Method for estimating motion state of floating type wind turbine

Abstract

The invention discloses a floating type wind turbine motion state estimation method, and belongs to the technical field of marine floating type wind turbine motion response, and the method comprises the following steps: S1, constructing a two-dimensional simplified dynamic model; the method specifically comprises the following steps: S1.1, constructing a dynamic architecture of the wind turbine unit; s1.2, analyzing aerodynamic force, aerodynamic torque, still water recovery torque, hydrodynamic force, hydrodynamic torque and mooring torque; s1.3, obtaining an overall simplified dynamic model of the wind turbine unit; s2, optimizing the additional mass parameter matrix A, the damping coefficient matrix B and the rigidity coefficient matrix C, and respectively recording as Abest, Bbest and Cbest; and S3, estimating the surging motion displacement and the pitching angle of the wind turbine at the future moment by adopting the Abest, the Bbest and the Cbest based on a Kalman filter. The method can guarantee the prediction precision while improving the calculation efficiency.

Description

Technical Field

[0001] The invention belongs to the technical field of motion response of offshore floating wind turbines, and in particular relates to a method for estimating the motion state of a floating wind turbine. Background Art

[0002] Offshore wind power is a renewable energy source with huge development potential. Currently, floating offshore wind power technology is regarded as the main technology for future deep-sea offshore wind power development. Since there is no fixed connection between the bottom of the floating offshore wind turbine and the seabed, its dynamic response is more significant. Stable and accurate estimation of the wind turbine motion response in complex wind and wave environments is of great significance to unit control, performance estimation and prediction.

[0003] Existing methods often require more iterations to approach the optimal solution, especially when the objective function is complex and there are multiple local optimal points. It is easy to fall into the local optimal solution, which is not enough to take into account both the accuracy of the simplified dynamic model and the computational efficiency of the operating response. In the face of complex wind and wave environments, existing methods find it difficult to maintain a stable estimate of the motion state response of the wind turbine. Summary of the invention

[0004] A floating wind turbine motion state estimation method of the present invention can maintain stable estimation of the wind turbine motion state response.

[0005] A floating wind turbine motion state estimation method of the present invention comprises the following steps:

[0006] S1: Based on the dynamics framework of floating wind turbines, a simplified two-dimensional dynamic model for surge and pitch is constructed;

[0007] The specific steps include:

[0008] S1.1: Simplify the wind turbine into a rigidly connected mast buoy, tower, impeller and nacelle; simplify the mooring system into a linear spring; construct the dynamic architecture of the wind turbine;

[0009] S1.2: Analyze aerodynamic forces, aerodynamic moments, hydrostatic restoring moments, hydrodynamic forces, hydrodynamic moments, and mooring moments;

[0010] S1.3: Based on the law of rigid body dynamics, combined with the aerodynamic force, aerodynamic moment, hydrostatic restoring moment, hydrodynamic force, hydrodynamic moment and mooring moment in S1.2; set the wind turbine mass matrix M, additional mass parameter matrix A, damping coefficient matrix B and rigidity coefficient matrix C; obtain the overall simplified dynamic model of the wind turbine;

[0011] S2: Based on OpenFAST, the fitness function and constraint conditions are established to optimize the additional mass parameter matrix A, the damping coefficient matrix B and the stiffness coefficient matrix C. The optimized matrices are respectively denoted as A best , B best and C best ;

[0012] S3: Adopt A best , B best and C best , Based on the Kalman filter, the longitudinal motion displacement and pitch angle of the wind turbine are estimated at the future moment.

[0013] Furthermore, in step S1.3, the wind turbine mass matrix M is:

[0014]

[0015] M 11 =m

[0016] M 12 =M 21 =m·z total

[0017] M 22 =I o

[0018]

[0019] Where m is the mass of the wind turbine; Z total is the center of mass position of the wind turbine; I floater is the moment of inertia of the mast buoy; I tower is the moment of inertia of the tower; M floater is the mass of the mast buoy; M tower is the mass of the tower; M nacelle is the impeller mass; M rotor for the cabin quality;

[0020] set up

[0021] The overall simplified dynamic model of the wind turbine is:

[0022]

[0023] Where x represents the wind turbine's longitudinal displacement; θ is the wind turbine's pitch angle; F Hydrodynamic F is water power; wind is the aerodynamic force; T Hydrodynamic is the hydrodynamic torque; T Wind is the aerodynamic torque.

