Ship dynamic positioning event trigger control method considering execution performance difference
Through the adaptive parameter model and multi-port event triggering mechanism, a virtual control law and first-order filter are constructed, which solves the problem of difference in the transmission pressure of the ship's signal and the performance of the actuator, and achieves more efficient control effects and reduced communication load.
Patent Information
- Application Number
- CN202411043432.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2024-11-15
AI Technical Summary
When the existing power positioning ship control system faces large winds and waves and the difference in the performance of the actuator, the signal transmission pressure has not been completely alleviated, and the traditional event triggering mechanism has failed to effectively reduce the communication load from the sensor to the actuator, affecting the control effect.
Using the execution performance difference model and multi-port event triggering mechanism based on adaptive parameters, a virtual control law is constructed and the first-order filter is determined through dynamic surface technology, and the actual control law vector is obtained to achieve consideration of execution performance differences and signal transmission pressure. The reduction.
It effectively reduces the system signal transmission pressure, takes into account the performance differences of actuators, improves the output performance of the control system, and is suitable for navigation and engineering applications.
Smart Images

Figure CN118963133A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of application of adaptively controlled dynamic positioning ship equipment, and in particular to a ship dynamic positioning event triggering control method taking execution performance differences into consideration. Background Art
[0002] Dynamically positioned vessels are widely used on oil drilling platforms or scientific research vessels, and are generally equipped with multiple thrusters including side thrusters and azimuth thrusters, which can stabilize at the desired position under the conditions of marine environmental disturbances.
[0003] In actual operation, dynamically positioned ships may encounter strong winds and waves, which puts forward strong robustness requirements for the design of adaptive controllers. In addition, in order to achieve error stabilization, the control signal sent from the controller to the actuator is a continuous signal. If a fixed-frequency signal zero-order holder (ZOH) is used, the signal transmission frequency can be reduced, but the control effect is weakened. Existing research results mainly use traditional event triggering mechanisms to reduce the communication load from the controller to the actuator, but the communication load from the sensor to the actuator has not been alleviated. In addition, existing research results do not consider the differences in the execution effects of different actuators, which is obviously not conducive to application in engineering practice.
[0004] Based on the above analysis, the dynamic positioning control algorithm based on the traditional event trigger mechanism has the following two main defects:
[0005] 1) Although the traditional event trigger mechanism reduces the signal transmission pressure between the controller and the actuator, continuous signals are still transmitted between the sensor and the controller, and the signal transmission pressure of the system has not been completely alleviated, which is obviously not conducive to engineering practice.
[0006] 2) In actual navigation, a full-drive ship is generally equipped with multiple actuators including main propellers, side propellers, and full-turn propellers. Due to the differences in execution performance between actuators, their influence on the control effect is coupled with each other, which seriously weakens the output performance of the control system. Summary of the invention
[0007] The invention discloses a ship dynamic positioning event triggering control method taking into account execution performance differences to overcome the above technical problems.
[0008] In order to achieve the above object, the technical solution of the present invention is:
[0009] A method for triggering control of ship dynamic positioning events taking into account execution performance differences comprises the following steps:
[0010] S1: Establish an execution performance difference model based on adaptive parameters and a mathematical model of a three-degree-of-freedom nonlinear all-drive ship with an adaptive law of actuator correlation coefficients;
[0011] S2: According to the three-degree-of-freedom nonlinear full-drive ship mathematical model, obtain the system output error signal vector; based on the multi-port event trigger mechanism, obtain the error trigger value;
[0012] S3: According to the error trigger value and the three-degree-of-freedom nonlinear full-drive ship mathematical model,
[0013] Construct a virtual control law to determine the first-order filter based on dynamic surface technology;
[0014] S4: According to the first-order filter, the system output error signal vector and the error trigger value, an actual control law vector is obtained to implement trigger control of the ship dynamic positioning event taking into account the execution performance difference through the actual control law vector.
[0015] Furthermore, the method of obtaining the actual control law vector is as follows;
[0016]
[0017] Among them, α up =(α x ,α y ,α ψ ) T
[0018] Where: Φ(Ξ) is the robust neural damping term, φ Ξ represents the robust neural scaling, φ Ξ =1+||S Ξ (Ξ)||||β Ξ ||;||S Ξ (Ξ)|| F represents the basis function norm; α x represents the forward control law; α y represents the drift control law; α ψ represents the heading control law; k Ξn K is the neural damping design matrix, which is a diagonal matrix with all elements positive; Ξe is the velocity stabilization matrix; α up represents the actual control law vector.
