Full-drive ship dynamic positioning control method

Abstract

The invention discloses a dynamic positioning control method for a full-drive ship. The dynamic positioning control method comprises the following steps: (1) establishing a three-degree-of-freedom mathematical model of the full-drive ship; (2) based on the three-degree-of-freedom mathematical model, establishing a ship observer kinetic model, fusing data of position, speed and acceleration sensors, and outputting observation values including position, speed, acceleration and low-frequency load components; (3) designing an acceleration feedforward controller based on the state variable observation value, and designing an acceleration feedback controller based on a low-pass filter; and (4) calculating control input by adopting a composite PID (Proportion Integration Differentiation) controller combined with acceleration feedforward and acceleration feedback to realize full-drive ship dynamic positioning control. According to the invention, the resistance of the ship to external environment interference can be enhanced, and the control precision is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ship motion control, in particular to a full-drive ship power positioning control method. BACKGROUND

[0002] With the decrease of sea ice coverage, the ship navigation activities in polar regions are expanding, and some ships are required to have positioning control capability. The widely used positioning methods mainly include mooring system, dynamic positioning system and combination of the two. The dynamic positioning system is a process of achieving fixed position of the ship by receiving real-time sensing information of the ship position, heading, etc., calculating the thrust by using automatic control algorithm and automatically resisting environmental interference, which has the advantages of high positioning accuracy and strong flexibility.

[0003] Due to the harsh environment in the Arctic region, higher requirements are put forward for the propulsion system of polar ships, one of which is a full-drive ship with an omnidirectional pod system, which can realize 360° free rotation and quickly apply propulsion in any direction. The full-drive ship has the advantages of simple model, easy analysis and strong maneuvering performance, so it is of great significance to study the dynamic positioning of the full-drive ship.

[0004] Many researchers have explored various control methods to solve the dynamic positioning control problem, such as traditional PID controller combined with filter, control technology based on Kalman filter theory and sliding mode control technology. With the progress of inertial measurement unit (IMU), a control strategy using acceleration measurement has been proposed.

[0005] The Chinese patent document with publication number CN111427269A discloses a dynamic positioning model test control method based on fuzzy PID control, which sets a wave height instrument at the center of gravity position of the ship to monitor the wave height in real time; the ship position is measured by a position measurement system, and the position information is sent to a filter to estimate the low-frequency position and low-frequency speed of the ship, which is further sent to a central PID controller to calculate the required control force, and finally the thrust distribution module is used to distribute the thrust to each propeller of the dynamic positioning barge.

[0006] The Chinese patent document with publication number CN103217160A discloses a filtering method in a ship dynamic positioning system, which includes: (1) measuring the position and acceleration of the ship and converting the data to a unified coordinate system; (2) establishing a ship model containing high-frequency acceleration; (3) establishing a Kalman filter that can estimate position, speed and acceleration; (4) based on the measured ship position and acceleration, using the Kalman filter to filter out high-frequency interference and obtaining the low-frequency ship position, speed and acceleration required by the dynamic positioning system of the ship, and sending the data to the dynamic positioning system.

[0007] However, most of the current researches mainly focus on open water, and for dynamic positioning control in ice environment, people pay more attention to the influence of DP ship-ice interaction and ice management operation, and less attention to DP control algorithm. SUMMARY

[0008] The application provides a full-drive ship dynamic positioning control method, which can enhance the resistance of the ship to external environmental interference and improve the control precision.

[0009] A full-drive ship dynamic positioning control method, comprising:

[0010] (1) establishing a three-degree-of-freedom mathematical model of the full-drive ship;

[0011] (2) based on the three-degree-of-freedom mathematical model, establishing a ship observer dynamics model, fusing position, speed, acceleration sensor data, and outputting state variable observation values, including position, speed, acceleration and low-frequency estimated ice load observation values;

[0012] (3) based on the state variable observation values, designing an acceleration feedforward controller, and based on a low-pass filter, designing an acceleration feedback controller;

[0013] (4) using a composite PID controller combining acceleration feedforward and acceleration feedback, calculating control input, and realizing full-drive ship dynamic positioning control.

[0014] The application obtains ship state variable estimated values based on a state observer and combines acceleration direct measurement to obtain high and low frequency estimated load values; a composite controller is designed combining acceleration feedforward, acceleration feedback and PID control algorithm, and control resultant force is outputted to realize ship positioning in a predetermined safe area.