[0024] Furthermore, the formulas of fitness function and constraints are:

[0025]

[0026] In the formula, x s (i) is the sloshing displacement output by the simplified dynamic model at time i;

[0027] x r (i) is the surge displacement of the wind turbine at time i under OpenFAST analysis;

[0028] θ s (i) is the pitch motion angle of the wind turbine output by the simplified dynamic model at time i;

[0029] θ r (i) is the pitch motion angle of the wind turbine at time i under OpenFAST analysis;

[0030]

[0031] is the data simulated by OpenFAST based on different wind turbine structural parameters, wind and wave environment parameters and wind turbine initial state; through the iterative convergence of formula (1) and (2), the optimized A is assumed best , B best and C best for:

[0032] and

[0033] Step S3 includes the following steps:

[0034] Further,

[0035] S3.1: Convert equation (1) into a simplified state space model, the formula is:

[0036]

[0037] S3.2: Set the actual wind turbine oscillation displacement and wind turbine pitch angle output equations as:

[0038]

[0039] S3.3: Establish the Kalman filter equation:

[0040]

[0041] In the formula,

[0042] G k =e ZT ,H k=∫0 T e Z(tx-t) Ld τ ,N k =C best

[0043] T is the discrete period, tx is the simulation time, τ is the time from the beginning to time k; x k is the system state variable at time k; w k is the process noise vector at time k; v k is the measurement noise vector at time k; u k is the system input at time k; u k-1 is the system input at time k-1;

[0044]

[0045] In the formula, Q k w k The covariance matrix of k is the measurement noise vector v k The covariance matrix of

[0046] State update equation:

[0047]

[0048] Error gain equation:

[0049]

[0050] Among them, the actual measured output at time k is y true,k , calculated by formula (3);

[0051] Update the covariance matrix:

[0052]

[0053] I is the unit matrix; the Kalman equation is used to estimate the wind turbine's longitudinal motion displacement x and longitudinal angular velocity θ at a certain moment.

[0054] Furthermore, the calculation formulas for aerodynamic force and aerodynamic moment are:

[0055]

[0056]

[0057] In the formula, F wind is the aerodynamic force; ρ is the air density; V total is the relative wind speed of wind turbine blades; T Wind is the aerodynamic torque; is the inflow angle; Cl is the lift coefficient; C d is the gas drag coefficient; r is the distance from the blade element to the center of the wind turbine hub.

[0058] Furthermore, the calculation formulas for the hydrostatic restoring moment, hydrodynamic force and hydrodynamic moment are:

[0059]

[0060] Where, T Buoy is the total hydrostatic restoring moment of the wind turbine; m is the total mass of the wind turbine; g is the acceleration due to gravity; Z b is the buoyancy center of the wind turbine; Z g is the gravity center of the wind turbine; θ is the pitch motion angle of the wind turbine; ρ1 is the water density where the wind turbine is located; is the additional inertia of the wind turbine, D is the diameter of the mast buoy;

[0061]

[0062] Where, T Hydrodynamic is the hydrodynamic moment; w is the frequency of regular waves; u 水 The wave speed of the wave; C m is the inertia coefficient; C D is the water resistance coefficient; z is the height of the wind turbine; A1 is the end surface area of ​​the mast buoy; F Hydrodynamic For water power.

[0063] Furthermore, in the formula, F moor is the horizontal mooring force; K moor Wind turbine horizontal mooring stiffness coefficient matrix; Z moor is the distance from the mooring anchor point to the origin of the local coordinate system in the local coordinate system; x0 is the longitudinal displacement of the origin of the local coordinate system.