[0019] Furthermore, the mathematical model of the three-degree-of-freedom nonlinear full-drive ship is established as follows:
[0020]
[0021]
[0022] Where: represents the attitude vector of the ship, where x, y, ψ represent the abscissa of the ship's position in the inertial coordinate system, the ordinate of the ship's position and the heading angle of the ship respectively; T Represents the transpose of a matrix or vector; represents a three-dimensional vector space; R(ψ) is a rotation matrix;
[0023] represents the velocity vector, where u, v, and r represent the forward speed, drift speed, and yaw angular velocity of the ship respectively; M represents the inertia matrix; represents the first derivative of η; represents the first-order derivative of Ξ; N(Ξ)Ξ is the unknown nonlinear term of the dynamically positioned ship; is the maneuvering torque vector, where τ u ,τ v ,τ r They represent longitudinal thrust, transverse thrust and ship turning moment respectively; d ω represents the marine environmental disturbance, where d w =[d wu ,d wv ,d wr ] T , d wu Indicates the forward motion environment interference; d wv Indicates the environmental interference of drifting motion; d wr Indicates the environmental disturbance of the bow motion;
[0024] in,
[0025]
[0026]
[0027] τ=T(γ)κ(n)u p
[0028] Where: m represents the mass of the dynamically positioned ship; represents the hydrodynamic derivative of forward acceleration;
[0029] represents the hydrodynamic derivative of the drift acceleration; Y r represents the coupled velocity hydrodynamic derivative; represents the hydrodynamic derivative of coupled acceleration; x G Indicates the distance between the ship's center of mass and the origin of the attached coordinate system; N r I represents the hydrodynamic derivative of the bow speed; z represents the moment of inertia of a dynamically positioned ship; |·| represents absolute value operation;
[0030] T(γ) is the position matrix of the actuator; κ(n) is the gain matrix of the actuator; up To directly control the input vector, u p =[u p1 ,u p2 ,...,u pq ,...,u pQ ] T , where Q is the equivalent thruster, i.e. the number of actuators, and u pq =p q |p q | represents the direct control input vector of the qth equivalent thruster; p q is the pitch ratio of the qth equivalent propeller, q = 1, 2, ..., Q represents the index number of the equivalent propeller;
[0031] Among them, the direct control input vector u p Obtained according to the following execution performance difference model based on adaptive parameters:
[0032] Where: k p is the execution effect matrix; p is the intermediate input, is the execution bias vector;
[0033] Let g p =k(n)k p , Then θ=λsup{μ}l;
[0034] Where: g p is the effect gain matrix; μ is the deviation gain vector; θ is the scaling vector, λ is the effect compensation matrix, k(n) represents the gain matrix of the actuator; k p is the execution effect matrix; is the execution bias vector; l represents the Q row and column vector whose elements are all 1, where l = (1,1,...,1) T ; sup{μ} means taking the lower limit of the elements of μ;
[0035] Then, the adaptive parameters are introduced to estimate the effect compensation matrix λ and the scaling vector θ, so the intermediate input γ p for:
[0036]
[0037]
[0038] Where: γ p is the intermediate input of the design, are the adaptive parameters of the elements of the 1st, 2nd, …, qth, …, Qth equivalent actuators in the effect compensation matrix λ, respectively. are the adaptive parameters of the elements of the 1st, 2nd, …, qth, …, Qth equivalent actuators in the scaling vector θ respectively; Q is the number of equivalent thrusters; is the pseudo-inverse matrix of T(γ); T(γ) is the position matrix of the actuator; α up represents the actual control law vector; Θ represents the hyperbolic tangent vector; T1(γ) represents the first column vector of T(γ); T q (γ) represents the qth column vector of T(γ); ε represents the hyperbolic tangent epsilon; Ξ e represents the velocity error vector;
[0039] Ξ e =β Ξ -Ξ
[0040] Where: e represents the trigger speed error vector; β Ξ represents a first-order filter; Ξ represents a velocity vector;
[0041] Among them, the following actuator correlation coefficient adaptive law is established to obtain the adaptive parameters of the elements of the effect compensation matrix λ and the scaling vector θ:
[0042]
[0043]
[0044] Where: Γ λq , Γ θq , σ λq , σ θq These are design parameters. They are and The initial value of They are and The first derivative of ; express The elements of rows q and columns b of T aq (γ) represents the a-row, q-column element of T(γ), express The element in row a and column q, Ξ ea Indicates Ξ e The ath element of upb Represents α up The bth element of λq represents the λ gain parameter of the qth equivalent thruster, Γ θq represents the θ gain parameter of the qth equivalent thruster, σ λq represents the λ convergence parameter of the qth equivalent thruster, σ θqrepresents the θ convergence parameter of the qth equivalent thruster.
[0045] Furthermore, the error trigger value is obtained as follows:
[0046] First, obtain the system output error signal vector η e as follows:
[0047]
[0048] η e =(x e (t),y e (t),ψ e (t)) T
[0049] Where: x e (t),y e (t),ψ e (t) respectively represent the abscissa error, ordinate error and heading error of the ship's position in the inertial coordinate system at time t; Respectively expressed in t k -t k+1 The trigger value of the abscissa error, the trigger value of the ordinate error and the trigger value of the heading error of the ship's position within the interval; and They respectively represent the triggering time of the abscissa error, the ordinate error and the heading error of the ship's position; k represents the index number of the time; η e Represents the system output error signal vector;
[0050] The error trigger value for:
[0051]
[0052] Where: Indicates the error trigger value; K T represents the trigger matrix,
[0053] K T =diag{1+λ 1x δ x ,1+λ 1y δ y ,1+λ 1ψ δ ψ}; diag{·} means taking diagonal operation; λ 1x ,λ 1y ,λ 1ψ ,λ 2x ,λ 2y ,λ 2ψ All are trigger boundary parameters;x represents the forward trigger increment threshold; y Indicates the incremental threshold for triggering horizontal drift; ζ ψ Indicates the heading trigger increment threshold.