[0015] In step (1), the three-degree-of-freedom mathematical model of the full-drive ship is established, and the expression is:

[0016]

[0017] In the formula, η = [x, y, ψ] T represents the position and yaw angle of the ship in the fixed coordinate system; R(η) represents the coordinate conversion matrix between the world coordinate system and the body coordinate system; V = [u, v, r] T , u represents the surge linear velocity of the ship in the body coordinate system, v represents the sway linear velocity of the ship in the body coordinate system, and r represents the yaw angular velocity of the ship in the body coordinate system; M represents the mass inertia matrix; D represents the hydrodynamic damping matrix; τ ∈ R 3 represents the control input; b ∈ R 3 represents all slow-changing environmental disturbance models; τ ice_h ∈ R3 represents high-frequency ice load.

[0018] R(η), M, D are given by:

[0019]

[0020] where ψ is the yaw angle of the ship in the fixed coordinate system; m is the mass of the ship; I z is the moment of inertia about the ship axis z o ; X u , Y v , N v are linear fluid dynamic damping coefficients; r , Y r , N b are linear fluid added mass coefficients.

[0021] The slowly varying environmental disturbance model b is given by:

[0022]

[0023] where w b ∈ R3 is a zero-mean Gaussian white noise vector; T 3×3 ∈ R b is a time constant diagonal matrix; E 3×3 ∈ R ice_h is a diagonal scaling matrix.

[0024] The expression of the high frequency ice load τ ice_h is given by:

[0025]

[0026] τ e = Γτ e

[0027] Γ = [0 I3]

[0028] where τ 6×6 ∈ R6 is the designed simulated load; ω ∈ R6 is a zero-mean white noise; Ω ∈ R 6×6 , ∑ ∈ R f are positive gain matrices; I3 denotes the identity diagonal matrix.

[0029] In step (2), the ship observer dynamics model is established, which is given by:

[0030]

[0031] where x denotes the ship position and heading estimates; denotes the ship linear and angular velocity estimates; denotes the position and heading observation errors.V, ω represent ship linear and angular velocity observation errors; V, ω represent ship linear and angular velocity observation errors; V, ω represent ship linear and angular velocity observation errors;a f V, ω represent ship linear and angular velocity observation errors;

[0032] The low frequency ice load estimate is described as:

[0033]

[0034] where, V, ω represent ship linear and angular velocity observation errors;k a V, ω represent ship linear and angular velocity observation errors.

[0035] The high frequency ice load estimate is described as:

[0036]

[0037] where, V, ω represent ship linear and angular velocity observation errors; V, ω represent ship linear and angular velocity observation errors;K2, K3 are design parameters.

[0038] In step (3), the acceleration feedforward controller is designed based on the state variable observations and is represented as:

[0039]

[0040] where, ff ∈ R 3 V, ω represent ship linear and angular velocity observation errors.

[0041] In step (3), the acceleration feedback controller is designed and is represented as:

[0042]

[0043]

[0044] where, fb ∈ R 3 V, ω represent ship linear and angular velocity observation errors;K a , T f are design parameters; s is a complex variable.

[0045] In step (4), the compound PID controller is represented as:

[0046] τ = τ PID - τ fb - K ff τ ff

[0047] The control algorithm for the compound PID controller is:

[0048]

[0049] In the formula, tau PID ∈R 3 is the PID control input; eta e represents the error between the position and heading observation values and target values; V e represents the error between the linear velocity and angular velocity observation values and target values; K ff , K p , K i , K d are design parameters.

[0050] Compared with the prior art, the present application has the following beneficial effects:

[0051] 1. The traditional model filters waves, and can separate high-frequency and low-frequency components in waves. According to the characteristics of ice load action, the present application finds that the ice force spectrum is similar to the wave spectrum form, and analyzes the ice force by referring to the Neumann spectrum, and constructs a second-order ice load model.

[0052] 2. The present application constructs a filter observer, and innovatively integrates the data of position, velocity and acceleration sensors, and the observer can finally output the position, velocity, acceleration and low-frequency components in the ice load parameters.

[0053] 3. The present application feeds back the low-frequency ice load signal to the actuator through a feedforward path, so as to compensate for the low-frequency force in advance, offset the dynamic delay of the system, and improve the control precision.