[0064] Beneficial effects:

[0065] 1. This scheme adopts a simplified model combined with Kalman filtering to predict pitch and surge. Because of the simplified model, the amount of calculation is greatly reduced.

[0066] 2. Compared with the existing prediction model, the prediction accuracy of this scheme can remain basically consistent with the existing model. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 It is a schematic diagram of the dynamic structure of a floating wind turbine;

[0068] Figure 2 It is the free decay characteristic diagram of surge and pitch in the free decay test based on OpenFAST and simplified model;

[0069] Figure 3 This is a comparison of the surge and pitch responses between the simplified model and OpenFAST when there is wind but no waves;

[0070] Figure 4 This is a comparison of the surge and pitch responses between the simplified model and OpenFAST when there are waves but no wind;

[0071] Figure 5 It is a comparison chart of surge and pitch response between the simplified model and OpenFAST under the combined wind and wave (regular) condition.

[0072] Figure 6 It is a comparison chart of state estimation between the simplified model and OpenFAST when the Kalman filter is added under the combined wind and wave (regular) condition. DETAILED DESCRIPTION

[0073] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. 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.

[0074] A method for estimating the motion state of a floating wind turbine comprises the following steps:

[0075] S1: Based on the dynamics framework of floating wind turbines, a simplified two-dimensional dynamic model for surge and pitch is constructed;

[0076] The specific steps include:

[0077] S1.1: Dynamic architecture of floating wind turbine, i.e., simplify the wind turbine into a rigidly connected mast buoy, tower, impeller and nacelle; simplify the mooring system into a linear spring; construct the dynamic architecture of the wind turbine, see Figure 1 .

[0078] S1.2: Analyze the aerodynamic forces and moments of the wind turbine dynamic structure.

[0079] The specific steps include:

[0080] S1.2.1: Calculate the turbulence intensity S(f) at the hub of the wind turbine and calculate the time domain wind speed curve V based on the turbulence intensity S(f); the formula is:

[0081]

[0082] Where S(f) is the turbulence intensity at the hub of the wind turbine, which is the quantity to be determined; σ is the turbulence density, which is a known quantity; V hub is the wind speed at the hub of the wind turbine, which is a known quantity; f is the wind speed frequency, which is a known quantity; L is the turbulence integral scale, which is a known quantity;

[0083] According to the turbulence intensity S(f), the time domain wind speed curve V is calculated; the formula is:

[0084]

[0085] Among them, ξ is the phase;

[0086] S1.2.2: Based on the blade element momentum theory, solve the tangential and axial induction coefficients, and calculate the aerodynamic force and aerodynamic moment at the hub of the wind turbine based on the tangential and axial induction coefficients.

[0087] The tangential and axial induction coefficients are:

[0088]

[0089] In the formula, a b 、a' b , C x , C y is the intermediate parameter, which is a known quantity; α is the axial induction coefficient, which is a quantity to be determined; σ R is the turbulence intensity at the airfoil radius R, which is a known quantity; α' is the tangential induction coefficient, which is a quantity to be determined; Φ is the angle of attack, which is a known quantity;

[0090] a b =a / f,a' b =a' / f,

[0091] C x =C l cosΦ+C d sinΦ

[0092] C y =C l sinΦ-C d cosΦ

[0093] In the formula, C d is the gas resistance coefficient; C l is the lift coefficient; Φ is the angle of attack, which is a known quantity;

[0094] According to the axial induction coefficient α and the tangential induction coefficient α', the inflow angle is calculated The formula is:

[0095]

[0096] Where U∞ is the wind speed at infinity; Ω is the rotor speed; r is the distance from the blade element to the center of the wind turbine hub; λ r is the local tip speed ratio;

[0097] According to the inflow angle Calculate the relative wind speed V of wind turbine blades total , the formula is:

[0098]

[0099] Among them, V is the time domain wind speed curve; is the inflow angle.