[0054] Furthermore, the triggering rules of the multi-port event triggering mechanism are as follows:
[0055]
[0056] Where: inf{·} represents the lower bound operation; δ x Indicates the forward trigger coefficient threshold; δ y Indicates the drift trigger coefficient threshold; δ ψ Indicates the threshold of the yaw trigger coefficient; ζ x represents the forward trigger increment threshold; y Indicates the incremental threshold for triggering horizontal drift; ζ ψ Indicates the heading trigger increment threshold.
[0057] Furthermore, the virtual control law is constructed as follows:
[0058]
[0059] Where: α Ξ represents the virtual control law; R -1 (ψ) is the inverse matrix of the rotation matrix; K η is the attitude calming matrix; Indicates the error trigger value;
[0060] According to the virtual control law α Ξ , based on the dynamic surface technique, determine the first-order filter β Ξ , the formula used is as follows:
[0061]
[0062] Where: t Ξ is the time constant matrix, where t Ξ =diag{t1,t2,t3}, diag{·} represents diagonal operation, t1, t2, t3 represent the first component, the second component and the third component on the diagonal of the time constant matrix respectively; β Ξ represents a first-order filter; Represents β Ξ The first derivative of Ξ (0) represents the initial value of the first-order filter; α Ξ (0) represents the initial value of the virtual control law.
[0063] Beneficial effects: The present invention is a ship dynamic positioning event trigger control method that takes into account the difference in execution performance. According to the execution performance difference model based on adaptive parameters and the three-degree-of-freedom nonlinear full-drive ship mathematical model of the actuator correlation coefficient adaptive law, the system output error signal vector is obtained; based on the multi-port event trigger mechanism, the error trigger value is obtained; then a virtual control law is constructed, and a first-order filter is determined based on the dynamic surface technology; finally, the actual control law vector is obtained to achieve the control of the ship dynamic positioning event trigger. The present invention can solve the problem of execution performance difference in the ship dynamic positioning adaptive control method, and reduce the signal transmission pressure of the system by citing the multi-port event trigger mechanism; at the same time, the execution performance difference of the actuator is also considered, and the unknown difference and actuator gain are continuously compensated by the designed actuator correlation coefficient adaptive law, so that the present invention is closer to navigation practice and engineering application, and has theoretical guiding significance for the intelligent dynamic positioning ship to move towards engineering application. BRIEF DESCRIPTION OF THE DRAWINGS
[0064] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0065] Figure 1 This is a flow chart of a ship dynamic positioning event triggering control method considering execution performance differences of the present invention;
[0066] Figure 2 It is a wind speed and wind direction curve and a three-dimensional view of sea surface wind and waves in an embodiment of the present invention;
[0067] Figure 3 A schematic diagram comparing the ship trajectories under the control of an embodiment of the present invention and a traditional algorithm;
[0068] Figure 4 is a curve showing the change of the ship position coordinates and the bow heading in the embodiment of the present invention;
[0069] Figure 5 is a variation curve of the forward speed, the transverse drift speed and the bow angular velocity of the ship in the embodiment of the present invention;
[0070] Figure 6 is a schematic diagram of an actual input curve of an actuator of a ship actuator in an embodiment of the present invention;
[0071] Figure 7 A schematic diagram of a multi-port event triggering time interval in an embodiment of the present invention;
[0072] Figure 8 is the adaptive parameter in the embodiment of the present invention Change curve;
[0073] Fig. 9 is the adaptive parameter in the embodiment of the present invention Change curve. DETAILED DESCRIPTION
[0074] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0075] This embodiment introduces a ship dynamic positioning event trigger control method considering execution performance differences, such as Figure 1 As shown, the following steps are included:
[0076] S1: Establish an execution performance difference model based on adaptive parameters and a mathematical model of a three-degree-of-freedom nonlinear all-drive ship with an adaptive law of actuator correlation coefficients;
[0077] Preferably, the mathematical model of the three-degree-of-freedom nonlinear full-drive ship is established as shown in equations (1) and (2):
[0078]
[0079]
[0080] Where: represents the attitude vector of the ship, where x, y, ψ represent the abscissa of the ship's position in the inertial coordinate system, the ordinate of the ship's position and the heading angle of the ship respectively; T represents the transpose of the matrix or vector; represents a three-dimensional vector space; R(ψ) is a rotation matrix;
[0081] represents the velocity vector, where u, v, and r represent the forward speed, drift speed, and yaw angular velocity of the ship respectively; M represents the inertia matrix; represents the first derivative of η; represents the first-order derivative of Ξ; N(Ξ)Ξ is the unknown nonlinear term of the dynamically positioned ship, which is used to describe the inertial and viscous fluid dynamic (or torque) variables acting on the hull; is the maneuvering torque vector, where τ u ,τ v ,τ rThey represent longitudinal thrust, transverse thrust and ship turning moment respectively; d ω Represents the marine environment disturbance, which is immeasurable and is used to describe the disturbance force and torque of the external marine environment disturbance on the ship, where d w =[d wu ,d wv ,d wr ] T , d wu Indicates the forward motion environment interference; d wv Indicates the environmental interference of drifting motion; d wr Indicates the environmental disturbance of the bow motion;
[0082] Specifically, the inertia matrix M and the unknown nonlinear term N(Ξ)Ξ of the dynamically positioned ship can be used to describe the hydrodynamic derivatives.