[0054] 4. The present application adopts acceleration feedback, feeds back the low-frequency ice load signal to the actuator through a feedback path, and enhances the resistance of the ship to external environmental interference. BRIEF DESCRIPTION OF DRAWINGS

[0055] Figure 1 It is a flow chart of a full-drive ship dynamic positioning control method according to an embodiment of the present application.

[0056] Figure 2 It is a full-drive ship motion model diagram according to the present application.

[0057] Fig. 3(a) is a comparison between the actual position and the estimated position of the ship in the x direction.

[0058] Fig. 3(b) is a comparison between the actual position and the estimated position of the ship in the y direction.

[0059] Fig. 3(c) is a comparison between the actual heading and the estimated heading of the ship.

[0060] Fig. 3(d) is a comparison between the actual low-frequency ice load and the estimated low-frequency ice load.

[0061] Figure 3(e) is a plot of high frequency actual ice load versus high frequency estimated ice load.

[0062] Figure 4 is a plot of position and heading angle change of the present invention.

[0063] Figure 5 is a plot of position and heading angle error curve simulation of the present invention.

[0064] Figure 6 is a plot of control input curve simulation of the present invention. DETAILED DESCRIPTION

[0065] The present invention will be further described in conjunction with the accompanying drawings and examples, it should be noted that the following examples are intended to facilitate the understanding of the present invention and do not limit the present invention in any way.

[0066] As shown in Figure 1, a full drive ship power positioning control method mainly includes three parts of ship and disturbance model establishment, design of state observer and design of compound controller. Figure 1 The present invention studies the problem of power positioning control of full drive ship, as shown in Figure 2, the three-degree-of-freedom dynamic model of full drive ship is expressed as:

[0067] Figure 2

[0068]

[0069] In the formula, η = [x, y, ψ] T represents the position and yaw angle of the ship in the fixed coordinate system; R(η) represents the coordinate conversion matrix between the world coordinate system and the body coordinate system; V = [u, v, r] T represents the linear velocity and angular velocity of the ship in the body coordinate system; M represents the mass inertia matrix; D represents the hydrodynamic damping matrix; τ ∈ R 3 represents the control input; b ∈ R 3 represents all slow-varying environmental disturbances; τ ice_h ∈ R3 represents high-frequency ice load.

[0070] In the formula, R(η), M and D are expressed as:

[0071]

[0072]

[0073] The slow-varying environmental disturbance model in the present invention is expressed as:

[0074]

[0075] In the formula, w b ​​∈R3 is a zero-mean Gaussian white noise vector; T b ∈R 3×3 is a time constant diagonal matrix; E b ∈R 3×3 is a diagonal scaling matrix.

[0076] By the ice pressure analysis and the ice force analysis based on Bohai JZ-20 platform of the ice pressure analysis of the pile leg of the offshore platform in Bohai Sea, it is found that the ice force spectrum is close to the wave spectrum, and the ice load action characteristic state equation is designed based on the Neumann spectrum in the sea wave spectrum, and is described as:

[0077]

[0078] τ ice_h =Γτ e

[0079] Γ=[0 I3]

[0080] In the formula, τ e ∈R6 is a designed simulated ice load; ω∈R6 is a zero-mean white noise; Ω∈R 6×6 , ∑∈R 6×6 is a positive gain matrix.

[0081] The ship power positioning observer model is designed, and is described as:

[0082]

[0083] In the formula, represents the ship position and heading estimation value; represents the ship linear velocity and angular velocity estimation value; represents the position and heading observation error; represents the linear velocity and angular velocity observation error; represents the high-frequency estimated ice load; represents the slowly changing estimated load; a f represents the estimated acceleration.

[0084] The low-frequency estimated ice load based on the acceleration measurement is described as:

[0085]

[0086] The high-frequency estimated ice load is described as:

[0087]

[0088] In the formula, represents the observation value of the simulated load; represents the high-frequency load observation error.

[0089] The acceleration feedforward controller used in the application is described as:

[0090]

[0091] In the formula, τ ff ∈R 3 represents the acceleration feedforward control input.

[0092] The acceleration feedback controller used in the application is described as:

[0093]

[0094] In the formula, τ fb ∈R 3 is the acceleration feedback control input.