[0100] According to the relative wind speed V of the wind turbine blades total , the aerodynamic force and aerodynamic moment on the blade element of the wind turbine are calculated:

[0101]

[0102] In the formula, F wind is the aerodynamic force; ρ is the air density, which is a known quantity; V total is the relative wind speed of wind turbine blades; T Wind is the aerodynamic torque; is the inflow angle; C l is the lift coefficient; C d is the gas drag coefficient; r is the distance from the blade element to the center of the wind turbine hub.

[0103] S1.3: Analyze the hydrostatic and hydrodynamic forces in a two-dimensional simplified dynamics framework;

[0104] The specific steps include:

[0105] S1.3.1: Calculate the hydrostatic forces for a two-dimensional simplified dynamics framework; the hydrostatic forces include the restoring moments due to gravity and buoyancy and are given by:

[0106] T Buoy = mg(Z b -Z g )sin(θ)

[0107] Where, T Buoy is the hydrostatic restoring moment of the wind turbine; m is the total mass of the wind turbine, which is a known quantity; g is the acceleration due to gravity; Z b is the buoyancy center of the wind turbine, which is a known quantity; Z g is the center of gravity of the wind turbine, which is a known quantity; θ is the pitch motion angle of the wind turbine;

[0108] Assuming that the wind turbine is in a small motion, let sin(θ) = θ, then:

[0109] The total hydrostatic restoring moment of the wind turbine is:

[0110]

[0111] ρ1 is the water density where the wind turbine is located; is the additional inertia of the wind turbine; D is the diameter of the mast buoy.

[0112] S1.3.2: Calculate the hydrodynamic forces of a two-dimensional simplified dynamical framework;

[0113] Considering the waves as regular waves, the wave velocity and acceleration are calculated using the following formula:

[0114] u 水 =wh 水 / 2cosh 水 (k((Z b +h 水 )) / sinh 水 (kh 水 )cos(wt)

[0115] a 水 =-w 2 h 水 / 2cosh 水 (k((Z b +h 水 )) / sinh 水 (kh 水 )sin(wt)

[0116] w=2Πf1

[0117] Where f1 is the regular wave frequency; Z b is the buoyancy center of the wind turbine; w is the frequency of regular waves; t is time; h 水 is the water depth of the water area where the wind turbine is located; k is the wave number of regular waves; u 水 is the wave speed of regular waves; a 水 is the wave acceleration.

[0118] According to the wave speed u of regular waves 水 and the wave frequency w of regular waves, the hydrodynamic force and hydrodynamic moment are calculated using the formula:

[0119]

[0120] x(z)=x0+zθ

[0121]

[0122] Among them, u 水 is the wave speed; C mis the inertia coefficient; C D is the water resistance coefficient; x is the longitudinal displacement of the wind turbine; z is the height of the wind turbine; A1 is the end surface area of ​​the mast buoy; D is the diameter of the mast buoy. Hydrodynamic For water power; T Hydrodynamic is the hydrodynamic moment; x0 is the longitudinal motion displacement of the origin of the wind turbine.

[0123] S1.4: Analyze the mooring forces of the mooring system.

[0124] The specific steps include:

[0125] S1.4.1: Find the distance between the connection point and the anchor chain point;

[0126] l=(X p -X A ),h=(Y p -Y A )

[0127] Where l is the horizontal distance between the connection point and the anchor chain point; h is the vertical distance between the connection point and the anchor chain point; (X p , Y p ) is the coordinate of the connection point in the global coordinate system; (X A , Y A ) is the coordinate of the anchor point in the global coordinate system.

[0128] According to the relationship between force and projection distance, the horizontal force H acting on the foundation platform at the connection point is obtained. 水平 and the vertical force V 垂直 , the formula is:

[0129] l=(H 水平 L0 / EA)+(H 水平 L0 / W)[sinh -1 (V 垂直 / H 水平 )-sinh -1 ((V 垂直 -W) / H 水平 )]

[0130]

[0131] Where, L0 is the elongation length of the catenary spring; W is the unit mass of the catenary spring; EA is the stiffness of the catenary spring;

[0132] Define the wind turbine horizontal mooring stiffness coefficient matrix, the formula is:

[0133]

[0134] Calculate the horizontal mooring force Fmoor : and horizontal mooring moment T moor .