[0083] in,
[0084]
[0085]
[0086] The maneuvering moment τ is expressed by the following formula:
[0087] τ=T(γ)κ(n)u p
[0088] Where: m represents the mass of the dynamically positioned ship; represents the hydrodynamic derivative of forward acceleration;
[0089] represents the hydrodynamic derivative of the drift acceleration; Y r represents the coupled velocity hydrodynamic derivative; represents the hydrodynamic derivative of coupled acceleration; x G Indicates the distance between the ship's center of mass and the origin of the attached coordinate system; N r I represents the hydrodynamic derivative of the bow speed; z represents the moment of inertia of a dynamically positioned ship; |·| represents absolute value operation;
[0090] T(γ) is the position matrix of the actuator, which is determined by the torque length of each actuator and the direction angle γ of the full-turn propeller; κ(n) is the gain matrix of the actuator, which is a gain matrix in the form of a diagonal matrix, whose elements are determined by the propeller speed n, and this term represents the unknown gain of the actuator; u p To directly control the input vector, u p =[u p1 ,u p2 ,...,u pq ,...,u pQ ] T, where Q is the number of equivalent thrusters, u pq =p q |p q | represents the direct control input vector of the qth equivalent thruster; p q is the pitch ratio of the qth equivalent propeller, q=1,2,…,Q represents the index number of the equivalent propeller.
[0091] Since ships may encounter the problem of actuator performance differences in actual engineering, this embodiment introduces an actuator performance difference model to solve this problem. The model can be expressed by (4).
[0092] The execution performance difference model based on adaptive parameters is established as follows:
[0093]
[0094] Where: k p is the execution effect matrix, which has a diagonal form; p is the intermediate input, is the execution deviation vector. p and Together they represent unknown execution performance differences among the executors.
[0095] Specifically, in the embodiment, the execution performance difference model includes four modes:
[0096] 1. k pq =1, At this time, the actuator, that is, the equivalent thruster, operates normally.
[0097] 2.0 <k pq <1, At this point the actuator suffers a loss of efficiency.
[0098] 3. k pq =1, At this point the actuator is experiencing an execution deviation.
[0099] 4. k pq =0, which means that the actuator is completely ineffective.
[0100] Among them, k pq is the execution effect matrix k p The element corresponding to the qth equivalent actuator; Represents the execution deviation vector The element in corresponds to the qth equivalent actuator.
[0101] The execution performance difference model can obtain the direct control input vector u by simulating the unknown execution performance difference of the actuator. p Although the execution effect matrix kp and execution deviation vector The expression of is unknown, but the intermediate input Υ can be obtained by the following method p , so that k p and Compensate and obtain u p , as follows:
[0102] Let g p =k(n)k p , Then θ=λsup{μ}l;
[0103] Where: g p is the effect gain matrix; μ is the deviation gain vector; θ is the scaling vector, λ is the effect compensation matrix, k(n) represents the gain matrix of the actuator; k p is the execution effect matrix; is the execution bias vector; l represents the Q row and column vector whose elements are all 1, where l = (1,1,...,1) T ; sup{μ} means taking the lower limit for the elements of μ.
[0104] Based on the above design, the adaptive parameter and As the estimated value of the element corresponding to the qth actuator in the effect compensation matrix λ and the scaling vector θ, the effect compensation matrix λ and the scaling vector θ are estimated, and the intermediate input Υ p Designed for:
[0105]
[0106]
[0107] Where: p is the intermediate input of the design, are the adaptive parameters of the elements of the 1st, 2nd, …, qth, …, Qth equivalent actuators in the effect compensation matrix λ, respectively. are the adaptive parameters of the elements of the 1st, 2nd, …, qth, …, Qth equivalent actuators in the scaling vector θ respectively; Q is the number of equivalent thrusters; is the pseudo-inverse matrix of T(γ); T(γ) is the position matrix of the actuator; α up represents the actual control law vector; Θ represents the hyperbolic tangent vector; T1(γ) represents the first column vector of T(γ); T q (γ) represents the qth column vector of T(γ); ε represents the hyperbolic tangent epsilon; Ξ e represents the velocity error vector;
[0108] Among them, the velocity error vector Ξ e The construction is as follows;
[0109] Ξ e =β Ξ -Ξ (6)
[0110] Where: e represents the velocity error vector; β Ξ represents a first-order filter; Ξ represents a velocity vector;
[0111] Specifically, the actuator association coefficient adaptive law is expressed by (7) and (8). In order to obtain the estimated value of the element corresponding to the qth equivalent actuator in the effect compensation matrix λ and the scaling vector θ, the following actuator association coefficient adaptive law is established to obtain the adaptive parameters of the elements of the effect compensation matrix λ and the scaling vector θ:
[0112]
[0113]
[0114] where Γ λq , Γ θq , σ λq , σ θq These are design parameters. They are and The initial value of They are and The first derivative of ; express The elements of the q rows and b columns of express The element in row a and column q, Ξ ea Indicates Ξ e The ath element of upb Represents α up The bth element of Γ λq represents the λ gain parameter of the qth equivalent thruster, Γ θq represents the θ gain parameter of the qth equivalent thruster, σ λq represents the λ convergence parameter of the qth equivalent thruster, σ θq represents the θ convergence parameter of the qth equivalent thruster; a represents the first-dimensional index number; b represents the second-dimensional index number.