[0095] The PID control algorithm designed in the application is described as:

[0096]

[0097] In the formula, τ PID ∈R 3 is the PID control input; η e represents the error between the position and heading observation value and the target value; V e represents the error between the linear velocity and angular velocity observation value and the target value.

[0098] The composite controller designed in the application is described as:

[0099] τ = τ PID - τ fb - K ff τ ff

[0100] The core content of the application is: a complex disturbance load model containing high and low frequencies is established; a dynamic positioning state observer is designed based on direct acceleration measurement, to obtain the low-frequency state variable of the ship and the high and low frequency disturbance estimation value. An acceleration feedforward controller, an acceleration feedback controller and a PID controller are designed and combined to calculate the control force.

[0101] In summary, the dynamic positioning control algorithm for the fully driven ship proposed in the application obtains the state variable estimation value of the ship and the high and low frequency estimated load value based on the state observer; a composite PID controller is designed by combining the acceleration feedforward, acceleration feedback and PID control algorithms, to output the control force, so as to realize the positioning of the fully driven ship in the predetermined safe area.

[0102] To verify the effectiveness of the composite controller designed in the application, a simulation experiment is performed.

[0103] Table 1 gives the model parameters of the fully driven ship, as follows:

[0104] Table 1 Full-propulsion vessel model parameters

[0105]

[0106] Table 2 gives the full-propulsion vessel structure parameters and disturbance model related parameters as follows:

[0107] Table 2 Full-propulsion vessel structure parameters and disturbance model related parameters

[0108]

[0109] The observer and controller related parameters of the present application are designed as follows:

[0110] omegaoi = 0.8976; Ratio_omegaoi = 1; k 31 = 1 - ratio_omegaoi 2 ; ki = 0.175; k 21 = ki * (1 - k 31 ); k 22 = 10 * (-k 21 * omegaoi 2 ); k 32 = 0.7; k a = 4; Ki = ki * diag([1 1 1]); K2 = [k 21 * eye(3); k 22 * eye(3)]; K3 = [k 31 * eye(3); k 32 * eye(3)]; K4 = diag [25 25 25]; K = diag [25 25 25]; Omega = [zeros(3) eye(3); -omegaoi 2 * eye(3) - 2 * omegaoi * 0.1 * eye(3)]; K p = diag([57.75 57.75 75]) * le4; K i = diag([0 0 0]); K d = diag([78 78 300]) * le5; K a = diag([5.3122e6 8.2831e6 3.7454e9]) * 0.5; T f = 5 * eye(3); K ff = diag([1 1 1]).

[0111] Fig. 3(a) is a comparison between the actual and estimated positions of the ship in the x direction; Fig. 3(b) is a comparison between the actual and estimated positions of the ship in the y direction; Fig. 3(c) is a comparison between the actual and estimated headings of the ship; Fig. 3(d) is a comparison between the actual and estimated low-frequency ice loads; and Fig. 3(e) is a comparison between the actual and estimated high-frequency ice loads. As can be seen from the figures, the errors between the observed and actual values of the position and heading angle are small, and the high and low frequency loads are separated, and the observer is effective.

[0112] Figure 4 Fig. 4 is a position and heading angle variation diagram of the present application, and as can be seen from the figure, the position and heading angle of the ship gradually tend to the set target values and are finally stabilized around the target values.

[0113] Figure 5 Fig. 5 is a position and heading angle error curve simulation diagram of the present application, and as can be seen from the figure, the position and heading angle tracking error curve of the ship gradually tends to 0 and maintains a small steady-state error.

[0114] Figure 6 Fig. 6 is a control input curve simulation diagram of the present application, and as can be seen from the figure, the control signal is calculated by the algorithm to obtain the corresponding control force and control torque, which is rapidly increased in the initial stage, then gradually decreased, and finally stabilized and adjusted to ensure the stable convergence of the system error. Although there are uncertainties and disturbances, the dynamic positioning tracking error of the ship still converges to zero and is always maintained within the preset position safety area, proving the effectiveness of the present application.

[0115] The above-described embodiments have described the technical solutions and beneficial effects of the present application in detail, and it should be understood that the above-described only the specific embodiments of the present application, and is not intended to limit the present application, any modification, supplement and equivalent replacement made within the principle range of the present application should be included in the protection scope of the present application.