[0135] F moor =-K moor (x0+z moor sin(θ)

[0136] T moor =F moor ·z moor =-K moor (z moor ·θ+x0)·z moor

[0137] Among them, F moor is the horizontal mooring force; K moor The wind turbine horizontal mooring stiffness coefficient matrix; θ is the wind turbine pitching angle; Z moor is the distance from the mooring anchor point to the origin of the local coordinate system in the local coordinate system; x0 is the longitudinal displacement of the origin of the local coordinate system.

[0138] S1.5: Assuming that the main components of the floating wind turbine, the nacelle, the tower and the floating platform are rigidly connected, the law of rigid body dynamics can be used to obtain:

[0139]

[0140]

[0141] in,

[0142] In the formula, F i is the pneumatic, hydrodynamic and mooring force; M i is pneumatic, hydraulic, torque; r i is the distance from the component of the wind turbine to the origin of the local coordinate system; v0 is the velocity at the origin of the local coordinate system; ω is the angular velocity of the origin of the local coordinate system; m is the total mass of the wind turbine; N is the number of the wind turbine force-bearing parts; m i is the mass of each component of the wind turbine; v i is the velocity of the local coordinate system of the center of mass of each component of the wind turbine.

[0143] The total mass of the wind turbine is:

[0144] m=∑M components =M floater +M tower +M nacelle +M rotor ;

[0145] Among them, M floateris the mass of the mast buoy; M tower is the mass of the tower; M nacelle is the impeller mass; M rotor For cabin quality.

[0146] The formula for the location of the center of mass of the system is:

[0147]

[0148] In the formula, z floater is the center of mass position of the mast buoy; z tower is the center of mass position of the tower; z hub is the wheel hub center of mass position.

[0149] The formula for transposing the inertia of the center of mass to another axis parallel to it is:

[0150]

[0151] In the formula, I floater is the moment of inertia of the mast buoy; I tower is the moment of inertia of the tower;

[0152] Write the mass of the wind turbine in matrix form:

[0153]

[0154] M 11 =m

[0155] M 12 =M 21 =m·z total

[0156] M 22 =I o (2);

[0157] Where x0 is the longitudinal displacement of the origin of the local coordinate system, is the velocity vector of the local coordinate system origin, and θ is the pitch motion angle of the wind turbine; is the angular velocity vector.

[0158] Assume that the additional quality parameter matrix A is:

[0159]

[0160] Where ρ0 is the fluid density, C m is the inertia coefficient, D is the diameter of the mast buoy; z bot is the coordinate of the center of mass of the mast buoy.

[0161] Since the floating wind turbine will exert force on the surrounding fluid when it moves, causing the fluid motion state to change, and the change in motion state will produce a damping effect on the mast buoy. Considering this damping characteristic, the damping coefficient matrix B is introduced. Suppose the damping coefficient matrix B is:

[0162]

[0163] Assume the stiffness coefficient matrix C is:

[0164]

[0165] Combining equations (1) to (5), the simplified overall dynamic model of the wind turbine is obtained as follows:

[0166]

[0167] In the formula, F Hydrodynamic F is water power; wind is the aerodynamic force; T Hydrodynamic is the hydrodynamic torque; T Wind is the aerodynamic torque.

[0168] The simplified model includes the wind turbine longitudinal motion displacement x, the wind turbine pitch angle θ, the additional mass parameter matrix A, the damping coefficient matrix B and the stiffness coefficient matrix C.

[0169] S2: Establish a fitness function and optimize the additional mass parameter matrix A, the damping coefficient matrix B and the stiffness coefficient matrix C based on the artificial bee colony algorithm.

[0170]

[0171] Among them, x s (i) is the oscillating motion displacement output by the simplified dynamic model at time i; the right side of the equal sign in equation (6) is obtained by calculation in S1, and the parameters in the additional mass parameter matrix A, damping coefficient matrix B and stiffness coefficient matrix C on the left side of the equal sign are set by the constraint conditions of equation (7). Based on the iteration of the adaptation function, the additional mass parameter matrix A, damping coefficient matrix B and stiffness coefficient matrix C that meet the adaptation function can be iterated, denoted as A best , B best and C best .