[0115] In addition, the determination of adaptive parameters also includes and elements in T(γ), such as T aq (γ), For the actuator correlation coefficient adaptive law of the qth equivalent thruster, the elements taken in the matrix The corresponding position in T(γ) is related to the equivalent thruster index number q corresponding to the actuator correlation coefficient adaptive law. Therefore, the matrix elements involved in obtaining the actuator correlation coefficient adaptive law also reference the same index number q as the equivalent thruster.
[0116] S2: According to the three-degree-of-freedom nonlinear full-drive ship mathematical model, obtain the system output error signal vector η e ; Based on the multi-port event trigger mechanism, obtain the error trigger value
[0117] Preferably, the error trigger value Get as follows:
[0118] Specifically, based on the three-degree-of-freedom nonlinear full-drive ship mathematical model, the ship's position error and heading error x are constructed. e ,y e ,ψ e , and written in vector form η e , then η e =(x e (t),y e (t),ψ e (t)) T is the system output error signal vector, and its expression is η e =η d -η, η d is the expected attitude vector of the ship; η represents the attitude vector of the ship.
[0119] Specifically, in order to solve the problem of limited communication bandwidth, this embodiment introduces a multi-port event trigger mechanism. The sensor observes the system output error signal and transmits it to the controller only at each triggering moment, as shown in the following formula:
[0120] First, obtain the system output error signal vector η e as follows:
[0121]
[0122] η e =(x e (t),y e (t),ψ e (t)) T
[0123] Where: x e (t),y e (t),ψ e(t) respectively represent the abscissa error, ordinate error and heading error of the ship's position in the inertial coordinate system at time t; Respectively expressed in t k -t k+1 The trigger value of the abscissa error, the trigger value of the ordinate error and the trigger value of the heading error of the ship's position within the interval; and They represent the triggering time of the abscissa error, the ordinate error and the heading error of the ship’s position, respectively. They are k -t k+1 The value is constant within the interval until the next trigger time t k+1 Refresh; k represents the index number of the time; η e Represents the system output error signal vector;
[0124] Furthermore, based on the above design, the error trigger value is derived through the multi-port event trigger mechanism The expression is:
[0125]
[0126] Where: Indicates the error trigger value; K T represents the trigger matrix,
[0127] K T =diag{1+λ 1x δ x ,1+λ 1y δ y ,1+λ 1ψ δ ψ}; diag{·} means taking diagonal operation; λ 1x ,λ 1y ,λ 1ψ ,λ 2x ,λ 2y ,λ 2ψ All are trigger boundary parameters; x represents the forward trigger increment threshold; y Indicates the incremental threshold for triggering horizontal drift; ζ ψ Indicates the heading trigger increment threshold;
[0128] in, They are lateral, longitudinal and heading components of the ship;
[0129] Preferably, the triggering rules of the multi-port event triggering mechanism are as follows:
[0130]
[0131] in: As the main push trigger mechanism,
[0132] For the side push trigger mechanism,
[0133] It is a full-swing trigger mechanism.
[0134] Where: inf{·} represents the lower bound operation; δ x Indicates the forward trigger coefficient threshold; δ y Indicates the drift trigger coefficient threshold; δ ψ Indicates the threshold of the yaw trigger coefficient; ζ x represents the forward trigger increment threshold; y Indicates the incremental threshold for triggering horizontal drift; ζ ψ Indicates the heading trigger increment threshold; where 0<δ x , δ y , δ ψ <1,0<ζ x , y , ψ <1.
[0135] Specifically, in this embodiment, due to the unknown nonlinear term N of the dynamically positioned ship in the ship motion model, Ξ (Ξ) is an unknown smooth continuous function, so the radial basis function neural network F is used nn (Ξ) to fit.
[0136]
[0137] Where: F nn (Ξ) is a multivariate function about Ξ, used to fit N Ξ (Ξ); S Ξ (Ξ) represents the basis function vector, where S Ξ (Ξ) = diag{S u (Ξ),S v (Ξ),S r (Ξ)},S u (Ξ), S v (Ξ), S r (Ξ) represent the forward basis function, the lateral drift basis function, and the yaw basis function respectively; A Ξ is the weight matrix of the neural network, ε Ξ is the fitting error, b Ξ =||A Ξ || F As the norm value of the weight matrix, ||A Ξ || F Indicates AΞ The norm value of W Ξ It represents the normalized value of the weight matrix.
[0138] S3: According to the error trigger value and the three-degree-of-freedom nonlinear full-drive ship mathematical model, constructing the virtual control law α Ξ , to determine the first-order filter β based on the dynamic surface technique Ξ ;
[0139] Preferably, the virtual control law α Ξ The build is as follows:
[0140]
[0141] Where: α Ξ represents the virtual control law; R -1 (ψ) is the inverse matrix of the rotation matrix; K η is the attitude calming matrix; Indicates the error trigger value;
[0142] Specifically, based on the current research results, the ship controller will encounter the problem of "complexity explosion" in practical engineering applications. To solve this problem, the present invention introduces dynamic surface technology (DSC) by constructing a first-order filter β Ξ The virtual control law is filtered to avoid repeated derivation in the controller calculation process.
