method for encircling a target with an underactuated unmanned vehicle for a predefined time according to claim 5, characterized in that: The expression of the filtered output of the predefined time virtual control law in step S5 is: in and All are filtered outputs of predefined time virtual control laws; for The first derivative of for The first derivative of ; The expressions of the longitudinal velocity error and bow angular velocity error surrounding the unmanned boat are: where u e is the longitudinal velocity error of the surrounding unmanned boat, r e is the bow angular velocity error of the surrounding unmanned boat; e fu and e fr All of them are filtering errors surrounding the unmanned boat.
7. The control method for encircling a target with an underactuated unmanned vehicle for a predefined time according to claim 6, characterized in that: The specific method of step S7 is: The motion parameters of the unmanned boat at the previous moment, the predefined time virtual control law at the current moment, and the longitudinal velocity error and bow angular velocity error of the surrounding unmanned boat at the current moment are combined into vectors x1 and x2. Vectors x1 and x2 are used as the input of the adaptive RBF neural network, and the comprehensive uncertainty term is obtained through the approximation function of the adaptive RBF neural network. where x1=[u,α u ] T ; x2=[r,α r ] T ; The adaptive law of the adaptive RBF neural network is: F1, F2, Ψ1 and Ψ2 are all positive constants; for The first derivative of ; for The first derivative of ; and are the estimated weights of the adaptive RBF neural network, Z1(x1) and Z2(x2) represent Gaussian basis functions, b l and c l Represent the width vector and center vector of the Gaussian basis function, respectively, ||·|| represents the second norm, and e is a natural constant; for The first derivative of ; for The first derivative of ; and is the upper bound estimate of the approximation error of the adaptive RBF neural network.
8. The control method for encircling a target with an underactuated unmanned vehicle for a predefined time according to claim 7, characterized in that: The expression of the predefined time controller in step S8 is: in and is the original control input corresponding to the current moment; i=u,r, Indicates an offset fault caused by installation error, displacement, or eccentricity. and When , it indicates that the actuator works normally; k3 and k4 are both positive numbers; Λ u and Λ r Indicates the state of the auxiliary power system at a predefined time; k Λu and k Λr is a constant greater than 0.
9. The method for controlling target encirclement of an underactuated unmanned vehicle for a predefined time according to claim 8, characterized in that: The specific method of step S9 includes the following sub-steps: S9-1, according to the formula: Saturate the original control input corresponding to the current moment to obtain the saturated control input corresponding to the current moment in is the original control input corresponding to the current moment; and are the upper and lower bounds of the control input respectively; S9-2, according to the formula: Acquiring control inputs through actuator faults and input saturation That is, the actual control input at the current moment is obtained; i * =u * ,r * , is the actual control input corresponding to the longitudinal velocity direction at the current moment, is the actual control input corresponding to the yaw angular velocity direction at the current moment; S9-3. Control the surrounding unmanned boat according to the actual control input at the current moment.