[0172] x r (i) is the surging displacement of the wind turbine at time i under OpenFAST analysis, which is obtained by substituting the hydrodynamic force, hydrodynamic moment, hydrostatic moment, aerodynamic force, aerodynamic moment, and mooring moment calculated in step S1 into the OpenFAST simulation;

[0173] θ s(i) is the pitch motion angle of the wind turbine output by the simplified dynamic model at time i. The right side of the equal sign in equation (6) is obtained by calculation in S1. The parameters in the additional mass parameter matrix A, damping coefficient matrix B and stiffness coefficient matrix C on the left side of the equal sign are set by the constraint conditions in equation (7);

[0174] θ r (i) is the pitch motion angle of the wind turbine at time i under OpenFAST analysis, which is obtained by substituting the hydrodynamic force, hydrodynamic moment, hydrostatic moment, aerodynamic force, aerodynamic moment, and mooring moment calculated in step S1 into the OpenFAST simulation.

[0175]

[0176] These are all data simulated by OpenFAST based on different wind turbine structural parameters, wind and wave environment parameters and wind turbine initial states.

[0177] The initial data is obtained by fitting the actual working condition data simulated by OpenFAST, and then substituted into the simplified dynamic model simulated by ODE45. The obtained data are compared with the error to construct the fitness function to update the particle position and velocity until convergence, and the output A is obtained. best , B best and C best .

[0178] set up:

[0179]

[0180] S3: Design a Kalman filter to predict the displacement and pitch angle of the surge.

[0181] The simplified state space model of the wind turbine is constructed by formula (6):

[0182]

[0183] Where x is the wind turbine's longitudinal displacement; θ is the wind turbine's pitch angle;

[0184] Assume that the real x is the displacement of the wind turbine in the longitudinal motion; θ is the pitch angle of the wind turbine. The output equation is as follows:

[0185]

[0186] Kalman filter equation:

[0187]

[0188] In the formula

[0189] G k =e ZT ,Hk =∫0 T e Z(tx-τ )Ld τ ,N k =C best ;

[0190] Where T is the discrete period, tx is the simulation time, and τ is the time from the beginning to time k. k is the system state variable at time k; w k is the process noise vector at time k; v k is the measurement noise vector at time k; u k is the system input at time k; u k-1 is the system input at time k-1.

[0191] Assume that the process noise vector w k The covariance matrix is ​​Q k , the measurement noise vector v k The covariance matrix is ​​R k , process noise vector w k and the measurement noise vector v k Independent of each other, there are:

[0192]

[0193] State update equation:

[0194]

[0195] At time k, the actual measured output y is calculated by equation (7): true,k , based on the actual measured output y true,k And the output estimated by the Kalman filter model The information error e can be calculated k , using the new information error to correct the estimated internal state of the system:

[0196]

[0197] Update the covariance matrix:

[0198]

[0199] I is the identity matrix;

[0200] The Kalman equation is used to estimate the wind turbine's longitudinal displacement x and longitudinal angular velocity θ at a certain moment.

[0201] Simulation:

[0202] The working conditions, wind parameters, wave parameters and initial parameters are recorded in Table 1, and the simplified model and OpenFAST are used to simulate the parameters in Table 1. The simulation results are shown in Figure 2-Figure 6 .

[0203] Table 1 Parameter settings under different working conditions

[0204]

[0205] Figure 2 The free decay response characteristics of the surge and pitch of the simplified dynamic model and OpenFAST are shown. The surge displacement response of the simplified model is in good agreement with the results of OpenFAST. In addition, their natural frequencies are almost the same. This shows that the simplified model of this scheme can replace the traditional OpenFAST model for the prediction of surge and pitch.

[0206] Figure 3 The comparison of the surge and pitch motion responses of the simplified dynamic model and OpenFAST in windless conditions is shown. The maximum error of surge and pitch motion is about 5% in steady state, and the maximum error in the transition process is within 10%, which meets engineering requirements.