[0143] According to the virtual control law α Ξ , based on the dynamic surface technique, determine the first-order filter β Ξ , the formula used is as follows:
[0144]
[0145] Where: t Ξ is the time constant matrix. The matrix elements are selected based on experience and are usually in the range of 0.01 to 0.1. Ξ =diag{t1,t2,t3}, diag{·} represents diagonal operation, t1, t2, t3 represent the first component, the second component and the third component on the diagonal of the time constant matrix respectively; β Ξ represents a first-order filter; Represents β Ξ The first derivative of Ξ (0) represents the initial value of the first-order filter; α Ξ (0) represents the initial value of the virtual control law;
[0146] Among them, the virtual control law can be used to stabilize the abscissa error, ordinate error and heading error of the ship's position, that is, the DSC error:
[0147] q Ξ =α Ξ -β Ξ ,
[0148] Where: q Ξ is the dynamic surface error;
[0149] S4: According to the first-order filter β Ξ , speed error e 、System output error signal vector η e and error trigger value Get the actual control law vector α up , through the actual control law vector α up , to achieve control over the triggering of ship dynamic positioning events taking into account differences in execution performance.
[0150] Preferably, the actual control law vector α is obtained up Here is how;
[0151] Specifically, in order to stabilize the errors in the three degrees of freedom of forward movement, lateral drift and bow pitch, the actual control law vector α is designed up , the expression is as follows:
[0152]
[0153] Among them, α up =(α x ,α y ,α ψ ) T
[0154] Where: Φ(Ξ) is the robust neural damping term, φ Ξ represents the robust neural scaling, φ Ξ =1+||S Ξ (Ξ)||||β Ξ ||;||S Ξ (Ξ)|| F represents the basis function norm; α x represents the forward control law; α y represents the drift control law; α ψ represents the heading control law; k Ξn K is the neural damping design matrix, which is a diagonal matrix with all elements positive; Ξe is the velocity stabilization matrix; where k Ξn and K Ξe is a matrix of diagonal form chosen empirically.
[0155] A specific embodiment of the present invention is as follows:
[0156] In order to perform the dynamic positioning task, the initial position coordinates are selected as (10m, 10m), the initial heading is 156°, the expected position is set to (0m, 0m), the expected heading is 140°, and the task duration is 300s. Figure 2 It represents the simulation environment used on the MATLAB simulation platform, showing the 10m air wind speed and wind direction curve and the three-dimensional view of sea surface wind and waves under level 6 sea conditions. Figure 3 It is shown that the trajectory of the ship under the control of this embodiment and the traditional algorithm is compared. It can be seen that the ship can eventually stabilize at the desired position under both algorithms, but the trajectory of the ship under the algorithm proposed in this embodiment is obviously smoother. Figure 4 The variation curves of the ship position and heading are given, showing that the ship can overcome the performance differences of the actuators, starting from the initial position (10m, 10m) and the initial heading angle of 156°, and finally stabilizing to (0m, 0m) and the heading angle of 140°. Figure 5 The variation curves of the ship's forward speed, drift speed and bow angular velocity are given. The step-shaped signal is the signal after the multi-port event is triggered, indicating that the speed signal transmission from the sensor to the actuator has achieved intermittent triggering, which reduces the signal transmission frequency of the channel between the sensor and the controller. Figure 6 represents the actual input of the ship actuator, where the signal actually received by the actuator is in a step-like shape (such as p 2o ), indicating that the signal transmission frequency from the controller to the actuator is reduced, alleviating the communication burden. Since the curves of other actuators are similar to p 2o , so it is not expanded here. Figure 7 It represents the time interval diagram of multi-port event triggering. The three control channels (i.e., forward, drift, and yaw channels) are triggered independently. Each line represents a trigger. Signal transmission is performed only when an event is triggered. It can be seen that the trigger time interval is large, which reduces the signal transmission frequency. Figure 8 and Fig. 9 The adaptive parameter change curves designed in this embodiment are respectively shown, which can realize adaptive update of values to continuously compensate for the unknown gain and performance differences of the actuator, among which, Fig. 9 Adaptive parameters in The changes are large at the beginning of the simulation, but gradually become stable over time.
[0157] This embodiment has the following two main features:
[0158] (1) In terms of reducing the communication load, a multi-port event trigger mechanism is designed. A trigger rule based on the output signal is constructed on the dynamic positioning signal transmission channel (including three degrees of freedom: forward, lateral and yaw). Based on the backstepping recursive control design, the control law on the three degrees of freedom is obtained. This can simultaneously reduce the communication load between the sensor and the controller and between the controller and the actuator, taking into account both the guarantee of strong robustness and the saving of communication resources.