Claims

1. A method of dynamically positioning a fully driven marine vessel, characterized by, Comprise: (1) Establish a three-degree-of-freedom mathematical model of the fully-driven ship; (2) Based on the three-degree-of-freedom mathematical model, a ship observer dynamics model is established, which fuses position, velocity, acceleration sensor data, and outputs state variable observations, including position, velocity, acceleration and low-frequency estimated ice load observations; (3) Based on the state variable observations, an acceleration feedforward controller is designed, and an acceleration feedback controller is designed based on a low-pass filter; (4) A compound PID controller combining acceleration feedforward and acceleration feedback is used to calculate the control input, realizing dynamic positioning control of the fully-driven ship.

2. The fully-driven vessel dynamic positioning control method of claim 1, wherein, In step (1), a three-degree-of-freedom mathematical model of the fully-driven ship is established, and the expression is: where η = [x, y, ψ] T denotes the position and yaw angle of the ship in the fixed coordinate system; R(η) denotes the coordinate transformation matrix between the world coordinate system and the body coordinate system; V = [u, v, r] T denotes the surge linear velocity of the ship in the body coordinate system, v denotes the sway linear velocity of the ship in the body coordinate system, and r denotes the yaw angular velocity of the ship in the body coordinate system; M denotes the mass inertia matrix; D denotes the hydrodynamic damping matrix; τ ∈ R 3 denotes the control input; b ∈ R 3 denotes all slow-varying environmental disturbance models; τ ice_h ∈ R3denotes the high-frequency ice load.

3. The fully-driven vessel dynamic positioning control method of claim 2, wherein, The specific expressions of R(η), M, and D are: where ψ is the yaw angle of the ship in the fixed coordinate system; m is the mass of the ship; I z is the moment of inertia about the ship axis z o ; X u , Y v , N v , Y r , N r are linear hydrodynamic damping coefficients; are linear hydrodynamic added mass coefficients.

4. The fully-driven vessel dynamic positioning control method of claim 2, wherein, The expression of the slowly changing environmental disturbance model b is: where w b ∈ R3 is a zero-mean Gaussian white noise vector; T b ∈ R 3×3 is a diagonal matrix of time constants; E b ∈ R 3×3 is a diagonal scaling matrix; High frequency ice load τ ice_h The expression is: τ ice_h = Γτ e Γ = [0 3×3 I3] where τ e ∈ R6is the designed simulation load; ω ∈ R6is a zero-mean white noise; Ω ∈ R 6×6 ,∑ ∈ R 6×6 is a positive gain matrix; I3denotes a unit diagonal matrix.

5. The fully-driven vessel dynamic positioning control method of claim 2, wherein, In step (2), the ship observer dynamics model is established, and the expression is: where represents the ship position and heading estimate; represents the ship linear and angular velocity estimate; represents the position and heading observation error; represents the ship linear and angular velocity observation error; represents the high frequency estimated ice load; represents the slowly varying estimated load;a f represents the estimated acceleration; K, K1, K4, K5, k represent design parameters.

6. A method of dynamic positioning control of a fully driven marine vessel according to claim 5, characterized in that, The low-frequency estimated load is described as: wherein is the low frequency estimated load, k a is a design parameter.

7. The fully-driven vessel dynamic positioning control method of claim 5, wherein, The high-frequency estimated ice load is described as: wherein represents the observed value of the simulated load; represents the high-frequency load observation error; K2 and K3 are design parameters.

8. The fully-driven vessel dynamic positioning control method of claim 2, wherein, In step (3), an acceleration feedforward controller is designed based on the state variable observations, and is expressed as: where τ ff ∈ R 3 denotes the acceleration feedforward control input.

9. The fully-driven vessel dynamic positioning control method of claim 2, wherein, In step (3), an acceleration feedback controller is designed, and is expressed as: where τ fb ∈ R 3 is the acceleration feedback control input; K a , T f is a design parameter; s is a complex variable.

10. A method of dynamic positioning control of a fully driven marine vessel according to claim 9, characterized by, In step (4), the compound PID controller is expressed as: τ = τ PID -τ fb -K ff τ ff The control algorithm of the compound PID controller is: where τ PID ∈ R 3 is the PID control input; η e denotes the error between the position and heading observations and the target values; V e denotes the error between the linear and angular velocity observations and the target values; K ff , K p , K i , K d are design parameters.

Images