[0207] Figure 4 The comparison of the surge and pitch motion responses of the simplified dynamic model and OpenFAST in the case of waves but no wind is shown. The simplified model captures the key motion characteristics of surge and pitch.

[0208] Figure 5 The comparison of the surge and pitch response of the regular wind-wave combined simplified model and OpenFAST in the time and frequency domains is shown. The accuracy of the surge and pitch of the simplified model is reduced, especially in the time domain. This is because OpenFAST is highly nonlinear, and the simplified model sacrifices a certain degree of accuracy for time efficiency. Nevertheless, the main features of the motion response can be well captured by the simplified model. In addition, the power spectral density in the frequency domain is also very consistent with OpenFAST. For the frequency characteristics, the results of the simplified model are very consistent with OpenFAST, indicating the reliability of the simplified theoretical model.

[0209] Figure 6The state estimation performance of the Kalman filter based on the simplified model is demonstrated. Using the output of the pitch angular velocity and the input of the aerodynamic and hydrodynamic forces, the other three estimated states, namely the surge displacement, surge velocity and pitch angle, are expected to be almost the same as the output of the simplified model, which fully demonstrates that the Kalman filter is designed correctly and can reliably estimate the state. Note that the ideal condition means that the Kalman filter is discretized based on the simplified dynamic model. At the same time, the "Ture" output is also the simulation result of the simplified model. This shows that the state space model of the Kalman filter is almost the same as the simplified model, which can be used to test the accuracy of the Kalman filter. The wide applicability and good performance of the simplified dynamic model are demonstrated through multiple environmental tests. Compared with computationally intensive models, such as models based on computational fluid dynamics (CFD), the simplified dynamic model captures the key characteristics of the motion, surge and pitch, through reasonable simplification, which greatly reduces the computing time and resource requirements, making real-time or near-real-time dynamic analysis possible. At the same time, the traditional dynamic model can be combined with modern filtering technology, and only some measurable parameters are used to estimate other state variables that are difficult to measure directly, which has practical value in engineering applications.

[0210] Based on the above ideal embodiments of the present invention, the relevant staff can make various changes and modifications without departing from the technical concept of the present invention through the above description. The technical scope of the present invention is not limited to the contents of the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A method for estimating the motion state of a floating wind turbine, characterized in that: The following steps are included: S1: Based on the dynamics framework of floating wind turbines, a simplified two-dimensional dynamic model for surge and pitch is constructed; The specific steps include: S1.1: Simplify the wind turbine into a rigidly connected mast buoy, tower, impeller and nacelle; simplify the mooring system into a linear spring; construct the dynamic architecture of the wind turbine; S1.2: Analyze aerodynamic forces, aerodynamic moments, hydrostatic restoring moments, hydrodynamic forces, hydrodynamic moments, and mooring moments; S1.3: Based on the law of rigid body dynamics, combined with the aerodynamic force, aerodynamic moment, hydrostatic restoring moment, hydrodynamic force, hydrodynamic moment and mooring moment in S1.2; set the wind turbine mass matrix M, additional mass parameter matrix A, damping coefficient matrix B and rigidity coefficient matrix C; The overall simplified dynamic model of the wind turbine is obtained; S2: Based on OpenFAST, the fitness function and constraint conditions are established to optimize the additional mass parameter matrix A, the damping coefficient matrix B and the stiffness coefficient matrix C. The optimized matrices are recorded as A best , B best and C best ; S3: Adopt A best , B best and C best , Based on the Kalman filter, the longitudinal motion displacement and pitch angle of the wind turbine are estimated at the future moment.