[0159] (2) Aiming at the actuator performance difference problem encountered by the actuator-driven full-drive ship in engineering practice, an actuator performance difference model is introduced, in which the unknown actuator performance difference and unknown gain are compensated by designing the actuator correlation coefficient adaptive law, and there is no need to adjust the parameters online. In this algorithm, robust neural damping technology and dynamic surface technology are used to approximate the model structure nonlinearity and parameter uncertainty in the dynamic positioning ship model.
[0160] Combining the prior art, controller design and simulation test, the present invention has the following two beneficial effects in the field of ship dynamic positioning:
[0161] 1) The multi-port event trigger mechanism and execution performance difference model adopted by the present invention can reduce the communication load while ensuring good control effect. The traditional event trigger mechanism can only reduce the communication load from the controller to the actuator. The trigger mechanism adopted by the present invention can also reduce the communication load from the sensor to the controller. In addition, the execution performance difference of the actuator is also taken into account. The actuator correlation coefficient adaptive law is designed to continuously compensate for the unknown performance difference and actuator gain, making this embodiment closer to navigation practice and engineering application.
[0162] 2) A full-drive ship path tracking simulation experiment was conducted in a simulated ocean environment to verify the effectiveness of the control algorithm proposed in the present invention. The robust neural damping technology and dynamic surface technology used are used to deal with unknown nonlinear terms and "complexity explosion" problems respectively. The application prospects of dynamic positioning in the marine field are very important. This embodiment will be able to provide a design method for intelligent control for surface operating platforms such as scientific research vessels, oil drilling, and floating docks.
[0163] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements 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 ship dynamic positioning event triggering control method considering execution performance differences, characterized in that: The steps include: S1: Establish a mathematical model of a three-degree-of-freedom nonlinear all-drive ship based on the execution performance difference model of adaptive parameters and the actuator correlation coefficient adaptive law; S2: According to the three-degree-of-freedom nonlinear full-drive ship mathematical model, obtain the system output error signal vector; based on the multi-port event trigger mechanism, obtain the error trigger value; S3: constructing a virtual control law according to the error trigger value and the three-degree-of-freedom nonlinear full-drive ship mathematical model to determine a first-order filter based on a dynamic surface technology; S4: According to the first-order filter, the system output error signal vector and the error trigger value, an actual control law vector is obtained to implement trigger control of the ship dynamic positioning event taking into account the execution performance difference through the actual control law vector.
2. A ship dynamic positioning event triggering control method considering execution performance differences according to claim 1, characterized in that: The method to obtain the actual control law vector is as follows; Among them, a up =(a x ,a y ,a ψ ) T Where: Φ(Ξ) is the robust neural damping term, φ Ξ represents the robust neural scaling, φ Ξ =1+||S Ξ (Ξ)||||β Ξ ||;||S Ξ (Ξ)|| F represents the basis function norm; α x represents the forward control law; α y represents the drift control law; α ψ represents the heading control law; k Ξn K is the neural damping design matrix, which is a diagonal matrix with all elements positive; Ξe is the velocity stabilization matrix; α up represents the actual control law vector; β Ξ represents a first-order filter; e represents the trigger velocity error vector; R(ψ) is the rotation matrix; Represents the error trigger value; η e Represents the system output error signal vector.
3. A ship dynamic positioning event triggering control method considering execution performance differences according to claim 1, characterized in that: The mathematical model of the three-degree-of-freedom nonlinear all-drive ship is established as follows: Where: represents the attitude vector of the ship, where x, y, ψ represent the abscissa of the ship's position in the inertial coordinate system, the ordinate of the ship's position and the heading angle of the ship respectively; T Represents the transpose of a matrix or vector; represents a three-dimensional vector space; R(ψ) is a rotation matrix; represents the velocity vector, where u, v, and r represent the forward speed, drift speed, and yaw angular velocity of the ship respectively; M represents the inertia matrix; represents the first derivative of η; represents the first-order derivative of Ξ; N(Ξ)Ξ is the unknown nonlinear term of the dynamically positioned ship; is the maneuvering torque vector, where τ u ,τ v ,τ r They represent longitudinal thrust, transverse thrust and ship turning moment respectively; d ω represents the marine environmental disturbance, where d w =[d wu ,d wv ,d wr ] T , d wu Indicates the forward motion environment interference; d wv Indicates the environmental interference of drifting motion; d wr Indicates the environmental disturbance of the bow motion; in, τ=T(γ)κ(n)u p Where: m represents the mass of the dynamically positioned ship; represents the hydrodynamic derivative of forward acceleration; represents the hydrodynamic derivative of the drift acceleration; Y r represents the coupled velocity hydrodynamic derivative; represents the hydrodynamic derivative of coupled acceleration; x G Indicates the distance between the ship's center of mass and the origin of the attached coordinate system; N r I represents the hydrodynamic derivative of the bow speed; z represents the moment of inertia of a dynamically positioned ship; |·| represents absolute value