2. A floating wind turbine motion state estimation method according to claim 1, characterized in that: In step S1.3, the wind turbine mass matrix M is: M 11 =m M 12 =M 21 =m×z total M 22 =I o Where m is the mass of the wind turbine; Z total is the center of mass position of the wind turbine; I floater is the moment of inertia of the mast buoy; I tower is the moment of inertia of the tower; M floater is the mass of the mast buoy; M tower is the mass of the tower; M nacelle is the impeller mass; M rotor for the cabin quality; set up The overall simplified dynamic model of the wind turbine is: Where x represents the wind turbine's longitudinal displacement; θ is the wind turbine's pitch angle; F Hydr odynamic is hydrodynamic; F wind is the aerodynamic force; T Hydrodynamic is the hydrodynamic torque; T Wind is the aerodynamic torque.

3. A floating wind turbine motion state estimation method according to claim 2, characterized in that: In step S2, the formulas of fitness function and constraint conditions are: In the formula, x s (i) is the sloshing displacement output by the simplified dynamic model at time i; x r (i) is the surge displacement of the wind turbine at time i under OpenFAST analysis; θ s (i) is the pitch motion angle of the wind turbine output by the simplified dynamic model at time i; θ r (i) is the pitch motion angle of the wind turbine at time i under OpenFAST analysis; is the data simulated by OpenFAST based on different wind turbine structural parameters, wind and wave environment parameters and wind turbine initial state; through the iterative convergence of formula (1) and (2), the optimized A best , B best and C best for: and 4. A floating wind turbine motion state estimation method according to claim 3, characterized in that: Step S3 includes the following steps: S3.1: Convert equation (1) into a simplified state space model, the formula is: S3.2: Set the actual wind turbine oscillation displacement and wind turbine pitch angle output equations as: S3.3: Establish the Kalman filter equation: In the formula, G k =e ZT ,H k =∫0 T it is Z(tx-t) Ldt,N k =C best T is the discrete period, tx is the simulation time, t is the time from the beginning to time k; x k is the system state variable at time k; w k is the process noise vector at time k; v k is the measurement noise vector at time k; u k is the system input at time k; u k-1 is the system input at time k-1; In the formula, Q k is the process noise vector w k The covariance matrix of k is the measurement noise vector v k The covariance matrix of State update equation: Error gain equation: Among them, the actual measured output at time k is y true,k , calculated by formula (3); Update the covariance matrix: I is the unit matrix; the Kalman equation is used to estimate the wind turbine's longitudinal motion displacement x and longitudinal angular velocity θ at a certain moment.

5. A floating wind turbine motion state estimation method according to claim 3, characterized in that: The calculation formulas for aerodynamic force and aerodynamic moment are: In the formula, F wind is the aerodynamic force; r is the air density; V total is the relative wind speed of wind turbine blades; T Wind is the aerodynamic torque; is the inflow angle; C l is the lift coefficient; C d is the gas drag coefficient; r is the distance from the blade element to the center of the wind turbine hub.

6. A floating wind turbine motion state estimation method according to claim 3, characterized in that: The calculation formulas for the hydrostatic restoring moment, hydrodynamic force and hydrodynamic moment are: Where, T Buoy is the total hydrostatic restoring moment of the wind turbine; m is the total mass of the wind turbine; g is the acceleration due to gravity; Z b is the buoyancy center of the wind turbine; Z g is the gravity center of the wind turbine; θ is the pitch motion angle of the wind turbine; ρ1 is the water density where the wind turbine is located; is the additional inertia of the wind turbine, D is the diameter of the mast buoy; Where, T Hydrodynamic is the hydrodynamic moment; w is the frequency of regular waves; u 水 The wave speed of the wave; C m is the inertia coefficient; C D is the water resistance coefficient; z is the height of the wind turbine; A1 is the end surface area of ​​the mast buoy; F Hydr odynamic means hydrodynamic.

7. A floating wind turbine motion state estimation method according to claim 3, characterized in that: The mooring moment is: T moor =F moor ×z moor =-K moor (z moor ×q+x0)×z moor ; In the formula, F moor is the horizontal mooring force; K moor Wind turbine horizontal mooring stiffness coefficient matrix; Z moor is the distance from the mooring anchor point to the origin of the local coordinate system in the local coordinate system; x0 is the longitudinal displacement of the origin of the local coordinate system.

Images