operation; T(γ) is the position matrix of the actuator; κ(n) is the gain matrix of the actuator; u p To directly control the input vector, u p =[u p1 ,u p2 ,...,u pq ,...,u pQ ] T , where Q is the equivalent thruster, i.e. the number of actuators, and u pq =p q |p q | represents the direct control input vector of the qth equivalent thruster; p q is the pitch ratio of the qth equivalent propeller, q = 1, 2, ..., Q represents the index number of the equivalent propeller; Among them, the direct control input vector u p Obtained according to the following execution performance difference model based on adaptive parameters: Where: k p is the execution effect matrix; p is the intermediate input, is the execution bias vector; Let g p =k(n)k p , Then θ=λsup{μ}l; Where: g p is the effect gain matrix; μ is the deviation gain vector; θ is the scaling vector, λ is the effect compensation matrix, k(n) represents the gain matrix of the actuator; k p is the execution effect matrix; is the execution bias vector; l represents the Q row and column vector whose elements are all 1, where l = (1,1,...,1) T ; sup{μ} means taking the lower limit of the elements of μ; Then, the adaptive parameters are introduced to estimate the effect compensation matrix λ and the scaling vector θ, so the intermediate input γ p for: Where: γ p is the intermediate input of the design, are the adaptive parameters of the elements of the 1st, 2nd, …, qth, …, Qth equivalent actuators in the effect compensation matrix λ, respectively. are the adaptive parameters of the elements of the 1st, 2nd, …, qth, …, Qth equivalent actuators in the scaling vector θ respectively; Q is the number of equivalent thrusters; is the pseudo-inverse matrix of T(γ); T(γ) is the position matrix of the actuator; α up represents the actual control law vector; Θ represents the hyperbolic tangent vector; T1(γ) represents the first column vector of T(γ); T q (γ) represents the qth column vector of T(γ); ε represents the hyperbolic tangent epsilon; Ξ e represents the velocity error vector; X e =b Ξ -X Where: e represents the trigger speed error vector; β Ξ represents a first-order filter; Ξ represents a velocity vector; Among them, the following actuator correlation coefficient adaptive law is established to obtain the adaptive parameters of the elements of the effect compensation matrix λ and the scaling vector θ: Where: λq , Γ θq , σ λq , σ θq These are design parameters. They are and The initial value of They are and The first derivative of ; express The elements of rows q and columns b of T aq (γ) represents the a-row, q-column element of T(γ), express The element in row a and column q, Ξ ea Indicates Ξ e The ath element of upb Represents α up The bth element of λq represents the λ gain parameter of the qth equivalent thruster, Γ θq represents the θ gain parameter of the qth equivalent thruster, σ λq represents the λ convergence parameter of the qth equivalent thruster, σ θq represents the θ convergence parameter of the qth equivalent thruster.
4. A ship dynamic positioning event triggering control method considering execution performance differences according to claim 1, characterized in that: The error trigger value is obtained as follows: First, obtain the system output error signal vector η e as follows: or e =(x e (t),y e (t),ψ e (t)) T Where: x e (t),y e (t),ψ e (t) respectively represent the abscissa error, ordinate error and heading error of the ship's position in the inertial coordinate system at time t; Respectively expressed in t k -t k+1 The trigger value of the abscissa error, the trigger value of the ordinate error and the trigger value of the heading error of the ship's position within the interval; and They respectively represent the triggering time of the abscissa error, the triggering time of the ordinate error and the triggering time of the heading error of the ship's position; k represents the index number of the time; η e Represents the system output error signal vector; The error trigger value for: Where: Indicates the error trigger value; K T represents the trigger matrix, K T =diag{1+λ 1x δ x ,1+λ 1y δ y ,1+λ 1ψ δ ψ }; diag{·} means taking diagonal operation; λ 1x ,λ 1y ,λ 1ψ ,λ 2x ,λ 2y ,λ 2ψ All are trigger boundary parameters; x represents the forward trigger increment threshold; y Indicates the incremental threshold for triggering horizontal drift; ζ ψ Indicates the heading trigger increment threshold.
5. The method for triggering control of ship dynamic positioning events considering execution performance differences according to claim 1, characterized in that: The triggering rules of the multi-port event triggering mechanism are as follows: Where: inf{·} represents the lower bound operation; δ x Indicates the forward trigger coefficient threshold; δ y Indicates the drift trigger coefficient threshold; δ ψ Indicates the threshold of the yaw trigger coefficient; ζ x represents the forward trigger increment threshold; y Indicates the incremental threshold for triggering horizontal drift; ζ ψ Indicates the heading trigger increment threshold.
6. A ship dynamic positioning event triggering control method considering execution performance differences according to claim 1, characterized in that: The virtual control law is constructed as follows: Where: α Ξ represents the virtual control law; R -1 (ψ) is the inverse matrix of the rotation matrix; K η is the attitude calming matrix; Indicates the error trigger value; According to the virtual control law α Ξ , based on the dynamic surface technique, determine the first-order filter β Ξ , the formula used is as follows: β Ξ (0)=α Ξ (0) Where: t Ξ is the time constant matrix, where t Ξ =diag{t1,t2,t3}, diag{·} represents diagonal operation, t1, t2, t3 represent the first component, the second component and the third component on the diagonal of the time constant matrix respectively; β Ξ represents a first-order filter; Represents β Ξ The first derivative of Ξ (0) represents the initial value of the first-order filter; α Ξ (0) represents the initial value of the virtual control law.
Citation Information
Cited By
Ship dynamic positioning anti-interference control method considering vertical dynamic optimization
CN122284342A
A ship dynamic positioning anti-disturbance control method considering vertical dynamic optimization
CN122284342B
